diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,42034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7225259507237302,
+  "eval_steps": 500,
+  "global_step": 30000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00012042099178728837,
+      "grad_norm": 3.4532573223114014,
+      "learning_rate": 1.9258545979778525e-09,
+      "loss": 1.2769,
+      "step": 5
+    },
+    {
+      "epoch": 0.00024084198357457673,
+      "grad_norm": 3.2849907875061035,
+      "learning_rate": 4.333172845450169e-09,
+      "loss": 1.2562,
+      "step": 10
+    },
+    {
+      "epoch": 0.00036126297536186506,
+      "grad_norm": 3.0698747634887695,
+      "learning_rate": 6.740491092922484e-09,
+      "loss": 1.2727,
+      "step": 15
+    },
+    {
+      "epoch": 0.00048168396714915346,
+      "grad_norm": 3.3526523113250732,
+      "learning_rate": 9.1478093403948e-09,
+      "loss": 1.2471,
+      "step": 20
+    },
+    {
+      "epoch": 0.0006021049589364418,
+      "grad_norm": 3.5746824741363525,
+      "learning_rate": 1.1555127587867116e-08,
+      "loss": 1.2575,
+      "step": 25
+    },
+    {
+      "epoch": 0.0007225259507237301,
+      "grad_norm": 3.2632179260253906,
+      "learning_rate": 1.3962445835339432e-08,
+      "loss": 1.1786,
+      "step": 30
+    },
+    {
+      "epoch": 0.0008429469425110186,
+      "grad_norm": 3.0086328983306885,
+      "learning_rate": 1.636976408281175e-08,
+      "loss": 1.4168,
+      "step": 35
+    },
+    {
+      "epoch": 0.0009633679342983069,
+      "grad_norm": 3.5532827377319336,
+      "learning_rate": 1.8777082330284063e-08,
+      "loss": 1.2169,
+      "step": 40
+    },
+    {
+      "epoch": 0.0010837889260855952,
+      "grad_norm": 3.0027787685394287,
+      "learning_rate": 2.118440057775638e-08,
+      "loss": 1.2513,
+      "step": 45
+    },
+    {
+      "epoch": 0.0012042099178728835,
+      "grad_norm": 3.6059303283691406,
+      "learning_rate": 2.3591718825228694e-08,
+      "loss": 1.2505,
+      "step": 50
+    },
+    {
+      "epoch": 0.0013246309096601719,
+      "grad_norm": 2.860373020172119,
+      "learning_rate": 2.599903707270101e-08,
+      "loss": 1.2457,
+      "step": 55
+    },
+    {
+      "epoch": 0.0014450519014474602,
+      "grad_norm": 3.1921334266662598,
+      "learning_rate": 2.8406355320173326e-08,
+      "loss": 1.2627,
+      "step": 60
+    },
+    {
+      "epoch": 0.0015654728932347486,
+      "grad_norm": 3.080982208251953,
+      "learning_rate": 3.081367356764564e-08,
+      "loss": 1.1995,
+      "step": 65
+    },
+    {
+      "epoch": 0.0016858938850220371,
+      "grad_norm": 3.90539288520813,
+      "learning_rate": 3.3220991815117954e-08,
+      "loss": 1.3055,
+      "step": 70
+    },
+    {
+      "epoch": 0.0018063148768093255,
+      "grad_norm": 3.3694944381713867,
+      "learning_rate": 3.5628310062590275e-08,
+      "loss": 1.2396,
+      "step": 75
+    },
+    {
+      "epoch": 0.0019267358685966139,
+      "grad_norm": 3.218294143676758,
+      "learning_rate": 3.803562831006259e-08,
+      "loss": 1.2292,
+      "step": 80
+    },
+    {
+      "epoch": 0.002047156860383902,
+      "grad_norm": 3.063572406768799,
+      "learning_rate": 4.04429465575349e-08,
+      "loss": 1.253,
+      "step": 85
+    },
+    {
+      "epoch": 0.0021675778521711903,
+      "grad_norm": 3.4931082725524902,
+      "learning_rate": 4.285026480500722e-08,
+      "loss": 1.2709,
+      "step": 90
+    },
+    {
+      "epoch": 0.002287998843958479,
+      "grad_norm": 3.7916605472564697,
+      "learning_rate": 4.525758305247954e-08,
+      "loss": 1.2878,
+      "step": 95
+    },
+    {
+      "epoch": 0.002408419835745767,
+      "grad_norm": 2.3481788635253906,
+      "learning_rate": 4.766490129995185e-08,
+      "loss": 1.2535,
+      "step": 100
+    },
+    {
+      "epoch": 0.0025288408275330556,
+      "grad_norm": 2.8467800617218018,
+      "learning_rate": 5.0072219547424166e-08,
+      "loss": 1.2573,
+      "step": 105
+    },
+    {
+      "epoch": 0.0026492618193203438,
+      "grad_norm": 3.1305465698242188,
+      "learning_rate": 5.2479537794896487e-08,
+      "loss": 1.2112,
+      "step": 110
+    },
+    {
+      "epoch": 0.0027696828111076323,
+      "grad_norm": 2.8494081497192383,
+      "learning_rate": 5.48868560423688e-08,
+      "loss": 1.2003,
+      "step": 115
+    },
+    {
+      "epoch": 0.0028901038028949205,
+      "grad_norm": 3.01526141166687,
+      "learning_rate": 5.7294174289841115e-08,
+      "loss": 1.2025,
+      "step": 120
+    },
+    {
+      "epoch": 0.003010524794682209,
+      "grad_norm": 3.138826370239258,
+      "learning_rate": 5.970149253731343e-08,
+      "loss": 1.2197,
+      "step": 125
+    },
+    {
+      "epoch": 0.003130945786469497,
+      "grad_norm": 3.5005788803100586,
+      "learning_rate": 6.210881078478574e-08,
+      "loss": 1.2479,
+      "step": 130
+    },
+    {
+      "epoch": 0.0032513667782567857,
+      "grad_norm": 3.51312255859375,
+      "learning_rate": 6.451612903225806e-08,
+      "loss": 1.2142,
+      "step": 135
+    },
+    {
+      "epoch": 0.0033717877700440743,
+      "grad_norm": 2.6358306407928467,
+      "learning_rate": 6.692344727973037e-08,
+      "loss": 1.2545,
+      "step": 140
+    },
+    {
+      "epoch": 0.0034922087618313624,
+      "grad_norm": 3.6043667793273926,
+      "learning_rate": 6.93307655272027e-08,
+      "loss": 1.2427,
+      "step": 145
+    },
+    {
+      "epoch": 0.003612629753618651,
+      "grad_norm": 3.161194324493408,
+      "learning_rate": 7.173808377467501e-08,
+      "loss": 1.2686,
+      "step": 150
+    },
+    {
+      "epoch": 0.003733050745405939,
+      "grad_norm": 3.4229471683502197,
+      "learning_rate": 7.414540202214733e-08,
+      "loss": 1.3099,
+      "step": 155
+    },
+    {
+      "epoch": 0.0038534717371932277,
+      "grad_norm": 3.3181447982788086,
+      "learning_rate": 7.655272026961963e-08,
+      "loss": 1.2736,
+      "step": 160
+    },
+    {
+      "epoch": 0.003973892728980516,
+      "grad_norm": 3.1612415313720703,
+      "learning_rate": 7.896003851709195e-08,
+      "loss": 1.2001,
+      "step": 165
+    },
+    {
+      "epoch": 0.004094313720767804,
+      "grad_norm": 2.884935140609741,
+      "learning_rate": 8.136735676456427e-08,
+      "loss": 1.2524,
+      "step": 170
+    },
+    {
+      "epoch": 0.0042147347125550925,
+      "grad_norm": 2.7151708602905273,
+      "learning_rate": 8.377467501203658e-08,
+      "loss": 1.2547,
+      "step": 175
+    },
+    {
+      "epoch": 0.004335155704342381,
+      "grad_norm": 3.2015910148620605,
+      "learning_rate": 8.618199325950891e-08,
+      "loss": 1.2492,
+      "step": 180
+    },
+    {
+      "epoch": 0.00445557669612967,
+      "grad_norm": 3.480416774749756,
+      "learning_rate": 8.858931150698122e-08,
+      "loss": 1.2558,
+      "step": 185
+    },
+    {
+      "epoch": 0.004575997687916958,
+      "grad_norm": 3.1630730628967285,
+      "learning_rate": 9.099662975445354e-08,
+      "loss": 1.2078,
+      "step": 190
+    },
+    {
+      "epoch": 0.004696418679704246,
+      "grad_norm": 3.1624763011932373,
+      "learning_rate": 9.340394800192584e-08,
+      "loss": 1.2021,
+      "step": 195
+    },
+    {
+      "epoch": 0.004816839671491534,
+      "grad_norm": 2.6630239486694336,
+      "learning_rate": 9.581126624939817e-08,
+      "loss": 1.1915,
+      "step": 200
+    },
+    {
+      "epoch": 0.004937260663278823,
+      "grad_norm": 2.7018954753875732,
+      "learning_rate": 9.821858449687048e-08,
+      "loss": 1.2762,
+      "step": 205
+    },
+    {
+      "epoch": 0.005057681655066111,
+      "grad_norm": 3.0080630779266357,
+      "learning_rate": 1.006259027443428e-07,
+      "loss": 1.2456,
+      "step": 210
+    },
+    {
+      "epoch": 0.005178102646853399,
+      "grad_norm": 2.8256232738494873,
+      "learning_rate": 1.0303322099181512e-07,
+      "loss": 1.218,
+      "step": 215
+    },
+    {
+      "epoch": 0.0052985236386406875,
+      "grad_norm": 3.1638636589050293,
+      "learning_rate": 1.0544053923928744e-07,
+      "loss": 1.2508,
+      "step": 220
+    },
+    {
+      "epoch": 0.0054189446304279765,
+      "grad_norm": 2.64296555519104,
+      "learning_rate": 1.0784785748675975e-07,
+      "loss": 1.2768,
+      "step": 225
+    },
+    {
+      "epoch": 0.005539365622215265,
+      "grad_norm": 2.2767012119293213,
+      "learning_rate": 1.1025517573423205e-07,
+      "loss": 1.1927,
+      "step": 230
+    },
+    {
+      "epoch": 0.005659786614002553,
+      "grad_norm": 2.666334629058838,
+      "learning_rate": 1.1266249398170438e-07,
+      "loss": 1.1981,
+      "step": 235
+    },
+    {
+      "epoch": 0.005780207605789841,
+      "grad_norm": 3.0129833221435547,
+      "learning_rate": 1.1506981222917669e-07,
+      "loss": 1.1702,
+      "step": 240
+    },
+    {
+      "epoch": 0.00590062859757713,
+      "grad_norm": 2.534895896911621,
+      "learning_rate": 1.17477130476649e-07,
+      "loss": 1.2273,
+      "step": 245
+    },
+    {
+      "epoch": 0.006021049589364418,
+      "grad_norm": 3.0159912109375,
+      "learning_rate": 1.1988444872412132e-07,
+      "loss": 1.1828,
+      "step": 250
+    },
+    {
+      "epoch": 0.006141470581151706,
+      "grad_norm": 2.4064695835113525,
+      "learning_rate": 1.2229176697159363e-07,
+      "loss": 1.1927,
+      "step": 255
+    },
+    {
+      "epoch": 0.006261891572938994,
+      "grad_norm": 2.506269693374634,
+      "learning_rate": 1.2469908521906595e-07,
+      "loss": 1.2452,
+      "step": 260
+    },
+    {
+      "epoch": 0.006382312564726283,
+      "grad_norm": 2.26957631111145,
+      "learning_rate": 1.271064034665383e-07,
+      "loss": 1.2155,
+      "step": 265
+    },
+    {
+      "epoch": 0.0065027335565135715,
+      "grad_norm": 2.604719638824463,
+      "learning_rate": 1.295137217140106e-07,
+      "loss": 1.1983,
+      "step": 270
+    },
+    {
+      "epoch": 0.00662315454830086,
+      "grad_norm": 2.3419127464294434,
+      "learning_rate": 1.3192103996148292e-07,
+      "loss": 1.2137,
+      "step": 275
+    },
+    {
+      "epoch": 0.006743575540088149,
+      "grad_norm": 2.5202889442443848,
+      "learning_rate": 1.343283582089552e-07,
+      "loss": 1.2272,
+      "step": 280
+    },
+    {
+      "epoch": 0.006863996531875437,
+      "grad_norm": 2.3413245677948,
+      "learning_rate": 1.3673567645642752e-07,
+      "loss": 1.2049,
+      "step": 285
+    },
+    {
+      "epoch": 0.006984417523662725,
+      "grad_norm": 2.6377112865448,
+      "learning_rate": 1.3914299470389983e-07,
+      "loss": 1.2673,
+      "step": 290
+    },
+    {
+      "epoch": 0.007104838515450013,
+      "grad_norm": 2.182490825653076,
+      "learning_rate": 1.4155031295137215e-07,
+      "loss": 1.2062,
+      "step": 295
+    },
+    {
+      "epoch": 0.007225259507237302,
+      "grad_norm": 2.1873981952667236,
+      "learning_rate": 1.439576311988445e-07,
+      "loss": 1.1677,
+      "step": 300
+    },
+    {
+      "epoch": 0.00734568049902459,
+      "grad_norm": 2.436870574951172,
+      "learning_rate": 1.463649494463168e-07,
+      "loss": 1.1813,
+      "step": 305
+    },
+    {
+      "epoch": 0.007466101490811878,
+      "grad_norm": 2.0647244453430176,
+      "learning_rate": 1.4877226769378912e-07,
+      "loss": 1.1143,
+      "step": 310
+    },
+    {
+      "epoch": 0.007586522482599166,
+      "grad_norm": 2.045997381210327,
+      "learning_rate": 1.5117958594126143e-07,
+      "loss": 1.2223,
+      "step": 315
+    },
+    {
+      "epoch": 0.007706943474386455,
+      "grad_norm": 2.0606868267059326,
+      "learning_rate": 1.5358690418873374e-07,
+      "loss": 1.184,
+      "step": 320
+    },
+    {
+      "epoch": 0.007827364466173744,
+      "grad_norm": 2.2403106689453125,
+      "learning_rate": 1.5599422243620606e-07,
+      "loss": 1.1599,
+      "step": 325
+    },
+    {
+      "epoch": 0.007947785457961033,
+      "grad_norm": 1.9159278869628906,
+      "learning_rate": 1.5840154068367837e-07,
+      "loss": 1.1863,
+      "step": 330
+    },
+    {
+      "epoch": 0.00806820644974832,
+      "grad_norm": 2.374685525894165,
+      "learning_rate": 1.608088589311507e-07,
+      "loss": 1.2206,
+      "step": 335
+    },
+    {
+      "epoch": 0.008188627441535609,
+      "grad_norm": 2.1852266788482666,
+      "learning_rate": 1.6321617717862303e-07,
+      "loss": 1.165,
+      "step": 340
+    },
+    {
+      "epoch": 0.008309048433322896,
+      "grad_norm": 2.137631416320801,
+      "learning_rate": 1.6562349542609534e-07,
+      "loss": 1.177,
+      "step": 345
+    },
+    {
+      "epoch": 0.008429469425110185,
+      "grad_norm": 2.00748348236084,
+      "learning_rate": 1.6803081367356763e-07,
+      "loss": 1.0761,
+      "step": 350
+    },
+    {
+      "epoch": 0.008549890416897474,
+      "grad_norm": 1.9672129154205322,
+      "learning_rate": 1.7043813192103994e-07,
+      "loss": 1.1491,
+      "step": 355
+    },
+    {
+      "epoch": 0.008670311408684761,
+      "grad_norm": 2.3764731884002686,
+      "learning_rate": 1.7284545016851226e-07,
+      "loss": 1.2437,
+      "step": 360
+    },
+    {
+      "epoch": 0.00879073240047205,
+      "grad_norm": 2.0609636306762695,
+      "learning_rate": 1.7525276841598457e-07,
+      "loss": 1.1461,
+      "step": 365
+    },
+    {
+      "epoch": 0.00891115339225934,
+      "grad_norm": 1.9100985527038574,
+      "learning_rate": 1.7766008666345688e-07,
+      "loss": 1.1708,
+      "step": 370
+    },
+    {
+      "epoch": 0.009031574384046627,
+      "grad_norm": 1.9871859550476074,
+      "learning_rate": 1.8006740491092923e-07,
+      "loss": 1.1183,
+      "step": 375
+    },
+    {
+      "epoch": 0.009151995375833916,
+      "grad_norm": 2.0929617881774902,
+      "learning_rate": 1.8247472315840154e-07,
+      "loss": 1.2154,
+      "step": 380
+    },
+    {
+      "epoch": 0.009272416367621203,
+      "grad_norm": 2.0678470134735107,
+      "learning_rate": 1.8488204140587385e-07,
+      "loss": 1.1872,
+      "step": 385
+    },
+    {
+      "epoch": 0.009392837359408492,
+      "grad_norm": 2.0919220447540283,
+      "learning_rate": 1.8728935965334617e-07,
+      "loss": 1.1782,
+      "step": 390
+    },
+    {
+      "epoch": 0.009513258351195781,
+      "grad_norm": 1.8138477802276611,
+      "learning_rate": 1.8969667790081848e-07,
+      "loss": 1.1245,
+      "step": 395
+    },
+    {
+      "epoch": 0.009633679342983068,
+      "grad_norm": 2.0728824138641357,
+      "learning_rate": 1.921039961482908e-07,
+      "loss": 1.1498,
+      "step": 400
+    },
+    {
+      "epoch": 0.009754100334770357,
+      "grad_norm": 2.1343467235565186,
+      "learning_rate": 1.945113143957631e-07,
+      "loss": 1.2057,
+      "step": 405
+    },
+    {
+      "epoch": 0.009874521326557646,
+      "grad_norm": 1.8334988355636597,
+      "learning_rate": 1.9691863264323545e-07,
+      "loss": 1.1354,
+      "step": 410
+    },
+    {
+      "epoch": 0.009994942318344933,
+      "grad_norm": 1.8179374933242798,
+      "learning_rate": 1.9932595089070774e-07,
+      "loss": 1.172,
+      "step": 415
+    },
+    {
+      "epoch": 0.010115363310132222,
+      "grad_norm": 1.8935538530349731,
+      "learning_rate": 2.0173326913818005e-07,
+      "loss": 1.2059,
+      "step": 420
+    },
+    {
+      "epoch": 0.010235784301919511,
+      "grad_norm": 2.025412082672119,
+      "learning_rate": 2.0414058738565237e-07,
+      "loss": 1.1687,
+      "step": 425
+    },
+    {
+      "epoch": 0.010356205293706799,
+      "grad_norm": 1.9169843196868896,
+      "learning_rate": 2.0654790563312468e-07,
+      "loss": 1.2163,
+      "step": 430
+    },
+    {
+      "epoch": 0.010476626285494088,
+      "grad_norm": 1.9890685081481934,
+      "learning_rate": 2.08955223880597e-07,
+      "loss": 1.1798,
+      "step": 435
+    },
+    {
+      "epoch": 0.010597047277281375,
+      "grad_norm": 1.7384068965911865,
+      "learning_rate": 2.113625421280693e-07,
+      "loss": 1.103,
+      "step": 440
+    },
+    {
+      "epoch": 0.010717468269068664,
+      "grad_norm": 1.8030136823654175,
+      "learning_rate": 2.1376986037554165e-07,
+      "loss": 1.2152,
+      "step": 445
+    },
+    {
+      "epoch": 0.010837889260855953,
+      "grad_norm": 1.7656060457229614,
+      "learning_rate": 2.1617717862301396e-07,
+      "loss": 1.1988,
+      "step": 450
+    },
+    {
+      "epoch": 0.01095831025264324,
+      "grad_norm": 2.1653716564178467,
+      "learning_rate": 2.1858449687048628e-07,
+      "loss": 1.1923,
+      "step": 455
+    },
+    {
+      "epoch": 0.01107873124443053,
+      "grad_norm": 2.042724132537842,
+      "learning_rate": 2.209918151179586e-07,
+      "loss": 1.216,
+      "step": 460
+    },
+    {
+      "epoch": 0.011199152236217818,
+      "grad_norm": 1.65592360496521,
+      "learning_rate": 2.233991333654309e-07,
+      "loss": 1.1506,
+      "step": 465
+    },
+    {
+      "epoch": 0.011319573228005106,
+      "grad_norm": 1.6511238813400269,
+      "learning_rate": 2.2580645161290322e-07,
+      "loss": 1.1547,
+      "step": 470
+    },
+    {
+      "epoch": 0.011439994219792395,
+      "grad_norm": 1.8964343070983887,
+      "learning_rate": 2.2821376986037553e-07,
+      "loss": 1.1387,
+      "step": 475
+    },
+    {
+      "epoch": 0.011560415211579682,
+      "grad_norm": 1.615464210510254,
+      "learning_rate": 2.3062108810784782e-07,
+      "loss": 1.1492,
+      "step": 480
+    },
+    {
+      "epoch": 0.01168083620336697,
+      "grad_norm": 2.1037256717681885,
+      "learning_rate": 2.3302840635532016e-07,
+      "loss": 1.1664,
+      "step": 485
+    },
+    {
+      "epoch": 0.01180125719515426,
+      "grad_norm": 1.4680767059326172,
+      "learning_rate": 2.3543572460279248e-07,
+      "loss": 1.1456,
+      "step": 490
+    },
+    {
+      "epoch": 0.011921678186941547,
+      "grad_norm": 1.7871757745742798,
+      "learning_rate": 2.378430428502648e-07,
+      "loss": 1.1777,
+      "step": 495
+    },
+    {
+      "epoch": 0.012042099178728836,
+      "grad_norm": 1.7084572315216064,
+      "learning_rate": 2.402503610977371e-07,
+      "loss": 1.1512,
+      "step": 500
+    },
+    {
+      "epoch": 0.012162520170516125,
+      "grad_norm": 1.815544605255127,
+      "learning_rate": 2.426576793452094e-07,
+      "loss": 1.153,
+      "step": 505
+    },
+    {
+      "epoch": 0.012282941162303412,
+      "grad_norm": 1.6143004894256592,
+      "learning_rate": 2.4506499759268173e-07,
+      "loss": 1.131,
+      "step": 510
+    },
+    {
+      "epoch": 0.012403362154090701,
+      "grad_norm": 2.2032082080841064,
+      "learning_rate": 2.4747231584015405e-07,
+      "loss": 1.1916,
+      "step": 515
+    },
+    {
+      "epoch": 0.012523783145877989,
+      "grad_norm": 1.9207594394683838,
+      "learning_rate": 2.4987963408762636e-07,
+      "loss": 1.1398,
+      "step": 520
+    },
+    {
+      "epoch": 0.012644204137665278,
+      "grad_norm": 1.954811930656433,
+      "learning_rate": 2.522869523350987e-07,
+      "loss": 1.162,
+      "step": 525
+    },
+    {
+      "epoch": 0.012764625129452567,
+      "grad_norm": 1.676818609237671,
+      "learning_rate": 2.54694270582571e-07,
+      "loss": 1.1526,
+      "step": 530
+    },
+    {
+      "epoch": 0.012885046121239854,
+      "grad_norm": 1.7171730995178223,
+      "learning_rate": 2.5710158883004335e-07,
+      "loss": 1.1389,
+      "step": 535
+    },
+    {
+      "epoch": 0.013005467113027143,
+      "grad_norm": 1.7267348766326904,
+      "learning_rate": 2.5950890707751567e-07,
+      "loss": 1.1892,
+      "step": 540
+    },
+    {
+      "epoch": 0.013125888104814432,
+      "grad_norm": 1.7431941032409668,
+      "learning_rate": 2.61916225324988e-07,
+      "loss": 1.0527,
+      "step": 545
+    },
+    {
+      "epoch": 0.01324630909660172,
+      "grad_norm": 1.5393019914627075,
+      "learning_rate": 2.643235435724603e-07,
+      "loss": 1.1091,
+      "step": 550
+    },
+    {
+      "epoch": 0.013366730088389008,
+      "grad_norm": 1.6455602645874023,
+      "learning_rate": 2.667308618199326e-07,
+      "loss": 1.108,
+      "step": 555
+    },
+    {
+      "epoch": 0.013487151080176297,
+      "grad_norm": 1.696184754371643,
+      "learning_rate": 2.691381800674049e-07,
+      "loss": 1.1133,
+      "step": 560
+    },
+    {
+      "epoch": 0.013607572071963584,
+      "grad_norm": 1.7459622621536255,
+      "learning_rate": 2.7154549831487724e-07,
+      "loss": 1.1074,
+      "step": 565
+    },
+    {
+      "epoch": 0.013727993063750873,
+      "grad_norm": 1.5549553632736206,
+      "learning_rate": 2.7395281656234955e-07,
+      "loss": 1.1855,
+      "step": 570
+    },
+    {
+      "epoch": 0.01384841405553816,
+      "grad_norm": 1.6191879510879517,
+      "learning_rate": 2.7636013480982187e-07,
+      "loss": 1.075,
+      "step": 575
+    },
+    {
+      "epoch": 0.01396883504732545,
+      "grad_norm": 1.6304470300674438,
+      "learning_rate": 2.787674530572942e-07,
+      "loss": 1.1123,
+      "step": 580
+    },
+    {
+      "epoch": 0.014089256039112739,
+      "grad_norm": 1.7658319473266602,
+      "learning_rate": 2.8117477130476644e-07,
+      "loss": 1.1768,
+      "step": 585
+    },
+    {
+      "epoch": 0.014209677030900026,
+      "grad_norm": 1.5940040349960327,
+      "learning_rate": 2.8358208955223876e-07,
+      "loss": 1.1377,
+      "step": 590
+    },
+    {
+      "epoch": 0.014330098022687315,
+      "grad_norm": 1.5074690580368042,
+      "learning_rate": 2.8598940779971107e-07,
+      "loss": 1.1538,
+      "step": 595
+    },
+    {
+      "epoch": 0.014450519014474604,
+      "grad_norm": 1.7436341047286987,
+      "learning_rate": 2.883967260471834e-07,
+      "loss": 1.0854,
+      "step": 600
+    },
+    {
+      "epoch": 0.014570940006261891,
+      "grad_norm": 1.6860382556915283,
+      "learning_rate": 2.908040442946557e-07,
+      "loss": 1.1191,
+      "step": 605
+    },
+    {
+      "epoch": 0.01469136099804918,
+      "grad_norm": 1.799471378326416,
+      "learning_rate": 2.9321136254212807e-07,
+      "loss": 1.1518,
+      "step": 610
+    },
+    {
+      "epoch": 0.014811781989836468,
+      "grad_norm": 1.7893131971359253,
+      "learning_rate": 2.956186807896004e-07,
+      "loss": 1.1027,
+      "step": 615
+    },
+    {
+      "epoch": 0.014932202981623757,
+      "grad_norm": 1.6366047859191895,
+      "learning_rate": 2.980259990370727e-07,
+      "loss": 1.1171,
+      "step": 620
+    },
+    {
+      "epoch": 0.015052623973411046,
+      "grad_norm": 1.6394248008728027,
+      "learning_rate": 3.00433317284545e-07,
+      "loss": 1.101,
+      "step": 625
+    },
+    {
+      "epoch": 0.015173044965198333,
+      "grad_norm": 1.658291220664978,
+      "learning_rate": 3.028406355320173e-07,
+      "loss": 1.144,
+      "step": 630
+    },
+    {
+      "epoch": 0.015293465956985622,
+      "grad_norm": 1.7101342678070068,
+      "learning_rate": 3.0524795377948964e-07,
+      "loss": 1.0963,
+      "step": 635
+    },
+    {
+      "epoch": 0.01541388694877291,
+      "grad_norm": 2.0829336643218994,
+      "learning_rate": 3.0765527202696195e-07,
+      "loss": 1.1316,
+      "step": 640
+    },
+    {
+      "epoch": 0.015534307940560198,
+      "grad_norm": 1.643818736076355,
+      "learning_rate": 3.1006259027443426e-07,
+      "loss": 1.1402,
+      "step": 645
+    },
+    {
+      "epoch": 0.015654728932347487,
+      "grad_norm": 1.5390375852584839,
+      "learning_rate": 3.124699085219066e-07,
+      "loss": 1.1286,
+      "step": 650
+    },
+    {
+      "epoch": 0.015775149924134774,
+      "grad_norm": 1.5295696258544922,
+      "learning_rate": 3.148772267693789e-07,
+      "loss": 1.0913,
+      "step": 655
+    },
+    {
+      "epoch": 0.015895570915922065,
+      "grad_norm": 1.7846722602844238,
+      "learning_rate": 3.172845450168512e-07,
+      "loss": 1.1475,
+      "step": 660
+    },
+    {
+      "epoch": 0.016015991907709352,
+      "grad_norm": 1.5267643928527832,
+      "learning_rate": 3.196918632643235e-07,
+      "loss": 1.147,
+      "step": 665
+    },
+    {
+      "epoch": 0.01613641289949664,
+      "grad_norm": 1.5753979682922363,
+      "learning_rate": 3.2209918151179583e-07,
+      "loss": 1.1521,
+      "step": 670
+    },
+    {
+      "epoch": 0.016256833891283927,
+      "grad_norm": 1.597781777381897,
+      "learning_rate": 3.2450649975926815e-07,
+      "loss": 1.1069,
+      "step": 675
+    },
+    {
+      "epoch": 0.016377254883071218,
+      "grad_norm": 1.9727338552474976,
+      "learning_rate": 3.2691381800674046e-07,
+      "loss": 1.1125,
+      "step": 680
+    },
+    {
+      "epoch": 0.016497675874858505,
+      "grad_norm": 1.6481128931045532,
+      "learning_rate": 3.2932113625421283e-07,
+      "loss": 1.1188,
+      "step": 685
+    },
+    {
+      "epoch": 0.016618096866645792,
+      "grad_norm": 1.6717129945755005,
+      "learning_rate": 3.3172845450168514e-07,
+      "loss": 1.1382,
+      "step": 690
+    },
+    {
+      "epoch": 0.016738517858433083,
+      "grad_norm": 1.7000445127487183,
+      "learning_rate": 3.3413577274915746e-07,
+      "loss": 1.0947,
+      "step": 695
+    },
+    {
+      "epoch": 0.01685893885022037,
+      "grad_norm": 1.8228042125701904,
+      "learning_rate": 3.3654309099662977e-07,
+      "loss": 1.109,
+      "step": 700
+    },
+    {
+      "epoch": 0.016979359842007657,
+      "grad_norm": 1.5345516204833984,
+      "learning_rate": 3.389504092441021e-07,
+      "loss": 1.1233,
+      "step": 705
+    },
+    {
+      "epoch": 0.017099780833794948,
+      "grad_norm": 1.7012462615966797,
+      "learning_rate": 3.413577274915744e-07,
+      "loss": 1.091,
+      "step": 710
+    },
+    {
+      "epoch": 0.017220201825582235,
+      "grad_norm": 1.6733700037002563,
+      "learning_rate": 3.437650457390467e-07,
+      "loss": 1.1274,
+      "step": 715
+    },
+    {
+      "epoch": 0.017340622817369523,
+      "grad_norm": 1.6725404262542725,
+      "learning_rate": 3.46172363986519e-07,
+      "loss": 1.1021,
+      "step": 720
+    },
+    {
+      "epoch": 0.017461043809156813,
+      "grad_norm": 1.902208924293518,
+      "learning_rate": 3.485796822339913e-07,
+      "loss": 1.1609,
+      "step": 725
+    },
+    {
+      "epoch": 0.0175814648009441,
+      "grad_norm": 1.8290274143218994,
+      "learning_rate": 3.509870004814636e-07,
+      "loss": 1.0894,
+      "step": 730
+    },
+    {
+      "epoch": 0.017701885792731388,
+      "grad_norm": 1.551406979560852,
+      "learning_rate": 3.533943187289359e-07,
+      "loss": 1.0974,
+      "step": 735
+    },
+    {
+      "epoch": 0.01782230678451868,
+      "grad_norm": 1.6504721641540527,
+      "learning_rate": 3.5580163697640823e-07,
+      "loss": 1.0495,
+      "step": 740
+    },
+    {
+      "epoch": 0.017942727776305966,
+      "grad_norm": 1.7599427700042725,
+      "learning_rate": 3.5820895522388055e-07,
+      "loss": 1.0976,
+      "step": 745
+    },
+    {
+      "epoch": 0.018063148768093253,
+      "grad_norm": 1.696821689605713,
+      "learning_rate": 3.6061627347135286e-07,
+      "loss": 1.1485,
+      "step": 750
+    },
+    {
+      "epoch": 0.018183569759880544,
+      "grad_norm": 1.7811964750289917,
+      "learning_rate": 3.6302359171882523e-07,
+      "loss": 1.0991,
+      "step": 755
+    },
+    {
+      "epoch": 0.01830399075166783,
+      "grad_norm": 1.7591949701309204,
+      "learning_rate": 3.6543090996629754e-07,
+      "loss": 1.1346,
+      "step": 760
+    },
+    {
+      "epoch": 0.01842441174345512,
+      "grad_norm": 1.4767632484436035,
+      "learning_rate": 3.6783822821376986e-07,
+      "loss": 1.0937,
+      "step": 765
+    },
+    {
+      "epoch": 0.018544832735242406,
+      "grad_norm": 1.8599060773849487,
+      "learning_rate": 3.7024554646124217e-07,
+      "loss": 1.0742,
+      "step": 770
+    },
+    {
+      "epoch": 0.018665253727029697,
+      "grad_norm": 1.5535491704940796,
+      "learning_rate": 3.726528647087145e-07,
+      "loss": 1.1141,
+      "step": 775
+    },
+    {
+      "epoch": 0.018785674718816984,
+      "grad_norm": 1.7345106601715088,
+      "learning_rate": 3.750601829561868e-07,
+      "loss": 1.0477,
+      "step": 780
+    },
+    {
+      "epoch": 0.01890609571060427,
+      "grad_norm": 1.6146178245544434,
+      "learning_rate": 3.774675012036591e-07,
+      "loss": 1.0765,
+      "step": 785
+    },
+    {
+      "epoch": 0.019026516702391562,
+      "grad_norm": 1.61898672580719,
+      "learning_rate": 3.798748194511314e-07,
+      "loss": 1.0599,
+      "step": 790
+    },
+    {
+      "epoch": 0.01914693769417885,
+      "grad_norm": 1.8047658205032349,
+      "learning_rate": 3.8228213769860374e-07,
+      "loss": 1.1565,
+      "step": 795
+    },
+    {
+      "epoch": 0.019267358685966136,
+      "grad_norm": 1.7502000331878662,
+      "learning_rate": 3.8468945594607605e-07,
+      "loss": 1.1029,
+      "step": 800
+    },
+    {
+      "epoch": 0.019387779677753427,
+      "grad_norm": 1.7901040315628052,
+      "learning_rate": 3.8709677419354837e-07,
+      "loss": 1.0321,
+      "step": 805
+    },
+    {
+      "epoch": 0.019508200669540714,
+      "grad_norm": 1.710308313369751,
+      "learning_rate": 3.895040924410207e-07,
+      "loss": 1.1007,
+      "step": 810
+    },
+    {
+      "epoch": 0.019628621661328,
+      "grad_norm": 1.6469848155975342,
+      "learning_rate": 3.91911410688493e-07,
+      "loss": 1.0919,
+      "step": 815
+    },
+    {
+      "epoch": 0.019749042653115292,
+      "grad_norm": 1.9024361371994019,
+      "learning_rate": 3.943187289359653e-07,
+      "loss": 1.1408,
+      "step": 820
+    },
+    {
+      "epoch": 0.01986946364490258,
+      "grad_norm": 1.4699581861495972,
+      "learning_rate": 3.967260471834376e-07,
+      "loss": 1.0799,
+      "step": 825
+    },
+    {
+      "epoch": 0.019989884636689867,
+      "grad_norm": 1.7625906467437744,
+      "learning_rate": 3.9913336543091e-07,
+      "loss": 1.0956,
+      "step": 830
+    },
+    {
+      "epoch": 0.020110305628477158,
+      "grad_norm": 1.45587158203125,
+      "learning_rate": 4.015406836783823e-07,
+      "loss": 1.1026,
+      "step": 835
+    },
+    {
+      "epoch": 0.020230726620264445,
+      "grad_norm": 1.5896087884902954,
+      "learning_rate": 4.039480019258546e-07,
+      "loss": 1.1299,
+      "step": 840
+    },
+    {
+      "epoch": 0.020351147612051732,
+      "grad_norm": 1.5886921882629395,
+      "learning_rate": 4.0635532017332693e-07,
+      "loss": 1.0504,
+      "step": 845
+    },
+    {
+      "epoch": 0.020471568603839023,
+      "grad_norm": 1.5642247200012207,
+      "learning_rate": 4.0876263842079925e-07,
+      "loss": 1.0425,
+      "step": 850
+    },
+    {
+      "epoch": 0.02059198959562631,
+      "grad_norm": 1.7490164041519165,
+      "learning_rate": 4.111699566682715e-07,
+      "loss": 1.047,
+      "step": 855
+    },
+    {
+      "epoch": 0.020712410587413597,
+      "grad_norm": 1.6565054655075073,
+      "learning_rate": 4.135772749157438e-07,
+      "loss": 1.0785,
+      "step": 860
+    },
+    {
+      "epoch": 0.020832831579200885,
+      "grad_norm": 1.748234510421753,
+      "learning_rate": 4.1598459316321614e-07,
+      "loss": 1.09,
+      "step": 865
+    },
+    {
+      "epoch": 0.020953252570988175,
+      "grad_norm": 1.5565547943115234,
+      "learning_rate": 4.1839191141068845e-07,
+      "loss": 1.043,
+      "step": 870
+    },
+    {
+      "epoch": 0.021073673562775463,
+      "grad_norm": 1.482420802116394,
+      "learning_rate": 4.2079922965816076e-07,
+      "loss": 1.0658,
+      "step": 875
+    },
+    {
+      "epoch": 0.02119409455456275,
+      "grad_norm": 1.6103140115737915,
+      "learning_rate": 4.232065479056331e-07,
+      "loss": 1.0916,
+      "step": 880
+    },
+    {
+      "epoch": 0.02131451554635004,
+      "grad_norm": 1.6495821475982666,
+      "learning_rate": 4.256138661531054e-07,
+      "loss": 1.0948,
+      "step": 885
+    },
+    {
+      "epoch": 0.021434936538137328,
+      "grad_norm": 1.6755701303482056,
+      "learning_rate": 4.280211844005777e-07,
+      "loss": 1.0383,
+      "step": 890
+    },
+    {
+      "epoch": 0.021555357529924615,
+      "grad_norm": 1.5431663990020752,
+      "learning_rate": 4.3042850264805e-07,
+      "loss": 1.1197,
+      "step": 895
+    },
+    {
+      "epoch": 0.021675778521711906,
+      "grad_norm": 1.7741978168487549,
+      "learning_rate": 4.3283582089552234e-07,
+      "loss": 1.0617,
+      "step": 900
+    },
+    {
+      "epoch": 0.021796199513499193,
+      "grad_norm": 1.4438221454620361,
+      "learning_rate": 4.352431391429947e-07,
+      "loss": 1.0777,
+      "step": 905
+    },
+    {
+      "epoch": 0.02191662050528648,
+      "grad_norm": 1.625065565109253,
+      "learning_rate": 4.37650457390467e-07,
+      "loss": 1.01,
+      "step": 910
+    },
+    {
+      "epoch": 0.02203704149707377,
+      "grad_norm": 1.5660688877105713,
+      "learning_rate": 4.4005777563793933e-07,
+      "loss": 1.0914,
+      "step": 915
+    },
+    {
+      "epoch": 0.02215746248886106,
+      "grad_norm": 1.5976072549819946,
+      "learning_rate": 4.4246509388541164e-07,
+      "loss": 1.0577,
+      "step": 920
+    },
+    {
+      "epoch": 0.022277883480648346,
+      "grad_norm": 1.7111833095550537,
+      "learning_rate": 4.4487241213288396e-07,
+      "loss": 1.0814,
+      "step": 925
+    },
+    {
+      "epoch": 0.022398304472435637,
+      "grad_norm": 1.7777079343795776,
+      "learning_rate": 4.4727973038035627e-07,
+      "loss": 1.0339,
+      "step": 930
+    },
+    {
+      "epoch": 0.022518725464222924,
+      "grad_norm": 1.6317858695983887,
+      "learning_rate": 4.496870486278286e-07,
+      "loss": 1.0387,
+      "step": 935
+    },
+    {
+      "epoch": 0.02263914645601021,
+      "grad_norm": 1.6453614234924316,
+      "learning_rate": 4.520943668753009e-07,
+      "loss": 1.078,
+      "step": 940
+    },
+    {
+      "epoch": 0.0227595674477975,
+      "grad_norm": 1.5312925577163696,
+      "learning_rate": 4.545016851227732e-07,
+      "loss": 1.0616,
+      "step": 945
+    },
+    {
+      "epoch": 0.02287998843958479,
+      "grad_norm": 1.4723235368728638,
+      "learning_rate": 4.5690900337024553e-07,
+      "loss": 1.054,
+      "step": 950
+    },
+    {
+      "epoch": 0.023000409431372076,
+      "grad_norm": 1.6957050561904907,
+      "learning_rate": 4.5931632161771784e-07,
+      "loss": 1.0612,
+      "step": 955
+    },
+    {
+      "epoch": 0.023120830423159364,
+      "grad_norm": 1.5759074687957764,
+      "learning_rate": 4.6172363986519016e-07,
+      "loss": 1.005,
+      "step": 960
+    },
+    {
+      "epoch": 0.023241251414946654,
+      "grad_norm": 1.7994741201400757,
+      "learning_rate": 4.6413095811266247e-07,
+      "loss": 1.115,
+      "step": 965
+    },
+    {
+      "epoch": 0.02336167240673394,
+      "grad_norm": 1.5642887353897095,
+      "learning_rate": 4.665382763601348e-07,
+      "loss": 1.1111,
+      "step": 970
+    },
+    {
+      "epoch": 0.02348209339852123,
+      "grad_norm": 1.8060336112976074,
+      "learning_rate": 4.6894559460760715e-07,
+      "loss": 1.0129,
+      "step": 975
+    },
+    {
+      "epoch": 0.02360251439030852,
+      "grad_norm": 1.6655142307281494,
+      "learning_rate": 4.7135291285507947e-07,
+      "loss": 1.0414,
+      "step": 980
+    },
+    {
+      "epoch": 0.023722935382095807,
+      "grad_norm": 1.7621194124221802,
+      "learning_rate": 4.737602311025518e-07,
+      "loss": 1.1035,
+      "step": 985
+    },
+    {
+      "epoch": 0.023843356373883094,
+      "grad_norm": 1.5201829671859741,
+      "learning_rate": 4.7616754935002404e-07,
+      "loss": 1.0852,
+      "step": 990
+    },
+    {
+      "epoch": 0.023963777365670385,
+      "grad_norm": 1.6027559041976929,
+      "learning_rate": 4.785748675974964e-07,
+      "loss": 1.0891,
+      "step": 995
+    },
+    {
+      "epoch": 0.024084198357457672,
+      "grad_norm": 1.7190868854522705,
+      "learning_rate": 4.809821858449687e-07,
+      "loss": 1.092,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02420461934924496,
+      "grad_norm": 1.8872677087783813,
+      "learning_rate": 4.83389504092441e-07,
+      "loss": 1.0614,
+      "step": 1005
+    },
+    {
+      "epoch": 0.02432504034103225,
+      "grad_norm": 1.4623500108718872,
+      "learning_rate": 4.857968223399134e-07,
+      "loss": 1.0281,
+      "step": 1010
+    },
+    {
+      "epoch": 0.024445461332819537,
+      "grad_norm": 1.6613128185272217,
+      "learning_rate": 4.882041405873856e-07,
+      "loss": 1.0937,
+      "step": 1015
+    },
+    {
+      "epoch": 0.024565882324606825,
+      "grad_norm": 1.533061146736145,
+      "learning_rate": 4.90611458834858e-07,
+      "loss": 1.0469,
+      "step": 1020
+    },
+    {
+      "epoch": 0.024686303316394115,
+      "grad_norm": 1.6154255867004395,
+      "learning_rate": 4.930187770823302e-07,
+      "loss": 1.0317,
+      "step": 1025
+    },
+    {
+      "epoch": 0.024806724308181403,
+      "grad_norm": 2.1915018558502197,
+      "learning_rate": 4.954260953298026e-07,
+      "loss": 1.0589,
+      "step": 1030
+    },
+    {
+      "epoch": 0.02492714529996869,
+      "grad_norm": 1.678594708442688,
+      "learning_rate": 4.978334135772749e-07,
+      "loss": 0.9996,
+      "step": 1035
+    },
+    {
+      "epoch": 0.025047566291755977,
+      "grad_norm": 2.124476909637451,
+      "learning_rate": 5.002407318247472e-07,
+      "loss": 1.0824,
+      "step": 1040
+    },
+    {
+      "epoch": 0.025167987283543268,
+      "grad_norm": 1.5182719230651855,
+      "learning_rate": 5.026480500722196e-07,
+      "loss": 1.0825,
+      "step": 1045
+    },
+    {
+      "epoch": 0.025288408275330555,
+      "grad_norm": 2.0846803188323975,
+      "learning_rate": 5.050553683196919e-07,
+      "loss": 1.037,
+      "step": 1050
+    },
+    {
+      "epoch": 0.025408829267117843,
+      "grad_norm": 1.5982701778411865,
+      "learning_rate": 5.074626865671642e-07,
+      "loss": 0.988,
+      "step": 1055
+    },
+    {
+      "epoch": 0.025529250258905133,
+      "grad_norm": 1.4213273525238037,
+      "learning_rate": 5.098700048146365e-07,
+      "loss": 1.0748,
+      "step": 1060
+    },
+    {
+      "epoch": 0.02564967125069242,
+      "grad_norm": 1.5166436433792114,
+      "learning_rate": 5.122773230621088e-07,
+      "loss": 1.0678,
+      "step": 1065
+    },
+    {
+      "epoch": 0.025770092242479708,
+      "grad_norm": 1.5300348997116089,
+      "learning_rate": 5.146846413095811e-07,
+      "loss": 1.0731,
+      "step": 1070
+    },
+    {
+      "epoch": 0.025890513234267,
+      "grad_norm": 1.6487699747085571,
+      "learning_rate": 5.170919595570534e-07,
+      "loss": 1.0892,
+      "step": 1075
+    },
+    {
+      "epoch": 0.026010934226054286,
+      "grad_norm": 1.641617774963379,
+      "learning_rate": 5.194992778045257e-07,
+      "loss": 1.0668,
+      "step": 1080
+    },
+    {
+      "epoch": 0.026131355217841573,
+      "grad_norm": 1.632634162902832,
+      "learning_rate": 5.21906596051998e-07,
+      "loss": 1.1131,
+      "step": 1085
+    },
+    {
+      "epoch": 0.026251776209628864,
+      "grad_norm": 1.6787406206130981,
+      "learning_rate": 5.243139142994704e-07,
+      "loss": 1.0292,
+      "step": 1090
+    },
+    {
+      "epoch": 0.02637219720141615,
+      "grad_norm": 1.5313917398452759,
+      "learning_rate": 5.267212325469426e-07,
+      "loss": 1.0467,
+      "step": 1095
+    },
+    {
+      "epoch": 0.02649261819320344,
+      "grad_norm": 1.5146945714950562,
+      "learning_rate": 5.29128550794415e-07,
+      "loss": 1.097,
+      "step": 1100
+    },
+    {
+      "epoch": 0.02661303918499073,
+      "grad_norm": 1.5477287769317627,
+      "learning_rate": 5.315358690418873e-07,
+      "loss": 1.0578,
+      "step": 1105
+    },
+    {
+      "epoch": 0.026733460176778016,
+      "grad_norm": 1.5698777437210083,
+      "learning_rate": 5.339431872893596e-07,
+      "loss": 1.0324,
+      "step": 1110
+    },
+    {
+      "epoch": 0.026853881168565304,
+      "grad_norm": 1.6078376770019531,
+      "learning_rate": 5.363505055368319e-07,
+      "loss": 1.0931,
+      "step": 1115
+    },
+    {
+      "epoch": 0.026974302160352594,
+      "grad_norm": 1.6052172183990479,
+      "learning_rate": 5.387578237843043e-07,
+      "loss": 0.9986,
+      "step": 1120
+    },
+    {
+      "epoch": 0.02709472315213988,
+      "grad_norm": 1.492550015449524,
+      "learning_rate": 5.411651420317765e-07,
+      "loss": 1.0325,
+      "step": 1125
+    },
+    {
+      "epoch": 0.02721514414392717,
+      "grad_norm": 1.6259959936141968,
+      "learning_rate": 5.435724602792489e-07,
+      "loss": 1.0123,
+      "step": 1130
+    },
+    {
+      "epoch": 0.027335565135714456,
+      "grad_norm": 1.3507508039474487,
+      "learning_rate": 5.459797785267211e-07,
+      "loss": 1.0907,
+      "step": 1135
+    },
+    {
+      "epoch": 0.027455986127501747,
+      "grad_norm": 1.743194818496704,
+      "learning_rate": 5.483870967741935e-07,
+      "loss": 1.0432,
+      "step": 1140
+    },
+    {
+      "epoch": 0.027576407119289034,
+      "grad_norm": 1.5899051427841187,
+      "learning_rate": 5.507944150216658e-07,
+      "loss": 1.0519,
+      "step": 1145
+    },
+    {
+      "epoch": 0.02769682811107632,
+      "grad_norm": 1.5947670936584473,
+      "learning_rate": 5.532017332691381e-07,
+      "loss": 1.1119,
+      "step": 1150
+    },
+    {
+      "epoch": 0.027817249102863612,
+      "grad_norm": 1.6205607652664185,
+      "learning_rate": 5.556090515166104e-07,
+      "loss": 1.0232,
+      "step": 1155
+    },
+    {
+      "epoch": 0.0279376700946509,
+      "grad_norm": 1.6683999300003052,
+      "learning_rate": 5.580163697640829e-07,
+      "loss": 1.0367,
+      "step": 1160
+    },
+    {
+      "epoch": 0.028058091086438187,
+      "grad_norm": 1.691907525062561,
+      "learning_rate": 5.604236880115551e-07,
+      "loss": 1.0982,
+      "step": 1165
+    },
+    {
+      "epoch": 0.028178512078225477,
+      "grad_norm": 1.7351024150848389,
+      "learning_rate": 5.628310062590275e-07,
+      "loss": 1.0043,
+      "step": 1170
+    },
+    {
+      "epoch": 0.028298933070012765,
+      "grad_norm": 1.624670386314392,
+      "learning_rate": 5.652383245064998e-07,
+      "loss": 1.0359,
+      "step": 1175
+    },
+    {
+      "epoch": 0.028419354061800052,
+      "grad_norm": 1.611996054649353,
+      "learning_rate": 5.676456427539721e-07,
+      "loss": 1.0703,
+      "step": 1180
+    },
+    {
+      "epoch": 0.028539775053587343,
+      "grad_norm": 1.5940327644348145,
+      "learning_rate": 5.700529610014444e-07,
+      "loss": 1.0034,
+      "step": 1185
+    },
+    {
+      "epoch": 0.02866019604537463,
+      "grad_norm": 1.5508549213409424,
+      "learning_rate": 5.724602792489168e-07,
+      "loss": 0.9925,
+      "step": 1190
+    },
+    {
+      "epoch": 0.028780617037161917,
+      "grad_norm": 1.7031129598617554,
+      "learning_rate": 5.74867597496389e-07,
+      "loss": 1.0722,
+      "step": 1195
+    },
+    {
+      "epoch": 0.028901038028949208,
+      "grad_norm": 1.542175531387329,
+      "learning_rate": 5.772749157438613e-07,
+      "loss": 1.0558,
+      "step": 1200
+    },
+    {
+      "epoch": 0.029021459020736495,
+      "grad_norm": 1.6221057176589966,
+      "learning_rate": 5.796822339913337e-07,
+      "loss": 0.9962,
+      "step": 1205
+    },
+    {
+      "epoch": 0.029141880012523783,
+      "grad_norm": 1.7214235067367554,
+      "learning_rate": 5.820895522388059e-07,
+      "loss": 1.0642,
+      "step": 1210
+    },
+    {
+      "epoch": 0.02926230100431107,
+      "grad_norm": 1.7093576192855835,
+      "learning_rate": 5.844968704862783e-07,
+      "loss": 1.0311,
+      "step": 1215
+    },
+    {
+      "epoch": 0.02938272199609836,
+      "grad_norm": 1.5216920375823975,
+      "learning_rate": 5.869041887337505e-07,
+      "loss": 1.0486,
+      "step": 1220
+    },
+    {
+      "epoch": 0.029503142987885648,
+      "grad_norm": 1.604498028755188,
+      "learning_rate": 5.893115069812229e-07,
+      "loss": 1.0437,
+      "step": 1225
+    },
+    {
+      "epoch": 0.029623563979672935,
+      "grad_norm": 1.562648057937622,
+      "learning_rate": 5.917188252286952e-07,
+      "loss": 1.1011,
+      "step": 1230
+    },
+    {
+      "epoch": 0.029743984971460226,
+      "grad_norm": 1.6788482666015625,
+      "learning_rate": 5.941261434761675e-07,
+      "loss": 1.0703,
+      "step": 1235
+    },
+    {
+      "epoch": 0.029864405963247513,
+      "grad_norm": 1.5976691246032715,
+      "learning_rate": 5.965334617236398e-07,
+      "loss": 1.0202,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0299848269550348,
+      "grad_norm": 1.937235713005066,
+      "learning_rate": 5.989407799711122e-07,
+      "loss": 1.0541,
+      "step": 1245
+    },
+    {
+      "epoch": 0.03010524794682209,
+      "grad_norm": 1.5281552076339722,
+      "learning_rate": 6.013480982185844e-07,
+      "loss": 1.0317,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03022566893860938,
+      "grad_norm": 1.8217495679855347,
+      "learning_rate": 6.037554164660568e-07,
+      "loss": 1.0604,
+      "step": 1255
+    },
+    {
+      "epoch": 0.030346089930396666,
+      "grad_norm": 1.394348382949829,
+      "learning_rate": 6.061627347135291e-07,
+      "loss": 1.0598,
+      "step": 1260
+    },
+    {
+      "epoch": 0.030466510922183956,
+      "grad_norm": 1.6202285289764404,
+      "learning_rate": 6.085700529610014e-07,
+      "loss": 1.0583,
+      "step": 1265
+    },
+    {
+      "epoch": 0.030586931913971244,
+      "grad_norm": 1.7116400003433228,
+      "learning_rate": 6.109773712084737e-07,
+      "loss": 1.0783,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03070735290575853,
+      "grad_norm": 1.7912710905075073,
+      "learning_rate": 6.13384689455946e-07,
+      "loss": 1.0799,
+      "step": 1275
+    },
+    {
+      "epoch": 0.03082777389754582,
+      "grad_norm": 1.6622925996780396,
+      "learning_rate": 6.157920077034183e-07,
+      "loss": 1.0426,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03094819488933311,
+      "grad_norm": 1.9057449102401733,
+      "learning_rate": 6.181993259508907e-07,
+      "loss": 1.064,
+      "step": 1285
+    },
+    {
+      "epoch": 0.031068615881120396,
+      "grad_norm": 1.7786000967025757,
+      "learning_rate": 6.206066441983629e-07,
+      "loss": 1.062,
+      "step": 1290
+    },
+    {
+      "epoch": 0.031189036872907687,
+      "grad_norm": 1.4491337537765503,
+      "learning_rate": 6.230139624458353e-07,
+      "loss": 1.0659,
+      "step": 1295
+    },
+    {
+      "epoch": 0.031309457864694974,
+      "grad_norm": 1.662503957748413,
+      "learning_rate": 6.254212806933076e-07,
+      "loss": 1.0827,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03142987885648226,
+      "grad_norm": 1.7556451559066772,
+      "learning_rate": 6.2782859894078e-07,
+      "loss": 1.0,
+      "step": 1305
+    },
+    {
+      "epoch": 0.03155029984826955,
+      "grad_norm": 1.7393630743026733,
+      "learning_rate": 6.302359171882523e-07,
+      "loss": 1.046,
+      "step": 1310
+    },
+    {
+      "epoch": 0.031670720840056836,
+      "grad_norm": 1.614888310432434,
+      "learning_rate": 6.326432354357247e-07,
+      "loss": 1.04,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03179114183184413,
+      "grad_norm": 1.588411808013916,
+      "learning_rate": 6.350505536831969e-07,
+      "loss": 1.1276,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03191156282363142,
+      "grad_norm": 1.5537513494491577,
+      "learning_rate": 6.374578719306693e-07,
+      "loss": 1.0437,
+      "step": 1325
+    },
+    {
+      "epoch": 0.032031983815418705,
+      "grad_norm": 1.452207326889038,
+      "learning_rate": 6.398651901781416e-07,
+      "loss": 1.0394,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03215240480720599,
+      "grad_norm": 1.6652417182922363,
+      "learning_rate": 6.422725084256138e-07,
+      "loss": 1.021,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03227282579899328,
+      "grad_norm": 1.5968703031539917,
+      "learning_rate": 6.446798266730862e-07,
+      "loss": 1.0361,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03239324679078057,
+      "grad_norm": 1.6294167041778564,
+      "learning_rate": 6.470871449205584e-07,
+      "loss": 1.0288,
+      "step": 1345
+    },
+    {
+      "epoch": 0.032513667782567854,
+      "grad_norm": 2.108187675476074,
+      "learning_rate": 6.494944631680308e-07,
+      "loss": 1.1072,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03263408877435515,
+      "grad_norm": 1.5419865846633911,
+      "learning_rate": 6.519017814155031e-07,
+      "loss": 0.9997,
+      "step": 1355
+    },
+    {
+      "epoch": 0.032754509766142435,
+      "grad_norm": 1.8960027694702148,
+      "learning_rate": 6.543090996629754e-07,
+      "loss": 1.0657,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03287493075792972,
+      "grad_norm": 1.8205565214157104,
+      "learning_rate": 6.567164179104477e-07,
+      "loss": 1.0377,
+      "step": 1365
+    },
+    {
+      "epoch": 0.03299535174971701,
+      "grad_norm": 1.4616715908050537,
+      "learning_rate": 6.591237361579201e-07,
+      "loss": 1.0416,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0331157727415043,
+      "grad_norm": 1.6263893842697144,
+      "learning_rate": 6.615310544053923e-07,
+      "loss": 1.0303,
+      "step": 1375
+    },
+    {
+      "epoch": 0.033236193733291584,
+      "grad_norm": 1.5705087184906006,
+      "learning_rate": 6.639383726528647e-07,
+      "loss": 1.0414,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03335661472507888,
+      "grad_norm": 1.559043526649475,
+      "learning_rate": 6.66345690900337e-07,
+      "loss": 1.0892,
+      "step": 1385
+    },
+    {
+      "epoch": 0.033477035716866166,
+      "grad_norm": 1.5846097469329834,
+      "learning_rate": 6.687530091478093e-07,
+      "loss": 1.0132,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03359745670865345,
+      "grad_norm": 1.6560180187225342,
+      "learning_rate": 6.711603273952816e-07,
+      "loss": 1.0384,
+      "step": 1395
+    },
+    {
+      "epoch": 0.03371787770044074,
+      "grad_norm": 1.416507601737976,
+      "learning_rate": 6.73567645642754e-07,
+      "loss": 0.9696,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03383829869222803,
+      "grad_norm": 1.463869333267212,
+      "learning_rate": 6.759749638902262e-07,
+      "loss": 1.0099,
+      "step": 1405
+    },
+    {
+      "epoch": 0.033958719684015315,
+      "grad_norm": 1.4500248432159424,
+      "learning_rate": 6.783822821376986e-07,
+      "loss": 1.0786,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03407914067580261,
+      "grad_norm": 1.8260598182678223,
+      "learning_rate": 6.807896003851708e-07,
+      "loss": 1.0813,
+      "step": 1415
+    },
+    {
+      "epoch": 0.034199561667589896,
+      "grad_norm": 1.6602531671524048,
+      "learning_rate": 6.831969186326432e-07,
+      "loss": 0.9867,
+      "step": 1420
+    },
+    {
+      "epoch": 0.034319982659377184,
+      "grad_norm": 1.6317486763000488,
+      "learning_rate": 6.856042368801155e-07,
+      "loss": 0.9797,
+      "step": 1425
+    },
+    {
+      "epoch": 0.03444040365116447,
+      "grad_norm": 1.5111490488052368,
+      "learning_rate": 6.880115551275878e-07,
+      "loss": 1.0456,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03456082464295176,
+      "grad_norm": 1.5642411708831787,
+      "learning_rate": 6.904188733750601e-07,
+      "loss": 0.9404,
+      "step": 1435
+    },
+    {
+      "epoch": 0.034681245634739045,
+      "grad_norm": 1.9178805351257324,
+      "learning_rate": 6.928261916225325e-07,
+      "loss": 1.0056,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03480166662652633,
+      "grad_norm": 1.6757012605667114,
+      "learning_rate": 6.952335098700047e-07,
+      "loss": 1.0829,
+      "step": 1445
+    },
+    {
+      "epoch": 0.03492208761831363,
+      "grad_norm": 1.7237210273742676,
+      "learning_rate": 6.976408281174772e-07,
+      "loss": 1.0273,
+      "step": 1450
+    },
+    {
+      "epoch": 0.035042508610100914,
+      "grad_norm": 2.0043551921844482,
+      "learning_rate": 7.000481463649495e-07,
+      "loss": 1.0238,
+      "step": 1455
+    },
+    {
+      "epoch": 0.0351629296018882,
+      "grad_norm": 1.5664793252944946,
+      "learning_rate": 7.024554646124218e-07,
+      "loss": 0.9788,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03528335059367549,
+      "grad_norm": 1.595779299736023,
+      "learning_rate": 7.048627828598941e-07,
+      "loss": 1.0589,
+      "step": 1465
+    },
+    {
+      "epoch": 0.035403771585462776,
+      "grad_norm": 1.4906885623931885,
+      "learning_rate": 7.072701011073664e-07,
+      "loss": 1.0575,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03552419257725006,
+      "grad_norm": 1.6824171543121338,
+      "learning_rate": 7.096774193548387e-07,
+      "loss": 1.074,
+      "step": 1475
+    },
+    {
+      "epoch": 0.03564461356903736,
+      "grad_norm": 1.4851865768432617,
+      "learning_rate": 7.12084737602311e-07,
+      "loss": 1.0499,
+      "step": 1480
+    },
+    {
+      "epoch": 0.035765034560824645,
+      "grad_norm": 1.6125143766403198,
+      "learning_rate": 7.144920558497833e-07,
+      "loss": 1.0288,
+      "step": 1485
+    },
+    {
+      "epoch": 0.03588545555261193,
+      "grad_norm": 1.6875344514846802,
+      "learning_rate": 7.168993740972556e-07,
+      "loss": 1.0351,
+      "step": 1490
+    },
+    {
+      "epoch": 0.03600587654439922,
+      "grad_norm": 1.6106306314468384,
+      "learning_rate": 7.19306692344728e-07,
+      "loss": 1.0325,
+      "step": 1495
+    },
+    {
+      "epoch": 0.03612629753618651,
+      "grad_norm": 1.4459638595581055,
+      "learning_rate": 7.217140105922002e-07,
+      "loss": 1.0201,
+      "step": 1500
+    },
+    {
+      "epoch": 0.036246718527973794,
+      "grad_norm": 1.626741647720337,
+      "learning_rate": 7.241213288396726e-07,
+      "loss": 1.0355,
+      "step": 1505
+    },
+    {
+      "epoch": 0.03636713951976109,
+      "grad_norm": 1.4596201181411743,
+      "learning_rate": 7.265286470871449e-07,
+      "loss": 1.0033,
+      "step": 1510
+    },
+    {
+      "epoch": 0.036487560511548375,
+      "grad_norm": 1.3791264295578003,
+      "learning_rate": 7.289359653346172e-07,
+      "loss": 1.0504,
+      "step": 1515
+    },
+    {
+      "epoch": 0.03660798150333566,
+      "grad_norm": 1.399931788444519,
+      "learning_rate": 7.313432835820895e-07,
+      "loss": 1.0174,
+      "step": 1520
+    },
+    {
+      "epoch": 0.03672840249512295,
+      "grad_norm": 1.4853743314743042,
+      "learning_rate": 7.337506018295619e-07,
+      "loss": 1.0472,
+      "step": 1525
+    },
+    {
+      "epoch": 0.03684882348691024,
+      "grad_norm": 1.5996602773666382,
+      "learning_rate": 7.361579200770341e-07,
+      "loss": 1.0267,
+      "step": 1530
+    },
+    {
+      "epoch": 0.036969244478697524,
+      "grad_norm": 1.67296302318573,
+      "learning_rate": 7.385652383245065e-07,
+      "loss": 1.0361,
+      "step": 1535
+    },
+    {
+      "epoch": 0.03708966547048481,
+      "grad_norm": 1.6642286777496338,
+      "learning_rate": 7.409725565719787e-07,
+      "loss": 1.0457,
+      "step": 1540
+    },
+    {
+      "epoch": 0.037210086462272106,
+      "grad_norm": 1.4878441095352173,
+      "learning_rate": 7.433798748194511e-07,
+      "loss": 1.0817,
+      "step": 1545
+    },
+    {
+      "epoch": 0.03733050745405939,
+      "grad_norm": 1.816658854484558,
+      "learning_rate": 7.457871930669234e-07,
+      "loss": 1.0329,
+      "step": 1550
+    },
+    {
+      "epoch": 0.03745092844584668,
+      "grad_norm": 1.480438470840454,
+      "learning_rate": 7.481945113143957e-07,
+      "loss": 1.0022,
+      "step": 1555
+    },
+    {
+      "epoch": 0.03757134943763397,
+      "grad_norm": 1.4642038345336914,
+      "learning_rate": 7.50601829561868e-07,
+      "loss": 1.0261,
+      "step": 1560
+    },
+    {
+      "epoch": 0.037691770429421255,
+      "grad_norm": 1.5570502281188965,
+      "learning_rate": 7.530091478093404e-07,
+      "loss": 1.0731,
+      "step": 1565
+    },
+    {
+      "epoch": 0.03781219142120854,
+      "grad_norm": 1.5301499366760254,
+      "learning_rate": 7.554164660568126e-07,
+      "loss": 1.0498,
+      "step": 1570
+    },
+    {
+      "epoch": 0.037932612412995836,
+      "grad_norm": 1.3333464860916138,
+      "learning_rate": 7.57823784304285e-07,
+      "loss": 0.9946,
+      "step": 1575
+    },
+    {
+      "epoch": 0.038053033404783124,
+      "grad_norm": 1.6789193153381348,
+      "learning_rate": 7.602311025517573e-07,
+      "loss": 0.9817,
+      "step": 1580
+    },
+    {
+      "epoch": 0.03817345439657041,
+      "grad_norm": 1.4923834800720215,
+      "learning_rate": 7.626384207992296e-07,
+      "loss": 1.0567,
+      "step": 1585
+    },
+    {
+      "epoch": 0.0382938753883577,
+      "grad_norm": 1.7668758630752563,
+      "learning_rate": 7.650457390467019e-07,
+      "loss": 0.984,
+      "step": 1590
+    },
+    {
+      "epoch": 0.038414296380144985,
+      "grad_norm": 1.4018738269805908,
+      "learning_rate": 7.674530572941742e-07,
+      "loss": 0.9654,
+      "step": 1595
+    },
+    {
+      "epoch": 0.03853471737193227,
+      "grad_norm": 1.733635425567627,
+      "learning_rate": 7.698603755416466e-07,
+      "loss": 1.0245,
+      "step": 1600
+    },
+    {
+      "epoch": 0.03865513836371957,
+      "grad_norm": 1.5820351839065552,
+      "learning_rate": 7.722676937891189e-07,
+      "loss": 1.0502,
+      "step": 1605
+    },
+    {
+      "epoch": 0.038775559355506854,
+      "grad_norm": 1.7299373149871826,
+      "learning_rate": 7.746750120365913e-07,
+      "loss": 1.0013,
+      "step": 1610
+    },
+    {
+      "epoch": 0.03889598034729414,
+      "grad_norm": 1.5306358337402344,
+      "learning_rate": 7.770823302840635e-07,
+      "loss": 1.0434,
+      "step": 1615
+    },
+    {
+      "epoch": 0.03901640133908143,
+      "grad_norm": 1.6972306966781616,
+      "learning_rate": 7.794896485315359e-07,
+      "loss": 1.0443,
+      "step": 1620
+    },
+    {
+      "epoch": 0.039136822330868716,
+      "grad_norm": 1.8630884885787964,
+      "learning_rate": 7.818969667790081e-07,
+      "loss": 0.9966,
+      "step": 1625
+    },
+    {
+      "epoch": 0.039257243322656,
+      "grad_norm": 1.4289274215698242,
+      "learning_rate": 7.843042850264805e-07,
+      "loss": 1.0051,
+      "step": 1630
+    },
+    {
+      "epoch": 0.03937766431444329,
+      "grad_norm": 1.3741828203201294,
+      "learning_rate": 7.867116032739528e-07,
+      "loss": 0.9737,
+      "step": 1635
+    },
+    {
+      "epoch": 0.039498085306230585,
+      "grad_norm": 1.6088579893112183,
+      "learning_rate": 7.891189215214251e-07,
+      "loss": 1.0336,
+      "step": 1640
+    },
+    {
+      "epoch": 0.03961850629801787,
+      "grad_norm": 1.3409250974655151,
+      "learning_rate": 7.915262397688974e-07,
+      "loss": 1.0477,
+      "step": 1645
+    },
+    {
+      "epoch": 0.03973892728980516,
+      "grad_norm": 1.3570778369903564,
+      "learning_rate": 7.939335580163698e-07,
+      "loss": 0.9611,
+      "step": 1650
+    },
+    {
+      "epoch": 0.03985934828159245,
+      "grad_norm": 1.4607723951339722,
+      "learning_rate": 7.96340876263842e-07,
+      "loss": 1.0476,
+      "step": 1655
+    },
+    {
+      "epoch": 0.039979769273379734,
+      "grad_norm": 1.4103513956069946,
+      "learning_rate": 7.987481945113144e-07,
+      "loss": 1.0038,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04010019026516702,
+      "grad_norm": 1.7692028284072876,
+      "learning_rate": 8.011555127587867e-07,
+      "loss": 1.0486,
+      "step": 1665
+    },
+    {
+      "epoch": 0.040220611256954315,
+      "grad_norm": 1.7130166292190552,
+      "learning_rate": 8.03562831006259e-07,
+      "loss": 1.0355,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0403410322487416,
+      "grad_norm": 1.7267684936523438,
+      "learning_rate": 8.059701492537313e-07,
+      "loss": 0.9746,
+      "step": 1675
+    },
+    {
+      "epoch": 0.04046145324052889,
+      "grad_norm": 1.5484259128570557,
+      "learning_rate": 8.083774675012036e-07,
+      "loss": 1.08,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04058187423231618,
+      "grad_norm": 1.6294459104537964,
+      "learning_rate": 8.107847857486759e-07,
+      "loss": 1.0619,
+      "step": 1685
+    },
+    {
+      "epoch": 0.040702295224103464,
+      "grad_norm": 1.4222997426986694,
+      "learning_rate": 8.131921039961483e-07,
+      "loss": 1.0091,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04082271621589075,
+      "grad_norm": 1.7014527320861816,
+      "learning_rate": 8.155994222436205e-07,
+      "loss": 1.0323,
+      "step": 1695
+    },
+    {
+      "epoch": 0.040943137207678046,
+      "grad_norm": 1.7980411052703857,
+      "learning_rate": 8.180067404910929e-07,
+      "loss": 1.064,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04106355819946533,
+      "grad_norm": 1.468579888343811,
+      "learning_rate": 8.204140587385652e-07,
+      "loss": 0.9758,
+      "step": 1705
+    },
+    {
+      "epoch": 0.04118397919125262,
+      "grad_norm": 1.6334229707717896,
+      "learning_rate": 8.228213769860375e-07,
+      "loss": 1.0022,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04130440018303991,
+      "grad_norm": 1.5733845233917236,
+      "learning_rate": 8.252286952335098e-07,
+      "loss": 1.0361,
+      "step": 1715
+    },
+    {
+      "epoch": 0.041424821174827195,
+      "grad_norm": 1.8664427995681763,
+      "learning_rate": 8.276360134809822e-07,
+      "loss": 1.0545,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04154524216661448,
+      "grad_norm": 1.562538743019104,
+      "learning_rate": 8.300433317284544e-07,
+      "loss": 1.0639,
+      "step": 1725
+    },
+    {
+      "epoch": 0.04166566315840177,
+      "grad_norm": 1.882519245147705,
+      "learning_rate": 8.324506499759267e-07,
+      "loss": 0.9948,
+      "step": 1730
+    },
+    {
+      "epoch": 0.041786084150189064,
+      "grad_norm": 1.424924373626709,
+      "learning_rate": 8.34857968223399e-07,
+      "loss": 1.0339,
+      "step": 1735
+    },
+    {
+      "epoch": 0.04190650514197635,
+      "grad_norm": 1.9680640697479248,
+      "learning_rate": 8.372652864708713e-07,
+      "loss": 1.0502,
+      "step": 1740
+    },
+    {
+      "epoch": 0.04202692613376364,
+      "grad_norm": 1.8039034605026245,
+      "learning_rate": 8.396726047183438e-07,
+      "loss": 1.0537,
+      "step": 1745
+    },
+    {
+      "epoch": 0.042147347125550925,
+      "grad_norm": 1.5843199491500854,
+      "learning_rate": 8.42079922965816e-07,
+      "loss": 1.0155,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04226776811733821,
+      "grad_norm": 1.2623687982559204,
+      "learning_rate": 8.444872412132884e-07,
+      "loss": 0.992,
+      "step": 1755
+    },
+    {
+      "epoch": 0.0423881891091255,
+      "grad_norm": 2.0166735649108887,
+      "learning_rate": 8.468945594607607e-07,
+      "loss": 1.0148,
+      "step": 1760
+    },
+    {
+      "epoch": 0.042508610100912794,
+      "grad_norm": 1.8024189472198486,
+      "learning_rate": 8.49301877708233e-07,
+      "loss": 1.0526,
+      "step": 1765
+    },
+    {
+      "epoch": 0.04262903109270008,
+      "grad_norm": 1.444460391998291,
+      "learning_rate": 8.517091959557053e-07,
+      "loss": 1.0287,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04274945208448737,
+      "grad_norm": 1.9382981061935425,
+      "learning_rate": 8.541165142031777e-07,
+      "loss": 1.0336,
+      "step": 1775
+    },
+    {
+      "epoch": 0.042869873076274656,
+      "grad_norm": 1.5555084943771362,
+      "learning_rate": 8.565238324506499e-07,
+      "loss": 1.0116,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04299029406806194,
+      "grad_norm": 1.4860131740570068,
+      "learning_rate": 8.589311506981223e-07,
+      "loss": 1.0343,
+      "step": 1785
+    },
+    {
+      "epoch": 0.04311071505984923,
+      "grad_norm": 1.5422332286834717,
+      "learning_rate": 8.613384689455946e-07,
+      "loss": 1.0307,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04323113605163652,
+      "grad_norm": 1.5868923664093018,
+      "learning_rate": 8.637457871930669e-07,
+      "loss": 1.0291,
+      "step": 1795
+    },
+    {
+      "epoch": 0.04335155704342381,
+      "grad_norm": 1.5933997631072998,
+      "learning_rate": 8.661531054405392e-07,
+      "loss": 1.0708,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0434719780352111,
+      "grad_norm": 1.8912562131881714,
+      "learning_rate": 8.685604236880116e-07,
+      "loss": 1.0921,
+      "step": 1805
+    },
+    {
+      "epoch": 0.04359239902699839,
+      "grad_norm": 1.5516586303710938,
+      "learning_rate": 8.709677419354838e-07,
+      "loss": 1.0165,
+      "step": 1810
+    },
+    {
+      "epoch": 0.043712820018785674,
+      "grad_norm": 1.5522100925445557,
+      "learning_rate": 8.733750601829562e-07,
+      "loss": 1.0371,
+      "step": 1815
+    },
+    {
+      "epoch": 0.04383324101057296,
+      "grad_norm": 1.924184799194336,
+      "learning_rate": 8.757823784304284e-07,
+      "loss": 1.0591,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04395366200236025,
+      "grad_norm": 1.5833402872085571,
+      "learning_rate": 8.781896966779008e-07,
+      "loss": 1.0504,
+      "step": 1825
+    },
+    {
+      "epoch": 0.04407408299414754,
+      "grad_norm": 1.7274823188781738,
+      "learning_rate": 8.805970149253731e-07,
+      "loss": 1.0242,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04419450398593483,
+      "grad_norm": 1.5076647996902466,
+      "learning_rate": 8.830043331728454e-07,
+      "loss": 0.9765,
+      "step": 1835
+    },
+    {
+      "epoch": 0.04431492497772212,
+      "grad_norm": 1.5085982084274292,
+      "learning_rate": 8.854116514203177e-07,
+      "loss": 1.0185,
+      "step": 1840
+    },
+    {
+      "epoch": 0.044435345969509404,
+      "grad_norm": 1.6915234327316284,
+      "learning_rate": 8.878189696677901e-07,
+      "loss": 1.0761,
+      "step": 1845
+    },
+    {
+      "epoch": 0.04455576696129669,
+      "grad_norm": 1.6578340530395508,
+      "learning_rate": 8.902262879152623e-07,
+      "loss": 1.0643,
+      "step": 1850
+    },
+    {
+      "epoch": 0.04467618795308398,
+      "grad_norm": 1.6258350610733032,
+      "learning_rate": 8.926336061627347e-07,
+      "loss": 1.0495,
+      "step": 1855
+    },
+    {
+      "epoch": 0.04479660894487127,
+      "grad_norm": 1.6774884462356567,
+      "learning_rate": 8.95040924410207e-07,
+      "loss": 1.0767,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04491702993665856,
+      "grad_norm": 1.6079150438308716,
+      "learning_rate": 8.974482426576792e-07,
+      "loss": 0.9649,
+      "step": 1865
+    },
+    {
+      "epoch": 0.04503745092844585,
+      "grad_norm": 1.6209561824798584,
+      "learning_rate": 8.998555609051516e-07,
+      "loss": 0.9927,
+      "step": 1870
+    },
+    {
+      "epoch": 0.045157871920233135,
+      "grad_norm": 1.641926884651184,
+      "learning_rate": 9.022628791526238e-07,
+      "loss": 1.029,
+      "step": 1875
+    },
+    {
+      "epoch": 0.04527829291202042,
+      "grad_norm": 1.6065603494644165,
+      "learning_rate": 9.046701974000962e-07,
+      "loss": 1.0018,
+      "step": 1880
+    },
+    {
+      "epoch": 0.04539871390380771,
+      "grad_norm": 1.5539517402648926,
+      "learning_rate": 9.070775156475685e-07,
+      "loss": 1.0798,
+      "step": 1885
+    },
+    {
+      "epoch": 0.045519134895595,
+      "grad_norm": 1.6610114574432373,
+      "learning_rate": 9.094848338950409e-07,
+      "loss": 1.0386,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04563955588738229,
+      "grad_norm": 1.5111974477767944,
+      "learning_rate": 9.118921521425132e-07,
+      "loss": 1.0406,
+      "step": 1895
+    },
+    {
+      "epoch": 0.04575997687916958,
+      "grad_norm": 2.0480804443359375,
+      "learning_rate": 9.142994703899856e-07,
+      "loss": 1.0007,
+      "step": 1900
+    },
+    {
+      "epoch": 0.045880397870956866,
+      "grad_norm": 1.7117605209350586,
+      "learning_rate": 9.167067886374578e-07,
+      "loss": 0.9986,
+      "step": 1905
+    },
+    {
+      "epoch": 0.04600081886274415,
+      "grad_norm": 1.505280613899231,
+      "learning_rate": 9.191141068849302e-07,
+      "loss": 1.0267,
+      "step": 1910
+    },
+    {
+      "epoch": 0.04612123985453144,
+      "grad_norm": 1.8697929382324219,
+      "learning_rate": 9.215214251324025e-07,
+      "loss": 1.0041,
+      "step": 1915
+    },
+    {
+      "epoch": 0.04624166084631873,
+      "grad_norm": 1.4194835424423218,
+      "learning_rate": 9.239287433798748e-07,
+      "loss": 1.0155,
+      "step": 1920
+    },
+    {
+      "epoch": 0.04636208183810602,
+      "grad_norm": 1.8681567907333374,
+      "learning_rate": 9.263360616273471e-07,
+      "loss": 1.0041,
+      "step": 1925
+    },
+    {
+      "epoch": 0.04648250282989331,
+      "grad_norm": 1.5799516439437866,
+      "learning_rate": 9.287433798748195e-07,
+      "loss": 0.996,
+      "step": 1930
+    },
+    {
+      "epoch": 0.046602923821680596,
+      "grad_norm": 1.585576057434082,
+      "learning_rate": 9.311506981222917e-07,
+      "loss": 1.0328,
+      "step": 1935
+    },
+    {
+      "epoch": 0.04672334481346788,
+      "grad_norm": 1.6536821126937866,
+      "learning_rate": 9.335580163697641e-07,
+      "loss": 1.0476,
+      "step": 1940
+    },
+    {
+      "epoch": 0.04684376580525517,
+      "grad_norm": 1.5552594661712646,
+      "learning_rate": 9.359653346172363e-07,
+      "loss": 1.0186,
+      "step": 1945
+    },
+    {
+      "epoch": 0.04696418679704246,
+      "grad_norm": 1.459959864616394,
+      "learning_rate": 9.383726528647087e-07,
+      "loss": 1.0039,
+      "step": 1950
+    },
+    {
+      "epoch": 0.04708460778882975,
+      "grad_norm": 1.63764226436615,
+      "learning_rate": 9.40779971112181e-07,
+      "loss": 1.0606,
+      "step": 1955
+    },
+    {
+      "epoch": 0.04720502878061704,
+      "grad_norm": 1.6388280391693115,
+      "learning_rate": 9.431872893596533e-07,
+      "loss": 0.9829,
+      "step": 1960
+    },
+    {
+      "epoch": 0.04732544977240433,
+      "grad_norm": 1.6173887252807617,
+      "learning_rate": 9.455946076071256e-07,
+      "loss": 1.0445,
+      "step": 1965
+    },
+    {
+      "epoch": 0.047445870764191614,
+      "grad_norm": 1.4738174676895142,
+      "learning_rate": 9.48001925854598e-07,
+      "loss": 1.0223,
+      "step": 1970
+    },
+    {
+      "epoch": 0.0475662917559789,
+      "grad_norm": 1.7925279140472412,
+      "learning_rate": 9.504092441020702e-07,
+      "loss": 1.0145,
+      "step": 1975
+    },
+    {
+      "epoch": 0.04768671274776619,
+      "grad_norm": 1.609230399131775,
+      "learning_rate": 9.528165623495426e-07,
+      "loss": 1.0355,
+      "step": 1980
+    },
+    {
+      "epoch": 0.047807133739553476,
+      "grad_norm": 1.5456503629684448,
+      "learning_rate": 9.552238805970149e-07,
+      "loss": 0.9816,
+      "step": 1985
+    },
+    {
+      "epoch": 0.04792755473134077,
+      "grad_norm": 1.7006056308746338,
+      "learning_rate": 9.576311988444872e-07,
+      "loss": 1.0325,
+      "step": 1990
+    },
+    {
+      "epoch": 0.04804797572312806,
+      "grad_norm": 2.056192636489868,
+      "learning_rate": 9.600385170919594e-07,
+      "loss": 1.0202,
+      "step": 1995
+    },
+    {
+      "epoch": 0.048168396714915344,
+      "grad_norm": 1.6051387786865234,
+      "learning_rate": 9.624458353394317e-07,
+      "loss": 0.9992,
+      "step": 2000
+    },
+    {
+      "epoch": 0.04828881770670263,
+      "grad_norm": 1.550959587097168,
+      "learning_rate": 9.648531535869041e-07,
+      "loss": 0.9809,
+      "step": 2005
+    },
+    {
+      "epoch": 0.04840923869848992,
+      "grad_norm": 1.8543497323989868,
+      "learning_rate": 9.672604718343765e-07,
+      "loss": 1.0177,
+      "step": 2010
+    },
+    {
+      "epoch": 0.048529659690277206,
+      "grad_norm": 1.6538995504379272,
+      "learning_rate": 9.696677900818486e-07,
+      "loss": 1.0336,
+      "step": 2015
+    },
+    {
+      "epoch": 0.0486500806820645,
+      "grad_norm": 1.763832926750183,
+      "learning_rate": 9.72075108329321e-07,
+      "loss": 0.9986,
+      "step": 2020
+    },
+    {
+      "epoch": 0.04877050167385179,
+      "grad_norm": 1.794128656387329,
+      "learning_rate": 9.744824265767934e-07,
+      "loss": 1.0349,
+      "step": 2025
+    },
+    {
+      "epoch": 0.048890922665639075,
+      "grad_norm": 1.6576578617095947,
+      "learning_rate": 9.768897448242657e-07,
+      "loss": 1.0408,
+      "step": 2030
+    },
+    {
+      "epoch": 0.04901134365742636,
+      "grad_norm": 1.539139986038208,
+      "learning_rate": 9.792970630717381e-07,
+      "loss": 0.9983,
+      "step": 2035
+    },
+    {
+      "epoch": 0.04913176464921365,
+      "grad_norm": 1.5743350982666016,
+      "learning_rate": 9.817043813192105e-07,
+      "loss": 1.0335,
+      "step": 2040
+    },
+    {
+      "epoch": 0.04925218564100094,
+      "grad_norm": 1.6326196193695068,
+      "learning_rate": 9.841116995666826e-07,
+      "loss": 1.0018,
+      "step": 2045
+    },
+    {
+      "epoch": 0.04937260663278823,
+      "grad_norm": 1.6246923208236694,
+      "learning_rate": 9.86519017814155e-07,
+      "loss": 0.9942,
+      "step": 2050
+    },
+    {
+      "epoch": 0.04949302762457552,
+      "grad_norm": 1.564854621887207,
+      "learning_rate": 9.889263360616274e-07,
+      "loss": 1.0211,
+      "step": 2055
+    },
+    {
+      "epoch": 0.049613448616362806,
+      "grad_norm": 1.5778380632400513,
+      "learning_rate": 9.913336543090997e-07,
+      "loss": 1.0246,
+      "step": 2060
+    },
+    {
+      "epoch": 0.04973386960815009,
+      "grad_norm": 1.7073304653167725,
+      "learning_rate": 9.937409725565719e-07,
+      "loss": 1.0102,
+      "step": 2065
+    },
+    {
+      "epoch": 0.04985429059993738,
+      "grad_norm": 1.4603596925735474,
+      "learning_rate": 9.961482908040443e-07,
+      "loss": 0.9871,
+      "step": 2070
+    },
+    {
+      "epoch": 0.04997471159172467,
+      "grad_norm": 1.6223366260528564,
+      "learning_rate": 9.985556090515166e-07,
+      "loss": 1.0269,
+      "step": 2075
+    },
+    {
+      "epoch": 0.050095132583511955,
+      "grad_norm": 1.8474642038345337,
+      "learning_rate": 9.999492952033262e-07,
+      "loss": 0.9398,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05021555357529925,
+      "grad_norm": 1.6487759351730347,
+      "learning_rate": 9.998225332116418e-07,
+      "loss": 0.9937,
+      "step": 2085
+    },
+    {
+      "epoch": 0.050335974567086536,
+      "grad_norm": 1.697744607925415,
+      "learning_rate": 9.996957712199575e-07,
+      "loss": 0.9984,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05045639555887382,
+      "grad_norm": 1.4020692110061646,
+      "learning_rate": 9.995690092282729e-07,
+      "loss": 0.9828,
+      "step": 2095
+    },
+    {
+      "epoch": 0.05057681655066111,
+      "grad_norm": 1.589512586593628,
+      "learning_rate": 9.994422472365885e-07,
+      "loss": 0.9955,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0506972375424484,
+      "grad_norm": 1.4988205432891846,
+      "learning_rate": 9.993154852449042e-07,
+      "loss": 1.0136,
+      "step": 2105
+    },
+    {
+      "epoch": 0.050817658534235685,
+      "grad_norm": 1.6529805660247803,
+      "learning_rate": 9.991887232532196e-07,
+      "loss": 0.9957,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05093807952602298,
+      "grad_norm": 1.5760482549667358,
+      "learning_rate": 9.990619612615353e-07,
+      "loss": 0.9742,
+      "step": 2115
+    },
+    {
+      "epoch": 0.05105850051781027,
+      "grad_norm": 1.7098240852355957,
+      "learning_rate": 9.98935199269851e-07,
+      "loss": 0.9826,
+      "step": 2120
+    },
+    {
+      "epoch": 0.051178921509597554,
+      "grad_norm": 1.4694714546203613,
+      "learning_rate": 9.988084372781663e-07,
+      "loss": 1.0453,
+      "step": 2125
+    },
+    {
+      "epoch": 0.05129934250138484,
+      "grad_norm": 1.719585657119751,
+      "learning_rate": 9.98681675286482e-07,
+      "loss": 1.0561,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05141976349317213,
+      "grad_norm": 1.5589990615844727,
+      "learning_rate": 9.985549132947976e-07,
+      "loss": 0.9995,
+      "step": 2135
+    },
+    {
+      "epoch": 0.051540184484959416,
+      "grad_norm": 1.570941686630249,
+      "learning_rate": 9.984281513031133e-07,
+      "loss": 1.0074,
+      "step": 2140
+    },
+    {
+      "epoch": 0.05166060547674671,
+      "grad_norm": 1.4263135194778442,
+      "learning_rate": 9.98301389311429e-07,
+      "loss": 0.9939,
+      "step": 2145
+    },
+    {
+      "epoch": 0.051781026468534,
+      "grad_norm": 1.6503669023513794,
+      "learning_rate": 9.981746273197443e-07,
+      "loss": 0.9384,
+      "step": 2150
+    },
+    {
+      "epoch": 0.051901447460321284,
+      "grad_norm": 1.5674190521240234,
+      "learning_rate": 9.9804786532806e-07,
+      "loss": 1.0988,
+      "step": 2155
+    },
+    {
+      "epoch": 0.05202186845210857,
+      "grad_norm": 1.8552207946777344,
+      "learning_rate": 9.979211033363756e-07,
+      "loss": 0.9471,
+      "step": 2160
+    },
+    {
+      "epoch": 0.05214228944389586,
+      "grad_norm": 1.5317758321762085,
+      "learning_rate": 9.977943413446913e-07,
+      "loss": 1.0026,
+      "step": 2165
+    },
+    {
+      "epoch": 0.052262710435683146,
+      "grad_norm": 1.5184342861175537,
+      "learning_rate": 9.976675793530067e-07,
+      "loss": 0.9657,
+      "step": 2170
+    },
+    {
+      "epoch": 0.052383131427470433,
+      "grad_norm": 2.0550878047943115,
+      "learning_rate": 9.975408173613224e-07,
+      "loss": 1.0551,
+      "step": 2175
+    },
+    {
+      "epoch": 0.05250355241925773,
+      "grad_norm": 1.670849084854126,
+      "learning_rate": 9.97414055369638e-07,
+      "loss": 1.0307,
+      "step": 2180
+    },
+    {
+      "epoch": 0.052623973411045015,
+      "grad_norm": 1.8367807865142822,
+      "learning_rate": 9.972872933779534e-07,
+      "loss": 1.0552,
+      "step": 2185
+    },
+    {
+      "epoch": 0.0527443944028323,
+      "grad_norm": 1.5710960626602173,
+      "learning_rate": 9.97160531386269e-07,
+      "loss": 1.026,
+      "step": 2190
+    },
+    {
+      "epoch": 0.05286481539461959,
+      "grad_norm": 1.4356889724731445,
+      "learning_rate": 9.970337693945847e-07,
+      "loss": 0.9868,
+      "step": 2195
+    },
+    {
+      "epoch": 0.05298523638640688,
+      "grad_norm": 1.4902724027633667,
+      "learning_rate": 9.969070074029002e-07,
+      "loss": 1.0166,
+      "step": 2200
+    },
+    {
+      "epoch": 0.053105657378194164,
+      "grad_norm": 1.6974903345108032,
+      "learning_rate": 9.967802454112158e-07,
+      "loss": 1.0271,
+      "step": 2205
+    },
+    {
+      "epoch": 0.05322607836998146,
+      "grad_norm": 1.4314632415771484,
+      "learning_rate": 9.966534834195315e-07,
+      "loss": 0.983,
+      "step": 2210
+    },
+    {
+      "epoch": 0.053346499361768746,
+      "grad_norm": 1.472601294517517,
+      "learning_rate": 9.96526721427847e-07,
+      "loss": 1.024,
+      "step": 2215
+    },
+    {
+      "epoch": 0.05346692035355603,
+      "grad_norm": 1.6418883800506592,
+      "learning_rate": 9.963999594361625e-07,
+      "loss": 1.0011,
+      "step": 2220
+    },
+    {
+      "epoch": 0.05358734134534332,
+      "grad_norm": 1.492514729499817,
+      "learning_rate": 9.962731974444782e-07,
+      "loss": 1.0287,
+      "step": 2225
+    },
+    {
+      "epoch": 0.05370776233713061,
+      "grad_norm": 1.6566858291625977,
+      "learning_rate": 9.961464354527938e-07,
+      "loss": 1.0115,
+      "step": 2230
+    },
+    {
+      "epoch": 0.053828183328917895,
+      "grad_norm": 2.257582187652588,
+      "learning_rate": 9.960196734611095e-07,
+      "loss": 1.0276,
+      "step": 2235
+    },
+    {
+      "epoch": 0.05394860432070519,
+      "grad_norm": 1.6120405197143555,
+      "learning_rate": 9.95892911469425e-07,
+      "loss": 1.0478,
+      "step": 2240
+    },
+    {
+      "epoch": 0.054069025312492476,
+      "grad_norm": 1.6881752014160156,
+      "learning_rate": 9.957661494777405e-07,
+      "loss": 1.0088,
+      "step": 2245
+    },
+    {
+      "epoch": 0.05418944630427976,
+      "grad_norm": 1.7539931535720825,
+      "learning_rate": 9.956393874860562e-07,
+      "loss": 1.0064,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05430986729606705,
+      "grad_norm": 1.7022444009780884,
+      "learning_rate": 9.955126254943718e-07,
+      "loss": 1.0547,
+      "step": 2255
+    },
+    {
+      "epoch": 0.05443028828785434,
+      "grad_norm": 1.4357643127441406,
+      "learning_rate": 9.953858635026873e-07,
+      "loss": 1.0532,
+      "step": 2260
+    },
+    {
+      "epoch": 0.054550709279641625,
+      "grad_norm": 1.4944684505462646,
+      "learning_rate": 9.95259101511003e-07,
+      "loss": 0.996,
+      "step": 2265
+    },
+    {
+      "epoch": 0.05467113027142891,
+      "grad_norm": 1.4375295639038086,
+      "learning_rate": 9.951323395193186e-07,
+      "loss": 1.0292,
+      "step": 2270
+    },
+    {
+      "epoch": 0.05479155126321621,
+      "grad_norm": 1.7429416179656982,
+      "learning_rate": 9.95005577527634e-07,
+      "loss": 0.9918,
+      "step": 2275
+    },
+    {
+      "epoch": 0.054911972255003494,
+      "grad_norm": 1.5450338125228882,
+      "learning_rate": 9.948788155359496e-07,
+      "loss": 1.0145,
+      "step": 2280
+    },
+    {
+      "epoch": 0.05503239324679078,
+      "grad_norm": 1.7032642364501953,
+      "learning_rate": 9.947520535442653e-07,
+      "loss": 1.0643,
+      "step": 2285
+    },
+    {
+      "epoch": 0.05515281423857807,
+      "grad_norm": 1.4787014722824097,
+      "learning_rate": 9.946252915525807e-07,
+      "loss": 0.972,
+      "step": 2290
+    },
+    {
+      "epoch": 0.055273235230365356,
+      "grad_norm": 1.6551034450531006,
+      "learning_rate": 9.944985295608964e-07,
+      "loss": 1.0394,
+      "step": 2295
+    },
+    {
+      "epoch": 0.05539365622215264,
+      "grad_norm": 1.5106158256530762,
+      "learning_rate": 9.94371767569212e-07,
+      "loss": 1.0251,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05551407721393994,
+      "grad_norm": 1.7415740489959717,
+      "learning_rate": 9.942450055775276e-07,
+      "loss": 1.0066,
+      "step": 2305
+    },
+    {
+      "epoch": 0.055634498205727224,
+      "grad_norm": 1.625821590423584,
+      "learning_rate": 9.94118243585843e-07,
+      "loss": 0.9502,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05575491919751451,
+      "grad_norm": 1.480774164199829,
+      "learning_rate": 9.939914815941587e-07,
+      "loss": 0.9622,
+      "step": 2315
+    },
+    {
+      "epoch": 0.0558753401893018,
+      "grad_norm": 1.4830437898635864,
+      "learning_rate": 9.938647196024744e-07,
+      "loss": 0.9566,
+      "step": 2320
+    },
+    {
+      "epoch": 0.055995761181089086,
+      "grad_norm": 1.5242736339569092,
+      "learning_rate": 9.9373795761079e-07,
+      "loss": 0.9981,
+      "step": 2325
+    },
+    {
+      "epoch": 0.056116182172876374,
+      "grad_norm": 1.4055622816085815,
+      "learning_rate": 9.936111956191057e-07,
+      "loss": 1.0229,
+      "step": 2330
+    },
+    {
+      "epoch": 0.05623660316466366,
+      "grad_norm": 2.0330963134765625,
+      "learning_rate": 9.93484433627421e-07,
+      "loss": 1.0014,
+      "step": 2335
+    },
+    {
+      "epoch": 0.056357024156450955,
+      "grad_norm": 1.824340581893921,
+      "learning_rate": 9.933576716357367e-07,
+      "loss": 0.9525,
+      "step": 2340
+    },
+    {
+      "epoch": 0.05647744514823824,
+      "grad_norm": 1.5875630378723145,
+      "learning_rate": 9.932309096440524e-07,
+      "loss": 0.9849,
+      "step": 2345
+    },
+    {
+      "epoch": 0.05659786614002553,
+      "grad_norm": 1.5911989212036133,
+      "learning_rate": 9.931041476523678e-07,
+      "loss": 1.0125,
+      "step": 2350
+    },
+    {
+      "epoch": 0.05671828713181282,
+      "grad_norm": 1.567030906677246,
+      "learning_rate": 9.929773856606835e-07,
+      "loss": 1.0127,
+      "step": 2355
+    },
+    {
+      "epoch": 0.056838708123600104,
+      "grad_norm": 1.8666136264801025,
+      "learning_rate": 9.92850623668999e-07,
+      "loss": 1.0164,
+      "step": 2360
+    },
+    {
+      "epoch": 0.05695912911538739,
+      "grad_norm": 1.4268726110458374,
+      "learning_rate": 9.927238616773145e-07,
+      "loss": 0.9955,
+      "step": 2365
+    },
+    {
+      "epoch": 0.057079550107174686,
+      "grad_norm": 1.6934856176376343,
+      "learning_rate": 9.925970996856302e-07,
+      "loss": 0.978,
+      "step": 2370
+    },
+    {
+      "epoch": 0.05719997109896197,
+      "grad_norm": 1.5638610124588013,
+      "learning_rate": 9.924703376939458e-07,
+      "loss": 1.0284,
+      "step": 2375
+    },
+    {
+      "epoch": 0.05732039209074926,
+      "grad_norm": 1.6345804929733276,
+      "learning_rate": 9.923435757022613e-07,
+      "loss": 1.0061,
+      "step": 2380
+    },
+    {
+      "epoch": 0.05744081308253655,
+      "grad_norm": 1.7583974599838257,
+      "learning_rate": 9.92216813710577e-07,
+      "loss": 1.0355,
+      "step": 2385
+    },
+    {
+      "epoch": 0.057561234074323835,
+      "grad_norm": 1.576433539390564,
+      "learning_rate": 9.920900517188925e-07,
+      "loss": 0.9556,
+      "step": 2390
+    },
+    {
+      "epoch": 0.05768165506611112,
+      "grad_norm": 1.4995986223220825,
+      "learning_rate": 9.919632897272082e-07,
+      "loss": 0.9899,
+      "step": 2395
+    },
+    {
+      "epoch": 0.057802076057898416,
+      "grad_norm": 1.8596760034561157,
+      "learning_rate": 9.918365277355238e-07,
+      "loss": 1.0413,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0579224970496857,
+      "grad_norm": 2.1781084537506104,
+      "learning_rate": 9.917097657438393e-07,
+      "loss": 1.0664,
+      "step": 2405
+    },
+    {
+      "epoch": 0.05804291804147299,
+      "grad_norm": 1.719193458557129,
+      "learning_rate": 9.91583003752155e-07,
+      "loss": 1.0398,
+      "step": 2410
+    },
+    {
+      "epoch": 0.05816333903326028,
+      "grad_norm": 1.5527453422546387,
+      "learning_rate": 9.914562417604706e-07,
+      "loss": 0.8899,
+      "step": 2415
+    },
+    {
+      "epoch": 0.058283760025047565,
+      "grad_norm": 2.198544979095459,
+      "learning_rate": 9.913294797687862e-07,
+      "loss": 1.0026,
+      "step": 2420
+    },
+    {
+      "epoch": 0.05840418101683485,
+      "grad_norm": 1.8652313947677612,
+      "learning_rate": 9.912027177771016e-07,
+      "loss": 1.0415,
+      "step": 2425
+    },
+    {
+      "epoch": 0.05852460200862214,
+      "grad_norm": 1.5211336612701416,
+      "learning_rate": 9.910759557854173e-07,
+      "loss": 0.9695,
+      "step": 2430
+    },
+    {
+      "epoch": 0.058645023000409434,
+      "grad_norm": 1.603398323059082,
+      "learning_rate": 9.90949193793733e-07,
+      "loss": 0.973,
+      "step": 2435
+    },
+    {
+      "epoch": 0.05876544399219672,
+      "grad_norm": 1.6345428228378296,
+      "learning_rate": 9.908224318020484e-07,
+      "loss": 0.9492,
+      "step": 2440
+    },
+    {
+      "epoch": 0.05888586498398401,
+      "grad_norm": 1.4910589456558228,
+      "learning_rate": 9.90695669810364e-07,
+      "loss": 0.9742,
+      "step": 2445
+    },
+    {
+      "epoch": 0.059006285975771296,
+      "grad_norm": 1.7999168634414673,
+      "learning_rate": 9.905689078186796e-07,
+      "loss": 0.9763,
+      "step": 2450
+    },
+    {
+      "epoch": 0.05912670696755858,
+      "grad_norm": 1.7445992231369019,
+      "learning_rate": 9.90442145826995e-07,
+      "loss": 1.0366,
+      "step": 2455
+    },
+    {
+      "epoch": 0.05924712795934587,
+      "grad_norm": 1.6648802757263184,
+      "learning_rate": 9.903153838353107e-07,
+      "loss": 1.0165,
+      "step": 2460
+    },
+    {
+      "epoch": 0.059367548951133164,
+      "grad_norm": 1.5104074478149414,
+      "learning_rate": 9.901886218436264e-07,
+      "loss": 0.9897,
+      "step": 2465
+    },
+    {
+      "epoch": 0.05948796994292045,
+      "grad_norm": 1.7375459671020508,
+      "learning_rate": 9.900618598519418e-07,
+      "loss": 1.0129,
+      "step": 2470
+    },
+    {
+      "epoch": 0.05960839093470774,
+      "grad_norm": 1.4074209928512573,
+      "learning_rate": 9.899350978602574e-07,
+      "loss": 1.0411,
+      "step": 2475
+    },
+    {
+      "epoch": 0.059728811926495026,
+      "grad_norm": 1.6192816495895386,
+      "learning_rate": 9.89808335868573e-07,
+      "loss": 0.995,
+      "step": 2480
+    },
+    {
+      "epoch": 0.059849232918282314,
+      "grad_norm": 1.6760953664779663,
+      "learning_rate": 9.896815738768887e-07,
+      "loss": 1.0046,
+      "step": 2485
+    },
+    {
+      "epoch": 0.0599696539100696,
+      "grad_norm": 1.579380989074707,
+      "learning_rate": 9.895548118852044e-07,
+      "loss": 0.9926,
+      "step": 2490
+    },
+    {
+      "epoch": 0.060090074901856895,
+      "grad_norm": 1.7261121273040771,
+      "learning_rate": 9.894280498935198e-07,
+      "loss": 0.9732,
+      "step": 2495
+    },
+    {
+      "epoch": 0.06021049589364418,
+      "grad_norm": 1.5446956157684326,
+      "learning_rate": 9.893012879018355e-07,
+      "loss": 1.0288,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06033091688543147,
+      "grad_norm": 1.4820080995559692,
+      "learning_rate": 9.89174525910151e-07,
+      "loss": 1.0583,
+      "step": 2505
+    },
+    {
+      "epoch": 0.06045133787721876,
+      "grad_norm": 1.5800060033798218,
+      "learning_rate": 9.890477639184667e-07,
+      "loss": 1.0293,
+      "step": 2510
+    },
+    {
+      "epoch": 0.060571758869006044,
+      "grad_norm": 1.6209781169891357,
+      "learning_rate": 9.889210019267822e-07,
+      "loss": 1.0297,
+      "step": 2515
+    },
+    {
+      "epoch": 0.06069217986079333,
+      "grad_norm": 1.3738642930984497,
+      "learning_rate": 9.887942399350978e-07,
+      "loss": 1.0049,
+      "step": 2520
+    },
+    {
+      "epoch": 0.06081260085258062,
+      "grad_norm": 1.6053427457809448,
+      "learning_rate": 9.886674779434135e-07,
+      "loss": 0.9941,
+      "step": 2525
+    },
+    {
+      "epoch": 0.06093302184436791,
+      "grad_norm": 1.4712995290756226,
+      "learning_rate": 9.88540715951729e-07,
+      "loss": 0.9638,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0610534428361552,
+      "grad_norm": 1.7111667394638062,
+      "learning_rate": 9.884139539600445e-07,
+      "loss": 1.0174,
+      "step": 2535
+    },
+    {
+      "epoch": 0.06117386382794249,
+      "grad_norm": 1.4965132474899292,
+      "learning_rate": 9.882871919683602e-07,
+      "loss": 0.9886,
+      "step": 2540
+    },
+    {
+      "epoch": 0.061294284819729775,
+      "grad_norm": 1.615824580192566,
+      "learning_rate": 9.881604299766756e-07,
+      "loss": 0.9891,
+      "step": 2545
+    },
+    {
+      "epoch": 0.06141470581151706,
+      "grad_norm": 1.4951409101486206,
+      "learning_rate": 9.880336679849913e-07,
+      "loss": 1.0106,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06153512680330435,
+      "grad_norm": 1.630304217338562,
+      "learning_rate": 9.87906905993307e-07,
+      "loss": 0.937,
+      "step": 2555
+    },
+    {
+      "epoch": 0.06165554779509164,
+      "grad_norm": 1.622524380683899,
+      "learning_rate": 9.877801440016226e-07,
+      "loss": 0.9855,
+      "step": 2560
+    },
+    {
+      "epoch": 0.06177596878687893,
+      "grad_norm": 1.665725588798523,
+      "learning_rate": 9.87653382009938e-07,
+      "loss": 1.0499,
+      "step": 2565
+    },
+    {
+      "epoch": 0.06189638977866622,
+      "grad_norm": 1.747391700744629,
+      "learning_rate": 9.875266200182536e-07,
+      "loss": 0.9747,
+      "step": 2570
+    },
+    {
+      "epoch": 0.062016810770453505,
+      "grad_norm": 1.7850977182388306,
+      "learning_rate": 9.873998580265693e-07,
+      "loss": 0.9773,
+      "step": 2575
+    },
+    {
+      "epoch": 0.06213723176224079,
+      "grad_norm": 1.7369146347045898,
+      "learning_rate": 9.87273096034885e-07,
+      "loss": 1.0316,
+      "step": 2580
+    },
+    {
+      "epoch": 0.06225765275402808,
+      "grad_norm": 1.810365915298462,
+      "learning_rate": 9.871463340432006e-07,
+      "loss": 0.9694,
+      "step": 2585
+    },
+    {
+      "epoch": 0.062378073745815374,
+      "grad_norm": 1.6777886152267456,
+      "learning_rate": 9.87019572051516e-07,
+      "loss": 0.9939,
+      "step": 2590
+    },
+    {
+      "epoch": 0.06249849473760266,
+      "grad_norm": 1.7758405208587646,
+      "learning_rate": 9.868928100598316e-07,
+      "loss": 1.0482,
+      "step": 2595
+    },
+    {
+      "epoch": 0.06261891572938995,
+      "grad_norm": 1.699005365371704,
+      "learning_rate": 9.867660480681473e-07,
+      "loss": 1.0163,
+      "step": 2600
+    },
+    {
+      "epoch": 0.06273933672117724,
+      "grad_norm": 1.5089378356933594,
+      "learning_rate": 9.866392860764627e-07,
+      "loss": 0.9912,
+      "step": 2605
+    },
+    {
+      "epoch": 0.06285975771296452,
+      "grad_norm": 1.4108537435531616,
+      "learning_rate": 9.865125240847784e-07,
+      "loss": 1.0056,
+      "step": 2610
+    },
+    {
+      "epoch": 0.06298017870475181,
+      "grad_norm": 1.6560865640640259,
+      "learning_rate": 9.86385762093094e-07,
+      "loss": 1.024,
+      "step": 2615
+    },
+    {
+      "epoch": 0.0631005996965391,
+      "grad_norm": 1.54850172996521,
+      "learning_rate": 9.862590001014094e-07,
+      "loss": 1.0085,
+      "step": 2620
+    },
+    {
+      "epoch": 0.06322102068832638,
+      "grad_norm": 1.6445883512496948,
+      "learning_rate": 9.86132238109725e-07,
+      "loss": 1.0119,
+      "step": 2625
+    },
+    {
+      "epoch": 0.06334144168011367,
+      "grad_norm": 1.811767816543579,
+      "learning_rate": 9.860054761180407e-07,
+      "loss": 0.9318,
+      "step": 2630
+    },
+    {
+      "epoch": 0.06346186267190096,
+      "grad_norm": 1.6554980278015137,
+      "learning_rate": 9.858787141263562e-07,
+      "loss": 1.0083,
+      "step": 2635
+    },
+    {
+      "epoch": 0.06358228366368826,
+      "grad_norm": 1.773821473121643,
+      "learning_rate": 9.857519521346718e-07,
+      "loss": 1.0159,
+      "step": 2640
+    },
+    {
+      "epoch": 0.06370270465547555,
+      "grad_norm": 1.6003626585006714,
+      "learning_rate": 9.856251901429875e-07,
+      "loss": 0.9948,
+      "step": 2645
+    },
+    {
+      "epoch": 0.06382312564726283,
+      "grad_norm": 1.6465872526168823,
+      "learning_rate": 9.85498428151303e-07,
+      "loss": 1.0038,
+      "step": 2650
+    },
+    {
+      "epoch": 0.06394354663905012,
+      "grad_norm": 1.4264850616455078,
+      "learning_rate": 9.853716661596185e-07,
+      "loss": 1.0058,
+      "step": 2655
+    },
+    {
+      "epoch": 0.06406396763083741,
+      "grad_norm": 1.4336827993392944,
+      "learning_rate": 9.852449041679342e-07,
+      "loss": 1.0192,
+      "step": 2660
+    },
+    {
+      "epoch": 0.0641843886226247,
+      "grad_norm": 1.6807986497879028,
+      "learning_rate": 9.851181421762498e-07,
+      "loss": 0.9417,
+      "step": 2665
+    },
+    {
+      "epoch": 0.06430480961441198,
+      "grad_norm": 1.5016063451766968,
+      "learning_rate": 9.849913801845655e-07,
+      "loss": 0.9898,
+      "step": 2670
+    },
+    {
+      "epoch": 0.06442523060619927,
+      "grad_norm": 1.5583468675613403,
+      "learning_rate": 9.848646181928811e-07,
+      "loss": 1.0263,
+      "step": 2675
+    },
+    {
+      "epoch": 0.06454565159798656,
+      "grad_norm": 1.9095042943954468,
+      "learning_rate": 9.847378562011966e-07,
+      "loss": 0.9524,
+      "step": 2680
+    },
+    {
+      "epoch": 0.06466607258977385,
+      "grad_norm": 1.6604355573654175,
+      "learning_rate": 9.846110942095122e-07,
+      "loss": 0.9932,
+      "step": 2685
+    },
+    {
+      "epoch": 0.06478649358156113,
+      "grad_norm": 1.5880812406539917,
+      "learning_rate": 9.844843322178278e-07,
+      "loss": 0.9439,
+      "step": 2690
+    },
+    {
+      "epoch": 0.06490691457334842,
+      "grad_norm": 1.512092113494873,
+      "learning_rate": 9.843575702261433e-07,
+      "loss": 0.9756,
+      "step": 2695
+    },
+    {
+      "epoch": 0.06502733556513571,
+      "grad_norm": 1.510452151298523,
+      "learning_rate": 9.84230808234459e-07,
+      "loss": 0.9915,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06514775655692301,
+      "grad_norm": 1.5895094871520996,
+      "learning_rate": 9.841040462427746e-07,
+      "loss": 1.0063,
+      "step": 2705
+    },
+    {
+      "epoch": 0.0652681775487103,
+      "grad_norm": 1.780158519744873,
+      "learning_rate": 9.8397728425109e-07,
+      "loss": 1.0395,
+      "step": 2710
+    },
+    {
+      "epoch": 0.06538859854049758,
+      "grad_norm": 1.534387469291687,
+      "learning_rate": 9.838505222594056e-07,
+      "loss": 0.9337,
+      "step": 2715
+    },
+    {
+      "epoch": 0.06550901953228487,
+      "grad_norm": 1.6160880327224731,
+      "learning_rate": 9.837237602677213e-07,
+      "loss": 0.9846,
+      "step": 2720
+    },
+    {
+      "epoch": 0.06562944052407216,
+      "grad_norm": 1.616960883140564,
+      "learning_rate": 9.835969982760367e-07,
+      "loss": 1.0247,
+      "step": 2725
+    },
+    {
+      "epoch": 0.06574986151585945,
+      "grad_norm": 1.6236367225646973,
+      "learning_rate": 9.834702362843524e-07,
+      "loss": 0.9983,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06587028250764673,
+      "grad_norm": 2.1704113483428955,
+      "learning_rate": 9.83343474292668e-07,
+      "loss": 1.0474,
+      "step": 2735
+    },
+    {
+      "epoch": 0.06599070349943402,
+      "grad_norm": 1.4857268333435059,
+      "learning_rate": 9.832167123009837e-07,
+      "loss": 0.9496,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06611112449122131,
+      "grad_norm": 1.6319400072097778,
+      "learning_rate": 9.830899503092993e-07,
+      "loss": 0.972,
+      "step": 2745
+    },
+    {
+      "epoch": 0.0662315454830086,
+      "grad_norm": 1.406339168548584,
+      "learning_rate": 9.829631883176147e-07,
+      "loss": 0.9842,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06635196647479588,
+      "grad_norm": 1.6158806085586548,
+      "learning_rate": 9.828364263259304e-07,
+      "loss": 0.9952,
+      "step": 2755
+    },
+    {
+      "epoch": 0.06647238746658317,
+      "grad_norm": 1.6224943399429321,
+      "learning_rate": 9.82709664334246e-07,
+      "loss": 0.961,
+      "step": 2760
+    },
+    {
+      "epoch": 0.06659280845837047,
+      "grad_norm": 1.630590558052063,
+      "learning_rate": 9.825829023425617e-07,
+      "loss": 1.0236,
+      "step": 2765
+    },
+    {
+      "epoch": 0.06671322945015776,
+      "grad_norm": 1.5676782131195068,
+      "learning_rate": 9.82456140350877e-07,
+      "loss": 0.9279,
+      "step": 2770
+    },
+    {
+      "epoch": 0.06683365044194504,
+      "grad_norm": 1.5544155836105347,
+      "learning_rate": 9.823293783591927e-07,
+      "loss": 0.9843,
+      "step": 2775
+    },
+    {
+      "epoch": 0.06695407143373233,
+      "grad_norm": 1.5769226551055908,
+      "learning_rate": 9.822026163675084e-07,
+      "loss": 1.0193,
+      "step": 2780
+    },
+    {
+      "epoch": 0.06707449242551962,
+      "grad_norm": 1.502962350845337,
+      "learning_rate": 9.820758543758238e-07,
+      "loss": 1.0003,
+      "step": 2785
+    },
+    {
+      "epoch": 0.0671949134173069,
+      "grad_norm": 1.9905481338500977,
+      "learning_rate": 9.819490923841395e-07,
+      "loss": 0.9875,
+      "step": 2790
+    },
+    {
+      "epoch": 0.0673153344090942,
+      "grad_norm": 1.6774263381958008,
+      "learning_rate": 9.818223303924551e-07,
+      "loss": 1.0727,
+      "step": 2795
+    },
+    {
+      "epoch": 0.06743575540088148,
+      "grad_norm": 1.4511786699295044,
+      "learning_rate": 9.816955684007708e-07,
+      "loss": 0.9355,
+      "step": 2800
+    },
+    {
+      "epoch": 0.06755617639266877,
+      "grad_norm": 1.5723068714141846,
+      "learning_rate": 9.815688064090862e-07,
+      "loss": 0.9647,
+      "step": 2805
+    },
+    {
+      "epoch": 0.06767659738445606,
+      "grad_norm": 1.5582698583602905,
+      "learning_rate": 9.814420444174018e-07,
+      "loss": 0.9908,
+      "step": 2810
+    },
+    {
+      "epoch": 0.06779701837624334,
+      "grad_norm": 1.5775694847106934,
+      "learning_rate": 9.813152824257175e-07,
+      "loss": 0.9866,
+      "step": 2815
+    },
+    {
+      "epoch": 0.06791743936803063,
+      "grad_norm": 1.4566597938537598,
+      "learning_rate": 9.81188520434033e-07,
+      "loss": 1.0293,
+      "step": 2820
+    },
+    {
+      "epoch": 0.06803786035981792,
+      "grad_norm": 1.9800313711166382,
+      "learning_rate": 9.810617584423486e-07,
+      "loss": 0.8688,
+      "step": 2825
+    },
+    {
+      "epoch": 0.06815828135160522,
+      "grad_norm": 1.5676807165145874,
+      "learning_rate": 9.809349964506642e-07,
+      "loss": 1.027,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0682787023433925,
+      "grad_norm": 1.5596281290054321,
+      "learning_rate": 9.808082344589798e-07,
+      "loss": 0.9814,
+      "step": 2835
+    },
+    {
+      "epoch": 0.06839912333517979,
+      "grad_norm": 1.5297009944915771,
+      "learning_rate": 9.806814724672953e-07,
+      "loss": 0.9466,
+      "step": 2840
+    },
+    {
+      "epoch": 0.06851954432696708,
+      "grad_norm": 1.5337249040603638,
+      "learning_rate": 9.80554710475611e-07,
+      "loss": 0.9952,
+      "step": 2845
+    },
+    {
+      "epoch": 0.06863996531875437,
+      "grad_norm": 1.7600308656692505,
+      "learning_rate": 9.804279484839266e-07,
+      "loss": 0.9941,
+      "step": 2850
+    },
+    {
+      "epoch": 0.06876038631054165,
+      "grad_norm": 1.5565025806427002,
+      "learning_rate": 9.803011864922422e-07,
+      "loss": 0.9536,
+      "step": 2855
+    },
+    {
+      "epoch": 0.06888080730232894,
+      "grad_norm": 1.4753719568252563,
+      "learning_rate": 9.801744245005579e-07,
+      "loss": 0.9562,
+      "step": 2860
+    },
+    {
+      "epoch": 0.06900122829411623,
+      "grad_norm": 1.6010030508041382,
+      "learning_rate": 9.800476625088733e-07,
+      "loss": 1.0187,
+      "step": 2865
+    },
+    {
+      "epoch": 0.06912164928590352,
+      "grad_norm": 1.723764181137085,
+      "learning_rate": 9.79920900517189e-07,
+      "loss": 0.9999,
+      "step": 2870
+    },
+    {
+      "epoch": 0.0692420702776908,
+      "grad_norm": 1.480420470237732,
+      "learning_rate": 9.797941385255046e-07,
+      "loss": 1.0608,
+      "step": 2875
+    },
+    {
+      "epoch": 0.06936249126947809,
+      "grad_norm": 1.699066400527954,
+      "learning_rate": 9.7966737653382e-07,
+      "loss": 1.0086,
+      "step": 2880
+    },
+    {
+      "epoch": 0.06948291226126538,
+      "grad_norm": 1.6982381343841553,
+      "learning_rate": 9.795406145421357e-07,
+      "loss": 0.9616,
+      "step": 2885
+    },
+    {
+      "epoch": 0.06960333325305267,
+      "grad_norm": 1.5258110761642456,
+      "learning_rate": 9.794138525504513e-07,
+      "loss": 0.9569,
+      "step": 2890
+    },
+    {
+      "epoch": 0.06972375424483997,
+      "grad_norm": 1.3778504133224487,
+      "learning_rate": 9.792870905587667e-07,
+      "loss": 0.949,
+      "step": 2895
+    },
+    {
+      "epoch": 0.06984417523662725,
+      "grad_norm": 1.5082603693008423,
+      "learning_rate": 9.791603285670824e-07,
+      "loss": 0.9212,
+      "step": 2900
+    },
+    {
+      "epoch": 0.06996459622841454,
+      "grad_norm": 1.683454155921936,
+      "learning_rate": 9.79033566575398e-07,
+      "loss": 1.0048,
+      "step": 2905
+    },
+    {
+      "epoch": 0.07008501722020183,
+      "grad_norm": 1.6343798637390137,
+      "learning_rate": 9.789068045837135e-07,
+      "loss": 0.9412,
+      "step": 2910
+    },
+    {
+      "epoch": 0.07020543821198912,
+      "grad_norm": 2.1352267265319824,
+      "learning_rate": 9.78780042592029e-07,
+      "loss": 1.0182,
+      "step": 2915
+    },
+    {
+      "epoch": 0.0703258592037764,
+      "grad_norm": 1.613693118095398,
+      "learning_rate": 9.786532806003447e-07,
+      "loss": 0.9761,
+      "step": 2920
+    },
+    {
+      "epoch": 0.07044628019556369,
+      "grad_norm": 1.7342922687530518,
+      "learning_rate": 9.785265186086604e-07,
+      "loss": 0.9448,
+      "step": 2925
+    },
+    {
+      "epoch": 0.07056670118735098,
+      "grad_norm": 1.664732575416565,
+      "learning_rate": 9.78399756616976e-07,
+      "loss": 1.026,
+      "step": 2930
+    },
+    {
+      "epoch": 0.07068712217913826,
+      "grad_norm": 1.4946379661560059,
+      "learning_rate": 9.782729946252915e-07,
+      "loss": 0.935,
+      "step": 2935
+    },
+    {
+      "epoch": 0.07080754317092555,
+      "grad_norm": 1.7126940488815308,
+      "learning_rate": 9.781462326336071e-07,
+      "loss": 0.9844,
+      "step": 2940
+    },
+    {
+      "epoch": 0.07092796416271284,
+      "grad_norm": 1.847761869430542,
+      "learning_rate": 9.780194706419228e-07,
+      "loss": 1.0524,
+      "step": 2945
+    },
+    {
+      "epoch": 0.07104838515450013,
+      "grad_norm": 1.6522741317749023,
+      "learning_rate": 9.778927086502384e-07,
+      "loss": 0.9787,
+      "step": 2950
+    },
+    {
+      "epoch": 0.07116880614628743,
+      "grad_norm": 1.5920339822769165,
+      "learning_rate": 9.777659466585538e-07,
+      "loss": 0.9595,
+      "step": 2955
+    },
+    {
+      "epoch": 0.07128922713807472,
+      "grad_norm": 1.5285476446151733,
+      "learning_rate": 9.776391846668695e-07,
+      "loss": 0.9854,
+      "step": 2960
+    },
+    {
+      "epoch": 0.071409648129862,
+      "grad_norm": 1.3394943475723267,
+      "learning_rate": 9.775124226751851e-07,
+      "loss": 0.956,
+      "step": 2965
+    },
+    {
+      "epoch": 0.07153006912164929,
+      "grad_norm": 1.6529779434204102,
+      "learning_rate": 9.773856606835006e-07,
+      "loss": 1.0041,
+      "step": 2970
+    },
+    {
+      "epoch": 0.07165049011343658,
+      "grad_norm": 1.643385887145996,
+      "learning_rate": 9.772588986918162e-07,
+      "loss": 1.0442,
+      "step": 2975
+    },
+    {
+      "epoch": 0.07177091110522386,
+      "grad_norm": 1.5586971044540405,
+      "learning_rate": 9.771321367001318e-07,
+      "loss": 0.9751,
+      "step": 2980
+    },
+    {
+      "epoch": 0.07189133209701115,
+      "grad_norm": 1.46442711353302,
+      "learning_rate": 9.770053747084473e-07,
+      "loss": 1.0208,
+      "step": 2985
+    },
+    {
+      "epoch": 0.07201175308879844,
+      "grad_norm": 1.5164238214492798,
+      "learning_rate": 9.76878612716763e-07,
+      "loss": 1.0347,
+      "step": 2990
+    },
+    {
+      "epoch": 0.07213217408058573,
+      "grad_norm": 1.6594327688217163,
+      "learning_rate": 9.767518507250786e-07,
+      "loss": 1.0027,
+      "step": 2995
+    },
+    {
+      "epoch": 0.07225259507237301,
+      "grad_norm": 1.7102785110473633,
+      "learning_rate": 9.766250887333942e-07,
+      "loss": 1.0035,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0723730160641603,
+      "grad_norm": 1.5333776473999023,
+      "learning_rate": 9.764983267417096e-07,
+      "loss": 0.9626,
+      "step": 3005
+    },
+    {
+      "epoch": 0.07249343705594759,
+      "grad_norm": 2.1526007652282715,
+      "learning_rate": 9.763715647500253e-07,
+      "loss": 0.9312,
+      "step": 3010
+    },
+    {
+      "epoch": 0.07261385804773487,
+      "grad_norm": 1.6529592275619507,
+      "learning_rate": 9.76244802758341e-07,
+      "loss": 0.9714,
+      "step": 3015
+    },
+    {
+      "epoch": 0.07273427903952218,
+      "grad_norm": 1.569183349609375,
+      "learning_rate": 9.761180407666566e-07,
+      "loss": 0.9682,
+      "step": 3020
+    },
+    {
+      "epoch": 0.07285470003130946,
+      "grad_norm": 1.4431354999542236,
+      "learning_rate": 9.75991278774972e-07,
+      "loss": 1.0084,
+      "step": 3025
+    },
+    {
+      "epoch": 0.07297512102309675,
+      "grad_norm": 1.7939332723617554,
+      "learning_rate": 9.758645167832877e-07,
+      "loss": 0.9695,
+      "step": 3030
+    },
+    {
+      "epoch": 0.07309554201488404,
+      "grad_norm": 1.7346773147583008,
+      "learning_rate": 9.757377547916033e-07,
+      "loss": 0.9873,
+      "step": 3035
+    },
+    {
+      "epoch": 0.07321596300667133,
+      "grad_norm": 1.540472388267517,
+      "learning_rate": 9.75610992799919e-07,
+      "loss": 0.9653,
+      "step": 3040
+    },
+    {
+      "epoch": 0.07333638399845861,
+      "grad_norm": 1.637434720993042,
+      "learning_rate": 9.754842308082344e-07,
+      "loss": 0.9974,
+      "step": 3045
+    },
+    {
+      "epoch": 0.0734568049902459,
+      "grad_norm": 1.5349056720733643,
+      "learning_rate": 9.7535746881655e-07,
+      "loss": 0.9389,
+      "step": 3050
+    },
+    {
+      "epoch": 0.07357722598203319,
+      "grad_norm": 1.4215797185897827,
+      "learning_rate": 9.752307068248657e-07,
+      "loss": 0.9786,
+      "step": 3055
+    },
+    {
+      "epoch": 0.07369764697382047,
+      "grad_norm": 1.6199102401733398,
+      "learning_rate": 9.75103944833181e-07,
+      "loss": 0.9213,
+      "step": 3060
+    },
+    {
+      "epoch": 0.07381806796560776,
+      "grad_norm": 1.6182783842086792,
+      "learning_rate": 9.749771828414967e-07,
+      "loss": 0.9973,
+      "step": 3065
+    },
+    {
+      "epoch": 0.07393848895739505,
+      "grad_norm": 1.4855315685272217,
+      "learning_rate": 9.748504208498124e-07,
+      "loss": 0.9413,
+      "step": 3070
+    },
+    {
+      "epoch": 0.07405890994918234,
+      "grad_norm": 1.6804499626159668,
+      "learning_rate": 9.747236588581278e-07,
+      "loss": 0.9306,
+      "step": 3075
+    },
+    {
+      "epoch": 0.07417933094096962,
+      "grad_norm": 1.6445069313049316,
+      "learning_rate": 9.745968968664435e-07,
+      "loss": 0.9883,
+      "step": 3080
+    },
+    {
+      "epoch": 0.07429975193275692,
+      "grad_norm": 1.3424928188323975,
+      "learning_rate": 9.744701348747591e-07,
+      "loss": 0.9506,
+      "step": 3085
+    },
+    {
+      "epoch": 0.07442017292454421,
+      "grad_norm": 1.3568273782730103,
+      "learning_rate": 9.743433728830748e-07,
+      "loss": 1.0278,
+      "step": 3090
+    },
+    {
+      "epoch": 0.0745405939163315,
+      "grad_norm": 1.8795121908187866,
+      "learning_rate": 9.742166108913902e-07,
+      "loss": 0.9812,
+      "step": 3095
+    },
+    {
+      "epoch": 0.07466101490811879,
+      "grad_norm": 1.6932713985443115,
+      "learning_rate": 9.740898488997058e-07,
+      "loss": 0.9878,
+      "step": 3100
+    },
+    {
+      "epoch": 0.07478143589990607,
+      "grad_norm": 1.490128517150879,
+      "learning_rate": 9.739630869080215e-07,
+      "loss": 0.927,
+      "step": 3105
+    },
+    {
+      "epoch": 0.07490185689169336,
+      "grad_norm": 1.718519687652588,
+      "learning_rate": 9.738363249163371e-07,
+      "loss": 1.0131,
+      "step": 3110
+    },
+    {
+      "epoch": 0.07502227788348065,
+      "grad_norm": 1.6648001670837402,
+      "learning_rate": 9.737095629246528e-07,
+      "loss": 0.9911,
+      "step": 3115
+    },
+    {
+      "epoch": 0.07514269887526794,
+      "grad_norm": 1.541821837425232,
+      "learning_rate": 9.735828009329682e-07,
+      "loss": 1.0154,
+      "step": 3120
+    },
+    {
+      "epoch": 0.07526311986705522,
+      "grad_norm": 1.4682984352111816,
+      "learning_rate": 9.734560389412838e-07,
+      "loss": 0.981,
+      "step": 3125
+    },
+    {
+      "epoch": 0.07538354085884251,
+      "grad_norm": 1.4823180437088013,
+      "learning_rate": 9.733292769495995e-07,
+      "loss": 1.0265,
+      "step": 3130
+    },
+    {
+      "epoch": 0.0755039618506298,
+      "grad_norm": 1.359682559967041,
+      "learning_rate": 9.73202514957915e-07,
+      "loss": 0.9137,
+      "step": 3135
+    },
+    {
+      "epoch": 0.07562438284241708,
+      "grad_norm": 1.5946718454360962,
+      "learning_rate": 9.730757529662306e-07,
+      "loss": 0.9695,
+      "step": 3140
+    },
+    {
+      "epoch": 0.07574480383420437,
+      "grad_norm": 1.5661665201187134,
+      "learning_rate": 9.729489909745462e-07,
+      "loss": 0.9655,
+      "step": 3145
+    },
+    {
+      "epoch": 0.07586522482599167,
+      "grad_norm": 1.5131149291992188,
+      "learning_rate": 9.728222289828616e-07,
+      "loss": 0.9829,
+      "step": 3150
+    },
+    {
+      "epoch": 0.07598564581777896,
+      "grad_norm": 1.6737961769104004,
+      "learning_rate": 9.726954669911773e-07,
+      "loss": 0.9591,
+      "step": 3155
+    },
+    {
+      "epoch": 0.07610606680956625,
+      "grad_norm": 1.9588265419006348,
+      "learning_rate": 9.72568704999493e-07,
+      "loss": 0.9509,
+      "step": 3160
+    },
+    {
+      "epoch": 0.07622648780135353,
+      "grad_norm": 1.6424431800842285,
+      "learning_rate": 9.724419430078084e-07,
+      "loss": 0.9529,
+      "step": 3165
+    },
+    {
+      "epoch": 0.07634690879314082,
+      "grad_norm": 1.5378398895263672,
+      "learning_rate": 9.72315181016124e-07,
+      "loss": 1.002,
+      "step": 3170
+    },
+    {
+      "epoch": 0.07646732978492811,
+      "grad_norm": 1.6808853149414062,
+      "learning_rate": 9.721884190244397e-07,
+      "loss": 1.004,
+      "step": 3175
+    },
+    {
+      "epoch": 0.0765877507767154,
+      "grad_norm": 1.6004494428634644,
+      "learning_rate": 9.720616570327553e-07,
+      "loss": 1.0206,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07670817176850268,
+      "grad_norm": 1.5458874702453613,
+      "learning_rate": 9.71934895041071e-07,
+      "loss": 0.972,
+      "step": 3185
+    },
+    {
+      "epoch": 0.07682859276028997,
+      "grad_norm": 1.4363664388656616,
+      "learning_rate": 9.718081330493864e-07,
+      "loss": 1.0765,
+      "step": 3190
+    },
+    {
+      "epoch": 0.07694901375207726,
+      "grad_norm": 1.6539244651794434,
+      "learning_rate": 9.71681371057702e-07,
+      "loss": 1.0203,
+      "step": 3195
+    },
+    {
+      "epoch": 0.07706943474386455,
+      "grad_norm": 1.4915926456451416,
+      "learning_rate": 9.715546090660177e-07,
+      "loss": 0.9945,
+      "step": 3200
+    },
+    {
+      "epoch": 0.07718985573565183,
+      "grad_norm": 1.519199013710022,
+      "learning_rate": 9.714278470743333e-07,
+      "loss": 0.9774,
+      "step": 3205
+    },
+    {
+      "epoch": 0.07731027672743913,
+      "grad_norm": 1.5417412519454956,
+      "learning_rate": 9.713010850826488e-07,
+      "loss": 0.9993,
+      "step": 3210
+    },
+    {
+      "epoch": 0.07743069771922642,
+      "grad_norm": 1.4990864992141724,
+      "learning_rate": 9.711743230909644e-07,
+      "loss": 0.9397,
+      "step": 3215
+    },
+    {
+      "epoch": 0.07755111871101371,
+      "grad_norm": 2.056589126586914,
+      "learning_rate": 9.7104756109928e-07,
+      "loss": 1.0334,
+      "step": 3220
+    },
+    {
+      "epoch": 0.077671539702801,
+      "grad_norm": 1.4858630895614624,
+      "learning_rate": 9.709207991075955e-07,
+      "loss": 0.9464,
+      "step": 3225
+    },
+    {
+      "epoch": 0.07779196069458828,
+      "grad_norm": 1.5432004928588867,
+      "learning_rate": 9.707940371159111e-07,
+      "loss": 0.9859,
+      "step": 3230
+    },
+    {
+      "epoch": 0.07791238168637557,
+      "grad_norm": 1.7187660932540894,
+      "learning_rate": 9.706672751242268e-07,
+      "loss": 0.995,
+      "step": 3235
+    },
+    {
+      "epoch": 0.07803280267816286,
+      "grad_norm": 1.5984174013137817,
+      "learning_rate": 9.705405131325422e-07,
+      "loss": 1.0028,
+      "step": 3240
+    },
+    {
+      "epoch": 0.07815322366995014,
+      "grad_norm": 1.5194472074508667,
+      "learning_rate": 9.704137511408578e-07,
+      "loss": 1.0039,
+      "step": 3245
+    },
+    {
+      "epoch": 0.07827364466173743,
+      "grad_norm": 1.6343015432357788,
+      "learning_rate": 9.702869891491735e-07,
+      "loss": 1.0423,
+      "step": 3250
+    },
+    {
+      "epoch": 0.07839406565352472,
+      "grad_norm": 1.5536481142044067,
+      "learning_rate": 9.70160227157489e-07,
+      "loss": 0.9829,
+      "step": 3255
+    },
+    {
+      "epoch": 0.078514486645312,
+      "grad_norm": 1.6728906631469727,
+      "learning_rate": 9.700334651658046e-07,
+      "loss": 0.9838,
+      "step": 3260
+    },
+    {
+      "epoch": 0.0786349076370993,
+      "grad_norm": 1.785743236541748,
+      "learning_rate": 9.699067031741202e-07,
+      "loss": 1.0017,
+      "step": 3265
+    },
+    {
+      "epoch": 0.07875532862888658,
+      "grad_norm": 1.5994573831558228,
+      "learning_rate": 9.697799411824359e-07,
+      "loss": 1.0219,
+      "step": 3270
+    },
+    {
+      "epoch": 0.07887574962067388,
+      "grad_norm": 1.6276570558547974,
+      "learning_rate": 9.696531791907515e-07,
+      "loss": 0.9886,
+      "step": 3275
+    },
+    {
+      "epoch": 0.07899617061246117,
+      "grad_norm": 1.6569063663482666,
+      "learning_rate": 9.69526417199067e-07,
+      "loss": 0.9907,
+      "step": 3280
+    },
+    {
+      "epoch": 0.07911659160424846,
+      "grad_norm": 1.6754655838012695,
+      "learning_rate": 9.693996552073826e-07,
+      "loss": 0.9841,
+      "step": 3285
+    },
+    {
+      "epoch": 0.07923701259603574,
+      "grad_norm": 1.7406693696975708,
+      "learning_rate": 9.692728932156982e-07,
+      "loss": 1.0043,
+      "step": 3290
+    },
+    {
+      "epoch": 0.07935743358782303,
+      "grad_norm": 1.4916058778762817,
+      "learning_rate": 9.691461312240139e-07,
+      "loss": 0.9928,
+      "step": 3295
+    },
+    {
+      "epoch": 0.07947785457961032,
+      "grad_norm": 1.5197107791900635,
+      "learning_rate": 9.690193692323293e-07,
+      "loss": 0.9679,
+      "step": 3300
+    },
+    {
+      "epoch": 0.0795982755713976,
+      "grad_norm": 1.7245734930038452,
+      "learning_rate": 9.68892607240645e-07,
+      "loss": 0.9643,
+      "step": 3305
+    },
+    {
+      "epoch": 0.0797186965631849,
+      "grad_norm": 1.7159137725830078,
+      "learning_rate": 9.687658452489606e-07,
+      "loss": 0.9778,
+      "step": 3310
+    },
+    {
+      "epoch": 0.07983911755497218,
+      "grad_norm": 1.564124345779419,
+      "learning_rate": 9.68639083257276e-07,
+      "loss": 0.9566,
+      "step": 3315
+    },
+    {
+      "epoch": 0.07995953854675947,
+      "grad_norm": 1.5709251165390015,
+      "learning_rate": 9.685123212655917e-07,
+      "loss": 0.9781,
+      "step": 3320
+    },
+    {
+      "epoch": 0.08007995953854675,
+      "grad_norm": 1.6757510900497437,
+      "learning_rate": 9.683855592739073e-07,
+      "loss": 1.0377,
+      "step": 3325
+    },
+    {
+      "epoch": 0.08020038053033404,
+      "grad_norm": 1.6469229459762573,
+      "learning_rate": 9.682587972822227e-07,
+      "loss": 0.9878,
+      "step": 3330
+    },
+    {
+      "epoch": 0.08032080152212133,
+      "grad_norm": 1.445010781288147,
+      "learning_rate": 9.681320352905384e-07,
+      "loss": 0.9342,
+      "step": 3335
+    },
+    {
+      "epoch": 0.08044122251390863,
+      "grad_norm": 2.0901291370391846,
+      "learning_rate": 9.68005273298854e-07,
+      "loss": 0.9842,
+      "step": 3340
+    },
+    {
+      "epoch": 0.08056164350569592,
+      "grad_norm": 1.6954982280731201,
+      "learning_rate": 9.678785113071697e-07,
+      "loss": 0.9938,
+      "step": 3345
+    },
+    {
+      "epoch": 0.0806820644974832,
+      "grad_norm": 1.504270076751709,
+      "learning_rate": 9.677517493154851e-07,
+      "loss": 0.906,
+      "step": 3350
+    },
+    {
+      "epoch": 0.08080248548927049,
+      "grad_norm": 1.5847922563552856,
+      "learning_rate": 9.676249873238008e-07,
+      "loss": 0.939,
+      "step": 3355
+    },
+    {
+      "epoch": 0.08092290648105778,
+      "grad_norm": 1.6328890323638916,
+      "learning_rate": 9.674982253321164e-07,
+      "loss": 0.9739,
+      "step": 3360
+    },
+    {
+      "epoch": 0.08104332747284507,
+      "grad_norm": 1.4796109199523926,
+      "learning_rate": 9.67371463340432e-07,
+      "loss": 0.9276,
+      "step": 3365
+    },
+    {
+      "epoch": 0.08116374846463235,
+      "grad_norm": 1.4965159893035889,
+      "learning_rate": 9.672447013487477e-07,
+      "loss": 0.9853,
+      "step": 3370
+    },
+    {
+      "epoch": 0.08128416945641964,
+      "grad_norm": 1.587471842765808,
+      "learning_rate": 9.671179393570631e-07,
+      "loss": 0.9829,
+      "step": 3375
+    },
+    {
+      "epoch": 0.08140459044820693,
+      "grad_norm": 1.473659873008728,
+      "learning_rate": 9.669911773653788e-07,
+      "loss": 0.9669,
+      "step": 3380
+    },
+    {
+      "epoch": 0.08152501143999422,
+      "grad_norm": 1.6593642234802246,
+      "learning_rate": 9.668644153736944e-07,
+      "loss": 0.9811,
+      "step": 3385
+    },
+    {
+      "epoch": 0.0816454324317815,
+      "grad_norm": 1.6420273780822754,
+      "learning_rate": 9.667376533820098e-07,
+      "loss": 0.9429,
+      "step": 3390
+    },
+    {
+      "epoch": 0.08176585342356879,
+      "grad_norm": 1.4429707527160645,
+      "learning_rate": 9.666108913903255e-07,
+      "loss": 0.8926,
+      "step": 3395
+    },
+    {
+      "epoch": 0.08188627441535609,
+      "grad_norm": 1.7320809364318848,
+      "learning_rate": 9.664841293986411e-07,
+      "loss": 1.0105,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08200669540714338,
+      "grad_norm": 1.5669728517532349,
+      "learning_rate": 9.663573674069566e-07,
+      "loss": 1.0101,
+      "step": 3405
+    },
+    {
+      "epoch": 0.08212711639893067,
+      "grad_norm": 1.6901319026947021,
+      "learning_rate": 9.662306054152722e-07,
+      "loss": 0.9212,
+      "step": 3410
+    },
+    {
+      "epoch": 0.08224753739071795,
+      "grad_norm": 1.5522770881652832,
+      "learning_rate": 9.661038434235879e-07,
+      "loss": 1.0259,
+      "step": 3415
+    },
+    {
+      "epoch": 0.08236795838250524,
+      "grad_norm": 1.5578824281692505,
+      "learning_rate": 9.659770814319033e-07,
+      "loss": 0.9705,
+      "step": 3420
+    },
+    {
+      "epoch": 0.08248837937429253,
+      "grad_norm": 1.5385942459106445,
+      "learning_rate": 9.65850319440219e-07,
+      "loss": 0.9389,
+      "step": 3425
+    },
+    {
+      "epoch": 0.08260880036607982,
+      "grad_norm": 1.453081488609314,
+      "learning_rate": 9.657235574485346e-07,
+      "loss": 0.917,
+      "step": 3430
+    },
+    {
+      "epoch": 0.0827292213578671,
+      "grad_norm": 1.5530083179473877,
+      "learning_rate": 9.655967954568502e-07,
+      "loss": 0.989,
+      "step": 3435
+    },
+    {
+      "epoch": 0.08284964234965439,
+      "grad_norm": 1.7280933856964111,
+      "learning_rate": 9.654700334651657e-07,
+      "loss": 0.9389,
+      "step": 3440
+    },
+    {
+      "epoch": 0.08297006334144168,
+      "grad_norm": 1.5677049160003662,
+      "learning_rate": 9.653432714734813e-07,
+      "loss": 0.9811,
+      "step": 3445
+    },
+    {
+      "epoch": 0.08309048433322896,
+      "grad_norm": 1.63473379611969,
+      "learning_rate": 9.65216509481797e-07,
+      "loss": 0.9095,
+      "step": 3450
+    },
+    {
+      "epoch": 0.08321090532501625,
+      "grad_norm": 1.622157335281372,
+      "learning_rate": 9.650897474901126e-07,
+      "loss": 0.9601,
+      "step": 3455
+    },
+    {
+      "epoch": 0.08333132631680354,
+      "grad_norm": 1.4803037643432617,
+      "learning_rate": 9.649629854984282e-07,
+      "loss": 0.9547,
+      "step": 3460
+    },
+    {
+      "epoch": 0.08345174730859084,
+      "grad_norm": 1.5026421546936035,
+      "learning_rate": 9.648362235067437e-07,
+      "loss": 1.0317,
+      "step": 3465
+    },
+    {
+      "epoch": 0.08357216830037813,
+      "grad_norm": 1.8309855461120605,
+      "learning_rate": 9.647094615150593e-07,
+      "loss": 0.956,
+      "step": 3470
+    },
+    {
+      "epoch": 0.08369258929216541,
+      "grad_norm": 1.4832587242126465,
+      "learning_rate": 9.64582699523375e-07,
+      "loss": 0.939,
+      "step": 3475
+    },
+    {
+      "epoch": 0.0838130102839527,
+      "grad_norm": 1.4295638799667358,
+      "learning_rate": 9.644559375316904e-07,
+      "loss": 1.0362,
+      "step": 3480
+    },
+    {
+      "epoch": 0.08393343127573999,
+      "grad_norm": 1.5575592517852783,
+      "learning_rate": 9.64329175540006e-07,
+      "loss": 1.0136,
+      "step": 3485
+    },
+    {
+      "epoch": 0.08405385226752728,
+      "grad_norm": 1.6031485795974731,
+      "learning_rate": 9.642024135483217e-07,
+      "loss": 0.9677,
+      "step": 3490
+    },
+    {
+      "epoch": 0.08417427325931456,
+      "grad_norm": 1.580514669418335,
+      "learning_rate": 9.640756515566371e-07,
+      "loss": 0.96,
+      "step": 3495
+    },
+    {
+      "epoch": 0.08429469425110185,
+      "grad_norm": 1.5536184310913086,
+      "learning_rate": 9.639488895649528e-07,
+      "loss": 0.9376,
+      "step": 3500
+    },
+    {
+      "epoch": 0.08441511524288914,
+      "grad_norm": 1.4722299575805664,
+      "learning_rate": 9.638221275732684e-07,
+      "loss": 0.9752,
+      "step": 3505
+    },
+    {
+      "epoch": 0.08453553623467643,
+      "grad_norm": 1.56247878074646,
+      "learning_rate": 9.636953655815838e-07,
+      "loss": 0.9566,
+      "step": 3510
+    },
+    {
+      "epoch": 0.08465595722646371,
+      "grad_norm": 1.4400372505187988,
+      "learning_rate": 9.635686035898995e-07,
+      "loss": 0.9702,
+      "step": 3515
+    },
+    {
+      "epoch": 0.084776378218251,
+      "grad_norm": 1.4799162149429321,
+      "learning_rate": 9.634418415982151e-07,
+      "loss": 0.9713,
+      "step": 3520
+    },
+    {
+      "epoch": 0.08489679921003829,
+      "grad_norm": 1.4603667259216309,
+      "learning_rate": 9.633150796065308e-07,
+      "loss": 0.9461,
+      "step": 3525
+    },
+    {
+      "epoch": 0.08501722020182559,
+      "grad_norm": 1.6365811824798584,
+      "learning_rate": 9.631883176148464e-07,
+      "loss": 0.9602,
+      "step": 3530
+    },
+    {
+      "epoch": 0.08513764119361288,
+      "grad_norm": 1.5613988637924194,
+      "learning_rate": 9.630615556231618e-07,
+      "loss": 0.9674,
+      "step": 3535
+    },
+    {
+      "epoch": 0.08525806218540016,
+      "grad_norm": 1.7290339469909668,
+      "learning_rate": 9.629347936314775e-07,
+      "loss": 0.9317,
+      "step": 3540
+    },
+    {
+      "epoch": 0.08537848317718745,
+      "grad_norm": 1.7438316345214844,
+      "learning_rate": 9.628080316397931e-07,
+      "loss": 0.957,
+      "step": 3545
+    },
+    {
+      "epoch": 0.08549890416897474,
+      "grad_norm": 1.7681242227554321,
+      "learning_rate": 9.626812696481088e-07,
+      "loss": 0.9307,
+      "step": 3550
+    },
+    {
+      "epoch": 0.08561932516076202,
+      "grad_norm": 1.491039752960205,
+      "learning_rate": 9.625545076564242e-07,
+      "loss": 0.9786,
+      "step": 3555
+    },
+    {
+      "epoch": 0.08573974615254931,
+      "grad_norm": 1.6438604593276978,
+      "learning_rate": 9.624277456647399e-07,
+      "loss": 0.9276,
+      "step": 3560
+    },
+    {
+      "epoch": 0.0858601671443366,
+      "grad_norm": 1.577164649963379,
+      "learning_rate": 9.623009836730555e-07,
+      "loss": 1.0229,
+      "step": 3565
+    },
+    {
+      "epoch": 0.08598058813612389,
+      "grad_norm": 1.5834629535675049,
+      "learning_rate": 9.62174221681371e-07,
+      "loss": 1.0192,
+      "step": 3570
+    },
+    {
+      "epoch": 0.08610100912791117,
+      "grad_norm": 1.7259495258331299,
+      "learning_rate": 9.620474596896866e-07,
+      "loss": 0.9634,
+      "step": 3575
+    },
+    {
+      "epoch": 0.08622143011969846,
+      "grad_norm": 1.504324197769165,
+      "learning_rate": 9.619206976980022e-07,
+      "loss": 0.9599,
+      "step": 3580
+    },
+    {
+      "epoch": 0.08634185111148575,
+      "grad_norm": 1.4879755973815918,
+      "learning_rate": 9.617939357063177e-07,
+      "loss": 0.9769,
+      "step": 3585
+    },
+    {
+      "epoch": 0.08646227210327304,
+      "grad_norm": 1.4818052053451538,
+      "learning_rate": 9.616671737146333e-07,
+      "loss": 0.9764,
+      "step": 3590
+    },
+    {
+      "epoch": 0.08658269309506034,
+      "grad_norm": 1.4573898315429688,
+      "learning_rate": 9.61540411722949e-07,
+      "loss": 0.9719,
+      "step": 3595
+    },
+    {
+      "epoch": 0.08670311408684762,
+      "grad_norm": 1.512553334236145,
+      "learning_rate": 9.614136497312644e-07,
+      "loss": 0.9713,
+      "step": 3600
+    },
+    {
+      "epoch": 0.08682353507863491,
+      "grad_norm": 1.518481969833374,
+      "learning_rate": 9.6128688773958e-07,
+      "loss": 1.0032,
+      "step": 3605
+    },
+    {
+      "epoch": 0.0869439560704222,
+      "grad_norm": 1.7753887176513672,
+      "learning_rate": 9.611601257478957e-07,
+      "loss": 0.9889,
+      "step": 3610
+    },
+    {
+      "epoch": 0.08706437706220949,
+      "grad_norm": 1.4080522060394287,
+      "learning_rate": 9.610333637562113e-07,
+      "loss": 0.9666,
+      "step": 3615
+    },
+    {
+      "epoch": 0.08718479805399677,
+      "grad_norm": 1.6740397214889526,
+      "learning_rate": 9.60906601764527e-07,
+      "loss": 0.9517,
+      "step": 3620
+    },
+    {
+      "epoch": 0.08730521904578406,
+      "grad_norm": 1.6367727518081665,
+      "learning_rate": 9.607798397728424e-07,
+      "loss": 0.9712,
+      "step": 3625
+    },
+    {
+      "epoch": 0.08742564003757135,
+      "grad_norm": 1.6772522926330566,
+      "learning_rate": 9.60653077781158e-07,
+      "loss": 0.9832,
+      "step": 3630
+    },
+    {
+      "epoch": 0.08754606102935863,
+      "grad_norm": 1.4191092252731323,
+      "learning_rate": 9.605263157894737e-07,
+      "loss": 0.947,
+      "step": 3635
+    },
+    {
+      "epoch": 0.08766648202114592,
+      "grad_norm": 1.4476778507232666,
+      "learning_rate": 9.603995537977893e-07,
+      "loss": 0.957,
+      "step": 3640
+    },
+    {
+      "epoch": 0.08778690301293321,
+      "grad_norm": 1.9380329847335815,
+      "learning_rate": 9.602727918061048e-07,
+      "loss": 1.0579,
+      "step": 3645
+    },
+    {
+      "epoch": 0.0879073240047205,
+      "grad_norm": 1.3948523998260498,
+      "learning_rate": 9.601460298144204e-07,
+      "loss": 0.9996,
+      "step": 3650
+    },
+    {
+      "epoch": 0.0880277449965078,
+      "grad_norm": 1.8436353206634521,
+      "learning_rate": 9.60019267822736e-07,
+      "loss": 0.9496,
+      "step": 3655
+    },
+    {
+      "epoch": 0.08814816598829509,
+      "grad_norm": 1.5903769731521606,
+      "learning_rate": 9.598925058310515e-07,
+      "loss": 0.993,
+      "step": 3660
+    },
+    {
+      "epoch": 0.08826858698008237,
+      "grad_norm": 1.574457049369812,
+      "learning_rate": 9.597657438393671e-07,
+      "loss": 1.0448,
+      "step": 3665
+    },
+    {
+      "epoch": 0.08838900797186966,
+      "grad_norm": 1.6137624979019165,
+      "learning_rate": 9.596389818476828e-07,
+      "loss": 1.0311,
+      "step": 3670
+    },
+    {
+      "epoch": 0.08850942896365695,
+      "grad_norm": 1.7495733499526978,
+      "learning_rate": 9.595122198559982e-07,
+      "loss": 1.0024,
+      "step": 3675
+    },
+    {
+      "epoch": 0.08862984995544423,
+      "grad_norm": 1.5811641216278076,
+      "learning_rate": 9.593854578643138e-07,
+      "loss": 0.9636,
+      "step": 3680
+    },
+    {
+      "epoch": 0.08875027094723152,
+      "grad_norm": 1.7564260959625244,
+      "learning_rate": 9.592586958726295e-07,
+      "loss": 0.9737,
+      "step": 3685
+    },
+    {
+      "epoch": 0.08887069193901881,
+      "grad_norm": 1.8012256622314453,
+      "learning_rate": 9.591319338809451e-07,
+      "loss": 0.991,
+      "step": 3690
+    },
+    {
+      "epoch": 0.0889911129308061,
+      "grad_norm": 1.920172095298767,
+      "learning_rate": 9.590051718892606e-07,
+      "loss": 1.0374,
+      "step": 3695
+    },
+    {
+      "epoch": 0.08911153392259338,
+      "grad_norm": 1.4977744817733765,
+      "learning_rate": 9.588784098975762e-07,
+      "loss": 0.9509,
+      "step": 3700
+    },
+    {
+      "epoch": 0.08923195491438067,
+      "grad_norm": 1.7665150165557861,
+      "learning_rate": 9.587516479058919e-07,
+      "loss": 0.9735,
+      "step": 3705
+    },
+    {
+      "epoch": 0.08935237590616796,
+      "grad_norm": 1.5105448961257935,
+      "learning_rate": 9.586248859142075e-07,
+      "loss": 1.0022,
+      "step": 3710
+    },
+    {
+      "epoch": 0.08947279689795525,
+      "grad_norm": 1.488258957862854,
+      "learning_rate": 9.584981239225232e-07,
+      "loss": 1.0177,
+      "step": 3715
+    },
+    {
+      "epoch": 0.08959321788974255,
+      "grad_norm": 1.6075270175933838,
+      "learning_rate": 9.583713619308386e-07,
+      "loss": 0.9024,
+      "step": 3720
+    },
+    {
+      "epoch": 0.08971363888152983,
+      "grad_norm": 1.5839999914169312,
+      "learning_rate": 9.582445999391542e-07,
+      "loss": 0.9843,
+      "step": 3725
+    },
+    {
+      "epoch": 0.08983405987331712,
+      "grad_norm": 1.4419506788253784,
+      "learning_rate": 9.581178379474699e-07,
+      "loss": 0.9383,
+      "step": 3730
+    },
+    {
+      "epoch": 0.08995448086510441,
+      "grad_norm": 1.660599946975708,
+      "learning_rate": 9.579910759557853e-07,
+      "loss": 0.9855,
+      "step": 3735
+    },
+    {
+      "epoch": 0.0900749018568917,
+      "grad_norm": 1.6551530361175537,
+      "learning_rate": 9.57864313964101e-07,
+      "loss": 0.9816,
+      "step": 3740
+    },
+    {
+      "epoch": 0.09019532284867898,
+      "grad_norm": 1.4094029664993286,
+      "learning_rate": 9.577375519724166e-07,
+      "loss": 0.9733,
+      "step": 3745
+    },
+    {
+      "epoch": 0.09031574384046627,
+      "grad_norm": 1.5762144327163696,
+      "learning_rate": 9.57610789980732e-07,
+      "loss": 0.9718,
+      "step": 3750
+    },
+    {
+      "epoch": 0.09043616483225356,
+      "grad_norm": 1.6266120672225952,
+      "learning_rate": 9.574840279890477e-07,
+      "loss": 0.9761,
+      "step": 3755
+    },
+    {
+      "epoch": 0.09055658582404084,
+      "grad_norm": 1.5777689218521118,
+      "learning_rate": 9.573572659973633e-07,
+      "loss": 1.0,
+      "step": 3760
+    },
+    {
+      "epoch": 0.09067700681582813,
+      "grad_norm": 1.5066728591918945,
+      "learning_rate": 9.572305040056788e-07,
+      "loss": 0.9302,
+      "step": 3765
+    },
+    {
+      "epoch": 0.09079742780761542,
+      "grad_norm": 1.7050273418426514,
+      "learning_rate": 9.571037420139944e-07,
+      "loss": 0.953,
+      "step": 3770
+    },
+    {
+      "epoch": 0.0909178487994027,
+      "grad_norm": 1.7422499656677246,
+      "learning_rate": 9.5697698002231e-07,
+      "loss": 1.003,
+      "step": 3775
+    },
+    {
+      "epoch": 0.09103826979119,
+      "grad_norm": 1.5633741617202759,
+      "learning_rate": 9.568502180306257e-07,
+      "loss": 0.9682,
+      "step": 3780
+    },
+    {
+      "epoch": 0.0911586907829773,
+      "grad_norm": 1.7610095739364624,
+      "learning_rate": 9.567234560389411e-07,
+      "loss": 0.9223,
+      "step": 3785
+    },
+    {
+      "epoch": 0.09127911177476458,
+      "grad_norm": 1.6257444620132446,
+      "learning_rate": 9.565966940472568e-07,
+      "loss": 0.9416,
+      "step": 3790
+    },
+    {
+      "epoch": 0.09139953276655187,
+      "grad_norm": 1.455300211906433,
+      "learning_rate": 9.564699320555724e-07,
+      "loss": 1.0258,
+      "step": 3795
+    },
+    {
+      "epoch": 0.09151995375833916,
+      "grad_norm": 1.596089243888855,
+      "learning_rate": 9.56343170063888e-07,
+      "loss": 1.001,
+      "step": 3800
+    },
+    {
+      "epoch": 0.09164037475012644,
+      "grad_norm": 1.5101442337036133,
+      "learning_rate": 9.562164080722037e-07,
+      "loss": 1.0037,
+      "step": 3805
+    },
+    {
+      "epoch": 0.09176079574191373,
+      "grad_norm": 1.5162341594696045,
+      "learning_rate": 9.560896460805191e-07,
+      "loss": 1.0177,
+      "step": 3810
+    },
+    {
+      "epoch": 0.09188121673370102,
+      "grad_norm": 1.6944547891616821,
+      "learning_rate": 9.559628840888348e-07,
+      "loss": 0.9239,
+      "step": 3815
+    },
+    {
+      "epoch": 0.0920016377254883,
+      "grad_norm": 1.5876063108444214,
+      "learning_rate": 9.558361220971504e-07,
+      "loss": 0.9869,
+      "step": 3820
+    },
+    {
+      "epoch": 0.09212205871727559,
+      "grad_norm": 1.616463303565979,
+      "learning_rate": 9.55709360105466e-07,
+      "loss": 1.03,
+      "step": 3825
+    },
+    {
+      "epoch": 0.09224247970906288,
+      "grad_norm": 1.567016363143921,
+      "learning_rate": 9.555825981137815e-07,
+      "loss": 0.9364,
+      "step": 3830
+    },
+    {
+      "epoch": 0.09236290070085017,
+      "grad_norm": 1.412610650062561,
+      "learning_rate": 9.554558361220971e-07,
+      "loss": 0.9713,
+      "step": 3835
+    },
+    {
+      "epoch": 0.09248332169263745,
+      "grad_norm": 1.6522778272628784,
+      "learning_rate": 9.553290741304128e-07,
+      "loss": 0.9959,
+      "step": 3840
+    },
+    {
+      "epoch": 0.09260374268442476,
+      "grad_norm": 1.5945355892181396,
+      "learning_rate": 9.552023121387282e-07,
+      "loss": 0.9766,
+      "step": 3845
+    },
+    {
+      "epoch": 0.09272416367621204,
+      "grad_norm": 1.6672556400299072,
+      "learning_rate": 9.550755501470439e-07,
+      "loss": 0.9743,
+      "step": 3850
+    },
+    {
+      "epoch": 0.09284458466799933,
+      "grad_norm": 1.7128037214279175,
+      "learning_rate": 9.549487881553595e-07,
+      "loss": 0.982,
+      "step": 3855
+    },
+    {
+      "epoch": 0.09296500565978662,
+      "grad_norm": 1.6894267797470093,
+      "learning_rate": 9.54822026163675e-07,
+      "loss": 0.9418,
+      "step": 3860
+    },
+    {
+      "epoch": 0.0930854266515739,
+      "grad_norm": 1.9653635025024414,
+      "learning_rate": 9.546952641719906e-07,
+      "loss": 1.0066,
+      "step": 3865
+    },
+    {
+      "epoch": 0.09320584764336119,
+      "grad_norm": 1.4464094638824463,
+      "learning_rate": 9.545685021803062e-07,
+      "loss": 0.9741,
+      "step": 3870
+    },
+    {
+      "epoch": 0.09332626863514848,
+      "grad_norm": 1.7934719324111938,
+      "learning_rate": 9.544417401886219e-07,
+      "loss": 0.9716,
+      "step": 3875
+    },
+    {
+      "epoch": 0.09344668962693577,
+      "grad_norm": 1.5995069742202759,
+      "learning_rate": 9.543149781969373e-07,
+      "loss": 0.9481,
+      "step": 3880
+    },
+    {
+      "epoch": 0.09356711061872305,
+      "grad_norm": 1.2907156944274902,
+      "learning_rate": 9.54188216205253e-07,
+      "loss": 0.8987,
+      "step": 3885
+    },
+    {
+      "epoch": 0.09368753161051034,
+      "grad_norm": 1.5202016830444336,
+      "learning_rate": 9.540614542135686e-07,
+      "loss": 0.897,
+      "step": 3890
+    },
+    {
+      "epoch": 0.09380795260229763,
+      "grad_norm": 1.5255939960479736,
+      "learning_rate": 9.539346922218842e-07,
+      "loss": 0.9924,
+      "step": 3895
+    },
+    {
+      "epoch": 0.09392837359408492,
+      "grad_norm": 1.9490069150924683,
+      "learning_rate": 9.538079302301999e-07,
+      "loss": 1.005,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0940487945858722,
+      "grad_norm": 1.5038361549377441,
+      "learning_rate": 9.536811682385153e-07,
+      "loss": 1.0007,
+      "step": 3905
+    },
+    {
+      "epoch": 0.0941692155776595,
+      "grad_norm": 1.6892212629318237,
+      "learning_rate": 9.53554406246831e-07,
+      "loss": 0.9309,
+      "step": 3910
+    },
+    {
+      "epoch": 0.09428963656944679,
+      "grad_norm": 1.5403926372528076,
+      "learning_rate": 9.534276442551465e-07,
+      "loss": 0.9412,
+      "step": 3915
+    },
+    {
+      "epoch": 0.09441005756123408,
+      "grad_norm": 1.5528895854949951,
+      "learning_rate": 9.53300882263462e-07,
+      "loss": 0.9814,
+      "step": 3920
+    },
+    {
+      "epoch": 0.09453047855302137,
+      "grad_norm": 1.7345143556594849,
+      "learning_rate": 9.531741202717777e-07,
+      "loss": 1.0062,
+      "step": 3925
+    },
+    {
+      "epoch": 0.09465089954480865,
+      "grad_norm": 1.5540204048156738,
+      "learning_rate": 9.530473582800933e-07,
+      "loss": 0.9077,
+      "step": 3930
+    },
+    {
+      "epoch": 0.09477132053659594,
+      "grad_norm": 1.5319840908050537,
+      "learning_rate": 9.529205962884088e-07,
+      "loss": 0.9675,
+      "step": 3935
+    },
+    {
+      "epoch": 0.09489174152838323,
+      "grad_norm": 1.5255876779556274,
+      "learning_rate": 9.527938342967244e-07,
+      "loss": 0.9512,
+      "step": 3940
+    },
+    {
+      "epoch": 0.09501216252017052,
+      "grad_norm": 1.393852710723877,
+      "learning_rate": 9.526670723050401e-07,
+      "loss": 0.9679,
+      "step": 3945
+    },
+    {
+      "epoch": 0.0951325835119578,
+      "grad_norm": 1.5980697870254517,
+      "learning_rate": 9.525403103133556e-07,
+      "loss": 0.9291,
+      "step": 3950
+    },
+    {
+      "epoch": 0.09525300450374509,
+      "grad_norm": 1.8581165075302124,
+      "learning_rate": 9.524135483216712e-07,
+      "loss": 0.8802,
+      "step": 3955
+    },
+    {
+      "epoch": 0.09537342549553238,
+      "grad_norm": 1.8296480178833008,
+      "learning_rate": 9.522867863299868e-07,
+      "loss": 0.9591,
+      "step": 3960
+    },
+    {
+      "epoch": 0.09549384648731966,
+      "grad_norm": 1.7010616064071655,
+      "learning_rate": 9.521600243383023e-07,
+      "loss": 0.9367,
+      "step": 3965
+    },
+    {
+      "epoch": 0.09561426747910695,
+      "grad_norm": 1.459513545036316,
+      "learning_rate": 9.52033262346618e-07,
+      "loss": 0.9193,
+      "step": 3970
+    },
+    {
+      "epoch": 0.09573468847089425,
+      "grad_norm": 1.5484166145324707,
+      "learning_rate": 9.519065003549336e-07,
+      "loss": 0.97,
+      "step": 3975
+    },
+    {
+      "epoch": 0.09585510946268154,
+      "grad_norm": 1.3898035287857056,
+      "learning_rate": 9.517797383632491e-07,
+      "loss": 1.0005,
+      "step": 3980
+    },
+    {
+      "epoch": 0.09597553045446883,
+      "grad_norm": 1.6612249612808228,
+      "learning_rate": 9.516529763715647e-07,
+      "loss": 0.9713,
+      "step": 3985
+    },
+    {
+      "epoch": 0.09609595144625611,
+      "grad_norm": 1.581580638885498,
+      "learning_rate": 9.515262143798803e-07,
+      "loss": 0.9966,
+      "step": 3990
+    },
+    {
+      "epoch": 0.0962163724380434,
+      "grad_norm": 1.6020958423614502,
+      "learning_rate": 9.513994523881959e-07,
+      "loss": 0.9582,
+      "step": 3995
+    },
+    {
+      "epoch": 0.09633679342983069,
+      "grad_norm": 1.6202304363250732,
+      "learning_rate": 9.512726903965115e-07,
+      "loss": 0.9673,
+      "step": 4000
+    },
+    {
+      "epoch": 0.09645721442161798,
+      "grad_norm": 1.4806082248687744,
+      "learning_rate": 9.51145928404827e-07,
+      "loss": 0.9414,
+      "step": 4005
+    },
+    {
+      "epoch": 0.09657763541340526,
+      "grad_norm": 1.4153348207473755,
+      "learning_rate": 9.510191664131426e-07,
+      "loss": 1.0094,
+      "step": 4010
+    },
+    {
+      "epoch": 0.09669805640519255,
+      "grad_norm": 1.5099906921386719,
+      "learning_rate": 9.508924044214582e-07,
+      "loss": 1.0002,
+      "step": 4015
+    },
+    {
+      "epoch": 0.09681847739697984,
+      "grad_norm": 1.4537502527236938,
+      "learning_rate": 9.507656424297739e-07,
+      "loss": 0.9604,
+      "step": 4020
+    },
+    {
+      "epoch": 0.09693889838876713,
+      "grad_norm": 1.5901180505752563,
+      "learning_rate": 9.506388804380894e-07,
+      "loss": 0.9957,
+      "step": 4025
+    },
+    {
+      "epoch": 0.09705931938055441,
+      "grad_norm": 1.530016303062439,
+      "learning_rate": 9.50512118446405e-07,
+      "loss": 0.9687,
+      "step": 4030
+    },
+    {
+      "epoch": 0.0971797403723417,
+      "grad_norm": 1.5487666130065918,
+      "learning_rate": 9.503853564547206e-07,
+      "loss": 0.964,
+      "step": 4035
+    },
+    {
+      "epoch": 0.097300161364129,
+      "grad_norm": 1.649968147277832,
+      "learning_rate": 9.502585944630361e-07,
+      "loss": 0.995,
+      "step": 4040
+    },
+    {
+      "epoch": 0.09742058235591629,
+      "grad_norm": 1.5910958051681519,
+      "learning_rate": 9.501318324713518e-07,
+      "loss": 0.9521,
+      "step": 4045
+    },
+    {
+      "epoch": 0.09754100334770358,
+      "grad_norm": 1.509109377861023,
+      "learning_rate": 9.500050704796674e-07,
+      "loss": 0.9757,
+      "step": 4050
+    },
+    {
+      "epoch": 0.09766142433949086,
+      "grad_norm": 1.4290871620178223,
+      "learning_rate": 9.498783084879829e-07,
+      "loss": 0.9602,
+      "step": 4055
+    },
+    {
+      "epoch": 0.09778184533127815,
+      "grad_norm": 1.5802230834960938,
+      "learning_rate": 9.497515464962985e-07,
+      "loss": 0.9717,
+      "step": 4060
+    },
+    {
+      "epoch": 0.09790226632306544,
+      "grad_norm": 1.4167861938476562,
+      "learning_rate": 9.496247845046142e-07,
+      "loss": 0.9509,
+      "step": 4065
+    },
+    {
+      "epoch": 0.09802268731485272,
+      "grad_norm": 1.5323904752731323,
+      "learning_rate": 9.494980225129297e-07,
+      "loss": 0.9557,
+      "step": 4070
+    },
+    {
+      "epoch": 0.09814310830664001,
+      "grad_norm": 1.2644950151443481,
+      "learning_rate": 9.493712605212452e-07,
+      "loss": 0.9239,
+      "step": 4075
+    },
+    {
+      "epoch": 0.0982635292984273,
+      "grad_norm": 1.574794054031372,
+      "learning_rate": 9.492444985295609e-07,
+      "loss": 0.9739,
+      "step": 4080
+    },
+    {
+      "epoch": 0.09838395029021459,
+      "grad_norm": 1.424659252166748,
+      "learning_rate": 9.491177365378764e-07,
+      "loss": 0.9867,
+      "step": 4085
+    },
+    {
+      "epoch": 0.09850437128200187,
+      "grad_norm": 1.5145343542099,
+      "learning_rate": 9.489909745461921e-07,
+      "loss": 0.9996,
+      "step": 4090
+    },
+    {
+      "epoch": 0.09862479227378916,
+      "grad_norm": 1.4513949155807495,
+      "learning_rate": 9.488642125545077e-07,
+      "loss": 0.9537,
+      "step": 4095
+    },
+    {
+      "epoch": 0.09874521326557646,
+      "grad_norm": 1.501413345336914,
+      "learning_rate": 9.487374505628231e-07,
+      "loss": 0.9569,
+      "step": 4100
+    },
+    {
+      "epoch": 0.09886563425736375,
+      "grad_norm": 1.536135196685791,
+      "learning_rate": 9.486106885711388e-07,
+      "loss": 0.9226,
+      "step": 4105
+    },
+    {
+      "epoch": 0.09898605524915104,
+      "grad_norm": 1.2886914014816284,
+      "learning_rate": 9.484839265794544e-07,
+      "loss": 0.9593,
+      "step": 4110
+    },
+    {
+      "epoch": 0.09910647624093832,
+      "grad_norm": 1.841726303100586,
+      "learning_rate": 9.4835716458777e-07,
+      "loss": 1.0258,
+      "step": 4115
+    },
+    {
+      "epoch": 0.09922689723272561,
+      "grad_norm": 1.5284498929977417,
+      "learning_rate": 9.482304025960855e-07,
+      "loss": 0.9678,
+      "step": 4120
+    },
+    {
+      "epoch": 0.0993473182245129,
+      "grad_norm": 3.6049418449401855,
+      "learning_rate": 9.481036406044011e-07,
+      "loss": 0.9033,
+      "step": 4125
+    },
+    {
+      "epoch": 0.09946773921630019,
+      "grad_norm": 1.560465931892395,
+      "learning_rate": 9.479768786127167e-07,
+      "loss": 0.9126,
+      "step": 4130
+    },
+    {
+      "epoch": 0.09958816020808747,
+      "grad_norm": 1.5048307180404663,
+      "learning_rate": 9.478501166210323e-07,
+      "loss": 0.9472,
+      "step": 4135
+    },
+    {
+      "epoch": 0.09970858119987476,
+      "grad_norm": 1.4856950044631958,
+      "learning_rate": 9.47723354629348e-07,
+      "loss": 0.9714,
+      "step": 4140
+    },
+    {
+      "epoch": 0.09982900219166205,
+      "grad_norm": 1.6258323192596436,
+      "learning_rate": 9.475965926376634e-07,
+      "loss": 0.9564,
+      "step": 4145
+    },
+    {
+      "epoch": 0.09994942318344933,
+      "grad_norm": 1.552347183227539,
+      "learning_rate": 9.474698306459791e-07,
+      "loss": 0.9473,
+      "step": 4150
+    },
+    {
+      "epoch": 0.10006984417523662,
+      "grad_norm": 1.8179153203964233,
+      "learning_rate": 9.473430686542947e-07,
+      "loss": 0.9837,
+      "step": 4155
+    },
+    {
+      "epoch": 0.10019026516702391,
+      "grad_norm": 1.4746826887130737,
+      "learning_rate": 9.472163066626102e-07,
+      "loss": 0.9202,
+      "step": 4160
+    },
+    {
+      "epoch": 0.10031068615881121,
+      "grad_norm": 1.7761483192443848,
+      "learning_rate": 9.470895446709259e-07,
+      "loss": 1.0175,
+      "step": 4165
+    },
+    {
+      "epoch": 0.1004311071505985,
+      "grad_norm": 1.3966078758239746,
+      "learning_rate": 9.469627826792414e-07,
+      "loss": 0.9678,
+      "step": 4170
+    },
+    {
+      "epoch": 0.10055152814238578,
+      "grad_norm": 1.6143893003463745,
+      "learning_rate": 9.46836020687557e-07,
+      "loss": 0.9992,
+      "step": 4175
+    },
+    {
+      "epoch": 0.10067194913417307,
+      "grad_norm": 1.681153655052185,
+      "learning_rate": 9.467092586958726e-07,
+      "loss": 0.965,
+      "step": 4180
+    },
+    {
+      "epoch": 0.10079237012596036,
+      "grad_norm": 1.4772725105285645,
+      "learning_rate": 9.465824967041883e-07,
+      "loss": 1.012,
+      "step": 4185
+    },
+    {
+      "epoch": 0.10091279111774765,
+      "grad_norm": 1.9717448949813843,
+      "learning_rate": 9.464557347125037e-07,
+      "loss": 1.0125,
+      "step": 4190
+    },
+    {
+      "epoch": 0.10103321210953493,
+      "grad_norm": 1.4508270025253296,
+      "learning_rate": 9.463289727208193e-07,
+      "loss": 0.9695,
+      "step": 4195
+    },
+    {
+      "epoch": 0.10115363310132222,
+      "grad_norm": 1.4963538646697998,
+      "learning_rate": 9.46202210729135e-07,
+      "loss": 1.0119,
+      "step": 4200
+    },
+    {
+      "epoch": 0.10127405409310951,
+      "grad_norm": 1.5520143508911133,
+      "learning_rate": 9.460754487374505e-07,
+      "loss": 0.9696,
+      "step": 4205
+    },
+    {
+      "epoch": 0.1013944750848968,
+      "grad_norm": 1.6408535242080688,
+      "learning_rate": 9.459486867457662e-07,
+      "loss": 0.9551,
+      "step": 4210
+    },
+    {
+      "epoch": 0.10151489607668408,
+      "grad_norm": 1.5083121061325073,
+      "learning_rate": 9.458219247540817e-07,
+      "loss": 0.9692,
+      "step": 4215
+    },
+    {
+      "epoch": 0.10163531706847137,
+      "grad_norm": 1.5416884422302246,
+      "learning_rate": 9.456951627623972e-07,
+      "loss": 0.9981,
+      "step": 4220
+    },
+    {
+      "epoch": 0.10175573806025866,
+      "grad_norm": 1.4579458236694336,
+      "learning_rate": 9.455684007707129e-07,
+      "loss": 0.9631,
+      "step": 4225
+    },
+    {
+      "epoch": 0.10187615905204596,
+      "grad_norm": 1.7778942584991455,
+      "learning_rate": 9.454416387790285e-07,
+      "loss": 0.9359,
+      "step": 4230
+    },
+    {
+      "epoch": 0.10199658004383325,
+      "grad_norm": 1.5496573448181152,
+      "learning_rate": 9.45314876787344e-07,
+      "loss": 0.9948,
+      "step": 4235
+    },
+    {
+      "epoch": 0.10211700103562053,
+      "grad_norm": 1.6221665143966675,
+      "learning_rate": 9.451881147956596e-07,
+      "loss": 0.9882,
+      "step": 4240
+    },
+    {
+      "epoch": 0.10223742202740782,
+      "grad_norm": 2.9152276515960693,
+      "learning_rate": 9.450613528039752e-07,
+      "loss": 0.9812,
+      "step": 4245
+    },
+    {
+      "epoch": 0.10235784301919511,
+      "grad_norm": 1.5360208749771118,
+      "learning_rate": 9.449345908122908e-07,
+      "loss": 0.944,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1024782640109824,
+      "grad_norm": 1.5541127920150757,
+      "learning_rate": 9.448078288206064e-07,
+      "loss": 0.9994,
+      "step": 4255
+    },
+    {
+      "epoch": 0.10259868500276968,
+      "grad_norm": 1.4846186637878418,
+      "learning_rate": 9.44681066828922e-07,
+      "loss": 0.9377,
+      "step": 4260
+    },
+    {
+      "epoch": 0.10271910599455697,
+      "grad_norm": 1.5796730518341064,
+      "learning_rate": 9.445543048372375e-07,
+      "loss": 0.903,
+      "step": 4265
+    },
+    {
+      "epoch": 0.10283952698634426,
+      "grad_norm": 1.6794935464859009,
+      "learning_rate": 9.444275428455532e-07,
+      "loss": 0.9928,
+      "step": 4270
+    },
+    {
+      "epoch": 0.10295994797813154,
+      "grad_norm": 1.5276776552200317,
+      "learning_rate": 9.443007808538688e-07,
+      "loss": 0.9501,
+      "step": 4275
+    },
+    {
+      "epoch": 0.10308036896991883,
+      "grad_norm": 1.5344417095184326,
+      "learning_rate": 9.441740188621843e-07,
+      "loss": 0.9711,
+      "step": 4280
+    },
+    {
+      "epoch": 0.10320078996170612,
+      "grad_norm": 1.7467256784439087,
+      "learning_rate": 9.440472568704999e-07,
+      "loss": 0.9535,
+      "step": 4285
+    },
+    {
+      "epoch": 0.10332121095349342,
+      "grad_norm": 1.38759183883667,
+      "learning_rate": 9.439204948788155e-07,
+      "loss": 0.9555,
+      "step": 4290
+    },
+    {
+      "epoch": 0.10344163194528071,
+      "grad_norm": 1.624312162399292,
+      "learning_rate": 9.437937328871311e-07,
+      "loss": 1.0224,
+      "step": 4295
+    },
+    {
+      "epoch": 0.103562052937068,
+      "grad_norm": 1.673833966255188,
+      "learning_rate": 9.436669708954467e-07,
+      "loss": 0.9612,
+      "step": 4300
+    },
+    {
+      "epoch": 0.10368247392885528,
+      "grad_norm": 1.6893752813339233,
+      "learning_rate": 9.435402089037622e-07,
+      "loss": 0.9885,
+      "step": 4305
+    },
+    {
+      "epoch": 0.10380289492064257,
+      "grad_norm": 1.6171820163726807,
+      "learning_rate": 9.434134469120778e-07,
+      "loss": 1.0103,
+      "step": 4310
+    },
+    {
+      "epoch": 0.10392331591242986,
+      "grad_norm": 1.706410527229309,
+      "learning_rate": 9.432866849203934e-07,
+      "loss": 0.9841,
+      "step": 4315
+    },
+    {
+      "epoch": 0.10404373690421714,
+      "grad_norm": 1.5108826160430908,
+      "learning_rate": 9.431599229287091e-07,
+      "loss": 0.9814,
+      "step": 4320
+    },
+    {
+      "epoch": 0.10416415789600443,
+      "grad_norm": 1.5877901315689087,
+      "learning_rate": 9.430331609370246e-07,
+      "loss": 0.9791,
+      "step": 4325
+    },
+    {
+      "epoch": 0.10428457888779172,
+      "grad_norm": 1.650411605834961,
+      "learning_rate": 9.429063989453401e-07,
+      "loss": 0.9648,
+      "step": 4330
+    },
+    {
+      "epoch": 0.104404999879579,
+      "grad_norm": 1.6145753860473633,
+      "learning_rate": 9.427796369536558e-07,
+      "loss": 0.9916,
+      "step": 4335
+    },
+    {
+      "epoch": 0.10452542087136629,
+      "grad_norm": 1.7058037519454956,
+      "learning_rate": 9.426528749619713e-07,
+      "loss": 0.9741,
+      "step": 4340
+    },
+    {
+      "epoch": 0.10464584186315358,
+      "grad_norm": 1.5574986934661865,
+      "learning_rate": 9.42526112970287e-07,
+      "loss": 0.9752,
+      "step": 4345
+    },
+    {
+      "epoch": 0.10476626285494087,
+      "grad_norm": 1.51565420627594,
+      "learning_rate": 9.423993509786026e-07,
+      "loss": 0.968,
+      "step": 4350
+    },
+    {
+      "epoch": 0.10488668384672817,
+      "grad_norm": 1.8395243883132935,
+      "learning_rate": 9.422725889869181e-07,
+      "loss": 0.953,
+      "step": 4355
+    },
+    {
+      "epoch": 0.10500710483851546,
+      "grad_norm": 1.5606211423873901,
+      "learning_rate": 9.421458269952337e-07,
+      "loss": 0.957,
+      "step": 4360
+    },
+    {
+      "epoch": 0.10512752583030274,
+      "grad_norm": 1.6414927244186401,
+      "learning_rate": 9.420190650035493e-07,
+      "loss": 0.9739,
+      "step": 4365
+    },
+    {
+      "epoch": 0.10524794682209003,
+      "grad_norm": 1.4883346557617188,
+      "learning_rate": 9.418923030118649e-07,
+      "loss": 0.9875,
+      "step": 4370
+    },
+    {
+      "epoch": 0.10536836781387732,
+      "grad_norm": 1.480233073234558,
+      "learning_rate": 9.417655410201804e-07,
+      "loss": 0.9975,
+      "step": 4375
+    },
+    {
+      "epoch": 0.1054887888056646,
+      "grad_norm": 1.6190452575683594,
+      "learning_rate": 9.416387790284961e-07,
+      "loss": 1.0012,
+      "step": 4380
+    },
+    {
+      "epoch": 0.10560920979745189,
+      "grad_norm": 1.6386826038360596,
+      "learning_rate": 9.415120170368116e-07,
+      "loss": 0.9849,
+      "step": 4385
+    },
+    {
+      "epoch": 0.10572963078923918,
+      "grad_norm": 1.881450891494751,
+      "learning_rate": 9.413852550451272e-07,
+      "loss": 0.923,
+      "step": 4390
+    },
+    {
+      "epoch": 0.10585005178102647,
+      "grad_norm": 1.536907434463501,
+      "learning_rate": 9.412584930534429e-07,
+      "loss": 0.9716,
+      "step": 4395
+    },
+    {
+      "epoch": 0.10597047277281375,
+      "grad_norm": 2.033087968826294,
+      "learning_rate": 9.411317310617583e-07,
+      "loss": 0.945,
+      "step": 4400
+    },
+    {
+      "epoch": 0.10609089376460104,
+      "grad_norm": 1.5700300931930542,
+      "learning_rate": 9.41004969070074e-07,
+      "loss": 0.9874,
+      "step": 4405
+    },
+    {
+      "epoch": 0.10621131475638833,
+      "grad_norm": 1.6118327379226685,
+      "learning_rate": 9.408782070783896e-07,
+      "loss": 0.9513,
+      "step": 4410
+    },
+    {
+      "epoch": 0.10633173574817562,
+      "grad_norm": 1.5397894382476807,
+      "learning_rate": 9.407514450867052e-07,
+      "loss": 0.9445,
+      "step": 4415
+    },
+    {
+      "epoch": 0.10645215673996292,
+      "grad_norm": 1.7848899364471436,
+      "learning_rate": 9.406246830950207e-07,
+      "loss": 0.9626,
+      "step": 4420
+    },
+    {
+      "epoch": 0.1065725777317502,
+      "grad_norm": 1.5513091087341309,
+      "learning_rate": 9.404979211033363e-07,
+      "loss": 0.9744,
+      "step": 4425
+    },
+    {
+      "epoch": 0.10669299872353749,
+      "grad_norm": 1.535009741783142,
+      "learning_rate": 9.403711591116519e-07,
+      "loss": 0.9383,
+      "step": 4430
+    },
+    {
+      "epoch": 0.10681341971532478,
+      "grad_norm": 1.59181547164917,
+      "learning_rate": 9.402443971199675e-07,
+      "loss": 0.9882,
+      "step": 4435
+    },
+    {
+      "epoch": 0.10693384070711207,
+      "grad_norm": 1.7125526666641235,
+      "learning_rate": 9.401176351282832e-07,
+      "loss": 0.979,
+      "step": 4440
+    },
+    {
+      "epoch": 0.10705426169889935,
+      "grad_norm": 1.619505524635315,
+      "learning_rate": 9.399908731365986e-07,
+      "loss": 0.9604,
+      "step": 4445
+    },
+    {
+      "epoch": 0.10717468269068664,
+      "grad_norm": 1.9453794956207275,
+      "learning_rate": 9.398641111449142e-07,
+      "loss": 0.9507,
+      "step": 4450
+    },
+    {
+      "epoch": 0.10729510368247393,
+      "grad_norm": 1.5830106735229492,
+      "learning_rate": 9.397373491532299e-07,
+      "loss": 0.9718,
+      "step": 4455
+    },
+    {
+      "epoch": 0.10741552467426121,
+      "grad_norm": 1.6598644256591797,
+      "learning_rate": 9.396105871615454e-07,
+      "loss": 1.0022,
+      "step": 4460
+    },
+    {
+      "epoch": 0.1075359456660485,
+      "grad_norm": 1.6971837282180786,
+      "learning_rate": 9.394838251698611e-07,
+      "loss": 1.0149,
+      "step": 4465
+    },
+    {
+      "epoch": 0.10765636665783579,
+      "grad_norm": 1.7452025413513184,
+      "learning_rate": 9.393570631781766e-07,
+      "loss": 0.9395,
+      "step": 4470
+    },
+    {
+      "epoch": 0.10777678764962308,
+      "grad_norm": 1.655214786529541,
+      "learning_rate": 9.392303011864922e-07,
+      "loss": 0.9422,
+      "step": 4475
+    },
+    {
+      "epoch": 0.10789720864141038,
+      "grad_norm": 1.542333960533142,
+      "learning_rate": 9.391035391948078e-07,
+      "loss": 1.021,
+      "step": 4480
+    },
+    {
+      "epoch": 0.10801762963319766,
+      "grad_norm": 1.7552878856658936,
+      "learning_rate": 9.389767772031234e-07,
+      "loss": 0.9563,
+      "step": 4485
+    },
+    {
+      "epoch": 0.10813805062498495,
+      "grad_norm": 1.8255192041397095,
+      "learning_rate": 9.388500152114389e-07,
+      "loss": 0.9953,
+      "step": 4490
+    },
+    {
+      "epoch": 0.10825847161677224,
+      "grad_norm": 1.5070478916168213,
+      "learning_rate": 9.387232532197545e-07,
+      "loss": 0.9908,
+      "step": 4495
+    },
+    {
+      "epoch": 0.10837889260855953,
+      "grad_norm": 1.4872255325317383,
+      "learning_rate": 9.385964912280702e-07,
+      "loss": 0.9305,
+      "step": 4500
+    },
+    {
+      "epoch": 0.10849931360034681,
+      "grad_norm": 1.501749873161316,
+      "learning_rate": 9.384697292363857e-07,
+      "loss": 0.9861,
+      "step": 4505
+    },
+    {
+      "epoch": 0.1086197345921341,
+      "grad_norm": 1.7855017185211182,
+      "learning_rate": 9.383429672447013e-07,
+      "loss": 0.9322,
+      "step": 4510
+    },
+    {
+      "epoch": 0.10874015558392139,
+      "grad_norm": 1.3943707942962646,
+      "learning_rate": 9.382162052530169e-07,
+      "loss": 0.9588,
+      "step": 4515
+    },
+    {
+      "epoch": 0.10886057657570868,
+      "grad_norm": 1.5438886880874634,
+      "learning_rate": 9.380894432613324e-07,
+      "loss": 0.9358,
+      "step": 4520
+    },
+    {
+      "epoch": 0.10898099756749596,
+      "grad_norm": 1.761030673980713,
+      "learning_rate": 9.379626812696481e-07,
+      "loss": 0.9704,
+      "step": 4525
+    },
+    {
+      "epoch": 0.10910141855928325,
+      "grad_norm": 1.877665638923645,
+      "learning_rate": 9.378359192779637e-07,
+      "loss": 1.0007,
+      "step": 4530
+    },
+    {
+      "epoch": 0.10922183955107054,
+      "grad_norm": 1.6068402528762817,
+      "learning_rate": 9.377091572862791e-07,
+      "loss": 0.9647,
+      "step": 4535
+    },
+    {
+      "epoch": 0.10934226054285782,
+      "grad_norm": 1.6051069498062134,
+      "learning_rate": 9.375823952945948e-07,
+      "loss": 0.9498,
+      "step": 4540
+    },
+    {
+      "epoch": 0.10946268153464513,
+      "grad_norm": 1.457535982131958,
+      "learning_rate": 9.374556333029104e-07,
+      "loss": 0.9663,
+      "step": 4545
+    },
+    {
+      "epoch": 0.10958310252643241,
+      "grad_norm": 1.6678305864334106,
+      "learning_rate": 9.37328871311226e-07,
+      "loss": 1.0178,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1097035235182197,
+      "grad_norm": 1.4807989597320557,
+      "learning_rate": 9.372021093195416e-07,
+      "loss": 0.8829,
+      "step": 4555
+    },
+    {
+      "epoch": 0.10982394451000699,
+      "grad_norm": 1.5890085697174072,
+      "learning_rate": 9.370753473278572e-07,
+      "loss": 0.9951,
+      "step": 4560
+    },
+    {
+      "epoch": 0.10994436550179428,
+      "grad_norm": 1.8150923252105713,
+      "learning_rate": 9.369485853361727e-07,
+      "loss": 0.8929,
+      "step": 4565
+    },
+    {
+      "epoch": 0.11006478649358156,
+      "grad_norm": 1.5665993690490723,
+      "learning_rate": 9.368218233444883e-07,
+      "loss": 0.9729,
+      "step": 4570
+    },
+    {
+      "epoch": 0.11018520748536885,
+      "grad_norm": 1.9737279415130615,
+      "learning_rate": 9.36695061352804e-07,
+      "loss": 1.0065,
+      "step": 4575
+    },
+    {
+      "epoch": 0.11030562847715614,
+      "grad_norm": 1.4127687215805054,
+      "learning_rate": 9.365682993611195e-07,
+      "loss": 0.9853,
+      "step": 4580
+    },
+    {
+      "epoch": 0.11042604946894342,
+      "grad_norm": 1.557056188583374,
+      "learning_rate": 9.364415373694351e-07,
+      "loss": 0.9623,
+      "step": 4585
+    },
+    {
+      "epoch": 0.11054647046073071,
+      "grad_norm": 1.6299985647201538,
+      "learning_rate": 9.363147753777507e-07,
+      "loss": 0.9772,
+      "step": 4590
+    },
+    {
+      "epoch": 0.110666891452518,
+      "grad_norm": 1.655799388885498,
+      "learning_rate": 9.361880133860662e-07,
+      "loss": 1.0002,
+      "step": 4595
+    },
+    {
+      "epoch": 0.11078731244430529,
+      "grad_norm": 1.546984314918518,
+      "learning_rate": 9.360612513943819e-07,
+      "loss": 0.8815,
+      "step": 4600
+    },
+    {
+      "epoch": 0.11090773343609257,
+      "grad_norm": 1.4920119047164917,
+      "learning_rate": 9.359344894026974e-07,
+      "loss": 0.9835,
+      "step": 4605
+    },
+    {
+      "epoch": 0.11102815442787987,
+      "grad_norm": 1.5333186388015747,
+      "learning_rate": 9.35807727411013e-07,
+      "loss": 0.9534,
+      "step": 4610
+    },
+    {
+      "epoch": 0.11114857541966716,
+      "grad_norm": 1.6916743516921997,
+      "learning_rate": 9.356809654193286e-07,
+      "loss": 0.9304,
+      "step": 4615
+    },
+    {
+      "epoch": 0.11126899641145445,
+      "grad_norm": 1.8019620180130005,
+      "learning_rate": 9.355542034276443e-07,
+      "loss": 1.0045,
+      "step": 4620
+    },
+    {
+      "epoch": 0.11138941740324174,
+      "grad_norm": 1.4818156957626343,
+      "learning_rate": 9.354274414359598e-07,
+      "loss": 0.9803,
+      "step": 4625
+    },
+    {
+      "epoch": 0.11150983839502902,
+      "grad_norm": 1.5333963632583618,
+      "learning_rate": 9.353006794442753e-07,
+      "loss": 0.9324,
+      "step": 4630
+    },
+    {
+      "epoch": 0.11163025938681631,
+      "grad_norm": 1.7001012563705444,
+      "learning_rate": 9.35173917452591e-07,
+      "loss": 0.8934,
+      "step": 4635
+    },
+    {
+      "epoch": 0.1117506803786036,
+      "grad_norm": 1.4808404445648193,
+      "learning_rate": 9.350471554609065e-07,
+      "loss": 0.9245,
+      "step": 4640
+    },
+    {
+      "epoch": 0.11187110137039089,
+      "grad_norm": 1.511756181716919,
+      "learning_rate": 9.349203934692222e-07,
+      "loss": 0.9419,
+      "step": 4645
+    },
+    {
+      "epoch": 0.11199152236217817,
+      "grad_norm": 1.5101591348648071,
+      "learning_rate": 9.347936314775378e-07,
+      "loss": 0.9783,
+      "step": 4650
+    },
+    {
+      "epoch": 0.11211194335396546,
+      "grad_norm": 1.513121247291565,
+      "learning_rate": 9.346668694858532e-07,
+      "loss": 1.0224,
+      "step": 4655
+    },
+    {
+      "epoch": 0.11223236434575275,
+      "grad_norm": 1.4988608360290527,
+      "learning_rate": 9.345401074941689e-07,
+      "loss": 0.9689,
+      "step": 4660
+    },
+    {
+      "epoch": 0.11235278533754003,
+      "grad_norm": 1.5721975564956665,
+      "learning_rate": 9.344133455024845e-07,
+      "loss": 0.9555,
+      "step": 4665
+    },
+    {
+      "epoch": 0.11247320632932732,
+      "grad_norm": 1.696729302406311,
+      "learning_rate": 9.342865835108001e-07,
+      "loss": 0.9495,
+      "step": 4670
+    },
+    {
+      "epoch": 0.11259362732111462,
+      "grad_norm": 1.4797080755233765,
+      "learning_rate": 9.341598215191156e-07,
+      "loss": 0.9938,
+      "step": 4675
+    },
+    {
+      "epoch": 0.11271404831290191,
+      "grad_norm": 1.5340688228607178,
+      "learning_rate": 9.340330595274313e-07,
+      "loss": 0.9271,
+      "step": 4680
+    },
+    {
+      "epoch": 0.1128344693046892,
+      "grad_norm": 1.5662153959274292,
+      "learning_rate": 9.339062975357468e-07,
+      "loss": 0.98,
+      "step": 4685
+    },
+    {
+      "epoch": 0.11295489029647648,
+      "grad_norm": 1.5172141790390015,
+      "learning_rate": 9.337795355440624e-07,
+      "loss": 0.958,
+      "step": 4690
+    },
+    {
+      "epoch": 0.11307531128826377,
+      "grad_norm": 1.6660382747650146,
+      "learning_rate": 9.336527735523781e-07,
+      "loss": 0.9115,
+      "step": 4695
+    },
+    {
+      "epoch": 0.11319573228005106,
+      "grad_norm": 1.740270972251892,
+      "learning_rate": 9.335260115606935e-07,
+      "loss": 0.9381,
+      "step": 4700
+    },
+    {
+      "epoch": 0.11331615327183835,
+      "grad_norm": 1.401108980178833,
+      "learning_rate": 9.333992495690092e-07,
+      "loss": 0.9772,
+      "step": 4705
+    },
+    {
+      "epoch": 0.11343657426362563,
+      "grad_norm": 1.6223112344741821,
+      "learning_rate": 9.332724875773248e-07,
+      "loss": 0.9462,
+      "step": 4710
+    },
+    {
+      "epoch": 0.11355699525541292,
+      "grad_norm": 1.654981255531311,
+      "learning_rate": 9.331457255856403e-07,
+      "loss": 0.9887,
+      "step": 4715
+    },
+    {
+      "epoch": 0.11367741624720021,
+      "grad_norm": 1.5551520586013794,
+      "learning_rate": 9.330189635939559e-07,
+      "loss": 0.9886,
+      "step": 4720
+    },
+    {
+      "epoch": 0.1137978372389875,
+      "grad_norm": 2.040355920791626,
+      "learning_rate": 9.328922016022715e-07,
+      "loss": 0.9203,
+      "step": 4725
+    },
+    {
+      "epoch": 0.11391825823077478,
+      "grad_norm": 1.7608866691589355,
+      "learning_rate": 9.327654396105871e-07,
+      "loss": 0.962,
+      "step": 4730
+    },
+    {
+      "epoch": 0.11403867922256208,
+      "grad_norm": 2.0532608032226562,
+      "learning_rate": 9.326386776189027e-07,
+      "loss": 0.9713,
+      "step": 4735
+    },
+    {
+      "epoch": 0.11415910021434937,
+      "grad_norm": 1.3905706405639648,
+      "learning_rate": 9.325119156272184e-07,
+      "loss": 0.9415,
+      "step": 4740
+    },
+    {
+      "epoch": 0.11427952120613666,
+      "grad_norm": 1.5101875066757202,
+      "learning_rate": 9.323851536355338e-07,
+      "loss": 1.0173,
+      "step": 4745
+    },
+    {
+      "epoch": 0.11439994219792395,
+      "grad_norm": 1.4829469919204712,
+      "learning_rate": 9.322583916438494e-07,
+      "loss": 1.0155,
+      "step": 4750
+    },
+    {
+      "epoch": 0.11452036318971123,
+      "grad_norm": 1.5244263410568237,
+      "learning_rate": 9.321316296521651e-07,
+      "loss": 0.9863,
+      "step": 4755
+    },
+    {
+      "epoch": 0.11464078418149852,
+      "grad_norm": 1.638397216796875,
+      "learning_rate": 9.320048676604806e-07,
+      "loss": 1.004,
+      "step": 4760
+    },
+    {
+      "epoch": 0.11476120517328581,
+      "grad_norm": 1.5108238458633423,
+      "learning_rate": 9.318781056687963e-07,
+      "loss": 0.9795,
+      "step": 4765
+    },
+    {
+      "epoch": 0.1148816261650731,
+      "grad_norm": 1.6074522733688354,
+      "learning_rate": 9.317513436771118e-07,
+      "loss": 1.0123,
+      "step": 4770
+    },
+    {
+      "epoch": 0.11500204715686038,
+      "grad_norm": 1.4743024110794067,
+      "learning_rate": 9.316245816854273e-07,
+      "loss": 0.9394,
+      "step": 4775
+    },
+    {
+      "epoch": 0.11512246814864767,
+      "grad_norm": 1.5913125276565552,
+      "learning_rate": 9.31497819693743e-07,
+      "loss": 0.9358,
+      "step": 4780
+    },
+    {
+      "epoch": 0.11524288914043496,
+      "grad_norm": 1.4869250059127808,
+      "learning_rate": 9.313710577020586e-07,
+      "loss": 0.9499,
+      "step": 4785
+    },
+    {
+      "epoch": 0.11536331013222224,
+      "grad_norm": 1.6868833303451538,
+      "learning_rate": 9.312442957103741e-07,
+      "loss": 0.9614,
+      "step": 4790
+    },
+    {
+      "epoch": 0.11548373112400953,
+      "grad_norm": 1.4460880756378174,
+      "learning_rate": 9.311175337186897e-07,
+      "loss": 0.8808,
+      "step": 4795
+    },
+    {
+      "epoch": 0.11560415211579683,
+      "grad_norm": 1.3952046632766724,
+      "learning_rate": 9.309907717270054e-07,
+      "loss": 0.9999,
+      "step": 4800
+    },
+    {
+      "epoch": 0.11572457310758412,
+      "grad_norm": 1.459100365638733,
+      "learning_rate": 9.308640097353209e-07,
+      "loss": 0.9508,
+      "step": 4805
+    },
+    {
+      "epoch": 0.1158449940993714,
+      "grad_norm": 1.4393701553344727,
+      "learning_rate": 9.307372477436365e-07,
+      "loss": 0.9974,
+      "step": 4810
+    },
+    {
+      "epoch": 0.1159654150911587,
+      "grad_norm": 1.6273918151855469,
+      "learning_rate": 9.306104857519521e-07,
+      "loss": 0.9248,
+      "step": 4815
+    },
+    {
+      "epoch": 0.11608583608294598,
+      "grad_norm": 1.478814959526062,
+      "learning_rate": 9.304837237602676e-07,
+      "loss": 0.9576,
+      "step": 4820
+    },
+    {
+      "epoch": 0.11620625707473327,
+      "grad_norm": 1.8840432167053223,
+      "learning_rate": 9.303569617685833e-07,
+      "loss": 1.0513,
+      "step": 4825
+    },
+    {
+      "epoch": 0.11632667806652056,
+      "grad_norm": 1.6974865198135376,
+      "learning_rate": 9.302301997768989e-07,
+      "loss": 0.9699,
+      "step": 4830
+    },
+    {
+      "epoch": 0.11644709905830784,
+      "grad_norm": 1.5943163633346558,
+      "learning_rate": 9.301034377852143e-07,
+      "loss": 0.946,
+      "step": 4835
+    },
+    {
+      "epoch": 0.11656752005009513,
+      "grad_norm": 1.5880036354064941,
+      "learning_rate": 9.2997667579353e-07,
+      "loss": 0.9744,
+      "step": 4840
+    },
+    {
+      "epoch": 0.11668794104188242,
+      "grad_norm": 1.8085670471191406,
+      "learning_rate": 9.298499138018456e-07,
+      "loss": 0.9736,
+      "step": 4845
+    },
+    {
+      "epoch": 0.1168083620336697,
+      "grad_norm": 1.504788875579834,
+      "learning_rate": 9.297231518101612e-07,
+      "loss": 0.9505,
+      "step": 4850
+    },
+    {
+      "epoch": 0.11692878302545699,
+      "grad_norm": 1.4562904834747314,
+      "learning_rate": 9.295963898184768e-07,
+      "loss": 0.9379,
+      "step": 4855
+    },
+    {
+      "epoch": 0.11704920401724428,
+      "grad_norm": 1.4931610822677612,
+      "learning_rate": 9.294696278267923e-07,
+      "loss": 0.9242,
+      "step": 4860
+    },
+    {
+      "epoch": 0.11716962500903158,
+      "grad_norm": 1.4892380237579346,
+      "learning_rate": 9.29342865835108e-07,
+      "loss": 0.9444,
+      "step": 4865
+    },
+    {
+      "epoch": 0.11729004600081887,
+      "grad_norm": 1.8332098722457886,
+      "learning_rate": 9.292161038434235e-07,
+      "loss": 1.0237,
+      "step": 4870
+    },
+    {
+      "epoch": 0.11741046699260616,
+      "grad_norm": 1.5724458694458008,
+      "learning_rate": 9.290893418517392e-07,
+      "loss": 0.8696,
+      "step": 4875
+    },
+    {
+      "epoch": 0.11753088798439344,
+      "grad_norm": 1.5204848051071167,
+      "learning_rate": 9.289625798600548e-07,
+      "loss": 0.9876,
+      "step": 4880
+    },
+    {
+      "epoch": 0.11765130897618073,
+      "grad_norm": 1.428585410118103,
+      "learning_rate": 9.288358178683703e-07,
+      "loss": 0.935,
+      "step": 4885
+    },
+    {
+      "epoch": 0.11777172996796802,
+      "grad_norm": 1.7297282218933105,
+      "learning_rate": 9.287090558766859e-07,
+      "loss": 0.9904,
+      "step": 4890
+    },
+    {
+      "epoch": 0.1178921509597553,
+      "grad_norm": 1.5793259143829346,
+      "learning_rate": 9.285822938850015e-07,
+      "loss": 0.9694,
+      "step": 4895
+    },
+    {
+      "epoch": 0.11801257195154259,
+      "grad_norm": 1.7904987335205078,
+      "learning_rate": 9.284555318933171e-07,
+      "loss": 0.9852,
+      "step": 4900
+    },
+    {
+      "epoch": 0.11813299294332988,
+      "grad_norm": 1.4578253030776978,
+      "learning_rate": 9.283287699016326e-07,
+      "loss": 0.9786,
+      "step": 4905
+    },
+    {
+      "epoch": 0.11825341393511717,
+      "grad_norm": 1.439958930015564,
+      "learning_rate": 9.282020079099483e-07,
+      "loss": 0.9291,
+      "step": 4910
+    },
+    {
+      "epoch": 0.11837383492690445,
+      "grad_norm": 1.6253753900527954,
+      "learning_rate": 9.280752459182638e-07,
+      "loss": 0.9948,
+      "step": 4915
+    },
+    {
+      "epoch": 0.11849425591869174,
+      "grad_norm": 1.617333173751831,
+      "learning_rate": 9.279484839265794e-07,
+      "loss": 0.9229,
+      "step": 4920
+    },
+    {
+      "epoch": 0.11861467691047904,
+      "grad_norm": 1.7600502967834473,
+      "learning_rate": 9.278217219348951e-07,
+      "loss": 0.9895,
+      "step": 4925
+    },
+    {
+      "epoch": 0.11873509790226633,
+      "grad_norm": 1.636748194694519,
+      "learning_rate": 9.276949599432105e-07,
+      "loss": 0.9766,
+      "step": 4930
+    },
+    {
+      "epoch": 0.11885551889405362,
+      "grad_norm": 1.6138620376586914,
+      "learning_rate": 9.275681979515262e-07,
+      "loss": 0.9108,
+      "step": 4935
+    },
+    {
+      "epoch": 0.1189759398858409,
+      "grad_norm": 1.5964850187301636,
+      "learning_rate": 9.274414359598418e-07,
+      "loss": 0.9817,
+      "step": 4940
+    },
+    {
+      "epoch": 0.11909636087762819,
+      "grad_norm": 1.6752839088439941,
+      "learning_rate": 9.273146739681574e-07,
+      "loss": 0.9545,
+      "step": 4945
+    },
+    {
+      "epoch": 0.11921678186941548,
+      "grad_norm": 1.6496039628982544,
+      "learning_rate": 9.271879119764729e-07,
+      "loss": 0.9907,
+      "step": 4950
+    },
+    {
+      "epoch": 0.11933720286120277,
+      "grad_norm": 1.4864082336425781,
+      "learning_rate": 9.270611499847885e-07,
+      "loss": 0.9827,
+      "step": 4955
+    },
+    {
+      "epoch": 0.11945762385299005,
+      "grad_norm": 1.483738660812378,
+      "learning_rate": 9.269343879931041e-07,
+      "loss": 0.9609,
+      "step": 4960
+    },
+    {
+      "epoch": 0.11957804484477734,
+      "grad_norm": 1.645350694656372,
+      "learning_rate": 9.268076260014197e-07,
+      "loss": 0.9546,
+      "step": 4965
+    },
+    {
+      "epoch": 0.11969846583656463,
+      "grad_norm": 1.473766803741455,
+      "learning_rate": 9.266808640097354e-07,
+      "loss": 0.9515,
+      "step": 4970
+    },
+    {
+      "epoch": 0.11981888682835191,
+      "grad_norm": 1.4186217784881592,
+      "learning_rate": 9.265541020180508e-07,
+      "loss": 1.0048,
+      "step": 4975
+    },
+    {
+      "epoch": 0.1199393078201392,
+      "grad_norm": 1.6372003555297852,
+      "learning_rate": 9.264273400263664e-07,
+      "loss": 0.9712,
+      "step": 4980
+    },
+    {
+      "epoch": 0.12005972881192649,
+      "grad_norm": 2.0134482383728027,
+      "learning_rate": 9.263005780346821e-07,
+      "loss": 0.9544,
+      "step": 4985
+    },
+    {
+      "epoch": 0.12018014980371379,
+      "grad_norm": 1.6138123273849487,
+      "learning_rate": 9.261738160429976e-07,
+      "loss": 0.9704,
+      "step": 4990
+    },
+    {
+      "epoch": 0.12030057079550108,
+      "grad_norm": 1.5256341695785522,
+      "learning_rate": 9.260470540513133e-07,
+      "loss": 0.9477,
+      "step": 4995
+    },
+    {
+      "epoch": 0.12042099178728836,
+      "grad_norm": 1.558856725692749,
+      "learning_rate": 9.259202920596288e-07,
+      "loss": 0.9431,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12054141277907565,
+      "grad_norm": 1.8143106698989868,
+      "learning_rate": 9.257935300679444e-07,
+      "loss": 0.9618,
+      "step": 5005
+    },
+    {
+      "epoch": 0.12066183377086294,
+      "grad_norm": 1.6698336601257324,
+      "learning_rate": 9.2566676807626e-07,
+      "loss": 0.9523,
+      "step": 5010
+    },
+    {
+      "epoch": 0.12078225476265023,
+      "grad_norm": 1.6376757621765137,
+      "learning_rate": 9.255400060845756e-07,
+      "loss": 0.8967,
+      "step": 5015
+    },
+    {
+      "epoch": 0.12090267575443751,
+      "grad_norm": 1.6007685661315918,
+      "learning_rate": 9.254132440928911e-07,
+      "loss": 0.9103,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1210230967462248,
+      "grad_norm": 1.5988738536834717,
+      "learning_rate": 9.252864821012067e-07,
+      "loss": 0.9777,
+      "step": 5025
+    },
+    {
+      "epoch": 0.12114351773801209,
+      "grad_norm": 1.5461716651916504,
+      "learning_rate": 9.251597201095224e-07,
+      "loss": 0.9269,
+      "step": 5030
+    },
+    {
+      "epoch": 0.12126393872979938,
+      "grad_norm": 1.6856049299240112,
+      "learning_rate": 9.250329581178379e-07,
+      "loss": 0.9482,
+      "step": 5035
+    },
+    {
+      "epoch": 0.12138435972158666,
+      "grad_norm": 1.4944417476654053,
+      "learning_rate": 9.249061961261535e-07,
+      "loss": 0.9646,
+      "step": 5040
+    },
+    {
+      "epoch": 0.12150478071337395,
+      "grad_norm": 1.87632155418396,
+      "learning_rate": 9.247794341344691e-07,
+      "loss": 0.9717,
+      "step": 5045
+    },
+    {
+      "epoch": 0.12162520170516124,
+      "grad_norm": 1.8691729307174683,
+      "learning_rate": 9.246526721427846e-07,
+      "loss": 0.929,
+      "step": 5050
+    },
+    {
+      "epoch": 0.12174562269694854,
+      "grad_norm": 1.5722233057022095,
+      "learning_rate": 9.245259101511003e-07,
+      "loss": 0.9637,
+      "step": 5055
+    },
+    {
+      "epoch": 0.12186604368873583,
+      "grad_norm": 1.6120997667312622,
+      "learning_rate": 9.243991481594159e-07,
+      "loss": 1.0072,
+      "step": 5060
+    },
+    {
+      "epoch": 0.12198646468052311,
+      "grad_norm": 1.6073572635650635,
+      "learning_rate": 9.242723861677313e-07,
+      "loss": 0.9322,
+      "step": 5065
+    },
+    {
+      "epoch": 0.1221068856723104,
+      "grad_norm": 1.6781420707702637,
+      "learning_rate": 9.24145624176047e-07,
+      "loss": 0.9434,
+      "step": 5070
+    },
+    {
+      "epoch": 0.12222730666409769,
+      "grad_norm": 1.8128712177276611,
+      "learning_rate": 9.240188621843626e-07,
+      "loss": 0.918,
+      "step": 5075
+    },
+    {
+      "epoch": 0.12234772765588497,
+      "grad_norm": 1.5147145986557007,
+      "learning_rate": 9.238921001926782e-07,
+      "loss": 0.9762,
+      "step": 5080
+    },
+    {
+      "epoch": 0.12246814864767226,
+      "grad_norm": 1.45526123046875,
+      "learning_rate": 9.237653382009938e-07,
+      "loss": 0.9586,
+      "step": 5085
+    },
+    {
+      "epoch": 0.12258856963945955,
+      "grad_norm": 1.5992141962051392,
+      "learning_rate": 9.236385762093094e-07,
+      "loss": 0.933,
+      "step": 5090
+    },
+    {
+      "epoch": 0.12270899063124684,
+      "grad_norm": 1.419451117515564,
+      "learning_rate": 9.235118142176249e-07,
+      "loss": 0.9963,
+      "step": 5095
+    },
+    {
+      "epoch": 0.12282941162303412,
+      "grad_norm": 1.6753170490264893,
+      "learning_rate": 9.233850522259405e-07,
+      "loss": 0.9369,
+      "step": 5100
+    },
+    {
+      "epoch": 0.12294983261482141,
+      "grad_norm": 1.876473307609558,
+      "learning_rate": 9.232582902342562e-07,
+      "loss": 0.9491,
+      "step": 5105
+    },
+    {
+      "epoch": 0.1230702536066087,
+      "grad_norm": 1.5998787879943848,
+      "learning_rate": 9.231315282425717e-07,
+      "loss": 0.9891,
+      "step": 5110
+    },
+    {
+      "epoch": 0.12319067459839599,
+      "grad_norm": 1.7187938690185547,
+      "learning_rate": 9.230047662508873e-07,
+      "loss": 0.9506,
+      "step": 5115
+    },
+    {
+      "epoch": 0.12331109559018329,
+      "grad_norm": 1.3768280744552612,
+      "learning_rate": 9.228780042592029e-07,
+      "loss": 0.9974,
+      "step": 5120
+    },
+    {
+      "epoch": 0.12343151658197057,
+      "grad_norm": 1.4900585412979126,
+      "learning_rate": 9.227512422675184e-07,
+      "loss": 0.9339,
+      "step": 5125
+    },
+    {
+      "epoch": 0.12355193757375786,
+      "grad_norm": 1.7107560634613037,
+      "learning_rate": 9.226244802758341e-07,
+      "loss": 0.9153,
+      "step": 5130
+    },
+    {
+      "epoch": 0.12367235856554515,
+      "grad_norm": 2.090656042098999,
+      "learning_rate": 9.224977182841496e-07,
+      "loss": 0.895,
+      "step": 5135
+    },
+    {
+      "epoch": 0.12379277955733244,
+      "grad_norm": 1.7037839889526367,
+      "learning_rate": 9.223709562924652e-07,
+      "loss": 1.0013,
+      "step": 5140
+    },
+    {
+      "epoch": 0.12391320054911972,
+      "grad_norm": 1.5398280620574951,
+      "learning_rate": 9.222441943007808e-07,
+      "loss": 0.9445,
+      "step": 5145
+    },
+    {
+      "epoch": 0.12403362154090701,
+      "grad_norm": 1.6887266635894775,
+      "learning_rate": 9.221174323090965e-07,
+      "loss": 0.9767,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1241540425326943,
+      "grad_norm": 1.401342511177063,
+      "learning_rate": 9.21990670317412e-07,
+      "loss": 0.8865,
+      "step": 5155
+    },
+    {
+      "epoch": 0.12427446352448158,
+      "grad_norm": 1.3718047142028809,
+      "learning_rate": 9.218639083257275e-07,
+      "loss": 0.9621,
+      "step": 5160
+    },
+    {
+      "epoch": 0.12439488451626887,
+      "grad_norm": 1.6709517240524292,
+      "learning_rate": 9.217371463340432e-07,
+      "loss": 0.967,
+      "step": 5165
+    },
+    {
+      "epoch": 0.12451530550805616,
+      "grad_norm": 1.510709285736084,
+      "learning_rate": 9.216103843423587e-07,
+      "loss": 0.9674,
+      "step": 5170
+    },
+    {
+      "epoch": 0.12463572649984345,
+      "grad_norm": 1.467331886291504,
+      "learning_rate": 9.214836223506744e-07,
+      "loss": 0.9637,
+      "step": 5175
+    },
+    {
+      "epoch": 0.12475614749163075,
+      "grad_norm": 1.5031909942626953,
+      "learning_rate": 9.2135686035899e-07,
+      "loss": 0.9485,
+      "step": 5180
+    },
+    {
+      "epoch": 0.12487656848341804,
+      "grad_norm": 1.759961485862732,
+      "learning_rate": 9.212300983673054e-07,
+      "loss": 1.0108,
+      "step": 5185
+    },
+    {
+      "epoch": 0.12499698947520532,
+      "grad_norm": 1.519850492477417,
+      "learning_rate": 9.211033363756211e-07,
+      "loss": 0.9442,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1251174104669926,
+      "grad_norm": 1.8199149370193481,
+      "learning_rate": 9.209765743839367e-07,
+      "loss": 0.9235,
+      "step": 5195
+    },
+    {
+      "epoch": 0.1252378314587799,
+      "grad_norm": 1.6225138902664185,
+      "learning_rate": 9.208498123922523e-07,
+      "loss": 0.9856,
+      "step": 5200
+    },
+    {
+      "epoch": 0.12535825245056717,
+      "grad_norm": 1.6960610151290894,
+      "learning_rate": 9.207230504005678e-07,
+      "loss": 1.0328,
+      "step": 5205
+    },
+    {
+      "epoch": 0.12547867344235447,
+      "grad_norm": 1.9315990209579468,
+      "learning_rate": 9.205962884088835e-07,
+      "loss": 0.9509,
+      "step": 5210
+    },
+    {
+      "epoch": 0.12559909443414177,
+      "grad_norm": 1.6579340696334839,
+      "learning_rate": 9.20469526417199e-07,
+      "loss": 0.9352,
+      "step": 5215
+    },
+    {
+      "epoch": 0.12571951542592905,
+      "grad_norm": 1.6070151329040527,
+      "learning_rate": 9.203427644255146e-07,
+      "loss": 0.9809,
+      "step": 5220
+    },
+    {
+      "epoch": 0.12583993641771635,
+      "grad_norm": 1.4353399276733398,
+      "learning_rate": 9.202160024338303e-07,
+      "loss": 0.9512,
+      "step": 5225
+    },
+    {
+      "epoch": 0.12596035740950362,
+      "grad_norm": 1.5951476097106934,
+      "learning_rate": 9.200892404421457e-07,
+      "loss": 0.9959,
+      "step": 5230
+    },
+    {
+      "epoch": 0.12608077840129092,
+      "grad_norm": 1.600804090499878,
+      "learning_rate": 9.199624784504614e-07,
+      "loss": 0.9749,
+      "step": 5235
+    },
+    {
+      "epoch": 0.1262011993930782,
+      "grad_norm": 1.6272754669189453,
+      "learning_rate": 9.19835716458777e-07,
+      "loss": 0.9287,
+      "step": 5240
+    },
+    {
+      "epoch": 0.1263216203848655,
+      "grad_norm": 1.556373953819275,
+      "learning_rate": 9.197089544670925e-07,
+      "loss": 0.9435,
+      "step": 5245
+    },
+    {
+      "epoch": 0.12644204137665277,
+      "grad_norm": 1.7332128286361694,
+      "learning_rate": 9.195821924754081e-07,
+      "loss": 0.9235,
+      "step": 5250
+    },
+    {
+      "epoch": 0.12656246236844007,
+      "grad_norm": 1.6399269104003906,
+      "learning_rate": 9.194554304837237e-07,
+      "loss": 0.9436,
+      "step": 5255
+    },
+    {
+      "epoch": 0.12668288336022734,
+      "grad_norm": 1.756186842918396,
+      "learning_rate": 9.193286684920393e-07,
+      "loss": 0.9137,
+      "step": 5260
+    },
+    {
+      "epoch": 0.12680330435201465,
+      "grad_norm": 1.9386248588562012,
+      "learning_rate": 9.192019065003549e-07,
+      "loss": 0.9831,
+      "step": 5265
+    },
+    {
+      "epoch": 0.12692372534380192,
+      "grad_norm": 1.8038384914398193,
+      "learning_rate": 9.190751445086706e-07,
+      "loss": 0.8885,
+      "step": 5270
+    },
+    {
+      "epoch": 0.12704414633558922,
+      "grad_norm": 1.5861470699310303,
+      "learning_rate": 9.18948382516986e-07,
+      "loss": 0.9518,
+      "step": 5275
+    },
+    {
+      "epoch": 0.12716456732737652,
+      "grad_norm": 1.5452990531921387,
+      "learning_rate": 9.188216205253016e-07,
+      "loss": 1.0352,
+      "step": 5280
+    },
+    {
+      "epoch": 0.1272849883191638,
+      "grad_norm": 1.6214449405670166,
+      "learning_rate": 9.186948585336173e-07,
+      "loss": 1.0022,
+      "step": 5285
+    },
+    {
+      "epoch": 0.1274054093109511,
+      "grad_norm": 1.5629680156707764,
+      "learning_rate": 9.185680965419328e-07,
+      "loss": 0.9569,
+      "step": 5290
+    },
+    {
+      "epoch": 0.12752583030273837,
+      "grad_norm": 1.5142781734466553,
+      "learning_rate": 9.184413345502485e-07,
+      "loss": 0.9449,
+      "step": 5295
+    },
+    {
+      "epoch": 0.12764625129452567,
+      "grad_norm": 1.5961127281188965,
+      "learning_rate": 9.18314572558564e-07,
+      "loss": 0.9837,
+      "step": 5300
+    },
+    {
+      "epoch": 0.12776667228631294,
+      "grad_norm": 1.583109974861145,
+      "learning_rate": 9.181878105668795e-07,
+      "loss": 0.9623,
+      "step": 5305
+    },
+    {
+      "epoch": 0.12788709327810024,
+      "grad_norm": 1.740462303161621,
+      "learning_rate": 9.180610485751952e-07,
+      "loss": 0.9572,
+      "step": 5310
+    },
+    {
+      "epoch": 0.12800751426988752,
+      "grad_norm": 1.6290134191513062,
+      "learning_rate": 9.179342865835108e-07,
+      "loss": 0.9361,
+      "step": 5315
+    },
+    {
+      "epoch": 0.12812793526167482,
+      "grad_norm": 1.3808825016021729,
+      "learning_rate": 9.178075245918263e-07,
+      "loss": 0.9961,
+      "step": 5320
+    },
+    {
+      "epoch": 0.1282483562534621,
+      "grad_norm": 1.8814772367477417,
+      "learning_rate": 9.176807626001419e-07,
+      "loss": 0.9583,
+      "step": 5325
+    },
+    {
+      "epoch": 0.1283687772452494,
+      "grad_norm": 1.5854055881500244,
+      "learning_rate": 9.175540006084576e-07,
+      "loss": 0.9677,
+      "step": 5330
+    },
+    {
+      "epoch": 0.12848919823703667,
+      "grad_norm": 1.688438892364502,
+      "learning_rate": 9.174272386167731e-07,
+      "loss": 0.9858,
+      "step": 5335
+    },
+    {
+      "epoch": 0.12860961922882397,
+      "grad_norm": 1.4895740747451782,
+      "learning_rate": 9.173004766250887e-07,
+      "loss": 0.9622,
+      "step": 5340
+    },
+    {
+      "epoch": 0.12873004022061127,
+      "grad_norm": 1.4580119848251343,
+      "learning_rate": 9.171737146334043e-07,
+      "loss": 0.9619,
+      "step": 5345
+    },
+    {
+      "epoch": 0.12885046121239854,
+      "grad_norm": 1.5655674934387207,
+      "learning_rate": 9.170469526417198e-07,
+      "loss": 0.9412,
+      "step": 5350
+    },
+    {
+      "epoch": 0.12897088220418584,
+      "grad_norm": 1.5745370388031006,
+      "learning_rate": 9.169201906500355e-07,
+      "loss": 0.9373,
+      "step": 5355
+    },
+    {
+      "epoch": 0.12909130319597312,
+      "grad_norm": 1.4132782220840454,
+      "learning_rate": 9.167934286583511e-07,
+      "loss": 0.9671,
+      "step": 5360
+    },
+    {
+      "epoch": 0.12921172418776042,
+      "grad_norm": 1.5725817680358887,
+      "learning_rate": 9.166666666666665e-07,
+      "loss": 0.9964,
+      "step": 5365
+    },
+    {
+      "epoch": 0.1293321451795477,
+      "grad_norm": 1.7641382217407227,
+      "learning_rate": 9.165399046749822e-07,
+      "loss": 0.9692,
+      "step": 5370
+    },
+    {
+      "epoch": 0.129452566171335,
+      "grad_norm": 1.7528362274169922,
+      "learning_rate": 9.164131426832978e-07,
+      "loss": 0.938,
+      "step": 5375
+    },
+    {
+      "epoch": 0.12957298716312227,
+      "grad_norm": 1.7025678157806396,
+      "learning_rate": 9.162863806916134e-07,
+      "loss": 0.9336,
+      "step": 5380
+    },
+    {
+      "epoch": 0.12969340815490957,
+      "grad_norm": 1.630113959312439,
+      "learning_rate": 9.16159618699929e-07,
+      "loss": 1.0092,
+      "step": 5385
+    },
+    {
+      "epoch": 0.12981382914669684,
+      "grad_norm": 1.58848237991333,
+      "learning_rate": 9.160328567082445e-07,
+      "loss": 0.9254,
+      "step": 5390
+    },
+    {
+      "epoch": 0.12993425013848414,
+      "grad_norm": 1.4365260601043701,
+      "learning_rate": 9.159060947165601e-07,
+      "loss": 0.9769,
+      "step": 5395
+    },
+    {
+      "epoch": 0.13005467113027142,
+      "grad_norm": 1.4236555099487305,
+      "learning_rate": 9.157793327248757e-07,
+      "loss": 0.9208,
+      "step": 5400
+    },
+    {
+      "epoch": 0.13017509212205872,
+      "grad_norm": 1.6881691217422485,
+      "learning_rate": 9.156525707331914e-07,
+      "loss": 0.9772,
+      "step": 5405
+    },
+    {
+      "epoch": 0.13029551311384602,
+      "grad_norm": 1.568904995918274,
+      "learning_rate": 9.155258087415069e-07,
+      "loss": 0.9121,
+      "step": 5410
+    },
+    {
+      "epoch": 0.1304159341056333,
+      "grad_norm": 1.721828818321228,
+      "learning_rate": 9.153990467498225e-07,
+      "loss": 0.9717,
+      "step": 5415
+    },
+    {
+      "epoch": 0.1305363550974206,
+      "grad_norm": 1.6432464122772217,
+      "learning_rate": 9.152722847581381e-07,
+      "loss": 0.9182,
+      "step": 5420
+    },
+    {
+      "epoch": 0.13065677608920787,
+      "grad_norm": 1.5352879762649536,
+      "learning_rate": 9.151455227664536e-07,
+      "loss": 0.9659,
+      "step": 5425
+    },
+    {
+      "epoch": 0.13077719708099517,
+      "grad_norm": 1.8074719905853271,
+      "learning_rate": 9.150187607747693e-07,
+      "loss": 0.9903,
+      "step": 5430
+    },
+    {
+      "epoch": 0.13089761807278244,
+      "grad_norm": 1.5905964374542236,
+      "learning_rate": 9.148919987830848e-07,
+      "loss": 0.9566,
+      "step": 5435
+    },
+    {
+      "epoch": 0.13101803906456974,
+      "grad_norm": 1.4441380500793457,
+      "learning_rate": 9.147652367914004e-07,
+      "loss": 0.9314,
+      "step": 5440
+    },
+    {
+      "epoch": 0.13113846005635701,
+      "grad_norm": 1.55857253074646,
+      "learning_rate": 9.14638474799716e-07,
+      "loss": 0.9121,
+      "step": 5445
+    },
+    {
+      "epoch": 0.13125888104814432,
+      "grad_norm": 1.6368721723556519,
+      "learning_rate": 9.145117128080317e-07,
+      "loss": 0.9487,
+      "step": 5450
+    },
+    {
+      "epoch": 0.1313793020399316,
+      "grad_norm": 1.7417676448822021,
+      "learning_rate": 9.143849508163472e-07,
+      "loss": 0.9368,
+      "step": 5455
+    },
+    {
+      "epoch": 0.1314997230317189,
+      "grad_norm": 1.3386549949645996,
+      "learning_rate": 9.142581888246627e-07,
+      "loss": 0.9817,
+      "step": 5460
+    },
+    {
+      "epoch": 0.1316201440235062,
+      "grad_norm": 1.568506121635437,
+      "learning_rate": 9.141314268329784e-07,
+      "loss": 0.9796,
+      "step": 5465
+    },
+    {
+      "epoch": 0.13174056501529346,
+      "grad_norm": 1.5980567932128906,
+      "learning_rate": 9.140046648412939e-07,
+      "loss": 0.9429,
+      "step": 5470
+    },
+    {
+      "epoch": 0.13186098600708077,
+      "grad_norm": 1.4643741846084595,
+      "learning_rate": 9.138779028496096e-07,
+      "loss": 0.9005,
+      "step": 5475
+    },
+    {
+      "epoch": 0.13198140699886804,
+      "grad_norm": 1.727617621421814,
+      "learning_rate": 9.137511408579252e-07,
+      "loss": 1.002,
+      "step": 5480
+    },
+    {
+      "epoch": 0.13210182799065534,
+      "grad_norm": 1.4297009706497192,
+      "learning_rate": 9.136243788662406e-07,
+      "loss": 0.9769,
+      "step": 5485
+    },
+    {
+      "epoch": 0.13222224898244261,
+      "grad_norm": 1.4668093919754028,
+      "learning_rate": 9.134976168745563e-07,
+      "loss": 1.007,
+      "step": 5490
+    },
+    {
+      "epoch": 0.13234266997422992,
+      "grad_norm": 1.4713921546936035,
+      "learning_rate": 9.133708548828719e-07,
+      "loss": 0.923,
+      "step": 5495
+    },
+    {
+      "epoch": 0.1324630909660172,
+      "grad_norm": 1.6667219400405884,
+      "learning_rate": 9.132440928911875e-07,
+      "loss": 0.9222,
+      "step": 5500
+    },
+    {
+      "epoch": 0.1325835119578045,
+      "grad_norm": 1.664129376411438,
+      "learning_rate": 9.13117330899503e-07,
+      "loss": 0.9041,
+      "step": 5505
+    },
+    {
+      "epoch": 0.13270393294959176,
+      "grad_norm": 1.6742534637451172,
+      "learning_rate": 9.129905689078186e-07,
+      "loss": 0.9022,
+      "step": 5510
+    },
+    {
+      "epoch": 0.13282435394137906,
+      "grad_norm": 1.4654879570007324,
+      "learning_rate": 9.128638069161342e-07,
+      "loss": 0.95,
+      "step": 5515
+    },
+    {
+      "epoch": 0.13294477493316634,
+      "grad_norm": 1.748233675956726,
+      "learning_rate": 9.127370449244498e-07,
+      "loss": 0.957,
+      "step": 5520
+    },
+    {
+      "epoch": 0.13306519592495364,
+      "grad_norm": 1.5935348272323608,
+      "learning_rate": 9.126102829327655e-07,
+      "loss": 0.9468,
+      "step": 5525
+    },
+    {
+      "epoch": 0.13318561691674094,
+      "grad_norm": 1.6781808137893677,
+      "learning_rate": 9.124835209410809e-07,
+      "loss": 0.9734,
+      "step": 5530
+    },
+    {
+      "epoch": 0.1333060379085282,
+      "grad_norm": 1.6739495992660522,
+      "learning_rate": 9.123567589493966e-07,
+      "loss": 0.9759,
+      "step": 5535
+    },
+    {
+      "epoch": 0.13342645890031551,
+      "grad_norm": 1.5237581729888916,
+      "learning_rate": 9.122299969577122e-07,
+      "loss": 0.9928,
+      "step": 5540
+    },
+    {
+      "epoch": 0.1335468798921028,
+      "grad_norm": 1.8118046522140503,
+      "learning_rate": 9.121032349660277e-07,
+      "loss": 0.9387,
+      "step": 5545
+    },
+    {
+      "epoch": 0.1336673008838901,
+      "grad_norm": 1.53839910030365,
+      "learning_rate": 9.119764729743433e-07,
+      "loss": 0.9461,
+      "step": 5550
+    },
+    {
+      "epoch": 0.13378772187567736,
+      "grad_norm": 1.5928133726119995,
+      "learning_rate": 9.118497109826589e-07,
+      "loss": 0.9554,
+      "step": 5555
+    },
+    {
+      "epoch": 0.13390814286746466,
+      "grad_norm": 1.980698585510254,
+      "learning_rate": 9.117229489909745e-07,
+      "loss": 0.9315,
+      "step": 5560
+    },
+    {
+      "epoch": 0.13402856385925194,
+      "grad_norm": 1.5155107975006104,
+      "learning_rate": 9.115961869992901e-07,
+      "loss": 0.9696,
+      "step": 5565
+    },
+    {
+      "epoch": 0.13414898485103924,
+      "grad_norm": 1.5934728384017944,
+      "learning_rate": 9.114694250076057e-07,
+      "loss": 0.9984,
+      "step": 5570
+    },
+    {
+      "epoch": 0.1342694058428265,
+      "grad_norm": 1.5468522310256958,
+      "learning_rate": 9.113426630159212e-07,
+      "loss": 0.931,
+      "step": 5575
+    },
+    {
+      "epoch": 0.1343898268346138,
+      "grad_norm": 1.7637958526611328,
+      "learning_rate": 9.112159010242368e-07,
+      "loss": 0.9437,
+      "step": 5580
+    },
+    {
+      "epoch": 0.13451024782640109,
+      "grad_norm": 1.5114047527313232,
+      "learning_rate": 9.110891390325525e-07,
+      "loss": 0.9994,
+      "step": 5585
+    },
+    {
+      "epoch": 0.1346306688181884,
+      "grad_norm": 1.611918330192566,
+      "learning_rate": 9.10962377040868e-07,
+      "loss": 0.9207,
+      "step": 5590
+    },
+    {
+      "epoch": 0.1347510898099757,
+      "grad_norm": 1.6683274507522583,
+      "learning_rate": 9.108356150491837e-07,
+      "loss": 0.9548,
+      "step": 5595
+    },
+    {
+      "epoch": 0.13487151080176296,
+      "grad_norm": 1.4041019678115845,
+      "learning_rate": 9.107088530574992e-07,
+      "loss": 0.9524,
+      "step": 5600
+    },
+    {
+      "epoch": 0.13499193179355026,
+      "grad_norm": 1.4725638628005981,
+      "learning_rate": 9.105820910658147e-07,
+      "loss": 0.9636,
+      "step": 5605
+    },
+    {
+      "epoch": 0.13511235278533754,
+      "grad_norm": 1.5536869764328003,
+      "learning_rate": 9.104553290741304e-07,
+      "loss": 0.9784,
+      "step": 5610
+    },
+    {
+      "epoch": 0.13523277377712484,
+      "grad_norm": 1.4738366603851318,
+      "learning_rate": 9.10328567082446e-07,
+      "loss": 0.9641,
+      "step": 5615
+    },
+    {
+      "epoch": 0.1353531947689121,
+      "grad_norm": 1.5258055925369263,
+      "learning_rate": 9.102018050907615e-07,
+      "loss": 0.9735,
+      "step": 5620
+    },
+    {
+      "epoch": 0.1354736157606994,
+      "grad_norm": 1.5292253494262695,
+      "learning_rate": 9.100750430990771e-07,
+      "loss": 0.9347,
+      "step": 5625
+    },
+    {
+      "epoch": 0.13559403675248669,
+      "grad_norm": 1.5491033792495728,
+      "learning_rate": 9.099482811073927e-07,
+      "loss": 1.0053,
+      "step": 5630
+    },
+    {
+      "epoch": 0.135714457744274,
+      "grad_norm": 1.5524201393127441,
+      "learning_rate": 9.098215191157083e-07,
+      "loss": 0.9651,
+      "step": 5635
+    },
+    {
+      "epoch": 0.13583487873606126,
+      "grad_norm": 1.6256749629974365,
+      "learning_rate": 9.096947571240239e-07,
+      "loss": 0.9522,
+      "step": 5640
+    },
+    {
+      "epoch": 0.13595529972784856,
+      "grad_norm": 1.6116081476211548,
+      "learning_rate": 9.095679951323395e-07,
+      "loss": 0.9779,
+      "step": 5645
+    },
+    {
+      "epoch": 0.13607572071963583,
+      "grad_norm": 1.7412595748901367,
+      "learning_rate": 9.09441233140655e-07,
+      "loss": 0.9511,
+      "step": 5650
+    },
+    {
+      "epoch": 0.13619614171142314,
+      "grad_norm": 1.6832035779953003,
+      "learning_rate": 9.093144711489706e-07,
+      "loss": 0.9324,
+      "step": 5655
+    },
+    {
+      "epoch": 0.13631656270321044,
+      "grad_norm": 1.6584666967391968,
+      "learning_rate": 9.091877091572863e-07,
+      "loss": 0.9868,
+      "step": 5660
+    },
+    {
+      "epoch": 0.1364369836949977,
+      "grad_norm": 1.8463393449783325,
+      "learning_rate": 9.090609471656017e-07,
+      "loss": 0.989,
+      "step": 5665
+    },
+    {
+      "epoch": 0.136557404686785,
+      "grad_norm": 1.6989035606384277,
+      "learning_rate": 9.089341851739174e-07,
+      "loss": 0.9642,
+      "step": 5670
+    },
+    {
+      "epoch": 0.13667782567857228,
+      "grad_norm": 1.5193052291870117,
+      "learning_rate": 9.08807423182233e-07,
+      "loss": 0.9344,
+      "step": 5675
+    },
+    {
+      "epoch": 0.13679824667035959,
+      "grad_norm": 1.4853129386901855,
+      "learning_rate": 9.086806611905486e-07,
+      "loss": 0.9027,
+      "step": 5680
+    },
+    {
+      "epoch": 0.13691866766214686,
+      "grad_norm": 1.9148509502410889,
+      "learning_rate": 9.085538991988642e-07,
+      "loss": 0.9396,
+      "step": 5685
+    },
+    {
+      "epoch": 0.13703908865393416,
+      "grad_norm": 1.6306005716323853,
+      "learning_rate": 9.084271372071797e-07,
+      "loss": 0.9338,
+      "step": 5690
+    },
+    {
+      "epoch": 0.13715950964572143,
+      "grad_norm": 1.5270458459854126,
+      "learning_rate": 9.083003752154953e-07,
+      "loss": 0.8957,
+      "step": 5695
+    },
+    {
+      "epoch": 0.13727993063750873,
+      "grad_norm": 1.4417778253555298,
+      "learning_rate": 9.081736132238109e-07,
+      "loss": 0.9709,
+      "step": 5700
+    },
+    {
+      "epoch": 0.137400351629296,
+      "grad_norm": 1.6104899644851685,
+      "learning_rate": 9.080468512321266e-07,
+      "loss": 0.9205,
+      "step": 5705
+    },
+    {
+      "epoch": 0.1375207726210833,
+      "grad_norm": 1.6175036430358887,
+      "learning_rate": 9.079200892404421e-07,
+      "loss": 0.9469,
+      "step": 5710
+    },
+    {
+      "epoch": 0.13764119361287058,
+      "grad_norm": 1.5673390626907349,
+      "learning_rate": 9.077933272487576e-07,
+      "loss": 0.9618,
+      "step": 5715
+    },
+    {
+      "epoch": 0.13776161460465788,
+      "grad_norm": 1.7446295022964478,
+      "learning_rate": 9.076665652570733e-07,
+      "loss": 0.925,
+      "step": 5720
+    },
+    {
+      "epoch": 0.13788203559644518,
+      "grad_norm": 1.4070531129837036,
+      "learning_rate": 9.075398032653888e-07,
+      "loss": 0.9859,
+      "step": 5725
+    },
+    {
+      "epoch": 0.13800245658823246,
+      "grad_norm": 1.5243618488311768,
+      "learning_rate": 9.074130412737045e-07,
+      "loss": 0.9412,
+      "step": 5730
+    },
+    {
+      "epoch": 0.13812287758001976,
+      "grad_norm": 1.5630604028701782,
+      "learning_rate": 9.0728627928202e-07,
+      "loss": 0.9375,
+      "step": 5735
+    },
+    {
+      "epoch": 0.13824329857180703,
+      "grad_norm": 1.7350032329559326,
+      "learning_rate": 9.071595172903356e-07,
+      "loss": 0.9522,
+      "step": 5740
+    },
+    {
+      "epoch": 0.13836371956359433,
+      "grad_norm": 1.8089555501937866,
+      "learning_rate": 9.070327552986512e-07,
+      "loss": 0.9314,
+      "step": 5745
+    },
+    {
+      "epoch": 0.1384841405553816,
+      "grad_norm": 1.511834740638733,
+      "learning_rate": 9.069059933069668e-07,
+      "loss": 1.0104,
+      "step": 5750
+    },
+    {
+      "epoch": 0.1386045615471689,
+      "grad_norm": 1.5103285312652588,
+      "learning_rate": 9.067792313152824e-07,
+      "loss": 0.9809,
+      "step": 5755
+    },
+    {
+      "epoch": 0.13872498253895618,
+      "grad_norm": 1.5690333843231201,
+      "learning_rate": 9.066524693235979e-07,
+      "loss": 0.996,
+      "step": 5760
+    },
+    {
+      "epoch": 0.13884540353074348,
+      "grad_norm": 1.577641248703003,
+      "learning_rate": 9.065257073319136e-07,
+      "loss": 0.9487,
+      "step": 5765
+    },
+    {
+      "epoch": 0.13896582452253076,
+      "grad_norm": 1.4639778137207031,
+      "learning_rate": 9.063989453402291e-07,
+      "loss": 0.971,
+      "step": 5770
+    },
+    {
+      "epoch": 0.13908624551431806,
+      "grad_norm": 1.6108697652816772,
+      "learning_rate": 9.062721833485447e-07,
+      "loss": 0.9708,
+      "step": 5775
+    },
+    {
+      "epoch": 0.13920666650610533,
+      "grad_norm": 1.603767991065979,
+      "learning_rate": 9.061454213568604e-07,
+      "loss": 0.8838,
+      "step": 5780
+    },
+    {
+      "epoch": 0.13932708749789263,
+      "grad_norm": 1.5478912591934204,
+      "learning_rate": 9.060186593651758e-07,
+      "loss": 0.9728,
+      "step": 5785
+    },
+    {
+      "epoch": 0.13944750848967993,
+      "grad_norm": 1.4290289878845215,
+      "learning_rate": 9.058918973734915e-07,
+      "loss": 0.863,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1395679294814672,
+      "grad_norm": 1.6638075113296509,
+      "learning_rate": 9.057651353818071e-07,
+      "loss": 0.9495,
+      "step": 5795
+    },
+    {
+      "epoch": 0.1396883504732545,
+      "grad_norm": 1.6155606508255005,
+      "learning_rate": 9.056383733901227e-07,
+      "loss": 0.9387,
+      "step": 5800
+    },
+    {
+      "epoch": 0.13980877146504178,
+      "grad_norm": 1.5489352941513062,
+      "learning_rate": 9.055116113984382e-07,
+      "loss": 0.9324,
+      "step": 5805
+    },
+    {
+      "epoch": 0.13992919245682908,
+      "grad_norm": 1.4846972227096558,
+      "learning_rate": 9.053848494067538e-07,
+      "loss": 0.9736,
+      "step": 5810
+    },
+    {
+      "epoch": 0.14004961344861636,
+      "grad_norm": 1.772792935371399,
+      "learning_rate": 9.052580874150694e-07,
+      "loss": 0.9563,
+      "step": 5815
+    },
+    {
+      "epoch": 0.14017003444040366,
+      "grad_norm": 1.4382880926132202,
+      "learning_rate": 9.05131325423385e-07,
+      "loss": 0.9554,
+      "step": 5820
+    },
+    {
+      "epoch": 0.14029045543219093,
+      "grad_norm": 1.7088966369628906,
+      "learning_rate": 9.050045634317007e-07,
+      "loss": 1.0169,
+      "step": 5825
+    },
+    {
+      "epoch": 0.14041087642397823,
+      "grad_norm": 1.3709522485733032,
+      "learning_rate": 9.048778014400161e-07,
+      "loss": 0.9589,
+      "step": 5830
+    },
+    {
+      "epoch": 0.1405312974157655,
+      "grad_norm": 1.8668538331985474,
+      "learning_rate": 9.047510394483317e-07,
+      "loss": 0.9706,
+      "step": 5835
+    },
+    {
+      "epoch": 0.1406517184075528,
+      "grad_norm": 1.7845852375030518,
+      "learning_rate": 9.046242774566474e-07,
+      "loss": 0.9407,
+      "step": 5840
+    },
+    {
+      "epoch": 0.14077213939934008,
+      "grad_norm": 1.5803017616271973,
+      "learning_rate": 9.044975154649629e-07,
+      "loss": 0.9663,
+      "step": 5845
+    },
+    {
+      "epoch": 0.14089256039112738,
+      "grad_norm": 1.6961511373519897,
+      "learning_rate": 9.043707534732785e-07,
+      "loss": 0.9783,
+      "step": 5850
+    },
+    {
+      "epoch": 0.14101298138291468,
+      "grad_norm": 1.4206011295318604,
+      "learning_rate": 9.042439914815941e-07,
+      "loss": 0.9357,
+      "step": 5855
+    },
+    {
+      "epoch": 0.14113340237470196,
+      "grad_norm": 1.9127111434936523,
+      "learning_rate": 9.041172294899096e-07,
+      "loss": 0.9295,
+      "step": 5860
+    },
+    {
+      "epoch": 0.14125382336648926,
+      "grad_norm": 1.4716366529464722,
+      "learning_rate": 9.039904674982253e-07,
+      "loss": 0.9509,
+      "step": 5865
+    },
+    {
+      "epoch": 0.14137424435827653,
+      "grad_norm": 1.47152841091156,
+      "learning_rate": 9.038637055065409e-07,
+      "loss": 0.9803,
+      "step": 5870
+    },
+    {
+      "epoch": 0.14149466535006383,
+      "grad_norm": 1.5282686948776245,
+      "learning_rate": 9.037369435148564e-07,
+      "loss": 0.9496,
+      "step": 5875
+    },
+    {
+      "epoch": 0.1416150863418511,
+      "grad_norm": 1.6607931852340698,
+      "learning_rate": 9.03610181523172e-07,
+      "loss": 0.9748,
+      "step": 5880
+    },
+    {
+      "epoch": 0.1417355073336384,
+      "grad_norm": 1.5049351453781128,
+      "learning_rate": 9.034834195314877e-07,
+      "loss": 0.9323,
+      "step": 5885
+    },
+    {
+      "epoch": 0.14185592832542568,
+      "grad_norm": 1.4687178134918213,
+      "learning_rate": 9.033566575398033e-07,
+      "loss": 0.953,
+      "step": 5890
+    },
+    {
+      "epoch": 0.14197634931721298,
+      "grad_norm": 1.6209310293197632,
+      "learning_rate": 9.032298955481188e-07,
+      "loss": 0.927,
+      "step": 5895
+    },
+    {
+      "epoch": 0.14209677030900025,
+      "grad_norm": 1.742845892906189,
+      "learning_rate": 9.031031335564344e-07,
+      "loss": 0.9633,
+      "step": 5900
+    },
+    {
+      "epoch": 0.14221719130078755,
+      "grad_norm": 1.5639408826828003,
+      "learning_rate": 9.0297637156475e-07,
+      "loss": 0.9433,
+      "step": 5905
+    },
+    {
+      "epoch": 0.14233761229257486,
+      "grad_norm": 1.7477506399154663,
+      "learning_rate": 9.028496095730656e-07,
+      "loss": 0.9144,
+      "step": 5910
+    },
+    {
+      "epoch": 0.14245803328436213,
+      "grad_norm": 1.8129380941390991,
+      "learning_rate": 9.027228475813812e-07,
+      "loss": 0.9531,
+      "step": 5915
+    },
+    {
+      "epoch": 0.14257845427614943,
+      "grad_norm": 1.5080280303955078,
+      "learning_rate": 9.025960855896967e-07,
+      "loss": 0.987,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1426988752679367,
+      "grad_norm": 1.4849168062210083,
+      "learning_rate": 9.024693235980123e-07,
+      "loss": 0.9883,
+      "step": 5925
+    },
+    {
+      "epoch": 0.142819296259724,
+      "grad_norm": 1.9817742109298706,
+      "learning_rate": 9.023425616063279e-07,
+      "loss": 0.9211,
+      "step": 5930
+    },
+    {
+      "epoch": 0.14293971725151128,
+      "grad_norm": 1.593101143836975,
+      "learning_rate": 9.022157996146436e-07,
+      "loss": 0.9607,
+      "step": 5935
+    },
+    {
+      "epoch": 0.14306013824329858,
+      "grad_norm": 2.151980400085449,
+      "learning_rate": 9.020890376229591e-07,
+      "loss": 0.9711,
+      "step": 5940
+    },
+    {
+      "epoch": 0.14318055923508585,
+      "grad_norm": 1.5875486135482788,
+      "learning_rate": 9.019622756312747e-07,
+      "loss": 0.9618,
+      "step": 5945
+    },
+    {
+      "epoch": 0.14330098022687315,
+      "grad_norm": 1.7066106796264648,
+      "learning_rate": 9.018355136395903e-07,
+      "loss": 0.9255,
+      "step": 5950
+    },
+    {
+      "epoch": 0.14342140121866043,
+      "grad_norm": 1.424856424331665,
+      "learning_rate": 9.017087516479058e-07,
+      "loss": 0.9177,
+      "step": 5955
+    },
+    {
+      "epoch": 0.14354182221044773,
+      "grad_norm": 1.345221996307373,
+      "learning_rate": 9.015819896562215e-07,
+      "loss": 0.9523,
+      "step": 5960
+    },
+    {
+      "epoch": 0.143662243202235,
+      "grad_norm": 1.4497827291488647,
+      "learning_rate": 9.014552276645371e-07,
+      "loss": 0.9444,
+      "step": 5965
+    },
+    {
+      "epoch": 0.1437826641940223,
+      "grad_norm": 1.830645203590393,
+      "learning_rate": 9.013284656728526e-07,
+      "loss": 0.9451,
+      "step": 5970
+    },
+    {
+      "epoch": 0.1439030851858096,
+      "grad_norm": 1.7493947744369507,
+      "learning_rate": 9.012017036811682e-07,
+      "loss": 0.9562,
+      "step": 5975
+    },
+    {
+      "epoch": 0.14402350617759688,
+      "grad_norm": 1.5937539339065552,
+      "learning_rate": 9.010749416894839e-07,
+      "loss": 1.0019,
+      "step": 5980
+    },
+    {
+      "epoch": 0.14414392716938418,
+      "grad_norm": 1.4270418882369995,
+      "learning_rate": 9.009481796977994e-07,
+      "loss": 0.9903,
+      "step": 5985
+    },
+    {
+      "epoch": 0.14426434816117145,
+      "grad_norm": 1.5541683435440063,
+      "learning_rate": 9.008214177061149e-07,
+      "loss": 0.9484,
+      "step": 5990
+    },
+    {
+      "epoch": 0.14438476915295875,
+      "grad_norm": 1.7125719785690308,
+      "learning_rate": 9.006946557144306e-07,
+      "loss": 0.9652,
+      "step": 5995
+    },
+    {
+      "epoch": 0.14450519014474603,
+      "grad_norm": 1.6315988302230835,
+      "learning_rate": 9.005678937227461e-07,
+      "loss": 0.9524,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14462561113653333,
+      "grad_norm": 1.493573546409607,
+      "learning_rate": 9.004411317310618e-07,
+      "loss": 0.8963,
+      "step": 6005
+    },
+    {
+      "epoch": 0.1447460321283206,
+      "grad_norm": 1.3541276454925537,
+      "learning_rate": 9.003143697393774e-07,
+      "loss": 1.0076,
+      "step": 6010
+    },
+    {
+      "epoch": 0.1448664531201079,
+      "grad_norm": 1.7591745853424072,
+      "learning_rate": 9.001876077476928e-07,
+      "loss": 0.9745,
+      "step": 6015
+    },
+    {
+      "epoch": 0.14498687411189518,
+      "grad_norm": 1.5763760805130005,
+      "learning_rate": 9.000608457560085e-07,
+      "loss": 0.972,
+      "step": 6020
+    },
+    {
+      "epoch": 0.14510729510368248,
+      "grad_norm": 1.6166871786117554,
+      "learning_rate": 8.999340837643241e-07,
+      "loss": 0.9752,
+      "step": 6025
+    },
+    {
+      "epoch": 0.14522771609546975,
+      "grad_norm": 1.7187212705612183,
+      "learning_rate": 8.998073217726397e-07,
+      "loss": 0.9813,
+      "step": 6030
+    },
+    {
+      "epoch": 0.14534813708725705,
+      "grad_norm": 1.5862038135528564,
+      "learning_rate": 8.996805597809552e-07,
+      "loss": 0.9084,
+      "step": 6035
+    },
+    {
+      "epoch": 0.14546855807904435,
+      "grad_norm": 1.7693798542022705,
+      "learning_rate": 8.995537977892708e-07,
+      "loss": 0.8757,
+      "step": 6040
+    },
+    {
+      "epoch": 0.14558897907083163,
+      "grad_norm": 1.809888482093811,
+      "learning_rate": 8.994270357975864e-07,
+      "loss": 0.937,
+      "step": 6045
+    },
+    {
+      "epoch": 0.14570940006261893,
+      "grad_norm": 1.5466762781143188,
+      "learning_rate": 8.99300273805902e-07,
+      "loss": 0.9364,
+      "step": 6050
+    },
+    {
+      "epoch": 0.1458298210544062,
+      "grad_norm": 1.6745027303695679,
+      "learning_rate": 8.991735118142177e-07,
+      "loss": 0.9831,
+      "step": 6055
+    },
+    {
+      "epoch": 0.1459502420461935,
+      "grad_norm": 1.58736252784729,
+      "learning_rate": 8.990467498225331e-07,
+      "loss": 0.971,
+      "step": 6060
+    },
+    {
+      "epoch": 0.14607066303798077,
+      "grad_norm": 1.7206737995147705,
+      "learning_rate": 8.989199878308488e-07,
+      "loss": 0.9727,
+      "step": 6065
+    },
+    {
+      "epoch": 0.14619108402976808,
+      "grad_norm": 1.7877613306045532,
+      "learning_rate": 8.987932258391644e-07,
+      "loss": 0.9706,
+      "step": 6070
+    },
+    {
+      "epoch": 0.14631150502155535,
+      "grad_norm": 1.5204682350158691,
+      "learning_rate": 8.986664638474799e-07,
+      "loss": 0.8874,
+      "step": 6075
+    },
+    {
+      "epoch": 0.14643192601334265,
+      "grad_norm": 1.5084939002990723,
+      "learning_rate": 8.985397018557956e-07,
+      "loss": 0.9579,
+      "step": 6080
+    },
+    {
+      "epoch": 0.14655234700512992,
+      "grad_norm": 1.596987247467041,
+      "learning_rate": 8.984129398641111e-07,
+      "loss": 0.9857,
+      "step": 6085
+    },
+    {
+      "epoch": 0.14667276799691722,
+      "grad_norm": 1.6701748371124268,
+      "learning_rate": 8.982861778724267e-07,
+      "loss": 0.9154,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1467931889887045,
+      "grad_norm": 1.676477074623108,
+      "learning_rate": 8.981594158807423e-07,
+      "loss": 0.8887,
+      "step": 6095
+    },
+    {
+      "epoch": 0.1469136099804918,
+      "grad_norm": 1.4097827672958374,
+      "learning_rate": 8.98032653889058e-07,
+      "loss": 0.9449,
+      "step": 6100
+    },
+    {
+      "epoch": 0.1470340309722791,
+      "grad_norm": 1.5947428941726685,
+      "learning_rate": 8.979058918973734e-07,
+      "loss": 0.9222,
+      "step": 6105
+    },
+    {
+      "epoch": 0.14715445196406637,
+      "grad_norm": 1.4588384628295898,
+      "learning_rate": 8.97779129905689e-07,
+      "loss": 0.9933,
+      "step": 6110
+    },
+    {
+      "epoch": 0.14727487295585368,
+      "grad_norm": 1.5385102033615112,
+      "learning_rate": 8.976523679140047e-07,
+      "loss": 0.9369,
+      "step": 6115
+    },
+    {
+      "epoch": 0.14739529394764095,
+      "grad_norm": 1.8313229084014893,
+      "learning_rate": 8.975256059223202e-07,
+      "loss": 0.9278,
+      "step": 6120
+    },
+    {
+      "epoch": 0.14751571493942825,
+      "grad_norm": 1.6547642946243286,
+      "learning_rate": 8.973988439306359e-07,
+      "loss": 0.9718,
+      "step": 6125
+    },
+    {
+      "epoch": 0.14763613593121552,
+      "grad_norm": 1.3734008073806763,
+      "learning_rate": 8.972720819389514e-07,
+      "loss": 0.8818,
+      "step": 6130
+    },
+    {
+      "epoch": 0.14775655692300282,
+      "grad_norm": 1.4598811864852905,
+      "learning_rate": 8.971453199472669e-07,
+      "loss": 0.9302,
+      "step": 6135
+    },
+    {
+      "epoch": 0.1478769779147901,
+      "grad_norm": 1.5441845655441284,
+      "learning_rate": 8.970185579555826e-07,
+      "loss": 0.8982,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1479973989065774,
+      "grad_norm": 1.6037344932556152,
+      "learning_rate": 8.968917959638982e-07,
+      "loss": 0.9638,
+      "step": 6145
+    },
+    {
+      "epoch": 0.14811781989836467,
+      "grad_norm": 1.66959810256958,
+      "learning_rate": 8.967650339722137e-07,
+      "loss": 0.9568,
+      "step": 6150
+    },
+    {
+      "epoch": 0.14823824089015197,
+      "grad_norm": 1.5276808738708496,
+      "learning_rate": 8.966382719805293e-07,
+      "loss": 0.9364,
+      "step": 6155
+    },
+    {
+      "epoch": 0.14835866188193925,
+      "grad_norm": 1.5752981901168823,
+      "learning_rate": 8.965115099888449e-07,
+      "loss": 0.9797,
+      "step": 6160
+    },
+    {
+      "epoch": 0.14847908287372655,
+      "grad_norm": 1.6811892986297607,
+      "learning_rate": 8.963847479971605e-07,
+      "loss": 0.9202,
+      "step": 6165
+    },
+    {
+      "epoch": 0.14859950386551385,
+      "grad_norm": 1.6610900163650513,
+      "learning_rate": 8.962579860054761e-07,
+      "loss": 0.9118,
+      "step": 6170
+    },
+    {
+      "epoch": 0.14871992485730112,
+      "grad_norm": 1.4799003601074219,
+      "learning_rate": 8.961312240137917e-07,
+      "loss": 0.9738,
+      "step": 6175
+    },
+    {
+      "epoch": 0.14884034584908842,
+      "grad_norm": 1.5117777585983276,
+      "learning_rate": 8.960044620221072e-07,
+      "loss": 0.9589,
+      "step": 6180
+    },
+    {
+      "epoch": 0.1489607668408757,
+      "grad_norm": 1.5448501110076904,
+      "learning_rate": 8.958777000304228e-07,
+      "loss": 0.9391,
+      "step": 6185
+    },
+    {
+      "epoch": 0.149081187832663,
+      "grad_norm": 1.6339643001556396,
+      "learning_rate": 8.957509380387385e-07,
+      "loss": 0.947,
+      "step": 6190
+    },
+    {
+      "epoch": 0.14920160882445027,
+      "grad_norm": 1.4847960472106934,
+      "learning_rate": 8.95624176047054e-07,
+      "loss": 1.0043,
+      "step": 6195
+    },
+    {
+      "epoch": 0.14932202981623757,
+      "grad_norm": 1.7087690830230713,
+      "learning_rate": 8.954974140553696e-07,
+      "loss": 1.0199,
+      "step": 6200
+    },
+    {
+      "epoch": 0.14944245080802485,
+      "grad_norm": 1.7362369298934937,
+      "learning_rate": 8.953706520636852e-07,
+      "loss": 0.906,
+      "step": 6205
+    },
+    {
+      "epoch": 0.14956287179981215,
+      "grad_norm": 1.414970874786377,
+      "learning_rate": 8.952438900720008e-07,
+      "loss": 0.9886,
+      "step": 6210
+    },
+    {
+      "epoch": 0.14968329279159942,
+      "grad_norm": 1.4704195261001587,
+      "learning_rate": 8.951171280803164e-07,
+      "loss": 0.9758,
+      "step": 6215
+    },
+    {
+      "epoch": 0.14980371378338672,
+      "grad_norm": 1.5782198905944824,
+      "learning_rate": 8.949903660886319e-07,
+      "loss": 1.0132,
+      "step": 6220
+    },
+    {
+      "epoch": 0.149924134775174,
+      "grad_norm": 1.6264985799789429,
+      "learning_rate": 8.948636040969475e-07,
+      "loss": 0.9184,
+      "step": 6225
+    },
+    {
+      "epoch": 0.1500445557669613,
+      "grad_norm": 1.6577425003051758,
+      "learning_rate": 8.947368421052631e-07,
+      "loss": 0.9571,
+      "step": 6230
+    },
+    {
+      "epoch": 0.1501649767587486,
+      "grad_norm": 1.3714252710342407,
+      "learning_rate": 8.946100801135788e-07,
+      "loss": 0.9219,
+      "step": 6235
+    },
+    {
+      "epoch": 0.15028539775053587,
+      "grad_norm": 1.5763870477676392,
+      "learning_rate": 8.944833181218943e-07,
+      "loss": 0.9337,
+      "step": 6240
+    },
+    {
+      "epoch": 0.15040581874232317,
+      "grad_norm": 1.5407581329345703,
+      "learning_rate": 8.943565561302098e-07,
+      "loss": 0.9042,
+      "step": 6245
+    },
+    {
+      "epoch": 0.15052623973411045,
+      "grad_norm": 1.387267827987671,
+      "learning_rate": 8.942297941385255e-07,
+      "loss": 0.9513,
+      "step": 6250
+    },
+    {
+      "epoch": 0.15064666072589775,
+      "grad_norm": 1.731074571609497,
+      "learning_rate": 8.94103032146841e-07,
+      "loss": 0.9562,
+      "step": 6255
+    },
+    {
+      "epoch": 0.15076708171768502,
+      "grad_norm": 1.5412241220474243,
+      "learning_rate": 8.939762701551567e-07,
+      "loss": 0.9866,
+      "step": 6260
+    },
+    {
+      "epoch": 0.15088750270947232,
+      "grad_norm": 1.3752002716064453,
+      "learning_rate": 8.938495081634723e-07,
+      "loss": 0.9786,
+      "step": 6265
+    },
+    {
+      "epoch": 0.1510079237012596,
+      "grad_norm": 1.5093090534210205,
+      "learning_rate": 8.937227461717878e-07,
+      "loss": 0.8847,
+      "step": 6270
+    },
+    {
+      "epoch": 0.1511283446930469,
+      "grad_norm": 1.5734307765960693,
+      "learning_rate": 8.935959841801034e-07,
+      "loss": 0.9767,
+      "step": 6275
+    },
+    {
+      "epoch": 0.15124876568483417,
+      "grad_norm": 1.5351828336715698,
+      "learning_rate": 8.93469222188419e-07,
+      "loss": 0.9378,
+      "step": 6280
+    },
+    {
+      "epoch": 0.15136918667662147,
+      "grad_norm": 1.6842786073684692,
+      "learning_rate": 8.933424601967346e-07,
+      "loss": 0.9786,
+      "step": 6285
+    },
+    {
+      "epoch": 0.15148960766840874,
+      "grad_norm": 1.781619906425476,
+      "learning_rate": 8.932156982050501e-07,
+      "loss": 0.9514,
+      "step": 6290
+    },
+    {
+      "epoch": 0.15161002866019604,
+      "grad_norm": 1.6095762252807617,
+      "learning_rate": 8.930889362133658e-07,
+      "loss": 0.9219,
+      "step": 6295
+    },
+    {
+      "epoch": 0.15173044965198335,
+      "grad_norm": 1.392642855644226,
+      "learning_rate": 8.929621742216813e-07,
+      "loss": 0.9286,
+      "step": 6300
+    },
+    {
+      "epoch": 0.15185087064377062,
+      "grad_norm": 1.4566751718521118,
+      "learning_rate": 8.928354122299969e-07,
+      "loss": 0.9776,
+      "step": 6305
+    },
+    {
+      "epoch": 0.15197129163555792,
+      "grad_norm": 1.5674692392349243,
+      "learning_rate": 8.927086502383126e-07,
+      "loss": 0.9211,
+      "step": 6310
+    },
+    {
+      "epoch": 0.1520917126273452,
+      "grad_norm": 1.6276681423187256,
+      "learning_rate": 8.92581888246628e-07,
+      "loss": 0.961,
+      "step": 6315
+    },
+    {
+      "epoch": 0.1522121336191325,
+      "grad_norm": 1.4062598943710327,
+      "learning_rate": 8.924551262549437e-07,
+      "loss": 0.9144,
+      "step": 6320
+    },
+    {
+      "epoch": 0.15233255461091977,
+      "grad_norm": 1.5248562097549438,
+      "learning_rate": 8.923283642632593e-07,
+      "loss": 0.986,
+      "step": 6325
+    },
+    {
+      "epoch": 0.15245297560270707,
+      "grad_norm": 1.496514916419983,
+      "learning_rate": 8.922016022715749e-07,
+      "loss": 0.9464,
+      "step": 6330
+    },
+    {
+      "epoch": 0.15257339659449434,
+      "grad_norm": 1.688599944114685,
+      "learning_rate": 8.920748402798904e-07,
+      "loss": 0.9531,
+      "step": 6335
+    },
+    {
+      "epoch": 0.15269381758628164,
+      "grad_norm": 1.6141420602798462,
+      "learning_rate": 8.91948078288206e-07,
+      "loss": 0.9509,
+      "step": 6340
+    },
+    {
+      "epoch": 0.15281423857806892,
+      "grad_norm": 1.5606366395950317,
+      "learning_rate": 8.918213162965216e-07,
+      "loss": 1.0047,
+      "step": 6345
+    },
+    {
+      "epoch": 0.15293465956985622,
+      "grad_norm": 1.5859097242355347,
+      "learning_rate": 8.916945543048372e-07,
+      "loss": 0.9504,
+      "step": 6350
+    },
+    {
+      "epoch": 0.15305508056164352,
+      "grad_norm": 1.4285775423049927,
+      "learning_rate": 8.915677923131529e-07,
+      "loss": 0.9887,
+      "step": 6355
+    },
+    {
+      "epoch": 0.1531755015534308,
+      "grad_norm": 1.529550313949585,
+      "learning_rate": 8.914410303214683e-07,
+      "loss": 0.9791,
+      "step": 6360
+    },
+    {
+      "epoch": 0.1532959225452181,
+      "grad_norm": 1.4174224138259888,
+      "learning_rate": 8.913142683297839e-07,
+      "loss": 0.9772,
+      "step": 6365
+    },
+    {
+      "epoch": 0.15341634353700537,
+      "grad_norm": 1.4871257543563843,
+      "learning_rate": 8.911875063380996e-07,
+      "loss": 0.8747,
+      "step": 6370
+    },
+    {
+      "epoch": 0.15353676452879267,
+      "grad_norm": 1.4553872346878052,
+      "learning_rate": 8.910607443464151e-07,
+      "loss": 0.9605,
+      "step": 6375
+    },
+    {
+      "epoch": 0.15365718552057994,
+      "grad_norm": 1.5385500192642212,
+      "learning_rate": 8.909339823547307e-07,
+      "loss": 0.9069,
+      "step": 6380
+    },
+    {
+      "epoch": 0.15377760651236724,
+      "grad_norm": 1.5096920728683472,
+      "learning_rate": 8.908072203630463e-07,
+      "loss": 0.9155,
+      "step": 6385
+    },
+    {
+      "epoch": 0.15389802750415452,
+      "grad_norm": 1.5188795328140259,
+      "learning_rate": 8.906804583713618e-07,
+      "loss": 0.9971,
+      "step": 6390
+    },
+    {
+      "epoch": 0.15401844849594182,
+      "grad_norm": 1.3419337272644043,
+      "learning_rate": 8.905536963796775e-07,
+      "loss": 0.9307,
+      "step": 6395
+    },
+    {
+      "epoch": 0.1541388694877291,
+      "grad_norm": 1.4751125574111938,
+      "learning_rate": 8.904269343879931e-07,
+      "loss": 0.8893,
+      "step": 6400
+    },
+    {
+      "epoch": 0.1542592904795164,
+      "grad_norm": 1.5324145555496216,
+      "learning_rate": 8.903001723963086e-07,
+      "loss": 0.966,
+      "step": 6405
+    },
+    {
+      "epoch": 0.15437971147130367,
+      "grad_norm": 1.554923176765442,
+      "learning_rate": 8.901734104046242e-07,
+      "loss": 0.9722,
+      "step": 6410
+    },
+    {
+      "epoch": 0.15450013246309097,
+      "grad_norm": 1.5499253273010254,
+      "learning_rate": 8.900466484129399e-07,
+      "loss": 0.9327,
+      "step": 6415
+    },
+    {
+      "epoch": 0.15462055345487827,
+      "grad_norm": 1.539252519607544,
+      "learning_rate": 8.899198864212554e-07,
+      "loss": 0.9854,
+      "step": 6420
+    },
+    {
+      "epoch": 0.15474097444666554,
+      "grad_norm": 1.5881555080413818,
+      "learning_rate": 8.89793124429571e-07,
+      "loss": 1.0085,
+      "step": 6425
+    },
+    {
+      "epoch": 0.15486139543845284,
+      "grad_norm": 1.6186283826828003,
+      "learning_rate": 8.896663624378866e-07,
+      "loss": 0.9865,
+      "step": 6430
+    },
+    {
+      "epoch": 0.15498181643024012,
+      "grad_norm": 1.7926065921783447,
+      "learning_rate": 8.895396004462021e-07,
+      "loss": 1.0098,
+      "step": 6435
+    },
+    {
+      "epoch": 0.15510223742202742,
+      "grad_norm": 1.7826013565063477,
+      "learning_rate": 8.894128384545178e-07,
+      "loss": 0.9576,
+      "step": 6440
+    },
+    {
+      "epoch": 0.1552226584138147,
+      "grad_norm": 1.736943006515503,
+      "learning_rate": 8.892860764628334e-07,
+      "loss": 0.9785,
+      "step": 6445
+    },
+    {
+      "epoch": 0.155343079405602,
+      "grad_norm": 1.822184681892395,
+      "learning_rate": 8.891593144711488e-07,
+      "loss": 0.9646,
+      "step": 6450
+    },
+    {
+      "epoch": 0.15546350039738926,
+      "grad_norm": 1.5065011978149414,
+      "learning_rate": 8.890325524794645e-07,
+      "loss": 0.9547,
+      "step": 6455
+    },
+    {
+      "epoch": 0.15558392138917657,
+      "grad_norm": 1.5549193620681763,
+      "learning_rate": 8.889057904877801e-07,
+      "loss": 0.9702,
+      "step": 6460
+    },
+    {
+      "epoch": 0.15570434238096384,
+      "grad_norm": 1.4823460578918457,
+      "learning_rate": 8.887790284960957e-07,
+      "loss": 0.9393,
+      "step": 6465
+    },
+    {
+      "epoch": 0.15582476337275114,
+      "grad_norm": 1.5357725620269775,
+      "learning_rate": 8.886522665044113e-07,
+      "loss": 1.0052,
+      "step": 6470
+    },
+    {
+      "epoch": 0.15594518436453841,
+      "grad_norm": 1.395545244216919,
+      "learning_rate": 8.885255045127269e-07,
+      "loss": 0.927,
+      "step": 6475
+    },
+    {
+      "epoch": 0.15606560535632572,
+      "grad_norm": 1.5250643491744995,
+      "learning_rate": 8.883987425210424e-07,
+      "loss": 0.984,
+      "step": 6480
+    },
+    {
+      "epoch": 0.15618602634811302,
+      "grad_norm": 1.5090745687484741,
+      "learning_rate": 8.88271980529358e-07,
+      "loss": 0.9602,
+      "step": 6485
+    },
+    {
+      "epoch": 0.1563064473399003,
+      "grad_norm": 1.5585459470748901,
+      "learning_rate": 8.881452185376737e-07,
+      "loss": 0.9441,
+      "step": 6490
+    },
+    {
+      "epoch": 0.1564268683316876,
+      "grad_norm": 1.7480950355529785,
+      "learning_rate": 8.880184565459891e-07,
+      "loss": 0.9616,
+      "step": 6495
+    },
+    {
+      "epoch": 0.15654728932347486,
+      "grad_norm": 1.503509759902954,
+      "learning_rate": 8.878916945543048e-07,
+      "loss": 0.9515,
+      "step": 6500
+    },
+    {
+      "epoch": 0.15666771031526217,
+      "grad_norm": 1.5634033679962158,
+      "learning_rate": 8.877649325626204e-07,
+      "loss": 1.0264,
+      "step": 6505
+    },
+    {
+      "epoch": 0.15678813130704944,
+      "grad_norm": 1.530869960784912,
+      "learning_rate": 8.876381705709359e-07,
+      "loss": 0.984,
+      "step": 6510
+    },
+    {
+      "epoch": 0.15690855229883674,
+      "grad_norm": 1.5853240489959717,
+      "learning_rate": 8.875114085792516e-07,
+      "loss": 1.0086,
+      "step": 6515
+    },
+    {
+      "epoch": 0.157028973290624,
+      "grad_norm": 1.508705496788025,
+      "learning_rate": 8.873846465875671e-07,
+      "loss": 0.9516,
+      "step": 6520
+    },
+    {
+      "epoch": 0.15714939428241131,
+      "grad_norm": 1.5178998708724976,
+      "learning_rate": 8.872578845958827e-07,
+      "loss": 0.9062,
+      "step": 6525
+    },
+    {
+      "epoch": 0.1572698152741986,
+      "grad_norm": 1.529740810394287,
+      "learning_rate": 8.871311226041983e-07,
+      "loss": 0.9721,
+      "step": 6530
+    },
+    {
+      "epoch": 0.1573902362659859,
+      "grad_norm": 1.5887424945831299,
+      "learning_rate": 8.87004360612514e-07,
+      "loss": 0.9015,
+      "step": 6535
+    },
+    {
+      "epoch": 0.15751065725777316,
+      "grad_norm": 2.0315704345703125,
+      "learning_rate": 8.868775986208295e-07,
+      "loss": 0.9583,
+      "step": 6540
+    },
+    {
+      "epoch": 0.15763107824956046,
+      "grad_norm": 1.5251991748809814,
+      "learning_rate": 8.86750836629145e-07,
+      "loss": 0.9607,
+      "step": 6545
+    },
+    {
+      "epoch": 0.15775149924134776,
+      "grad_norm": 1.5378426313400269,
+      "learning_rate": 8.866240746374607e-07,
+      "loss": 1.0108,
+      "step": 6550
+    },
+    {
+      "epoch": 0.15787192023313504,
+      "grad_norm": 1.4061293601989746,
+      "learning_rate": 8.864973126457762e-07,
+      "loss": 0.8955,
+      "step": 6555
+    },
+    {
+      "epoch": 0.15799234122492234,
+      "grad_norm": 1.381935477256775,
+      "learning_rate": 8.863705506540919e-07,
+      "loss": 0.9266,
+      "step": 6560
+    },
+    {
+      "epoch": 0.1581127622167096,
+      "grad_norm": 1.6147645711898804,
+      "learning_rate": 8.862437886624074e-07,
+      "loss": 0.9526,
+      "step": 6565
+    },
+    {
+      "epoch": 0.1582331832084969,
+      "grad_norm": 1.4142693281173706,
+      "learning_rate": 8.861170266707229e-07,
+      "loss": 0.9508,
+      "step": 6570
+    },
+    {
+      "epoch": 0.1583536042002842,
+      "grad_norm": 1.6666898727416992,
+      "learning_rate": 8.859902646790386e-07,
+      "loss": 0.9025,
+      "step": 6575
+    },
+    {
+      "epoch": 0.1584740251920715,
+      "grad_norm": 1.609081745147705,
+      "learning_rate": 8.858635026873542e-07,
+      "loss": 0.9337,
+      "step": 6580
+    },
+    {
+      "epoch": 0.15859444618385876,
+      "grad_norm": 1.6501929759979248,
+      "learning_rate": 8.857367406956698e-07,
+      "loss": 0.8925,
+      "step": 6585
+    },
+    {
+      "epoch": 0.15871486717564606,
+      "grad_norm": 1.5381691455841064,
+      "learning_rate": 8.856099787039853e-07,
+      "loss": 0.9307,
+      "step": 6590
+    },
+    {
+      "epoch": 0.15883528816743334,
+      "grad_norm": 1.5835388898849487,
+      "learning_rate": 8.85483216712301e-07,
+      "loss": 0.9817,
+      "step": 6595
+    },
+    {
+      "epoch": 0.15895570915922064,
+      "grad_norm": 3.076845407485962,
+      "learning_rate": 8.853564547206165e-07,
+      "loss": 0.9501,
+      "step": 6600
+    },
+    {
+      "epoch": 0.1590761301510079,
+      "grad_norm": 1.6130303144454956,
+      "learning_rate": 8.852296927289321e-07,
+      "loss": 0.9755,
+      "step": 6605
+    },
+    {
+      "epoch": 0.1591965511427952,
+      "grad_norm": 1.55154287815094,
+      "learning_rate": 8.851029307372478e-07,
+      "loss": 0.9794,
+      "step": 6610
+    },
+    {
+      "epoch": 0.1593169721345825,
+      "grad_norm": 1.5113273859024048,
+      "learning_rate": 8.849761687455632e-07,
+      "loss": 0.9276,
+      "step": 6615
+    },
+    {
+      "epoch": 0.1594373931263698,
+      "grad_norm": 1.5491209030151367,
+      "learning_rate": 8.848494067538789e-07,
+      "loss": 1.0023,
+      "step": 6620
+    },
+    {
+      "epoch": 0.1595578141181571,
+      "grad_norm": 1.6405673027038574,
+      "learning_rate": 8.847226447621945e-07,
+      "loss": 0.9533,
+      "step": 6625
+    },
+    {
+      "epoch": 0.15967823510994436,
+      "grad_norm": 1.6780866384506226,
+      "learning_rate": 8.8459588277051e-07,
+      "loss": 0.959,
+      "step": 6630
+    },
+    {
+      "epoch": 0.15979865610173166,
+      "grad_norm": 1.3100441694259644,
+      "learning_rate": 8.844691207788256e-07,
+      "loss": 1.0109,
+      "step": 6635
+    },
+    {
+      "epoch": 0.15991907709351894,
+      "grad_norm": 1.7152460813522339,
+      "learning_rate": 8.843423587871412e-07,
+      "loss": 0.9075,
+      "step": 6640
+    },
+    {
+      "epoch": 0.16003949808530624,
+      "grad_norm": 1.6447054147720337,
+      "learning_rate": 8.842155967954568e-07,
+      "loss": 0.937,
+      "step": 6645
+    },
+    {
+      "epoch": 0.1601599190770935,
+      "grad_norm": 1.7059388160705566,
+      "learning_rate": 8.840888348037724e-07,
+      "loss": 0.9517,
+      "step": 6650
+    },
+    {
+      "epoch": 0.1602803400688808,
+      "grad_norm": 1.547312617301941,
+      "learning_rate": 8.839620728120881e-07,
+      "loss": 0.8843,
+      "step": 6655
+    },
+    {
+      "epoch": 0.16040076106066808,
+      "grad_norm": 1.666043758392334,
+      "learning_rate": 8.838353108204035e-07,
+      "loss": 0.9763,
+      "step": 6660
+    },
+    {
+      "epoch": 0.16052118205245539,
+      "grad_norm": 1.5424047708511353,
+      "learning_rate": 8.837085488287191e-07,
+      "loss": 0.9772,
+      "step": 6665
+    },
+    {
+      "epoch": 0.16064160304424266,
+      "grad_norm": 1.7461053133010864,
+      "learning_rate": 8.835817868370348e-07,
+      "loss": 0.9384,
+      "step": 6670
+    },
+    {
+      "epoch": 0.16076202403602996,
+      "grad_norm": 1.5185924768447876,
+      "learning_rate": 8.834550248453503e-07,
+      "loss": 0.9413,
+      "step": 6675
+    },
+    {
+      "epoch": 0.16088244502781726,
+      "grad_norm": 1.5015673637390137,
+      "learning_rate": 8.833282628536659e-07,
+      "loss": 0.9503,
+      "step": 6680
+    },
+    {
+      "epoch": 0.16100286601960453,
+      "grad_norm": 1.4696515798568726,
+      "learning_rate": 8.832015008619815e-07,
+      "loss": 0.9829,
+      "step": 6685
+    },
+    {
+      "epoch": 0.16112328701139184,
+      "grad_norm": 1.4103220701217651,
+      "learning_rate": 8.83074738870297e-07,
+      "loss": 0.9143,
+      "step": 6690
+    },
+    {
+      "epoch": 0.1612437080031791,
+      "grad_norm": 1.5549726486206055,
+      "learning_rate": 8.829479768786127e-07,
+      "loss": 0.9381,
+      "step": 6695
+    },
+    {
+      "epoch": 0.1613641289949664,
+      "grad_norm": 1.6166776418685913,
+      "learning_rate": 8.828212148869283e-07,
+      "loss": 0.9271,
+      "step": 6700
+    },
+    {
+      "epoch": 0.16148454998675368,
+      "grad_norm": 1.652779459953308,
+      "learning_rate": 8.826944528952438e-07,
+      "loss": 0.9363,
+      "step": 6705
+    },
+    {
+      "epoch": 0.16160497097854098,
+      "grad_norm": 1.5943504571914673,
+      "learning_rate": 8.825676909035594e-07,
+      "loss": 0.9475,
+      "step": 6710
+    },
+    {
+      "epoch": 0.16172539197032826,
+      "grad_norm": 1.6199977397918701,
+      "learning_rate": 8.82440928911875e-07,
+      "loss": 0.9631,
+      "step": 6715
+    },
+    {
+      "epoch": 0.16184581296211556,
+      "grad_norm": 1.6312978267669678,
+      "learning_rate": 8.823141669201906e-07,
+      "loss": 0.9328,
+      "step": 6720
+    },
+    {
+      "epoch": 0.16196623395390283,
+      "grad_norm": 1.6854356527328491,
+      "learning_rate": 8.821874049285062e-07,
+      "loss": 1.0102,
+      "step": 6725
+    },
+    {
+      "epoch": 0.16208665494569013,
+      "grad_norm": 1.5627775192260742,
+      "learning_rate": 8.820606429368218e-07,
+      "loss": 1.0122,
+      "step": 6730
+    },
+    {
+      "epoch": 0.1622070759374774,
+      "grad_norm": 1.4499491453170776,
+      "learning_rate": 8.819338809451373e-07,
+      "loss": 0.9514,
+      "step": 6735
+    },
+    {
+      "epoch": 0.1623274969292647,
+      "grad_norm": 1.5392065048217773,
+      "learning_rate": 8.81807118953453e-07,
+      "loss": 0.9317,
+      "step": 6740
+    },
+    {
+      "epoch": 0.162447917921052,
+      "grad_norm": 1.5162497758865356,
+      "learning_rate": 8.816803569617686e-07,
+      "loss": 0.9393,
+      "step": 6745
+    },
+    {
+      "epoch": 0.16256833891283928,
+      "grad_norm": 1.8237184286117554,
+      "learning_rate": 8.81553594970084e-07,
+      "loss": 0.9112,
+      "step": 6750
+    },
+    {
+      "epoch": 0.16268875990462658,
+      "grad_norm": 1.511694073677063,
+      "learning_rate": 8.814268329783997e-07,
+      "loss": 0.9365,
+      "step": 6755
+    },
+    {
+      "epoch": 0.16280918089641386,
+      "grad_norm": 1.4274612665176392,
+      "learning_rate": 8.813000709867153e-07,
+      "loss": 0.9378,
+      "step": 6760
+    },
+    {
+      "epoch": 0.16292960188820116,
+      "grad_norm": 1.5141350030899048,
+      "learning_rate": 8.811733089950309e-07,
+      "loss": 0.9561,
+      "step": 6765
+    },
+    {
+      "epoch": 0.16305002287998843,
+      "grad_norm": 1.6848194599151611,
+      "learning_rate": 8.810465470033465e-07,
+      "loss": 0.9606,
+      "step": 6770
+    },
+    {
+      "epoch": 0.16317044387177573,
+      "grad_norm": 1.487057089805603,
+      "learning_rate": 8.80919785011662e-07,
+      "loss": 0.9161,
+      "step": 6775
+    },
+    {
+      "epoch": 0.163290864863563,
+      "grad_norm": 1.4446853399276733,
+      "learning_rate": 8.807930230199776e-07,
+      "loss": 0.9692,
+      "step": 6780
+    },
+    {
+      "epoch": 0.1634112858553503,
+      "grad_norm": 2.0542047023773193,
+      "learning_rate": 8.806662610282932e-07,
+      "loss": 0.9392,
+      "step": 6785
+    },
+    {
+      "epoch": 0.16353170684713758,
+      "grad_norm": 1.7039285898208618,
+      "learning_rate": 8.805394990366089e-07,
+      "loss": 0.9278,
+      "step": 6790
+    },
+    {
+      "epoch": 0.16365212783892488,
+      "grad_norm": 1.584222435951233,
+      "learning_rate": 8.804127370449243e-07,
+      "loss": 0.932,
+      "step": 6795
+    },
+    {
+      "epoch": 0.16377254883071218,
+      "grad_norm": 1.6305924654006958,
+      "learning_rate": 8.8028597505324e-07,
+      "loss": 0.9655,
+      "step": 6800
+    },
+    {
+      "epoch": 0.16389296982249946,
+      "grad_norm": 1.6200332641601562,
+      "learning_rate": 8.801592130615556e-07,
+      "loss": 0.96,
+      "step": 6805
+    },
+    {
+      "epoch": 0.16401339081428676,
+      "grad_norm": 1.4420777559280396,
+      "learning_rate": 8.800324510698711e-07,
+      "loss": 0.9659,
+      "step": 6810
+    },
+    {
+      "epoch": 0.16413381180607403,
+      "grad_norm": 1.7470130920410156,
+      "learning_rate": 8.799056890781868e-07,
+      "loss": 0.924,
+      "step": 6815
+    },
+    {
+      "epoch": 0.16425423279786133,
+      "grad_norm": 1.611938714981079,
+      "learning_rate": 8.797789270865023e-07,
+      "loss": 0.981,
+      "step": 6820
+    },
+    {
+      "epoch": 0.1643746537896486,
+      "grad_norm": 1.7520917654037476,
+      "learning_rate": 8.796521650948179e-07,
+      "loss": 0.9833,
+      "step": 6825
+    },
+    {
+      "epoch": 0.1644950747814359,
+      "grad_norm": 1.7744758129119873,
+      "learning_rate": 8.795254031031335e-07,
+      "loss": 0.9853,
+      "step": 6830
+    },
+    {
+      "epoch": 0.16461549577322318,
+      "grad_norm": 1.465274691581726,
+      "learning_rate": 8.793986411114491e-07,
+      "loss": 0.9643,
+      "step": 6835
+    },
+    {
+      "epoch": 0.16473591676501048,
+      "grad_norm": 1.4554154872894287,
+      "learning_rate": 8.792718791197647e-07,
+      "loss": 0.9115,
+      "step": 6840
+    },
+    {
+      "epoch": 0.16485633775679776,
+      "grad_norm": 1.8319108486175537,
+      "learning_rate": 8.791451171280802e-07,
+      "loss": 0.9758,
+      "step": 6845
+    },
+    {
+      "epoch": 0.16497675874858506,
+      "grad_norm": 1.3700926303863525,
+      "learning_rate": 8.790183551363959e-07,
+      "loss": 0.987,
+      "step": 6850
+    },
+    {
+      "epoch": 0.16509717974037233,
+      "grad_norm": 1.4417375326156616,
+      "learning_rate": 8.788915931447114e-07,
+      "loss": 0.9503,
+      "step": 6855
+    },
+    {
+      "epoch": 0.16521760073215963,
+      "grad_norm": 1.8359936475753784,
+      "learning_rate": 8.787648311530271e-07,
+      "loss": 0.939,
+      "step": 6860
+    },
+    {
+      "epoch": 0.16533802172394693,
+      "grad_norm": 1.694645643234253,
+      "learning_rate": 8.786380691613426e-07,
+      "loss": 0.9382,
+      "step": 6865
+    },
+    {
+      "epoch": 0.1654584427157342,
+      "grad_norm": 1.778012752532959,
+      "learning_rate": 8.785113071696581e-07,
+      "loss": 0.9851,
+      "step": 6870
+    },
+    {
+      "epoch": 0.1655788637075215,
+      "grad_norm": 1.6528106927871704,
+      "learning_rate": 8.783845451779738e-07,
+      "loss": 0.9749,
+      "step": 6875
+    },
+    {
+      "epoch": 0.16569928469930878,
+      "grad_norm": 1.4038870334625244,
+      "learning_rate": 8.782577831862894e-07,
+      "loss": 0.9657,
+      "step": 6880
+    },
+    {
+      "epoch": 0.16581970569109608,
+      "grad_norm": 1.967231035232544,
+      "learning_rate": 8.78131021194605e-07,
+      "loss": 0.9531,
+      "step": 6885
+    },
+    {
+      "epoch": 0.16594012668288335,
+      "grad_norm": 1.5113967657089233,
+      "learning_rate": 8.780042592029205e-07,
+      "loss": 0.9842,
+      "step": 6890
+    },
+    {
+      "epoch": 0.16606054767467066,
+      "grad_norm": 1.4383405447006226,
+      "learning_rate": 8.778774972112361e-07,
+      "loss": 0.9339,
+      "step": 6895
+    },
+    {
+      "epoch": 0.16618096866645793,
+      "grad_norm": 1.5188686847686768,
+      "learning_rate": 8.777507352195517e-07,
+      "loss": 0.994,
+      "step": 6900
+    },
+    {
+      "epoch": 0.16630138965824523,
+      "grad_norm": 1.4522814750671387,
+      "learning_rate": 8.776239732278673e-07,
+      "loss": 0.9889,
+      "step": 6905
+    },
+    {
+      "epoch": 0.1664218106500325,
+      "grad_norm": 1.40193510055542,
+      "learning_rate": 8.77497211236183e-07,
+      "loss": 0.9585,
+      "step": 6910
+    },
+    {
+      "epoch": 0.1665422316418198,
+      "grad_norm": 1.6679961681365967,
+      "learning_rate": 8.773704492444985e-07,
+      "loss": 0.9333,
+      "step": 6915
+    },
+    {
+      "epoch": 0.16666265263360708,
+      "grad_norm": 1.6849275827407837,
+      "learning_rate": 8.77243687252814e-07,
+      "loss": 0.9663,
+      "step": 6920
+    },
+    {
+      "epoch": 0.16678307362539438,
+      "grad_norm": 1.5537941455841064,
+      "learning_rate": 8.771169252611297e-07,
+      "loss": 0.969,
+      "step": 6925
+    },
+    {
+      "epoch": 0.16690349461718168,
+      "grad_norm": 1.6917616128921509,
+      "learning_rate": 8.769901632694453e-07,
+      "loss": 0.9102,
+      "step": 6930
+    },
+    {
+      "epoch": 0.16702391560896895,
+      "grad_norm": 1.4796494245529175,
+      "learning_rate": 8.768634012777608e-07,
+      "loss": 0.9696,
+      "step": 6935
+    },
+    {
+      "epoch": 0.16714433660075625,
+      "grad_norm": 1.639575481414795,
+      "learning_rate": 8.767366392860764e-07,
+      "loss": 0.9919,
+      "step": 6940
+    },
+    {
+      "epoch": 0.16726475759254353,
+      "grad_norm": 1.4652765989303589,
+      "learning_rate": 8.766098772943921e-07,
+      "loss": 0.9667,
+      "step": 6945
+    },
+    {
+      "epoch": 0.16738517858433083,
+      "grad_norm": 1.4565825462341309,
+      "learning_rate": 8.764831153027076e-07,
+      "loss": 0.9219,
+      "step": 6950
+    },
+    {
+      "epoch": 0.1675055995761181,
+      "grad_norm": 1.7221384048461914,
+      "learning_rate": 8.763563533110232e-07,
+      "loss": 0.9487,
+      "step": 6955
+    },
+    {
+      "epoch": 0.1676260205679054,
+      "grad_norm": 1.753455638885498,
+      "learning_rate": 8.762295913193388e-07,
+      "loss": 0.91,
+      "step": 6960
+    },
+    {
+      "epoch": 0.16774644155969268,
+      "grad_norm": 1.534593939781189,
+      "learning_rate": 8.761028293276543e-07,
+      "loss": 0.9343,
+      "step": 6965
+    },
+    {
+      "epoch": 0.16786686255147998,
+      "grad_norm": 1.6360231637954712,
+      "learning_rate": 8.7597606733597e-07,
+      "loss": 0.9446,
+      "step": 6970
+    },
+    {
+      "epoch": 0.16798728354326725,
+      "grad_norm": 1.6140836477279663,
+      "learning_rate": 8.758493053442856e-07,
+      "loss": 0.9353,
+      "step": 6975
+    },
+    {
+      "epoch": 0.16810770453505455,
+      "grad_norm": 1.7073837518692017,
+      "learning_rate": 8.75722543352601e-07,
+      "loss": 0.9966,
+      "step": 6980
+    },
+    {
+      "epoch": 0.16822812552684183,
+      "grad_norm": 1.6094011068344116,
+      "learning_rate": 8.755957813609167e-07,
+      "loss": 0.9944,
+      "step": 6985
+    },
+    {
+      "epoch": 0.16834854651862913,
+      "grad_norm": 1.502415418624878,
+      "learning_rate": 8.754690193692323e-07,
+      "loss": 0.9864,
+      "step": 6990
+    },
+    {
+      "epoch": 0.16846896751041643,
+      "grad_norm": 1.5931711196899414,
+      "learning_rate": 8.753422573775479e-07,
+      "loss": 0.8975,
+      "step": 6995
+    },
+    {
+      "epoch": 0.1685893885022037,
+      "grad_norm": 1.5415033102035522,
+      "learning_rate": 8.752154953858635e-07,
+      "loss": 0.9557,
+      "step": 7000
+    },
+    {
+      "epoch": 0.168709809493991,
+      "grad_norm": 1.4606413841247559,
+      "learning_rate": 8.750887333941791e-07,
+      "loss": 0.9077,
+      "step": 7005
+    },
+    {
+      "epoch": 0.16883023048577828,
+      "grad_norm": 1.534355640411377,
+      "learning_rate": 8.749619714024946e-07,
+      "loss": 0.9229,
+      "step": 7010
+    },
+    {
+      "epoch": 0.16895065147756558,
+      "grad_norm": 1.5767483711242676,
+      "learning_rate": 8.748352094108102e-07,
+      "loss": 1.0249,
+      "step": 7015
+    },
+    {
+      "epoch": 0.16907107246935285,
+      "grad_norm": 1.371443748474121,
+      "learning_rate": 8.747084474191259e-07,
+      "loss": 0.9365,
+      "step": 7020
+    },
+    {
+      "epoch": 0.16919149346114015,
+      "grad_norm": 1.5228232145309448,
+      "learning_rate": 8.745816854274414e-07,
+      "loss": 0.9641,
+      "step": 7025
+    },
+    {
+      "epoch": 0.16931191445292743,
+      "grad_norm": 1.520755648612976,
+      "learning_rate": 8.74454923435757e-07,
+      "loss": 0.9721,
+      "step": 7030
+    },
+    {
+      "epoch": 0.16943233544471473,
+      "grad_norm": 1.4273567199707031,
+      "learning_rate": 8.743281614440726e-07,
+      "loss": 0.902,
+      "step": 7035
+    },
+    {
+      "epoch": 0.169552756436502,
+      "grad_norm": 1.4603309631347656,
+      "learning_rate": 8.742013994523881e-07,
+      "loss": 0.9388,
+      "step": 7040
+    },
+    {
+      "epoch": 0.1696731774282893,
+      "grad_norm": 3.162548303604126,
+      "learning_rate": 8.740746374607038e-07,
+      "loss": 0.9625,
+      "step": 7045
+    },
+    {
+      "epoch": 0.16979359842007657,
+      "grad_norm": 1.5650300979614258,
+      "learning_rate": 8.739478754690193e-07,
+      "loss": 0.9908,
+      "step": 7050
+    },
+    {
+      "epoch": 0.16991401941186388,
+      "grad_norm": 1.647933006286621,
+      "learning_rate": 8.738211134773349e-07,
+      "loss": 0.9675,
+      "step": 7055
+    },
+    {
+      "epoch": 0.17003444040365118,
+      "grad_norm": 1.657855749130249,
+      "learning_rate": 8.736943514856505e-07,
+      "loss": 0.9611,
+      "step": 7060
+    },
+    {
+      "epoch": 0.17015486139543845,
+      "grad_norm": 1.6878466606140137,
+      "learning_rate": 8.735675894939662e-07,
+      "loss": 0.9491,
+      "step": 7065
+    },
+    {
+      "epoch": 0.17027528238722575,
+      "grad_norm": 1.745612621307373,
+      "learning_rate": 8.734408275022817e-07,
+      "loss": 0.9741,
+      "step": 7070
+    },
+    {
+      "epoch": 0.17039570337901302,
+      "grad_norm": 1.6257210969924927,
+      "learning_rate": 8.733140655105972e-07,
+      "loss": 0.9407,
+      "step": 7075
+    },
+    {
+      "epoch": 0.17051612437080033,
+      "grad_norm": 1.6022166013717651,
+      "learning_rate": 8.731873035189129e-07,
+      "loss": 0.8677,
+      "step": 7080
+    },
+    {
+      "epoch": 0.1706365453625876,
+      "grad_norm": 1.5448520183563232,
+      "learning_rate": 8.730605415272284e-07,
+      "loss": 0.985,
+      "step": 7085
+    },
+    {
+      "epoch": 0.1707569663543749,
+      "grad_norm": 1.5920443534851074,
+      "learning_rate": 8.729337795355441e-07,
+      "loss": 0.9544,
+      "step": 7090
+    },
+    {
+      "epoch": 0.17087738734616217,
+      "grad_norm": 1.6187429428100586,
+      "learning_rate": 8.728070175438597e-07,
+      "loss": 0.961,
+      "step": 7095
+    },
+    {
+      "epoch": 0.17099780833794948,
+      "grad_norm": 1.5467497110366821,
+      "learning_rate": 8.726802555521751e-07,
+      "loss": 0.9439,
+      "step": 7100
+    },
+    {
+      "epoch": 0.17111822932973675,
+      "grad_norm": 1.6352099180221558,
+      "learning_rate": 8.725534935604908e-07,
+      "loss": 0.945,
+      "step": 7105
+    },
+    {
+      "epoch": 0.17123865032152405,
+      "grad_norm": 1.3417030572891235,
+      "learning_rate": 8.724267315688064e-07,
+      "loss": 0.9185,
+      "step": 7110
+    },
+    {
+      "epoch": 0.17135907131331132,
+      "grad_norm": 1.594738245010376,
+      "learning_rate": 8.72299969577122e-07,
+      "loss": 0.9201,
+      "step": 7115
+    },
+    {
+      "epoch": 0.17147949230509862,
+      "grad_norm": 1.7222914695739746,
+      "learning_rate": 8.721732075854375e-07,
+      "loss": 0.927,
+      "step": 7120
+    },
+    {
+      "epoch": 0.17159991329688593,
+      "grad_norm": 2.0674643516540527,
+      "learning_rate": 8.720464455937532e-07,
+      "loss": 0.9905,
+      "step": 7125
+    },
+    {
+      "epoch": 0.1717203342886732,
+      "grad_norm": 1.612656831741333,
+      "learning_rate": 8.719196836020687e-07,
+      "loss": 0.9274,
+      "step": 7130
+    },
+    {
+      "epoch": 0.1718407552804605,
+      "grad_norm": 1.4258354902267456,
+      "learning_rate": 8.717929216103843e-07,
+      "loss": 0.971,
+      "step": 7135
+    },
+    {
+      "epoch": 0.17196117627224777,
+      "grad_norm": 1.4199203252792358,
+      "learning_rate": 8.716661596187e-07,
+      "loss": 0.9387,
+      "step": 7140
+    },
+    {
+      "epoch": 0.17208159726403507,
+      "grad_norm": 1.5075029134750366,
+      "learning_rate": 8.715393976270154e-07,
+      "loss": 0.9702,
+      "step": 7145
+    },
+    {
+      "epoch": 0.17220201825582235,
+      "grad_norm": 1.5064207315444946,
+      "learning_rate": 8.714126356353311e-07,
+      "loss": 0.8867,
+      "step": 7150
+    },
+    {
+      "epoch": 0.17232243924760965,
+      "grad_norm": 1.5721375942230225,
+      "learning_rate": 8.712858736436467e-07,
+      "loss": 0.956,
+      "step": 7155
+    },
+    {
+      "epoch": 0.17244286023939692,
+      "grad_norm": 1.6020773649215698,
+      "learning_rate": 8.711591116519622e-07,
+      "loss": 0.94,
+      "step": 7160
+    },
+    {
+      "epoch": 0.17256328123118422,
+      "grad_norm": 1.4800697565078735,
+      "learning_rate": 8.710323496602778e-07,
+      "loss": 0.9859,
+      "step": 7165
+    },
+    {
+      "epoch": 0.1726837022229715,
+      "grad_norm": 1.772230625152588,
+      "learning_rate": 8.709055876685934e-07,
+      "loss": 0.9597,
+      "step": 7170
+    },
+    {
+      "epoch": 0.1728041232147588,
+      "grad_norm": 1.602545142173767,
+      "learning_rate": 8.70778825676909e-07,
+      "loss": 0.9412,
+      "step": 7175
+    },
+    {
+      "epoch": 0.17292454420654607,
+      "grad_norm": 1.5732465982437134,
+      "learning_rate": 8.706520636852246e-07,
+      "loss": 0.9165,
+      "step": 7180
+    },
+    {
+      "epoch": 0.17304496519833337,
+      "grad_norm": 1.5865917205810547,
+      "learning_rate": 8.705253016935403e-07,
+      "loss": 0.9101,
+      "step": 7185
+    },
+    {
+      "epoch": 0.17316538619012067,
+      "grad_norm": 1.4857765436172485,
+      "learning_rate": 8.703985397018557e-07,
+      "loss": 0.9447,
+      "step": 7190
+    },
+    {
+      "epoch": 0.17328580718190795,
+      "grad_norm": 1.5675077438354492,
+      "learning_rate": 8.702717777101713e-07,
+      "loss": 0.9187,
+      "step": 7195
+    },
+    {
+      "epoch": 0.17340622817369525,
+      "grad_norm": 1.5084619522094727,
+      "learning_rate": 8.70145015718487e-07,
+      "loss": 0.966,
+      "step": 7200
+    },
+    {
+      "epoch": 0.17352664916548252,
+      "grad_norm": 1.5520141124725342,
+      "learning_rate": 8.700182537268025e-07,
+      "loss": 0.9421,
+      "step": 7205
+    },
+    {
+      "epoch": 0.17364707015726982,
+      "grad_norm": 1.9939416646957397,
+      "learning_rate": 8.698914917351182e-07,
+      "loss": 0.9815,
+      "step": 7210
+    },
+    {
+      "epoch": 0.1737674911490571,
+      "grad_norm": 1.4251264333724976,
+      "learning_rate": 8.697647297434337e-07,
+      "loss": 0.9374,
+      "step": 7215
+    },
+    {
+      "epoch": 0.1738879121408444,
+      "grad_norm": 1.5417747497558594,
+      "learning_rate": 8.696379677517492e-07,
+      "loss": 0.9117,
+      "step": 7220
+    },
+    {
+      "epoch": 0.17400833313263167,
+      "grad_norm": 1.4149068593978882,
+      "learning_rate": 8.695112057600649e-07,
+      "loss": 0.897,
+      "step": 7225
+    },
+    {
+      "epoch": 0.17412875412441897,
+      "grad_norm": 1.4917889833450317,
+      "learning_rate": 8.693844437683805e-07,
+      "loss": 0.9394,
+      "step": 7230
+    },
+    {
+      "epoch": 0.17424917511620625,
+      "grad_norm": 1.6822994947433472,
+      "learning_rate": 8.69257681776696e-07,
+      "loss": 0.8761,
+      "step": 7235
+    },
+    {
+      "epoch": 0.17436959610799355,
+      "grad_norm": 1.4337525367736816,
+      "learning_rate": 8.691309197850116e-07,
+      "loss": 0.8911,
+      "step": 7240
+    },
+    {
+      "epoch": 0.17449001709978085,
+      "grad_norm": 1.5922350883483887,
+      "learning_rate": 8.690041577933273e-07,
+      "loss": 0.9328,
+      "step": 7245
+    },
+    {
+      "epoch": 0.17461043809156812,
+      "grad_norm": 1.6190274953842163,
+      "learning_rate": 8.688773958016428e-07,
+      "loss": 0.952,
+      "step": 7250
+    },
+    {
+      "epoch": 0.17473085908335542,
+      "grad_norm": 1.870335578918457,
+      "learning_rate": 8.687506338099584e-07,
+      "loss": 0.9857,
+      "step": 7255
+    },
+    {
+      "epoch": 0.1748512800751427,
+      "grad_norm": 1.4689422845840454,
+      "learning_rate": 8.68623871818274e-07,
+      "loss": 0.8954,
+      "step": 7260
+    },
+    {
+      "epoch": 0.17497170106693,
+      "grad_norm": 1.566745400428772,
+      "learning_rate": 8.684971098265895e-07,
+      "loss": 0.9357,
+      "step": 7265
+    },
+    {
+      "epoch": 0.17509212205871727,
+      "grad_norm": 1.6396536827087402,
+      "learning_rate": 8.683703478349052e-07,
+      "loss": 0.9431,
+      "step": 7270
+    },
+    {
+      "epoch": 0.17521254305050457,
+      "grad_norm": 1.8179210424423218,
+      "learning_rate": 8.682435858432208e-07,
+      "loss": 0.9463,
+      "step": 7275
+    },
+    {
+      "epoch": 0.17533296404229184,
+      "grad_norm": 2.6301631927490234,
+      "learning_rate": 8.681168238515362e-07,
+      "loss": 0.9387,
+      "step": 7280
+    },
+    {
+      "epoch": 0.17545338503407915,
+      "grad_norm": 1.613966941833496,
+      "learning_rate": 8.679900618598519e-07,
+      "loss": 0.9599,
+      "step": 7285
+    },
+    {
+      "epoch": 0.17557380602586642,
+      "grad_norm": 1.6236636638641357,
+      "learning_rate": 8.678632998681675e-07,
+      "loss": 0.9078,
+      "step": 7290
+    },
+    {
+      "epoch": 0.17569422701765372,
+      "grad_norm": 1.417366862297058,
+      "learning_rate": 8.677365378764831e-07,
+      "loss": 0.9107,
+      "step": 7295
+    },
+    {
+      "epoch": 0.175814648009441,
+      "grad_norm": 1.6419947147369385,
+      "learning_rate": 8.676097758847987e-07,
+      "loss": 0.9224,
+      "step": 7300
+    },
+    {
+      "epoch": 0.1759350690012283,
+      "grad_norm": 1.635498285293579,
+      "learning_rate": 8.674830138931142e-07,
+      "loss": 1.0329,
+      "step": 7305
+    },
+    {
+      "epoch": 0.1760554899930156,
+      "grad_norm": 1.461501121520996,
+      "learning_rate": 8.673562519014298e-07,
+      "loss": 0.9704,
+      "step": 7310
+    },
+    {
+      "epoch": 0.17617591098480287,
+      "grad_norm": 2.0301854610443115,
+      "learning_rate": 8.672294899097454e-07,
+      "loss": 0.957,
+      "step": 7315
+    },
+    {
+      "epoch": 0.17629633197659017,
+      "grad_norm": 1.6653895378112793,
+      "learning_rate": 8.671027279180611e-07,
+      "loss": 1.0023,
+      "step": 7320
+    },
+    {
+      "epoch": 0.17641675296837744,
+      "grad_norm": 1.5719858407974243,
+      "learning_rate": 8.669759659263766e-07,
+      "loss": 0.9648,
+      "step": 7325
+    },
+    {
+      "epoch": 0.17653717396016474,
+      "grad_norm": 1.5789759159088135,
+      "learning_rate": 8.668492039346922e-07,
+      "loss": 0.9642,
+      "step": 7330
+    },
+    {
+      "epoch": 0.17665759495195202,
+      "grad_norm": 1.499147891998291,
+      "learning_rate": 8.667224419430078e-07,
+      "loss": 0.9344,
+      "step": 7335
+    },
+    {
+      "epoch": 0.17677801594373932,
+      "grad_norm": 1.4714581966400146,
+      "learning_rate": 8.665956799513233e-07,
+      "loss": 0.9036,
+      "step": 7340
+    },
+    {
+      "epoch": 0.1768984369355266,
+      "grad_norm": 1.5490918159484863,
+      "learning_rate": 8.66468917959639e-07,
+      "loss": 1.0156,
+      "step": 7345
+    },
+    {
+      "epoch": 0.1770188579273139,
+      "grad_norm": 1.5557063817977905,
+      "learning_rate": 8.663421559679545e-07,
+      "loss": 0.9306,
+      "step": 7350
+    },
+    {
+      "epoch": 0.17713927891910117,
+      "grad_norm": 1.6066994667053223,
+      "learning_rate": 8.662153939762701e-07,
+      "loss": 0.9709,
+      "step": 7355
+    },
+    {
+      "epoch": 0.17725969991088847,
+      "grad_norm": 1.4699970483779907,
+      "learning_rate": 8.660886319845857e-07,
+      "loss": 0.9337,
+      "step": 7360
+    },
+    {
+      "epoch": 0.17738012090267574,
+      "grad_norm": 1.4439870119094849,
+      "learning_rate": 8.659618699929013e-07,
+      "loss": 0.9177,
+      "step": 7365
+    },
+    {
+      "epoch": 0.17750054189446304,
+      "grad_norm": 1.50325608253479,
+      "learning_rate": 8.658351080012169e-07,
+      "loss": 0.9252,
+      "step": 7370
+    },
+    {
+      "epoch": 0.17762096288625034,
+      "grad_norm": 1.4546009302139282,
+      "learning_rate": 8.657083460095324e-07,
+      "loss": 0.9483,
+      "step": 7375
+    },
+    {
+      "epoch": 0.17774138387803762,
+      "grad_norm": 1.5476328134536743,
+      "learning_rate": 8.655815840178481e-07,
+      "loss": 0.9219,
+      "step": 7380
+    },
+    {
+      "epoch": 0.17786180486982492,
+      "grad_norm": 1.4623545408248901,
+      "learning_rate": 8.654548220261636e-07,
+      "loss": 0.9562,
+      "step": 7385
+    },
+    {
+      "epoch": 0.1779822258616122,
+      "grad_norm": 1.5667556524276733,
+      "learning_rate": 8.653280600344793e-07,
+      "loss": 0.9165,
+      "step": 7390
+    },
+    {
+      "epoch": 0.1781026468533995,
+      "grad_norm": 1.402410626411438,
+      "learning_rate": 8.652012980427949e-07,
+      "loss": 0.9527,
+      "step": 7395
+    },
+    {
+      "epoch": 0.17822306784518677,
+      "grad_norm": 1.6034663915634155,
+      "learning_rate": 8.650745360511103e-07,
+      "loss": 0.9819,
+      "step": 7400
+    },
+    {
+      "epoch": 0.17834348883697407,
+      "grad_norm": 1.5540601015090942,
+      "learning_rate": 8.64947774059426e-07,
+      "loss": 0.9365,
+      "step": 7405
+    },
+    {
+      "epoch": 0.17846390982876134,
+      "grad_norm": 1.5132516622543335,
+      "learning_rate": 8.648210120677416e-07,
+      "loss": 0.9918,
+      "step": 7410
+    },
+    {
+      "epoch": 0.17858433082054864,
+      "grad_norm": 1.5507590770721436,
+      "learning_rate": 8.646942500760572e-07,
+      "loss": 0.9234,
+      "step": 7415
+    },
+    {
+      "epoch": 0.17870475181233592,
+      "grad_norm": 1.5630336999893188,
+      "learning_rate": 8.645674880843727e-07,
+      "loss": 0.94,
+      "step": 7420
+    },
+    {
+      "epoch": 0.17882517280412322,
+      "grad_norm": 1.7237917184829712,
+      "learning_rate": 8.644407260926883e-07,
+      "loss": 0.9653,
+      "step": 7425
+    },
+    {
+      "epoch": 0.1789455937959105,
+      "grad_norm": 1.4693753719329834,
+      "learning_rate": 8.643139641010039e-07,
+      "loss": 0.9563,
+      "step": 7430
+    },
+    {
+      "epoch": 0.1790660147876978,
+      "grad_norm": 1.636465072631836,
+      "learning_rate": 8.641872021093195e-07,
+      "loss": 0.9552,
+      "step": 7435
+    },
+    {
+      "epoch": 0.1791864357794851,
+      "grad_norm": 1.3864063024520874,
+      "learning_rate": 8.640604401176352e-07,
+      "loss": 0.9553,
+      "step": 7440
+    },
+    {
+      "epoch": 0.17930685677127237,
+      "grad_norm": 1.6570584774017334,
+      "learning_rate": 8.639336781259506e-07,
+      "loss": 0.9229,
+      "step": 7445
+    },
+    {
+      "epoch": 0.17942727776305967,
+      "grad_norm": 1.4525080919265747,
+      "learning_rate": 8.638069161342662e-07,
+      "loss": 0.9361,
+      "step": 7450
+    },
+    {
+      "epoch": 0.17954769875484694,
+      "grad_norm": 1.5122401714324951,
+      "learning_rate": 8.636801541425819e-07,
+      "loss": 0.9189,
+      "step": 7455
+    },
+    {
+      "epoch": 0.17966811974663424,
+      "grad_norm": 1.5133171081542969,
+      "learning_rate": 8.635533921508974e-07,
+      "loss": 0.9435,
+      "step": 7460
+    },
+    {
+      "epoch": 0.17978854073842152,
+      "grad_norm": 1.5368260145187378,
+      "learning_rate": 8.63426630159213e-07,
+      "loss": 0.9743,
+      "step": 7465
+    },
+    {
+      "epoch": 0.17990896173020882,
+      "grad_norm": 1.7313311100006104,
+      "learning_rate": 8.632998681675286e-07,
+      "loss": 0.916,
+      "step": 7470
+    },
+    {
+      "epoch": 0.1800293827219961,
+      "grad_norm": 1.5005427598953247,
+      "learning_rate": 8.631731061758442e-07,
+      "loss": 0.9366,
+      "step": 7475
+    },
+    {
+      "epoch": 0.1801498037137834,
+      "grad_norm": 1.3697391748428345,
+      "learning_rate": 8.630463441841598e-07,
+      "loss": 0.9251,
+      "step": 7480
+    },
+    {
+      "epoch": 0.18027022470557066,
+      "grad_norm": 1.6520494222640991,
+      "learning_rate": 8.629195821924754e-07,
+      "loss": 0.9587,
+      "step": 7485
+    },
+    {
+      "epoch": 0.18039064569735797,
+      "grad_norm": 1.641788363456726,
+      "learning_rate": 8.627928202007909e-07,
+      "loss": 0.9348,
+      "step": 7490
+    },
+    {
+      "epoch": 0.18051106668914524,
+      "grad_norm": 1.6810346841812134,
+      "learning_rate": 8.626660582091065e-07,
+      "loss": 0.9282,
+      "step": 7495
+    },
+    {
+      "epoch": 0.18063148768093254,
+      "grad_norm": 1.5090734958648682,
+      "learning_rate": 8.625392962174222e-07,
+      "loss": 0.9644,
+      "step": 7500
+    },
+    {
+      "epoch": 0.18075190867271984,
+      "grad_norm": 1.4409652948379517,
+      "learning_rate": 8.624125342257377e-07,
+      "loss": 0.8875,
+      "step": 7505
+    },
+    {
+      "epoch": 0.18087232966450711,
+      "grad_norm": 1.6661747694015503,
+      "learning_rate": 8.622857722340534e-07,
+      "loss": 0.9867,
+      "step": 7510
+    },
+    {
+      "epoch": 0.18099275065629442,
+      "grad_norm": 1.784266471862793,
+      "learning_rate": 8.621590102423689e-07,
+      "loss": 0.9067,
+      "step": 7515
+    },
+    {
+      "epoch": 0.1811131716480817,
+      "grad_norm": 1.497228741645813,
+      "learning_rate": 8.620322482506844e-07,
+      "loss": 0.9663,
+      "step": 7520
+    },
+    {
+      "epoch": 0.181233592639869,
+      "grad_norm": 1.8661032915115356,
+      "learning_rate": 8.619054862590001e-07,
+      "loss": 0.9239,
+      "step": 7525
+    },
+    {
+      "epoch": 0.18135401363165626,
+      "grad_norm": 1.5789985656738281,
+      "learning_rate": 8.617787242673157e-07,
+      "loss": 1.0048,
+      "step": 7530
+    },
+    {
+      "epoch": 0.18147443462344356,
+      "grad_norm": 1.6937061548233032,
+      "learning_rate": 8.616519622756312e-07,
+      "loss": 0.9206,
+      "step": 7535
+    },
+    {
+      "epoch": 0.18159485561523084,
+      "grad_norm": 1.5689666271209717,
+      "learning_rate": 8.615252002839468e-07,
+      "loss": 0.9583,
+      "step": 7540
+    },
+    {
+      "epoch": 0.18171527660701814,
+      "grad_norm": 1.628058671951294,
+      "learning_rate": 8.613984382922624e-07,
+      "loss": 0.9669,
+      "step": 7545
+    },
+    {
+      "epoch": 0.1818356975988054,
+      "grad_norm": 1.5606167316436768,
+      "learning_rate": 8.61271676300578e-07,
+      "loss": 0.9188,
+      "step": 7550
+    },
+    {
+      "epoch": 0.1819561185905927,
+      "grad_norm": 1.4785312414169312,
+      "learning_rate": 8.611449143088936e-07,
+      "loss": 0.9194,
+      "step": 7555
+    },
+    {
+      "epoch": 0.18207653958238,
+      "grad_norm": 1.4869170188903809,
+      "learning_rate": 8.610181523172092e-07,
+      "loss": 0.9779,
+      "step": 7560
+    },
+    {
+      "epoch": 0.1821969605741673,
+      "grad_norm": 1.5507047176361084,
+      "learning_rate": 8.608913903255247e-07,
+      "loss": 0.9377,
+      "step": 7565
+    },
+    {
+      "epoch": 0.1823173815659546,
+      "grad_norm": 1.3956730365753174,
+      "learning_rate": 8.607646283338403e-07,
+      "loss": 0.9261,
+      "step": 7570
+    },
+    {
+      "epoch": 0.18243780255774186,
+      "grad_norm": 1.5166481733322144,
+      "learning_rate": 8.60637866342156e-07,
+      "loss": 0.9878,
+      "step": 7575
+    },
+    {
+      "epoch": 0.18255822354952916,
+      "grad_norm": 1.4868241548538208,
+      "learning_rate": 8.605111043504714e-07,
+      "loss": 0.8817,
+      "step": 7580
+    },
+    {
+      "epoch": 0.18267864454131644,
+      "grad_norm": 1.660776138305664,
+      "learning_rate": 8.603843423587871e-07,
+      "loss": 0.9763,
+      "step": 7585
+    },
+    {
+      "epoch": 0.18279906553310374,
+      "grad_norm": 1.8276817798614502,
+      "learning_rate": 8.602575803671027e-07,
+      "loss": 0.9639,
+      "step": 7590
+    },
+    {
+      "epoch": 0.182919486524891,
+      "grad_norm": 1.818725347518921,
+      "learning_rate": 8.601308183754183e-07,
+      "loss": 0.9588,
+      "step": 7595
+    },
+    {
+      "epoch": 0.1830399075166783,
+      "grad_norm": 1.521095871925354,
+      "learning_rate": 8.600040563837339e-07,
+      "loss": 0.9631,
+      "step": 7600
+    },
+    {
+      "epoch": 0.1831603285084656,
+      "grad_norm": 1.6158627271652222,
+      "learning_rate": 8.598772943920494e-07,
+      "loss": 0.9082,
+      "step": 7605
+    },
+    {
+      "epoch": 0.1832807495002529,
+      "grad_norm": 1.6289167404174805,
+      "learning_rate": 8.59750532400365e-07,
+      "loss": 0.9901,
+      "step": 7610
+    },
+    {
+      "epoch": 0.18340117049204016,
+      "grad_norm": 1.6304922103881836,
+      "learning_rate": 8.596237704086806e-07,
+      "loss": 0.9861,
+      "step": 7615
+    },
+    {
+      "epoch": 0.18352159148382746,
+      "grad_norm": 1.5539021492004395,
+      "learning_rate": 8.594970084169963e-07,
+      "loss": 0.9694,
+      "step": 7620
+    },
+    {
+      "epoch": 0.18364201247561474,
+      "grad_norm": 1.6585330963134766,
+      "learning_rate": 8.593702464253118e-07,
+      "loss": 0.9466,
+      "step": 7625
+    },
+    {
+      "epoch": 0.18376243346740204,
+      "grad_norm": 1.7435808181762695,
+      "learning_rate": 8.592434844336273e-07,
+      "loss": 0.9547,
+      "step": 7630
+    },
+    {
+      "epoch": 0.18388285445918934,
+      "grad_norm": 1.4508938789367676,
+      "learning_rate": 8.59116722441943e-07,
+      "loss": 0.9705,
+      "step": 7635
+    },
+    {
+      "epoch": 0.1840032754509766,
+      "grad_norm": 1.6086461544036865,
+      "learning_rate": 8.589899604502585e-07,
+      "loss": 0.8958,
+      "step": 7640
+    },
+    {
+      "epoch": 0.1841236964427639,
+      "grad_norm": 1.4307724237442017,
+      "learning_rate": 8.588631984585742e-07,
+      "loss": 0.9562,
+      "step": 7645
+    },
+    {
+      "epoch": 0.18424411743455119,
+      "grad_norm": 1.552018642425537,
+      "learning_rate": 8.587364364668897e-07,
+      "loss": 0.933,
+      "step": 7650
+    },
+    {
+      "epoch": 0.1843645384263385,
+      "grad_norm": 1.5186680555343628,
+      "learning_rate": 8.586096744752052e-07,
+      "loss": 0.9275,
+      "step": 7655
+    },
+    {
+      "epoch": 0.18448495941812576,
+      "grad_norm": 1.5889519453048706,
+      "learning_rate": 8.584829124835209e-07,
+      "loss": 0.9499,
+      "step": 7660
+    },
+    {
+      "epoch": 0.18460538040991306,
+      "grad_norm": 1.6468712091445923,
+      "learning_rate": 8.583561504918365e-07,
+      "loss": 0.9632,
+      "step": 7665
+    },
+    {
+      "epoch": 0.18472580140170033,
+      "grad_norm": 1.4589017629623413,
+      "learning_rate": 8.582293885001521e-07,
+      "loss": 0.9375,
+      "step": 7670
+    },
+    {
+      "epoch": 0.18484622239348764,
+      "grad_norm": 1.484344482421875,
+      "learning_rate": 8.581026265084676e-07,
+      "loss": 0.9317,
+      "step": 7675
+    },
+    {
+      "epoch": 0.1849666433852749,
+      "grad_norm": 1.5107464790344238,
+      "learning_rate": 8.579758645167833e-07,
+      "loss": 0.9496,
+      "step": 7680
+    },
+    {
+      "epoch": 0.1850870643770622,
+      "grad_norm": 1.4735205173492432,
+      "learning_rate": 8.578491025250988e-07,
+      "loss": 0.9515,
+      "step": 7685
+    },
+    {
+      "epoch": 0.1852074853688495,
+      "grad_norm": 1.405005931854248,
+      "learning_rate": 8.577223405334144e-07,
+      "loss": 0.9233,
+      "step": 7690
+    },
+    {
+      "epoch": 0.18532790636063678,
+      "grad_norm": 1.4967010021209717,
+      "learning_rate": 8.575955785417301e-07,
+      "loss": 0.9095,
+      "step": 7695
+    },
+    {
+      "epoch": 0.18544832735242409,
+      "grad_norm": 1.5843182802200317,
+      "learning_rate": 8.574688165500455e-07,
+      "loss": 0.9234,
+      "step": 7700
+    },
+    {
+      "epoch": 0.18556874834421136,
+      "grad_norm": 1.563092827796936,
+      "learning_rate": 8.573420545583612e-07,
+      "loss": 0.9395,
+      "step": 7705
+    },
+    {
+      "epoch": 0.18568916933599866,
+      "grad_norm": 1.7785170078277588,
+      "learning_rate": 8.572152925666768e-07,
+      "loss": 0.9331,
+      "step": 7710
+    },
+    {
+      "epoch": 0.18580959032778593,
+      "grad_norm": 1.48356294631958,
+      "learning_rate": 8.570885305749924e-07,
+      "loss": 0.9174,
+      "step": 7715
+    },
+    {
+      "epoch": 0.18593001131957324,
+      "grad_norm": 1.706422209739685,
+      "learning_rate": 8.569617685833079e-07,
+      "loss": 0.9301,
+      "step": 7720
+    },
+    {
+      "epoch": 0.1860504323113605,
+      "grad_norm": 1.7067279815673828,
+      "learning_rate": 8.568350065916235e-07,
+      "loss": 0.9196,
+      "step": 7725
+    },
+    {
+      "epoch": 0.1861708533031478,
+      "grad_norm": 1.6843154430389404,
+      "learning_rate": 8.567082445999391e-07,
+      "loss": 0.9272,
+      "step": 7730
+    },
+    {
+      "epoch": 0.18629127429493508,
+      "grad_norm": 1.6786079406738281,
+      "learning_rate": 8.565814826082547e-07,
+      "loss": 0.9272,
+      "step": 7735
+    },
+    {
+      "epoch": 0.18641169528672238,
+      "grad_norm": 1.850604772567749,
+      "learning_rate": 8.564547206165704e-07,
+      "loss": 0.9236,
+      "step": 7740
+    },
+    {
+      "epoch": 0.18653211627850966,
+      "grad_norm": 1.6428070068359375,
+      "learning_rate": 8.563279586248858e-07,
+      "loss": 0.9401,
+      "step": 7745
+    },
+    {
+      "epoch": 0.18665253727029696,
+      "grad_norm": 1.5928632020950317,
+      "learning_rate": 8.562011966332014e-07,
+      "loss": 0.9904,
+      "step": 7750
+    },
+    {
+      "epoch": 0.18677295826208426,
+      "grad_norm": 1.7247638702392578,
+      "learning_rate": 8.560744346415171e-07,
+      "loss": 0.9355,
+      "step": 7755
+    },
+    {
+      "epoch": 0.18689337925387153,
+      "grad_norm": 1.6918985843658447,
+      "learning_rate": 8.559476726498326e-07,
+      "loss": 0.9348,
+      "step": 7760
+    },
+    {
+      "epoch": 0.18701380024565883,
+      "grad_norm": 1.6758086681365967,
+      "learning_rate": 8.558209106581482e-07,
+      "loss": 0.9397,
+      "step": 7765
+    },
+    {
+      "epoch": 0.1871342212374461,
+      "grad_norm": 1.3978140354156494,
+      "learning_rate": 8.556941486664638e-07,
+      "loss": 0.9691,
+      "step": 7770
+    },
+    {
+      "epoch": 0.1872546422292334,
+      "grad_norm": 1.5308520793914795,
+      "learning_rate": 8.555673866747793e-07,
+      "loss": 0.9205,
+      "step": 7775
+    },
+    {
+      "epoch": 0.18737506322102068,
+      "grad_norm": 1.3909852504730225,
+      "learning_rate": 8.55440624683095e-07,
+      "loss": 0.9421,
+      "step": 7780
+    },
+    {
+      "epoch": 0.18749548421280798,
+      "grad_norm": 1.5220245122909546,
+      "learning_rate": 8.553138626914106e-07,
+      "loss": 0.9511,
+      "step": 7785
+    },
+    {
+      "epoch": 0.18761590520459526,
+      "grad_norm": 1.8534610271453857,
+      "learning_rate": 8.551871006997261e-07,
+      "loss": 0.9701,
+      "step": 7790
+    },
+    {
+      "epoch": 0.18773632619638256,
+      "grad_norm": 1.5921791791915894,
+      "learning_rate": 8.550603387080417e-07,
+      "loss": 0.9832,
+      "step": 7795
+    },
+    {
+      "epoch": 0.18785674718816983,
+      "grad_norm": 1.5496329069137573,
+      "learning_rate": 8.549335767163574e-07,
+      "loss": 0.9646,
+      "step": 7800
+    },
+    {
+      "epoch": 0.18797716817995713,
+      "grad_norm": 1.608414888381958,
+      "learning_rate": 8.548068147246729e-07,
+      "loss": 0.9486,
+      "step": 7805
+    },
+    {
+      "epoch": 0.1880975891717444,
+      "grad_norm": 1.4896926879882812,
+      "learning_rate": 8.546800527329884e-07,
+      "loss": 0.9662,
+      "step": 7810
+    },
+    {
+      "epoch": 0.1882180101635317,
+      "grad_norm": 1.7178910970687866,
+      "learning_rate": 8.545532907413041e-07,
+      "loss": 0.949,
+      "step": 7815
+    },
+    {
+      "epoch": 0.188338431155319,
+      "grad_norm": 1.6171796321868896,
+      "learning_rate": 8.544265287496196e-07,
+      "loss": 0.9203,
+      "step": 7820
+    },
+    {
+      "epoch": 0.18845885214710628,
+      "grad_norm": 1.8074618577957153,
+      "learning_rate": 8.542997667579353e-07,
+      "loss": 0.925,
+      "step": 7825
+    },
+    {
+      "epoch": 0.18857927313889358,
+      "grad_norm": 1.5316404104232788,
+      "learning_rate": 8.541730047662509e-07,
+      "loss": 0.9347,
+      "step": 7830
+    },
+    {
+      "epoch": 0.18869969413068086,
+      "grad_norm": 1.57062566280365,
+      "learning_rate": 8.540462427745663e-07,
+      "loss": 0.9009,
+      "step": 7835
+    },
+    {
+      "epoch": 0.18882011512246816,
+      "grad_norm": 1.4911035299301147,
+      "learning_rate": 8.53919480782882e-07,
+      "loss": 0.9158,
+      "step": 7840
+    },
+    {
+      "epoch": 0.18894053611425543,
+      "grad_norm": 1.526751160621643,
+      "learning_rate": 8.537927187911976e-07,
+      "loss": 0.9323,
+      "step": 7845
+    },
+    {
+      "epoch": 0.18906095710604273,
+      "grad_norm": 1.5808849334716797,
+      "learning_rate": 8.536659567995132e-07,
+      "loss": 0.9654,
+      "step": 7850
+    },
+    {
+      "epoch": 0.18918137809783,
+      "grad_norm": 1.5629630088806152,
+      "learning_rate": 8.535391948078288e-07,
+      "loss": 0.9316,
+      "step": 7855
+    },
+    {
+      "epoch": 0.1893017990896173,
+      "grad_norm": 1.66767156124115,
+      "learning_rate": 8.534124328161444e-07,
+      "loss": 0.9788,
+      "step": 7860
+    },
+    {
+      "epoch": 0.18942222008140458,
+      "grad_norm": 1.495460033416748,
+      "learning_rate": 8.532856708244599e-07,
+      "loss": 0.8883,
+      "step": 7865
+    },
+    {
+      "epoch": 0.18954264107319188,
+      "grad_norm": 1.5910550355911255,
+      "learning_rate": 8.531589088327755e-07,
+      "loss": 0.9311,
+      "step": 7870
+    },
+    {
+      "epoch": 0.18966306206497915,
+      "grad_norm": 1.4263969659805298,
+      "learning_rate": 8.530321468410912e-07,
+      "loss": 0.9181,
+      "step": 7875
+    },
+    {
+      "epoch": 0.18978348305676646,
+      "grad_norm": 1.524699330329895,
+      "learning_rate": 8.529053848494066e-07,
+      "loss": 0.9138,
+      "step": 7880
+    },
+    {
+      "epoch": 0.18990390404855376,
+      "grad_norm": 1.612310767173767,
+      "learning_rate": 8.527786228577223e-07,
+      "loss": 0.9042,
+      "step": 7885
+    },
+    {
+      "epoch": 0.19002432504034103,
+      "grad_norm": 1.527094841003418,
+      "learning_rate": 8.526518608660379e-07,
+      "loss": 0.9604,
+      "step": 7890
+    },
+    {
+      "epoch": 0.19014474603212833,
+      "grad_norm": 1.7105774879455566,
+      "learning_rate": 8.525250988743534e-07,
+      "loss": 0.975,
+      "step": 7895
+    },
+    {
+      "epoch": 0.1902651670239156,
+      "grad_norm": 1.6466574668884277,
+      "learning_rate": 8.523983368826691e-07,
+      "loss": 0.944,
+      "step": 7900
+    },
+    {
+      "epoch": 0.1903855880157029,
+      "grad_norm": 1.4922170639038086,
+      "learning_rate": 8.522715748909846e-07,
+      "loss": 0.9308,
+      "step": 7905
+    },
+    {
+      "epoch": 0.19050600900749018,
+      "grad_norm": 1.6497938632965088,
+      "learning_rate": 8.521448128993002e-07,
+      "loss": 0.9541,
+      "step": 7910
+    },
+    {
+      "epoch": 0.19062642999927748,
+      "grad_norm": 1.5365254878997803,
+      "learning_rate": 8.520180509076158e-07,
+      "loss": 0.9478,
+      "step": 7915
+    },
+    {
+      "epoch": 0.19074685099106475,
+      "grad_norm": 1.4882341623306274,
+      "learning_rate": 8.518912889159315e-07,
+      "loss": 0.984,
+      "step": 7920
+    },
+    {
+      "epoch": 0.19086727198285205,
+      "grad_norm": 1.5878196954727173,
+      "learning_rate": 8.517645269242469e-07,
+      "loss": 0.9844,
+      "step": 7925
+    },
+    {
+      "epoch": 0.19098769297463933,
+      "grad_norm": 1.6832592487335205,
+      "learning_rate": 8.516377649325625e-07,
+      "loss": 0.9382,
+      "step": 7930
+    },
+    {
+      "epoch": 0.19110811396642663,
+      "grad_norm": 1.73088800907135,
+      "learning_rate": 8.515110029408782e-07,
+      "loss": 0.9062,
+      "step": 7935
+    },
+    {
+      "epoch": 0.1912285349582139,
+      "grad_norm": 1.6562163829803467,
+      "learning_rate": 8.513842409491938e-07,
+      "loss": 0.9283,
+      "step": 7940
+    },
+    {
+      "epoch": 0.1913489559500012,
+      "grad_norm": 1.4478557109832764,
+      "learning_rate": 8.512574789575094e-07,
+      "loss": 0.9725,
+      "step": 7945
+    },
+    {
+      "epoch": 0.1914693769417885,
+      "grad_norm": 1.61317777633667,
+      "learning_rate": 8.511307169658249e-07,
+      "loss": 0.9291,
+      "step": 7950
+    },
+    {
+      "epoch": 0.19158979793357578,
+      "grad_norm": 1.579496145248413,
+      "learning_rate": 8.510039549741405e-07,
+      "loss": 0.9744,
+      "step": 7955
+    },
+    {
+      "epoch": 0.19171021892536308,
+      "grad_norm": 1.7063086032867432,
+      "learning_rate": 8.508771929824561e-07,
+      "loss": 0.9523,
+      "step": 7960
+    },
+    {
+      "epoch": 0.19183063991715035,
+      "grad_norm": 1.4805172681808472,
+      "learning_rate": 8.507504309907717e-07,
+      "loss": 0.9667,
+      "step": 7965
+    },
+    {
+      "epoch": 0.19195106090893765,
+      "grad_norm": 1.4771111011505127,
+      "learning_rate": 8.506236689990874e-07,
+      "loss": 0.9738,
+      "step": 7970
+    },
+    {
+      "epoch": 0.19207148190072493,
+      "grad_norm": 1.4734350442886353,
+      "learning_rate": 8.504969070074028e-07,
+      "loss": 0.9327,
+      "step": 7975
+    },
+    {
+      "epoch": 0.19219190289251223,
+      "grad_norm": 1.5960177183151245,
+      "learning_rate": 8.503701450157185e-07,
+      "loss": 0.9299,
+      "step": 7980
+    },
+    {
+      "epoch": 0.1923123238842995,
+      "grad_norm": 1.7017934322357178,
+      "learning_rate": 8.502433830240341e-07,
+      "loss": 0.9444,
+      "step": 7985
+    },
+    {
+      "epoch": 0.1924327448760868,
+      "grad_norm": 1.5495920181274414,
+      "learning_rate": 8.501166210323496e-07,
+      "loss": 0.9296,
+      "step": 7990
+    },
+    {
+      "epoch": 0.19255316586787408,
+      "grad_norm": 1.8156256675720215,
+      "learning_rate": 8.499898590406652e-07,
+      "loss": 0.9251,
+      "step": 7995
+    },
+    {
+      "epoch": 0.19267358685966138,
+      "grad_norm": 1.6993896961212158,
+      "learning_rate": 8.498630970489808e-07,
+      "loss": 0.9591,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19279400785144865,
+      "grad_norm": 1.6826540231704712,
+      "learning_rate": 8.497363350572964e-07,
+      "loss": 0.9337,
+      "step": 8005
+    },
+    {
+      "epoch": 0.19291442884323595,
+      "grad_norm": 1.5303553342819214,
+      "learning_rate": 8.49609573065612e-07,
+      "loss": 0.9317,
+      "step": 8010
+    },
+    {
+      "epoch": 0.19303484983502325,
+      "grad_norm": 1.7423923015594482,
+      "learning_rate": 8.494828110739276e-07,
+      "loss": 0.9019,
+      "step": 8015
+    },
+    {
+      "epoch": 0.19315527082681053,
+      "grad_norm": 1.838249921798706,
+      "learning_rate": 8.493560490822431e-07,
+      "loss": 0.988,
+      "step": 8020
+    },
+    {
+      "epoch": 0.19327569181859783,
+      "grad_norm": 1.5801361799240112,
+      "learning_rate": 8.492292870905587e-07,
+      "loss": 0.9737,
+      "step": 8025
+    },
+    {
+      "epoch": 0.1933961128103851,
+      "grad_norm": 1.6283152103424072,
+      "learning_rate": 8.491025250988744e-07,
+      "loss": 0.9168,
+      "step": 8030
+    },
+    {
+      "epoch": 0.1935165338021724,
+      "grad_norm": 1.3490309715270996,
+      "learning_rate": 8.489757631071899e-07,
+      "loss": 0.9394,
+      "step": 8035
+    },
+    {
+      "epoch": 0.19363695479395968,
+      "grad_norm": 1.5663366317749023,
+      "learning_rate": 8.488490011155056e-07,
+      "loss": 0.906,
+      "step": 8040
+    },
+    {
+      "epoch": 0.19375737578574698,
+      "grad_norm": 1.4401100873947144,
+      "learning_rate": 8.487222391238211e-07,
+      "loss": 0.9805,
+      "step": 8045
+    },
+    {
+      "epoch": 0.19387779677753425,
+      "grad_norm": 1.521753191947937,
+      "learning_rate": 8.485954771321366e-07,
+      "loss": 0.9311,
+      "step": 8050
+    },
+    {
+      "epoch": 0.19399821776932155,
+      "grad_norm": 1.6854158639907837,
+      "learning_rate": 8.484687151404523e-07,
+      "loss": 0.9865,
+      "step": 8055
+    },
+    {
+      "epoch": 0.19411863876110882,
+      "grad_norm": 1.4435174465179443,
+      "learning_rate": 8.483419531487679e-07,
+      "loss": 0.9494,
+      "step": 8060
+    },
+    {
+      "epoch": 0.19423905975289613,
+      "grad_norm": 1.5255597829818726,
+      "learning_rate": 8.482151911570834e-07,
+      "loss": 0.9435,
+      "step": 8065
+    },
+    {
+      "epoch": 0.1943594807446834,
+      "grad_norm": 1.3990375995635986,
+      "learning_rate": 8.48088429165399e-07,
+      "loss": 0.9662,
+      "step": 8070
+    },
+    {
+      "epoch": 0.1944799017364707,
+      "grad_norm": 1.5630102157592773,
+      "learning_rate": 8.479616671737146e-07,
+      "loss": 0.9869,
+      "step": 8075
+    },
+    {
+      "epoch": 0.194600322728258,
+      "grad_norm": 1.6120853424072266,
+      "learning_rate": 8.478349051820302e-07,
+      "loss": 0.9512,
+      "step": 8080
+    },
+    {
+      "epoch": 0.19472074372004528,
+      "grad_norm": 1.422527551651001,
+      "learning_rate": 8.477081431903458e-07,
+      "loss": 0.9422,
+      "step": 8085
+    },
+    {
+      "epoch": 0.19484116471183258,
+      "grad_norm": 1.715245246887207,
+      "learning_rate": 8.475813811986614e-07,
+      "loss": 0.9268,
+      "step": 8090
+    },
+    {
+      "epoch": 0.19496158570361985,
+      "grad_norm": 1.4960979223251343,
+      "learning_rate": 8.474546192069769e-07,
+      "loss": 0.9693,
+      "step": 8095
+    },
+    {
+      "epoch": 0.19508200669540715,
+      "grad_norm": 1.811741828918457,
+      "learning_rate": 8.473278572152925e-07,
+      "loss": 0.9967,
+      "step": 8100
+    },
+    {
+      "epoch": 0.19520242768719442,
+      "grad_norm": 1.8520315885543823,
+      "learning_rate": 8.472010952236082e-07,
+      "loss": 0.9514,
+      "step": 8105
+    },
+    {
+      "epoch": 0.19532284867898173,
+      "grad_norm": 1.4953160285949707,
+      "learning_rate": 8.470743332319236e-07,
+      "loss": 0.9414,
+      "step": 8110
+    },
+    {
+      "epoch": 0.195443269670769,
+      "grad_norm": 1.5047308206558228,
+      "learning_rate": 8.469475712402393e-07,
+      "loss": 0.9273,
+      "step": 8115
+    },
+    {
+      "epoch": 0.1955636906625563,
+      "grad_norm": 1.5107359886169434,
+      "learning_rate": 8.468208092485549e-07,
+      "loss": 0.9351,
+      "step": 8120
+    },
+    {
+      "epoch": 0.19568411165434357,
+      "grad_norm": 1.3560688495635986,
+      "learning_rate": 8.466940472568705e-07,
+      "loss": 0.9283,
+      "step": 8125
+    },
+    {
+      "epoch": 0.19580453264613087,
+      "grad_norm": 2.0308735370635986,
+      "learning_rate": 8.465672852651861e-07,
+      "loss": 0.9829,
+      "step": 8130
+    },
+    {
+      "epoch": 0.19592495363791818,
+      "grad_norm": 1.367697834968567,
+      "learning_rate": 8.464405232735016e-07,
+      "loss": 0.9557,
+      "step": 8135
+    },
+    {
+      "epoch": 0.19604537462970545,
+      "grad_norm": 1.5483380556106567,
+      "learning_rate": 8.463137612818172e-07,
+      "loss": 0.9502,
+      "step": 8140
+    },
+    {
+      "epoch": 0.19616579562149275,
+      "grad_norm": 1.533666729927063,
+      "learning_rate": 8.461869992901328e-07,
+      "loss": 0.9107,
+      "step": 8145
+    },
+    {
+      "epoch": 0.19628621661328002,
+      "grad_norm": 1.575505018234253,
+      "learning_rate": 8.460602372984485e-07,
+      "loss": 0.9449,
+      "step": 8150
+    },
+    {
+      "epoch": 0.19640663760506732,
+      "grad_norm": 1.605459213256836,
+      "learning_rate": 8.45933475306764e-07,
+      "loss": 0.9894,
+      "step": 8155
+    },
+    {
+      "epoch": 0.1965270585968546,
+      "grad_norm": 1.5962629318237305,
+      "learning_rate": 8.458067133150795e-07,
+      "loss": 0.8916,
+      "step": 8160
+    },
+    {
+      "epoch": 0.1966474795886419,
+      "grad_norm": 1.4973393678665161,
+      "learning_rate": 8.456799513233952e-07,
+      "loss": 0.9511,
+      "step": 8165
+    },
+    {
+      "epoch": 0.19676790058042917,
+      "grad_norm": 1.5095032453536987,
+      "learning_rate": 8.455531893317107e-07,
+      "loss": 0.9788,
+      "step": 8170
+    },
+    {
+      "epoch": 0.19688832157221647,
+      "grad_norm": 1.900133490562439,
+      "learning_rate": 8.454264273400264e-07,
+      "loss": 0.9538,
+      "step": 8175
+    },
+    {
+      "epoch": 0.19700874256400375,
+      "grad_norm": 1.530402421951294,
+      "learning_rate": 8.452996653483419e-07,
+      "loss": 0.9698,
+      "step": 8180
+    },
+    {
+      "epoch": 0.19712916355579105,
+      "grad_norm": 1.551681637763977,
+      "learning_rate": 8.451729033566574e-07,
+      "loss": 0.8823,
+      "step": 8185
+    },
+    {
+      "epoch": 0.19724958454757832,
+      "grad_norm": 1.61061429977417,
+      "learning_rate": 8.450461413649731e-07,
+      "loss": 0.9481,
+      "step": 8190
+    },
+    {
+      "epoch": 0.19737000553936562,
+      "grad_norm": 1.6920380592346191,
+      "learning_rate": 8.449193793732887e-07,
+      "loss": 0.9234,
+      "step": 8195
+    },
+    {
+      "epoch": 0.19749042653115292,
+      "grad_norm": 1.539293885231018,
+      "learning_rate": 8.447926173816043e-07,
+      "loss": 0.9832,
+      "step": 8200
+    },
+    {
+      "epoch": 0.1976108475229402,
+      "grad_norm": 1.5610792636871338,
+      "learning_rate": 8.446658553899198e-07,
+      "loss": 0.9322,
+      "step": 8205
+    },
+    {
+      "epoch": 0.1977312685147275,
+      "grad_norm": 1.4349342584609985,
+      "learning_rate": 8.445390933982355e-07,
+      "loss": 0.9551,
+      "step": 8210
+    },
+    {
+      "epoch": 0.19785168950651477,
+      "grad_norm": 1.6934919357299805,
+      "learning_rate": 8.44412331406551e-07,
+      "loss": 0.9032,
+      "step": 8215
+    },
+    {
+      "epoch": 0.19797211049830207,
+      "grad_norm": 1.657364845275879,
+      "learning_rate": 8.442855694148666e-07,
+      "loss": 0.9226,
+      "step": 8220
+    },
+    {
+      "epoch": 0.19809253149008935,
+      "grad_norm": 1.3018603324890137,
+      "learning_rate": 8.441588074231823e-07,
+      "loss": 0.9307,
+      "step": 8225
+    },
+    {
+      "epoch": 0.19821295248187665,
+      "grad_norm": 1.4067305326461792,
+      "learning_rate": 8.440320454314977e-07,
+      "loss": 0.9819,
+      "step": 8230
+    },
+    {
+      "epoch": 0.19833337347366392,
+      "grad_norm": 1.7177774906158447,
+      "learning_rate": 8.439052834398134e-07,
+      "loss": 0.9877,
+      "step": 8235
+    },
+    {
+      "epoch": 0.19845379446545122,
+      "grad_norm": 1.5552080869674683,
+      "learning_rate": 8.43778521448129e-07,
+      "loss": 0.9211,
+      "step": 8240
+    },
+    {
+      "epoch": 0.1985742154572385,
+      "grad_norm": 1.6212966442108154,
+      "learning_rate": 8.436517594564446e-07,
+      "loss": 0.882,
+      "step": 8245
+    },
+    {
+      "epoch": 0.1986946364490258,
+      "grad_norm": 1.4554401636123657,
+      "learning_rate": 8.435249974647601e-07,
+      "loss": 0.9842,
+      "step": 8250
+    },
+    {
+      "epoch": 0.19881505744081307,
+      "grad_norm": 1.6188297271728516,
+      "learning_rate": 8.433982354730757e-07,
+      "loss": 0.9164,
+      "step": 8255
+    },
+    {
+      "epoch": 0.19893547843260037,
+      "grad_norm": 1.503028154373169,
+      "learning_rate": 8.432714734813913e-07,
+      "loss": 0.9838,
+      "step": 8260
+    },
+    {
+      "epoch": 0.19905589942438767,
+      "grad_norm": 1.3847625255584717,
+      "learning_rate": 8.431447114897069e-07,
+      "loss": 0.9386,
+      "step": 8265
+    },
+    {
+      "epoch": 0.19917632041617495,
+      "grad_norm": 1.653942346572876,
+      "learning_rate": 8.430179494980226e-07,
+      "loss": 0.9559,
+      "step": 8270
+    },
+    {
+      "epoch": 0.19929674140796225,
+      "grad_norm": 1.5789453983306885,
+      "learning_rate": 8.42891187506338e-07,
+      "loss": 0.9565,
+      "step": 8275
+    },
+    {
+      "epoch": 0.19941716239974952,
+      "grad_norm": 1.4539376497268677,
+      "learning_rate": 8.427644255146536e-07,
+      "loss": 0.9697,
+      "step": 8280
+    },
+    {
+      "epoch": 0.19953758339153682,
+      "grad_norm": 1.746583342552185,
+      "learning_rate": 8.426376635229693e-07,
+      "loss": 1.0047,
+      "step": 8285
+    },
+    {
+      "epoch": 0.1996580043833241,
+      "grad_norm": 2.238571882247925,
+      "learning_rate": 8.425109015312848e-07,
+      "loss": 0.9372,
+      "step": 8290
+    },
+    {
+      "epoch": 0.1997784253751114,
+      "grad_norm": 1.5720535516738892,
+      "learning_rate": 8.423841395396004e-07,
+      "loss": 0.8936,
+      "step": 8295
+    },
+    {
+      "epoch": 0.19989884636689867,
+      "grad_norm": 1.8832333087921143,
+      "learning_rate": 8.42257377547916e-07,
+      "loss": 1.0061,
+      "step": 8300
+    },
+    {
+      "epoch": 0.20001926735868597,
+      "grad_norm": 1.635061264038086,
+      "learning_rate": 8.421306155562315e-07,
+      "loss": 0.98,
+      "step": 8305
+    },
+    {
+      "epoch": 0.20013968835047324,
+      "grad_norm": 1.7162737846374512,
+      "learning_rate": 8.420038535645472e-07,
+      "loss": 0.9042,
+      "step": 8310
+    },
+    {
+      "epoch": 0.20026010934226054,
+      "grad_norm": 1.5249680280685425,
+      "learning_rate": 8.418770915728628e-07,
+      "loss": 0.9422,
+      "step": 8315
+    },
+    {
+      "epoch": 0.20038053033404782,
+      "grad_norm": 1.5159083604812622,
+      "learning_rate": 8.417503295811783e-07,
+      "loss": 0.9585,
+      "step": 8320
+    },
+    {
+      "epoch": 0.20050095132583512,
+      "grad_norm": 1.547275424003601,
+      "learning_rate": 8.416235675894939e-07,
+      "loss": 0.9712,
+      "step": 8325
+    },
+    {
+      "epoch": 0.20062137231762242,
+      "grad_norm": 1.4447563886642456,
+      "learning_rate": 8.414968055978096e-07,
+      "loss": 0.9383,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2007417933094097,
+      "grad_norm": 1.4032526016235352,
+      "learning_rate": 8.413700436061251e-07,
+      "loss": 0.987,
+      "step": 8335
+    },
+    {
+      "epoch": 0.200862214301197,
+      "grad_norm": 1.5580028295516968,
+      "learning_rate": 8.412432816144407e-07,
+      "loss": 0.8971,
+      "step": 8340
+    },
+    {
+      "epoch": 0.20098263529298427,
+      "grad_norm": 1.5366116762161255,
+      "learning_rate": 8.411165196227563e-07,
+      "loss": 0.9864,
+      "step": 8345
+    },
+    {
+      "epoch": 0.20110305628477157,
+      "grad_norm": 1.590053677558899,
+      "learning_rate": 8.409897576310718e-07,
+      "loss": 0.9343,
+      "step": 8350
+    },
+    {
+      "epoch": 0.20122347727655884,
+      "grad_norm": 1.5261359214782715,
+      "learning_rate": 8.408629956393875e-07,
+      "loss": 0.9035,
+      "step": 8355
+    },
+    {
+      "epoch": 0.20134389826834614,
+      "grad_norm": 1.4733154773712158,
+      "learning_rate": 8.407362336477031e-07,
+      "loss": 0.9255,
+      "step": 8360
+    },
+    {
+      "epoch": 0.20146431926013342,
+      "grad_norm": 1.9009312391281128,
+      "learning_rate": 8.406094716560185e-07,
+      "loss": 0.979,
+      "step": 8365
+    },
+    {
+      "epoch": 0.20158474025192072,
+      "grad_norm": 1.560454249382019,
+      "learning_rate": 8.404827096643342e-07,
+      "loss": 0.9264,
+      "step": 8370
+    },
+    {
+      "epoch": 0.201705161243708,
+      "grad_norm": 1.6878447532653809,
+      "learning_rate": 8.403559476726498e-07,
+      "loss": 0.9107,
+      "step": 8375
+    },
+    {
+      "epoch": 0.2018255822354953,
+      "grad_norm": 1.3901766538619995,
+      "learning_rate": 8.402291856809654e-07,
+      "loss": 0.9724,
+      "step": 8380
+    },
+    {
+      "epoch": 0.20194600322728257,
+      "grad_norm": 1.402201771736145,
+      "learning_rate": 8.40102423689281e-07,
+      "loss": 0.9326,
+      "step": 8385
+    },
+    {
+      "epoch": 0.20206642421906987,
+      "grad_norm": 1.4999525547027588,
+      "learning_rate": 8.399756616975966e-07,
+      "loss": 0.983,
+      "step": 8390
+    },
+    {
+      "epoch": 0.20218684521085717,
+      "grad_norm": 1.554654598236084,
+      "learning_rate": 8.398488997059121e-07,
+      "loss": 0.9594,
+      "step": 8395
+    },
+    {
+      "epoch": 0.20230726620264444,
+      "grad_norm": 1.4805643558502197,
+      "learning_rate": 8.397221377142277e-07,
+      "loss": 0.9716,
+      "step": 8400
+    },
+    {
+      "epoch": 0.20242768719443174,
+      "grad_norm": 1.4133025407791138,
+      "learning_rate": 8.395953757225434e-07,
+      "loss": 0.9541,
+      "step": 8405
+    },
+    {
+      "epoch": 0.20254810818621902,
+      "grad_norm": 1.8007771968841553,
+      "learning_rate": 8.394686137308588e-07,
+      "loss": 0.9827,
+      "step": 8410
+    },
+    {
+      "epoch": 0.20266852917800632,
+      "grad_norm": 1.5795798301696777,
+      "learning_rate": 8.393418517391745e-07,
+      "loss": 0.9792,
+      "step": 8415
+    },
+    {
+      "epoch": 0.2027889501697936,
+      "grad_norm": 1.633016586303711,
+      "learning_rate": 8.392150897474901e-07,
+      "loss": 0.9379,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2029093711615809,
+      "grad_norm": 1.6277592182159424,
+      "learning_rate": 8.390883277558056e-07,
+      "loss": 0.9478,
+      "step": 8425
+    },
+    {
+      "epoch": 0.20302979215336817,
+      "grad_norm": 1.718751311302185,
+      "learning_rate": 8.389615657641213e-07,
+      "loss": 0.944,
+      "step": 8430
+    },
+    {
+      "epoch": 0.20315021314515547,
+      "grad_norm": 1.5914260149002075,
+      "learning_rate": 8.388348037724368e-07,
+      "loss": 0.9741,
+      "step": 8435
+    },
+    {
+      "epoch": 0.20327063413694274,
+      "grad_norm": 1.4470100402832031,
+      "learning_rate": 8.387080417807524e-07,
+      "loss": 0.9075,
+      "step": 8440
+    },
+    {
+      "epoch": 0.20339105512873004,
+      "grad_norm": 1.3547005653381348,
+      "learning_rate": 8.38581279789068e-07,
+      "loss": 0.9711,
+      "step": 8445
+    },
+    {
+      "epoch": 0.20351147612051732,
+      "grad_norm": 1.6130958795547485,
+      "learning_rate": 8.384545177973837e-07,
+      "loss": 0.8619,
+      "step": 8450
+    },
+    {
+      "epoch": 0.20363189711230462,
+      "grad_norm": 1.448756456375122,
+      "learning_rate": 8.383277558056992e-07,
+      "loss": 0.9311,
+      "step": 8455
+    },
+    {
+      "epoch": 0.20375231810409192,
+      "grad_norm": 1.7305628061294556,
+      "learning_rate": 8.382009938140147e-07,
+      "loss": 0.9222,
+      "step": 8460
+    },
+    {
+      "epoch": 0.2038727390958792,
+      "grad_norm": 1.4965406656265259,
+      "learning_rate": 8.380742318223304e-07,
+      "loss": 0.956,
+      "step": 8465
+    },
+    {
+      "epoch": 0.2039931600876665,
+      "grad_norm": 1.5525379180908203,
+      "learning_rate": 8.379474698306459e-07,
+      "loss": 0.9343,
+      "step": 8470
+    },
+    {
+      "epoch": 0.20411358107945377,
+      "grad_norm": 1.8234738111495972,
+      "learning_rate": 8.378207078389616e-07,
+      "loss": 0.9319,
+      "step": 8475
+    },
+    {
+      "epoch": 0.20423400207124107,
+      "grad_norm": 1.8297779560089111,
+      "learning_rate": 8.376939458472771e-07,
+      "loss": 0.9049,
+      "step": 8480
+    },
+    {
+      "epoch": 0.20435442306302834,
+      "grad_norm": 1.5790259838104248,
+      "learning_rate": 8.375671838555926e-07,
+      "loss": 0.9588,
+      "step": 8485
+    },
+    {
+      "epoch": 0.20447484405481564,
+      "grad_norm": 1.7919198274612427,
+      "learning_rate": 8.374404218639083e-07,
+      "loss": 0.9558,
+      "step": 8490
+    },
+    {
+      "epoch": 0.20459526504660291,
+      "grad_norm": 1.4658918380737305,
+      "learning_rate": 8.373136598722239e-07,
+      "loss": 0.9135,
+      "step": 8495
+    },
+    {
+      "epoch": 0.20471568603839022,
+      "grad_norm": 1.581298828125,
+      "learning_rate": 8.371868978805395e-07,
+      "loss": 0.931,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2048361070301775,
+      "grad_norm": 1.5665123462677002,
+      "learning_rate": 8.37060135888855e-07,
+      "loss": 0.9686,
+      "step": 8505
+    },
+    {
+      "epoch": 0.2049565280219648,
+      "grad_norm": 1.6349343061447144,
+      "learning_rate": 8.369333738971707e-07,
+      "loss": 0.9041,
+      "step": 8510
+    },
+    {
+      "epoch": 0.20507694901375206,
+      "grad_norm": 1.7548354864120483,
+      "learning_rate": 8.368066119054862e-07,
+      "loss": 0.9142,
+      "step": 8515
+    },
+    {
+      "epoch": 0.20519737000553936,
+      "grad_norm": 1.7395234107971191,
+      "learning_rate": 8.366798499138018e-07,
+      "loss": 0.8966,
+      "step": 8520
+    },
+    {
+      "epoch": 0.20531779099732667,
+      "grad_norm": 1.6489882469177246,
+      "learning_rate": 8.365530879221175e-07,
+      "loss": 0.9579,
+      "step": 8525
+    },
+    {
+      "epoch": 0.20543821198911394,
+      "grad_norm": 1.679075837135315,
+      "learning_rate": 8.364263259304329e-07,
+      "loss": 0.9257,
+      "step": 8530
+    },
+    {
+      "epoch": 0.20555863298090124,
+      "grad_norm": 1.3792853355407715,
+      "learning_rate": 8.362995639387486e-07,
+      "loss": 0.9361,
+      "step": 8535
+    },
+    {
+      "epoch": 0.2056790539726885,
+      "grad_norm": 1.6504968404769897,
+      "learning_rate": 8.361728019470642e-07,
+      "loss": 0.9498,
+      "step": 8540
+    },
+    {
+      "epoch": 0.20579947496447581,
+      "grad_norm": 1.645012378692627,
+      "learning_rate": 8.360460399553797e-07,
+      "loss": 0.9963,
+      "step": 8545
+    },
+    {
+      "epoch": 0.2059198959562631,
+      "grad_norm": 1.3971391916275024,
+      "learning_rate": 8.359192779636953e-07,
+      "loss": 0.9941,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2060403169480504,
+      "grad_norm": 1.675577998161316,
+      "learning_rate": 8.357925159720109e-07,
+      "loss": 0.9227,
+      "step": 8555
+    },
+    {
+      "epoch": 0.20616073793983766,
+      "grad_norm": 1.5664169788360596,
+      "learning_rate": 8.356657539803265e-07,
+      "loss": 0.9335,
+      "step": 8560
+    },
+    {
+      "epoch": 0.20628115893162496,
+      "grad_norm": 1.7027219533920288,
+      "learning_rate": 8.355389919886421e-07,
+      "loss": 0.961,
+      "step": 8565
+    },
+    {
+      "epoch": 0.20640157992341224,
+      "grad_norm": 1.5615417957305908,
+      "learning_rate": 8.354122299969578e-07,
+      "loss": 0.9518,
+      "step": 8570
+    },
+    {
+      "epoch": 0.20652200091519954,
+      "grad_norm": 1.8420103788375854,
+      "learning_rate": 8.352854680052732e-07,
+      "loss": 0.9987,
+      "step": 8575
+    },
+    {
+      "epoch": 0.20664242190698684,
+      "grad_norm": 1.625141978263855,
+      "learning_rate": 8.351587060135888e-07,
+      "loss": 0.8968,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2067628428987741,
+      "grad_norm": 1.6060346364974976,
+      "learning_rate": 8.350319440219045e-07,
+      "loss": 0.9154,
+      "step": 8585
+    },
+    {
+      "epoch": 0.20688326389056141,
+      "grad_norm": 1.5125768184661865,
+      "learning_rate": 8.3490518203022e-07,
+      "loss": 0.9727,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2070036848823487,
+      "grad_norm": 1.5166090726852417,
+      "learning_rate": 8.347784200385356e-07,
+      "loss": 0.9022,
+      "step": 8595
+    },
+    {
+      "epoch": 0.207124105874136,
+      "grad_norm": 1.4007935523986816,
+      "learning_rate": 8.346516580468512e-07,
+      "loss": 0.948,
+      "step": 8600
+    },
+    {
+      "epoch": 0.20724452686592326,
+      "grad_norm": 1.7391657829284668,
+      "learning_rate": 8.345248960551667e-07,
+      "loss": 0.8898,
+      "step": 8605
+    },
+    {
+      "epoch": 0.20736494785771056,
+      "grad_norm": 1.3972952365875244,
+      "learning_rate": 8.343981340634824e-07,
+      "loss": 0.8994,
+      "step": 8610
+    },
+    {
+      "epoch": 0.20748536884949784,
+      "grad_norm": 1.6144566535949707,
+      "learning_rate": 8.34271372071798e-07,
+      "loss": 0.9528,
+      "step": 8615
+    },
+    {
+      "epoch": 0.20760578984128514,
+      "grad_norm": 1.5001790523529053,
+      "learning_rate": 8.341446100801135e-07,
+      "loss": 0.9638,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2077262108330724,
+      "grad_norm": 1.884637713432312,
+      "learning_rate": 8.340178480884291e-07,
+      "loss": 0.9093,
+      "step": 8625
+    },
+    {
+      "epoch": 0.2078466318248597,
+      "grad_norm": 1.5699843168258667,
+      "learning_rate": 8.338910860967447e-07,
+      "loss": 0.973,
+      "step": 8630
+    },
+    {
+      "epoch": 0.20796705281664699,
+      "grad_norm": 1.5710140466690063,
+      "learning_rate": 8.337643241050603e-07,
+      "loss": 0.9331,
+      "step": 8635
+    },
+    {
+      "epoch": 0.2080874738084343,
+      "grad_norm": 1.4831452369689941,
+      "learning_rate": 8.336375621133759e-07,
+      "loss": 0.9384,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2082078948002216,
+      "grad_norm": 1.5688238143920898,
+      "learning_rate": 8.335108001216915e-07,
+      "loss": 0.9716,
+      "step": 8645
+    },
+    {
+      "epoch": 0.20832831579200886,
+      "grad_norm": 1.4611543416976929,
+      "learning_rate": 8.33384038130007e-07,
+      "loss": 0.8993,
+      "step": 8650
+    },
+    {
+      "epoch": 0.20844873678379616,
+      "grad_norm": 1.4965218305587769,
+      "learning_rate": 8.332572761383227e-07,
+      "loss": 0.9188,
+      "step": 8655
+    },
+    {
+      "epoch": 0.20856915777558344,
+      "grad_norm": 1.3796385526657104,
+      "learning_rate": 8.331305141466383e-07,
+      "loss": 0.8892,
+      "step": 8660
+    },
+    {
+      "epoch": 0.20868957876737074,
+      "grad_norm": 1.4881573915481567,
+      "learning_rate": 8.330037521549537e-07,
+      "loss": 0.8978,
+      "step": 8665
+    },
+    {
+      "epoch": 0.208809999759158,
+      "grad_norm": 1.588682770729065,
+      "learning_rate": 8.328769901632694e-07,
+      "loss": 0.9164,
+      "step": 8670
+    },
+    {
+      "epoch": 0.2089304207509453,
+      "grad_norm": 1.5392111539840698,
+      "learning_rate": 8.32750228171585e-07,
+      "loss": 0.9245,
+      "step": 8675
+    },
+    {
+      "epoch": 0.20905084174273258,
+      "grad_norm": 2.1696298122406006,
+      "learning_rate": 8.326234661799006e-07,
+      "loss": 0.9343,
+      "step": 8680
+    },
+    {
+      "epoch": 0.20917126273451989,
+      "grad_norm": 1.838498830795288,
+      "learning_rate": 8.324967041882162e-07,
+      "loss": 0.9351,
+      "step": 8685
+    },
+    {
+      "epoch": 0.20929168372630716,
+      "grad_norm": 1.9327776432037354,
+      "learning_rate": 8.323699421965317e-07,
+      "loss": 0.9452,
+      "step": 8690
+    },
+    {
+      "epoch": 0.20941210471809446,
+      "grad_norm": 1.5329734086990356,
+      "learning_rate": 8.322431802048473e-07,
+      "loss": 0.9778,
+      "step": 8695
+    },
+    {
+      "epoch": 0.20953252570988173,
+      "grad_norm": 1.5288251638412476,
+      "learning_rate": 8.321164182131629e-07,
+      "loss": 0.9253,
+      "step": 8700
+    },
+    {
+      "epoch": 0.20965294670166904,
+      "grad_norm": 1.4873708486557007,
+      "learning_rate": 8.319896562214786e-07,
+      "loss": 0.9109,
+      "step": 8705
+    },
+    {
+      "epoch": 0.20977336769345634,
+      "grad_norm": 1.5548856258392334,
+      "learning_rate": 8.31862894229794e-07,
+      "loss": 0.9195,
+      "step": 8710
+    },
+    {
+      "epoch": 0.2098937886852436,
+      "grad_norm": 1.4828455448150635,
+      "learning_rate": 8.317361322381096e-07,
+      "loss": 0.881,
+      "step": 8715
+    },
+    {
+      "epoch": 0.2100142096770309,
+      "grad_norm": 1.46074378490448,
+      "learning_rate": 8.316093702464253e-07,
+      "loss": 0.9704,
+      "step": 8720
+    },
+    {
+      "epoch": 0.21013463066881818,
+      "grad_norm": 1.7580933570861816,
+      "learning_rate": 8.314826082547408e-07,
+      "loss": 0.9254,
+      "step": 8725
+    },
+    {
+      "epoch": 0.21025505166060549,
+      "grad_norm": 1.614861011505127,
+      "learning_rate": 8.313558462630565e-07,
+      "loss": 0.9465,
+      "step": 8730
+    },
+    {
+      "epoch": 0.21037547265239276,
+      "grad_norm": 1.5930465459823608,
+      "learning_rate": 8.31229084271372e-07,
+      "loss": 0.9391,
+      "step": 8735
+    },
+    {
+      "epoch": 0.21049589364418006,
+      "grad_norm": 1.4870792627334595,
+      "learning_rate": 8.311023222796876e-07,
+      "loss": 0.8878,
+      "step": 8740
+    },
+    {
+      "epoch": 0.21061631463596733,
+      "grad_norm": 1.4152858257293701,
+      "learning_rate": 8.309755602880032e-07,
+      "loss": 0.9364,
+      "step": 8745
+    },
+    {
+      "epoch": 0.21073673562775463,
+      "grad_norm": 1.6814806461334229,
+      "learning_rate": 8.308487982963188e-07,
+      "loss": 0.92,
+      "step": 8750
+    },
+    {
+      "epoch": 0.2108571566195419,
+      "grad_norm": 1.6177480220794678,
+      "learning_rate": 8.307220363046344e-07,
+      "loss": 0.9439,
+      "step": 8755
+    },
+    {
+      "epoch": 0.2109775776113292,
+      "grad_norm": 1.4960063695907593,
+      "learning_rate": 8.305952743129499e-07,
+      "loss": 0.8827,
+      "step": 8760
+    },
+    {
+      "epoch": 0.21109799860311648,
+      "grad_norm": 1.6099573373794556,
+      "learning_rate": 8.304685123212656e-07,
+      "loss": 0.961,
+      "step": 8765
+    },
+    {
+      "epoch": 0.21121841959490378,
+      "grad_norm": 1.7052769660949707,
+      "learning_rate": 8.303417503295811e-07,
+      "loss": 0.9841,
+      "step": 8770
+    },
+    {
+      "epoch": 0.21133884058669108,
+      "grad_norm": 1.85787034034729,
+      "learning_rate": 8.302149883378968e-07,
+      "loss": 0.93,
+      "step": 8775
+    },
+    {
+      "epoch": 0.21145926157847836,
+      "grad_norm": 1.7271051406860352,
+      "learning_rate": 8.300882263462123e-07,
+      "loss": 0.9026,
+      "step": 8780
+    },
+    {
+      "epoch": 0.21157968257026566,
+      "grad_norm": 1.6201260089874268,
+      "learning_rate": 8.299614643545278e-07,
+      "loss": 0.9182,
+      "step": 8785
+    },
+    {
+      "epoch": 0.21170010356205293,
+      "grad_norm": 2.0843045711517334,
+      "learning_rate": 8.298347023628435e-07,
+      "loss": 1.0306,
+      "step": 8790
+    },
+    {
+      "epoch": 0.21182052455384023,
+      "grad_norm": 1.541614055633545,
+      "learning_rate": 8.297079403711591e-07,
+      "loss": 0.8982,
+      "step": 8795
+    },
+    {
+      "epoch": 0.2119409455456275,
+      "grad_norm": 1.7830393314361572,
+      "learning_rate": 8.295811783794747e-07,
+      "loss": 0.9132,
+      "step": 8800
+    },
+    {
+      "epoch": 0.2120613665374148,
+      "grad_norm": 1.6531665325164795,
+      "learning_rate": 8.294544163877902e-07,
+      "loss": 0.9826,
+      "step": 8805
+    },
+    {
+      "epoch": 0.21218178752920208,
+      "grad_norm": 1.539405107498169,
+      "learning_rate": 8.293276543961058e-07,
+      "loss": 0.9498,
+      "step": 8810
+    },
+    {
+      "epoch": 0.21230220852098938,
+      "grad_norm": 1.5371087789535522,
+      "learning_rate": 8.292008924044214e-07,
+      "loss": 0.8884,
+      "step": 8815
+    },
+    {
+      "epoch": 0.21242262951277666,
+      "grad_norm": 1.4828464984893799,
+      "learning_rate": 8.29074130412737e-07,
+      "loss": 0.9002,
+      "step": 8820
+    },
+    {
+      "epoch": 0.21254305050456396,
+      "grad_norm": 1.4968286752700806,
+      "learning_rate": 8.289473684210527e-07,
+      "loss": 0.9189,
+      "step": 8825
+    },
+    {
+      "epoch": 0.21266347149635123,
+      "grad_norm": 1.496974229812622,
+      "learning_rate": 8.288206064293681e-07,
+      "loss": 0.951,
+      "step": 8830
+    },
+    {
+      "epoch": 0.21278389248813853,
+      "grad_norm": 1.4990928173065186,
+      "learning_rate": 8.286938444376837e-07,
+      "loss": 0.977,
+      "step": 8835
+    },
+    {
+      "epoch": 0.21290431347992583,
+      "grad_norm": 1.6481170654296875,
+      "learning_rate": 8.285670824459994e-07,
+      "loss": 0.9582,
+      "step": 8840
+    },
+    {
+      "epoch": 0.2130247344717131,
+      "grad_norm": 1.5998990535736084,
+      "learning_rate": 8.284403204543149e-07,
+      "loss": 0.8715,
+      "step": 8845
+    },
+    {
+      "epoch": 0.2131451554635004,
+      "grad_norm": 1.5779908895492554,
+      "learning_rate": 8.283135584626305e-07,
+      "loss": 0.9167,
+      "step": 8850
+    },
+    {
+      "epoch": 0.21326557645528768,
+      "grad_norm": 1.4358510971069336,
+      "learning_rate": 8.281867964709461e-07,
+      "loss": 0.9238,
+      "step": 8855
+    },
+    {
+      "epoch": 0.21338599744707498,
+      "grad_norm": 1.5531116724014282,
+      "learning_rate": 8.280600344792617e-07,
+      "loss": 0.9415,
+      "step": 8860
+    },
+    {
+      "epoch": 0.21350641843886226,
+      "grad_norm": 1.5850262641906738,
+      "learning_rate": 8.279332724875773e-07,
+      "loss": 0.9163,
+      "step": 8865
+    },
+    {
+      "epoch": 0.21362683943064956,
+      "grad_norm": 1.3992431163787842,
+      "learning_rate": 8.278065104958929e-07,
+      "loss": 0.9199,
+      "step": 8870
+    },
+    {
+      "epoch": 0.21374726042243683,
+      "grad_norm": 1.2640268802642822,
+      "learning_rate": 8.276797485042084e-07,
+      "loss": 0.874,
+      "step": 8875
+    },
+    {
+      "epoch": 0.21386768141422413,
+      "grad_norm": 1.5796748399734497,
+      "learning_rate": 8.27552986512524e-07,
+      "loss": 0.8909,
+      "step": 8880
+    },
+    {
+      "epoch": 0.2139881024060114,
+      "grad_norm": 1.6204979419708252,
+      "learning_rate": 8.274262245208397e-07,
+      "loss": 0.8746,
+      "step": 8885
+    },
+    {
+      "epoch": 0.2141085233977987,
+      "grad_norm": 1.566311001777649,
+      "learning_rate": 8.272994625291552e-07,
+      "loss": 0.9035,
+      "step": 8890
+    },
+    {
+      "epoch": 0.21422894438958598,
+      "grad_norm": 1.482667326927185,
+      "learning_rate": 8.271727005374707e-07,
+      "loss": 0.9214,
+      "step": 8895
+    },
+    {
+      "epoch": 0.21434936538137328,
+      "grad_norm": 1.520044207572937,
+      "learning_rate": 8.270459385457864e-07,
+      "loss": 0.9096,
+      "step": 8900
+    },
+    {
+      "epoch": 0.21446978637316058,
+      "grad_norm": 1.5712969303131104,
+      "learning_rate": 8.269191765541019e-07,
+      "loss": 0.9333,
+      "step": 8905
+    },
+    {
+      "epoch": 0.21459020736494785,
+      "grad_norm": 1.6662484407424927,
+      "learning_rate": 8.267924145624176e-07,
+      "loss": 0.9577,
+      "step": 8910
+    },
+    {
+      "epoch": 0.21471062835673516,
+      "grad_norm": 1.487890362739563,
+      "learning_rate": 8.266656525707332e-07,
+      "loss": 0.9055,
+      "step": 8915
+    },
+    {
+      "epoch": 0.21483104934852243,
+      "grad_norm": 1.7255772352218628,
+      "learning_rate": 8.265388905790486e-07,
+      "loss": 0.9385,
+      "step": 8920
+    },
+    {
+      "epoch": 0.21495147034030973,
+      "grad_norm": 1.6064256429672241,
+      "learning_rate": 8.264121285873643e-07,
+      "loss": 0.9491,
+      "step": 8925
+    },
+    {
+      "epoch": 0.215071891332097,
+      "grad_norm": 1.5699553489685059,
+      "learning_rate": 8.262853665956799e-07,
+      "loss": 0.9202,
+      "step": 8930
+    },
+    {
+      "epoch": 0.2151923123238843,
+      "grad_norm": 1.543715238571167,
+      "learning_rate": 8.261586046039955e-07,
+      "loss": 0.9249,
+      "step": 8935
+    },
+    {
+      "epoch": 0.21531273331567158,
+      "grad_norm": 1.4051569700241089,
+      "learning_rate": 8.260318426123111e-07,
+      "loss": 0.9502,
+      "step": 8940
+    },
+    {
+      "epoch": 0.21543315430745888,
+      "grad_norm": 1.690550446510315,
+      "learning_rate": 8.259050806206267e-07,
+      "loss": 0.8807,
+      "step": 8945
+    },
+    {
+      "epoch": 0.21555357529924615,
+      "grad_norm": 1.3593227863311768,
+      "learning_rate": 8.257783186289422e-07,
+      "loss": 0.9861,
+      "step": 8950
+    },
+    {
+      "epoch": 0.21567399629103345,
+      "grad_norm": 1.652435064315796,
+      "learning_rate": 8.256515566372578e-07,
+      "loss": 0.9384,
+      "step": 8955
+    },
+    {
+      "epoch": 0.21579441728282076,
+      "grad_norm": 1.7051680088043213,
+      "learning_rate": 8.255247946455735e-07,
+      "loss": 0.9455,
+      "step": 8960
+    },
+    {
+      "epoch": 0.21591483827460803,
+      "grad_norm": 1.766378402709961,
+      "learning_rate": 8.25398032653889e-07,
+      "loss": 0.9533,
+      "step": 8965
+    },
+    {
+      "epoch": 0.21603525926639533,
+      "grad_norm": 1.6049004793167114,
+      "learning_rate": 8.252712706622046e-07,
+      "loss": 0.9435,
+      "step": 8970
+    },
+    {
+      "epoch": 0.2161556802581826,
+      "grad_norm": 1.7170352935791016,
+      "learning_rate": 8.251445086705202e-07,
+      "loss": 0.9371,
+      "step": 8975
+    },
+    {
+      "epoch": 0.2162761012499699,
+      "grad_norm": 1.506013035774231,
+      "learning_rate": 8.250177466788359e-07,
+      "loss": 0.8851,
+      "step": 8980
+    },
+    {
+      "epoch": 0.21639652224175718,
+      "grad_norm": 1.8347394466400146,
+      "learning_rate": 8.248909846871514e-07,
+      "loss": 0.9822,
+      "step": 8985
+    },
+    {
+      "epoch": 0.21651694323354448,
+      "grad_norm": 1.7963132858276367,
+      "learning_rate": 8.247642226954669e-07,
+      "loss": 0.9288,
+      "step": 8990
+    },
+    {
+      "epoch": 0.21663736422533175,
+      "grad_norm": 1.444070816040039,
+      "learning_rate": 8.246374607037826e-07,
+      "loss": 0.9085,
+      "step": 8995
+    },
+    {
+      "epoch": 0.21675778521711905,
+      "grad_norm": 1.5620007514953613,
+      "learning_rate": 8.245106987120981e-07,
+      "loss": 0.9667,
+      "step": 9000
+    },
+    {
+      "epoch": 0.21687820620890633,
+      "grad_norm": 1.4637254476547241,
+      "learning_rate": 8.243839367204138e-07,
+      "loss": 0.9519,
+      "step": 9005
+    },
+    {
+      "epoch": 0.21699862720069363,
+      "grad_norm": 1.7866486310958862,
+      "learning_rate": 8.242571747287294e-07,
+      "loss": 0.9099,
+      "step": 9010
+    },
+    {
+      "epoch": 0.2171190481924809,
+      "grad_norm": 1.5120880603790283,
+      "learning_rate": 8.241304127370448e-07,
+      "loss": 0.889,
+      "step": 9015
+    },
+    {
+      "epoch": 0.2172394691842682,
+      "grad_norm": 1.47304105758667,
+      "learning_rate": 8.240036507453605e-07,
+      "loss": 0.9659,
+      "step": 9020
+    },
+    {
+      "epoch": 0.2173598901760555,
+      "grad_norm": 1.4226914644241333,
+      "learning_rate": 8.238768887536761e-07,
+      "loss": 0.9371,
+      "step": 9025
+    },
+    {
+      "epoch": 0.21748031116784278,
+      "grad_norm": 1.4126009941101074,
+      "learning_rate": 8.237501267619917e-07,
+      "loss": 0.9464,
+      "step": 9030
+    },
+    {
+      "epoch": 0.21760073215963008,
+      "grad_norm": 1.5451222658157349,
+      "learning_rate": 8.236233647703072e-07,
+      "loss": 0.8829,
+      "step": 9035
+    },
+    {
+      "epoch": 0.21772115315141735,
+      "grad_norm": 1.4735143184661865,
+      "learning_rate": 8.234966027786229e-07,
+      "loss": 0.9358,
+      "step": 9040
+    },
+    {
+      "epoch": 0.21784157414320465,
+      "grad_norm": 1.7014250755310059,
+      "learning_rate": 8.233698407869384e-07,
+      "loss": 0.9346,
+      "step": 9045
+    },
+    {
+      "epoch": 0.21796199513499193,
+      "grad_norm": 1.54008150100708,
+      "learning_rate": 8.23243078795254e-07,
+      "loss": 0.9538,
+      "step": 9050
+    },
+    {
+      "epoch": 0.21808241612677923,
+      "grad_norm": 1.5994731187820435,
+      "learning_rate": 8.231163168035697e-07,
+      "loss": 0.9334,
+      "step": 9055
+    },
+    {
+      "epoch": 0.2182028371185665,
+      "grad_norm": 1.5312918424606323,
+      "learning_rate": 8.229895548118851e-07,
+      "loss": 0.9295,
+      "step": 9060
+    },
+    {
+      "epoch": 0.2183232581103538,
+      "grad_norm": 1.4426707029342651,
+      "learning_rate": 8.228627928202008e-07,
+      "loss": 0.9179,
+      "step": 9065
+    },
+    {
+      "epoch": 0.21844367910214108,
+      "grad_norm": 1.4576219320297241,
+      "learning_rate": 8.227360308285164e-07,
+      "loss": 0.9565,
+      "step": 9070
+    },
+    {
+      "epoch": 0.21856410009392838,
+      "grad_norm": 1.8166043758392334,
+      "learning_rate": 8.226092688368319e-07,
+      "loss": 0.9101,
+      "step": 9075
+    },
+    {
+      "epoch": 0.21868452108571565,
+      "grad_norm": 1.6293615102767944,
+      "learning_rate": 8.224825068451475e-07,
+      "loss": 0.8985,
+      "step": 9080
+    },
+    {
+      "epoch": 0.21880494207750295,
+      "grad_norm": 1.313899278640747,
+      "learning_rate": 8.223557448534631e-07,
+      "loss": 0.9632,
+      "step": 9085
+    },
+    {
+      "epoch": 0.21892536306929025,
+      "grad_norm": 1.5151069164276123,
+      "learning_rate": 8.222289828617787e-07,
+      "loss": 0.9532,
+      "step": 9090
+    },
+    {
+      "epoch": 0.21904578406107753,
+      "grad_norm": 1.794693946838379,
+      "learning_rate": 8.221022208700943e-07,
+      "loss": 0.9792,
+      "step": 9095
+    },
+    {
+      "epoch": 0.21916620505286483,
+      "grad_norm": 1.683070421218872,
+      "learning_rate": 8.2197545887841e-07,
+      "loss": 0.9168,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2192866260446521,
+      "grad_norm": 1.560676097869873,
+      "learning_rate": 8.218486968867254e-07,
+      "loss": 0.9712,
+      "step": 9105
+    },
+    {
+      "epoch": 0.2194070470364394,
+      "grad_norm": 1.6398364305496216,
+      "learning_rate": 8.21721934895041e-07,
+      "loss": 0.9973,
+      "step": 9110
+    },
+    {
+      "epoch": 0.21952746802822667,
+      "grad_norm": 1.6339504718780518,
+      "learning_rate": 8.215951729033567e-07,
+      "loss": 0.876,
+      "step": 9115
+    },
+    {
+      "epoch": 0.21964788902001398,
+      "grad_norm": 2.2558040618896484,
+      "learning_rate": 8.214684109116722e-07,
+      "loss": 0.985,
+      "step": 9120
+    },
+    {
+      "epoch": 0.21976831001180125,
+      "grad_norm": 1.5891695022583008,
+      "learning_rate": 8.213416489199878e-07,
+      "loss": 0.9204,
+      "step": 9125
+    },
+    {
+      "epoch": 0.21988873100358855,
+      "grad_norm": 1.626607894897461,
+      "learning_rate": 8.212148869283034e-07,
+      "loss": 0.9663,
+      "step": 9130
+    },
+    {
+      "epoch": 0.22000915199537582,
+      "grad_norm": 1.8692668676376343,
+      "learning_rate": 8.210881249366189e-07,
+      "loss": 0.9775,
+      "step": 9135
+    },
+    {
+      "epoch": 0.22012957298716312,
+      "grad_norm": 1.653274416923523,
+      "learning_rate": 8.209613629449346e-07,
+      "loss": 0.9183,
+      "step": 9140
+    },
+    {
+      "epoch": 0.2202499939789504,
+      "grad_norm": 1.647321105003357,
+      "learning_rate": 8.208346009532502e-07,
+      "loss": 0.9363,
+      "step": 9145
+    },
+    {
+      "epoch": 0.2203704149707377,
+      "grad_norm": 1.404857873916626,
+      "learning_rate": 8.207078389615657e-07,
+      "loss": 0.8812,
+      "step": 9150
+    },
+    {
+      "epoch": 0.220490835962525,
+      "grad_norm": 1.4598122835159302,
+      "learning_rate": 8.205810769698813e-07,
+      "loss": 0.9517,
+      "step": 9155
+    },
+    {
+      "epoch": 0.22061125695431227,
+      "grad_norm": 1.5059850215911865,
+      "learning_rate": 8.20454314978197e-07,
+      "loss": 0.886,
+      "step": 9160
+    },
+    {
+      "epoch": 0.22073167794609957,
+      "grad_norm": 1.731574535369873,
+      "learning_rate": 8.203275529865125e-07,
+      "loss": 0.9316,
+      "step": 9165
+    },
+    {
+      "epoch": 0.22085209893788685,
+      "grad_norm": 1.4054409265518188,
+      "learning_rate": 8.202007909948281e-07,
+      "loss": 0.9281,
+      "step": 9170
+    },
+    {
+      "epoch": 0.22097251992967415,
+      "grad_norm": 1.78800630569458,
+      "learning_rate": 8.200740290031437e-07,
+      "loss": 0.954,
+      "step": 9175
+    },
+    {
+      "epoch": 0.22109294092146142,
+      "grad_norm": 1.5038881301879883,
+      "learning_rate": 8.199472670114592e-07,
+      "loss": 0.9794,
+      "step": 9180
+    },
+    {
+      "epoch": 0.22121336191324872,
+      "grad_norm": 1.5534361600875854,
+      "learning_rate": 8.198205050197749e-07,
+      "loss": 0.9235,
+      "step": 9185
+    },
+    {
+      "epoch": 0.221333782905036,
+      "grad_norm": 1.6785494089126587,
+      "learning_rate": 8.196937430280905e-07,
+      "loss": 0.9028,
+      "step": 9190
+    },
+    {
+      "epoch": 0.2214542038968233,
+      "grad_norm": 1.5974268913269043,
+      "learning_rate": 8.195669810364059e-07,
+      "loss": 0.9303,
+      "step": 9195
+    },
+    {
+      "epoch": 0.22157462488861057,
+      "grad_norm": 1.4408422708511353,
+      "learning_rate": 8.194402190447216e-07,
+      "loss": 0.8687,
+      "step": 9200
+    },
+    {
+      "epoch": 0.22169504588039787,
+      "grad_norm": 1.5255826711654663,
+      "learning_rate": 8.193134570530372e-07,
+      "loss": 0.9982,
+      "step": 9205
+    },
+    {
+      "epoch": 0.22181546687218515,
+      "grad_norm": 1.4327969551086426,
+      "learning_rate": 8.191866950613528e-07,
+      "loss": 0.9542,
+      "step": 9210
+    },
+    {
+      "epoch": 0.22193588786397245,
+      "grad_norm": 1.6425248384475708,
+      "learning_rate": 8.190599330696684e-07,
+      "loss": 0.8825,
+      "step": 9215
+    },
+    {
+      "epoch": 0.22205630885575975,
+      "grad_norm": 1.54902184009552,
+      "learning_rate": 8.189331710779839e-07,
+      "loss": 0.9572,
+      "step": 9220
+    },
+    {
+      "epoch": 0.22217672984754702,
+      "grad_norm": 1.4737131595611572,
+      "learning_rate": 8.188064090862995e-07,
+      "loss": 0.9314,
+      "step": 9225
+    },
+    {
+      "epoch": 0.22229715083933432,
+      "grad_norm": 1.696393609046936,
+      "learning_rate": 8.186796470946151e-07,
+      "loss": 0.9247,
+      "step": 9230
+    },
+    {
+      "epoch": 0.2224175718311216,
+      "grad_norm": 1.816698431968689,
+      "learning_rate": 8.185528851029308e-07,
+      "loss": 0.9583,
+      "step": 9235
+    },
+    {
+      "epoch": 0.2225379928229089,
+      "grad_norm": 1.5856852531433105,
+      "learning_rate": 8.184261231112462e-07,
+      "loss": 0.9045,
+      "step": 9240
+    },
+    {
+      "epoch": 0.22265841381469617,
+      "grad_norm": 1.691108226776123,
+      "learning_rate": 8.182993611195619e-07,
+      "loss": 0.9046,
+      "step": 9245
+    },
+    {
+      "epoch": 0.22277883480648347,
+      "grad_norm": 1.5667667388916016,
+      "learning_rate": 8.181725991278775e-07,
+      "loss": 0.959,
+      "step": 9250
+    },
+    {
+      "epoch": 0.22289925579827075,
+      "grad_norm": 1.6249210834503174,
+      "learning_rate": 8.18045837136193e-07,
+      "loss": 0.982,
+      "step": 9255
+    },
+    {
+      "epoch": 0.22301967679005805,
+      "grad_norm": 1.4573348760604858,
+      "learning_rate": 8.179190751445087e-07,
+      "loss": 0.9418,
+      "step": 9260
+    },
+    {
+      "epoch": 0.22314009778184532,
+      "grad_norm": 1.4732943773269653,
+      "learning_rate": 8.177923131528242e-07,
+      "loss": 0.9852,
+      "step": 9265
+    },
+    {
+      "epoch": 0.22326051877363262,
+      "grad_norm": 1.6084719896316528,
+      "learning_rate": 8.176655511611398e-07,
+      "loss": 0.9526,
+      "step": 9270
+    },
+    {
+      "epoch": 0.2233809397654199,
+      "grad_norm": 1.8238649368286133,
+      "learning_rate": 8.175387891694554e-07,
+      "loss": 0.8735,
+      "step": 9275
+    },
+    {
+      "epoch": 0.2235013607572072,
+      "grad_norm": 1.6604654788970947,
+      "learning_rate": 8.17412027177771e-07,
+      "loss": 0.9414,
+      "step": 9280
+    },
+    {
+      "epoch": 0.2236217817489945,
+      "grad_norm": 1.5664421319961548,
+      "learning_rate": 8.172852651860866e-07,
+      "loss": 0.9376,
+      "step": 9285
+    },
+    {
+      "epoch": 0.22374220274078177,
+      "grad_norm": 1.5238959789276123,
+      "learning_rate": 8.171585031944021e-07,
+      "loss": 0.9551,
+      "step": 9290
+    },
+    {
+      "epoch": 0.22386262373256907,
+      "grad_norm": 1.7010811567306519,
+      "learning_rate": 8.170317412027178e-07,
+      "loss": 0.9277,
+      "step": 9295
+    },
+    {
+      "epoch": 0.22398304472435634,
+      "grad_norm": 1.8692560195922852,
+      "learning_rate": 8.169049792110333e-07,
+      "loss": 0.9838,
+      "step": 9300
+    },
+    {
+      "epoch": 0.22410346571614365,
+      "grad_norm": 1.478436827659607,
+      "learning_rate": 8.16778217219349e-07,
+      "loss": 0.9273,
+      "step": 9305
+    },
+    {
+      "epoch": 0.22422388670793092,
+      "grad_norm": 1.5085464715957642,
+      "learning_rate": 8.166514552276645e-07,
+      "loss": 0.9042,
+      "step": 9310
+    },
+    {
+      "epoch": 0.22434430769971822,
+      "grad_norm": 1.468819260597229,
+      "learning_rate": 8.1652469323598e-07,
+      "loss": 0.9537,
+      "step": 9315
+    },
+    {
+      "epoch": 0.2244647286915055,
+      "grad_norm": 1.4283561706542969,
+      "learning_rate": 8.163979312442957e-07,
+      "loss": 0.8855,
+      "step": 9320
+    },
+    {
+      "epoch": 0.2245851496832928,
+      "grad_norm": 1.5676054954528809,
+      "learning_rate": 8.162711692526113e-07,
+      "loss": 0.9456,
+      "step": 9325
+    },
+    {
+      "epoch": 0.22470557067508007,
+      "grad_norm": 1.615062952041626,
+      "learning_rate": 8.161444072609269e-07,
+      "loss": 0.9407,
+      "step": 9330
+    },
+    {
+      "epoch": 0.22482599166686737,
+      "grad_norm": 1.4833568334579468,
+      "learning_rate": 8.160176452692424e-07,
+      "loss": 0.9134,
+      "step": 9335
+    },
+    {
+      "epoch": 0.22494641265865464,
+      "grad_norm": 1.451080083847046,
+      "learning_rate": 8.15890883277558e-07,
+      "loss": 0.887,
+      "step": 9340
+    },
+    {
+      "epoch": 0.22506683365044194,
+      "grad_norm": 1.5666835308074951,
+      "learning_rate": 8.157641212858736e-07,
+      "loss": 0.958,
+      "step": 9345
+    },
+    {
+      "epoch": 0.22518725464222925,
+      "grad_norm": 1.4987585544586182,
+      "learning_rate": 8.156373592941892e-07,
+      "loss": 0.8869,
+      "step": 9350
+    },
+    {
+      "epoch": 0.22530767563401652,
+      "grad_norm": 1.6401864290237427,
+      "learning_rate": 8.155105973025049e-07,
+      "loss": 0.9254,
+      "step": 9355
+    },
+    {
+      "epoch": 0.22542809662580382,
+      "grad_norm": 3.9416544437408447,
+      "learning_rate": 8.153838353108203e-07,
+      "loss": 0.8688,
+      "step": 9360
+    },
+    {
+      "epoch": 0.2255485176175911,
+      "grad_norm": 3.244033098220825,
+      "learning_rate": 8.15257073319136e-07,
+      "loss": 0.8505,
+      "step": 9365
+    },
+    {
+      "epoch": 0.2256689386093784,
+      "grad_norm": 1.533593773841858,
+      "learning_rate": 8.151303113274516e-07,
+      "loss": 0.9816,
+      "step": 9370
+    },
+    {
+      "epoch": 0.22578935960116567,
+      "grad_norm": 1.6429873704910278,
+      "learning_rate": 8.150035493357671e-07,
+      "loss": 0.992,
+      "step": 9375
+    },
+    {
+      "epoch": 0.22590978059295297,
+      "grad_norm": 1.38799250125885,
+      "learning_rate": 8.148767873440827e-07,
+      "loss": 0.9864,
+      "step": 9380
+    },
+    {
+      "epoch": 0.22603020158474024,
+      "grad_norm": 1.5810966491699219,
+      "learning_rate": 8.147500253523983e-07,
+      "loss": 0.9026,
+      "step": 9385
+    },
+    {
+      "epoch": 0.22615062257652754,
+      "grad_norm": 1.738532543182373,
+      "learning_rate": 8.146232633607139e-07,
+      "loss": 0.9626,
+      "step": 9390
+    },
+    {
+      "epoch": 0.22627104356831482,
+      "grad_norm": 1.605966329574585,
+      "learning_rate": 8.144965013690295e-07,
+      "loss": 0.8974,
+      "step": 9395
+    },
+    {
+      "epoch": 0.22639146456010212,
+      "grad_norm": 1.7122306823730469,
+      "learning_rate": 8.143697393773451e-07,
+      "loss": 0.9531,
+      "step": 9400
+    },
+    {
+      "epoch": 0.22651188555188942,
+      "grad_norm": 1.4698325395584106,
+      "learning_rate": 8.142429773856606e-07,
+      "loss": 0.9264,
+      "step": 9405
+    },
+    {
+      "epoch": 0.2266323065436767,
+      "grad_norm": 1.5234861373901367,
+      "learning_rate": 8.141162153939762e-07,
+      "loss": 0.9472,
+      "step": 9410
+    },
+    {
+      "epoch": 0.226752727535464,
+      "grad_norm": 1.6981227397918701,
+      "learning_rate": 8.139894534022919e-07,
+      "loss": 0.9419,
+      "step": 9415
+    },
+    {
+      "epoch": 0.22687314852725127,
+      "grad_norm": 1.6163617372512817,
+      "learning_rate": 8.138626914106074e-07,
+      "loss": 0.9259,
+      "step": 9420
+    },
+    {
+      "epoch": 0.22699356951903857,
+      "grad_norm": 1.4136073589324951,
+      "learning_rate": 8.137359294189229e-07,
+      "loss": 0.9281,
+      "step": 9425
+    },
+    {
+      "epoch": 0.22711399051082584,
+      "grad_norm": 1.4144577980041504,
+      "learning_rate": 8.136091674272386e-07,
+      "loss": 0.9359,
+      "step": 9430
+    },
+    {
+      "epoch": 0.22723441150261314,
+      "grad_norm": 1.7528362274169922,
+      "learning_rate": 8.134824054355541e-07,
+      "loss": 0.9475,
+      "step": 9435
+    },
+    {
+      "epoch": 0.22735483249440042,
+      "grad_norm": 1.4299169778823853,
+      "learning_rate": 8.133556434438698e-07,
+      "loss": 0.9244,
+      "step": 9440
+    },
+    {
+      "epoch": 0.22747525348618772,
+      "grad_norm": 1.508216381072998,
+      "learning_rate": 8.132288814521854e-07,
+      "loss": 0.9444,
+      "step": 9445
+    },
+    {
+      "epoch": 0.227595674477975,
+      "grad_norm": 1.595263957977295,
+      "learning_rate": 8.131021194605008e-07,
+      "loss": 0.9829,
+      "step": 9450
+    },
+    {
+      "epoch": 0.2277160954697623,
+      "grad_norm": 1.7147514820098877,
+      "learning_rate": 8.129753574688165e-07,
+      "loss": 0.9648,
+      "step": 9455
+    },
+    {
+      "epoch": 0.22783651646154957,
+      "grad_norm": 1.4773589372634888,
+      "learning_rate": 8.128485954771321e-07,
+      "loss": 0.9089,
+      "step": 9460
+    },
+    {
+      "epoch": 0.22795693745333687,
+      "grad_norm": 1.7763646841049194,
+      "learning_rate": 8.127218334854477e-07,
+      "loss": 0.8933,
+      "step": 9465
+    },
+    {
+      "epoch": 0.22807735844512417,
+      "grad_norm": 1.7241098880767822,
+      "learning_rate": 8.125950714937633e-07,
+      "loss": 0.9405,
+      "step": 9470
+    },
+    {
+      "epoch": 0.22819777943691144,
+      "grad_norm": 1.6956807374954224,
+      "learning_rate": 8.124683095020789e-07,
+      "loss": 0.9229,
+      "step": 9475
+    },
+    {
+      "epoch": 0.22831820042869874,
+      "grad_norm": 1.601848840713501,
+      "learning_rate": 8.123415475103944e-07,
+      "loss": 0.9073,
+      "step": 9480
+    },
+    {
+      "epoch": 0.22843862142048602,
+      "grad_norm": 1.5310746431350708,
+      "learning_rate": 8.1221478551871e-07,
+      "loss": 0.9413,
+      "step": 9485
+    },
+    {
+      "epoch": 0.22855904241227332,
+      "grad_norm": 1.630176305770874,
+      "learning_rate": 8.120880235270257e-07,
+      "loss": 0.9737,
+      "step": 9490
+    },
+    {
+      "epoch": 0.2286794634040606,
+      "grad_norm": 1.371692180633545,
+      "learning_rate": 8.119612615353411e-07,
+      "loss": 0.9324,
+      "step": 9495
+    },
+    {
+      "epoch": 0.2287998843958479,
+      "grad_norm": 1.6516149044036865,
+      "learning_rate": 8.118344995436568e-07,
+      "loss": 0.9379,
+      "step": 9500
+    },
+    {
+      "epoch": 0.22892030538763516,
+      "grad_norm": 1.6909898519515991,
+      "learning_rate": 8.117077375519724e-07,
+      "loss": 0.9307,
+      "step": 9505
+    },
+    {
+      "epoch": 0.22904072637942247,
+      "grad_norm": 1.8337432146072388,
+      "learning_rate": 8.11580975560288e-07,
+      "loss": 0.9343,
+      "step": 9510
+    },
+    {
+      "epoch": 0.22916114737120974,
+      "grad_norm": 1.6035581827163696,
+      "learning_rate": 8.114542135686036e-07,
+      "loss": 0.9654,
+      "step": 9515
+    },
+    {
+      "epoch": 0.22928156836299704,
+      "grad_norm": 1.5184273719787598,
+      "learning_rate": 8.113274515769191e-07,
+      "loss": 0.8989,
+      "step": 9520
+    },
+    {
+      "epoch": 0.2294019893547843,
+      "grad_norm": 1.56526780128479,
+      "learning_rate": 8.112006895852347e-07,
+      "loss": 0.9954,
+      "step": 9525
+    },
+    {
+      "epoch": 0.22952241034657161,
+      "grad_norm": 1.6989341974258423,
+      "learning_rate": 8.110739275935503e-07,
+      "loss": 0.9337,
+      "step": 9530
+    },
+    {
+      "epoch": 0.22964283133835892,
+      "grad_norm": 1.5808727741241455,
+      "learning_rate": 8.10947165601866e-07,
+      "loss": 0.936,
+      "step": 9535
+    },
+    {
+      "epoch": 0.2297632523301462,
+      "grad_norm": 1.6222801208496094,
+      "learning_rate": 8.108204036101814e-07,
+      "loss": 0.9086,
+      "step": 9540
+    },
+    {
+      "epoch": 0.2298836733219335,
+      "grad_norm": 1.6468894481658936,
+      "learning_rate": 8.10693641618497e-07,
+      "loss": 0.904,
+      "step": 9545
+    },
+    {
+      "epoch": 0.23000409431372076,
+      "grad_norm": 1.5979857444763184,
+      "learning_rate": 8.105668796268127e-07,
+      "loss": 0.902,
+      "step": 9550
+    },
+    {
+      "epoch": 0.23012451530550806,
+      "grad_norm": 1.7124053239822388,
+      "learning_rate": 8.104401176351282e-07,
+      "loss": 0.9355,
+      "step": 9555
+    },
+    {
+      "epoch": 0.23024493629729534,
+      "grad_norm": 1.5121243000030518,
+      "learning_rate": 8.103133556434439e-07,
+      "loss": 0.96,
+      "step": 9560
+    },
+    {
+      "epoch": 0.23036535728908264,
+      "grad_norm": 1.6419447660446167,
+      "learning_rate": 8.101865936517594e-07,
+      "loss": 0.913,
+      "step": 9565
+    },
+    {
+      "epoch": 0.2304857782808699,
+      "grad_norm": 1.5802937746047974,
+      "learning_rate": 8.100598316600749e-07,
+      "loss": 0.9036,
+      "step": 9570
+    },
+    {
+      "epoch": 0.23060619927265721,
+      "grad_norm": 1.6120250225067139,
+      "learning_rate": 8.099330696683906e-07,
+      "loss": 1.0045,
+      "step": 9575
+    },
+    {
+      "epoch": 0.2307266202644445,
+      "grad_norm": 1.6308773756027222,
+      "learning_rate": 8.098063076767062e-07,
+      "loss": 0.9348,
+      "step": 9580
+    },
+    {
+      "epoch": 0.2308470412562318,
+      "grad_norm": 1.5396705865859985,
+      "learning_rate": 8.096795456850218e-07,
+      "loss": 0.903,
+      "step": 9585
+    },
+    {
+      "epoch": 0.23096746224801906,
+      "grad_norm": 1.4627436399459839,
+      "learning_rate": 8.095527836933373e-07,
+      "loss": 0.9531,
+      "step": 9590
+    },
+    {
+      "epoch": 0.23108788323980636,
+      "grad_norm": 1.521868109703064,
+      "learning_rate": 8.09426021701653e-07,
+      "loss": 0.8989,
+      "step": 9595
+    },
+    {
+      "epoch": 0.23120830423159366,
+      "grad_norm": 1.6635419130325317,
+      "learning_rate": 8.092992597099685e-07,
+      "loss": 0.9897,
+      "step": 9600
+    },
+    {
+      "epoch": 0.23132872522338094,
+      "grad_norm": 1.906615972518921,
+      "learning_rate": 8.091724977182841e-07,
+      "loss": 0.9235,
+      "step": 9605
+    },
+    {
+      "epoch": 0.23144914621516824,
+      "grad_norm": 1.4365830421447754,
+      "learning_rate": 8.090457357265997e-07,
+      "loss": 0.9304,
+      "step": 9610
+    },
+    {
+      "epoch": 0.2315695672069555,
+      "grad_norm": 1.482728362083435,
+      "learning_rate": 8.089189737349152e-07,
+      "loss": 0.9914,
+      "step": 9615
+    },
+    {
+      "epoch": 0.2316899881987428,
+      "grad_norm": 1.4797344207763672,
+      "learning_rate": 8.087922117432309e-07,
+      "loss": 0.9271,
+      "step": 9620
+    },
+    {
+      "epoch": 0.2318104091905301,
+      "grad_norm": 1.562041163444519,
+      "learning_rate": 8.086654497515465e-07,
+      "loss": 0.9269,
+      "step": 9625
+    },
+    {
+      "epoch": 0.2319308301823174,
+      "grad_norm": 1.515610694885254,
+      "learning_rate": 8.08538687759862e-07,
+      "loss": 0.9559,
+      "step": 9630
+    },
+    {
+      "epoch": 0.23205125117410466,
+      "grad_norm": 1.6904336214065552,
+      "learning_rate": 8.084119257681776e-07,
+      "loss": 0.9103,
+      "step": 9635
+    },
+    {
+      "epoch": 0.23217167216589196,
+      "grad_norm": 1.5302565097808838,
+      "learning_rate": 8.082851637764932e-07,
+      "loss": 0.9336,
+      "step": 9640
+    },
+    {
+      "epoch": 0.23229209315767924,
+      "grad_norm": 1.7606114149093628,
+      "learning_rate": 8.081584017848088e-07,
+      "loss": 0.8815,
+      "step": 9645
+    },
+    {
+      "epoch": 0.23241251414946654,
+      "grad_norm": 1.400367259979248,
+      "learning_rate": 8.080316397931244e-07,
+      "loss": 0.9513,
+      "step": 9650
+    },
+    {
+      "epoch": 0.2325329351412538,
+      "grad_norm": 1.7276761531829834,
+      "learning_rate": 8.079048778014401e-07,
+      "loss": 0.9295,
+      "step": 9655
+    },
+    {
+      "epoch": 0.2326533561330411,
+      "grad_norm": 1.7191526889801025,
+      "learning_rate": 8.077781158097555e-07,
+      "loss": 0.8973,
+      "step": 9660
+    },
+    {
+      "epoch": 0.2327737771248284,
+      "grad_norm": 1.4676998853683472,
+      "learning_rate": 8.076513538180711e-07,
+      "loss": 0.8989,
+      "step": 9665
+    },
+    {
+      "epoch": 0.23289419811661569,
+      "grad_norm": 1.5539352893829346,
+      "learning_rate": 8.075245918263868e-07,
+      "loss": 0.9475,
+      "step": 9670
+    },
+    {
+      "epoch": 0.233014619108403,
+      "grad_norm": 1.6929291486740112,
+      "learning_rate": 8.073978298347023e-07,
+      "loss": 0.943,
+      "step": 9675
+    },
+    {
+      "epoch": 0.23313504010019026,
+      "grad_norm": 1.5347380638122559,
+      "learning_rate": 8.072710678430179e-07,
+      "loss": 0.9108,
+      "step": 9680
+    },
+    {
+      "epoch": 0.23325546109197756,
+      "grad_norm": 1.5121805667877197,
+      "learning_rate": 8.071443058513335e-07,
+      "loss": 0.9501,
+      "step": 9685
+    },
+    {
+      "epoch": 0.23337588208376484,
+      "grad_norm": 1.4723517894744873,
+      "learning_rate": 8.07017543859649e-07,
+      "loss": 0.9007,
+      "step": 9690
+    },
+    {
+      "epoch": 0.23349630307555214,
+      "grad_norm": 1.4446206092834473,
+      "learning_rate": 8.068907818679647e-07,
+      "loss": 0.95,
+      "step": 9695
+    },
+    {
+      "epoch": 0.2336167240673394,
+      "grad_norm": 1.6316677331924438,
+      "learning_rate": 8.067640198762803e-07,
+      "loss": 0.9776,
+      "step": 9700
+    },
+    {
+      "epoch": 0.2337371450591267,
+      "grad_norm": 1.6081719398498535,
+      "learning_rate": 8.066372578845958e-07,
+      "loss": 0.9209,
+      "step": 9705
+    },
+    {
+      "epoch": 0.23385756605091398,
+      "grad_norm": 1.6020768880844116,
+      "learning_rate": 8.065104958929114e-07,
+      "loss": 0.9266,
+      "step": 9710
+    },
+    {
+      "epoch": 0.23397798704270129,
+      "grad_norm": 1.450314998626709,
+      "learning_rate": 8.063837339012271e-07,
+      "loss": 0.9394,
+      "step": 9715
+    },
+    {
+      "epoch": 0.23409840803448856,
+      "grad_norm": 1.5475916862487793,
+      "learning_rate": 8.062569719095426e-07,
+      "loss": 0.8923,
+      "step": 9720
+    },
+    {
+      "epoch": 0.23421882902627586,
+      "grad_norm": 1.7144578695297241,
+      "learning_rate": 8.061302099178581e-07,
+      "loss": 0.9136,
+      "step": 9725
+    },
+    {
+      "epoch": 0.23433925001806316,
+      "grad_norm": 1.5156981945037842,
+      "learning_rate": 8.060034479261738e-07,
+      "loss": 0.9153,
+      "step": 9730
+    },
+    {
+      "epoch": 0.23445967100985043,
+      "grad_norm": 1.4762510061264038,
+      "learning_rate": 8.058766859344893e-07,
+      "loss": 0.9274,
+      "step": 9735
+    },
+    {
+      "epoch": 0.23458009200163774,
+      "grad_norm": 1.5367364883422852,
+      "learning_rate": 8.05749923942805e-07,
+      "loss": 0.9553,
+      "step": 9740
+    },
+    {
+      "epoch": 0.234700512993425,
+      "grad_norm": 1.4681675434112549,
+      "learning_rate": 8.056231619511206e-07,
+      "loss": 0.9167,
+      "step": 9745
+    },
+    {
+      "epoch": 0.2348209339852123,
+      "grad_norm": 1.4918434619903564,
+      "learning_rate": 8.05496399959436e-07,
+      "loss": 0.963,
+      "step": 9750
+    },
+    {
+      "epoch": 0.23494135497699958,
+      "grad_norm": 1.7512080669403076,
+      "learning_rate": 8.053696379677517e-07,
+      "loss": 0.9161,
+      "step": 9755
+    },
+    {
+      "epoch": 0.23506177596878688,
+      "grad_norm": 1.8454680442810059,
+      "learning_rate": 8.052428759760673e-07,
+      "loss": 0.9616,
+      "step": 9760
+    },
+    {
+      "epoch": 0.23518219696057416,
+      "grad_norm": 1.601298213005066,
+      "learning_rate": 8.051161139843829e-07,
+      "loss": 0.9286,
+      "step": 9765
+    },
+    {
+      "epoch": 0.23530261795236146,
+      "grad_norm": 1.5492949485778809,
+      "learning_rate": 8.049893519926985e-07,
+      "loss": 0.9517,
+      "step": 9770
+    },
+    {
+      "epoch": 0.23542303894414873,
+      "grad_norm": 2.002368688583374,
+      "learning_rate": 8.04862590001014e-07,
+      "loss": 0.9469,
+      "step": 9775
+    },
+    {
+      "epoch": 0.23554345993593603,
+      "grad_norm": 1.533035159111023,
+      "learning_rate": 8.047358280093296e-07,
+      "loss": 0.9728,
+      "step": 9780
+    },
+    {
+      "epoch": 0.2356638809277233,
+      "grad_norm": 1.7091615200042725,
+      "learning_rate": 8.046090660176452e-07,
+      "loss": 0.915,
+      "step": 9785
+    },
+    {
+      "epoch": 0.2357843019195106,
+      "grad_norm": 1.8280829191207886,
+      "learning_rate": 8.044823040259609e-07,
+      "loss": 0.9154,
+      "step": 9790
+    },
+    {
+      "epoch": 0.2359047229112979,
+      "grad_norm": 1.37148916721344,
+      "learning_rate": 8.043555420342763e-07,
+      "loss": 0.9121,
+      "step": 9795
+    },
+    {
+      "epoch": 0.23602514390308518,
+      "grad_norm": 1.4839376211166382,
+      "learning_rate": 8.04228780042592e-07,
+      "loss": 0.8989,
+      "step": 9800
+    },
+    {
+      "epoch": 0.23614556489487248,
+      "grad_norm": 1.599485993385315,
+      "learning_rate": 8.041020180509076e-07,
+      "loss": 0.9382,
+      "step": 9805
+    },
+    {
+      "epoch": 0.23626598588665976,
+      "grad_norm": 1.343124270439148,
+      "learning_rate": 8.039752560592231e-07,
+      "loss": 0.9264,
+      "step": 9810
+    },
+    {
+      "epoch": 0.23638640687844706,
+      "grad_norm": 1.5841120481491089,
+      "learning_rate": 8.038484940675388e-07,
+      "loss": 0.9461,
+      "step": 9815
+    },
+    {
+      "epoch": 0.23650682787023433,
+      "grad_norm": 1.5181254148483276,
+      "learning_rate": 8.037217320758543e-07,
+      "loss": 0.9149,
+      "step": 9820
+    },
+    {
+      "epoch": 0.23662724886202163,
+      "grad_norm": 1.5661225318908691,
+      "learning_rate": 8.035949700841699e-07,
+      "loss": 0.9626,
+      "step": 9825
+    },
+    {
+      "epoch": 0.2367476698538089,
+      "grad_norm": 1.4676679372787476,
+      "learning_rate": 8.034682080924855e-07,
+      "loss": 0.968,
+      "step": 9830
+    },
+    {
+      "epoch": 0.2368680908455962,
+      "grad_norm": 1.5806342363357544,
+      "learning_rate": 8.033414461008012e-07,
+      "loss": 0.9607,
+      "step": 9835
+    },
+    {
+      "epoch": 0.23698851183738348,
+      "grad_norm": 1.432308316230774,
+      "learning_rate": 8.032146841091166e-07,
+      "loss": 0.9342,
+      "step": 9840
+    },
+    {
+      "epoch": 0.23710893282917078,
+      "grad_norm": 1.5676426887512207,
+      "learning_rate": 8.030879221174322e-07,
+      "loss": 0.96,
+      "step": 9845
+    },
+    {
+      "epoch": 0.23722935382095808,
+      "grad_norm": 1.8210023641586304,
+      "learning_rate": 8.029611601257479e-07,
+      "loss": 0.9001,
+      "step": 9850
+    },
+    {
+      "epoch": 0.23734977481274536,
+      "grad_norm": 1.5412278175354004,
+      "learning_rate": 8.028343981340634e-07,
+      "loss": 0.931,
+      "step": 9855
+    },
+    {
+      "epoch": 0.23747019580453266,
+      "grad_norm": 1.582090139389038,
+      "learning_rate": 8.027076361423791e-07,
+      "loss": 0.893,
+      "step": 9860
+    },
+    {
+      "epoch": 0.23759061679631993,
+      "grad_norm": 1.480709433555603,
+      "learning_rate": 8.025808741506946e-07,
+      "loss": 0.8721,
+      "step": 9865
+    },
+    {
+      "epoch": 0.23771103778810723,
+      "grad_norm": 1.7517716884613037,
+      "learning_rate": 8.024541121590101e-07,
+      "loss": 0.8898,
+      "step": 9870
+    },
+    {
+      "epoch": 0.2378314587798945,
+      "grad_norm": 1.5809065103530884,
+      "learning_rate": 8.023273501673258e-07,
+      "loss": 0.9582,
+      "step": 9875
+    },
+    {
+      "epoch": 0.2379518797716818,
+      "grad_norm": 1.604315161705017,
+      "learning_rate": 8.022005881756414e-07,
+      "loss": 0.9441,
+      "step": 9880
+    },
+    {
+      "epoch": 0.23807230076346908,
+      "grad_norm": 1.5863564014434814,
+      "learning_rate": 8.02073826183957e-07,
+      "loss": 0.9584,
+      "step": 9885
+    },
+    {
+      "epoch": 0.23819272175525638,
+      "grad_norm": 1.621246337890625,
+      "learning_rate": 8.019470641922725e-07,
+      "loss": 0.9417,
+      "step": 9890
+    },
+    {
+      "epoch": 0.23831314274704365,
+      "grad_norm": 2.1980340480804443,
+      "learning_rate": 8.018203022005881e-07,
+      "loss": 0.9585,
+      "step": 9895
+    },
+    {
+      "epoch": 0.23843356373883096,
+      "grad_norm": 1.7182064056396484,
+      "learning_rate": 8.016935402089037e-07,
+      "loss": 0.9086,
+      "step": 9900
+    },
+    {
+      "epoch": 0.23855398473061823,
+      "grad_norm": 1.6237313747406006,
+      "learning_rate": 8.015667782172193e-07,
+      "loss": 0.9415,
+      "step": 9905
+    },
+    {
+      "epoch": 0.23867440572240553,
+      "grad_norm": 1.5629863739013672,
+      "learning_rate": 8.014400162255349e-07,
+      "loss": 0.8983,
+      "step": 9910
+    },
+    {
+      "epoch": 0.23879482671419283,
+      "grad_norm": 1.561966896057129,
+      "learning_rate": 8.013132542338504e-07,
+      "loss": 0.9293,
+      "step": 9915
+    },
+    {
+      "epoch": 0.2389152477059801,
+      "grad_norm": 1.643320083618164,
+      "learning_rate": 8.011864922421661e-07,
+      "loss": 0.9716,
+      "step": 9920
+    },
+    {
+      "epoch": 0.2390356686977674,
+      "grad_norm": 1.4515066146850586,
+      "learning_rate": 8.010597302504817e-07,
+      "loss": 0.9052,
+      "step": 9925
+    },
+    {
+      "epoch": 0.23915608968955468,
+      "grad_norm": 1.4326725006103516,
+      "learning_rate": 8.009329682587972e-07,
+      "loss": 0.9318,
+      "step": 9930
+    },
+    {
+      "epoch": 0.23927651068134198,
+      "grad_norm": 1.5089689493179321,
+      "learning_rate": 8.008062062671128e-07,
+      "loss": 0.9239,
+      "step": 9935
+    },
+    {
+      "epoch": 0.23939693167312925,
+      "grad_norm": 1.687284231185913,
+      "learning_rate": 8.006794442754284e-07,
+      "loss": 0.9622,
+      "step": 9940
+    },
+    {
+      "epoch": 0.23951735266491656,
+      "grad_norm": 1.4424015283584595,
+      "learning_rate": 8.00552682283744e-07,
+      "loss": 0.928,
+      "step": 9945
+    },
+    {
+      "epoch": 0.23963777365670383,
+      "grad_norm": 1.5909708738327026,
+      "learning_rate": 8.004259202920596e-07,
+      "loss": 0.9228,
+      "step": 9950
+    },
+    {
+      "epoch": 0.23975819464849113,
+      "grad_norm": 1.5277528762817383,
+      "learning_rate": 8.002991583003752e-07,
+      "loss": 0.9105,
+      "step": 9955
+    },
+    {
+      "epoch": 0.2398786156402784,
+      "grad_norm": 1.580113172531128,
+      "learning_rate": 8.001723963086907e-07,
+      "loss": 0.9575,
+      "step": 9960
+    },
+    {
+      "epoch": 0.2399990366320657,
+      "grad_norm": 1.5940556526184082,
+      "learning_rate": 8.000456343170063e-07,
+      "loss": 0.9305,
+      "step": 9965
+    },
+    {
+      "epoch": 0.24011945762385298,
+      "grad_norm": 1.5910226106643677,
+      "learning_rate": 7.99918872325322e-07,
+      "loss": 1.0218,
+      "step": 9970
+    },
+    {
+      "epoch": 0.24023987861564028,
+      "grad_norm": 1.6788511276245117,
+      "learning_rate": 7.997921103336375e-07,
+      "loss": 0.9411,
+      "step": 9975
+    },
+    {
+      "epoch": 0.24036029960742758,
+      "grad_norm": 1.4881618022918701,
+      "learning_rate": 7.99665348341953e-07,
+      "loss": 0.8998,
+      "step": 9980
+    },
+    {
+      "epoch": 0.24048072059921485,
+      "grad_norm": 2.371474266052246,
+      "learning_rate": 7.995385863502687e-07,
+      "loss": 0.8955,
+      "step": 9985
+    },
+    {
+      "epoch": 0.24060114159100215,
+      "grad_norm": 1.70774245262146,
+      "learning_rate": 7.994118243585843e-07,
+      "loss": 0.987,
+      "step": 9990
+    },
+    {
+      "epoch": 0.24072156258278943,
+      "grad_norm": 1.4561922550201416,
+      "learning_rate": 7.992850623668999e-07,
+      "loss": 0.8907,
+      "step": 9995
+    },
+    {
+      "epoch": 0.24084198357457673,
+      "grad_norm": 2.2680671215057373,
+      "learning_rate": 7.991583003752155e-07,
+      "loss": 0.9029,
+      "step": 10000
+    },
+    {
+      "epoch": 0.240962404566364,
+      "grad_norm": 1.4901586771011353,
+      "learning_rate": 7.990315383835311e-07,
+      "loss": 0.9067,
+      "step": 10005
+    },
+    {
+      "epoch": 0.2410828255581513,
+      "grad_norm": 1.5713847875595093,
+      "learning_rate": 7.989047763918466e-07,
+      "loss": 0.9342,
+      "step": 10010
+    },
+    {
+      "epoch": 0.24120324654993858,
+      "grad_norm": 1.7380549907684326,
+      "learning_rate": 7.987780144001622e-07,
+      "loss": 0.9784,
+      "step": 10015
+    },
+    {
+      "epoch": 0.24132366754172588,
+      "grad_norm": 1.6125047206878662,
+      "learning_rate": 7.986512524084779e-07,
+      "loss": 0.9062,
+      "step": 10020
+    },
+    {
+      "epoch": 0.24144408853351315,
+      "grad_norm": 1.498434066772461,
+      "learning_rate": 7.985244904167933e-07,
+      "loss": 0.9183,
+      "step": 10025
+    },
+    {
+      "epoch": 0.24156450952530045,
+      "grad_norm": 1.5576609373092651,
+      "learning_rate": 7.98397728425109e-07,
+      "loss": 0.9168,
+      "step": 10030
+    },
+    {
+      "epoch": 0.24168493051708773,
+      "grad_norm": 1.649990439414978,
+      "learning_rate": 7.982709664334246e-07,
+      "loss": 0.972,
+      "step": 10035
+    },
+    {
+      "epoch": 0.24180535150887503,
+      "grad_norm": 1.5411405563354492,
+      "learning_rate": 7.981442044417402e-07,
+      "loss": 0.9177,
+      "step": 10040
+    },
+    {
+      "epoch": 0.24192577250066233,
+      "grad_norm": 1.6077088117599487,
+      "learning_rate": 7.980174424500558e-07,
+      "loss": 0.8901,
+      "step": 10045
+    },
+    {
+      "epoch": 0.2420461934924496,
+      "grad_norm": 1.8830534219741821,
+      "learning_rate": 7.978906804583713e-07,
+      "loss": 0.9117,
+      "step": 10050
+    },
+    {
+      "epoch": 0.2421666144842369,
+      "grad_norm": 1.7061887979507446,
+      "learning_rate": 7.977639184666869e-07,
+      "loss": 0.886,
+      "step": 10055
+    },
+    {
+      "epoch": 0.24228703547602418,
+      "grad_norm": 1.4365276098251343,
+      "learning_rate": 7.976371564750025e-07,
+      "loss": 0.8753,
+      "step": 10060
+    },
+    {
+      "epoch": 0.24240745646781148,
+      "grad_norm": 1.485558032989502,
+      "learning_rate": 7.975103944833182e-07,
+      "loss": 0.9064,
+      "step": 10065
+    },
+    {
+      "epoch": 0.24252787745959875,
+      "grad_norm": 1.5110009908676147,
+      "learning_rate": 7.973836324916337e-07,
+      "loss": 0.9667,
+      "step": 10070
+    },
+    {
+      "epoch": 0.24264829845138605,
+      "grad_norm": 1.6972064971923828,
+      "learning_rate": 7.972568704999492e-07,
+      "loss": 0.9475,
+      "step": 10075
+    },
+    {
+      "epoch": 0.24276871944317333,
+      "grad_norm": 2.102038621902466,
+      "learning_rate": 7.971301085082649e-07,
+      "loss": 0.9135,
+      "step": 10080
+    },
+    {
+      "epoch": 0.24288914043496063,
+      "grad_norm": 1.4368513822555542,
+      "learning_rate": 7.970033465165804e-07,
+      "loss": 0.9098,
+      "step": 10085
+    },
+    {
+      "epoch": 0.2430095614267479,
+      "grad_norm": 1.52772057056427,
+      "learning_rate": 7.968765845248961e-07,
+      "loss": 0.8988,
+      "step": 10090
+    },
+    {
+      "epoch": 0.2431299824185352,
+      "grad_norm": 1.4237383604049683,
+      "learning_rate": 7.967498225332116e-07,
+      "loss": 0.9366,
+      "step": 10095
+    },
+    {
+      "epoch": 0.24325040341032247,
+      "grad_norm": 1.4499160051345825,
+      "learning_rate": 7.966230605415271e-07,
+      "loss": 0.9305,
+      "step": 10100
+    },
+    {
+      "epoch": 0.24337082440210978,
+      "grad_norm": 1.6376252174377441,
+      "learning_rate": 7.964962985498428e-07,
+      "loss": 0.9434,
+      "step": 10105
+    },
+    {
+      "epoch": 0.24349124539389708,
+      "grad_norm": 1.7884248495101929,
+      "learning_rate": 7.963695365581584e-07,
+      "loss": 0.9995,
+      "step": 10110
+    },
+    {
+      "epoch": 0.24361166638568435,
+      "grad_norm": 1.5183720588684082,
+      "learning_rate": 7.96242774566474e-07,
+      "loss": 0.9227,
+      "step": 10115
+    },
+    {
+      "epoch": 0.24373208737747165,
+      "grad_norm": 1.5874552726745605,
+      "learning_rate": 7.961160125747895e-07,
+      "loss": 0.9153,
+      "step": 10120
+    },
+    {
+      "epoch": 0.24385250836925892,
+      "grad_norm": 1.6614394187927246,
+      "learning_rate": 7.959892505831052e-07,
+      "loss": 0.978,
+      "step": 10125
+    },
+    {
+      "epoch": 0.24397292936104623,
+      "grad_norm": 1.417792558670044,
+      "learning_rate": 7.958624885914207e-07,
+      "loss": 0.9653,
+      "step": 10130
+    },
+    {
+      "epoch": 0.2440933503528335,
+      "grad_norm": 1.610689401626587,
+      "learning_rate": 7.957357265997363e-07,
+      "loss": 0.9511,
+      "step": 10135
+    },
+    {
+      "epoch": 0.2442137713446208,
+      "grad_norm": 1.515844702720642,
+      "learning_rate": 7.95608964608052e-07,
+      "loss": 0.9546,
+      "step": 10140
+    },
+    {
+      "epoch": 0.24433419233640807,
+      "grad_norm": 1.5553603172302246,
+      "learning_rate": 7.954822026163674e-07,
+      "loss": 0.9197,
+      "step": 10145
+    },
+    {
+      "epoch": 0.24445461332819537,
+      "grad_norm": 1.635810375213623,
+      "learning_rate": 7.953554406246831e-07,
+      "loss": 0.8908,
+      "step": 10150
+    },
+    {
+      "epoch": 0.24457503431998265,
+      "grad_norm": 1.8756006956100464,
+      "learning_rate": 7.952286786329987e-07,
+      "loss": 0.9145,
+      "step": 10155
+    },
+    {
+      "epoch": 0.24469545531176995,
+      "grad_norm": 1.562914490699768,
+      "learning_rate": 7.951019166413142e-07,
+      "loss": 0.9254,
+      "step": 10160
+    },
+    {
+      "epoch": 0.24481587630355722,
+      "grad_norm": 1.4989361763000488,
+      "learning_rate": 7.949751546496298e-07,
+      "loss": 0.9285,
+      "step": 10165
+    },
+    {
+      "epoch": 0.24493629729534452,
+      "grad_norm": 1.4218214750289917,
+      "learning_rate": 7.948483926579454e-07,
+      "loss": 0.9399,
+      "step": 10170
+    },
+    {
+      "epoch": 0.24505671828713183,
+      "grad_norm": 1.5706713199615479,
+      "learning_rate": 7.94721630666261e-07,
+      "loss": 0.8914,
+      "step": 10175
+    },
+    {
+      "epoch": 0.2451771392789191,
+      "grad_norm": 1.4235652685165405,
+      "learning_rate": 7.945948686745766e-07,
+      "loss": 0.8733,
+      "step": 10180
+    },
+    {
+      "epoch": 0.2452975602707064,
+      "grad_norm": 1.735122561454773,
+      "learning_rate": 7.944681066828923e-07,
+      "loss": 0.9394,
+      "step": 10185
+    },
+    {
+      "epoch": 0.24541798126249367,
+      "grad_norm": 1.5530369281768799,
+      "learning_rate": 7.943413446912077e-07,
+      "loss": 0.9979,
+      "step": 10190
+    },
+    {
+      "epoch": 0.24553840225428097,
+      "grad_norm": 1.6291742324829102,
+      "learning_rate": 7.942145826995233e-07,
+      "loss": 0.9256,
+      "step": 10195
+    },
+    {
+      "epoch": 0.24565882324606825,
+      "grad_norm": 1.5617204904556274,
+      "learning_rate": 7.94087820707839e-07,
+      "loss": 0.9474,
+      "step": 10200
+    },
+    {
+      "epoch": 0.24577924423785555,
+      "grad_norm": 1.5999999046325684,
+      "learning_rate": 7.939610587161545e-07,
+      "loss": 0.9339,
+      "step": 10205
+    },
+    {
+      "epoch": 0.24589966522964282,
+      "grad_norm": 1.3905305862426758,
+      "learning_rate": 7.938342967244701e-07,
+      "loss": 0.8518,
+      "step": 10210
+    },
+    {
+      "epoch": 0.24602008622143012,
+      "grad_norm": 1.569653034210205,
+      "learning_rate": 7.937075347327857e-07,
+      "loss": 0.9487,
+      "step": 10215
+    },
+    {
+      "epoch": 0.2461405072132174,
+      "grad_norm": 1.578831434249878,
+      "learning_rate": 7.935807727411012e-07,
+      "loss": 0.9662,
+      "step": 10220
+    },
+    {
+      "epoch": 0.2462609282050047,
+      "grad_norm": 1.528179407119751,
+      "learning_rate": 7.934540107494169e-07,
+      "loss": 0.9369,
+      "step": 10225
+    },
+    {
+      "epoch": 0.24638134919679197,
+      "grad_norm": 1.5490789413452148,
+      "learning_rate": 7.933272487577325e-07,
+      "loss": 0.9889,
+      "step": 10230
+    },
+    {
+      "epoch": 0.24650177018857927,
+      "grad_norm": 1.6813278198242188,
+      "learning_rate": 7.93200486766048e-07,
+      "loss": 0.9001,
+      "step": 10235
+    },
+    {
+      "epoch": 0.24662219118036657,
+      "grad_norm": 1.4886800050735474,
+      "learning_rate": 7.930737247743636e-07,
+      "loss": 0.923,
+      "step": 10240
+    },
+    {
+      "epoch": 0.24674261217215385,
+      "grad_norm": 1.586333990097046,
+      "learning_rate": 7.929469627826793e-07,
+      "loss": 0.951,
+      "step": 10245
+    },
+    {
+      "epoch": 0.24686303316394115,
+      "grad_norm": 1.4950419664382935,
+      "learning_rate": 7.928202007909948e-07,
+      "loss": 0.9323,
+      "step": 10250
+    },
+    {
+      "epoch": 0.24698345415572842,
+      "grad_norm": 1.5455451011657715,
+      "learning_rate": 7.926934387993104e-07,
+      "loss": 0.9356,
+      "step": 10255
+    },
+    {
+      "epoch": 0.24710387514751572,
+      "grad_norm": 1.5218327045440674,
+      "learning_rate": 7.92566676807626e-07,
+      "loss": 0.9828,
+      "step": 10260
+    },
+    {
+      "epoch": 0.247224296139303,
+      "grad_norm": 1.6908859014511108,
+      "learning_rate": 7.924399148159415e-07,
+      "loss": 0.9144,
+      "step": 10265
+    },
+    {
+      "epoch": 0.2473447171310903,
+      "grad_norm": 1.6348868608474731,
+      "learning_rate": 7.923131528242572e-07,
+      "loss": 0.9528,
+      "step": 10270
+    },
+    {
+      "epoch": 0.24746513812287757,
+      "grad_norm": 1.3725874423980713,
+      "learning_rate": 7.921863908325728e-07,
+      "loss": 0.8838,
+      "step": 10275
+    },
+    {
+      "epoch": 0.24758555911466487,
+      "grad_norm": 1.5570913553237915,
+      "learning_rate": 7.920596288408882e-07,
+      "loss": 0.9295,
+      "step": 10280
+    },
+    {
+      "epoch": 0.24770598010645214,
+      "grad_norm": 1.600831151008606,
+      "learning_rate": 7.919328668492039e-07,
+      "loss": 0.9013,
+      "step": 10285
+    },
+    {
+      "epoch": 0.24782640109823945,
+      "grad_norm": 1.507856011390686,
+      "learning_rate": 7.918061048575195e-07,
+      "loss": 0.9489,
+      "step": 10290
+    },
+    {
+      "epoch": 0.24794682209002675,
+      "grad_norm": 1.6118507385253906,
+      "learning_rate": 7.916793428658351e-07,
+      "loss": 0.9631,
+      "step": 10295
+    },
+    {
+      "epoch": 0.24806724308181402,
+      "grad_norm": 1.5403980016708374,
+      "learning_rate": 7.915525808741507e-07,
+      "loss": 0.8982,
+      "step": 10300
+    },
+    {
+      "epoch": 0.24818766407360132,
+      "grad_norm": 1.4400405883789062,
+      "learning_rate": 7.914258188824663e-07,
+      "loss": 0.8841,
+      "step": 10305
+    },
+    {
+      "epoch": 0.2483080850653886,
+      "grad_norm": 1.6470706462860107,
+      "learning_rate": 7.912990568907818e-07,
+      "loss": 0.895,
+      "step": 10310
+    },
+    {
+      "epoch": 0.2484285060571759,
+      "grad_norm": 1.5725401639938354,
+      "learning_rate": 7.911722948990974e-07,
+      "loss": 0.9149,
+      "step": 10315
+    },
+    {
+      "epoch": 0.24854892704896317,
+      "grad_norm": 1.4755620956420898,
+      "learning_rate": 7.910455329074131e-07,
+      "loss": 0.9526,
+      "step": 10320
+    },
+    {
+      "epoch": 0.24866934804075047,
+      "grad_norm": 1.624035120010376,
+      "learning_rate": 7.909187709157285e-07,
+      "loss": 0.9191,
+      "step": 10325
+    },
+    {
+      "epoch": 0.24878976903253774,
+      "grad_norm": 1.6740853786468506,
+      "learning_rate": 7.907920089240442e-07,
+      "loss": 0.8927,
+      "step": 10330
+    },
+    {
+      "epoch": 0.24891019002432505,
+      "grad_norm": 1.940212607383728,
+      "learning_rate": 7.906652469323598e-07,
+      "loss": 0.9221,
+      "step": 10335
+    },
+    {
+      "epoch": 0.24903061101611232,
+      "grad_norm": 1.5053337812423706,
+      "learning_rate": 7.905384849406753e-07,
+      "loss": 0.8715,
+      "step": 10340
+    },
+    {
+      "epoch": 0.24915103200789962,
+      "grad_norm": 1.6379318237304688,
+      "learning_rate": 7.90411722948991e-07,
+      "loss": 0.9527,
+      "step": 10345
+    },
+    {
+      "epoch": 0.2492714529996869,
+      "grad_norm": 1.5555362701416016,
+      "learning_rate": 7.902849609573065e-07,
+      "loss": 0.9096,
+      "step": 10350
+    },
+    {
+      "epoch": 0.2493918739914742,
+      "grad_norm": 1.6231192350387573,
+      "learning_rate": 7.901581989656221e-07,
+      "loss": 0.9344,
+      "step": 10355
+    },
+    {
+      "epoch": 0.2495122949832615,
+      "grad_norm": 2.2351765632629395,
+      "learning_rate": 7.900314369739377e-07,
+      "loss": 0.9563,
+      "step": 10360
+    },
+    {
+      "epoch": 0.24963271597504877,
+      "grad_norm": 1.3904240131378174,
+      "learning_rate": 7.899046749822534e-07,
+      "loss": 0.9499,
+      "step": 10365
+    },
+    {
+      "epoch": 0.24975313696683607,
+      "grad_norm": 2.1326162815093994,
+      "learning_rate": 7.897779129905689e-07,
+      "loss": 0.922,
+      "step": 10370
+    },
+    {
+      "epoch": 0.24987355795862334,
+      "grad_norm": 1.416149377822876,
+      "learning_rate": 7.896511509988844e-07,
+      "loss": 0.9077,
+      "step": 10375
+    },
+    {
+      "epoch": 0.24999397895041064,
+      "grad_norm": 1.636558175086975,
+      "learning_rate": 7.895243890072001e-07,
+      "loss": 0.9489,
+      "step": 10380
+    },
+    {
+      "epoch": 0.2501143999421979,
+      "grad_norm": 1.5372464656829834,
+      "learning_rate": 7.893976270155156e-07,
+      "loss": 0.9633,
+      "step": 10385
+    },
+    {
+      "epoch": 0.2502348209339852,
+      "grad_norm": 1.4980571269989014,
+      "learning_rate": 7.892708650238313e-07,
+      "loss": 0.9452,
+      "step": 10390
+    },
+    {
+      "epoch": 0.2503552419257725,
+      "grad_norm": 1.3831257820129395,
+      "learning_rate": 7.891441030321468e-07,
+      "loss": 0.9252,
+      "step": 10395
+    },
+    {
+      "epoch": 0.2504756629175598,
+      "grad_norm": 1.4749293327331543,
+      "learning_rate": 7.890173410404623e-07,
+      "loss": 0.9759,
+      "step": 10400
+    },
+    {
+      "epoch": 0.25059608390934707,
+      "grad_norm": 1.7515690326690674,
+      "learning_rate": 7.88890579048778e-07,
+      "loss": 1.0089,
+      "step": 10405
+    },
+    {
+      "epoch": 0.25071650490113434,
+      "grad_norm": 1.518198013305664,
+      "learning_rate": 7.887638170570936e-07,
+      "loss": 0.9502,
+      "step": 10410
+    },
+    {
+      "epoch": 0.25083692589292167,
+      "grad_norm": 1.4665954113006592,
+      "learning_rate": 7.886370550654092e-07,
+      "loss": 0.9368,
+      "step": 10415
+    },
+    {
+      "epoch": 0.25095734688470894,
+      "grad_norm": 1.6156924962997437,
+      "learning_rate": 7.885102930737247e-07,
+      "loss": 0.9023,
+      "step": 10420
+    },
+    {
+      "epoch": 0.2510777678764962,
+      "grad_norm": 1.5096485614776611,
+      "learning_rate": 7.883835310820403e-07,
+      "loss": 0.9287,
+      "step": 10425
+    },
+    {
+      "epoch": 0.25119818886828355,
+      "grad_norm": 1.555070400238037,
+      "learning_rate": 7.882567690903559e-07,
+      "loss": 0.9033,
+      "step": 10430
+    },
+    {
+      "epoch": 0.2513186098600708,
+      "grad_norm": 1.5974732637405396,
+      "learning_rate": 7.881300070986715e-07,
+      "loss": 0.9553,
+      "step": 10435
+    },
+    {
+      "epoch": 0.2514390308518581,
+      "grad_norm": 1.684948205947876,
+      "learning_rate": 7.880032451069872e-07,
+      "loss": 0.9587,
+      "step": 10440
+    },
+    {
+      "epoch": 0.25155945184364537,
+      "grad_norm": 1.6145825386047363,
+      "learning_rate": 7.878764831153026e-07,
+      "loss": 0.9151,
+      "step": 10445
+    },
+    {
+      "epoch": 0.2516798728354327,
+      "grad_norm": 1.5612765550613403,
+      "learning_rate": 7.877497211236183e-07,
+      "loss": 0.9207,
+      "step": 10450
+    },
+    {
+      "epoch": 0.25180029382721997,
+      "grad_norm": 1.639469027519226,
+      "learning_rate": 7.876229591319339e-07,
+      "loss": 0.9319,
+      "step": 10455
+    },
+    {
+      "epoch": 0.25192071481900724,
+      "grad_norm": 1.57554292678833,
+      "learning_rate": 7.874961971402494e-07,
+      "loss": 0.8741,
+      "step": 10460
+    },
+    {
+      "epoch": 0.2520411358107945,
+      "grad_norm": 1.6669059991836548,
+      "learning_rate": 7.87369435148565e-07,
+      "loss": 0.9171,
+      "step": 10465
+    },
+    {
+      "epoch": 0.25216155680258184,
+      "grad_norm": 1.6609137058258057,
+      "learning_rate": 7.872426731568806e-07,
+      "loss": 0.947,
+      "step": 10470
+    },
+    {
+      "epoch": 0.2522819777943691,
+      "grad_norm": 1.5579173564910889,
+      "learning_rate": 7.871159111651962e-07,
+      "loss": 0.9089,
+      "step": 10475
+    },
+    {
+      "epoch": 0.2524023987861564,
+      "grad_norm": 1.6061656475067139,
+      "learning_rate": 7.869891491735118e-07,
+      "loss": 0.9144,
+      "step": 10480
+    },
+    {
+      "epoch": 0.25252281977794366,
+      "grad_norm": 1.5641448497772217,
+      "learning_rate": 7.868623871818275e-07,
+      "loss": 0.9066,
+      "step": 10485
+    },
+    {
+      "epoch": 0.252643240769731,
+      "grad_norm": 1.3542393445968628,
+      "learning_rate": 7.867356251901429e-07,
+      "loss": 0.9641,
+      "step": 10490
+    },
+    {
+      "epoch": 0.25276366176151827,
+      "grad_norm": 1.599341869354248,
+      "learning_rate": 7.866088631984585e-07,
+      "loss": 0.8958,
+      "step": 10495
+    },
+    {
+      "epoch": 0.25288408275330554,
+      "grad_norm": 1.5220602750778198,
+      "learning_rate": 7.864821012067742e-07,
+      "loss": 0.9044,
+      "step": 10500
+    },
+    {
+      "epoch": 0.25300450374509287,
+      "grad_norm": 1.562246561050415,
+      "learning_rate": 7.863553392150897e-07,
+      "loss": 0.8578,
+      "step": 10505
+    },
+    {
+      "epoch": 0.25312492473688014,
+      "grad_norm": 1.6219570636749268,
+      "learning_rate": 7.862285772234053e-07,
+      "loss": 0.9255,
+      "step": 10510
+    },
+    {
+      "epoch": 0.2532453457286674,
+      "grad_norm": 1.7990187406539917,
+      "learning_rate": 7.861018152317209e-07,
+      "loss": 0.9217,
+      "step": 10515
+    },
+    {
+      "epoch": 0.2533657667204547,
+      "grad_norm": 1.477607250213623,
+      "learning_rate": 7.859750532400364e-07,
+      "loss": 0.9559,
+      "step": 10520
+    },
+    {
+      "epoch": 0.253486187712242,
+      "grad_norm": 1.5525449514389038,
+      "learning_rate": 7.858482912483521e-07,
+      "loss": 0.8984,
+      "step": 10525
+    },
+    {
+      "epoch": 0.2536066087040293,
+      "grad_norm": 1.4910857677459717,
+      "learning_rate": 7.857215292566677e-07,
+      "loss": 0.8982,
+      "step": 10530
+    },
+    {
+      "epoch": 0.25372702969581656,
+      "grad_norm": 1.5159627199172974,
+      "learning_rate": 7.855947672649832e-07,
+      "loss": 0.9465,
+      "step": 10535
+    },
+    {
+      "epoch": 0.25384745068760384,
+      "grad_norm": 1.797110676765442,
+      "learning_rate": 7.854680052732988e-07,
+      "loss": 0.9194,
+      "step": 10540
+    },
+    {
+      "epoch": 0.25396787167939117,
+      "grad_norm": 1.5768470764160156,
+      "learning_rate": 7.853412432816144e-07,
+      "loss": 0.9684,
+      "step": 10545
+    },
+    {
+      "epoch": 0.25408829267117844,
+      "grad_norm": 1.5663419961929321,
+      "learning_rate": 7.8521448128993e-07,
+      "loss": 0.9466,
+      "step": 10550
+    },
+    {
+      "epoch": 0.2542087136629657,
+      "grad_norm": 1.6284027099609375,
+      "learning_rate": 7.850877192982455e-07,
+      "loss": 0.9652,
+      "step": 10555
+    },
+    {
+      "epoch": 0.25432913465475304,
+      "grad_norm": 1.629702091217041,
+      "learning_rate": 7.849609573065612e-07,
+      "loss": 0.9266,
+      "step": 10560
+    },
+    {
+      "epoch": 0.2544495556465403,
+      "grad_norm": 1.8558638095855713,
+      "learning_rate": 7.848341953148767e-07,
+      "loss": 0.9305,
+      "step": 10565
+    },
+    {
+      "epoch": 0.2545699766383276,
+      "grad_norm": 1.5336354970932007,
+      "learning_rate": 7.847074333231924e-07,
+      "loss": 0.9098,
+      "step": 10570
+    },
+    {
+      "epoch": 0.25469039763011486,
+      "grad_norm": 1.4641679525375366,
+      "learning_rate": 7.84580671331508e-07,
+      "loss": 0.9908,
+      "step": 10575
+    },
+    {
+      "epoch": 0.2548108186219022,
+      "grad_norm": 1.5101847648620605,
+      "learning_rate": 7.844539093398234e-07,
+      "loss": 0.9182,
+      "step": 10580
+    },
+    {
+      "epoch": 0.25493123961368946,
+      "grad_norm": 1.5551209449768066,
+      "learning_rate": 7.843271473481391e-07,
+      "loss": 0.9541,
+      "step": 10585
+    },
+    {
+      "epoch": 0.25505166060547674,
+      "grad_norm": 1.5243884325027466,
+      "learning_rate": 7.842003853564547e-07,
+      "loss": 0.9704,
+      "step": 10590
+    },
+    {
+      "epoch": 0.255172081597264,
+      "grad_norm": 1.464455485343933,
+      "learning_rate": 7.840736233647703e-07,
+      "loss": 0.8812,
+      "step": 10595
+    },
+    {
+      "epoch": 0.25529250258905134,
+      "grad_norm": 1.4903156757354736,
+      "learning_rate": 7.839468613730859e-07,
+      "loss": 0.8868,
+      "step": 10600
+    },
+    {
+      "epoch": 0.2554129235808386,
+      "grad_norm": 1.581946611404419,
+      "learning_rate": 7.838200993814014e-07,
+      "loss": 0.8508,
+      "step": 10605
+    },
+    {
+      "epoch": 0.2555333445726259,
+      "grad_norm": 1.4839967489242554,
+      "learning_rate": 7.83693337389717e-07,
+      "loss": 0.9232,
+      "step": 10610
+    },
+    {
+      "epoch": 0.2556537655644132,
+      "grad_norm": 1.7015328407287598,
+      "learning_rate": 7.835665753980326e-07,
+      "loss": 0.9166,
+      "step": 10615
+    },
+    {
+      "epoch": 0.2557741865562005,
+      "grad_norm": 1.6252477169036865,
+      "learning_rate": 7.834398134063483e-07,
+      "loss": 0.9375,
+      "step": 10620
+    },
+    {
+      "epoch": 0.25589460754798776,
+      "grad_norm": 1.5471183061599731,
+      "learning_rate": 7.833130514146637e-07,
+      "loss": 0.918,
+      "step": 10625
+    },
+    {
+      "epoch": 0.25601502853977504,
+      "grad_norm": 1.4048386812210083,
+      "learning_rate": 7.831862894229793e-07,
+      "loss": 0.9263,
+      "step": 10630
+    },
+    {
+      "epoch": 0.25613544953156236,
+      "grad_norm": 1.5877938270568848,
+      "learning_rate": 7.83059527431295e-07,
+      "loss": 0.9965,
+      "step": 10635
+    },
+    {
+      "epoch": 0.25625587052334964,
+      "grad_norm": 1.8080618381500244,
+      "learning_rate": 7.829327654396105e-07,
+      "loss": 0.9158,
+      "step": 10640
+    },
+    {
+      "epoch": 0.2563762915151369,
+      "grad_norm": 1.533905267715454,
+      "learning_rate": 7.828060034479262e-07,
+      "loss": 0.9173,
+      "step": 10645
+    },
+    {
+      "epoch": 0.2564967125069242,
+      "grad_norm": 1.73739492893219,
+      "learning_rate": 7.826792414562417e-07,
+      "loss": 0.9389,
+      "step": 10650
+    },
+    {
+      "epoch": 0.2566171334987115,
+      "grad_norm": 1.486688494682312,
+      "learning_rate": 7.825524794645573e-07,
+      "loss": 0.959,
+      "step": 10655
+    },
+    {
+      "epoch": 0.2567375544904988,
+      "grad_norm": 1.5035080909729004,
+      "learning_rate": 7.824257174728729e-07,
+      "loss": 0.9039,
+      "step": 10660
+    },
+    {
+      "epoch": 0.25685797548228606,
+      "grad_norm": 1.618054986000061,
+      "learning_rate": 7.822989554811885e-07,
+      "loss": 0.9412,
+      "step": 10665
+    },
+    {
+      "epoch": 0.25697839647407333,
+      "grad_norm": 1.6609909534454346,
+      "learning_rate": 7.82172193489504e-07,
+      "loss": 0.8911,
+      "step": 10670
+    },
+    {
+      "epoch": 0.25709881746586066,
+      "grad_norm": 1.6681872606277466,
+      "learning_rate": 7.820454314978196e-07,
+      "loss": 0.9248,
+      "step": 10675
+    },
+    {
+      "epoch": 0.25721923845764794,
+      "grad_norm": 1.5016465187072754,
+      "learning_rate": 7.819186695061353e-07,
+      "loss": 0.9441,
+      "step": 10680
+    },
+    {
+      "epoch": 0.2573396594494352,
+      "grad_norm": 1.589369297027588,
+      "learning_rate": 7.817919075144508e-07,
+      "loss": 0.9639,
+      "step": 10685
+    },
+    {
+      "epoch": 0.25746008044122254,
+      "grad_norm": 1.712123155593872,
+      "learning_rate": 7.816651455227664e-07,
+      "loss": 0.9627,
+      "step": 10690
+    },
+    {
+      "epoch": 0.2575805014330098,
+      "grad_norm": 1.4570205211639404,
+      "learning_rate": 7.81538383531082e-07,
+      "loss": 0.8918,
+      "step": 10695
+    },
+    {
+      "epoch": 0.2577009224247971,
+      "grad_norm": 1.4842736721038818,
+      "learning_rate": 7.814116215393975e-07,
+      "loss": 0.9268,
+      "step": 10700
+    },
+    {
+      "epoch": 0.25782134341658436,
+      "grad_norm": 1.4311498403549194,
+      "learning_rate": 7.812848595477132e-07,
+      "loss": 0.9513,
+      "step": 10705
+    },
+    {
+      "epoch": 0.2579417644083717,
+      "grad_norm": 1.791871190071106,
+      "learning_rate": 7.811580975560288e-07,
+      "loss": 0.9687,
+      "step": 10710
+    },
+    {
+      "epoch": 0.25806218540015896,
+      "grad_norm": 1.471676230430603,
+      "learning_rate": 7.810313355643444e-07,
+      "loss": 0.9213,
+      "step": 10715
+    },
+    {
+      "epoch": 0.25818260639194623,
+      "grad_norm": 1.6426857709884644,
+      "learning_rate": 7.809045735726599e-07,
+      "loss": 0.9445,
+      "step": 10720
+    },
+    {
+      "epoch": 0.2583030273837335,
+      "grad_norm": 1.7955909967422485,
+      "learning_rate": 7.807778115809755e-07,
+      "loss": 0.8952,
+      "step": 10725
+    },
+    {
+      "epoch": 0.25842344837552084,
+      "grad_norm": 1.5316821336746216,
+      "learning_rate": 7.806510495892911e-07,
+      "loss": 0.8781,
+      "step": 10730
+    },
+    {
+      "epoch": 0.2585438693673081,
+      "grad_norm": 1.5494688749313354,
+      "learning_rate": 7.805242875976067e-07,
+      "loss": 0.9542,
+      "step": 10735
+    },
+    {
+      "epoch": 0.2586642903590954,
+      "grad_norm": 1.5220109224319458,
+      "learning_rate": 7.803975256059223e-07,
+      "loss": 0.9104,
+      "step": 10740
+    },
+    {
+      "epoch": 0.2587847113508827,
+      "grad_norm": 1.5227622985839844,
+      "learning_rate": 7.802707636142378e-07,
+      "loss": 0.9549,
+      "step": 10745
+    },
+    {
+      "epoch": 0.25890513234267,
+      "grad_norm": 1.5300692319869995,
+      "learning_rate": 7.801440016225534e-07,
+      "loss": 0.9111,
+      "step": 10750
+    },
+    {
+      "epoch": 0.25902555333445726,
+      "grad_norm": 1.4417011737823486,
+      "learning_rate": 7.800172396308691e-07,
+      "loss": 0.8647,
+      "step": 10755
+    },
+    {
+      "epoch": 0.25914597432624453,
+      "grad_norm": 1.893277645111084,
+      "learning_rate": 7.798904776391846e-07,
+      "loss": 0.9442,
+      "step": 10760
+    },
+    {
+      "epoch": 0.25926639531803186,
+      "grad_norm": 1.6196260452270508,
+      "learning_rate": 7.797637156475002e-07,
+      "loss": 0.9151,
+      "step": 10765
+    },
+    {
+      "epoch": 0.25938681630981913,
+      "grad_norm": 1.7305158376693726,
+      "learning_rate": 7.796369536558158e-07,
+      "loss": 0.9412,
+      "step": 10770
+    },
+    {
+      "epoch": 0.2595072373016064,
+      "grad_norm": 1.6203160285949707,
+      "learning_rate": 7.795101916641314e-07,
+      "loss": 0.9353,
+      "step": 10775
+    },
+    {
+      "epoch": 0.2596276582933937,
+      "grad_norm": 1.3651759624481201,
+      "learning_rate": 7.79383429672447e-07,
+      "loss": 0.9142,
+      "step": 10780
+    },
+    {
+      "epoch": 0.259748079285181,
+      "grad_norm": 1.571115493774414,
+      "learning_rate": 7.792566676807626e-07,
+      "loss": 0.9168,
+      "step": 10785
+    },
+    {
+      "epoch": 0.2598685002769683,
+      "grad_norm": 1.6559398174285889,
+      "learning_rate": 7.791299056890781e-07,
+      "loss": 0.9628,
+      "step": 10790
+    },
+    {
+      "epoch": 0.25998892126875556,
+      "grad_norm": 1.8212372064590454,
+      "learning_rate": 7.790031436973937e-07,
+      "loss": 0.9273,
+      "step": 10795
+    },
+    {
+      "epoch": 0.26010934226054283,
+      "grad_norm": 1.6014947891235352,
+      "learning_rate": 7.788763817057094e-07,
+      "loss": 0.9179,
+      "step": 10800
+    },
+    {
+      "epoch": 0.26022976325233016,
+      "grad_norm": 2.0242059230804443,
+      "learning_rate": 7.787496197140249e-07,
+      "loss": 0.9219,
+      "step": 10805
+    },
+    {
+      "epoch": 0.26035018424411743,
+      "grad_norm": 1.4445858001708984,
+      "learning_rate": 7.786228577223404e-07,
+      "loss": 0.9384,
+      "step": 10810
+    },
+    {
+      "epoch": 0.2604706052359047,
+      "grad_norm": 1.4920601844787598,
+      "learning_rate": 7.784960957306561e-07,
+      "loss": 0.9833,
+      "step": 10815
+    },
+    {
+      "epoch": 0.26059102622769204,
+      "grad_norm": 1.5067740678787231,
+      "learning_rate": 7.783693337389716e-07,
+      "loss": 0.8776,
+      "step": 10820
+    },
+    {
+      "epoch": 0.2607114472194793,
+      "grad_norm": 1.4836264848709106,
+      "learning_rate": 7.782425717472873e-07,
+      "loss": 0.9002,
+      "step": 10825
+    },
+    {
+      "epoch": 0.2608318682112666,
+      "grad_norm": 1.4572845697402954,
+      "learning_rate": 7.781158097556029e-07,
+      "loss": 0.9199,
+      "step": 10830
+    },
+    {
+      "epoch": 0.26095228920305386,
+      "grad_norm": 1.5106346607208252,
+      "learning_rate": 7.779890477639183e-07,
+      "loss": 0.8979,
+      "step": 10835
+    },
+    {
+      "epoch": 0.2610727101948412,
+      "grad_norm": 1.5475988388061523,
+      "learning_rate": 7.77862285772234e-07,
+      "loss": 0.9266,
+      "step": 10840
+    },
+    {
+      "epoch": 0.26119313118662846,
+      "grad_norm": 1.5073845386505127,
+      "learning_rate": 7.777355237805496e-07,
+      "loss": 0.9067,
+      "step": 10845
+    },
+    {
+      "epoch": 0.26131355217841573,
+      "grad_norm": 1.6924643516540527,
+      "learning_rate": 7.776087617888652e-07,
+      "loss": 0.9067,
+      "step": 10850
+    },
+    {
+      "epoch": 0.261433973170203,
+      "grad_norm": 1.3849414587020874,
+      "learning_rate": 7.774819997971807e-07,
+      "loss": 0.9451,
+      "step": 10855
+    },
+    {
+      "epoch": 0.26155439416199033,
+      "grad_norm": 1.6200286149978638,
+      "learning_rate": 7.773552378054964e-07,
+      "loss": 0.9167,
+      "step": 10860
+    },
+    {
+      "epoch": 0.2616748151537776,
+      "grad_norm": 1.5841484069824219,
+      "learning_rate": 7.772284758138119e-07,
+      "loss": 0.9103,
+      "step": 10865
+    },
+    {
+      "epoch": 0.2617952361455649,
+      "grad_norm": 1.4382582902908325,
+      "learning_rate": 7.771017138221275e-07,
+      "loss": 0.9266,
+      "step": 10870
+    },
+    {
+      "epoch": 0.2619156571373522,
+      "grad_norm": 1.5426464080810547,
+      "learning_rate": 7.769749518304432e-07,
+      "loss": 0.8977,
+      "step": 10875
+    },
+    {
+      "epoch": 0.2620360781291395,
+      "grad_norm": 3.087616443634033,
+      "learning_rate": 7.768481898387586e-07,
+      "loss": 0.9505,
+      "step": 10880
+    },
+    {
+      "epoch": 0.26215649912092676,
+      "grad_norm": 1.3389363288879395,
+      "learning_rate": 7.767214278470743e-07,
+      "loss": 0.9173,
+      "step": 10885
+    },
+    {
+      "epoch": 0.26227692011271403,
+      "grad_norm": 1.723882794380188,
+      "learning_rate": 7.765946658553899e-07,
+      "loss": 0.9164,
+      "step": 10890
+    },
+    {
+      "epoch": 0.26239734110450136,
+      "grad_norm": 1.563045620918274,
+      "learning_rate": 7.764679038637054e-07,
+      "loss": 0.8643,
+      "step": 10895
+    },
+    {
+      "epoch": 0.26251776209628863,
+      "grad_norm": 1.4408845901489258,
+      "learning_rate": 7.763411418720211e-07,
+      "loss": 0.9516,
+      "step": 10900
+    },
+    {
+      "epoch": 0.2626381830880759,
+      "grad_norm": 1.5494115352630615,
+      "learning_rate": 7.762143798803366e-07,
+      "loss": 0.9362,
+      "step": 10905
+    },
+    {
+      "epoch": 0.2627586040798632,
+      "grad_norm": 1.701547384262085,
+      "learning_rate": 7.760876178886522e-07,
+      "loss": 0.9754,
+      "step": 10910
+    },
+    {
+      "epoch": 0.2628790250716505,
+      "grad_norm": 1.7622368335723877,
+      "learning_rate": 7.759608558969678e-07,
+      "loss": 0.9356,
+      "step": 10915
+    },
+    {
+      "epoch": 0.2629994460634378,
+      "grad_norm": 1.4967761039733887,
+      "learning_rate": 7.758340939052835e-07,
+      "loss": 0.9762,
+      "step": 10920
+    },
+    {
+      "epoch": 0.26311986705522505,
+      "grad_norm": 1.4345545768737793,
+      "learning_rate": 7.757073319135989e-07,
+      "loss": 0.9089,
+      "step": 10925
+    },
+    {
+      "epoch": 0.2632402880470124,
+      "grad_norm": 1.3745805025100708,
+      "learning_rate": 7.755805699219145e-07,
+      "loss": 0.9402,
+      "step": 10930
+    },
+    {
+      "epoch": 0.26336070903879966,
+      "grad_norm": 1.5181835889816284,
+      "learning_rate": 7.754538079302302e-07,
+      "loss": 0.9539,
+      "step": 10935
+    },
+    {
+      "epoch": 0.26348113003058693,
+      "grad_norm": 1.359045386314392,
+      "learning_rate": 7.753270459385457e-07,
+      "loss": 0.8787,
+      "step": 10940
+    },
+    {
+      "epoch": 0.2636015510223742,
+      "grad_norm": 1.57675302028656,
+      "learning_rate": 7.752002839468614e-07,
+      "loss": 0.9159,
+      "step": 10945
+    },
+    {
+      "epoch": 0.26372197201416153,
+      "grad_norm": 1.686819314956665,
+      "learning_rate": 7.750735219551769e-07,
+      "loss": 0.9341,
+      "step": 10950
+    },
+    {
+      "epoch": 0.2638423930059488,
+      "grad_norm": 1.331228256225586,
+      "learning_rate": 7.749467599634924e-07,
+      "loss": 0.9419,
+      "step": 10955
+    },
+    {
+      "epoch": 0.2639628139977361,
+      "grad_norm": 1.553196907043457,
+      "learning_rate": 7.748199979718081e-07,
+      "loss": 0.9446,
+      "step": 10960
+    },
+    {
+      "epoch": 0.26408323498952335,
+      "grad_norm": 1.725075125694275,
+      "learning_rate": 7.746932359801237e-07,
+      "loss": 0.9115,
+      "step": 10965
+    },
+    {
+      "epoch": 0.2642036559813107,
+      "grad_norm": 1.7521941661834717,
+      "learning_rate": 7.745664739884392e-07,
+      "loss": 0.9618,
+      "step": 10970
+    },
+    {
+      "epoch": 0.26432407697309795,
+      "grad_norm": 1.7503281831741333,
+      "learning_rate": 7.744397119967548e-07,
+      "loss": 0.9227,
+      "step": 10975
+    },
+    {
+      "epoch": 0.26444449796488523,
+      "grad_norm": 1.615695595741272,
+      "learning_rate": 7.743129500050705e-07,
+      "loss": 0.9125,
+      "step": 10980
+    },
+    {
+      "epoch": 0.2645649189566725,
+      "grad_norm": 1.4781994819641113,
+      "learning_rate": 7.74186188013386e-07,
+      "loss": 0.9136,
+      "step": 10985
+    },
+    {
+      "epoch": 0.26468533994845983,
+      "grad_norm": 1.5761228799819946,
+      "learning_rate": 7.740594260217016e-07,
+      "loss": 0.9014,
+      "step": 10990
+    },
+    {
+      "epoch": 0.2648057609402471,
+      "grad_norm": 1.7451871633529663,
+      "learning_rate": 7.739326640300172e-07,
+      "loss": 0.9402,
+      "step": 10995
+    },
+    {
+      "epoch": 0.2649261819320344,
+      "grad_norm": 1.504990577697754,
+      "learning_rate": 7.738059020383327e-07,
+      "loss": 0.9071,
+      "step": 11000
+    },
+    {
+      "epoch": 0.2650466029238217,
+      "grad_norm": 1.5758501291275024,
+      "learning_rate": 7.736791400466484e-07,
+      "loss": 0.9766,
+      "step": 11005
+    },
+    {
+      "epoch": 0.265167023915609,
+      "grad_norm": 1.4429376125335693,
+      "learning_rate": 7.73552378054964e-07,
+      "loss": 0.9063,
+      "step": 11010
+    },
+    {
+      "epoch": 0.26528744490739625,
+      "grad_norm": 1.5816013813018799,
+      "learning_rate": 7.734256160632797e-07,
+      "loss": 0.9162,
+      "step": 11015
+    },
+    {
+      "epoch": 0.2654078658991835,
+      "grad_norm": 1.6807634830474854,
+      "learning_rate": 7.732988540715951e-07,
+      "loss": 0.938,
+      "step": 11020
+    },
+    {
+      "epoch": 0.26552828689097085,
+      "grad_norm": 1.447342038154602,
+      "learning_rate": 7.731720920799107e-07,
+      "loss": 0.9361,
+      "step": 11025
+    },
+    {
+      "epoch": 0.26564870788275813,
+      "grad_norm": 1.5453587770462036,
+      "learning_rate": 7.730453300882264e-07,
+      "loss": 0.9549,
+      "step": 11030
+    },
+    {
+      "epoch": 0.2657691288745454,
+      "grad_norm": 1.658867597579956,
+      "learning_rate": 7.729185680965419e-07,
+      "loss": 0.9225,
+      "step": 11035
+    },
+    {
+      "epoch": 0.2658895498663327,
+      "grad_norm": 1.4685916900634766,
+      "learning_rate": 7.727918061048575e-07,
+      "loss": 0.8776,
+      "step": 11040
+    },
+    {
+      "epoch": 0.26600997085812,
+      "grad_norm": 1.5075732469558716,
+      "learning_rate": 7.726650441131731e-07,
+      "loss": 0.9686,
+      "step": 11045
+    },
+    {
+      "epoch": 0.2661303918499073,
+      "grad_norm": 1.5110318660736084,
+      "learning_rate": 7.725382821214886e-07,
+      "loss": 0.9034,
+      "step": 11050
+    },
+    {
+      "epoch": 0.26625081284169455,
+      "grad_norm": 1.489485740661621,
+      "learning_rate": 7.724115201298043e-07,
+      "loss": 0.9587,
+      "step": 11055
+    },
+    {
+      "epoch": 0.2663712338334819,
+      "grad_norm": 1.6650078296661377,
+      "learning_rate": 7.722847581381199e-07,
+      "loss": 0.9401,
+      "step": 11060
+    },
+    {
+      "epoch": 0.26649165482526915,
+      "grad_norm": 1.5072864294052124,
+      "learning_rate": 7.721579961464354e-07,
+      "loss": 0.8722,
+      "step": 11065
+    },
+    {
+      "epoch": 0.2666120758170564,
+      "grad_norm": 1.5710713863372803,
+      "learning_rate": 7.72031234154751e-07,
+      "loss": 0.9457,
+      "step": 11070
+    },
+    {
+      "epoch": 0.2667324968088437,
+      "grad_norm": 1.441070795059204,
+      "learning_rate": 7.719044721630666e-07,
+      "loss": 0.9234,
+      "step": 11075
+    },
+    {
+      "epoch": 0.26685291780063103,
+      "grad_norm": 1.599685549736023,
+      "learning_rate": 7.717777101713822e-07,
+      "loss": 0.939,
+      "step": 11080
+    },
+    {
+      "epoch": 0.2669733387924183,
+      "grad_norm": 1.5153743028640747,
+      "learning_rate": 7.716509481796978e-07,
+      "loss": 0.9616,
+      "step": 11085
+    },
+    {
+      "epoch": 0.2670937597842056,
+      "grad_norm": 1.5871284008026123,
+      "learning_rate": 7.715241861880134e-07,
+      "loss": 0.9913,
+      "step": 11090
+    },
+    {
+      "epoch": 0.26721418077599285,
+      "grad_norm": 1.5965393781661987,
+      "learning_rate": 7.713974241963289e-07,
+      "loss": 0.9124,
+      "step": 11095
+    },
+    {
+      "epoch": 0.2673346017677802,
+      "grad_norm": 1.5728003978729248,
+      "learning_rate": 7.712706622046446e-07,
+      "loss": 0.9236,
+      "step": 11100
+    },
+    {
+      "epoch": 0.26745502275956745,
+      "grad_norm": 1.4616875648498535,
+      "learning_rate": 7.711439002129602e-07,
+      "loss": 0.941,
+      "step": 11105
+    },
+    {
+      "epoch": 0.2675754437513547,
+      "grad_norm": 1.7237117290496826,
+      "learning_rate": 7.710171382212756e-07,
+      "loss": 0.9377,
+      "step": 11110
+    },
+    {
+      "epoch": 0.267695864743142,
+      "grad_norm": 1.513432502746582,
+      "learning_rate": 7.708903762295913e-07,
+      "loss": 0.9646,
+      "step": 11115
+    },
+    {
+      "epoch": 0.2678162857349293,
+      "grad_norm": 1.4135527610778809,
+      "learning_rate": 7.707636142379069e-07,
+      "loss": 0.9422,
+      "step": 11120
+    },
+    {
+      "epoch": 0.2679367067267166,
+      "grad_norm": 1.4501285552978516,
+      "learning_rate": 7.706368522462225e-07,
+      "loss": 0.9505,
+      "step": 11125
+    },
+    {
+      "epoch": 0.2680571277185039,
+      "grad_norm": 1.7850518226623535,
+      "learning_rate": 7.705100902545381e-07,
+      "loss": 0.9199,
+      "step": 11130
+    },
+    {
+      "epoch": 0.2681775487102912,
+      "grad_norm": 1.5002425909042358,
+      "learning_rate": 7.703833282628536e-07,
+      "loss": 0.9142,
+      "step": 11135
+    },
+    {
+      "epoch": 0.2682979697020785,
+      "grad_norm": 1.6325747966766357,
+      "learning_rate": 7.702565662711692e-07,
+      "loss": 0.9115,
+      "step": 11140
+    },
+    {
+      "epoch": 0.26841839069386575,
+      "grad_norm": 1.522478699684143,
+      "learning_rate": 7.701298042794848e-07,
+      "loss": 0.9923,
+      "step": 11145
+    },
+    {
+      "epoch": 0.268538811685653,
+      "grad_norm": 1.7870874404907227,
+      "learning_rate": 7.700030422878005e-07,
+      "loss": 0.9462,
+      "step": 11150
+    },
+    {
+      "epoch": 0.26865923267744035,
+      "grad_norm": 1.5112918615341187,
+      "learning_rate": 7.698762802961159e-07,
+      "loss": 0.8391,
+      "step": 11155
+    },
+    {
+      "epoch": 0.2687796536692276,
+      "grad_norm": 1.603967308998108,
+      "learning_rate": 7.697495183044315e-07,
+      "loss": 0.9103,
+      "step": 11160
+    },
+    {
+      "epoch": 0.2689000746610149,
+      "grad_norm": 1.631317377090454,
+      "learning_rate": 7.696227563127472e-07,
+      "loss": 0.9129,
+      "step": 11165
+    },
+    {
+      "epoch": 0.26902049565280217,
+      "grad_norm": 1.62999427318573,
+      "learning_rate": 7.694959943210627e-07,
+      "loss": 0.9031,
+      "step": 11170
+    },
+    {
+      "epoch": 0.2691409166445895,
+      "grad_norm": 1.3878083229064941,
+      "learning_rate": 7.693692323293784e-07,
+      "loss": 0.919,
+      "step": 11175
+    },
+    {
+      "epoch": 0.2692613376363768,
+      "grad_norm": 1.516343355178833,
+      "learning_rate": 7.692424703376939e-07,
+      "loss": 0.9456,
+      "step": 11180
+    },
+    {
+      "epoch": 0.26938175862816405,
+      "grad_norm": 1.4829399585723877,
+      "learning_rate": 7.691157083460095e-07,
+      "loss": 0.915,
+      "step": 11185
+    },
+    {
+      "epoch": 0.2695021796199514,
+      "grad_norm": 1.9153403043746948,
+      "learning_rate": 7.689889463543251e-07,
+      "loss": 0.9071,
+      "step": 11190
+    },
+    {
+      "epoch": 0.26962260061173865,
+      "grad_norm": 1.5353426933288574,
+      "learning_rate": 7.688621843626407e-07,
+      "loss": 0.9368,
+      "step": 11195
+    },
+    {
+      "epoch": 0.2697430216035259,
+      "grad_norm": 1.4356238842010498,
+      "learning_rate": 7.687354223709563e-07,
+      "loss": 0.9369,
+      "step": 11200
+    },
+    {
+      "epoch": 0.2698634425953132,
+      "grad_norm": 1.9013326168060303,
+      "learning_rate": 7.686086603792718e-07,
+      "loss": 0.9614,
+      "step": 11205
+    },
+    {
+      "epoch": 0.2699838635871005,
+      "grad_norm": 1.7938717603683472,
+      "learning_rate": 7.684818983875875e-07,
+      "loss": 0.9126,
+      "step": 11210
+    },
+    {
+      "epoch": 0.2701042845788878,
+      "grad_norm": 1.5766910314559937,
+      "learning_rate": 7.68355136395903e-07,
+      "loss": 0.9517,
+      "step": 11215
+    },
+    {
+      "epoch": 0.27022470557067507,
+      "grad_norm": 1.5627079010009766,
+      "learning_rate": 7.682283744042186e-07,
+      "loss": 0.9558,
+      "step": 11220
+    },
+    {
+      "epoch": 0.27034512656246235,
+      "grad_norm": 1.4922775030136108,
+      "learning_rate": 7.681016124125342e-07,
+      "loss": 0.9098,
+      "step": 11225
+    },
+    {
+      "epoch": 0.2704655475542497,
+      "grad_norm": 1.735756278038025,
+      "learning_rate": 7.679748504208497e-07,
+      "loss": 0.9373,
+      "step": 11230
+    },
+    {
+      "epoch": 0.27058596854603695,
+      "grad_norm": 1.7026416063308716,
+      "learning_rate": 7.678480884291654e-07,
+      "loss": 0.9146,
+      "step": 11235
+    },
+    {
+      "epoch": 0.2707063895378242,
+      "grad_norm": 1.3560659885406494,
+      "learning_rate": 7.67721326437481e-07,
+      "loss": 0.9555,
+      "step": 11240
+    },
+    {
+      "epoch": 0.2708268105296115,
+      "grad_norm": 1.432887315750122,
+      "learning_rate": 7.675945644457966e-07,
+      "loss": 0.9551,
+      "step": 11245
+    },
+    {
+      "epoch": 0.2709472315213988,
+      "grad_norm": 1.6378475427627563,
+      "learning_rate": 7.674678024541121e-07,
+      "loss": 0.8897,
+      "step": 11250
+    },
+    {
+      "epoch": 0.2710676525131861,
+      "grad_norm": 1.5451323986053467,
+      "learning_rate": 7.673410404624277e-07,
+      "loss": 0.9499,
+      "step": 11255
+    },
+    {
+      "epoch": 0.27118807350497337,
+      "grad_norm": 1.508756160736084,
+      "learning_rate": 7.672142784707433e-07,
+      "loss": 0.9058,
+      "step": 11260
+    },
+    {
+      "epoch": 0.2713084944967607,
+      "grad_norm": 1.5754610300064087,
+      "learning_rate": 7.670875164790589e-07,
+      "loss": 0.9651,
+      "step": 11265
+    },
+    {
+      "epoch": 0.271428915488548,
+      "grad_norm": 1.5599875450134277,
+      "learning_rate": 7.669607544873746e-07,
+      "loss": 0.9101,
+      "step": 11270
+    },
+    {
+      "epoch": 0.27154933648033525,
+      "grad_norm": 1.5504896640777588,
+      "learning_rate": 7.6683399249569e-07,
+      "loss": 0.9511,
+      "step": 11275
+    },
+    {
+      "epoch": 0.2716697574721225,
+      "grad_norm": 1.6675881147384644,
+      "learning_rate": 7.667072305040056e-07,
+      "loss": 0.9192,
+      "step": 11280
+    },
+    {
+      "epoch": 0.27179017846390985,
+      "grad_norm": 1.3830749988555908,
+      "learning_rate": 7.665804685123213e-07,
+      "loss": 0.922,
+      "step": 11285
+    },
+    {
+      "epoch": 0.2719105994556971,
+      "grad_norm": 1.6686891317367554,
+      "learning_rate": 7.664537065206368e-07,
+      "loss": 0.9337,
+      "step": 11290
+    },
+    {
+      "epoch": 0.2720310204474844,
+      "grad_norm": 1.8631843328475952,
+      "learning_rate": 7.663269445289524e-07,
+      "loss": 0.9199,
+      "step": 11295
+    },
+    {
+      "epoch": 0.27215144143927167,
+      "grad_norm": 1.4798780679702759,
+      "learning_rate": 7.66200182537268e-07,
+      "loss": 0.958,
+      "step": 11300
+    },
+    {
+      "epoch": 0.272271862431059,
+      "grad_norm": 1.3518550395965576,
+      "learning_rate": 7.660734205455836e-07,
+      "loss": 0.8929,
+      "step": 11305
+    },
+    {
+      "epoch": 0.27239228342284627,
+      "grad_norm": 1.6929333209991455,
+      "learning_rate": 7.659466585538992e-07,
+      "loss": 0.916,
+      "step": 11310
+    },
+    {
+      "epoch": 0.27251270441463354,
+      "grad_norm": 1.4561303853988647,
+      "learning_rate": 7.658198965622148e-07,
+      "loss": 0.9634,
+      "step": 11315
+    },
+    {
+      "epoch": 0.2726331254064209,
+      "grad_norm": 1.59073007106781,
+      "learning_rate": 7.656931345705303e-07,
+      "loss": 0.9048,
+      "step": 11320
+    },
+    {
+      "epoch": 0.27275354639820815,
+      "grad_norm": 1.526777744293213,
+      "learning_rate": 7.655663725788459e-07,
+      "loss": 0.9254,
+      "step": 11325
+    },
+    {
+      "epoch": 0.2728739673899954,
+      "grad_norm": 1.476694107055664,
+      "learning_rate": 7.654396105871616e-07,
+      "loss": 0.9151,
+      "step": 11330
+    },
+    {
+      "epoch": 0.2729943883817827,
+      "grad_norm": 1.5337563753128052,
+      "learning_rate": 7.653128485954771e-07,
+      "loss": 0.9102,
+      "step": 11335
+    },
+    {
+      "epoch": 0.27311480937357,
+      "grad_norm": 1.5907397270202637,
+      "learning_rate": 7.651860866037926e-07,
+      "loss": 0.9202,
+      "step": 11340
+    },
+    {
+      "epoch": 0.2732352303653573,
+      "grad_norm": 1.7442792654037476,
+      "learning_rate": 7.650593246121083e-07,
+      "loss": 0.9326,
+      "step": 11345
+    },
+    {
+      "epoch": 0.27335565135714457,
+      "grad_norm": 1.5198930501937866,
+      "learning_rate": 7.649325626204238e-07,
+      "loss": 0.9246,
+      "step": 11350
+    },
+    {
+      "epoch": 0.27347607234893184,
+      "grad_norm": 1.7051467895507812,
+      "learning_rate": 7.648058006287395e-07,
+      "loss": 0.9402,
+      "step": 11355
+    },
+    {
+      "epoch": 0.27359649334071917,
+      "grad_norm": 1.5837024450302124,
+      "learning_rate": 7.646790386370551e-07,
+      "loss": 0.8917,
+      "step": 11360
+    },
+    {
+      "epoch": 0.27371691433250644,
+      "grad_norm": 2.134157419204712,
+      "learning_rate": 7.645522766453705e-07,
+      "loss": 0.9517,
+      "step": 11365
+    },
+    {
+      "epoch": 0.2738373353242937,
+      "grad_norm": 1.6315033435821533,
+      "learning_rate": 7.644255146536862e-07,
+      "loss": 0.9348,
+      "step": 11370
+    },
+    {
+      "epoch": 0.27395775631608105,
+      "grad_norm": 1.4641317129135132,
+      "learning_rate": 7.642987526620018e-07,
+      "loss": 0.9088,
+      "step": 11375
+    },
+    {
+      "epoch": 0.2740781773078683,
+      "grad_norm": 1.6204805374145508,
+      "learning_rate": 7.641719906703174e-07,
+      "loss": 0.9057,
+      "step": 11380
+    },
+    {
+      "epoch": 0.2741985982996556,
+      "grad_norm": 1.5819618701934814,
+      "learning_rate": 7.64045228678633e-07,
+      "loss": 0.9122,
+      "step": 11385
+    },
+    {
+      "epoch": 0.27431901929144287,
+      "grad_norm": 1.5845000743865967,
+      "learning_rate": 7.639184666869486e-07,
+      "loss": 0.9032,
+      "step": 11390
+    },
+    {
+      "epoch": 0.2744394402832302,
+      "grad_norm": 1.9233126640319824,
+      "learning_rate": 7.637917046952641e-07,
+      "loss": 0.9111,
+      "step": 11395
+    },
+    {
+      "epoch": 0.27455986127501747,
+      "grad_norm": 1.5648930072784424,
+      "learning_rate": 7.636649427035797e-07,
+      "loss": 0.9242,
+      "step": 11400
+    },
+    {
+      "epoch": 0.27468028226680474,
+      "grad_norm": 1.3369144201278687,
+      "learning_rate": 7.635381807118954e-07,
+      "loss": 0.8929,
+      "step": 11405
+    },
+    {
+      "epoch": 0.274800703258592,
+      "grad_norm": 1.5446151494979858,
+      "learning_rate": 7.634114187202108e-07,
+      "loss": 0.9359,
+      "step": 11410
+    },
+    {
+      "epoch": 0.27492112425037935,
+      "grad_norm": 1.5810915231704712,
+      "learning_rate": 7.632846567285265e-07,
+      "loss": 0.9423,
+      "step": 11415
+    },
+    {
+      "epoch": 0.2750415452421666,
+      "grad_norm": 1.6881409883499146,
+      "learning_rate": 7.631578947368421e-07,
+      "loss": 0.9508,
+      "step": 11420
+    },
+    {
+      "epoch": 0.2751619662339539,
+      "grad_norm": 1.5262564420700073,
+      "learning_rate": 7.630311327451576e-07,
+      "loss": 0.9414,
+      "step": 11425
+    },
+    {
+      "epoch": 0.27528238722574117,
+      "grad_norm": 1.4334099292755127,
+      "learning_rate": 7.629043707534733e-07,
+      "loss": 0.9007,
+      "step": 11430
+    },
+    {
+      "epoch": 0.2754028082175285,
+      "grad_norm": 1.3691715002059937,
+      "learning_rate": 7.627776087617888e-07,
+      "loss": 0.8882,
+      "step": 11435
+    },
+    {
+      "epoch": 0.27552322920931577,
+      "grad_norm": 1.717207670211792,
+      "learning_rate": 7.626508467701044e-07,
+      "loss": 0.8711,
+      "step": 11440
+    },
+    {
+      "epoch": 0.27564365020110304,
+      "grad_norm": 1.5524109601974487,
+      "learning_rate": 7.6252408477842e-07,
+      "loss": 0.8835,
+      "step": 11445
+    },
+    {
+      "epoch": 0.27576407119289037,
+      "grad_norm": 1.6486324071884155,
+      "learning_rate": 7.623973227867357e-07,
+      "loss": 0.8512,
+      "step": 11450
+    },
+    {
+      "epoch": 0.27588449218467764,
+      "grad_norm": 1.8091490268707275,
+      "learning_rate": 7.622705607950511e-07,
+      "loss": 0.9348,
+      "step": 11455
+    },
+    {
+      "epoch": 0.2760049131764649,
+      "grad_norm": 1.6140121221542358,
+      "learning_rate": 7.621437988033667e-07,
+      "loss": 0.9328,
+      "step": 11460
+    },
+    {
+      "epoch": 0.2761253341682522,
+      "grad_norm": 1.4895670413970947,
+      "learning_rate": 7.620170368116824e-07,
+      "loss": 0.8896,
+      "step": 11465
+    },
+    {
+      "epoch": 0.2762457551600395,
+      "grad_norm": 1.525490403175354,
+      "learning_rate": 7.618902748199979e-07,
+      "loss": 0.969,
+      "step": 11470
+    },
+    {
+      "epoch": 0.2763661761518268,
+      "grad_norm": 1.5126816034317017,
+      "learning_rate": 7.617635128283136e-07,
+      "loss": 0.9383,
+      "step": 11475
+    },
+    {
+      "epoch": 0.27648659714361407,
+      "grad_norm": 1.466437578201294,
+      "learning_rate": 7.616367508366291e-07,
+      "loss": 0.9023,
+      "step": 11480
+    },
+    {
+      "epoch": 0.27660701813540134,
+      "grad_norm": 1.477810263633728,
+      "learning_rate": 7.615099888449446e-07,
+      "loss": 0.9952,
+      "step": 11485
+    },
+    {
+      "epoch": 0.27672743912718867,
+      "grad_norm": 1.444549798965454,
+      "learning_rate": 7.613832268532603e-07,
+      "loss": 0.969,
+      "step": 11490
+    },
+    {
+      "epoch": 0.27684786011897594,
+      "grad_norm": 1.5766974687576294,
+      "learning_rate": 7.612564648615759e-07,
+      "loss": 0.8708,
+      "step": 11495
+    },
+    {
+      "epoch": 0.2769682811107632,
+      "grad_norm": 1.6157164573669434,
+      "learning_rate": 7.611297028698915e-07,
+      "loss": 0.9892,
+      "step": 11500
+    },
+    {
+      "epoch": 0.27708870210255054,
+      "grad_norm": 1.6714296340942383,
+      "learning_rate": 7.61002940878207e-07,
+      "loss": 0.8914,
+      "step": 11505
+    },
+    {
+      "epoch": 0.2772091230943378,
+      "grad_norm": 1.6638739109039307,
+      "learning_rate": 7.608761788865227e-07,
+      "loss": 0.9758,
+      "step": 11510
+    },
+    {
+      "epoch": 0.2773295440861251,
+      "grad_norm": 1.5034291744232178,
+      "learning_rate": 7.607494168948382e-07,
+      "loss": 0.9094,
+      "step": 11515
+    },
+    {
+      "epoch": 0.27744996507791236,
+      "grad_norm": 1.6141936779022217,
+      "learning_rate": 7.606226549031538e-07,
+      "loss": 0.973,
+      "step": 11520
+    },
+    {
+      "epoch": 0.2775703860696997,
+      "grad_norm": 1.4471684694290161,
+      "learning_rate": 7.604958929114694e-07,
+      "loss": 0.934,
+      "step": 11525
+    },
+    {
+      "epoch": 0.27769080706148697,
+      "grad_norm": 1.5534322261810303,
+      "learning_rate": 7.603691309197849e-07,
+      "loss": 0.935,
+      "step": 11530
+    },
+    {
+      "epoch": 0.27781122805327424,
+      "grad_norm": 1.5422011613845825,
+      "learning_rate": 7.602423689281006e-07,
+      "loss": 0.8949,
+      "step": 11535
+    },
+    {
+      "epoch": 0.2779316490450615,
+      "grad_norm": 1.6572320461273193,
+      "learning_rate": 7.601156069364162e-07,
+      "loss": 0.9403,
+      "step": 11540
+    },
+    {
+      "epoch": 0.27805207003684884,
+      "grad_norm": 1.4555169343948364,
+      "learning_rate": 7.599888449447317e-07,
+      "loss": 0.9631,
+      "step": 11545
+    },
+    {
+      "epoch": 0.2781724910286361,
+      "grad_norm": 1.4223170280456543,
+      "learning_rate": 7.598620829530473e-07,
+      "loss": 0.9567,
+      "step": 11550
+    },
+    {
+      "epoch": 0.2782929120204234,
+      "grad_norm": 1.6159312725067139,
+      "learning_rate": 7.597353209613629e-07,
+      "loss": 0.9088,
+      "step": 11555
+    },
+    {
+      "epoch": 0.27841333301221066,
+      "grad_norm": 1.6263511180877686,
+      "learning_rate": 7.596085589696785e-07,
+      "loss": 0.9317,
+      "step": 11560
+    },
+    {
+      "epoch": 0.278533754003998,
+      "grad_norm": 1.8005136251449585,
+      "learning_rate": 7.594817969779941e-07,
+      "loss": 0.9596,
+      "step": 11565
+    },
+    {
+      "epoch": 0.27865417499578526,
+      "grad_norm": 1.4669032096862793,
+      "learning_rate": 7.593550349863098e-07,
+      "loss": 0.9032,
+      "step": 11570
+    },
+    {
+      "epoch": 0.27877459598757254,
+      "grad_norm": 1.738063097000122,
+      "learning_rate": 7.592282729946252e-07,
+      "loss": 0.8592,
+      "step": 11575
+    },
+    {
+      "epoch": 0.27889501697935987,
+      "grad_norm": 1.4538829326629639,
+      "learning_rate": 7.591015110029408e-07,
+      "loss": 0.8801,
+      "step": 11580
+    },
+    {
+      "epoch": 0.27901543797114714,
+      "grad_norm": 1.568163514137268,
+      "learning_rate": 7.589747490112565e-07,
+      "loss": 0.9436,
+      "step": 11585
+    },
+    {
+      "epoch": 0.2791358589629344,
+      "grad_norm": 1.8308765888214111,
+      "learning_rate": 7.58847987019572e-07,
+      "loss": 0.8998,
+      "step": 11590
+    },
+    {
+      "epoch": 0.2792562799547217,
+      "grad_norm": 1.699054479598999,
+      "learning_rate": 7.587212250278876e-07,
+      "loss": 0.9425,
+      "step": 11595
+    },
+    {
+      "epoch": 0.279376700946509,
+      "grad_norm": 1.5109950304031372,
+      "learning_rate": 7.585944630362032e-07,
+      "loss": 0.9135,
+      "step": 11600
+    },
+    {
+      "epoch": 0.2794971219382963,
+      "grad_norm": 1.4689977169036865,
+      "learning_rate": 7.584677010445187e-07,
+      "loss": 0.9104,
+      "step": 11605
+    },
+    {
+      "epoch": 0.27961754293008356,
+      "grad_norm": 1.4785994291305542,
+      "learning_rate": 7.583409390528344e-07,
+      "loss": 0.9532,
+      "step": 11610
+    },
+    {
+      "epoch": 0.27973796392187084,
+      "grad_norm": 1.7106760740280151,
+      "learning_rate": 7.5821417706115e-07,
+      "loss": 0.9276,
+      "step": 11615
+    },
+    {
+      "epoch": 0.27985838491365816,
+      "grad_norm": 1.4372215270996094,
+      "learning_rate": 7.580874150694655e-07,
+      "loss": 0.8959,
+      "step": 11620
+    },
+    {
+      "epoch": 0.27997880590544544,
+      "grad_norm": 1.5852820873260498,
+      "learning_rate": 7.579606530777811e-07,
+      "loss": 0.9093,
+      "step": 11625
+    },
+    {
+      "epoch": 0.2800992268972327,
+      "grad_norm": 1.6567399501800537,
+      "learning_rate": 7.578338910860968e-07,
+      "loss": 0.8969,
+      "step": 11630
+    },
+    {
+      "epoch": 0.28021964788902004,
+      "grad_norm": 1.4735957384109497,
+      "learning_rate": 7.577071290944123e-07,
+      "loss": 0.9294,
+      "step": 11635
+    },
+    {
+      "epoch": 0.2803400688808073,
+      "grad_norm": 1.548314094543457,
+      "learning_rate": 7.575803671027278e-07,
+      "loss": 0.9493,
+      "step": 11640
+    },
+    {
+      "epoch": 0.2804604898725946,
+      "grad_norm": 1.645434021949768,
+      "learning_rate": 7.574536051110435e-07,
+      "loss": 0.9196,
+      "step": 11645
+    },
+    {
+      "epoch": 0.28058091086438186,
+      "grad_norm": 1.3293720483779907,
+      "learning_rate": 7.57326843119359e-07,
+      "loss": 0.9154,
+      "step": 11650
+    },
+    {
+      "epoch": 0.2807013318561692,
+      "grad_norm": 1.407025694847107,
+      "learning_rate": 7.572000811276747e-07,
+      "loss": 0.9089,
+      "step": 11655
+    },
+    {
+      "epoch": 0.28082175284795646,
+      "grad_norm": 1.4637197256088257,
+      "learning_rate": 7.570733191359903e-07,
+      "loss": 0.9169,
+      "step": 11660
+    },
+    {
+      "epoch": 0.28094217383974374,
+      "grad_norm": 1.6624650955200195,
+      "learning_rate": 7.569465571443057e-07,
+      "loss": 0.9073,
+      "step": 11665
+    },
+    {
+      "epoch": 0.281062594831531,
+      "grad_norm": 1.5180774927139282,
+      "learning_rate": 7.568197951526214e-07,
+      "loss": 0.9752,
+      "step": 11670
+    },
+    {
+      "epoch": 0.28118301582331834,
+      "grad_norm": 1.5979740619659424,
+      "learning_rate": 7.56693033160937e-07,
+      "loss": 0.9037,
+      "step": 11675
+    },
+    {
+      "epoch": 0.2813034368151056,
+      "grad_norm": 1.5818020105361938,
+      "learning_rate": 7.565662711692526e-07,
+      "loss": 0.9473,
+      "step": 11680
+    },
+    {
+      "epoch": 0.2814238578068929,
+      "grad_norm": 1.7185585498809814,
+      "learning_rate": 7.564395091775682e-07,
+      "loss": 0.9279,
+      "step": 11685
+    },
+    {
+      "epoch": 0.28154427879868016,
+      "grad_norm": 1.6393964290618896,
+      "learning_rate": 7.563127471858837e-07,
+      "loss": 0.8911,
+      "step": 11690
+    },
+    {
+      "epoch": 0.2816646997904675,
+      "grad_norm": 1.623339056968689,
+      "learning_rate": 7.561859851941993e-07,
+      "loss": 0.9645,
+      "step": 11695
+    },
+    {
+      "epoch": 0.28178512078225476,
+      "grad_norm": 1.5230019092559814,
+      "learning_rate": 7.560592232025149e-07,
+      "loss": 0.9363,
+      "step": 11700
+    },
+    {
+      "epoch": 0.28190554177404203,
+      "grad_norm": 1.6126532554626465,
+      "learning_rate": 7.559324612108306e-07,
+      "loss": 0.9706,
+      "step": 11705
+    },
+    {
+      "epoch": 0.28202596276582936,
+      "grad_norm": 1.8683366775512695,
+      "learning_rate": 7.55805699219146e-07,
+      "loss": 0.9447,
+      "step": 11710
+    },
+    {
+      "epoch": 0.28214638375761664,
+      "grad_norm": 1.5551618337631226,
+      "learning_rate": 7.556789372274617e-07,
+      "loss": 0.9209,
+      "step": 11715
+    },
+    {
+      "epoch": 0.2822668047494039,
+      "grad_norm": 1.4535892009735107,
+      "learning_rate": 7.555521752357773e-07,
+      "loss": 0.9299,
+      "step": 11720
+    },
+    {
+      "epoch": 0.2823872257411912,
+      "grad_norm": 2.7465555667877197,
+      "learning_rate": 7.554254132440928e-07,
+      "loss": 0.9016,
+      "step": 11725
+    },
+    {
+      "epoch": 0.2825076467329785,
+      "grad_norm": 1.6401972770690918,
+      "learning_rate": 7.552986512524085e-07,
+      "loss": 0.8983,
+      "step": 11730
+    },
+    {
+      "epoch": 0.2826280677247658,
+      "grad_norm": 1.655357003211975,
+      "learning_rate": 7.55171889260724e-07,
+      "loss": 0.9406,
+      "step": 11735
+    },
+    {
+      "epoch": 0.28274848871655306,
+      "grad_norm": 1.4839775562286377,
+      "learning_rate": 7.550451272690396e-07,
+      "loss": 0.9209,
+      "step": 11740
+    },
+    {
+      "epoch": 0.28286890970834033,
+      "grad_norm": 1.6104551553726196,
+      "learning_rate": 7.549183652773552e-07,
+      "loss": 0.9115,
+      "step": 11745
+    },
+    {
+      "epoch": 0.28298933070012766,
+      "grad_norm": 1.6464667320251465,
+      "learning_rate": 7.547916032856709e-07,
+      "loss": 0.9345,
+      "step": 11750
+    },
+    {
+      "epoch": 0.28310975169191493,
+      "grad_norm": 1.4681965112686157,
+      "learning_rate": 7.546648412939863e-07,
+      "loss": 0.9639,
+      "step": 11755
+    },
+    {
+      "epoch": 0.2832301726837022,
+      "grad_norm": 1.4175541400909424,
+      "learning_rate": 7.545380793023019e-07,
+      "loss": 0.938,
+      "step": 11760
+    },
+    {
+      "epoch": 0.28335059367548954,
+      "grad_norm": 1.8389374017715454,
+      "learning_rate": 7.544113173106176e-07,
+      "loss": 0.9199,
+      "step": 11765
+    },
+    {
+      "epoch": 0.2834710146672768,
+      "grad_norm": 1.4778343439102173,
+      "learning_rate": 7.542845553189331e-07,
+      "loss": 0.9656,
+      "step": 11770
+    },
+    {
+      "epoch": 0.2835914356590641,
+      "grad_norm": 1.595827341079712,
+      "learning_rate": 7.541577933272488e-07,
+      "loss": 0.9093,
+      "step": 11775
+    },
+    {
+      "epoch": 0.28371185665085136,
+      "grad_norm": 1.5474449396133423,
+      "learning_rate": 7.540310313355643e-07,
+      "loss": 0.9245,
+      "step": 11780
+    },
+    {
+      "epoch": 0.2838322776426387,
+      "grad_norm": 1.8880990743637085,
+      "learning_rate": 7.539042693438798e-07,
+      "loss": 0.9379,
+      "step": 11785
+    },
+    {
+      "epoch": 0.28395269863442596,
+      "grad_norm": 1.5933219194412231,
+      "learning_rate": 7.537775073521955e-07,
+      "loss": 0.9089,
+      "step": 11790
+    },
+    {
+      "epoch": 0.28407311962621323,
+      "grad_norm": 1.6476610898971558,
+      "learning_rate": 7.536507453605111e-07,
+      "loss": 0.947,
+      "step": 11795
+    },
+    {
+      "epoch": 0.2841935406180005,
+      "grad_norm": 1.3580931425094604,
+      "learning_rate": 7.535239833688267e-07,
+      "loss": 0.935,
+      "step": 11800
+    },
+    {
+      "epoch": 0.28431396160978784,
+      "grad_norm": 1.3850847482681274,
+      "learning_rate": 7.533972213771422e-07,
+      "loss": 0.9051,
+      "step": 11805
+    },
+    {
+      "epoch": 0.2844343826015751,
+      "grad_norm": 1.4011439085006714,
+      "learning_rate": 7.532704593854578e-07,
+      "loss": 0.8799,
+      "step": 11810
+    },
+    {
+      "epoch": 0.2845548035933624,
+      "grad_norm": 1.6950008869171143,
+      "learning_rate": 7.531436973937734e-07,
+      "loss": 0.9325,
+      "step": 11815
+    },
+    {
+      "epoch": 0.2846752245851497,
+      "grad_norm": 1.5124635696411133,
+      "learning_rate": 7.53016935402089e-07,
+      "loss": 0.9198,
+      "step": 11820
+    },
+    {
+      "epoch": 0.284795645576937,
+      "grad_norm": 1.481153964996338,
+      "learning_rate": 7.528901734104046e-07,
+      "loss": 0.9484,
+      "step": 11825
+    },
+    {
+      "epoch": 0.28491606656872426,
+      "grad_norm": 1.6342970132827759,
+      "learning_rate": 7.527634114187201e-07,
+      "loss": 0.9259,
+      "step": 11830
+    },
+    {
+      "epoch": 0.28503648756051153,
+      "grad_norm": 1.6247309446334839,
+      "learning_rate": 7.526366494270358e-07,
+      "loss": 0.9414,
+      "step": 11835
+    },
+    {
+      "epoch": 0.28515690855229886,
+      "grad_norm": 1.7598034143447876,
+      "learning_rate": 7.525098874353514e-07,
+      "loss": 0.9578,
+      "step": 11840
+    },
+    {
+      "epoch": 0.28527732954408613,
+      "grad_norm": 1.5112534761428833,
+      "learning_rate": 7.523831254436669e-07,
+      "loss": 0.9242,
+      "step": 11845
+    },
+    {
+      "epoch": 0.2853977505358734,
+      "grad_norm": 1.48063063621521,
+      "learning_rate": 7.522563634519825e-07,
+      "loss": 0.9513,
+      "step": 11850
+    },
+    {
+      "epoch": 0.2855181715276607,
+      "grad_norm": 1.4742642641067505,
+      "learning_rate": 7.521296014602981e-07,
+      "loss": 0.9896,
+      "step": 11855
+    },
+    {
+      "epoch": 0.285638592519448,
+      "grad_norm": 1.7979505062103271,
+      "learning_rate": 7.520028394686137e-07,
+      "loss": 0.9469,
+      "step": 11860
+    },
+    {
+      "epoch": 0.2857590135112353,
+      "grad_norm": 1.6058517694473267,
+      "learning_rate": 7.518760774769293e-07,
+      "loss": 0.9196,
+      "step": 11865
+    },
+    {
+      "epoch": 0.28587943450302256,
+      "grad_norm": 1.6035789251327515,
+      "learning_rate": 7.51749315485245e-07,
+      "loss": 0.9473,
+      "step": 11870
+    },
+    {
+      "epoch": 0.28599985549480983,
+      "grad_norm": 1.6348637342453003,
+      "learning_rate": 7.516225534935604e-07,
+      "loss": 0.9306,
+      "step": 11875
+    },
+    {
+      "epoch": 0.28612027648659716,
+      "grad_norm": 1.7751243114471436,
+      "learning_rate": 7.51495791501876e-07,
+      "loss": 0.967,
+      "step": 11880
+    },
+    {
+      "epoch": 0.28624069747838443,
+      "grad_norm": 1.681087613105774,
+      "learning_rate": 7.513690295101917e-07,
+      "loss": 0.9591,
+      "step": 11885
+    },
+    {
+      "epoch": 0.2863611184701717,
+      "grad_norm": 1.6238495111465454,
+      "learning_rate": 7.512422675185072e-07,
+      "loss": 0.9472,
+      "step": 11890
+    },
+    {
+      "epoch": 0.28648153946195903,
+      "grad_norm": 1.5437794923782349,
+      "learning_rate": 7.511155055268227e-07,
+      "loss": 0.928,
+      "step": 11895
+    },
+    {
+      "epoch": 0.2866019604537463,
+      "grad_norm": 1.565635323524475,
+      "learning_rate": 7.509887435351384e-07,
+      "loss": 0.9666,
+      "step": 11900
+    },
+    {
+      "epoch": 0.2867223814455336,
+      "grad_norm": 1.6704597473144531,
+      "learning_rate": 7.508619815434539e-07,
+      "loss": 0.9399,
+      "step": 11905
+    },
+    {
+      "epoch": 0.28684280243732085,
+      "grad_norm": 1.522324800491333,
+      "learning_rate": 7.507352195517696e-07,
+      "loss": 0.9282,
+      "step": 11910
+    },
+    {
+      "epoch": 0.2869632234291082,
+      "grad_norm": 1.4775621891021729,
+      "learning_rate": 7.506084575600852e-07,
+      "loss": 0.8827,
+      "step": 11915
+    },
+    {
+      "epoch": 0.28708364442089546,
+      "grad_norm": 1.5191771984100342,
+      "learning_rate": 7.504816955684007e-07,
+      "loss": 0.887,
+      "step": 11920
+    },
+    {
+      "epoch": 0.28720406541268273,
+      "grad_norm": 1.6595885753631592,
+      "learning_rate": 7.503549335767163e-07,
+      "loss": 0.9344,
+      "step": 11925
+    },
+    {
+      "epoch": 0.28732448640447,
+      "grad_norm": 1.7235771417617798,
+      "learning_rate": 7.502281715850319e-07,
+      "loss": 0.967,
+      "step": 11930
+    },
+    {
+      "epoch": 0.28744490739625733,
+      "grad_norm": 1.4997406005859375,
+      "learning_rate": 7.501014095933475e-07,
+      "loss": 0.9614,
+      "step": 11935
+    },
+    {
+      "epoch": 0.2875653283880446,
+      "grad_norm": 1.6935820579528809,
+      "learning_rate": 7.49974647601663e-07,
+      "loss": 0.9519,
+      "step": 11940
+    },
+    {
+      "epoch": 0.2876857493798319,
+      "grad_norm": 1.5619289875030518,
+      "learning_rate": 7.498478856099787e-07,
+      "loss": 0.8883,
+      "step": 11945
+    },
+    {
+      "epoch": 0.2878061703716192,
+      "grad_norm": 1.4663188457489014,
+      "learning_rate": 7.497211236182942e-07,
+      "loss": 0.9096,
+      "step": 11950
+    },
+    {
+      "epoch": 0.2879265913634065,
+      "grad_norm": 1.4758613109588623,
+      "learning_rate": 7.495943616266098e-07,
+      "loss": 0.9472,
+      "step": 11955
+    },
+    {
+      "epoch": 0.28804701235519375,
+      "grad_norm": 1.6588231325149536,
+      "learning_rate": 7.494675996349255e-07,
+      "loss": 0.8967,
+      "step": 11960
+    },
+    {
+      "epoch": 0.28816743334698103,
+      "grad_norm": 1.676204800605774,
+      "learning_rate": 7.493408376432409e-07,
+      "loss": 0.9067,
+      "step": 11965
+    },
+    {
+      "epoch": 0.28828785433876836,
+      "grad_norm": 1.6175498962402344,
+      "learning_rate": 7.492140756515566e-07,
+      "loss": 0.9315,
+      "step": 11970
+    },
+    {
+      "epoch": 0.28840827533055563,
+      "grad_norm": 1.8844404220581055,
+      "learning_rate": 7.490873136598722e-07,
+      "loss": 0.9453,
+      "step": 11975
+    },
+    {
+      "epoch": 0.2885286963223429,
+      "grad_norm": 1.6034772396087646,
+      "learning_rate": 7.489605516681878e-07,
+      "loss": 0.9564,
+      "step": 11980
+    },
+    {
+      "epoch": 0.2886491173141302,
+      "grad_norm": 1.7410510778427124,
+      "learning_rate": 7.488337896765033e-07,
+      "loss": 0.952,
+      "step": 11985
+    },
+    {
+      "epoch": 0.2887695383059175,
+      "grad_norm": 1.5769463777542114,
+      "learning_rate": 7.487070276848189e-07,
+      "loss": 0.9601,
+      "step": 11990
+    },
+    {
+      "epoch": 0.2888899592977048,
+      "grad_norm": 1.3784805536270142,
+      "learning_rate": 7.485802656931345e-07,
+      "loss": 0.9725,
+      "step": 11995
+    },
+    {
+      "epoch": 0.28901038028949205,
+      "grad_norm": 1.6138496398925781,
+      "learning_rate": 7.484535037014501e-07,
+      "loss": 0.8706,
+      "step": 12000
+    },
+    {
+      "epoch": 0.2891308012812793,
+      "grad_norm": 1.6166455745697021,
+      "learning_rate": 7.483267417097658e-07,
+      "loss": 0.9151,
+      "step": 12005
+    },
+    {
+      "epoch": 0.28925122227306665,
+      "grad_norm": 1.4987436532974243,
+      "learning_rate": 7.481999797180812e-07,
+      "loss": 0.95,
+      "step": 12010
+    },
+    {
+      "epoch": 0.28937164326485393,
+      "grad_norm": 1.3807342052459717,
+      "learning_rate": 7.480732177263968e-07,
+      "loss": 0.8848,
+      "step": 12015
+    },
+    {
+      "epoch": 0.2894920642566412,
+      "grad_norm": 1.5243955850601196,
+      "learning_rate": 7.479464557347125e-07,
+      "loss": 0.9122,
+      "step": 12020
+    },
+    {
+      "epoch": 0.28961248524842853,
+      "grad_norm": 1.4616668224334717,
+      "learning_rate": 7.47819693743028e-07,
+      "loss": 0.971,
+      "step": 12025
+    },
+    {
+      "epoch": 0.2897329062402158,
+      "grad_norm": 1.71182119846344,
+      "learning_rate": 7.476929317513437e-07,
+      "loss": 0.9404,
+      "step": 12030
+    },
+    {
+      "epoch": 0.2898533272320031,
+      "grad_norm": 1.7082267999649048,
+      "learning_rate": 7.475661697596592e-07,
+      "loss": 0.8648,
+      "step": 12035
+    },
+    {
+      "epoch": 0.28997374822379035,
+      "grad_norm": 1.566318154335022,
+      "learning_rate": 7.474394077679748e-07,
+      "loss": 0.9647,
+      "step": 12040
+    },
+    {
+      "epoch": 0.2900941692155777,
+      "grad_norm": 1.5414106845855713,
+      "learning_rate": 7.473126457762904e-07,
+      "loss": 0.927,
+      "step": 12045
+    },
+    {
+      "epoch": 0.29021459020736495,
+      "grad_norm": 1.4850926399230957,
+      "learning_rate": 7.47185883784606e-07,
+      "loss": 0.9644,
+      "step": 12050
+    },
+    {
+      "epoch": 0.2903350111991522,
+      "grad_norm": 1.467564582824707,
+      "learning_rate": 7.470591217929216e-07,
+      "loss": 0.9193,
+      "step": 12055
+    },
+    {
+      "epoch": 0.2904554321909395,
+      "grad_norm": 1.3166462182998657,
+      "learning_rate": 7.469323598012371e-07,
+      "loss": 0.9112,
+      "step": 12060
+    },
+    {
+      "epoch": 0.29057585318272683,
+      "grad_norm": 1.6410185098648071,
+      "learning_rate": 7.468055978095528e-07,
+      "loss": 0.9226,
+      "step": 12065
+    },
+    {
+      "epoch": 0.2906962741745141,
+      "grad_norm": 1.5618233680725098,
+      "learning_rate": 7.466788358178684e-07,
+      "loss": 0.9354,
+      "step": 12070
+    },
+    {
+      "epoch": 0.2908166951663014,
+      "grad_norm": 1.4823178052902222,
+      "learning_rate": 7.465520738261839e-07,
+      "loss": 0.9549,
+      "step": 12075
+    },
+    {
+      "epoch": 0.2909371161580887,
+      "grad_norm": 1.6021852493286133,
+      "learning_rate": 7.464253118344995e-07,
+      "loss": 0.9438,
+      "step": 12080
+    },
+    {
+      "epoch": 0.291057537149876,
+      "grad_norm": 1.3809022903442383,
+      "learning_rate": 7.462985498428151e-07,
+      "loss": 0.9323,
+      "step": 12085
+    },
+    {
+      "epoch": 0.29117795814166325,
+      "grad_norm": 1.5327422618865967,
+      "learning_rate": 7.461717878511307e-07,
+      "loss": 0.9262,
+      "step": 12090
+    },
+    {
+      "epoch": 0.2912983791334505,
+      "grad_norm": 1.8754101991653442,
+      "learning_rate": 7.460450258594463e-07,
+      "loss": 0.9813,
+      "step": 12095
+    },
+    {
+      "epoch": 0.29141880012523785,
+      "grad_norm": 1.5512574911117554,
+      "learning_rate": 7.45918263867762e-07,
+      "loss": 0.9313,
+      "step": 12100
+    },
+    {
+      "epoch": 0.2915392211170251,
+      "grad_norm": 1.3583546876907349,
+      "learning_rate": 7.457915018760774e-07,
+      "loss": 0.9547,
+      "step": 12105
+    },
+    {
+      "epoch": 0.2916596421088124,
+      "grad_norm": 1.700033187866211,
+      "learning_rate": 7.45664739884393e-07,
+      "loss": 0.9332,
+      "step": 12110
+    },
+    {
+      "epoch": 0.2917800631005997,
+      "grad_norm": 1.4785394668579102,
+      "learning_rate": 7.455379778927087e-07,
+      "loss": 0.9547,
+      "step": 12115
+    },
+    {
+      "epoch": 0.291900484092387,
+      "grad_norm": 1.5970126390457153,
+      "learning_rate": 7.454112159010242e-07,
+      "loss": 0.9075,
+      "step": 12120
+    },
+    {
+      "epoch": 0.2920209050841743,
+      "grad_norm": 1.6902731657028198,
+      "learning_rate": 7.452844539093398e-07,
+      "loss": 0.8942,
+      "step": 12125
+    },
+    {
+      "epoch": 0.29214132607596155,
+      "grad_norm": 1.7784210443496704,
+      "learning_rate": 7.451576919176554e-07,
+      "loss": 0.9149,
+      "step": 12130
+    },
+    {
+      "epoch": 0.2922617470677488,
+      "grad_norm": 1.578518033027649,
+      "learning_rate": 7.450309299259709e-07,
+      "loss": 0.8796,
+      "step": 12135
+    },
+    {
+      "epoch": 0.29238216805953615,
+      "grad_norm": 1.5227102041244507,
+      "learning_rate": 7.449041679342866e-07,
+      "loss": 0.9589,
+      "step": 12140
+    },
+    {
+      "epoch": 0.2925025890513234,
+      "grad_norm": 1.3288702964782715,
+      "learning_rate": 7.447774059426022e-07,
+      "loss": 0.9111,
+      "step": 12145
+    },
+    {
+      "epoch": 0.2926230100431107,
+      "grad_norm": 1.8506990671157837,
+      "learning_rate": 7.446506439509177e-07,
+      "loss": 0.8961,
+      "step": 12150
+    },
+    {
+      "epoch": 0.292743431034898,
+      "grad_norm": 1.7175493240356445,
+      "learning_rate": 7.445238819592333e-07,
+      "loss": 0.8911,
+      "step": 12155
+    },
+    {
+      "epoch": 0.2928638520266853,
+      "grad_norm": 1.4701457023620605,
+      "learning_rate": 7.44397119967549e-07,
+      "loss": 0.9094,
+      "step": 12160
+    },
+    {
+      "epoch": 0.2929842730184726,
+      "grad_norm": 1.4599250555038452,
+      "learning_rate": 7.442703579758645e-07,
+      "loss": 0.9085,
+      "step": 12165
+    },
+    {
+      "epoch": 0.29310469401025985,
+      "grad_norm": 1.5617862939834595,
+      "learning_rate": 7.4414359598418e-07,
+      "loss": 0.9693,
+      "step": 12170
+    },
+    {
+      "epoch": 0.2932251150020472,
+      "grad_norm": 1.5455378293991089,
+      "learning_rate": 7.440168339924957e-07,
+      "loss": 0.9117,
+      "step": 12175
+    },
+    {
+      "epoch": 0.29334553599383445,
+      "grad_norm": 1.577228307723999,
+      "learning_rate": 7.438900720008112e-07,
+      "loss": 0.9267,
+      "step": 12180
+    },
+    {
+      "epoch": 0.2934659569856217,
+      "grad_norm": 1.498317003250122,
+      "learning_rate": 7.437633100091269e-07,
+      "loss": 0.9736,
+      "step": 12185
+    },
+    {
+      "epoch": 0.293586377977409,
+      "grad_norm": 1.5183560848236084,
+      "learning_rate": 7.436365480174425e-07,
+      "loss": 0.9262,
+      "step": 12190
+    },
+    {
+      "epoch": 0.2937067989691963,
+      "grad_norm": 1.533486247062683,
+      "learning_rate": 7.435097860257579e-07,
+      "loss": 0.9778,
+      "step": 12195
+    },
+    {
+      "epoch": 0.2938272199609836,
+      "grad_norm": 1.602159023284912,
+      "learning_rate": 7.433830240340736e-07,
+      "loss": 0.9087,
+      "step": 12200
+    },
+    {
+      "epoch": 0.29394764095277087,
+      "grad_norm": 1.51285719871521,
+      "learning_rate": 7.432562620423892e-07,
+      "loss": 0.9398,
+      "step": 12205
+    },
+    {
+      "epoch": 0.2940680619445582,
+      "grad_norm": 1.9584453105926514,
+      "learning_rate": 7.431295000507048e-07,
+      "loss": 0.9404,
+      "step": 12210
+    },
+    {
+      "epoch": 0.2941884829363455,
+      "grad_norm": 1.566316843032837,
+      "learning_rate": 7.430027380590204e-07,
+      "loss": 0.8733,
+      "step": 12215
+    },
+    {
+      "epoch": 0.29430890392813275,
+      "grad_norm": 1.4096190929412842,
+      "learning_rate": 7.42875976067336e-07,
+      "loss": 0.8925,
+      "step": 12220
+    },
+    {
+      "epoch": 0.29442932491992,
+      "grad_norm": 1.6779398918151855,
+      "learning_rate": 7.427492140756515e-07,
+      "loss": 0.9056,
+      "step": 12225
+    },
+    {
+      "epoch": 0.29454974591170735,
+      "grad_norm": 1.5147545337677002,
+      "learning_rate": 7.426224520839671e-07,
+      "loss": 0.8572,
+      "step": 12230
+    },
+    {
+      "epoch": 0.2946701669034946,
+      "grad_norm": 1.5745984315872192,
+      "learning_rate": 7.424956900922828e-07,
+      "loss": 0.9432,
+      "step": 12235
+    },
+    {
+      "epoch": 0.2947905878952819,
+      "grad_norm": 1.381061315536499,
+      "learning_rate": 7.423689281005982e-07,
+      "loss": 0.9103,
+      "step": 12240
+    },
+    {
+      "epoch": 0.29491100888706917,
+      "grad_norm": 1.6397510766983032,
+      "learning_rate": 7.422421661089139e-07,
+      "loss": 0.9932,
+      "step": 12245
+    },
+    {
+      "epoch": 0.2950314298788565,
+      "grad_norm": 1.387656807899475,
+      "learning_rate": 7.421154041172295e-07,
+      "loss": 0.9609,
+      "step": 12250
+    },
+    {
+      "epoch": 0.2951518508706438,
+      "grad_norm": 1.619707465171814,
+      "learning_rate": 7.41988642125545e-07,
+      "loss": 0.9056,
+      "step": 12255
+    },
+    {
+      "epoch": 0.29527227186243105,
+      "grad_norm": 1.9158905744552612,
+      "learning_rate": 7.418618801338607e-07,
+      "loss": 0.987,
+      "step": 12260
+    },
+    {
+      "epoch": 0.2953926928542184,
+      "grad_norm": 1.4746280908584595,
+      "learning_rate": 7.417351181421762e-07,
+      "loss": 0.9398,
+      "step": 12265
+    },
+    {
+      "epoch": 0.29551311384600565,
+      "grad_norm": 1.4023077487945557,
+      "learning_rate": 7.416083561504918e-07,
+      "loss": 0.9538,
+      "step": 12270
+    },
+    {
+      "epoch": 0.2956335348377929,
+      "grad_norm": 1.656822681427002,
+      "learning_rate": 7.414815941588074e-07,
+      "loss": 0.9088,
+      "step": 12275
+    },
+    {
+      "epoch": 0.2957539558295802,
+      "grad_norm": 1.841209888458252,
+      "learning_rate": 7.41354832167123e-07,
+      "loss": 0.9502,
+      "step": 12280
+    },
+    {
+      "epoch": 0.2958743768213675,
+      "grad_norm": 1.6077429056167603,
+      "learning_rate": 7.412280701754385e-07,
+      "loss": 0.8769,
+      "step": 12285
+    },
+    {
+      "epoch": 0.2959947978131548,
+      "grad_norm": 1.752013087272644,
+      "learning_rate": 7.411013081837541e-07,
+      "loss": 0.8655,
+      "step": 12290
+    },
+    {
+      "epoch": 0.29611521880494207,
+      "grad_norm": 1.6575478315353394,
+      "learning_rate": 7.409745461920698e-07,
+      "loss": 0.9049,
+      "step": 12295
+    },
+    {
+      "epoch": 0.29623563979672934,
+      "grad_norm": 1.7278908491134644,
+      "learning_rate": 7.408477842003853e-07,
+      "loss": 0.8964,
+      "step": 12300
+    },
+    {
+      "epoch": 0.2963560607885167,
+      "grad_norm": 1.4890385866165161,
+      "learning_rate": 7.40721022208701e-07,
+      "loss": 0.9283,
+      "step": 12305
+    },
+    {
+      "epoch": 0.29647648178030395,
+      "grad_norm": 1.5512890815734863,
+      "learning_rate": 7.405942602170165e-07,
+      "loss": 0.9398,
+      "step": 12310
+    },
+    {
+      "epoch": 0.2965969027720912,
+      "grad_norm": 1.420446753501892,
+      "learning_rate": 7.40467498225332e-07,
+      "loss": 0.9375,
+      "step": 12315
+    },
+    {
+      "epoch": 0.2967173237638785,
+      "grad_norm": 1.551830768585205,
+      "learning_rate": 7.403407362336477e-07,
+      "loss": 0.9162,
+      "step": 12320
+    },
+    {
+      "epoch": 0.2968377447556658,
+      "grad_norm": 1.4006000757217407,
+      "learning_rate": 7.402139742419633e-07,
+      "loss": 0.937,
+      "step": 12325
+    },
+    {
+      "epoch": 0.2969581657474531,
+      "grad_norm": 1.6034669876098633,
+      "learning_rate": 7.400872122502789e-07,
+      "loss": 0.9142,
+      "step": 12330
+    },
+    {
+      "epoch": 0.29707858673924037,
+      "grad_norm": 1.5593807697296143,
+      "learning_rate": 7.399604502585944e-07,
+      "loss": 0.949,
+      "step": 12335
+    },
+    {
+      "epoch": 0.2971990077310277,
+      "grad_norm": 1.5804564952850342,
+      "learning_rate": 7.3983368826691e-07,
+      "loss": 0.8958,
+      "step": 12340
+    },
+    {
+      "epoch": 0.29731942872281497,
+      "grad_norm": 1.548058271408081,
+      "learning_rate": 7.397069262752256e-07,
+      "loss": 0.9193,
+      "step": 12345
+    },
+    {
+      "epoch": 0.29743984971460224,
+      "grad_norm": 1.42953622341156,
+      "learning_rate": 7.395801642835412e-07,
+      "loss": 0.8941,
+      "step": 12350
+    },
+    {
+      "epoch": 0.2975602707063895,
+      "grad_norm": 1.4340226650238037,
+      "learning_rate": 7.394534022918568e-07,
+      "loss": 0.8934,
+      "step": 12355
+    },
+    {
+      "epoch": 0.29768069169817685,
+      "grad_norm": 2.00081205368042,
+      "learning_rate": 7.393266403001723e-07,
+      "loss": 0.9587,
+      "step": 12360
+    },
+    {
+      "epoch": 0.2978011126899641,
+      "grad_norm": 1.5126882791519165,
+      "learning_rate": 7.39199878308488e-07,
+      "loss": 0.9114,
+      "step": 12365
+    },
+    {
+      "epoch": 0.2979215336817514,
+      "grad_norm": 2.0931479930877686,
+      "learning_rate": 7.390731163168036e-07,
+      "loss": 0.9315,
+      "step": 12370
+    },
+    {
+      "epoch": 0.29804195467353867,
+      "grad_norm": 1.5068527460098267,
+      "learning_rate": 7.389463543251191e-07,
+      "loss": 0.966,
+      "step": 12375
+    },
+    {
+      "epoch": 0.298162375665326,
+      "grad_norm": 1.7116059064865112,
+      "learning_rate": 7.388195923334347e-07,
+      "loss": 0.9632,
+      "step": 12380
+    },
+    {
+      "epoch": 0.29828279665711327,
+      "grad_norm": 1.4662964344024658,
+      "learning_rate": 7.386928303417503e-07,
+      "loss": 0.9452,
+      "step": 12385
+    },
+    {
+      "epoch": 0.29840321764890054,
+      "grad_norm": 2.1836767196655273,
+      "learning_rate": 7.385660683500659e-07,
+      "loss": 0.9214,
+      "step": 12390
+    },
+    {
+      "epoch": 0.29852363864068787,
+      "grad_norm": 1.7127418518066406,
+      "learning_rate": 7.384393063583815e-07,
+      "loss": 0.9549,
+      "step": 12395
+    },
+    {
+      "epoch": 0.29864405963247515,
+      "grad_norm": 2.890199661254883,
+      "learning_rate": 7.383125443666971e-07,
+      "loss": 0.9384,
+      "step": 12400
+    },
+    {
+      "epoch": 0.2987644806242624,
+      "grad_norm": 1.4727007150650024,
+      "learning_rate": 7.381857823750126e-07,
+      "loss": 0.9184,
+      "step": 12405
+    },
+    {
+      "epoch": 0.2988849016160497,
+      "grad_norm": 1.4787065982818604,
+      "learning_rate": 7.380590203833282e-07,
+      "loss": 0.9226,
+      "step": 12410
+    },
+    {
+      "epoch": 0.299005322607837,
+      "grad_norm": 1.5422483682632446,
+      "learning_rate": 7.379322583916439e-07,
+      "loss": 0.926,
+      "step": 12415
+    },
+    {
+      "epoch": 0.2991257435996243,
+      "grad_norm": 1.5323668718338013,
+      "learning_rate": 7.378054963999594e-07,
+      "loss": 0.8974,
+      "step": 12420
+    },
+    {
+      "epoch": 0.29924616459141157,
+      "grad_norm": 1.539766550064087,
+      "learning_rate": 7.37678734408275e-07,
+      "loss": 0.9356,
+      "step": 12425
+    },
+    {
+      "epoch": 0.29936658558319884,
+      "grad_norm": 1.5156326293945312,
+      "learning_rate": 7.375519724165906e-07,
+      "loss": 0.9192,
+      "step": 12430
+    },
+    {
+      "epoch": 0.29948700657498617,
+      "grad_norm": 1.5541387796401978,
+      "learning_rate": 7.374252104249061e-07,
+      "loss": 0.9117,
+      "step": 12435
+    },
+    {
+      "epoch": 0.29960742756677344,
+      "grad_norm": 1.493133544921875,
+      "learning_rate": 7.372984484332218e-07,
+      "loss": 0.9163,
+      "step": 12440
+    },
+    {
+      "epoch": 0.2997278485585607,
+      "grad_norm": 1.497205376625061,
+      "learning_rate": 7.371716864415374e-07,
+      "loss": 0.9502,
+      "step": 12445
+    },
+    {
+      "epoch": 0.299848269550348,
+      "grad_norm": 1.4042925834655762,
+      "learning_rate": 7.370449244498529e-07,
+      "loss": 0.9456,
+      "step": 12450
+    },
+    {
+      "epoch": 0.2999686905421353,
+      "grad_norm": 1.5551249980926514,
+      "learning_rate": 7.369181624581685e-07,
+      "loss": 0.9541,
+      "step": 12455
+    },
+    {
+      "epoch": 0.3000891115339226,
+      "grad_norm": 1.5501880645751953,
+      "learning_rate": 7.367914004664841e-07,
+      "loss": 0.9713,
+      "step": 12460
+    },
+    {
+      "epoch": 0.30020953252570987,
+      "grad_norm": 1.3974348306655884,
+      "learning_rate": 7.366646384747997e-07,
+      "loss": 0.935,
+      "step": 12465
+    },
+    {
+      "epoch": 0.3003299535174972,
+      "grad_norm": 1.6827278137207031,
+      "learning_rate": 7.365378764831152e-07,
+      "loss": 0.9721,
+      "step": 12470
+    },
+    {
+      "epoch": 0.30045037450928447,
+      "grad_norm": 1.3156495094299316,
+      "learning_rate": 7.364111144914309e-07,
+      "loss": 0.9203,
+      "step": 12475
+    },
+    {
+      "epoch": 0.30057079550107174,
+      "grad_norm": 1.7625117301940918,
+      "learning_rate": 7.362843524997464e-07,
+      "loss": 0.9274,
+      "step": 12480
+    },
+    {
+      "epoch": 0.300691216492859,
+      "grad_norm": 1.4372049570083618,
+      "learning_rate": 7.36157590508062e-07,
+      "loss": 0.925,
+      "step": 12485
+    },
+    {
+      "epoch": 0.30081163748464634,
+      "grad_norm": 1.524330496788025,
+      "learning_rate": 7.360308285163777e-07,
+      "loss": 0.8828,
+      "step": 12490
+    },
+    {
+      "epoch": 0.3009320584764336,
+      "grad_norm": 1.4534857273101807,
+      "learning_rate": 7.359040665246931e-07,
+      "loss": 0.9535,
+      "step": 12495
+    },
+    {
+      "epoch": 0.3010524794682209,
+      "grad_norm": 1.603531837463379,
+      "learning_rate": 7.357773045330088e-07,
+      "loss": 0.9084,
+      "step": 12500
+    },
+    {
+      "epoch": 0.30117290046000816,
+      "grad_norm": 1.6054507493972778,
+      "learning_rate": 7.356505425413244e-07,
+      "loss": 0.9002,
+      "step": 12505
+    },
+    {
+      "epoch": 0.3012933214517955,
+      "grad_norm": 1.4217936992645264,
+      "learning_rate": 7.3552378054964e-07,
+      "loss": 0.9002,
+      "step": 12510
+    },
+    {
+      "epoch": 0.30141374244358277,
+      "grad_norm": 1.3850017786026,
+      "learning_rate": 7.353970185579556e-07,
+      "loss": 0.9223,
+      "step": 12515
+    },
+    {
+      "epoch": 0.30153416343537004,
+      "grad_norm": 1.639332890510559,
+      "learning_rate": 7.352702565662711e-07,
+      "loss": 0.8923,
+      "step": 12520
+    },
+    {
+      "epoch": 0.30165458442715737,
+      "grad_norm": 1.4566148519515991,
+      "learning_rate": 7.351434945745867e-07,
+      "loss": 0.9832,
+      "step": 12525
+    },
+    {
+      "epoch": 0.30177500541894464,
+      "grad_norm": 1.6120548248291016,
+      "learning_rate": 7.350167325829023e-07,
+      "loss": 0.9499,
+      "step": 12530
+    },
+    {
+      "epoch": 0.3018954264107319,
+      "grad_norm": 1.4781494140625,
+      "learning_rate": 7.34889970591218e-07,
+      "loss": 0.9534,
+      "step": 12535
+    },
+    {
+      "epoch": 0.3020158474025192,
+      "grad_norm": 1.6403977870941162,
+      "learning_rate": 7.347632085995334e-07,
+      "loss": 0.9437,
+      "step": 12540
+    },
+    {
+      "epoch": 0.3021362683943065,
+      "grad_norm": 1.8433960676193237,
+      "learning_rate": 7.34636446607849e-07,
+      "loss": 0.9331,
+      "step": 12545
+    },
+    {
+      "epoch": 0.3022566893860938,
+      "grad_norm": 1.4172022342681885,
+      "learning_rate": 7.345096846161647e-07,
+      "loss": 0.9138,
+      "step": 12550
+    },
+    {
+      "epoch": 0.30237711037788106,
+      "grad_norm": 1.539608359336853,
+      "learning_rate": 7.343829226244802e-07,
+      "loss": 0.965,
+      "step": 12555
+    },
+    {
+      "epoch": 0.30249753136966834,
+      "grad_norm": 1.4750511646270752,
+      "learning_rate": 7.342561606327959e-07,
+      "loss": 0.9426,
+      "step": 12560
+    },
+    {
+      "epoch": 0.30261795236145567,
+      "grad_norm": 1.6245813369750977,
+      "learning_rate": 7.341293986411114e-07,
+      "loss": 0.9602,
+      "step": 12565
+    },
+    {
+      "epoch": 0.30273837335324294,
+      "grad_norm": 1.6092920303344727,
+      "learning_rate": 7.34002636649427e-07,
+      "loss": 0.9319,
+      "step": 12570
+    },
+    {
+      "epoch": 0.3028587943450302,
+      "grad_norm": 1.5803841352462769,
+      "learning_rate": 7.338758746577426e-07,
+      "loss": 0.9257,
+      "step": 12575
+    },
+    {
+      "epoch": 0.3029792153368175,
+      "grad_norm": 1.3965332508087158,
+      "learning_rate": 7.337491126660582e-07,
+      "loss": 0.9672,
+      "step": 12580
+    },
+    {
+      "epoch": 0.3030996363286048,
+      "grad_norm": 1.4964956045150757,
+      "learning_rate": 7.336223506743737e-07,
+      "loss": 0.9366,
+      "step": 12585
+    },
+    {
+      "epoch": 0.3032200573203921,
+      "grad_norm": 1.4547592401504517,
+      "learning_rate": 7.334955886826893e-07,
+      "loss": 0.966,
+      "step": 12590
+    },
+    {
+      "epoch": 0.30334047831217936,
+      "grad_norm": 1.5723978281021118,
+      "learning_rate": 7.33368826691005e-07,
+      "loss": 0.9346,
+      "step": 12595
+    },
+    {
+      "epoch": 0.3034608993039667,
+      "grad_norm": 1.736131191253662,
+      "learning_rate": 7.332420646993205e-07,
+      "loss": 0.9028,
+      "step": 12600
+    },
+    {
+      "epoch": 0.30358132029575396,
+      "grad_norm": 1.5331920385360718,
+      "learning_rate": 7.331153027076361e-07,
+      "loss": 0.951,
+      "step": 12605
+    },
+    {
+      "epoch": 0.30370174128754124,
+      "grad_norm": 1.4614812135696411,
+      "learning_rate": 7.329885407159517e-07,
+      "loss": 0.9038,
+      "step": 12610
+    },
+    {
+      "epoch": 0.3038221622793285,
+      "grad_norm": 1.489661693572998,
+      "learning_rate": 7.328617787242672e-07,
+      "loss": 0.9479,
+      "step": 12615
+    },
+    {
+      "epoch": 0.30394258327111584,
+      "grad_norm": 1.7694226503372192,
+      "learning_rate": 7.327350167325829e-07,
+      "loss": 0.9877,
+      "step": 12620
+    },
+    {
+      "epoch": 0.3040630042629031,
+      "grad_norm": 1.6576279401779175,
+      "learning_rate": 7.326082547408985e-07,
+      "loss": 0.9251,
+      "step": 12625
+    },
+    {
+      "epoch": 0.3041834252546904,
+      "grad_norm": 1.6266634464263916,
+      "learning_rate": 7.324814927492141e-07,
+      "loss": 0.9477,
+      "step": 12630
+    },
+    {
+      "epoch": 0.30430384624647766,
+      "grad_norm": 1.8012726306915283,
+      "learning_rate": 7.323547307575296e-07,
+      "loss": 0.9414,
+      "step": 12635
+    },
+    {
+      "epoch": 0.304424267238265,
+      "grad_norm": 1.5876845121383667,
+      "learning_rate": 7.322279687658452e-07,
+      "loss": 0.8884,
+      "step": 12640
+    },
+    {
+      "epoch": 0.30454468823005226,
+      "grad_norm": 1.5976060628890991,
+      "learning_rate": 7.321012067741608e-07,
+      "loss": 0.9195,
+      "step": 12645
+    },
+    {
+      "epoch": 0.30466510922183954,
+      "grad_norm": 1.7384992837905884,
+      "learning_rate": 7.319744447824764e-07,
+      "loss": 0.8957,
+      "step": 12650
+    },
+    {
+      "epoch": 0.30478553021362687,
+      "grad_norm": 1.5005650520324707,
+      "learning_rate": 7.31847682790792e-07,
+      "loss": 0.8999,
+      "step": 12655
+    },
+    {
+      "epoch": 0.30490595120541414,
+      "grad_norm": 1.669329047203064,
+      "learning_rate": 7.317209207991075e-07,
+      "loss": 0.9812,
+      "step": 12660
+    },
+    {
+      "epoch": 0.3050263721972014,
+      "grad_norm": 2.4113380908966064,
+      "learning_rate": 7.315941588074231e-07,
+      "loss": 0.937,
+      "step": 12665
+    },
+    {
+      "epoch": 0.3051467931889887,
+      "grad_norm": 1.5560503005981445,
+      "learning_rate": 7.314673968157388e-07,
+      "loss": 0.9361,
+      "step": 12670
+    },
+    {
+      "epoch": 0.305267214180776,
+      "grad_norm": 1.586330533027649,
+      "learning_rate": 7.313406348240543e-07,
+      "loss": 0.8831,
+      "step": 12675
+    },
+    {
+      "epoch": 0.3053876351725633,
+      "grad_norm": 1.5432112216949463,
+      "learning_rate": 7.312138728323699e-07,
+      "loss": 0.9674,
+      "step": 12680
+    },
+    {
+      "epoch": 0.30550805616435056,
+      "grad_norm": 1.4705675840377808,
+      "learning_rate": 7.310871108406855e-07,
+      "loss": 0.9631,
+      "step": 12685
+    },
+    {
+      "epoch": 0.30562847715613783,
+      "grad_norm": 1.5707398653030396,
+      "learning_rate": 7.30960348849001e-07,
+      "loss": 0.9109,
+      "step": 12690
+    },
+    {
+      "epoch": 0.30574889814792516,
+      "grad_norm": 1.442728042602539,
+      "learning_rate": 7.308335868573167e-07,
+      "loss": 0.939,
+      "step": 12695
+    },
+    {
+      "epoch": 0.30586931913971244,
+      "grad_norm": 1.4560580253601074,
+      "learning_rate": 7.307068248656323e-07,
+      "loss": 0.9516,
+      "step": 12700
+    },
+    {
+      "epoch": 0.3059897401314997,
+      "grad_norm": 1.3931323289871216,
+      "learning_rate": 7.305800628739478e-07,
+      "loss": 0.9051,
+      "step": 12705
+    },
+    {
+      "epoch": 0.30611016112328704,
+      "grad_norm": 1.5336275100708008,
+      "learning_rate": 7.304533008822634e-07,
+      "loss": 0.9185,
+      "step": 12710
+    },
+    {
+      "epoch": 0.3062305821150743,
+      "grad_norm": 1.623144507408142,
+      "learning_rate": 7.303265388905791e-07,
+      "loss": 0.9126,
+      "step": 12715
+    },
+    {
+      "epoch": 0.3063510031068616,
+      "grad_norm": 1.3647394180297852,
+      "learning_rate": 7.301997768988946e-07,
+      "loss": 0.9169,
+      "step": 12720
+    },
+    {
+      "epoch": 0.30647142409864886,
+      "grad_norm": 1.5183027982711792,
+      "learning_rate": 7.300730149072101e-07,
+      "loss": 0.8957,
+      "step": 12725
+    },
+    {
+      "epoch": 0.3065918450904362,
+      "grad_norm": 1.7214516401290894,
+      "learning_rate": 7.299462529155258e-07,
+      "loss": 0.8836,
+      "step": 12730
+    },
+    {
+      "epoch": 0.30671226608222346,
+      "grad_norm": 1.7252259254455566,
+      "learning_rate": 7.298194909238413e-07,
+      "loss": 0.8943,
+      "step": 12735
+    },
+    {
+      "epoch": 0.30683268707401073,
+      "grad_norm": 1.7761805057525635,
+      "learning_rate": 7.29692728932157e-07,
+      "loss": 0.9436,
+      "step": 12740
+    },
+    {
+      "epoch": 0.306953108065798,
+      "grad_norm": 1.742750883102417,
+      "learning_rate": 7.295659669404726e-07,
+      "loss": 0.9304,
+      "step": 12745
+    },
+    {
+      "epoch": 0.30707352905758534,
+      "grad_norm": 1.8857698440551758,
+      "learning_rate": 7.29439204948788e-07,
+      "loss": 0.9366,
+      "step": 12750
+    },
+    {
+      "epoch": 0.3071939500493726,
+      "grad_norm": 1.5318527221679688,
+      "learning_rate": 7.293124429571037e-07,
+      "loss": 0.9186,
+      "step": 12755
+    },
+    {
+      "epoch": 0.3073143710411599,
+      "grad_norm": 1.5236737728118896,
+      "learning_rate": 7.291856809654193e-07,
+      "loss": 0.9085,
+      "step": 12760
+    },
+    {
+      "epoch": 0.30743479203294716,
+      "grad_norm": 1.7529305219650269,
+      "learning_rate": 7.290589189737349e-07,
+      "loss": 0.8909,
+      "step": 12765
+    },
+    {
+      "epoch": 0.3075552130247345,
+      "grad_norm": 1.4512908458709717,
+      "learning_rate": 7.289321569820504e-07,
+      "loss": 0.9263,
+      "step": 12770
+    },
+    {
+      "epoch": 0.30767563401652176,
+      "grad_norm": 1.7084128856658936,
+      "learning_rate": 7.288053949903661e-07,
+      "loss": 0.9624,
+      "step": 12775
+    },
+    {
+      "epoch": 0.30779605500830903,
+      "grad_norm": 1.4374417066574097,
+      "learning_rate": 7.286786329986816e-07,
+      "loss": 0.9476,
+      "step": 12780
+    },
+    {
+      "epoch": 0.30791647600009636,
+      "grad_norm": 1.4318782091140747,
+      "learning_rate": 7.285518710069972e-07,
+      "loss": 0.9191,
+      "step": 12785
+    },
+    {
+      "epoch": 0.30803689699188364,
+      "grad_norm": 1.5451619625091553,
+      "learning_rate": 7.284251090153129e-07,
+      "loss": 0.8945,
+      "step": 12790
+    },
+    {
+      "epoch": 0.3081573179836709,
+      "grad_norm": 1.6542329788208008,
+      "learning_rate": 7.282983470236283e-07,
+      "loss": 0.8807,
+      "step": 12795
+    },
+    {
+      "epoch": 0.3082777389754582,
+      "grad_norm": 1.6969656944274902,
+      "learning_rate": 7.28171585031944e-07,
+      "loss": 0.9141,
+      "step": 12800
+    },
+    {
+      "epoch": 0.3083981599672455,
+      "grad_norm": 1.467103123664856,
+      "learning_rate": 7.280448230402596e-07,
+      "loss": 0.9163,
+      "step": 12805
+    },
+    {
+      "epoch": 0.3085185809590328,
+      "grad_norm": 1.5035455226898193,
+      "learning_rate": 7.279180610485751e-07,
+      "loss": 0.9495,
+      "step": 12810
+    },
+    {
+      "epoch": 0.30863900195082006,
+      "grad_norm": 1.504870891571045,
+      "learning_rate": 7.277912990568908e-07,
+      "loss": 0.8743,
+      "step": 12815
+    },
+    {
+      "epoch": 0.30875942294260733,
+      "grad_norm": 1.5313233137130737,
+      "learning_rate": 7.276645370652063e-07,
+      "loss": 0.907,
+      "step": 12820
+    },
+    {
+      "epoch": 0.30887984393439466,
+      "grad_norm": 1.7336504459381104,
+      "learning_rate": 7.275377750735219e-07,
+      "loss": 0.9576,
+      "step": 12825
+    },
+    {
+      "epoch": 0.30900026492618193,
+      "grad_norm": 1.4466485977172852,
+      "learning_rate": 7.274110130818375e-07,
+      "loss": 0.932,
+      "step": 12830
+    },
+    {
+      "epoch": 0.3091206859179692,
+      "grad_norm": 1.5821729898452759,
+      "learning_rate": 7.272842510901532e-07,
+      "loss": 0.8947,
+      "step": 12835
+    },
+    {
+      "epoch": 0.30924110690975654,
+      "grad_norm": 1.4757659435272217,
+      "learning_rate": 7.271574890984686e-07,
+      "loss": 0.9697,
+      "step": 12840
+    },
+    {
+      "epoch": 0.3093615279015438,
+      "grad_norm": 1.3930412530899048,
+      "learning_rate": 7.270307271067842e-07,
+      "loss": 0.8832,
+      "step": 12845
+    },
+    {
+      "epoch": 0.3094819488933311,
+      "grad_norm": 1.5699188709259033,
+      "learning_rate": 7.269039651150999e-07,
+      "loss": 0.9627,
+      "step": 12850
+    },
+    {
+      "epoch": 0.30960236988511836,
+      "grad_norm": 1.5719423294067383,
+      "learning_rate": 7.267772031234154e-07,
+      "loss": 0.9564,
+      "step": 12855
+    },
+    {
+      "epoch": 0.3097227908769057,
+      "grad_norm": 1.3833011388778687,
+      "learning_rate": 7.266504411317311e-07,
+      "loss": 0.8952,
+      "step": 12860
+    },
+    {
+      "epoch": 0.30984321186869296,
+      "grad_norm": 1.3970744609832764,
+      "learning_rate": 7.265236791400466e-07,
+      "loss": 0.881,
+      "step": 12865
+    },
+    {
+      "epoch": 0.30996363286048023,
+      "grad_norm": 1.5236992835998535,
+      "learning_rate": 7.263969171483621e-07,
+      "loss": 0.9615,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3100840538522675,
+      "grad_norm": 1.4951473474502563,
+      "learning_rate": 7.262701551566778e-07,
+      "loss": 0.9406,
+      "step": 12875
+    },
+    {
+      "epoch": 0.31020447484405483,
+      "grad_norm": 1.4671101570129395,
+      "learning_rate": 7.261433931649934e-07,
+      "loss": 0.9372,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3103248958358421,
+      "grad_norm": 1.4430850744247437,
+      "learning_rate": 7.260166311733089e-07,
+      "loss": 0.9654,
+      "step": 12885
+    },
+    {
+      "epoch": 0.3104453168276294,
+      "grad_norm": 1.451224684715271,
+      "learning_rate": 7.258898691816245e-07,
+      "loss": 0.9165,
+      "step": 12890
+    },
+    {
+      "epoch": 0.31056573781941665,
+      "grad_norm": 1.6193621158599854,
+      "learning_rate": 7.257631071899402e-07,
+      "loss": 0.9557,
+      "step": 12895
+    },
+    {
+      "epoch": 0.310686158811204,
+      "grad_norm": 1.431778073310852,
+      "learning_rate": 7.256363451982557e-07,
+      "loss": 0.9045,
+      "step": 12900
+    },
+    {
+      "epoch": 0.31080657980299126,
+      "grad_norm": 1.5868239402770996,
+      "learning_rate": 7.255095832065713e-07,
+      "loss": 0.9615,
+      "step": 12905
+    },
+    {
+      "epoch": 0.31092700079477853,
+      "grad_norm": 1.6796597242355347,
+      "learning_rate": 7.253828212148869e-07,
+      "loss": 0.9218,
+      "step": 12910
+    },
+    {
+      "epoch": 0.31104742178656586,
+      "grad_norm": 1.586971402168274,
+      "learning_rate": 7.252560592232024e-07,
+      "loss": 0.9413,
+      "step": 12915
+    },
+    {
+      "epoch": 0.31116784277835313,
+      "grad_norm": 1.555291771888733,
+      "learning_rate": 7.251292972315181e-07,
+      "loss": 0.8944,
+      "step": 12920
+    },
+    {
+      "epoch": 0.3112882637701404,
+      "grad_norm": 1.624487042427063,
+      "learning_rate": 7.250025352398337e-07,
+      "loss": 0.9409,
+      "step": 12925
+    },
+    {
+      "epoch": 0.3114086847619277,
+      "grad_norm": 1.8175469636917114,
+      "learning_rate": 7.248757732481492e-07,
+      "loss": 0.8911,
+      "step": 12930
+    },
+    {
+      "epoch": 0.311529105753715,
+      "grad_norm": 1.4849705696105957,
+      "learning_rate": 7.247490112564648e-07,
+      "loss": 0.927,
+      "step": 12935
+    },
+    {
+      "epoch": 0.3116495267455023,
+      "grad_norm": 1.5028043985366821,
+      "learning_rate": 7.246222492647804e-07,
+      "loss": 0.9222,
+      "step": 12940
+    },
+    {
+      "epoch": 0.31176994773728955,
+      "grad_norm": 1.5761449337005615,
+      "learning_rate": 7.24495487273096e-07,
+      "loss": 0.9601,
+      "step": 12945
+    },
+    {
+      "epoch": 0.31189036872907683,
+      "grad_norm": 1.5496820211410522,
+      "learning_rate": 7.243687252814116e-07,
+      "loss": 0.9189,
+      "step": 12950
+    },
+    {
+      "epoch": 0.31201078972086416,
+      "grad_norm": 1.502267599105835,
+      "learning_rate": 7.242419632897271e-07,
+      "loss": 0.8692,
+      "step": 12955
+    },
+    {
+      "epoch": 0.31213121071265143,
+      "grad_norm": 1.561307668685913,
+      "learning_rate": 7.241152012980427e-07,
+      "loss": 0.9263,
+      "step": 12960
+    },
+    {
+      "epoch": 0.3122516317044387,
+      "grad_norm": 1.597617268562317,
+      "learning_rate": 7.239884393063583e-07,
+      "loss": 0.888,
+      "step": 12965
+    },
+    {
+      "epoch": 0.31237205269622603,
+      "grad_norm": 1.7717195749282837,
+      "learning_rate": 7.23861677314674e-07,
+      "loss": 0.893,
+      "step": 12970
+    },
+    {
+      "epoch": 0.3124924736880133,
+      "grad_norm": 1.3559247255325317,
+      "learning_rate": 7.237349153229895e-07,
+      "loss": 0.9217,
+      "step": 12975
+    },
+    {
+      "epoch": 0.3126128946798006,
+      "grad_norm": 1.5530009269714355,
+      "learning_rate": 7.236081533313051e-07,
+      "loss": 0.9354,
+      "step": 12980
+    },
+    {
+      "epoch": 0.31273331567158785,
+      "grad_norm": 1.5920103788375854,
+      "learning_rate": 7.234813913396207e-07,
+      "loss": 0.906,
+      "step": 12985
+    },
+    {
+      "epoch": 0.3128537366633752,
+      "grad_norm": 1.5397907495498657,
+      "learning_rate": 7.233546293479362e-07,
+      "loss": 0.8861,
+      "step": 12990
+    },
+    {
+      "epoch": 0.31297415765516245,
+      "grad_norm": 1.7374628782272339,
+      "learning_rate": 7.232278673562519e-07,
+      "loss": 0.9622,
+      "step": 12995
+    },
+    {
+      "epoch": 0.31309457864694973,
+      "grad_norm": 1.6311469078063965,
+      "learning_rate": 7.231011053645675e-07,
+      "loss": 0.9128,
+      "step": 13000
+    },
+    {
+      "epoch": 0.313214999638737,
+      "grad_norm": 1.3971755504608154,
+      "learning_rate": 7.22974343372883e-07,
+      "loss": 0.9522,
+      "step": 13005
+    },
+    {
+      "epoch": 0.31333542063052433,
+      "grad_norm": 1.5353749990463257,
+      "learning_rate": 7.228475813811986e-07,
+      "loss": 0.9066,
+      "step": 13010
+    },
+    {
+      "epoch": 0.3134558416223116,
+      "grad_norm": 1.4577460289001465,
+      "learning_rate": 7.227208193895143e-07,
+      "loss": 0.9182,
+      "step": 13015
+    },
+    {
+      "epoch": 0.3135762626140989,
+      "grad_norm": 1.6157535314559937,
+      "learning_rate": 7.225940573978298e-07,
+      "loss": 0.904,
+      "step": 13020
+    },
+    {
+      "epoch": 0.31369668360588615,
+      "grad_norm": 1.75568687915802,
+      "learning_rate": 7.224672954061453e-07,
+      "loss": 0.9106,
+      "step": 13025
+    },
+    {
+      "epoch": 0.3138171045976735,
+      "grad_norm": 1.5280760526657104,
+      "learning_rate": 7.22340533414461e-07,
+      "loss": 0.923,
+      "step": 13030
+    },
+    {
+      "epoch": 0.31393752558946075,
+      "grad_norm": 1.4735020399093628,
+      "learning_rate": 7.222137714227765e-07,
+      "loss": 0.914,
+      "step": 13035
+    },
+    {
+      "epoch": 0.314057946581248,
+      "grad_norm": 1.794739842414856,
+      "learning_rate": 7.220870094310922e-07,
+      "loss": 0.9401,
+      "step": 13040
+    },
+    {
+      "epoch": 0.31417836757303536,
+      "grad_norm": 1.6184120178222656,
+      "learning_rate": 7.219602474394078e-07,
+      "loss": 0.9148,
+      "step": 13045
+    },
+    {
+      "epoch": 0.31429878856482263,
+      "grad_norm": 1.521453857421875,
+      "learning_rate": 7.218334854477232e-07,
+      "loss": 0.9631,
+      "step": 13050
+    },
+    {
+      "epoch": 0.3144192095566099,
+      "grad_norm": 1.5748240947723389,
+      "learning_rate": 7.217067234560389e-07,
+      "loss": 0.933,
+      "step": 13055
+    },
+    {
+      "epoch": 0.3145396305483972,
+      "grad_norm": 1.590860366821289,
+      "learning_rate": 7.215799614643545e-07,
+      "loss": 0.8997,
+      "step": 13060
+    },
+    {
+      "epoch": 0.3146600515401845,
+      "grad_norm": 1.5484305620193481,
+      "learning_rate": 7.214531994726701e-07,
+      "loss": 0.9044,
+      "step": 13065
+    },
+    {
+      "epoch": 0.3147804725319718,
+      "grad_norm": 1.5679107904434204,
+      "learning_rate": 7.213264374809856e-07,
+      "loss": 0.8809,
+      "step": 13070
+    },
+    {
+      "epoch": 0.31490089352375905,
+      "grad_norm": 1.368302345275879,
+      "learning_rate": 7.211996754893012e-07,
+      "loss": 0.9278,
+      "step": 13075
+    },
+    {
+      "epoch": 0.3150213145155463,
+      "grad_norm": 1.55647873878479,
+      "learning_rate": 7.210729134976169e-07,
+      "loss": 0.9217,
+      "step": 13080
+    },
+    {
+      "epoch": 0.31514173550733365,
+      "grad_norm": 1.5681729316711426,
+      "learning_rate": 7.209461515059324e-07,
+      "loss": 0.9554,
+      "step": 13085
+    },
+    {
+      "epoch": 0.3152621564991209,
+      "grad_norm": 1.511460781097412,
+      "learning_rate": 7.208193895142481e-07,
+      "loss": 0.9247,
+      "step": 13090
+    },
+    {
+      "epoch": 0.3153825774909082,
+      "grad_norm": 1.5184860229492188,
+      "learning_rate": 7.206926275225636e-07,
+      "loss": 0.8765,
+      "step": 13095
+    },
+    {
+      "epoch": 0.31550299848269553,
+      "grad_norm": 1.7019352912902832,
+      "learning_rate": 7.205658655308792e-07,
+      "loss": 0.9122,
+      "step": 13100
+    },
+    {
+      "epoch": 0.3156234194744828,
+      "grad_norm": 1.5765482187271118,
+      "learning_rate": 7.204391035391948e-07,
+      "loss": 0.9133,
+      "step": 13105
+    },
+    {
+      "epoch": 0.3157438404662701,
+      "grad_norm": 1.5094279050827026,
+      "learning_rate": 7.203123415475104e-07,
+      "loss": 0.9337,
+      "step": 13110
+    },
+    {
+      "epoch": 0.31586426145805735,
+      "grad_norm": 1.4930437803268433,
+      "learning_rate": 7.20185579555826e-07,
+      "loss": 0.9392,
+      "step": 13115
+    },
+    {
+      "epoch": 0.3159846824498447,
+      "grad_norm": 1.6247256994247437,
+      "learning_rate": 7.200588175641415e-07,
+      "loss": 0.9644,
+      "step": 13120
+    },
+    {
+      "epoch": 0.31610510344163195,
+      "grad_norm": 2.296663522720337,
+      "learning_rate": 7.199320555724572e-07,
+      "loss": 0.9872,
+      "step": 13125
+    },
+    {
+      "epoch": 0.3162255244334192,
+      "grad_norm": 1.671225666999817,
+      "learning_rate": 7.198052935807727e-07,
+      "loss": 0.9924,
+      "step": 13130
+    },
+    {
+      "epoch": 0.3163459454252065,
+      "grad_norm": 1.6024426221847534,
+      "learning_rate": 7.196785315890883e-07,
+      "loss": 0.9553,
+      "step": 13135
+    },
+    {
+      "epoch": 0.3164663664169938,
+      "grad_norm": 1.5475425720214844,
+      "learning_rate": 7.195517695974039e-07,
+      "loss": 0.913,
+      "step": 13140
+    },
+    {
+      "epoch": 0.3165867874087811,
+      "grad_norm": 1.7463332414627075,
+      "learning_rate": 7.194250076057194e-07,
+      "loss": 0.9436,
+      "step": 13145
+    },
+    {
+      "epoch": 0.3167072084005684,
+      "grad_norm": 1.461932897567749,
+      "learning_rate": 7.192982456140351e-07,
+      "loss": 0.9176,
+      "step": 13150
+    },
+    {
+      "epoch": 0.3168276293923557,
+      "grad_norm": 1.4646632671356201,
+      "learning_rate": 7.191714836223507e-07,
+      "loss": 0.9003,
+      "step": 13155
+    },
+    {
+      "epoch": 0.316948050384143,
+      "grad_norm": 1.5874431133270264,
+      "learning_rate": 7.190447216306663e-07,
+      "loss": 0.9587,
+      "step": 13160
+    },
+    {
+      "epoch": 0.31706847137593025,
+      "grad_norm": 1.8062708377838135,
+      "learning_rate": 7.189179596389818e-07,
+      "loss": 0.921,
+      "step": 13165
+    },
+    {
+      "epoch": 0.3171888923677175,
+      "grad_norm": 1.5187337398529053,
+      "learning_rate": 7.187911976472974e-07,
+      "loss": 0.9286,
+      "step": 13170
+    },
+    {
+      "epoch": 0.31730931335950485,
+      "grad_norm": 1.4751627445220947,
+      "learning_rate": 7.18664435655613e-07,
+      "loss": 0.8909,
+      "step": 13175
+    },
+    {
+      "epoch": 0.3174297343512921,
+      "grad_norm": 1.4710909128189087,
+      "learning_rate": 7.185376736639286e-07,
+      "loss": 0.8811,
+      "step": 13180
+    },
+    {
+      "epoch": 0.3175501553430794,
+      "grad_norm": 1.7347071170806885,
+      "learning_rate": 7.184109116722443e-07,
+      "loss": 0.9272,
+      "step": 13185
+    },
+    {
+      "epoch": 0.31767057633486667,
+      "grad_norm": 1.5981749296188354,
+      "learning_rate": 7.182841496805597e-07,
+      "loss": 0.9722,
+      "step": 13190
+    },
+    {
+      "epoch": 0.317790997326654,
+      "grad_norm": 1.7842888832092285,
+      "learning_rate": 7.181573876888753e-07,
+      "loss": 0.9258,
+      "step": 13195
+    },
+    {
+      "epoch": 0.3179114183184413,
+      "grad_norm": 1.4398106336593628,
+      "learning_rate": 7.18030625697191e-07,
+      "loss": 0.8923,
+      "step": 13200
+    },
+    {
+      "epoch": 0.31803183931022855,
+      "grad_norm": 1.4745246171951294,
+      "learning_rate": 7.179038637055065e-07,
+      "loss": 0.9105,
+      "step": 13205
+    },
+    {
+      "epoch": 0.3181522603020158,
+      "grad_norm": 1.5541917085647583,
+      "learning_rate": 7.177771017138221e-07,
+      "loss": 0.8613,
+      "step": 13210
+    },
+    {
+      "epoch": 0.31827268129380315,
+      "grad_norm": 1.4687206745147705,
+      "learning_rate": 7.176503397221377e-07,
+      "loss": 0.9219,
+      "step": 13215
+    },
+    {
+      "epoch": 0.3183931022855904,
+      "grad_norm": 1.6779239177703857,
+      "learning_rate": 7.175235777304532e-07,
+      "loss": 0.8582,
+      "step": 13220
+    },
+    {
+      "epoch": 0.3185135232773777,
+      "grad_norm": 1.4923917055130005,
+      "learning_rate": 7.173968157387689e-07,
+      "loss": 0.9351,
+      "step": 13225
+    },
+    {
+      "epoch": 0.318633944269165,
+      "grad_norm": 1.669708490371704,
+      "learning_rate": 7.172700537470845e-07,
+      "loss": 0.973,
+      "step": 13230
+    },
+    {
+      "epoch": 0.3187543652609523,
+      "grad_norm": 1.9078813791275024,
+      "learning_rate": 7.171432917554e-07,
+      "loss": 0.9391,
+      "step": 13235
+    },
+    {
+      "epoch": 0.3188747862527396,
+      "grad_norm": 1.4691177606582642,
+      "learning_rate": 7.170165297637156e-07,
+      "loss": 0.9232,
+      "step": 13240
+    },
+    {
+      "epoch": 0.31899520724452685,
+      "grad_norm": 1.5904854536056519,
+      "learning_rate": 7.168897677720313e-07,
+      "loss": 0.9105,
+      "step": 13245
+    },
+    {
+      "epoch": 0.3191156282363142,
+      "grad_norm": 1.7188918590545654,
+      "learning_rate": 7.167630057803468e-07,
+      "loss": 0.884,
+      "step": 13250
+    },
+    {
+      "epoch": 0.31923604922810145,
+      "grad_norm": 1.4673360586166382,
+      "learning_rate": 7.166362437886623e-07,
+      "loss": 0.9632,
+      "step": 13255
+    },
+    {
+      "epoch": 0.3193564702198887,
+      "grad_norm": 1.5699816942214966,
+      "learning_rate": 7.16509481796978e-07,
+      "loss": 0.9771,
+      "step": 13260
+    },
+    {
+      "epoch": 0.319476891211676,
+      "grad_norm": 1.598327398300171,
+      "learning_rate": 7.163827198052935e-07,
+      "loss": 0.9507,
+      "step": 13265
+    },
+    {
+      "epoch": 0.3195973122034633,
+      "grad_norm": 1.6673810482025146,
+      "learning_rate": 7.162559578136092e-07,
+      "loss": 0.9501,
+      "step": 13270
+    },
+    {
+      "epoch": 0.3197177331952506,
+      "grad_norm": 1.5441638231277466,
+      "learning_rate": 7.161291958219248e-07,
+      "loss": 0.9359,
+      "step": 13275
+    },
+    {
+      "epoch": 0.31983815418703787,
+      "grad_norm": 1.3806934356689453,
+      "learning_rate": 7.160024338302402e-07,
+      "loss": 0.8852,
+      "step": 13280
+    },
+    {
+      "epoch": 0.3199585751788252,
+      "grad_norm": 1.5654473304748535,
+      "learning_rate": 7.158756718385559e-07,
+      "loss": 0.9018,
+      "step": 13285
+    },
+    {
+      "epoch": 0.3200789961706125,
+      "grad_norm": 1.5995250940322876,
+      "learning_rate": 7.157489098468715e-07,
+      "loss": 0.9273,
+      "step": 13290
+    },
+    {
+      "epoch": 0.32019941716239975,
+      "grad_norm": 1.6036864519119263,
+      "learning_rate": 7.156221478551871e-07,
+      "loss": 0.9197,
+      "step": 13295
+    },
+    {
+      "epoch": 0.320319838154187,
+      "grad_norm": 1.4564919471740723,
+      "learning_rate": 7.154953858635027e-07,
+      "loss": 0.9494,
+      "step": 13300
+    },
+    {
+      "epoch": 0.32044025914597435,
+      "grad_norm": 1.441354513168335,
+      "learning_rate": 7.153686238718183e-07,
+      "loss": 0.9084,
+      "step": 13305
+    },
+    {
+      "epoch": 0.3205606801377616,
+      "grad_norm": 1.6930876970291138,
+      "learning_rate": 7.152418618801338e-07,
+      "loss": 0.9256,
+      "step": 13310
+    },
+    {
+      "epoch": 0.3206811011295489,
+      "grad_norm": 1.5789000988006592,
+      "learning_rate": 7.151150998884494e-07,
+      "loss": 0.918,
+      "step": 13315
+    },
+    {
+      "epoch": 0.32080152212133617,
+      "grad_norm": 1.6201459169387817,
+      "learning_rate": 7.149883378967651e-07,
+      "loss": 0.8801,
+      "step": 13320
+    },
+    {
+      "epoch": 0.3209219431131235,
+      "grad_norm": 1.62224543094635,
+      "learning_rate": 7.148615759050805e-07,
+      "loss": 0.8973,
+      "step": 13325
+    },
+    {
+      "epoch": 0.32104236410491077,
+      "grad_norm": 1.5501794815063477,
+      "learning_rate": 7.147348139133962e-07,
+      "loss": 0.9312,
+      "step": 13330
+    },
+    {
+      "epoch": 0.32116278509669804,
+      "grad_norm": 1.4192733764648438,
+      "learning_rate": 7.146080519217118e-07,
+      "loss": 0.9125,
+      "step": 13335
+    },
+    {
+      "epoch": 0.3212832060884853,
+      "grad_norm": 1.6199774742126465,
+      "learning_rate": 7.144812899300273e-07,
+      "loss": 0.9498,
+      "step": 13340
+    },
+    {
+      "epoch": 0.32140362708027265,
+      "grad_norm": 1.7698010206222534,
+      "learning_rate": 7.14354527938343e-07,
+      "loss": 0.9673,
+      "step": 13345
+    },
+    {
+      "epoch": 0.3215240480720599,
+      "grad_norm": 1.632602572441101,
+      "learning_rate": 7.142277659466585e-07,
+      "loss": 0.8909,
+      "step": 13350
+    },
+    {
+      "epoch": 0.3216444690638472,
+      "grad_norm": 1.4607270956039429,
+      "learning_rate": 7.141010039549741e-07,
+      "loss": 0.8579,
+      "step": 13355
+    },
+    {
+      "epoch": 0.3217648900556345,
+      "grad_norm": 1.6992939710617065,
+      "learning_rate": 7.139742419632897e-07,
+      "loss": 0.9084,
+      "step": 13360
+    },
+    {
+      "epoch": 0.3218853110474218,
+      "grad_norm": 1.440292239189148,
+      "learning_rate": 7.138474799716054e-07,
+      "loss": 0.8978,
+      "step": 13365
+    },
+    {
+      "epoch": 0.32200573203920907,
+      "grad_norm": 1.411969542503357,
+      "learning_rate": 7.137207179799208e-07,
+      "loss": 0.9203,
+      "step": 13370
+    },
+    {
+      "epoch": 0.32212615303099634,
+      "grad_norm": 1.7011206150054932,
+      "learning_rate": 7.135939559882364e-07,
+      "loss": 0.9135,
+      "step": 13375
+    },
+    {
+      "epoch": 0.32224657402278367,
+      "grad_norm": 1.639039158821106,
+      "learning_rate": 7.134671939965521e-07,
+      "loss": 0.9575,
+      "step": 13380
+    },
+    {
+      "epoch": 0.32236699501457095,
+      "grad_norm": 1.7481632232666016,
+      "learning_rate": 7.133404320048676e-07,
+      "loss": 0.9525,
+      "step": 13385
+    },
+    {
+      "epoch": 0.3224874160063582,
+      "grad_norm": 1.509905219078064,
+      "learning_rate": 7.132136700131833e-07,
+      "loss": 0.9806,
+      "step": 13390
+    },
+    {
+      "epoch": 0.3226078369981455,
+      "grad_norm": 1.5179061889648438,
+      "learning_rate": 7.130869080214988e-07,
+      "loss": 0.9796,
+      "step": 13395
+    },
+    {
+      "epoch": 0.3227282579899328,
+      "grad_norm": 1.6003575325012207,
+      "learning_rate": 7.129601460298143e-07,
+      "loss": 0.9339,
+      "step": 13400
+    },
+    {
+      "epoch": 0.3228486789817201,
+      "grad_norm": 1.5749802589416504,
+      "learning_rate": 7.1283338403813e-07,
+      "loss": 0.9196,
+      "step": 13405
+    },
+    {
+      "epoch": 0.32296909997350737,
+      "grad_norm": 1.4835052490234375,
+      "learning_rate": 7.127066220464456e-07,
+      "loss": 0.9415,
+      "step": 13410
+    },
+    {
+      "epoch": 0.3230895209652947,
+      "grad_norm": 1.6921086311340332,
+      "learning_rate": 7.125798600547611e-07,
+      "loss": 0.8908,
+      "step": 13415
+    },
+    {
+      "epoch": 0.32320994195708197,
+      "grad_norm": 1.3966692686080933,
+      "learning_rate": 7.124530980630767e-07,
+      "loss": 0.9145,
+      "step": 13420
+    },
+    {
+      "epoch": 0.32333036294886924,
+      "grad_norm": 1.4754492044448853,
+      "learning_rate": 7.123263360713924e-07,
+      "loss": 0.898,
+      "step": 13425
+    },
+    {
+      "epoch": 0.3234507839406565,
+      "grad_norm": 1.5174185037612915,
+      "learning_rate": 7.121995740797079e-07,
+      "loss": 0.8697,
+      "step": 13430
+    },
+    {
+      "epoch": 0.32357120493244385,
+      "grad_norm": 1.631089210510254,
+      "learning_rate": 7.120728120880235e-07,
+      "loss": 0.9433,
+      "step": 13435
+    },
+    {
+      "epoch": 0.3236916259242311,
+      "grad_norm": 1.4840291738510132,
+      "learning_rate": 7.119460500963391e-07,
+      "loss": 0.932,
+      "step": 13440
+    },
+    {
+      "epoch": 0.3238120469160184,
+      "grad_norm": 1.6019710302352905,
+      "learning_rate": 7.118192881046546e-07,
+      "loss": 0.9055,
+      "step": 13445
+    },
+    {
+      "epoch": 0.32393246790780567,
+      "grad_norm": 1.71696937084198,
+      "learning_rate": 7.116925261129703e-07,
+      "loss": 0.9434,
+      "step": 13450
+    },
+    {
+      "epoch": 0.324052888899593,
+      "grad_norm": 1.519882082939148,
+      "learning_rate": 7.115657641212859e-07,
+      "loss": 0.9167,
+      "step": 13455
+    },
+    {
+      "epoch": 0.32417330989138027,
+      "grad_norm": 1.4360700845718384,
+      "learning_rate": 7.114390021296014e-07,
+      "loss": 0.9351,
+      "step": 13460
+    },
+    {
+      "epoch": 0.32429373088316754,
+      "grad_norm": 1.5094038248062134,
+      "learning_rate": 7.11312240137917e-07,
+      "loss": 0.9051,
+      "step": 13465
+    },
+    {
+      "epoch": 0.3244141518749548,
+      "grad_norm": 1.4464014768600464,
+      "learning_rate": 7.111854781462326e-07,
+      "loss": 0.9252,
+      "step": 13470
+    },
+    {
+      "epoch": 0.32453457286674214,
+      "grad_norm": 1.5232619047164917,
+      "learning_rate": 7.110587161545482e-07,
+      "loss": 0.8695,
+      "step": 13475
+    },
+    {
+      "epoch": 0.3246549938585294,
+      "grad_norm": 1.4510326385498047,
+      "learning_rate": 7.109319541628638e-07,
+      "loss": 0.9222,
+      "step": 13480
+    },
+    {
+      "epoch": 0.3247754148503167,
+      "grad_norm": 1.5552394390106201,
+      "learning_rate": 7.108051921711793e-07,
+      "loss": 0.8811,
+      "step": 13485
+    },
+    {
+      "epoch": 0.324895835842104,
+      "grad_norm": 1.6308211088180542,
+      "learning_rate": 7.106784301794949e-07,
+      "loss": 0.8948,
+      "step": 13490
+    },
+    {
+      "epoch": 0.3250162568338913,
+      "grad_norm": 1.4746640920639038,
+      "learning_rate": 7.105516681878105e-07,
+      "loss": 0.9311,
+      "step": 13495
+    },
+    {
+      "epoch": 0.32513667782567857,
+      "grad_norm": 1.5032947063446045,
+      "learning_rate": 7.104249061961262e-07,
+      "loss": 0.9029,
+      "step": 13500
+    },
+    {
+      "epoch": 0.32525709881746584,
+      "grad_norm": 1.6702075004577637,
+      "learning_rate": 7.102981442044417e-07,
+      "loss": 0.9339,
+      "step": 13505
+    },
+    {
+      "epoch": 0.32537751980925317,
+      "grad_norm": 1.8243368864059448,
+      "learning_rate": 7.101713822127573e-07,
+      "loss": 0.9171,
+      "step": 13510
+    },
+    {
+      "epoch": 0.32549794080104044,
+      "grad_norm": 1.6092954874038696,
+      "learning_rate": 7.100446202210729e-07,
+      "loss": 0.9102,
+      "step": 13515
+    },
+    {
+      "epoch": 0.3256183617928277,
+      "grad_norm": 1.7775665521621704,
+      "learning_rate": 7.099178582293884e-07,
+      "loss": 0.9326,
+      "step": 13520
+    },
+    {
+      "epoch": 0.325738782784615,
+      "grad_norm": 1.5000969171524048,
+      "learning_rate": 7.097910962377041e-07,
+      "loss": 0.8927,
+      "step": 13525
+    },
+    {
+      "epoch": 0.3258592037764023,
+      "grad_norm": 2.0762746334075928,
+      "learning_rate": 7.096643342460197e-07,
+      "loss": 0.8871,
+      "step": 13530
+    },
+    {
+      "epoch": 0.3259796247681896,
+      "grad_norm": 1.554837942123413,
+      "learning_rate": 7.095375722543352e-07,
+      "loss": 0.9493,
+      "step": 13535
+    },
+    {
+      "epoch": 0.32610004575997686,
+      "grad_norm": 1.3311814069747925,
+      "learning_rate": 7.094108102626508e-07,
+      "loss": 0.9113,
+      "step": 13540
+    },
+    {
+      "epoch": 0.3262204667517642,
+      "grad_norm": 1.598663568496704,
+      "learning_rate": 7.092840482709665e-07,
+      "loss": 0.9669,
+      "step": 13545
+    },
+    {
+      "epoch": 0.32634088774355147,
+      "grad_norm": 1.6727991104125977,
+      "learning_rate": 7.09157286279282e-07,
+      "loss": 0.8797,
+      "step": 13550
+    },
+    {
+      "epoch": 0.32646130873533874,
+      "grad_norm": 1.5206135511398315,
+      "learning_rate": 7.090305242875975e-07,
+      "loss": 0.9103,
+      "step": 13555
+    },
+    {
+      "epoch": 0.326581729727126,
+      "grad_norm": 1.4802076816558838,
+      "learning_rate": 7.089037622959132e-07,
+      "loss": 0.9366,
+      "step": 13560
+    },
+    {
+      "epoch": 0.32670215071891334,
+      "grad_norm": 1.6493333578109741,
+      "learning_rate": 7.087770003042287e-07,
+      "loss": 0.8865,
+      "step": 13565
+    },
+    {
+      "epoch": 0.3268225717107006,
+      "grad_norm": 1.5122278928756714,
+      "learning_rate": 7.086502383125444e-07,
+      "loss": 0.9311,
+      "step": 13570
+    },
+    {
+      "epoch": 0.3269429927024879,
+      "grad_norm": 1.5008400678634644,
+      "learning_rate": 7.0852347632086e-07,
+      "loss": 0.9301,
+      "step": 13575
+    },
+    {
+      "epoch": 0.32706341369427516,
+      "grad_norm": 1.6388431787490845,
+      "learning_rate": 7.083967143291754e-07,
+      "loss": 0.9092,
+      "step": 13580
+    },
+    {
+      "epoch": 0.3271838346860625,
+      "grad_norm": 1.8835076093673706,
+      "learning_rate": 7.082699523374911e-07,
+      "loss": 0.8953,
+      "step": 13585
+    },
+    {
+      "epoch": 0.32730425567784976,
+      "grad_norm": 1.381044864654541,
+      "learning_rate": 7.081431903458067e-07,
+      "loss": 0.9351,
+      "step": 13590
+    },
+    {
+      "epoch": 0.32742467666963704,
+      "grad_norm": 1.7384233474731445,
+      "learning_rate": 7.080164283541223e-07,
+      "loss": 0.8984,
+      "step": 13595
+    },
+    {
+      "epoch": 0.32754509766142437,
+      "grad_norm": 1.2871006727218628,
+      "learning_rate": 7.078896663624378e-07,
+      "loss": 0.8785,
+      "step": 13600
+    },
+    {
+      "epoch": 0.32766551865321164,
+      "grad_norm": 1.4323610067367554,
+      "learning_rate": 7.077629043707534e-07,
+      "loss": 0.9649,
+      "step": 13605
+    },
+    {
+      "epoch": 0.3277859396449989,
+      "grad_norm": 1.8085834980010986,
+      "learning_rate": 7.07636142379069e-07,
+      "loss": 0.956,
+      "step": 13610
+    },
+    {
+      "epoch": 0.3279063606367862,
+      "grad_norm": 1.494591236114502,
+      "learning_rate": 7.075093803873846e-07,
+      "loss": 0.8899,
+      "step": 13615
+    },
+    {
+      "epoch": 0.3280267816285735,
+      "grad_norm": 1.4227632284164429,
+      "learning_rate": 7.073826183957003e-07,
+      "loss": 0.948,
+      "step": 13620
+    },
+    {
+      "epoch": 0.3281472026203608,
+      "grad_norm": 1.6277165412902832,
+      "learning_rate": 7.072558564040157e-07,
+      "loss": 0.9471,
+      "step": 13625
+    },
+    {
+      "epoch": 0.32826762361214806,
+      "grad_norm": 1.4908299446105957,
+      "learning_rate": 7.071290944123314e-07,
+      "loss": 0.9433,
+      "step": 13630
+    },
+    {
+      "epoch": 0.32838804460393534,
+      "grad_norm": 1.476320505142212,
+      "learning_rate": 7.07002332420647e-07,
+      "loss": 0.9705,
+      "step": 13635
+    },
+    {
+      "epoch": 0.32850846559572267,
+      "grad_norm": 1.515441656112671,
+      "learning_rate": 7.068755704289625e-07,
+      "loss": 0.9863,
+      "step": 13640
+    },
+    {
+      "epoch": 0.32862888658750994,
+      "grad_norm": 1.490149736404419,
+      "learning_rate": 7.067488084372782e-07,
+      "loss": 0.911,
+      "step": 13645
+    },
+    {
+      "epoch": 0.3287493075792972,
+      "grad_norm": 1.5658856630325317,
+      "learning_rate": 7.066220464455937e-07,
+      "loss": 0.942,
+      "step": 13650
+    },
+    {
+      "epoch": 0.3288697285710845,
+      "grad_norm": 1.8892005681991577,
+      "learning_rate": 7.064952844539093e-07,
+      "loss": 0.9446,
+      "step": 13655
+    },
+    {
+      "epoch": 0.3289901495628718,
+      "grad_norm": 1.5404114723205566,
+      "learning_rate": 7.063685224622249e-07,
+      "loss": 0.926,
+      "step": 13660
+    },
+    {
+      "epoch": 0.3291105705546591,
+      "grad_norm": 1.7159450054168701,
+      "learning_rate": 7.062417604705405e-07,
+      "loss": 0.9088,
+      "step": 13665
+    },
+    {
+      "epoch": 0.32923099154644636,
+      "grad_norm": 1.5221284627914429,
+      "learning_rate": 7.06114998478856e-07,
+      "loss": 0.9194,
+      "step": 13670
+    },
+    {
+      "epoch": 0.3293514125382337,
+      "grad_norm": 1.5362894535064697,
+      "learning_rate": 7.059882364871716e-07,
+      "loss": 0.9113,
+      "step": 13675
+    },
+    {
+      "epoch": 0.32947183353002096,
+      "grad_norm": 1.454923391342163,
+      "learning_rate": 7.058614744954873e-07,
+      "loss": 0.8754,
+      "step": 13680
+    },
+    {
+      "epoch": 0.32959225452180824,
+      "grad_norm": 1.5516047477722168,
+      "learning_rate": 7.057347125038028e-07,
+      "loss": 0.914,
+      "step": 13685
+    },
+    {
+      "epoch": 0.3297126755135955,
+      "grad_norm": 1.3942458629608154,
+      "learning_rate": 7.056079505121185e-07,
+      "loss": 0.9446,
+      "step": 13690
+    },
+    {
+      "epoch": 0.32983309650538284,
+      "grad_norm": 1.4205833673477173,
+      "learning_rate": 7.05481188520434e-07,
+      "loss": 0.9282,
+      "step": 13695
+    },
+    {
+      "epoch": 0.3299535174971701,
+      "grad_norm": 1.4916503429412842,
+      "learning_rate": 7.053544265287495e-07,
+      "loss": 0.8933,
+      "step": 13700
+    },
+    {
+      "epoch": 0.3300739384889574,
+      "grad_norm": 1.4028732776641846,
+      "learning_rate": 7.052276645370652e-07,
+      "loss": 0.921,
+      "step": 13705
+    },
+    {
+      "epoch": 0.33019435948074466,
+      "grad_norm": 1.5253241062164307,
+      "learning_rate": 7.051009025453808e-07,
+      "loss": 0.942,
+      "step": 13710
+    },
+    {
+      "epoch": 0.330314780472532,
+      "grad_norm": 1.5972962379455566,
+      "learning_rate": 7.049741405536963e-07,
+      "loss": 0.9471,
+      "step": 13715
+    },
+    {
+      "epoch": 0.33043520146431926,
+      "grad_norm": 1.4718220233917236,
+      "learning_rate": 7.048473785620119e-07,
+      "loss": 0.9449,
+      "step": 13720
+    },
+    {
+      "epoch": 0.33055562245610653,
+      "grad_norm": 1.5282396078109741,
+      "learning_rate": 7.047206165703275e-07,
+      "loss": 0.9375,
+      "step": 13725
+    },
+    {
+      "epoch": 0.33067604344789386,
+      "grad_norm": 1.509559988975525,
+      "learning_rate": 7.045938545786431e-07,
+      "loss": 0.9044,
+      "step": 13730
+    },
+    {
+      "epoch": 0.33079646443968114,
+      "grad_norm": 1.661907434463501,
+      "learning_rate": 7.044670925869587e-07,
+      "loss": 0.9749,
+      "step": 13735
+    },
+    {
+      "epoch": 0.3309168854314684,
+      "grad_norm": 1.8139874935150146,
+      "learning_rate": 7.043403305952743e-07,
+      "loss": 0.8787,
+      "step": 13740
+    },
+    {
+      "epoch": 0.3310373064232557,
+      "grad_norm": 1.4727627038955688,
+      "learning_rate": 7.042135686035898e-07,
+      "loss": 0.8722,
+      "step": 13745
+    },
+    {
+      "epoch": 0.331157727415043,
+      "grad_norm": 1.3403947353363037,
+      "learning_rate": 7.040868066119054e-07,
+      "loss": 0.9231,
+      "step": 13750
+    },
+    {
+      "epoch": 0.3312781484068303,
+      "grad_norm": 1.4505927562713623,
+      "learning_rate": 7.039600446202211e-07,
+      "loss": 0.8517,
+      "step": 13755
+    },
+    {
+      "epoch": 0.33139856939861756,
+      "grad_norm": 1.5560884475708008,
+      "learning_rate": 7.038332826285366e-07,
+      "loss": 0.9006,
+      "step": 13760
+    },
+    {
+      "epoch": 0.33151899039040483,
+      "grad_norm": 1.6754164695739746,
+      "learning_rate": 7.037065206368522e-07,
+      "loss": 0.9583,
+      "step": 13765
+    },
+    {
+      "epoch": 0.33163941138219216,
+      "grad_norm": 1.5612949132919312,
+      "learning_rate": 7.035797586451678e-07,
+      "loss": 0.9091,
+      "step": 13770
+    },
+    {
+      "epoch": 0.33175983237397944,
+      "grad_norm": 1.49845290184021,
+      "learning_rate": 7.034529966534834e-07,
+      "loss": 0.9324,
+      "step": 13775
+    },
+    {
+      "epoch": 0.3318802533657667,
+      "grad_norm": 1.5018976926803589,
+      "learning_rate": 7.03326234661799e-07,
+      "loss": 0.9384,
+      "step": 13780
+    },
+    {
+      "epoch": 0.332000674357554,
+      "grad_norm": 1.5326520204544067,
+      "learning_rate": 7.031994726701145e-07,
+      "loss": 0.8921,
+      "step": 13785
+    },
+    {
+      "epoch": 0.3321210953493413,
+      "grad_norm": 1.4276926517486572,
+      "learning_rate": 7.030727106784301e-07,
+      "loss": 0.9469,
+      "step": 13790
+    },
+    {
+      "epoch": 0.3322415163411286,
+      "grad_norm": 1.6255600452423096,
+      "learning_rate": 7.029459486867457e-07,
+      "loss": 0.9313,
+      "step": 13795
+    },
+    {
+      "epoch": 0.33236193733291586,
+      "grad_norm": 1.544748067855835,
+      "learning_rate": 7.028191866950614e-07,
+      "loss": 0.9412,
+      "step": 13800
+    },
+    {
+      "epoch": 0.3324823583247032,
+      "grad_norm": 1.56434965133667,
+      "learning_rate": 7.026924247033769e-07,
+      "loss": 0.9771,
+      "step": 13805
+    },
+    {
+      "epoch": 0.33260277931649046,
+      "grad_norm": 1.5131580829620361,
+      "learning_rate": 7.025656627116924e-07,
+      "loss": 0.91,
+      "step": 13810
+    },
+    {
+      "epoch": 0.33272320030827773,
+      "grad_norm": 1.5495069026947021,
+      "learning_rate": 7.024389007200081e-07,
+      "loss": 0.921,
+      "step": 13815
+    },
+    {
+      "epoch": 0.332843621300065,
+      "grad_norm": 1.6015516519546509,
+      "learning_rate": 7.023121387283236e-07,
+      "loss": 0.8956,
+      "step": 13820
+    },
+    {
+      "epoch": 0.33296404229185234,
+      "grad_norm": 1.6354554891586304,
+      "learning_rate": 7.021853767366393e-07,
+      "loss": 0.9254,
+      "step": 13825
+    },
+    {
+      "epoch": 0.3330844632836396,
+      "grad_norm": 1.6220519542694092,
+      "learning_rate": 7.020586147449549e-07,
+      "loss": 0.9387,
+      "step": 13830
+    },
+    {
+      "epoch": 0.3332048842754269,
+      "grad_norm": 1.7443767786026,
+      "learning_rate": 7.019318527532704e-07,
+      "loss": 0.9206,
+      "step": 13835
+    },
+    {
+      "epoch": 0.33332530526721416,
+      "grad_norm": 1.4180527925491333,
+      "learning_rate": 7.01805090761586e-07,
+      "loss": 0.9414,
+      "step": 13840
+    },
+    {
+      "epoch": 0.3334457262590015,
+      "grad_norm": 1.4445128440856934,
+      "learning_rate": 7.016783287699016e-07,
+      "loss": 0.9222,
+      "step": 13845
+    },
+    {
+      "epoch": 0.33356614725078876,
+      "grad_norm": 1.522389531135559,
+      "learning_rate": 7.015515667782172e-07,
+      "loss": 0.9301,
+      "step": 13850
+    },
+    {
+      "epoch": 0.33368656824257603,
+      "grad_norm": 1.559030294418335,
+      "learning_rate": 7.014248047865327e-07,
+      "loss": 0.9337,
+      "step": 13855
+    },
+    {
+      "epoch": 0.33380698923436336,
+      "grad_norm": 1.6620439291000366,
+      "learning_rate": 7.012980427948484e-07,
+      "loss": 0.8898,
+      "step": 13860
+    },
+    {
+      "epoch": 0.33392741022615063,
+      "grad_norm": 1.3690235614776611,
+      "learning_rate": 7.011712808031639e-07,
+      "loss": 0.939,
+      "step": 13865
+    },
+    {
+      "epoch": 0.3340478312179379,
+      "grad_norm": 1.57186758518219,
+      "learning_rate": 7.010445188114795e-07,
+      "loss": 0.9139,
+      "step": 13870
+    },
+    {
+      "epoch": 0.3341682522097252,
+      "grad_norm": 1.8457096815109253,
+      "learning_rate": 7.009177568197952e-07,
+      "loss": 0.9582,
+      "step": 13875
+    },
+    {
+      "epoch": 0.3342886732015125,
+      "grad_norm": 1.5629444122314453,
+      "learning_rate": 7.007909948281106e-07,
+      "loss": 0.9404,
+      "step": 13880
+    },
+    {
+      "epoch": 0.3344090941932998,
+      "grad_norm": 1.7172342538833618,
+      "learning_rate": 7.006642328364263e-07,
+      "loss": 0.8399,
+      "step": 13885
+    },
+    {
+      "epoch": 0.33452951518508706,
+      "grad_norm": 1.6273506879806519,
+      "learning_rate": 7.005374708447419e-07,
+      "loss": 0.8748,
+      "step": 13890
+    },
+    {
+      "epoch": 0.33464993617687433,
+      "grad_norm": 1.6080079078674316,
+      "learning_rate": 7.004107088530575e-07,
+      "loss": 0.9539,
+      "step": 13895
+    },
+    {
+      "epoch": 0.33477035716866166,
+      "grad_norm": 1.5368893146514893,
+      "learning_rate": 7.00283946861373e-07,
+      "loss": 0.9064,
+      "step": 13900
+    },
+    {
+      "epoch": 0.33489077816044893,
+      "grad_norm": 1.4585624933242798,
+      "learning_rate": 7.001571848696886e-07,
+      "loss": 0.9208,
+      "step": 13905
+    },
+    {
+      "epoch": 0.3350111991522362,
+      "grad_norm": 1.532059669494629,
+      "learning_rate": 7.000304228780042e-07,
+      "loss": 0.9143,
+      "step": 13910
+    },
+    {
+      "epoch": 0.3351316201440235,
+      "grad_norm": 1.5856126546859741,
+      "learning_rate": 6.999036608863198e-07,
+      "loss": 0.9381,
+      "step": 13915
+    },
+    {
+      "epoch": 0.3352520411358108,
+      "grad_norm": 1.5512603521347046,
+      "learning_rate": 6.997768988946355e-07,
+      "loss": 0.9217,
+      "step": 13920
+    },
+    {
+      "epoch": 0.3353724621275981,
+      "grad_norm": 1.4288643598556519,
+      "learning_rate": 6.996501369029509e-07,
+      "loss": 0.881,
+      "step": 13925
+    },
+    {
+      "epoch": 0.33549288311938535,
+      "grad_norm": 1.4540021419525146,
+      "learning_rate": 6.995233749112665e-07,
+      "loss": 0.893,
+      "step": 13930
+    },
+    {
+      "epoch": 0.3356133041111727,
+      "grad_norm": 1.457465410232544,
+      "learning_rate": 6.993966129195822e-07,
+      "loss": 0.9204,
+      "step": 13935
+    },
+    {
+      "epoch": 0.33573372510295996,
+      "grad_norm": 1.5519061088562012,
+      "learning_rate": 6.992698509278977e-07,
+      "loss": 0.9038,
+      "step": 13940
+    },
+    {
+      "epoch": 0.33585414609474723,
+      "grad_norm": 1.4745047092437744,
+      "learning_rate": 6.991430889362134e-07,
+      "loss": 0.9281,
+      "step": 13945
+    },
+    {
+      "epoch": 0.3359745670865345,
+      "grad_norm": 1.4566423892974854,
+      "learning_rate": 6.990163269445289e-07,
+      "loss": 0.8641,
+      "step": 13950
+    },
+    {
+      "epoch": 0.33609498807832183,
+      "grad_norm": 1.702518105506897,
+      "learning_rate": 6.988895649528444e-07,
+      "loss": 0.9277,
+      "step": 13955
+    },
+    {
+      "epoch": 0.3362154090701091,
+      "grad_norm": 1.5833892822265625,
+      "learning_rate": 6.987628029611601e-07,
+      "loss": 0.8841,
+      "step": 13960
+    },
+    {
+      "epoch": 0.3363358300618964,
+      "grad_norm": 1.4968024492263794,
+      "learning_rate": 6.986360409694757e-07,
+      "loss": 0.9054,
+      "step": 13965
+    },
+    {
+      "epoch": 0.33645625105368365,
+      "grad_norm": 1.5771639347076416,
+      "learning_rate": 6.985092789777912e-07,
+      "loss": 0.9448,
+      "step": 13970
+    },
+    {
+      "epoch": 0.336576672045471,
+      "grad_norm": 1.413447618484497,
+      "learning_rate": 6.983825169861068e-07,
+      "loss": 0.8788,
+      "step": 13975
+    },
+    {
+      "epoch": 0.33669709303725825,
+      "grad_norm": 1.5518314838409424,
+      "learning_rate": 6.982557549944225e-07,
+      "loss": 0.8855,
+      "step": 13980
+    },
+    {
+      "epoch": 0.33681751402904553,
+      "grad_norm": 1.6350971460342407,
+      "learning_rate": 6.98128993002738e-07,
+      "loss": 0.9442,
+      "step": 13985
+    },
+    {
+      "epoch": 0.33693793502083286,
+      "grad_norm": 1.3646329641342163,
+      "learning_rate": 6.980022310110536e-07,
+      "loss": 0.912,
+      "step": 13990
+    },
+    {
+      "epoch": 0.33705835601262013,
+      "grad_norm": 1.8672513961791992,
+      "learning_rate": 6.978754690193692e-07,
+      "loss": 0.9439,
+      "step": 13995
+    },
+    {
+      "epoch": 0.3371787770044074,
+      "grad_norm": 1.4845424890518188,
+      "learning_rate": 6.977487070276847e-07,
+      "loss": 0.9131,
+      "step": 14000
+    },
+    {
+      "epoch": 0.3372991979961947,
+      "grad_norm": 1.4043781757354736,
+      "learning_rate": 6.976219450360004e-07,
+      "loss": 0.9178,
+      "step": 14005
+    },
+    {
+      "epoch": 0.337419618987982,
+      "grad_norm": 1.4757076501846313,
+      "learning_rate": 6.97495183044316e-07,
+      "loss": 0.925,
+      "step": 14010
+    },
+    {
+      "epoch": 0.3375400399797693,
+      "grad_norm": 1.5454431772232056,
+      "learning_rate": 6.973684210526314e-07,
+      "loss": 0.9657,
+      "step": 14015
+    },
+    {
+      "epoch": 0.33766046097155655,
+      "grad_norm": 1.502697229385376,
+      "learning_rate": 6.972416590609471e-07,
+      "loss": 0.9258,
+      "step": 14020
+    },
+    {
+      "epoch": 0.3377808819633438,
+      "grad_norm": 1.6740292310714722,
+      "learning_rate": 6.971148970692627e-07,
+      "loss": 0.9542,
+      "step": 14025
+    },
+    {
+      "epoch": 0.33790130295513116,
+      "grad_norm": 1.615235447883606,
+      "learning_rate": 6.969881350775783e-07,
+      "loss": 0.9191,
+      "step": 14030
+    },
+    {
+      "epoch": 0.33802172394691843,
+      "grad_norm": 1.4535822868347168,
+      "learning_rate": 6.968613730858939e-07,
+      "loss": 0.9537,
+      "step": 14035
+    },
+    {
+      "epoch": 0.3381421449387057,
+      "grad_norm": 1.4042657613754272,
+      "learning_rate": 6.967346110942095e-07,
+      "loss": 0.9199,
+      "step": 14040
+    },
+    {
+      "epoch": 0.33826256593049303,
+      "grad_norm": 1.8075425624847412,
+      "learning_rate": 6.96607849102525e-07,
+      "loss": 0.924,
+      "step": 14045
+    },
+    {
+      "epoch": 0.3383829869222803,
+      "grad_norm": 1.711000680923462,
+      "learning_rate": 6.964810871108406e-07,
+      "loss": 0.9265,
+      "step": 14050
+    },
+    {
+      "epoch": 0.3385034079140676,
+      "grad_norm": 1.5463098287582397,
+      "learning_rate": 6.963543251191563e-07,
+      "loss": 0.9224,
+      "step": 14055
+    },
+    {
+      "epoch": 0.33862382890585485,
+      "grad_norm": 1.6100355386734009,
+      "learning_rate": 6.962275631274718e-07,
+      "loss": 0.9438,
+      "step": 14060
+    },
+    {
+      "epoch": 0.3387442498976422,
+      "grad_norm": 1.7484734058380127,
+      "learning_rate": 6.961008011357874e-07,
+      "loss": 0.9099,
+      "step": 14065
+    },
+    {
+      "epoch": 0.33886467088942945,
+      "grad_norm": 1.6006505489349365,
+      "learning_rate": 6.95974039144103e-07,
+      "loss": 0.9214,
+      "step": 14070
+    },
+    {
+      "epoch": 0.3389850918812167,
+      "grad_norm": 1.4968926906585693,
+      "learning_rate": 6.958472771524185e-07,
+      "loss": 0.942,
+      "step": 14075
+    },
+    {
+      "epoch": 0.339105512873004,
+      "grad_norm": 1.4391860961914062,
+      "learning_rate": 6.957205151607342e-07,
+      "loss": 0.9237,
+      "step": 14080
+    },
+    {
+      "epoch": 0.33922593386479133,
+      "grad_norm": 1.981737732887268,
+      "learning_rate": 6.955937531690497e-07,
+      "loss": 0.9219,
+      "step": 14085
+    },
+    {
+      "epoch": 0.3393463548565786,
+      "grad_norm": 2.0144782066345215,
+      "learning_rate": 6.954669911773653e-07,
+      "loss": 0.953,
+      "step": 14090
+    },
+    {
+      "epoch": 0.3394667758483659,
+      "grad_norm": 1.4978262186050415,
+      "learning_rate": 6.953402291856809e-07,
+      "loss": 0.8834,
+      "step": 14095
+    },
+    {
+      "epoch": 0.33958719684015315,
+      "grad_norm": 1.5561473369598389,
+      "learning_rate": 6.952134671939966e-07,
+      "loss": 0.8911,
+      "step": 14100
+    },
+    {
+      "epoch": 0.3397076178319405,
+      "grad_norm": 1.5292071104049683,
+      "learning_rate": 6.950867052023122e-07,
+      "loss": 0.9516,
+      "step": 14105
+    },
+    {
+      "epoch": 0.33982803882372775,
+      "grad_norm": 1.6524728536605835,
+      "learning_rate": 6.949599432106276e-07,
+      "loss": 0.9669,
+      "step": 14110
+    },
+    {
+      "epoch": 0.339948459815515,
+      "grad_norm": 1.861893653869629,
+      "learning_rate": 6.948331812189433e-07,
+      "loss": 0.8984,
+      "step": 14115
+    },
+    {
+      "epoch": 0.34006888080730235,
+      "grad_norm": 1.8193475008010864,
+      "learning_rate": 6.947064192272589e-07,
+      "loss": 0.9429,
+      "step": 14120
+    },
+    {
+      "epoch": 0.3401893017990896,
+      "grad_norm": 1.6743884086608887,
+      "learning_rate": 6.945796572355745e-07,
+      "loss": 0.9687,
+      "step": 14125
+    },
+    {
+      "epoch": 0.3403097227908769,
+      "grad_norm": 1.641255497932434,
+      "learning_rate": 6.944528952438901e-07,
+      "loss": 0.9117,
+      "step": 14130
+    },
+    {
+      "epoch": 0.3404301437826642,
+      "grad_norm": 1.6246029138565063,
+      "learning_rate": 6.943261332522056e-07,
+      "loss": 0.9413,
+      "step": 14135
+    },
+    {
+      "epoch": 0.3405505647744515,
+      "grad_norm": 1.459546685218811,
+      "learning_rate": 6.941993712605212e-07,
+      "loss": 0.9272,
+      "step": 14140
+    },
+    {
+      "epoch": 0.3406709857662388,
+      "grad_norm": 1.6321953535079956,
+      "learning_rate": 6.940726092688368e-07,
+      "loss": 0.938,
+      "step": 14145
+    },
+    {
+      "epoch": 0.34079140675802605,
+      "grad_norm": 1.5907163619995117,
+      "learning_rate": 6.939458472771525e-07,
+      "loss": 0.9595,
+      "step": 14150
+    },
+    {
+      "epoch": 0.3409118277498133,
+      "grad_norm": 1.456050157546997,
+      "learning_rate": 6.938190852854679e-07,
+      "loss": 0.9371,
+      "step": 14155
+    },
+    {
+      "epoch": 0.34103224874160065,
+      "grad_norm": 1.67795729637146,
+      "learning_rate": 6.936923232937836e-07,
+      "loss": 0.9938,
+      "step": 14160
+    },
+    {
+      "epoch": 0.3411526697333879,
+      "grad_norm": 1.7361115217208862,
+      "learning_rate": 6.935655613020992e-07,
+      "loss": 0.9387,
+      "step": 14165
+    },
+    {
+      "epoch": 0.3412730907251752,
+      "grad_norm": 1.8264391422271729,
+      "learning_rate": 6.934387993104147e-07,
+      "loss": 0.9648,
+      "step": 14170
+    },
+    {
+      "epoch": 0.3413935117169625,
+      "grad_norm": 1.6677944660186768,
+      "learning_rate": 6.933120373187304e-07,
+      "loss": 0.9156,
+      "step": 14175
+    },
+    {
+      "epoch": 0.3415139327087498,
+      "grad_norm": 1.4905892610549927,
+      "learning_rate": 6.931852753270459e-07,
+      "loss": 0.8835,
+      "step": 14180
+    },
+    {
+      "epoch": 0.3416343537005371,
+      "grad_norm": 1.5072036981582642,
+      "learning_rate": 6.930585133353615e-07,
+      "loss": 0.9174,
+      "step": 14185
+    },
+    {
+      "epoch": 0.34175477469232435,
+      "grad_norm": 1.64654541015625,
+      "learning_rate": 6.929317513436771e-07,
+      "loss": 0.9023,
+      "step": 14190
+    },
+    {
+      "epoch": 0.3418751956841117,
+      "grad_norm": 1.5622669458389282,
+      "learning_rate": 6.928049893519927e-07,
+      "loss": 0.9345,
+      "step": 14195
+    },
+    {
+      "epoch": 0.34199561667589895,
+      "grad_norm": 1.562902808189392,
+      "learning_rate": 6.926782273603082e-07,
+      "loss": 0.8823,
+      "step": 14200
+    },
+    {
+      "epoch": 0.3421160376676862,
+      "grad_norm": 1.4731025695800781,
+      "learning_rate": 6.925514653686238e-07,
+      "loss": 0.915,
+      "step": 14205
+    },
+    {
+      "epoch": 0.3422364586594735,
+      "grad_norm": 1.7186492681503296,
+      "learning_rate": 6.924247033769395e-07,
+      "loss": 0.9277,
+      "step": 14210
+    },
+    {
+      "epoch": 0.3423568796512608,
+      "grad_norm": 1.7451112270355225,
+      "learning_rate": 6.92297941385255e-07,
+      "loss": 0.9411,
+      "step": 14215
+    },
+    {
+      "epoch": 0.3424773006430481,
+      "grad_norm": 1.4915034770965576,
+      "learning_rate": 6.921711793935707e-07,
+      "loss": 0.9142,
+      "step": 14220
+    },
+    {
+      "epoch": 0.3425977216348354,
+      "grad_norm": 1.48759925365448,
+      "learning_rate": 6.920444174018862e-07,
+      "loss": 0.9394,
+      "step": 14225
+    },
+    {
+      "epoch": 0.34271814262662265,
+      "grad_norm": 1.5974241495132446,
+      "learning_rate": 6.919176554102017e-07,
+      "loss": 0.8681,
+      "step": 14230
+    },
+    {
+      "epoch": 0.34283856361841,
+      "grad_norm": 1.517372965812683,
+      "learning_rate": 6.917908934185174e-07,
+      "loss": 0.8845,
+      "step": 14235
+    },
+    {
+      "epoch": 0.34295898461019725,
+      "grad_norm": 1.6111204624176025,
+      "learning_rate": 6.91664131426833e-07,
+      "loss": 0.9087,
+      "step": 14240
+    },
+    {
+      "epoch": 0.3430794056019845,
+      "grad_norm": 1.4942491054534912,
+      "learning_rate": 6.915373694351486e-07,
+      "loss": 0.9885,
+      "step": 14245
+    },
+    {
+      "epoch": 0.34319982659377185,
+      "grad_norm": 1.5210905075073242,
+      "learning_rate": 6.914106074434641e-07,
+      "loss": 0.9222,
+      "step": 14250
+    },
+    {
+      "epoch": 0.3433202475855591,
+      "grad_norm": 1.6447272300720215,
+      "learning_rate": 6.912838454517797e-07,
+      "loss": 0.9016,
+      "step": 14255
+    },
+    {
+      "epoch": 0.3434406685773464,
+      "grad_norm": 1.4414374828338623,
+      "learning_rate": 6.911570834600953e-07,
+      "loss": 0.9526,
+      "step": 14260
+    },
+    {
+      "epoch": 0.34356108956913367,
+      "grad_norm": 1.6271892786026,
+      "learning_rate": 6.910303214684109e-07,
+      "loss": 0.9561,
+      "step": 14265
+    },
+    {
+      "epoch": 0.343681510560921,
+      "grad_norm": 1.5519522428512573,
+      "learning_rate": 6.909035594767265e-07,
+      "loss": 0.8733,
+      "step": 14270
+    },
+    {
+      "epoch": 0.3438019315527083,
+      "grad_norm": 1.8349835872650146,
+      "learning_rate": 6.90776797485042e-07,
+      "loss": 0.8791,
+      "step": 14275
+    },
+    {
+      "epoch": 0.34392235254449555,
+      "grad_norm": 1.4303919076919556,
+      "learning_rate": 6.906500354933577e-07,
+      "loss": 0.9214,
+      "step": 14280
+    },
+    {
+      "epoch": 0.3440427735362828,
+      "grad_norm": 1.4979641437530518,
+      "learning_rate": 6.905232735016733e-07,
+      "loss": 0.8764,
+      "step": 14285
+    },
+    {
+      "epoch": 0.34416319452807015,
+      "grad_norm": 1.6719623804092407,
+      "learning_rate": 6.903965115099888e-07,
+      "loss": 0.93,
+      "step": 14290
+    },
+    {
+      "epoch": 0.3442836155198574,
+      "grad_norm": 1.5701115131378174,
+      "learning_rate": 6.902697495183044e-07,
+      "loss": 0.8932,
+      "step": 14295
+    },
+    {
+      "epoch": 0.3444040365116447,
+      "grad_norm": 1.6487075090408325,
+      "learning_rate": 6.9014298752662e-07,
+      "loss": 0.9413,
+      "step": 14300
+    },
+    {
+      "epoch": 0.344524457503432,
+      "grad_norm": 1.575084924697876,
+      "learning_rate": 6.900162255349356e-07,
+      "loss": 0.9203,
+      "step": 14305
+    },
+    {
+      "epoch": 0.3446448784952193,
+      "grad_norm": 1.462643027305603,
+      "learning_rate": 6.898894635432512e-07,
+      "loss": 0.9005,
+      "step": 14310
+    },
+    {
+      "epoch": 0.34476529948700657,
+      "grad_norm": 1.448420763015747,
+      "learning_rate": 6.897627015515668e-07,
+      "loss": 0.9337,
+      "step": 14315
+    },
+    {
+      "epoch": 0.34488572047879384,
+      "grad_norm": 1.6062064170837402,
+      "learning_rate": 6.896359395598823e-07,
+      "loss": 0.9557,
+      "step": 14320
+    },
+    {
+      "epoch": 0.3450061414705812,
+      "grad_norm": 1.5350664854049683,
+      "learning_rate": 6.895091775681979e-07,
+      "loss": 0.9215,
+      "step": 14325
+    },
+    {
+      "epoch": 0.34512656246236845,
+      "grad_norm": 1.4015288352966309,
+      "learning_rate": 6.893824155765136e-07,
+      "loss": 0.9488,
+      "step": 14330
+    },
+    {
+      "epoch": 0.3452469834541557,
+      "grad_norm": 1.4283367395401,
+      "learning_rate": 6.892556535848291e-07,
+      "loss": 0.9227,
+      "step": 14335
+    },
+    {
+      "epoch": 0.345367404445943,
+      "grad_norm": 1.5763317346572876,
+      "learning_rate": 6.891288915931446e-07,
+      "loss": 0.9458,
+      "step": 14340
+    },
+    {
+      "epoch": 0.3454878254377303,
+      "grad_norm": 1.4506820440292358,
+      "learning_rate": 6.890021296014603e-07,
+      "loss": 0.9104,
+      "step": 14345
+    },
+    {
+      "epoch": 0.3456082464295176,
+      "grad_norm": 1.586639165878296,
+      "learning_rate": 6.888753676097758e-07,
+      "loss": 0.9296,
+      "step": 14350
+    },
+    {
+      "epoch": 0.34572866742130487,
+      "grad_norm": 1.5703184604644775,
+      "learning_rate": 6.887486056180915e-07,
+      "loss": 0.9198,
+      "step": 14355
+    },
+    {
+      "epoch": 0.34584908841309214,
+      "grad_norm": 1.518678903579712,
+      "learning_rate": 6.886218436264071e-07,
+      "loss": 0.9037,
+      "step": 14360
+    },
+    {
+      "epoch": 0.34596950940487947,
+      "grad_norm": 1.5263680219650269,
+      "learning_rate": 6.884950816347226e-07,
+      "loss": 0.9364,
+      "step": 14365
+    },
+    {
+      "epoch": 0.34608993039666675,
+      "grad_norm": 1.6340490579605103,
+      "learning_rate": 6.883683196430382e-07,
+      "loss": 0.9527,
+      "step": 14370
+    },
+    {
+      "epoch": 0.346210351388454,
+      "grad_norm": 1.4462147951126099,
+      "learning_rate": 6.882415576513538e-07,
+      "loss": 0.8904,
+      "step": 14375
+    },
+    {
+      "epoch": 0.34633077238024135,
+      "grad_norm": 1.6212999820709229,
+      "learning_rate": 6.881147956596694e-07,
+      "loss": 0.942,
+      "step": 14380
+    },
+    {
+      "epoch": 0.3464511933720286,
+      "grad_norm": 1.5607647895812988,
+      "learning_rate": 6.879880336679849e-07,
+      "loss": 0.9361,
+      "step": 14385
+    },
+    {
+      "epoch": 0.3465716143638159,
+      "grad_norm": 1.485691785812378,
+      "learning_rate": 6.878612716763006e-07,
+      "loss": 0.896,
+      "step": 14390
+    },
+    {
+      "epoch": 0.34669203535560317,
+      "grad_norm": 1.5243165493011475,
+      "learning_rate": 6.877345096846161e-07,
+      "loss": 0.9551,
+      "step": 14395
+    },
+    {
+      "epoch": 0.3468124563473905,
+      "grad_norm": 1.5864940881729126,
+      "learning_rate": 6.876077476929317e-07,
+      "loss": 0.8871,
+      "step": 14400
+    },
+    {
+      "epoch": 0.34693287733917777,
+      "grad_norm": 1.4279605150222778,
+      "learning_rate": 6.874809857012474e-07,
+      "loss": 0.9501,
+      "step": 14405
+    },
+    {
+      "epoch": 0.34705329833096504,
+      "grad_norm": 1.622212290763855,
+      "learning_rate": 6.873542237095628e-07,
+      "loss": 0.8701,
+      "step": 14410
+    },
+    {
+      "epoch": 0.3471737193227523,
+      "grad_norm": 1.7826794385910034,
+      "learning_rate": 6.872274617178785e-07,
+      "loss": 0.9761,
+      "step": 14415
+    },
+    {
+      "epoch": 0.34729414031453965,
+      "grad_norm": 1.470369577407837,
+      "learning_rate": 6.871006997261941e-07,
+      "loss": 0.8671,
+      "step": 14420
+    },
+    {
+      "epoch": 0.3474145613063269,
+      "grad_norm": 1.6227402687072754,
+      "learning_rate": 6.869739377345097e-07,
+      "loss": 0.9453,
+      "step": 14425
+    },
+    {
+      "epoch": 0.3475349822981142,
+      "grad_norm": 1.3772656917572021,
+      "learning_rate": 6.868471757428253e-07,
+      "loss": 0.9403,
+      "step": 14430
+    },
+    {
+      "epoch": 0.3476554032899015,
+      "grad_norm": 1.508234977722168,
+      "learning_rate": 6.867204137511408e-07,
+      "loss": 0.8758,
+      "step": 14435
+    },
+    {
+      "epoch": 0.3477758242816888,
+      "grad_norm": 1.557151198387146,
+      "learning_rate": 6.865936517594564e-07,
+      "loss": 0.944,
+      "step": 14440
+    },
+    {
+      "epoch": 0.34789624527347607,
+      "grad_norm": 1.6642489433288574,
+      "learning_rate": 6.86466889767772e-07,
+      "loss": 0.9506,
+      "step": 14445
+    },
+    {
+      "epoch": 0.34801666626526334,
+      "grad_norm": 1.665263056755066,
+      "learning_rate": 6.863401277760877e-07,
+      "loss": 0.9669,
+      "step": 14450
+    },
+    {
+      "epoch": 0.34813708725705067,
+      "grad_norm": 1.4337249994277954,
+      "learning_rate": 6.862133657844031e-07,
+      "loss": 0.9121,
+      "step": 14455
+    },
+    {
+      "epoch": 0.34825750824883794,
+      "grad_norm": 1.5387197732925415,
+      "learning_rate": 6.860866037927187e-07,
+      "loss": 0.918,
+      "step": 14460
+    },
+    {
+      "epoch": 0.3483779292406252,
+      "grad_norm": 1.3670315742492676,
+      "learning_rate": 6.859598418010344e-07,
+      "loss": 0.9419,
+      "step": 14465
+    },
+    {
+      "epoch": 0.3484983502324125,
+      "grad_norm": 1.4968301057815552,
+      "learning_rate": 6.858330798093499e-07,
+      "loss": 0.9089,
+      "step": 14470
+    },
+    {
+      "epoch": 0.3486187712241998,
+      "grad_norm": 1.4888393878936768,
+      "learning_rate": 6.857063178176656e-07,
+      "loss": 0.9157,
+      "step": 14475
+    },
+    {
+      "epoch": 0.3487391922159871,
+      "grad_norm": 1.4342647790908813,
+      "learning_rate": 6.855795558259811e-07,
+      "loss": 0.9848,
+      "step": 14480
+    },
+    {
+      "epoch": 0.34885961320777437,
+      "grad_norm": 1.4552973508834839,
+      "learning_rate": 6.854527938342966e-07,
+      "loss": 0.9555,
+      "step": 14485
+    },
+    {
+      "epoch": 0.3489800341995617,
+      "grad_norm": 1.503014087677002,
+      "learning_rate": 6.853260318426123e-07,
+      "loss": 0.9381,
+      "step": 14490
+    },
+    {
+      "epoch": 0.34910045519134897,
+      "grad_norm": 1.4178707599639893,
+      "learning_rate": 6.851992698509279e-07,
+      "loss": 0.9311,
+      "step": 14495
+    },
+    {
+      "epoch": 0.34922087618313624,
+      "grad_norm": 1.550925612449646,
+      "learning_rate": 6.850725078592434e-07,
+      "loss": 0.9277,
+      "step": 14500
+    },
+    {
+      "epoch": 0.3493412971749235,
+      "grad_norm": 1.6314831972122192,
+      "learning_rate": 6.84945745867559e-07,
+      "loss": 0.9252,
+      "step": 14505
+    },
+    {
+      "epoch": 0.34946171816671084,
+      "grad_norm": 1.5699819326400757,
+      "learning_rate": 6.848189838758747e-07,
+      "loss": 0.9059,
+      "step": 14510
+    },
+    {
+      "epoch": 0.3495821391584981,
+      "grad_norm": 1.4622288942337036,
+      "learning_rate": 6.846922218841902e-07,
+      "loss": 0.9703,
+      "step": 14515
+    },
+    {
+      "epoch": 0.3497025601502854,
+      "grad_norm": 1.5557868480682373,
+      "learning_rate": 6.845654598925058e-07,
+      "loss": 0.8926,
+      "step": 14520
+    },
+    {
+      "epoch": 0.34982298114207266,
+      "grad_norm": 1.5330698490142822,
+      "learning_rate": 6.844386979008214e-07,
+      "loss": 0.904,
+      "step": 14525
+    },
+    {
+      "epoch": 0.34994340213386,
+      "grad_norm": 1.5403659343719482,
+      "learning_rate": 6.843119359091369e-07,
+      "loss": 0.9054,
+      "step": 14530
+    },
+    {
+      "epoch": 0.35006382312564727,
+      "grad_norm": 1.4800559282302856,
+      "learning_rate": 6.841851739174526e-07,
+      "loss": 0.9392,
+      "step": 14535
+    },
+    {
+      "epoch": 0.35018424411743454,
+      "grad_norm": 1.378165602684021,
+      "learning_rate": 6.840584119257682e-07,
+      "loss": 0.9487,
+      "step": 14540
+    },
+    {
+      "epoch": 0.3503046651092218,
+      "grad_norm": 1.4545567035675049,
+      "learning_rate": 6.839316499340838e-07,
+      "loss": 0.9389,
+      "step": 14545
+    },
+    {
+      "epoch": 0.35042508610100914,
+      "grad_norm": 1.4694862365722656,
+      "learning_rate": 6.838048879423993e-07,
+      "loss": 0.9267,
+      "step": 14550
+    },
+    {
+      "epoch": 0.3505455070927964,
+      "grad_norm": 1.4605337381362915,
+      "learning_rate": 6.836781259507149e-07,
+      "loss": 0.8976,
+      "step": 14555
+    },
+    {
+      "epoch": 0.3506659280845837,
+      "grad_norm": 1.4146809577941895,
+      "learning_rate": 6.835513639590305e-07,
+      "loss": 0.938,
+      "step": 14560
+    },
+    {
+      "epoch": 0.350786349076371,
+      "grad_norm": 1.5950515270233154,
+      "learning_rate": 6.834246019673461e-07,
+      "loss": 0.9015,
+      "step": 14565
+    },
+    {
+      "epoch": 0.3509067700681583,
+      "grad_norm": 1.5792133808135986,
+      "learning_rate": 6.832978399756617e-07,
+      "loss": 0.8922,
+      "step": 14570
+    },
+    {
+      "epoch": 0.35102719105994556,
+      "grad_norm": 1.5362197160720825,
+      "learning_rate": 6.831710779839772e-07,
+      "loss": 0.9435,
+      "step": 14575
+    },
+    {
+      "epoch": 0.35114761205173284,
+      "grad_norm": 1.4740729331970215,
+      "learning_rate": 6.830443159922928e-07,
+      "loss": 0.926,
+      "step": 14580
+    },
+    {
+      "epoch": 0.35126803304352017,
+      "grad_norm": 1.5383095741271973,
+      "learning_rate": 6.829175540006085e-07,
+      "loss": 0.8675,
+      "step": 14585
+    },
+    {
+      "epoch": 0.35138845403530744,
+      "grad_norm": 1.6412146091461182,
+      "learning_rate": 6.82790792008924e-07,
+      "loss": 0.9271,
+      "step": 14590
+    },
+    {
+      "epoch": 0.3515088750270947,
+      "grad_norm": 1.4079954624176025,
+      "learning_rate": 6.826640300172396e-07,
+      "loss": 0.9501,
+      "step": 14595
+    },
+    {
+      "epoch": 0.351629296018882,
+      "grad_norm": 1.6104514598846436,
+      "learning_rate": 6.825372680255552e-07,
+      "loss": 0.9345,
+      "step": 14600
+    },
+    {
+      "epoch": 0.3517497170106693,
+      "grad_norm": 1.4540212154388428,
+      "learning_rate": 6.824105060338707e-07,
+      "loss": 0.9067,
+      "step": 14605
+    },
+    {
+      "epoch": 0.3518701380024566,
+      "grad_norm": 1.5886861085891724,
+      "learning_rate": 6.822837440421864e-07,
+      "loss": 0.9966,
+      "step": 14610
+    },
+    {
+      "epoch": 0.35199055899424386,
+      "grad_norm": 1.532599687576294,
+      "learning_rate": 6.82156982050502e-07,
+      "loss": 0.9118,
+      "step": 14615
+    },
+    {
+      "epoch": 0.3521109799860312,
+      "grad_norm": 1.4757113456726074,
+      "learning_rate": 6.820302200588175e-07,
+      "loss": 0.8775,
+      "step": 14620
+    },
+    {
+      "epoch": 0.35223140097781847,
+      "grad_norm": 1.3537869453430176,
+      "learning_rate": 6.819034580671331e-07,
+      "loss": 0.938,
+      "step": 14625
+    },
+    {
+      "epoch": 0.35235182196960574,
+      "grad_norm": 1.5233502388000488,
+      "learning_rate": 6.817766960754488e-07,
+      "loss": 0.9157,
+      "step": 14630
+    },
+    {
+      "epoch": 0.352472242961393,
+      "grad_norm": 1.6560343503952026,
+      "learning_rate": 6.816499340837643e-07,
+      "loss": 0.9447,
+      "step": 14635
+    },
+    {
+      "epoch": 0.35259266395318034,
+      "grad_norm": 1.452549695968628,
+      "learning_rate": 6.815231720920798e-07,
+      "loss": 0.8794,
+      "step": 14640
+    },
+    {
+      "epoch": 0.3527130849449676,
+      "grad_norm": 1.5324472188949585,
+      "learning_rate": 6.813964101003955e-07,
+      "loss": 0.9094,
+      "step": 14645
+    },
+    {
+      "epoch": 0.3528335059367549,
+      "grad_norm": 1.5266990661621094,
+      "learning_rate": 6.81269648108711e-07,
+      "loss": 0.9036,
+      "step": 14650
+    },
+    {
+      "epoch": 0.35295392692854216,
+      "grad_norm": 1.6903965473175049,
+      "learning_rate": 6.811428861170267e-07,
+      "loss": 0.9105,
+      "step": 14655
+    },
+    {
+      "epoch": 0.3530743479203295,
+      "grad_norm": 1.6886303424835205,
+      "learning_rate": 6.810161241253423e-07,
+      "loss": 0.9129,
+      "step": 14660
+    },
+    {
+      "epoch": 0.35319476891211676,
+      "grad_norm": 1.5531803369522095,
+      "learning_rate": 6.808893621336577e-07,
+      "loss": 0.9767,
+      "step": 14665
+    },
+    {
+      "epoch": 0.35331518990390404,
+      "grad_norm": 1.5922292470932007,
+      "learning_rate": 6.807626001419734e-07,
+      "loss": 0.9539,
+      "step": 14670
+    },
+    {
+      "epoch": 0.3534356108956913,
+      "grad_norm": 1.556394100189209,
+      "learning_rate": 6.80635838150289e-07,
+      "loss": 0.8524,
+      "step": 14675
+    },
+    {
+      "epoch": 0.35355603188747864,
+      "grad_norm": 1.4371875524520874,
+      "learning_rate": 6.805090761586046e-07,
+      "loss": 0.895,
+      "step": 14680
+    },
+    {
+      "epoch": 0.3536764528792659,
+      "grad_norm": 1.4753456115722656,
+      "learning_rate": 6.803823141669201e-07,
+      "loss": 0.9195,
+      "step": 14685
+    },
+    {
+      "epoch": 0.3537968738710532,
+      "grad_norm": 1.5675758123397827,
+      "learning_rate": 6.802555521752358e-07,
+      "loss": 0.8769,
+      "step": 14690
+    },
+    {
+      "epoch": 0.3539172948628405,
+      "grad_norm": 1.6256442070007324,
+      "learning_rate": 6.801287901835513e-07,
+      "loss": 0.8808,
+      "step": 14695
+    },
+    {
+      "epoch": 0.3540377158546278,
+      "grad_norm": 1.9164952039718628,
+      "learning_rate": 6.800020281918669e-07,
+      "loss": 0.9165,
+      "step": 14700
+    },
+    {
+      "epoch": 0.35415813684641506,
+      "grad_norm": 1.5501030683517456,
+      "learning_rate": 6.798752662001826e-07,
+      "loss": 0.8797,
+      "step": 14705
+    },
+    {
+      "epoch": 0.35427855783820233,
+      "grad_norm": 1.5340889692306519,
+      "learning_rate": 6.79748504208498e-07,
+      "loss": 0.9298,
+      "step": 14710
+    },
+    {
+      "epoch": 0.35439897882998966,
+      "grad_norm": 1.3141673803329468,
+      "learning_rate": 6.796217422168137e-07,
+      "loss": 0.9389,
+      "step": 14715
+    },
+    {
+      "epoch": 0.35451939982177694,
+      "grad_norm": 1.4161008596420288,
+      "learning_rate": 6.794949802251293e-07,
+      "loss": 0.9145,
+      "step": 14720
+    },
+    {
+      "epoch": 0.3546398208135642,
+      "grad_norm": 1.6499849557876587,
+      "learning_rate": 6.793682182334448e-07,
+      "loss": 0.9219,
+      "step": 14725
+    },
+    {
+      "epoch": 0.3547602418053515,
+      "grad_norm": 1.5385875701904297,
+      "learning_rate": 6.792414562417605e-07,
+      "loss": 0.9265,
+      "step": 14730
+    },
+    {
+      "epoch": 0.3548806627971388,
+      "grad_norm": 1.5018668174743652,
+      "learning_rate": 6.79114694250076e-07,
+      "loss": 1.0083,
+      "step": 14735
+    },
+    {
+      "epoch": 0.3550010837889261,
+      "grad_norm": 1.418845534324646,
+      "learning_rate": 6.789879322583916e-07,
+      "loss": 0.9134,
+      "step": 14740
+    },
+    {
+      "epoch": 0.35512150478071336,
+      "grad_norm": 1.488181471824646,
+      "learning_rate": 6.788611702667072e-07,
+      "loss": 0.9387,
+      "step": 14745
+    },
+    {
+      "epoch": 0.3552419257725007,
+      "grad_norm": 1.5587797164916992,
+      "learning_rate": 6.787344082750229e-07,
+      "loss": 0.9402,
+      "step": 14750
+    },
+    {
+      "epoch": 0.35536234676428796,
+      "grad_norm": 1.431835651397705,
+      "learning_rate": 6.786076462833383e-07,
+      "loss": 0.9412,
+      "step": 14755
+    },
+    {
+      "epoch": 0.35548276775607524,
+      "grad_norm": 1.7869625091552734,
+      "learning_rate": 6.784808842916539e-07,
+      "loss": 0.9299,
+      "step": 14760
+    },
+    {
+      "epoch": 0.3556031887478625,
+      "grad_norm": 1.3361440896987915,
+      "learning_rate": 6.783541222999696e-07,
+      "loss": 0.9497,
+      "step": 14765
+    },
+    {
+      "epoch": 0.35572360973964984,
+      "grad_norm": 1.3490222692489624,
+      "learning_rate": 6.782273603082851e-07,
+      "loss": 0.8876,
+      "step": 14770
+    },
+    {
+      "epoch": 0.3558440307314371,
+      "grad_norm": 1.5653305053710938,
+      "learning_rate": 6.781005983166008e-07,
+      "loss": 0.9391,
+      "step": 14775
+    },
+    {
+      "epoch": 0.3559644517232244,
+      "grad_norm": 1.583202600479126,
+      "learning_rate": 6.779738363249163e-07,
+      "loss": 0.9037,
+      "step": 14780
+    },
+    {
+      "epoch": 0.35608487271501166,
+      "grad_norm": 1.6768081188201904,
+      "learning_rate": 6.778470743332318e-07,
+      "loss": 0.943,
+      "step": 14785
+    },
+    {
+      "epoch": 0.356205293706799,
+      "grad_norm": 1.8261566162109375,
+      "learning_rate": 6.777203123415475e-07,
+      "loss": 0.9512,
+      "step": 14790
+    },
+    {
+      "epoch": 0.35632571469858626,
+      "grad_norm": 1.6274113655090332,
+      "learning_rate": 6.775935503498631e-07,
+      "loss": 0.9726,
+      "step": 14795
+    },
+    {
+      "epoch": 0.35644613569037353,
+      "grad_norm": 1.8121517896652222,
+      "learning_rate": 6.774667883581786e-07,
+      "loss": 0.8796,
+      "step": 14800
+    },
+    {
+      "epoch": 0.3565665566821608,
+      "grad_norm": 1.510184407234192,
+      "learning_rate": 6.773400263664942e-07,
+      "loss": 0.874,
+      "step": 14805
+    },
+    {
+      "epoch": 0.35668697767394814,
+      "grad_norm": 1.5640634298324585,
+      "learning_rate": 6.772132643748099e-07,
+      "loss": 0.9704,
+      "step": 14810
+    },
+    {
+      "epoch": 0.3568073986657354,
+      "grad_norm": 1.549045443534851,
+      "learning_rate": 6.770865023831254e-07,
+      "loss": 0.9432,
+      "step": 14815
+    },
+    {
+      "epoch": 0.3569278196575227,
+      "grad_norm": 1.5475493669509888,
+      "learning_rate": 6.76959740391441e-07,
+      "loss": 0.8924,
+      "step": 14820
+    },
+    {
+      "epoch": 0.35704824064931,
+      "grad_norm": 1.4325554370880127,
+      "learning_rate": 6.768329783997566e-07,
+      "loss": 0.8884,
+      "step": 14825
+    },
+    {
+      "epoch": 0.3571686616410973,
+      "grad_norm": 1.6210287809371948,
+      "learning_rate": 6.767062164080721e-07,
+      "loss": 0.9236,
+      "step": 14830
+    },
+    {
+      "epoch": 0.35728908263288456,
+      "grad_norm": 1.5211275815963745,
+      "learning_rate": 6.765794544163878e-07,
+      "loss": 0.9077,
+      "step": 14835
+    },
+    {
+      "epoch": 0.35740950362467183,
+      "grad_norm": 1.7513643503189087,
+      "learning_rate": 6.764526924247034e-07,
+      "loss": 0.8607,
+      "step": 14840
+    },
+    {
+      "epoch": 0.35752992461645916,
+      "grad_norm": 1.4754881858825684,
+      "learning_rate": 6.763259304330188e-07,
+      "loss": 0.9212,
+      "step": 14845
+    },
+    {
+      "epoch": 0.35765034560824643,
+      "grad_norm": 1.642807126045227,
+      "learning_rate": 6.761991684413345e-07,
+      "loss": 0.8888,
+      "step": 14850
+    },
+    {
+      "epoch": 0.3577707666000337,
+      "grad_norm": 1.4399112462997437,
+      "learning_rate": 6.760724064496501e-07,
+      "loss": 0.891,
+      "step": 14855
+    },
+    {
+      "epoch": 0.357891187591821,
+      "grad_norm": 1.6988621950149536,
+      "learning_rate": 6.759456444579657e-07,
+      "loss": 0.9361,
+      "step": 14860
+    },
+    {
+      "epoch": 0.3580116085836083,
+      "grad_norm": 1.4203370809555054,
+      "learning_rate": 6.758188824662813e-07,
+      "loss": 0.9133,
+      "step": 14865
+    },
+    {
+      "epoch": 0.3581320295753956,
+      "grad_norm": 1.418886661529541,
+      "learning_rate": 6.756921204745968e-07,
+      "loss": 0.9059,
+      "step": 14870
+    },
+    {
+      "epoch": 0.35825245056718286,
+      "grad_norm": 1.404044508934021,
+      "learning_rate": 6.755653584829124e-07,
+      "loss": 0.9271,
+      "step": 14875
+    },
+    {
+      "epoch": 0.3583728715589702,
+      "grad_norm": 1.5734264850616455,
+      "learning_rate": 6.75438596491228e-07,
+      "loss": 0.8857,
+      "step": 14880
+    },
+    {
+      "epoch": 0.35849329255075746,
+      "grad_norm": 1.4540066719055176,
+      "learning_rate": 6.753118344995437e-07,
+      "loss": 0.8624,
+      "step": 14885
+    },
+    {
+      "epoch": 0.35861371354254473,
+      "grad_norm": 1.578904390335083,
+      "learning_rate": 6.751850725078592e-07,
+      "loss": 0.9649,
+      "step": 14890
+    },
+    {
+      "epoch": 0.358734134534332,
+      "grad_norm": 1.5617300271987915,
+      "learning_rate": 6.750583105161748e-07,
+      "loss": 0.9249,
+      "step": 14895
+    },
+    {
+      "epoch": 0.35885455552611933,
+      "grad_norm": 1.649563193321228,
+      "learning_rate": 6.749315485244904e-07,
+      "loss": 0.9314,
+      "step": 14900
+    },
+    {
+      "epoch": 0.3589749765179066,
+      "grad_norm": 1.5435034036636353,
+      "learning_rate": 6.748047865328059e-07,
+      "loss": 0.9417,
+      "step": 14905
+    },
+    {
+      "epoch": 0.3590953975096939,
+      "grad_norm": 1.6014434099197388,
+      "learning_rate": 6.746780245411216e-07,
+      "loss": 0.9236,
+      "step": 14910
+    },
+    {
+      "epoch": 0.35921581850148115,
+      "grad_norm": 1.627638339996338,
+      "learning_rate": 6.745512625494371e-07,
+      "loss": 0.8733,
+      "step": 14915
+    },
+    {
+      "epoch": 0.3593362394932685,
+      "grad_norm": 1.7317113876342773,
+      "learning_rate": 6.744245005577527e-07,
+      "loss": 0.986,
+      "step": 14920
+    },
+    {
+      "epoch": 0.35945666048505576,
+      "grad_norm": 1.584110975265503,
+      "learning_rate": 6.742977385660683e-07,
+      "loss": 0.8878,
+      "step": 14925
+    },
+    {
+      "epoch": 0.35957708147684303,
+      "grad_norm": 1.5106953382492065,
+      "learning_rate": 6.74170976574384e-07,
+      "loss": 0.9117,
+      "step": 14930
+    },
+    {
+      "epoch": 0.35969750246863036,
+      "grad_norm": 1.4878908395767212,
+      "learning_rate": 6.740442145826995e-07,
+      "loss": 0.8723,
+      "step": 14935
+    },
+    {
+      "epoch": 0.35981792346041763,
+      "grad_norm": 1.6115652322769165,
+      "learning_rate": 6.73917452591015e-07,
+      "loss": 0.9179,
+      "step": 14940
+    },
+    {
+      "epoch": 0.3599383444522049,
+      "grad_norm": 1.5578006505966187,
+      "learning_rate": 6.737906905993307e-07,
+      "loss": 0.9446,
+      "step": 14945
+    },
+    {
+      "epoch": 0.3600587654439922,
+      "grad_norm": 1.6315845251083374,
+      "learning_rate": 6.736639286076462e-07,
+      "loss": 0.9004,
+      "step": 14950
+    },
+    {
+      "epoch": 0.3601791864357795,
+      "grad_norm": 1.4576815366744995,
+      "learning_rate": 6.735371666159619e-07,
+      "loss": 0.9585,
+      "step": 14955
+    },
+    {
+      "epoch": 0.3602996074275668,
+      "grad_norm": 1.4762134552001953,
+      "learning_rate": 6.734104046242775e-07,
+      "loss": 0.9007,
+      "step": 14960
+    },
+    {
+      "epoch": 0.36042002841935405,
+      "grad_norm": 1.4411206245422363,
+      "learning_rate": 6.732836426325929e-07,
+      "loss": 0.8957,
+      "step": 14965
+    },
+    {
+      "epoch": 0.36054044941114133,
+      "grad_norm": 1.505435824394226,
+      "learning_rate": 6.731568806409086e-07,
+      "loss": 0.9294,
+      "step": 14970
+    },
+    {
+      "epoch": 0.36066087040292866,
+      "grad_norm": 1.5263447761535645,
+      "learning_rate": 6.730301186492242e-07,
+      "loss": 0.9304,
+      "step": 14975
+    },
+    {
+      "epoch": 0.36078129139471593,
+      "grad_norm": 1.568605661392212,
+      "learning_rate": 6.729033566575398e-07,
+      "loss": 0.9157,
+      "step": 14980
+    },
+    {
+      "epoch": 0.3609017123865032,
+      "grad_norm": 1.693253517150879,
+      "learning_rate": 6.727765946658553e-07,
+      "loss": 0.9702,
+      "step": 14985
+    },
+    {
+      "epoch": 0.3610221333782905,
+      "grad_norm": 1.3735846281051636,
+      "learning_rate": 6.726498326741709e-07,
+      "loss": 0.8727,
+      "step": 14990
+    },
+    {
+      "epoch": 0.3611425543700778,
+      "grad_norm": 1.50700843334198,
+      "learning_rate": 6.725230706824865e-07,
+      "loss": 0.9141,
+      "step": 14995
+    },
+    {
+      "epoch": 0.3612629753618651,
+      "grad_norm": 1.4231067895889282,
+      "learning_rate": 6.723963086908021e-07,
+      "loss": 0.9359,
+      "step": 15000
+    },
+    {
+      "epoch": 0.36138339635365235,
+      "grad_norm": 1.4903537034988403,
+      "learning_rate": 6.722695466991178e-07,
+      "loss": 0.954,
+      "step": 15005
+    },
+    {
+      "epoch": 0.3615038173454397,
+      "grad_norm": 1.5555487871170044,
+      "learning_rate": 6.721427847074332e-07,
+      "loss": 0.8628,
+      "step": 15010
+    },
+    {
+      "epoch": 0.36162423833722696,
+      "grad_norm": 1.661105751991272,
+      "learning_rate": 6.720160227157488e-07,
+      "loss": 0.9473,
+      "step": 15015
+    },
+    {
+      "epoch": 0.36174465932901423,
+      "grad_norm": 1.5630122423171997,
+      "learning_rate": 6.718892607240645e-07,
+      "loss": 0.9535,
+      "step": 15020
+    },
+    {
+      "epoch": 0.3618650803208015,
+      "grad_norm": 1.8373959064483643,
+      "learning_rate": 6.7176249873238e-07,
+      "loss": 0.9357,
+      "step": 15025
+    },
+    {
+      "epoch": 0.36198550131258883,
+      "grad_norm": 1.462262749671936,
+      "learning_rate": 6.716357367406956e-07,
+      "loss": 0.8968,
+      "step": 15030
+    },
+    {
+      "epoch": 0.3621059223043761,
+      "grad_norm": 1.5995755195617676,
+      "learning_rate": 6.715089747490112e-07,
+      "loss": 0.9252,
+      "step": 15035
+    },
+    {
+      "epoch": 0.3622263432961634,
+      "grad_norm": 1.420731782913208,
+      "learning_rate": 6.713822127573268e-07,
+      "loss": 0.8925,
+      "step": 15040
+    },
+    {
+      "epoch": 0.36234676428795065,
+      "grad_norm": 1.6841161251068115,
+      "learning_rate": 6.712554507656424e-07,
+      "loss": 0.8825,
+      "step": 15045
+    },
+    {
+      "epoch": 0.362467185279738,
+      "grad_norm": 1.5006433725357056,
+      "learning_rate": 6.71128688773958e-07,
+      "loss": 0.934,
+      "step": 15050
+    },
+    {
+      "epoch": 0.36258760627152525,
+      "grad_norm": 1.5156694650650024,
+      "learning_rate": 6.710019267822735e-07,
+      "loss": 0.8896,
+      "step": 15055
+    },
+    {
+      "epoch": 0.3627080272633125,
+      "grad_norm": 1.449398159980774,
+      "learning_rate": 6.708751647905891e-07,
+      "loss": 0.8704,
+      "step": 15060
+    },
+    {
+      "epoch": 0.36282844825509986,
+      "grad_norm": 1.5191737413406372,
+      "learning_rate": 6.707484027989048e-07,
+      "loss": 0.959,
+      "step": 15065
+    },
+    {
+      "epoch": 0.36294886924688713,
+      "grad_norm": 1.6532554626464844,
+      "learning_rate": 6.706216408072203e-07,
+      "loss": 0.9403,
+      "step": 15070
+    },
+    {
+      "epoch": 0.3630692902386744,
+      "grad_norm": 1.6268415451049805,
+      "learning_rate": 6.70494878815536e-07,
+      "loss": 0.931,
+      "step": 15075
+    },
+    {
+      "epoch": 0.3631897112304617,
+      "grad_norm": 1.616729974746704,
+      "learning_rate": 6.703681168238515e-07,
+      "loss": 0.9523,
+      "step": 15080
+    },
+    {
+      "epoch": 0.363310132222249,
+      "grad_norm": 1.5711734294891357,
+      "learning_rate": 6.70241354832167e-07,
+      "loss": 0.8738,
+      "step": 15085
+    },
+    {
+      "epoch": 0.3634305532140363,
+      "grad_norm": 1.3926578760147095,
+      "learning_rate": 6.701145928404827e-07,
+      "loss": 0.8539,
+      "step": 15090
+    },
+    {
+      "epoch": 0.36355097420582355,
+      "grad_norm": 1.7388372421264648,
+      "learning_rate": 6.699878308487983e-07,
+      "loss": 0.9459,
+      "step": 15095
+    },
+    {
+      "epoch": 0.3636713951976108,
+      "grad_norm": 1.4930264949798584,
+      "learning_rate": 6.698610688571138e-07,
+      "loss": 0.9285,
+      "step": 15100
+    },
+    {
+      "epoch": 0.36379181618939815,
+      "grad_norm": 1.6159045696258545,
+      "learning_rate": 6.697343068654294e-07,
+      "loss": 0.8258,
+      "step": 15105
+    },
+    {
+      "epoch": 0.3639122371811854,
+      "grad_norm": 1.646911382675171,
+      "learning_rate": 6.69607544873745e-07,
+      "loss": 0.9706,
+      "step": 15110
+    },
+    {
+      "epoch": 0.3640326581729727,
+      "grad_norm": 1.7125273942947388,
+      "learning_rate": 6.694807828820606e-07,
+      "loss": 0.8925,
+      "step": 15115
+    },
+    {
+      "epoch": 0.36415307916476,
+      "grad_norm": 1.4787230491638184,
+      "learning_rate": 6.693540208903762e-07,
+      "loss": 0.8243,
+      "step": 15120
+    },
+    {
+      "epoch": 0.3642735001565473,
+      "grad_norm": 1.6219500303268433,
+      "learning_rate": 6.692272588986918e-07,
+      "loss": 0.9227,
+      "step": 15125
+    },
+    {
+      "epoch": 0.3643939211483346,
+      "grad_norm": 1.4867146015167236,
+      "learning_rate": 6.691004969070074e-07,
+      "loss": 0.8787,
+      "step": 15130
+    },
+    {
+      "epoch": 0.36451434214012185,
+      "grad_norm": 1.5020312070846558,
+      "learning_rate": 6.689737349153229e-07,
+      "loss": 0.9214,
+      "step": 15135
+    },
+    {
+      "epoch": 0.3646347631319092,
+      "grad_norm": 1.6953963041305542,
+      "learning_rate": 6.688469729236386e-07,
+      "loss": 0.9409,
+      "step": 15140
+    },
+    {
+      "epoch": 0.36475518412369645,
+      "grad_norm": 1.7892584800720215,
+      "learning_rate": 6.687202109319542e-07,
+      "loss": 0.894,
+      "step": 15145
+    },
+    {
+      "epoch": 0.3648756051154837,
+      "grad_norm": 1.7647579908370972,
+      "learning_rate": 6.685934489402697e-07,
+      "loss": 0.9145,
+      "step": 15150
+    },
+    {
+      "epoch": 0.364996026107271,
+      "grad_norm": 1.6124937534332275,
+      "learning_rate": 6.684666869485853e-07,
+      "loss": 0.961,
+      "step": 15155
+    },
+    {
+      "epoch": 0.3651164470990583,
+      "grad_norm": 1.5672036409378052,
+      "learning_rate": 6.68339924956901e-07,
+      "loss": 0.9143,
+      "step": 15160
+    },
+    {
+      "epoch": 0.3652368680908456,
+      "grad_norm": 1.5922040939331055,
+      "learning_rate": 6.682131629652165e-07,
+      "loss": 0.9337,
+      "step": 15165
+    },
+    {
+      "epoch": 0.3653572890826329,
+      "grad_norm": 1.4634770154953003,
+      "learning_rate": 6.68086400973532e-07,
+      "loss": 0.9371,
+      "step": 15170
+    },
+    {
+      "epoch": 0.36547771007442015,
+      "grad_norm": 1.745644211769104,
+      "learning_rate": 6.679596389818477e-07,
+      "loss": 0.894,
+      "step": 15175
+    },
+    {
+      "epoch": 0.3655981310662075,
+      "grad_norm": 1.4401402473449707,
+      "learning_rate": 6.678328769901632e-07,
+      "loss": 0.9414,
+      "step": 15180
+    },
+    {
+      "epoch": 0.36571855205799475,
+      "grad_norm": 1.5754197835922241,
+      "learning_rate": 6.677061149984789e-07,
+      "loss": 0.9305,
+      "step": 15185
+    },
+    {
+      "epoch": 0.365838973049782,
+      "grad_norm": 1.81437349319458,
+      "learning_rate": 6.675793530067945e-07,
+      "loss": 0.9554,
+      "step": 15190
+    },
+    {
+      "epoch": 0.36595939404156935,
+      "grad_norm": 1.5831947326660156,
+      "learning_rate": 6.674525910151099e-07,
+      "loss": 0.9147,
+      "step": 15195
+    },
+    {
+      "epoch": 0.3660798150333566,
+      "grad_norm": 1.520748257637024,
+      "learning_rate": 6.673258290234256e-07,
+      "loss": 0.8806,
+      "step": 15200
+    },
+    {
+      "epoch": 0.3662002360251439,
+      "grad_norm": 1.6147689819335938,
+      "learning_rate": 6.671990670317412e-07,
+      "loss": 0.9263,
+      "step": 15205
+    },
+    {
+      "epoch": 0.3663206570169312,
+      "grad_norm": 1.4286832809448242,
+      "learning_rate": 6.670723050400568e-07,
+      "loss": 0.9172,
+      "step": 15210
+    },
+    {
+      "epoch": 0.3664410780087185,
+      "grad_norm": 1.776050090789795,
+      "learning_rate": 6.669455430483723e-07,
+      "loss": 0.8837,
+      "step": 15215
+    },
+    {
+      "epoch": 0.3665614990005058,
+      "grad_norm": 1.509376883506775,
+      "learning_rate": 6.66818781056688e-07,
+      "loss": 0.8783,
+      "step": 15220
+    },
+    {
+      "epoch": 0.36668191999229305,
+      "grad_norm": 1.715714454650879,
+      "learning_rate": 6.666920190650035e-07,
+      "loss": 0.8801,
+      "step": 15225
+    },
+    {
+      "epoch": 0.3668023409840803,
+      "grad_norm": 1.6566847562789917,
+      "learning_rate": 6.665652570733191e-07,
+      "loss": 0.8783,
+      "step": 15230
+    },
+    {
+      "epoch": 0.36692276197586765,
+      "grad_norm": 1.720525860786438,
+      "learning_rate": 6.664384950816348e-07,
+      "loss": 0.9263,
+      "step": 15235
+    },
+    {
+      "epoch": 0.3670431829676549,
+      "grad_norm": 1.5244266986846924,
+      "learning_rate": 6.663117330899502e-07,
+      "loss": 0.8521,
+      "step": 15240
+    },
+    {
+      "epoch": 0.3671636039594422,
+      "grad_norm": 1.6485753059387207,
+      "learning_rate": 6.661849710982659e-07,
+      "loss": 0.8937,
+      "step": 15245
+    },
+    {
+      "epoch": 0.36728402495122947,
+      "grad_norm": 1.4862918853759766,
+      "learning_rate": 6.660582091065815e-07,
+      "loss": 0.9009,
+      "step": 15250
+    },
+    {
+      "epoch": 0.3674044459430168,
+      "grad_norm": 1.5611529350280762,
+      "learning_rate": 6.65931447114897e-07,
+      "loss": 0.8937,
+      "step": 15255
+    },
+    {
+      "epoch": 0.3675248669348041,
+      "grad_norm": 1.604556918144226,
+      "learning_rate": 6.658046851232127e-07,
+      "loss": 0.9168,
+      "step": 15260
+    },
+    {
+      "epoch": 0.36764528792659135,
+      "grad_norm": 1.7058712244033813,
+      "learning_rate": 6.656779231315282e-07,
+      "loss": 0.9182,
+      "step": 15265
+    },
+    {
+      "epoch": 0.3677657089183787,
+      "grad_norm": 1.5210764408111572,
+      "learning_rate": 6.655511611398438e-07,
+      "loss": 0.9549,
+      "step": 15270
+    },
+    {
+      "epoch": 0.36788612991016595,
+      "grad_norm": 1.4679126739501953,
+      "learning_rate": 6.654243991481594e-07,
+      "loss": 0.9445,
+      "step": 15275
+    },
+    {
+      "epoch": 0.3680065509019532,
+      "grad_norm": 1.5327095985412598,
+      "learning_rate": 6.652976371564751e-07,
+      "loss": 0.9276,
+      "step": 15280
+    },
+    {
+      "epoch": 0.3681269718937405,
+      "grad_norm": 1.4916969537734985,
+      "learning_rate": 6.651708751647905e-07,
+      "loss": 0.9069,
+      "step": 15285
+    },
+    {
+      "epoch": 0.3682473928855278,
+      "grad_norm": 1.508493185043335,
+      "learning_rate": 6.650441131731061e-07,
+      "loss": 0.9226,
+      "step": 15290
+    },
+    {
+      "epoch": 0.3683678138773151,
+      "grad_norm": 1.6520406007766724,
+      "learning_rate": 6.649173511814218e-07,
+      "loss": 0.8933,
+      "step": 15295
+    },
+    {
+      "epoch": 0.36848823486910237,
+      "grad_norm": 1.682572603225708,
+      "learning_rate": 6.647905891897373e-07,
+      "loss": 0.895,
+      "step": 15300
+    },
+    {
+      "epoch": 0.36860865586088964,
+      "grad_norm": 1.3576946258544922,
+      "learning_rate": 6.64663827198053e-07,
+      "loss": 0.8832,
+      "step": 15305
+    },
+    {
+      "epoch": 0.368729076852677,
+      "grad_norm": 1.4632619619369507,
+      "learning_rate": 6.645370652063685e-07,
+      "loss": 0.9272,
+      "step": 15310
+    },
+    {
+      "epoch": 0.36884949784446425,
+      "grad_norm": 1.3913495540618896,
+      "learning_rate": 6.64410303214684e-07,
+      "loss": 0.8963,
+      "step": 15315
+    },
+    {
+      "epoch": 0.3689699188362515,
+      "grad_norm": 1.4010612964630127,
+      "learning_rate": 6.642835412229997e-07,
+      "loss": 0.8897,
+      "step": 15320
+    },
+    {
+      "epoch": 0.36909033982803885,
+      "grad_norm": 1.4855096340179443,
+      "learning_rate": 6.641567792313153e-07,
+      "loss": 0.9235,
+      "step": 15325
+    },
+    {
+      "epoch": 0.3692107608198261,
+      "grad_norm": 1.5315731763839722,
+      "learning_rate": 6.640300172396308e-07,
+      "loss": 0.8996,
+      "step": 15330
+    },
+    {
+      "epoch": 0.3693311818116134,
+      "grad_norm": 1.444700002670288,
+      "learning_rate": 6.639032552479464e-07,
+      "loss": 0.8574,
+      "step": 15335
+    },
+    {
+      "epoch": 0.36945160280340067,
+      "grad_norm": 1.5219303369522095,
+      "learning_rate": 6.63776493256262e-07,
+      "loss": 0.8845,
+      "step": 15340
+    },
+    {
+      "epoch": 0.369572023795188,
+      "grad_norm": 1.5101958513259888,
+      "learning_rate": 6.636497312645776e-07,
+      "loss": 0.8634,
+      "step": 15345
+    },
+    {
+      "epoch": 0.36969244478697527,
+      "grad_norm": 1.630250334739685,
+      "learning_rate": 6.635229692728932e-07,
+      "loss": 0.908,
+      "step": 15350
+    },
+    {
+      "epoch": 0.36981286577876255,
+      "grad_norm": 1.55610990524292,
+      "learning_rate": 6.633962072812088e-07,
+      "loss": 0.9107,
+      "step": 15355
+    },
+    {
+      "epoch": 0.3699332867705498,
+      "grad_norm": 1.470426082611084,
+      "learning_rate": 6.632694452895243e-07,
+      "loss": 0.9563,
+      "step": 15360
+    },
+    {
+      "epoch": 0.37005370776233715,
+      "grad_norm": 1.4882392883300781,
+      "learning_rate": 6.6314268329784e-07,
+      "loss": 0.9542,
+      "step": 15365
+    },
+    {
+      "epoch": 0.3701741287541244,
+      "grad_norm": 1.565896987915039,
+      "learning_rate": 6.630159213061556e-07,
+      "loss": 0.973,
+      "step": 15370
+    },
+    {
+      "epoch": 0.3702945497459117,
+      "grad_norm": 1.5221936702728271,
+      "learning_rate": 6.628891593144711e-07,
+      "loss": 0.9164,
+      "step": 15375
+    },
+    {
+      "epoch": 0.370414970737699,
+      "grad_norm": 1.369215726852417,
+      "learning_rate": 6.627623973227867e-07,
+      "loss": 0.8889,
+      "step": 15380
+    },
+    {
+      "epoch": 0.3705353917294863,
+      "grad_norm": 1.3607157468795776,
+      "learning_rate": 6.626356353311023e-07,
+      "loss": 0.9203,
+      "step": 15385
+    },
+    {
+      "epoch": 0.37065581272127357,
+      "grad_norm": 1.4368348121643066,
+      "learning_rate": 6.625088733394179e-07,
+      "loss": 0.9226,
+      "step": 15390
+    },
+    {
+      "epoch": 0.37077623371306084,
+      "grad_norm": 1.6463371515274048,
+      "learning_rate": 6.623821113477335e-07,
+      "loss": 0.8857,
+      "step": 15395
+    },
+    {
+      "epoch": 0.37089665470484817,
+      "grad_norm": 1.6441726684570312,
+      "learning_rate": 6.62255349356049e-07,
+      "loss": 0.8931,
+      "step": 15400
+    },
+    {
+      "epoch": 0.37101707569663545,
+      "grad_norm": 1.3902664184570312,
+      "learning_rate": 6.621285873643646e-07,
+      "loss": 0.9283,
+      "step": 15405
+    },
+    {
+      "epoch": 0.3711374966884227,
+      "grad_norm": 1.4455829858779907,
+      "learning_rate": 6.620018253726802e-07,
+      "loss": 0.933,
+      "step": 15410
+    },
+    {
+      "epoch": 0.37125791768021,
+      "grad_norm": 1.4055064916610718,
+      "learning_rate": 6.618750633809959e-07,
+      "loss": 0.8331,
+      "step": 15415
+    },
+    {
+      "epoch": 0.3713783386719973,
+      "grad_norm": 1.5223290920257568,
+      "learning_rate": 6.617483013893114e-07,
+      "loss": 0.8939,
+      "step": 15420
+    },
+    {
+      "epoch": 0.3714987596637846,
+      "grad_norm": 1.7608566284179688,
+      "learning_rate": 6.61621539397627e-07,
+      "loss": 0.9388,
+      "step": 15425
+    },
+    {
+      "epoch": 0.37161918065557187,
+      "grad_norm": 1.5901376008987427,
+      "learning_rate": 6.614947774059426e-07,
+      "loss": 0.8834,
+      "step": 15430
+    },
+    {
+      "epoch": 0.37173960164735914,
+      "grad_norm": 1.581833839416504,
+      "learning_rate": 6.613680154142581e-07,
+      "loss": 0.9965,
+      "step": 15435
+    },
+    {
+      "epoch": 0.37186002263914647,
+      "grad_norm": 1.6572521924972534,
+      "learning_rate": 6.612412534225738e-07,
+      "loss": 0.8962,
+      "step": 15440
+    },
+    {
+      "epoch": 0.37198044363093374,
+      "grad_norm": 1.6382677555084229,
+      "learning_rate": 6.611144914308894e-07,
+      "loss": 0.9106,
+      "step": 15445
+    },
+    {
+      "epoch": 0.372100864622721,
+      "grad_norm": 1.4641860723495483,
+      "learning_rate": 6.609877294392049e-07,
+      "loss": 0.9181,
+      "step": 15450
+    },
+    {
+      "epoch": 0.37222128561450835,
+      "grad_norm": 1.5975430011749268,
+      "learning_rate": 6.608609674475205e-07,
+      "loss": 0.9305,
+      "step": 15455
+    },
+    {
+      "epoch": 0.3723417066062956,
+      "grad_norm": 1.375770092010498,
+      "learning_rate": 6.607342054558361e-07,
+      "loss": 0.9122,
+      "step": 15460
+    },
+    {
+      "epoch": 0.3724621275980829,
+      "grad_norm": 1.7418895959854126,
+      "learning_rate": 6.606074434641517e-07,
+      "loss": 0.8588,
+      "step": 15465
+    },
+    {
+      "epoch": 0.37258254858987017,
+      "grad_norm": 1.4999524354934692,
+      "learning_rate": 6.604806814724672e-07,
+      "loss": 0.953,
+      "step": 15470
+    },
+    {
+      "epoch": 0.3727029695816575,
+      "grad_norm": 1.607024908065796,
+      "learning_rate": 6.603539194807829e-07,
+      "loss": 0.9159,
+      "step": 15475
+    },
+    {
+      "epoch": 0.37282339057344477,
+      "grad_norm": 1.7114670276641846,
+      "learning_rate": 6.602271574890984e-07,
+      "loss": 0.9333,
+      "step": 15480
+    },
+    {
+      "epoch": 0.37294381156523204,
+      "grad_norm": 1.398600697517395,
+      "learning_rate": 6.601003954974141e-07,
+      "loss": 0.9032,
+      "step": 15485
+    },
+    {
+      "epoch": 0.3730642325570193,
+      "grad_norm": 1.5683618783950806,
+      "learning_rate": 6.599736335057297e-07,
+      "loss": 0.8666,
+      "step": 15490
+    },
+    {
+      "epoch": 0.37318465354880664,
+      "grad_norm": 1.7809679508209229,
+      "learning_rate": 6.598468715140451e-07,
+      "loss": 0.9364,
+      "step": 15495
+    },
+    {
+      "epoch": 0.3733050745405939,
+      "grad_norm": 1.5396602153778076,
+      "learning_rate": 6.597201095223608e-07,
+      "loss": 0.9189,
+      "step": 15500
+    },
+    {
+      "epoch": 0.3734254955323812,
+      "grad_norm": 1.375185251235962,
+      "learning_rate": 6.595933475306764e-07,
+      "loss": 0.9043,
+      "step": 15505
+    },
+    {
+      "epoch": 0.3735459165241685,
+      "grad_norm": 1.5217154026031494,
+      "learning_rate": 6.59466585538992e-07,
+      "loss": 0.9204,
+      "step": 15510
+    },
+    {
+      "epoch": 0.3736663375159558,
+      "grad_norm": 1.5808666944503784,
+      "learning_rate": 6.593398235473075e-07,
+      "loss": 0.9414,
+      "step": 15515
+    },
+    {
+      "epoch": 0.37378675850774307,
+      "grad_norm": 1.6650584936141968,
+      "learning_rate": 6.592130615556231e-07,
+      "loss": 0.8927,
+      "step": 15520
+    },
+    {
+      "epoch": 0.37390717949953034,
+      "grad_norm": 1.3403691053390503,
+      "learning_rate": 6.590862995639387e-07,
+      "loss": 0.9146,
+      "step": 15525
+    },
+    {
+      "epoch": 0.37402760049131767,
+      "grad_norm": 1.6455671787261963,
+      "learning_rate": 6.589595375722543e-07,
+      "loss": 0.887,
+      "step": 15530
+    },
+    {
+      "epoch": 0.37414802148310494,
+      "grad_norm": 1.5723843574523926,
+      "learning_rate": 6.5883277558057e-07,
+      "loss": 0.885,
+      "step": 15535
+    },
+    {
+      "epoch": 0.3742684424748922,
+      "grad_norm": 1.5224273204803467,
+      "learning_rate": 6.587060135888854e-07,
+      "loss": 0.9342,
+      "step": 15540
+    },
+    {
+      "epoch": 0.3743888634666795,
+      "grad_norm": 1.4680192470550537,
+      "learning_rate": 6.58579251597201e-07,
+      "loss": 0.915,
+      "step": 15545
+    },
+    {
+      "epoch": 0.3745092844584668,
+      "grad_norm": 1.478745937347412,
+      "learning_rate": 6.584524896055167e-07,
+      "loss": 0.9552,
+      "step": 15550
+    },
+    {
+      "epoch": 0.3746297054502541,
+      "grad_norm": 1.7124029397964478,
+      "learning_rate": 6.583257276138322e-07,
+      "loss": 0.9157,
+      "step": 15555
+    },
+    {
+      "epoch": 0.37475012644204136,
+      "grad_norm": 1.7640318870544434,
+      "learning_rate": 6.581989656221479e-07,
+      "loss": 0.9328,
+      "step": 15560
+    },
+    {
+      "epoch": 0.37487054743382864,
+      "grad_norm": 1.5102007389068604,
+      "learning_rate": 6.580722036304634e-07,
+      "loss": 0.9801,
+      "step": 15565
+    },
+    {
+      "epoch": 0.37499096842561597,
+      "grad_norm": 1.603365182876587,
+      "learning_rate": 6.57945441638779e-07,
+      "loss": 0.9126,
+      "step": 15570
+    },
+    {
+      "epoch": 0.37511138941740324,
+      "grad_norm": 1.7390234470367432,
+      "learning_rate": 6.578186796470946e-07,
+      "loss": 0.9553,
+      "step": 15575
+    },
+    {
+      "epoch": 0.3752318104091905,
+      "grad_norm": 1.5153967142105103,
+      "learning_rate": 6.576919176554102e-07,
+      "loss": 0.9007,
+      "step": 15580
+    },
+    {
+      "epoch": 0.37535223140097784,
+      "grad_norm": 1.5532500743865967,
+      "learning_rate": 6.575651556637257e-07,
+      "loss": 0.9155,
+      "step": 15585
+    },
+    {
+      "epoch": 0.3754726523927651,
+      "grad_norm": 1.6289955377578735,
+      "learning_rate": 6.574383936720413e-07,
+      "loss": 0.9619,
+      "step": 15590
+    },
+    {
+      "epoch": 0.3755930733845524,
+      "grad_norm": 1.5017445087432861,
+      "learning_rate": 6.57311631680357e-07,
+      "loss": 0.9287,
+      "step": 15595
+    },
+    {
+      "epoch": 0.37571349437633966,
+      "grad_norm": 1.3886905908584595,
+      "learning_rate": 6.571848696886725e-07,
+      "loss": 0.9289,
+      "step": 15600
+    },
+    {
+      "epoch": 0.375833915368127,
+      "grad_norm": 1.5643943548202515,
+      "learning_rate": 6.570581076969882e-07,
+      "loss": 0.9097,
+      "step": 15605
+    },
+    {
+      "epoch": 0.37595433635991427,
+      "grad_norm": 1.5563030242919922,
+      "learning_rate": 6.569313457053037e-07,
+      "loss": 0.8883,
+      "step": 15610
+    },
+    {
+      "epoch": 0.37607475735170154,
+      "grad_norm": 1.6308798789978027,
+      "learning_rate": 6.568045837136192e-07,
+      "loss": 0.9332,
+      "step": 15615
+    },
+    {
+      "epoch": 0.3761951783434888,
+      "grad_norm": 1.3513928651809692,
+      "learning_rate": 6.566778217219349e-07,
+      "loss": 0.9116,
+      "step": 15620
+    },
+    {
+      "epoch": 0.37631559933527614,
+      "grad_norm": 1.5536242723464966,
+      "learning_rate": 6.565510597302505e-07,
+      "loss": 0.9038,
+      "step": 15625
+    },
+    {
+      "epoch": 0.3764360203270634,
+      "grad_norm": 1.635443925857544,
+      "learning_rate": 6.56424297738566e-07,
+      "loss": 0.9306,
+      "step": 15630
+    },
+    {
+      "epoch": 0.3765564413188507,
+      "grad_norm": 1.6803452968597412,
+      "learning_rate": 6.562975357468816e-07,
+      "loss": 0.9142,
+      "step": 15635
+    },
+    {
+      "epoch": 0.376676862310638,
+      "grad_norm": 1.5191351175308228,
+      "learning_rate": 6.561707737551972e-07,
+      "loss": 0.9241,
+      "step": 15640
+    },
+    {
+      "epoch": 0.3767972833024253,
+      "grad_norm": 1.50296950340271,
+      "learning_rate": 6.560440117635128e-07,
+      "loss": 0.8873,
+      "step": 15645
+    },
+    {
+      "epoch": 0.37691770429421256,
+      "grad_norm": 1.504440426826477,
+      "learning_rate": 6.559172497718284e-07,
+      "loss": 0.9058,
+      "step": 15650
+    },
+    {
+      "epoch": 0.37703812528599984,
+      "grad_norm": 2.1617422103881836,
+      "learning_rate": 6.55790487780144e-07,
+      "loss": 0.967,
+      "step": 15655
+    },
+    {
+      "epoch": 0.37715854627778717,
+      "grad_norm": 1.6108896732330322,
+      "learning_rate": 6.556637257884595e-07,
+      "loss": 0.9132,
+      "step": 15660
+    },
+    {
+      "epoch": 0.37727896726957444,
+      "grad_norm": 1.5929207801818848,
+      "learning_rate": 6.555369637967751e-07,
+      "loss": 0.8985,
+      "step": 15665
+    },
+    {
+      "epoch": 0.3773993882613617,
+      "grad_norm": 1.4225010871887207,
+      "learning_rate": 6.554102018050908e-07,
+      "loss": 0.902,
+      "step": 15670
+    },
+    {
+      "epoch": 0.377519809253149,
+      "grad_norm": 1.5468358993530273,
+      "learning_rate": 6.552834398134063e-07,
+      "loss": 0.8869,
+      "step": 15675
+    },
+    {
+      "epoch": 0.3776402302449363,
+      "grad_norm": 1.748754858970642,
+      "learning_rate": 6.551566778217219e-07,
+      "loss": 0.9002,
+      "step": 15680
+    },
+    {
+      "epoch": 0.3777606512367236,
+      "grad_norm": 1.3694506883621216,
+      "learning_rate": 6.550299158300375e-07,
+      "loss": 0.9047,
+      "step": 15685
+    },
+    {
+      "epoch": 0.37788107222851086,
+      "grad_norm": 1.686044454574585,
+      "learning_rate": 6.549031538383531e-07,
+      "loss": 0.8809,
+      "step": 15690
+    },
+    {
+      "epoch": 0.37800149322029813,
+      "grad_norm": 1.3558310270309448,
+      "learning_rate": 6.547763918466687e-07,
+      "loss": 0.8826,
+      "step": 15695
+    },
+    {
+      "epoch": 0.37812191421208546,
+      "grad_norm": 1.455237865447998,
+      "learning_rate": 6.546496298549842e-07,
+      "loss": 0.8968,
+      "step": 15700
+    },
+    {
+      "epoch": 0.37824233520387274,
+      "grad_norm": 1.6350035667419434,
+      "learning_rate": 6.545228678632998e-07,
+      "loss": 0.9855,
+      "step": 15705
+    },
+    {
+      "epoch": 0.37836275619566,
+      "grad_norm": 1.4560292959213257,
+      "learning_rate": 6.543961058716154e-07,
+      "loss": 0.9257,
+      "step": 15710
+    },
+    {
+      "epoch": 0.37848317718744734,
+      "grad_norm": 1.6068018674850464,
+      "learning_rate": 6.542693438799311e-07,
+      "loss": 0.8658,
+      "step": 15715
+    },
+    {
+      "epoch": 0.3786035981792346,
+      "grad_norm": 1.4421308040618896,
+      "learning_rate": 6.541425818882466e-07,
+      "loss": 0.9409,
+      "step": 15720
+    },
+    {
+      "epoch": 0.3787240191710219,
+      "grad_norm": 1.6204299926757812,
+      "learning_rate": 6.540158198965621e-07,
+      "loss": 0.9064,
+      "step": 15725
+    },
+    {
+      "epoch": 0.37884444016280916,
+      "grad_norm": 1.4492141008377075,
+      "learning_rate": 6.538890579048778e-07,
+      "loss": 0.8868,
+      "step": 15730
+    },
+    {
+      "epoch": 0.3789648611545965,
+      "grad_norm": 1.4784122705459595,
+      "learning_rate": 6.537622959131933e-07,
+      "loss": 0.8887,
+      "step": 15735
+    },
+    {
+      "epoch": 0.37908528214638376,
+      "grad_norm": 1.5144078731536865,
+      "learning_rate": 6.53635533921509e-07,
+      "loss": 0.8895,
+      "step": 15740
+    },
+    {
+      "epoch": 0.37920570313817104,
+      "grad_norm": 1.4604182243347168,
+      "learning_rate": 6.535087719298246e-07,
+      "loss": 0.908,
+      "step": 15745
+    },
+    {
+      "epoch": 0.3793261241299583,
+      "grad_norm": 1.8133037090301514,
+      "learning_rate": 6.5338200993814e-07,
+      "loss": 0.8743,
+      "step": 15750
+    },
+    {
+      "epoch": 0.37944654512174564,
+      "grad_norm": 1.9104950428009033,
+      "learning_rate": 6.532552479464557e-07,
+      "loss": 0.9707,
+      "step": 15755
+    },
+    {
+      "epoch": 0.3795669661135329,
+      "grad_norm": 1.2851866483688354,
+      "learning_rate": 6.531284859547713e-07,
+      "loss": 0.8677,
+      "step": 15760
+    },
+    {
+      "epoch": 0.3796873871053202,
+      "grad_norm": 1.5550357103347778,
+      "learning_rate": 6.530017239630869e-07,
+      "loss": 0.9423,
+      "step": 15765
+    },
+    {
+      "epoch": 0.3798078080971075,
+      "grad_norm": 1.375746250152588,
+      "learning_rate": 6.528749619714024e-07,
+      "loss": 0.8621,
+      "step": 15770
+    },
+    {
+      "epoch": 0.3799282290888948,
+      "grad_norm": 1.5696123838424683,
+      "learning_rate": 6.527481999797181e-07,
+      "loss": 0.8835,
+      "step": 15775
+    },
+    {
+      "epoch": 0.38004865008068206,
+      "grad_norm": 1.5719820261001587,
+      "learning_rate": 6.526214379880336e-07,
+      "loss": 0.9317,
+      "step": 15780
+    },
+    {
+      "epoch": 0.38016907107246933,
+      "grad_norm": 1.4429283142089844,
+      "learning_rate": 6.524946759963492e-07,
+      "loss": 0.862,
+      "step": 15785
+    },
+    {
+      "epoch": 0.38028949206425666,
+      "grad_norm": 1.5603541135787964,
+      "learning_rate": 6.523679140046649e-07,
+      "loss": 0.9097,
+      "step": 15790
+    },
+    {
+      "epoch": 0.38040991305604394,
+      "grad_norm": 1.5130277872085571,
+      "learning_rate": 6.522411520129803e-07,
+      "loss": 0.9088,
+      "step": 15795
+    },
+    {
+      "epoch": 0.3805303340478312,
+      "grad_norm": 1.3764976263046265,
+      "learning_rate": 6.52114390021296e-07,
+      "loss": 0.8709,
+      "step": 15800
+    },
+    {
+      "epoch": 0.3806507550396185,
+      "grad_norm": 1.3912783861160278,
+      "learning_rate": 6.519876280296116e-07,
+      "loss": 0.9016,
+      "step": 15805
+    },
+    {
+      "epoch": 0.3807711760314058,
+      "grad_norm": 1.5324950218200684,
+      "learning_rate": 6.518608660379272e-07,
+      "loss": 0.913,
+      "step": 15810
+    },
+    {
+      "epoch": 0.3808915970231931,
+      "grad_norm": 1.5858607292175293,
+      "learning_rate": 6.517341040462427e-07,
+      "loss": 0.8764,
+      "step": 15815
+    },
+    {
+      "epoch": 0.38101201801498036,
+      "grad_norm": 1.5081491470336914,
+      "learning_rate": 6.516073420545583e-07,
+      "loss": 0.9308,
+      "step": 15820
+    },
+    {
+      "epoch": 0.3811324390067677,
+      "grad_norm": 1.554904580116272,
+      "learning_rate": 6.514805800628739e-07,
+      "loss": 0.9474,
+      "step": 15825
+    },
+    {
+      "epoch": 0.38125285999855496,
+      "grad_norm": 1.6217893362045288,
+      "learning_rate": 6.513538180711895e-07,
+      "loss": 0.887,
+      "step": 15830
+    },
+    {
+      "epoch": 0.38137328099034223,
+      "grad_norm": 1.6178269386291504,
+      "learning_rate": 6.512270560795052e-07,
+      "loss": 0.958,
+      "step": 15835
+    },
+    {
+      "epoch": 0.3814937019821295,
+      "grad_norm": 1.3969887495040894,
+      "learning_rate": 6.511002940878206e-07,
+      "loss": 0.948,
+      "step": 15840
+    },
+    {
+      "epoch": 0.38161412297391684,
+      "grad_norm": 1.5228673219680786,
+      "learning_rate": 6.509735320961362e-07,
+      "loss": 0.8732,
+      "step": 15845
+    },
+    {
+      "epoch": 0.3817345439657041,
+      "grad_norm": 1.427930235862732,
+      "learning_rate": 6.508467701044519e-07,
+      "loss": 0.9315,
+      "step": 15850
+    },
+    {
+      "epoch": 0.3818549649574914,
+      "grad_norm": 1.573591947555542,
+      "learning_rate": 6.507200081127674e-07,
+      "loss": 0.9077,
+      "step": 15855
+    },
+    {
+      "epoch": 0.38197538594927866,
+      "grad_norm": 1.5548583269119263,
+      "learning_rate": 6.505932461210831e-07,
+      "loss": 0.9134,
+      "step": 15860
+    },
+    {
+      "epoch": 0.382095806941066,
+      "grad_norm": 1.7557604312896729,
+      "learning_rate": 6.504664841293986e-07,
+      "loss": 0.9059,
+      "step": 15865
+    },
+    {
+      "epoch": 0.38221622793285326,
+      "grad_norm": 1.349873661994934,
+      "learning_rate": 6.503397221377141e-07,
+      "loss": 0.8972,
+      "step": 15870
+    },
+    {
+      "epoch": 0.38233664892464053,
+      "grad_norm": 1.7548439502716064,
+      "learning_rate": 6.502129601460298e-07,
+      "loss": 0.9036,
+      "step": 15875
+    },
+    {
+      "epoch": 0.3824570699164278,
+      "grad_norm": 1.759466290473938,
+      "learning_rate": 6.500861981543454e-07,
+      "loss": 0.9172,
+      "step": 15880
+    },
+    {
+      "epoch": 0.38257749090821513,
+      "grad_norm": 1.6105694770812988,
+      "learning_rate": 6.499594361626609e-07,
+      "loss": 0.89,
+      "step": 15885
+    },
+    {
+      "epoch": 0.3826979119000024,
+      "grad_norm": 1.4921594858169556,
+      "learning_rate": 6.498326741709765e-07,
+      "loss": 0.958,
+      "step": 15890
+    },
+    {
+      "epoch": 0.3828183328917897,
+      "grad_norm": 1.6352791786193848,
+      "learning_rate": 6.497059121792922e-07,
+      "loss": 0.8882,
+      "step": 15895
+    },
+    {
+      "epoch": 0.382938753883577,
+      "grad_norm": 1.5943946838378906,
+      "learning_rate": 6.495791501876077e-07,
+      "loss": 0.9128,
+      "step": 15900
+    },
+    {
+      "epoch": 0.3830591748753643,
+      "grad_norm": 1.497174859046936,
+      "learning_rate": 6.494523881959233e-07,
+      "loss": 0.9548,
+      "step": 15905
+    },
+    {
+      "epoch": 0.38317959586715156,
+      "grad_norm": 1.540085792541504,
+      "learning_rate": 6.493256262042389e-07,
+      "loss": 0.9293,
+      "step": 15910
+    },
+    {
+      "epoch": 0.38330001685893883,
+      "grad_norm": 1.5195233821868896,
+      "learning_rate": 6.491988642125544e-07,
+      "loss": 0.9311,
+      "step": 15915
+    },
+    {
+      "epoch": 0.38342043785072616,
+      "grad_norm": 1.3735129833221436,
+      "learning_rate": 6.490721022208701e-07,
+      "loss": 0.8887,
+      "step": 15920
+    },
+    {
+      "epoch": 0.38354085884251343,
+      "grad_norm": 1.440384030342102,
+      "learning_rate": 6.489453402291857e-07,
+      "loss": 0.8948,
+      "step": 15925
+    },
+    {
+      "epoch": 0.3836612798343007,
+      "grad_norm": 1.7172589302062988,
+      "learning_rate": 6.488185782375011e-07,
+      "loss": 0.8894,
+      "step": 15930
+    },
+    {
+      "epoch": 0.383781700826088,
+      "grad_norm": 1.529138207435608,
+      "learning_rate": 6.486918162458168e-07,
+      "loss": 0.9409,
+      "step": 15935
+    },
+    {
+      "epoch": 0.3839021218178753,
+      "grad_norm": 1.5386652946472168,
+      "learning_rate": 6.485650542541324e-07,
+      "loss": 0.9094,
+      "step": 15940
+    },
+    {
+      "epoch": 0.3840225428096626,
+      "grad_norm": 1.4735944271087646,
+      "learning_rate": 6.48438292262448e-07,
+      "loss": 0.8712,
+      "step": 15945
+    },
+    {
+      "epoch": 0.38414296380144985,
+      "grad_norm": 1.6935973167419434,
+      "learning_rate": 6.483115302707636e-07,
+      "loss": 0.8893,
+      "step": 15950
+    },
+    {
+      "epoch": 0.3842633847932372,
+      "grad_norm": 1.518571138381958,
+      "learning_rate": 6.481847682790792e-07,
+      "loss": 0.9172,
+      "step": 15955
+    },
+    {
+      "epoch": 0.38438380578502446,
+      "grad_norm": 1.3189793825149536,
+      "learning_rate": 6.480580062873947e-07,
+      "loss": 0.9205,
+      "step": 15960
+    },
+    {
+      "epoch": 0.38450422677681173,
+      "grad_norm": 1.464625358581543,
+      "learning_rate": 6.479312442957103e-07,
+      "loss": 0.8986,
+      "step": 15965
+    },
+    {
+      "epoch": 0.384624647768599,
+      "grad_norm": 1.9872775077819824,
+      "learning_rate": 6.47804482304026e-07,
+      "loss": 0.892,
+      "step": 15970
+    },
+    {
+      "epoch": 0.38474506876038633,
+      "grad_norm": 1.6058225631713867,
+      "learning_rate": 6.476777203123415e-07,
+      "loss": 0.9443,
+      "step": 15975
+    },
+    {
+      "epoch": 0.3848654897521736,
+      "grad_norm": 1.6584371328353882,
+      "learning_rate": 6.475509583206571e-07,
+      "loss": 0.8896,
+      "step": 15980
+    },
+    {
+      "epoch": 0.3849859107439609,
+      "grad_norm": 1.4320168495178223,
+      "learning_rate": 6.474241963289727e-07,
+      "loss": 0.8685,
+      "step": 15985
+    },
+    {
+      "epoch": 0.38510633173574815,
+      "grad_norm": 1.446475863456726,
+      "learning_rate": 6.472974343372882e-07,
+      "loss": 0.9195,
+      "step": 15990
+    },
+    {
+      "epoch": 0.3852267527275355,
+      "grad_norm": 1.3913068771362305,
+      "learning_rate": 6.471706723456039e-07,
+      "loss": 0.9159,
+      "step": 15995
+    },
+    {
+      "epoch": 0.38534717371932276,
+      "grad_norm": 1.4390864372253418,
+      "learning_rate": 6.470439103539194e-07,
+      "loss": 0.9049,
+      "step": 16000
+    },
+    {
+      "epoch": 0.38546759471111003,
+      "grad_norm": 1.8681691884994507,
+      "learning_rate": 6.46917148362235e-07,
+      "loss": 0.9117,
+      "step": 16005
+    },
+    {
+      "epoch": 0.3855880157028973,
+      "grad_norm": 1.505843997001648,
+      "learning_rate": 6.467903863705506e-07,
+      "loss": 0.9035,
+      "step": 16010
+    },
+    {
+      "epoch": 0.38570843669468463,
+      "grad_norm": 1.7058168649673462,
+      "learning_rate": 6.466636243788663e-07,
+      "loss": 0.8936,
+      "step": 16015
+    },
+    {
+      "epoch": 0.3858288576864719,
+      "grad_norm": 1.4852770566940308,
+      "learning_rate": 6.465368623871818e-07,
+      "loss": 0.9428,
+      "step": 16020
+    },
+    {
+      "epoch": 0.3859492786782592,
+      "grad_norm": 1.5499645471572876,
+      "learning_rate": 6.464101003954973e-07,
+      "loss": 0.9151,
+      "step": 16025
+    },
+    {
+      "epoch": 0.3860696996700465,
+      "grad_norm": 1.4708459377288818,
+      "learning_rate": 6.46283338403813e-07,
+      "loss": 0.9137,
+      "step": 16030
+    },
+    {
+      "epoch": 0.3861901206618338,
+      "grad_norm": 1.685199499130249,
+      "learning_rate": 6.461565764121285e-07,
+      "loss": 0.9266,
+      "step": 16035
+    },
+    {
+      "epoch": 0.38631054165362105,
+      "grad_norm": 2.1729648113250732,
+      "learning_rate": 6.460298144204442e-07,
+      "loss": 0.9238,
+      "step": 16040
+    },
+    {
+      "epoch": 0.3864309626454083,
+      "grad_norm": 1.7583898305892944,
+      "learning_rate": 6.459030524287598e-07,
+      "loss": 0.938,
+      "step": 16045
+    },
+    {
+      "epoch": 0.38655138363719566,
+      "grad_norm": 1.462072730064392,
+      "learning_rate": 6.457762904370752e-07,
+      "loss": 0.8998,
+      "step": 16050
+    },
+    {
+      "epoch": 0.38667180462898293,
+      "grad_norm": 1.5906111001968384,
+      "learning_rate": 6.456495284453909e-07,
+      "loss": 0.975,
+      "step": 16055
+    },
+    {
+      "epoch": 0.3867922256207702,
+      "grad_norm": 1.4166845083236694,
+      "learning_rate": 6.455227664537065e-07,
+      "loss": 0.8913,
+      "step": 16060
+    },
+    {
+      "epoch": 0.3869126466125575,
+      "grad_norm": 1.5081733465194702,
+      "learning_rate": 6.453960044620221e-07,
+      "loss": 0.9391,
+      "step": 16065
+    },
+    {
+      "epoch": 0.3870330676043448,
+      "grad_norm": 1.5960620641708374,
+      "learning_rate": 6.452692424703376e-07,
+      "loss": 0.9146,
+      "step": 16070
+    },
+    {
+      "epoch": 0.3871534885961321,
+      "grad_norm": 1.5286779403686523,
+      "learning_rate": 6.451424804786533e-07,
+      "loss": 0.9371,
+      "step": 16075
+    },
+    {
+      "epoch": 0.38727390958791935,
+      "grad_norm": 1.838586688041687,
+      "learning_rate": 6.450157184869688e-07,
+      "loss": 0.9116,
+      "step": 16080
+    },
+    {
+      "epoch": 0.3873943305797067,
+      "grad_norm": 1.3697798252105713,
+      "learning_rate": 6.448889564952844e-07,
+      "loss": 0.9188,
+      "step": 16085
+    },
+    {
+      "epoch": 0.38751475157149395,
+      "grad_norm": 1.5102379322052002,
+      "learning_rate": 6.447621945036001e-07,
+      "loss": 0.9338,
+      "step": 16090
+    },
+    {
+      "epoch": 0.3876351725632812,
+      "grad_norm": 1.5454723834991455,
+      "learning_rate": 6.446354325119155e-07,
+      "loss": 0.8167,
+      "step": 16095
+    },
+    {
+      "epoch": 0.3877555935550685,
+      "grad_norm": 1.4021121263504028,
+      "learning_rate": 6.445086705202312e-07,
+      "loss": 0.9009,
+      "step": 16100
+    },
+    {
+      "epoch": 0.38787601454685583,
+      "grad_norm": 1.4518719911575317,
+      "learning_rate": 6.443819085285468e-07,
+      "loss": 0.896,
+      "step": 16105
+    },
+    {
+      "epoch": 0.3879964355386431,
+      "grad_norm": 1.5615367889404297,
+      "learning_rate": 6.442551465368623e-07,
+      "loss": 0.8825,
+      "step": 16110
+    },
+    {
+      "epoch": 0.3881168565304304,
+      "grad_norm": 1.4544663429260254,
+      "learning_rate": 6.441283845451779e-07,
+      "loss": 0.9124,
+      "step": 16115
+    },
+    {
+      "epoch": 0.38823727752221765,
+      "grad_norm": 1.5126712322235107,
+      "learning_rate": 6.440016225534935e-07,
+      "loss": 0.9434,
+      "step": 16120
+    },
+    {
+      "epoch": 0.388357698514005,
+      "grad_norm": 1.5839651823043823,
+      "learning_rate": 6.438748605618091e-07,
+      "loss": 0.9702,
+      "step": 16125
+    },
+    {
+      "epoch": 0.38847811950579225,
+      "grad_norm": 2.018829107284546,
+      "learning_rate": 6.437480985701247e-07,
+      "loss": 0.9184,
+      "step": 16130
+    },
+    {
+      "epoch": 0.3885985404975795,
+      "grad_norm": 1.6250749826431274,
+      "learning_rate": 6.436213365784404e-07,
+      "loss": 0.9592,
+      "step": 16135
+    },
+    {
+      "epoch": 0.3887189614893668,
+      "grad_norm": 1.336804747581482,
+      "learning_rate": 6.434945745867558e-07,
+      "loss": 0.8839,
+      "step": 16140
+    },
+    {
+      "epoch": 0.3888393824811541,
+      "grad_norm": 1.6236604452133179,
+      "learning_rate": 6.433678125950714e-07,
+      "loss": 0.9167,
+      "step": 16145
+    },
+    {
+      "epoch": 0.3889598034729414,
+      "grad_norm": 1.5479767322540283,
+      "learning_rate": 6.432410506033871e-07,
+      "loss": 0.8679,
+      "step": 16150
+    },
+    {
+      "epoch": 0.3890802244647287,
+      "grad_norm": 1.5190855264663696,
+      "learning_rate": 6.431142886117027e-07,
+      "loss": 0.8982,
+      "step": 16155
+    },
+    {
+      "epoch": 0.389200645456516,
+      "grad_norm": 1.6366859674453735,
+      "learning_rate": 6.429875266200183e-07,
+      "loss": 0.9002,
+      "step": 16160
+    },
+    {
+      "epoch": 0.3893210664483033,
+      "grad_norm": 1.5403122901916504,
+      "learning_rate": 6.428607646283338e-07,
+      "loss": 0.8571,
+      "step": 16165
+    },
+    {
+      "epoch": 0.38944148744009055,
+      "grad_norm": 1.5745781660079956,
+      "learning_rate": 6.427340026366494e-07,
+      "loss": 0.9317,
+      "step": 16170
+    },
+    {
+      "epoch": 0.3895619084318778,
+      "grad_norm": 1.6703364849090576,
+      "learning_rate": 6.42607240644965e-07,
+      "loss": 0.927,
+      "step": 16175
+    },
+    {
+      "epoch": 0.38968232942366515,
+      "grad_norm": 1.4843415021896362,
+      "learning_rate": 6.424804786532806e-07,
+      "loss": 0.9106,
+      "step": 16180
+    },
+    {
+      "epoch": 0.3898027504154524,
+      "grad_norm": 1.4242477416992188,
+      "learning_rate": 6.423537166615962e-07,
+      "loss": 0.9481,
+      "step": 16185
+    },
+    {
+      "epoch": 0.3899231714072397,
+      "grad_norm": 1.34717857837677,
+      "learning_rate": 6.422269546699117e-07,
+      "loss": 0.9421,
+      "step": 16190
+    },
+    {
+      "epoch": 0.390043592399027,
+      "grad_norm": 1.734257698059082,
+      "learning_rate": 6.421001926782273e-07,
+      "loss": 0.9215,
+      "step": 16195
+    },
+    {
+      "epoch": 0.3901640133908143,
+      "grad_norm": 1.5538983345031738,
+      "learning_rate": 6.41973430686543e-07,
+      "loss": 0.9245,
+      "step": 16200
+    },
+    {
+      "epoch": 0.3902844343826016,
+      "grad_norm": 1.4236520528793335,
+      "learning_rate": 6.418466686948585e-07,
+      "loss": 0.9123,
+      "step": 16205
+    },
+    {
+      "epoch": 0.39040485537438885,
+      "grad_norm": 1.6775470972061157,
+      "learning_rate": 6.417199067031741e-07,
+      "loss": 0.9514,
+      "step": 16210
+    },
+    {
+      "epoch": 0.3905252763661762,
+      "grad_norm": 1.5894907712936401,
+      "learning_rate": 6.415931447114897e-07,
+      "loss": 0.8613,
+      "step": 16215
+    },
+    {
+      "epoch": 0.39064569735796345,
+      "grad_norm": 1.8615303039550781,
+      "learning_rate": 6.414663827198053e-07,
+      "loss": 0.8585,
+      "step": 16220
+    },
+    {
+      "epoch": 0.3907661183497507,
+      "grad_norm": 1.587877631187439,
+      "learning_rate": 6.413396207281209e-07,
+      "loss": 0.9591,
+      "step": 16225
+    },
+    {
+      "epoch": 0.390886539341538,
+      "grad_norm": 1.5446151494979858,
+      "learning_rate": 6.412128587364364e-07,
+      "loss": 0.9316,
+      "step": 16230
+    },
+    {
+      "epoch": 0.3910069603333253,
+      "grad_norm": 1.3091187477111816,
+      "learning_rate": 6.41086096744752e-07,
+      "loss": 0.9587,
+      "step": 16235
+    },
+    {
+      "epoch": 0.3911273813251126,
+      "grad_norm": 1.5175570249557495,
+      "learning_rate": 6.409593347530676e-07,
+      "loss": 0.9147,
+      "step": 16240
+    },
+    {
+      "epoch": 0.3912478023168999,
+      "grad_norm": 1.386093020439148,
+      "learning_rate": 6.408325727613833e-07,
+      "loss": 0.9803,
+      "step": 16245
+    },
+    {
+      "epoch": 0.39136822330868715,
+      "grad_norm": 1.4947230815887451,
+      "learning_rate": 6.407058107696988e-07,
+      "loss": 0.879,
+      "step": 16250
+    },
+    {
+      "epoch": 0.3914886443004745,
+      "grad_norm": 1.4012850522994995,
+      "learning_rate": 6.405790487780143e-07,
+      "loss": 0.8928,
+      "step": 16255
+    },
+    {
+      "epoch": 0.39160906529226175,
+      "grad_norm": 1.508278727531433,
+      "learning_rate": 6.4045228678633e-07,
+      "loss": 0.8992,
+      "step": 16260
+    },
+    {
+      "epoch": 0.391729486284049,
+      "grad_norm": 1.5345510244369507,
+      "learning_rate": 6.403255247946455e-07,
+      "loss": 0.9355,
+      "step": 16265
+    },
+    {
+      "epoch": 0.39184990727583635,
+      "grad_norm": 1.515647292137146,
+      "learning_rate": 6.401987628029612e-07,
+      "loss": 0.9122,
+      "step": 16270
+    },
+    {
+      "epoch": 0.3919703282676236,
+      "grad_norm": 1.5095194578170776,
+      "learning_rate": 6.400720008112768e-07,
+      "loss": 0.9869,
+      "step": 16275
+    },
+    {
+      "epoch": 0.3920907492594109,
+      "grad_norm": 1.6890863180160522,
+      "learning_rate": 6.399452388195922e-07,
+      "loss": 0.9405,
+      "step": 16280
+    },
+    {
+      "epoch": 0.39221117025119817,
+      "grad_norm": 1.5294418334960938,
+      "learning_rate": 6.398184768279079e-07,
+      "loss": 0.9011,
+      "step": 16285
+    },
+    {
+      "epoch": 0.3923315912429855,
+      "grad_norm": 1.4933300018310547,
+      "learning_rate": 6.396917148362235e-07,
+      "loss": 0.9059,
+      "step": 16290
+    },
+    {
+      "epoch": 0.3924520122347728,
+      "grad_norm": 1.4482684135437012,
+      "learning_rate": 6.395649528445391e-07,
+      "loss": 0.8861,
+      "step": 16295
+    },
+    {
+      "epoch": 0.39257243322656005,
+      "grad_norm": 1.3621864318847656,
+      "learning_rate": 6.394381908528546e-07,
+      "loss": 0.9167,
+      "step": 16300
+    },
+    {
+      "epoch": 0.3926928542183473,
+      "grad_norm": 1.451184868812561,
+      "learning_rate": 6.393114288611703e-07,
+      "loss": 0.9173,
+      "step": 16305
+    },
+    {
+      "epoch": 0.39281327521013465,
+      "grad_norm": 1.487359642982483,
+      "learning_rate": 6.391846668694858e-07,
+      "loss": 0.9596,
+      "step": 16310
+    },
+    {
+      "epoch": 0.3929336962019219,
+      "grad_norm": 1.5361380577087402,
+      "learning_rate": 6.390579048778014e-07,
+      "loss": 0.934,
+      "step": 16315
+    },
+    {
+      "epoch": 0.3930541171937092,
+      "grad_norm": 1.566884160041809,
+      "learning_rate": 6.389311428861171e-07,
+      "loss": 0.8644,
+      "step": 16320
+    },
+    {
+      "epoch": 0.39317453818549647,
+      "grad_norm": 1.6171588897705078,
+      "learning_rate": 6.388043808944325e-07,
+      "loss": 0.9374,
+      "step": 16325
+    },
+    {
+      "epoch": 0.3932949591772838,
+      "grad_norm": 1.5238803625106812,
+      "learning_rate": 6.386776189027482e-07,
+      "loss": 0.9359,
+      "step": 16330
+    },
+    {
+      "epoch": 0.39341538016907107,
+      "grad_norm": 1.6145451068878174,
+      "learning_rate": 6.385508569110638e-07,
+      "loss": 0.914,
+      "step": 16335
+    },
+    {
+      "epoch": 0.39353580116085835,
+      "grad_norm": 1.5774810314178467,
+      "learning_rate": 6.384240949193794e-07,
+      "loss": 0.8834,
+      "step": 16340
+    },
+    {
+      "epoch": 0.3936562221526457,
+      "grad_norm": 1.6705057621002197,
+      "learning_rate": 6.382973329276949e-07,
+      "loss": 0.8951,
+      "step": 16345
+    },
+    {
+      "epoch": 0.39377664314443295,
+      "grad_norm": 1.5039398670196533,
+      "learning_rate": 6.381705709360105e-07,
+      "loss": 0.9229,
+      "step": 16350
+    },
+    {
+      "epoch": 0.3938970641362202,
+      "grad_norm": 1.585025429725647,
+      "learning_rate": 6.380438089443261e-07,
+      "loss": 0.9039,
+      "step": 16355
+    },
+    {
+      "epoch": 0.3940174851280075,
+      "grad_norm": 1.6906098127365112,
+      "learning_rate": 6.379170469526417e-07,
+      "loss": 0.9368,
+      "step": 16360
+    },
+    {
+      "epoch": 0.3941379061197948,
+      "grad_norm": 1.5157690048217773,
+      "learning_rate": 6.377902849609574e-07,
+      "loss": 0.8906,
+      "step": 16365
+    },
+    {
+      "epoch": 0.3942583271115821,
+      "grad_norm": 1.4066121578216553,
+      "learning_rate": 6.376635229692728e-07,
+      "loss": 0.9337,
+      "step": 16370
+    },
+    {
+      "epoch": 0.39437874810336937,
+      "grad_norm": 1.4881788492202759,
+      "learning_rate": 6.375367609775884e-07,
+      "loss": 0.9241,
+      "step": 16375
+    },
+    {
+      "epoch": 0.39449916909515664,
+      "grad_norm": 1.500671148300171,
+      "learning_rate": 6.374099989859041e-07,
+      "loss": 0.97,
+      "step": 16380
+    },
+    {
+      "epoch": 0.39461959008694397,
+      "grad_norm": 1.5400736331939697,
+      "learning_rate": 6.372832369942196e-07,
+      "loss": 0.9208,
+      "step": 16385
+    },
+    {
+      "epoch": 0.39474001107873125,
+      "grad_norm": 1.468634009361267,
+      "learning_rate": 6.371564750025353e-07,
+      "loss": 0.9247,
+      "step": 16390
+    },
+    {
+      "epoch": 0.3948604320705185,
+      "grad_norm": 1.510955572128296,
+      "learning_rate": 6.370297130108508e-07,
+      "loss": 0.9041,
+      "step": 16395
+    },
+    {
+      "epoch": 0.39498085306230585,
+      "grad_norm": 1.528254747390747,
+      "learning_rate": 6.369029510191663e-07,
+      "loss": 0.9339,
+      "step": 16400
+    },
+    {
+      "epoch": 0.3951012740540931,
+      "grad_norm": 1.5330232381820679,
+      "learning_rate": 6.36776189027482e-07,
+      "loss": 0.897,
+      "step": 16405
+    },
+    {
+      "epoch": 0.3952216950458804,
+      "grad_norm": 1.4764940738677979,
+      "learning_rate": 6.366494270357976e-07,
+      "loss": 0.9394,
+      "step": 16410
+    },
+    {
+      "epoch": 0.39534211603766767,
+      "grad_norm": 1.4831757545471191,
+      "learning_rate": 6.365226650441131e-07,
+      "loss": 0.9418,
+      "step": 16415
+    },
+    {
+      "epoch": 0.395462537029455,
+      "grad_norm": 1.6943193674087524,
+      "learning_rate": 6.363959030524287e-07,
+      "loss": 0.8861,
+      "step": 16420
+    },
+    {
+      "epoch": 0.39558295802124227,
+      "grad_norm": 1.3982515335083008,
+      "learning_rate": 6.362691410607444e-07,
+      "loss": 0.791,
+      "step": 16425
+    },
+    {
+      "epoch": 0.39570337901302954,
+      "grad_norm": 1.4633845090866089,
+      "learning_rate": 6.361423790690599e-07,
+      "loss": 0.9638,
+      "step": 16430
+    },
+    {
+      "epoch": 0.3958238000048168,
+      "grad_norm": 1.807855486869812,
+      "learning_rate": 6.360156170773755e-07,
+      "loss": 0.941,
+      "step": 16435
+    },
+    {
+      "epoch": 0.39594422099660415,
+      "grad_norm": 1.523313283920288,
+      "learning_rate": 6.358888550856911e-07,
+      "loss": 0.9431,
+      "step": 16440
+    },
+    {
+      "epoch": 0.3960646419883914,
+      "grad_norm": 1.3463391065597534,
+      "learning_rate": 6.357620930940066e-07,
+      "loss": 0.9069,
+      "step": 16445
+    },
+    {
+      "epoch": 0.3961850629801787,
+      "grad_norm": 1.6524784564971924,
+      "learning_rate": 6.356353311023223e-07,
+      "loss": 0.9588,
+      "step": 16450
+    },
+    {
+      "epoch": 0.39630548397196597,
+      "grad_norm": 1.4540703296661377,
+      "learning_rate": 6.355085691106379e-07,
+      "loss": 0.8533,
+      "step": 16455
+    },
+    {
+      "epoch": 0.3964259049637533,
+      "grad_norm": 1.5041921138763428,
+      "learning_rate": 6.353818071189533e-07,
+      "loss": 0.9025,
+      "step": 16460
+    },
+    {
+      "epoch": 0.39654632595554057,
+      "grad_norm": 1.4892079830169678,
+      "learning_rate": 6.35255045127269e-07,
+      "loss": 0.948,
+      "step": 16465
+    },
+    {
+      "epoch": 0.39666674694732784,
+      "grad_norm": 1.909855604171753,
+      "learning_rate": 6.351282831355846e-07,
+      "loss": 0.9106,
+      "step": 16470
+    },
+    {
+      "epoch": 0.39678716793911517,
+      "grad_norm": 1.4165951013565063,
+      "learning_rate": 6.350015211439002e-07,
+      "loss": 0.9051,
+      "step": 16475
+    },
+    {
+      "epoch": 0.39690758893090244,
+      "grad_norm": 1.558835506439209,
+      "learning_rate": 6.348747591522158e-07,
+      "loss": 0.8895,
+      "step": 16480
+    },
+    {
+      "epoch": 0.3970280099226897,
+      "grad_norm": 1.5130789279937744,
+      "learning_rate": 6.347479971605314e-07,
+      "loss": 0.9011,
+      "step": 16485
+    },
+    {
+      "epoch": 0.397148430914477,
+      "grad_norm": 1.735569953918457,
+      "learning_rate": 6.346212351688469e-07,
+      "loss": 0.9343,
+      "step": 16490
+    },
+    {
+      "epoch": 0.3972688519062643,
+      "grad_norm": 1.4600125551223755,
+      "learning_rate": 6.344944731771625e-07,
+      "loss": 0.9466,
+      "step": 16495
+    },
+    {
+      "epoch": 0.3973892728980516,
+      "grad_norm": 1.612957239151001,
+      "learning_rate": 6.343677111854782e-07,
+      "loss": 0.9414,
+      "step": 16500
+    },
+    {
+      "epoch": 0.39750969388983887,
+      "grad_norm": 1.5158154964447021,
+      "learning_rate": 6.342409491937937e-07,
+      "loss": 0.9412,
+      "step": 16505
+    },
+    {
+      "epoch": 0.39763011488162614,
+      "grad_norm": 1.6656270027160645,
+      "learning_rate": 6.341141872021093e-07,
+      "loss": 0.8693,
+      "step": 16510
+    },
+    {
+      "epoch": 0.39775053587341347,
+      "grad_norm": 1.6811892986297607,
+      "learning_rate": 6.339874252104249e-07,
+      "loss": 0.9543,
+      "step": 16515
+    },
+    {
+      "epoch": 0.39787095686520074,
+      "grad_norm": 1.6302272081375122,
+      "learning_rate": 6.338606632187404e-07,
+      "loss": 0.9084,
+      "step": 16520
+    },
+    {
+      "epoch": 0.397991377856988,
+      "grad_norm": 1.7181906700134277,
+      "learning_rate": 6.337339012270561e-07,
+      "loss": 0.8922,
+      "step": 16525
+    },
+    {
+      "epoch": 0.39811179884877534,
+      "grad_norm": 1.523653268814087,
+      "learning_rate": 6.336071392353716e-07,
+      "loss": 0.8399,
+      "step": 16530
+    },
+    {
+      "epoch": 0.3982322198405626,
+      "grad_norm": 1.7446633577346802,
+      "learning_rate": 6.334803772436872e-07,
+      "loss": 0.9446,
+      "step": 16535
+    },
+    {
+      "epoch": 0.3983526408323499,
+      "grad_norm": 1.4260302782058716,
+      "learning_rate": 6.333536152520028e-07,
+      "loss": 0.9209,
+      "step": 16540
+    },
+    {
+      "epoch": 0.39847306182413716,
+      "grad_norm": 1.572892189025879,
+      "learning_rate": 6.332268532603185e-07,
+      "loss": 0.899,
+      "step": 16545
+    },
+    {
+      "epoch": 0.3985934828159245,
+      "grad_norm": 1.5617256164550781,
+      "learning_rate": 6.33100091268634e-07,
+      "loss": 0.909,
+      "step": 16550
+    },
+    {
+      "epoch": 0.39871390380771177,
+      "grad_norm": 1.680518388748169,
+      "learning_rate": 6.329733292769495e-07,
+      "loss": 0.9296,
+      "step": 16555
+    },
+    {
+      "epoch": 0.39883432479949904,
+      "grad_norm": 1.9393974542617798,
+      "learning_rate": 6.328465672852652e-07,
+      "loss": 0.8945,
+      "step": 16560
+    },
+    {
+      "epoch": 0.3989547457912863,
+      "grad_norm": 1.7089143991470337,
+      "learning_rate": 6.327198052935807e-07,
+      "loss": 0.9152,
+      "step": 16565
+    },
+    {
+      "epoch": 0.39907516678307364,
+      "grad_norm": 1.5115058422088623,
+      "learning_rate": 6.325930433018964e-07,
+      "loss": 0.8943,
+      "step": 16570
+    },
+    {
+      "epoch": 0.3991955877748609,
+      "grad_norm": 1.4281383752822876,
+      "learning_rate": 6.32466281310212e-07,
+      "loss": 0.9097,
+      "step": 16575
+    },
+    {
+      "epoch": 0.3993160087666482,
+      "grad_norm": 1.453593373298645,
+      "learning_rate": 6.323395193185274e-07,
+      "loss": 0.9352,
+      "step": 16580
+    },
+    {
+      "epoch": 0.39943642975843546,
+      "grad_norm": 1.7239865064620972,
+      "learning_rate": 6.322127573268431e-07,
+      "loss": 0.9259,
+      "step": 16585
+    },
+    {
+      "epoch": 0.3995568507502228,
+      "grad_norm": 1.3729093074798584,
+      "learning_rate": 6.320859953351587e-07,
+      "loss": 0.9445,
+      "step": 16590
+    },
+    {
+      "epoch": 0.39967727174201007,
+      "grad_norm": 1.594544529914856,
+      "learning_rate": 6.319592333434743e-07,
+      "loss": 0.9346,
+      "step": 16595
+    },
+    {
+      "epoch": 0.39979769273379734,
+      "grad_norm": 1.5372705459594727,
+      "learning_rate": 6.318324713517898e-07,
+      "loss": 0.937,
+      "step": 16600
+    },
+    {
+      "epoch": 0.39991811372558467,
+      "grad_norm": 1.740492343902588,
+      "learning_rate": 6.317057093601055e-07,
+      "loss": 0.955,
+      "step": 16605
+    },
+    {
+      "epoch": 0.40003853471737194,
+      "grad_norm": 1.7151694297790527,
+      "learning_rate": 6.31578947368421e-07,
+      "loss": 0.9433,
+      "step": 16610
+    },
+    {
+      "epoch": 0.4001589557091592,
+      "grad_norm": 1.511307954788208,
+      "learning_rate": 6.314521853767366e-07,
+      "loss": 0.8717,
+      "step": 16615
+    },
+    {
+      "epoch": 0.4002793767009465,
+      "grad_norm": 1.4371942281723022,
+      "learning_rate": 6.313254233850523e-07,
+      "loss": 0.9438,
+      "step": 16620
+    },
+    {
+      "epoch": 0.4003997976927338,
+      "grad_norm": 1.4800249338150024,
+      "learning_rate": 6.311986613933677e-07,
+      "loss": 0.92,
+      "step": 16625
+    },
+    {
+      "epoch": 0.4005202186845211,
+      "grad_norm": 1.5603761672973633,
+      "learning_rate": 6.310718994016834e-07,
+      "loss": 0.9322,
+      "step": 16630
+    },
+    {
+      "epoch": 0.40064063967630836,
+      "grad_norm": 1.6178046464920044,
+      "learning_rate": 6.30945137409999e-07,
+      "loss": 0.8687,
+      "step": 16635
+    },
+    {
+      "epoch": 0.40076106066809564,
+      "grad_norm": 1.4481092691421509,
+      "learning_rate": 6.308183754183145e-07,
+      "loss": 0.9405,
+      "step": 16640
+    },
+    {
+      "epoch": 0.40088148165988297,
+      "grad_norm": 1.5825382471084595,
+      "learning_rate": 6.306916134266301e-07,
+      "loss": 0.8999,
+      "step": 16645
+    },
+    {
+      "epoch": 0.40100190265167024,
+      "grad_norm": 1.569267749786377,
+      "learning_rate": 6.305648514349457e-07,
+      "loss": 0.8915,
+      "step": 16650
+    },
+    {
+      "epoch": 0.4011223236434575,
+      "grad_norm": 1.5229368209838867,
+      "learning_rate": 6.304380894432613e-07,
+      "loss": 0.9088,
+      "step": 16655
+    },
+    {
+      "epoch": 0.40124274463524484,
+      "grad_norm": 1.4642391204833984,
+      "learning_rate": 6.303113274515769e-07,
+      "loss": 0.9022,
+      "step": 16660
+    },
+    {
+      "epoch": 0.4013631656270321,
+      "grad_norm": 1.8505054712295532,
+      "learning_rate": 6.301845654598926e-07,
+      "loss": 0.9742,
+      "step": 16665
+    },
+    {
+      "epoch": 0.4014835866188194,
+      "grad_norm": 1.8605384826660156,
+      "learning_rate": 6.30057803468208e-07,
+      "loss": 0.8914,
+      "step": 16670
+    },
+    {
+      "epoch": 0.40160400761060666,
+      "grad_norm": 1.5216584205627441,
+      "learning_rate": 6.299310414765236e-07,
+      "loss": 0.9265,
+      "step": 16675
+    },
+    {
+      "epoch": 0.401724428602394,
+      "grad_norm": 1.6303132772445679,
+      "learning_rate": 6.298042794848393e-07,
+      "loss": 0.9063,
+      "step": 16680
+    },
+    {
+      "epoch": 0.40184484959418126,
+      "grad_norm": 1.4981777667999268,
+      "learning_rate": 6.296775174931548e-07,
+      "loss": 0.8603,
+      "step": 16685
+    },
+    {
+      "epoch": 0.40196527058596854,
+      "grad_norm": 1.6867685317993164,
+      "learning_rate": 6.295507555014705e-07,
+      "loss": 0.9356,
+      "step": 16690
+    },
+    {
+      "epoch": 0.4020856915777558,
+      "grad_norm": 1.4873652458190918,
+      "learning_rate": 6.29423993509786e-07,
+      "loss": 0.9087,
+      "step": 16695
+    },
+    {
+      "epoch": 0.40220611256954314,
+      "grad_norm": 1.558084487915039,
+      "learning_rate": 6.292972315181015e-07,
+      "loss": 0.9398,
+      "step": 16700
+    },
+    {
+      "epoch": 0.4023265335613304,
+      "grad_norm": 1.403309941291809,
+      "learning_rate": 6.291704695264172e-07,
+      "loss": 0.8758,
+      "step": 16705
+    },
+    {
+      "epoch": 0.4024469545531177,
+      "grad_norm": 1.7472859621047974,
+      "learning_rate": 6.290437075347328e-07,
+      "loss": 0.9222,
+      "step": 16710
+    },
+    {
+      "epoch": 0.402567375544905,
+      "grad_norm": 1.427155613899231,
+      "learning_rate": 6.289169455430483e-07,
+      "loss": 0.8999,
+      "step": 16715
+    },
+    {
+      "epoch": 0.4026877965366923,
+      "grad_norm": 1.5862841606140137,
+      "learning_rate": 6.287901835513639e-07,
+      "loss": 0.8766,
+      "step": 16720
+    },
+    {
+      "epoch": 0.40280821752847956,
+      "grad_norm": 1.6798174381256104,
+      "learning_rate": 6.286634215596795e-07,
+      "loss": 0.9728,
+      "step": 16725
+    },
+    {
+      "epoch": 0.40292863852026684,
+      "grad_norm": 1.4473795890808105,
+      "learning_rate": 6.285366595679951e-07,
+      "loss": 0.898,
+      "step": 16730
+    },
+    {
+      "epoch": 0.40304905951205416,
+      "grad_norm": 1.7120901346206665,
+      "learning_rate": 6.284098975763107e-07,
+      "loss": 0.8887,
+      "step": 16735
+    },
+    {
+      "epoch": 0.40316948050384144,
+      "grad_norm": 1.587019920349121,
+      "learning_rate": 6.282831355846263e-07,
+      "loss": 0.933,
+      "step": 16740
+    },
+    {
+      "epoch": 0.4032899014956287,
+      "grad_norm": 1.3541581630706787,
+      "learning_rate": 6.281563735929418e-07,
+      "loss": 0.8703,
+      "step": 16745
+    },
+    {
+      "epoch": 0.403410322487416,
+      "grad_norm": 1.5999221801757812,
+      "learning_rate": 6.280296116012575e-07,
+      "loss": 0.8818,
+      "step": 16750
+    },
+    {
+      "epoch": 0.4035307434792033,
+      "grad_norm": 1.6486579179763794,
+      "learning_rate": 6.279028496095731e-07,
+      "loss": 0.9107,
+      "step": 16755
+    },
+    {
+      "epoch": 0.4036511644709906,
+      "grad_norm": 1.8040155172348022,
+      "learning_rate": 6.277760876178885e-07,
+      "loss": 0.9138,
+      "step": 16760
+    },
+    {
+      "epoch": 0.40377158546277786,
+      "grad_norm": 1.6447283029556274,
+      "learning_rate": 6.276493256262042e-07,
+      "loss": 0.9212,
+      "step": 16765
+    },
+    {
+      "epoch": 0.40389200645456513,
+      "grad_norm": 1.5155274868011475,
+      "learning_rate": 6.275225636345198e-07,
+      "loss": 0.8841,
+      "step": 16770
+    },
+    {
+      "epoch": 0.40401242744635246,
+      "grad_norm": 1.705776333808899,
+      "learning_rate": 6.273958016428354e-07,
+      "loss": 0.9116,
+      "step": 16775
+    },
+    {
+      "epoch": 0.40413284843813974,
+      "grad_norm": 1.5871044397354126,
+      "learning_rate": 6.27269039651151e-07,
+      "loss": 0.9305,
+      "step": 16780
+    },
+    {
+      "epoch": 0.404253269429927,
+      "grad_norm": 1.5335018634796143,
+      "learning_rate": 6.271422776594665e-07,
+      "loss": 0.9243,
+      "step": 16785
+    },
+    {
+      "epoch": 0.40437369042171434,
+      "grad_norm": 1.5292437076568604,
+      "learning_rate": 6.270155156677821e-07,
+      "loss": 0.8909,
+      "step": 16790
+    },
+    {
+      "epoch": 0.4044941114135016,
+      "grad_norm": 1.6406742334365845,
+      "learning_rate": 6.268887536760977e-07,
+      "loss": 0.9501,
+      "step": 16795
+    },
+    {
+      "epoch": 0.4046145324052889,
+      "grad_norm": 1.462023377418518,
+      "learning_rate": 6.267619916844134e-07,
+      "loss": 0.8919,
+      "step": 16800
+    },
+    {
+      "epoch": 0.40473495339707616,
+      "grad_norm": 1.5037622451782227,
+      "learning_rate": 6.266352296927289e-07,
+      "loss": 0.9235,
+      "step": 16805
+    },
+    {
+      "epoch": 0.4048553743888635,
+      "grad_norm": 1.6942980289459229,
+      "learning_rate": 6.265084677010445e-07,
+      "loss": 0.9217,
+      "step": 16810
+    },
+    {
+      "epoch": 0.40497579538065076,
+      "grad_norm": 1.740375280380249,
+      "learning_rate": 6.263817057093601e-07,
+      "loss": 0.9559,
+      "step": 16815
+    },
+    {
+      "epoch": 0.40509621637243803,
+      "grad_norm": 1.5765180587768555,
+      "learning_rate": 6.262549437176756e-07,
+      "loss": 0.9235,
+      "step": 16820
+    },
+    {
+      "epoch": 0.4052166373642253,
+      "grad_norm": 1.4993369579315186,
+      "learning_rate": 6.261281817259913e-07,
+      "loss": 0.9422,
+      "step": 16825
+    },
+    {
+      "epoch": 0.40533705835601264,
+      "grad_norm": 1.3874576091766357,
+      "learning_rate": 6.260014197343068e-07,
+      "loss": 0.9643,
+      "step": 16830
+    },
+    {
+      "epoch": 0.4054574793477999,
+      "grad_norm": 1.4534552097320557,
+      "learning_rate": 6.258746577426224e-07,
+      "loss": 0.908,
+      "step": 16835
+    },
+    {
+      "epoch": 0.4055779003395872,
+      "grad_norm": 1.5588668584823608,
+      "learning_rate": 6.25747895750938e-07,
+      "loss": 0.8767,
+      "step": 16840
+    },
+    {
+      "epoch": 0.4056983213313745,
+      "grad_norm": 1.445541501045227,
+      "learning_rate": 6.256211337592536e-07,
+      "loss": 0.8934,
+      "step": 16845
+    },
+    {
+      "epoch": 0.4058187423231618,
+      "grad_norm": 1.5574666261672974,
+      "learning_rate": 6.254943717675692e-07,
+      "loss": 0.92,
+      "step": 16850
+    },
+    {
+      "epoch": 0.40593916331494906,
+      "grad_norm": 1.498324990272522,
+      "learning_rate": 6.253676097758847e-07,
+      "loss": 0.8976,
+      "step": 16855
+    },
+    {
+      "epoch": 0.40605958430673633,
+      "grad_norm": 1.4403085708618164,
+      "learning_rate": 6.252408477842004e-07,
+      "loss": 0.9248,
+      "step": 16860
+    },
+    {
+      "epoch": 0.40618000529852366,
+      "grad_norm": 1.6509935855865479,
+      "learning_rate": 6.251140857925159e-07,
+      "loss": 0.9398,
+      "step": 16865
+    },
+    {
+      "epoch": 0.40630042629031093,
+      "grad_norm": 1.5441958904266357,
+      "learning_rate": 6.249873238008316e-07,
+      "loss": 0.9183,
+      "step": 16870
+    },
+    {
+      "epoch": 0.4064208472820982,
+      "grad_norm": 1.4594789743423462,
+      "learning_rate": 6.248605618091472e-07,
+      "loss": 0.9768,
+      "step": 16875
+    },
+    {
+      "epoch": 0.4065412682738855,
+      "grad_norm": 1.571266770362854,
+      "learning_rate": 6.247337998174626e-07,
+      "loss": 0.8789,
+      "step": 16880
+    },
+    {
+      "epoch": 0.4066616892656728,
+      "grad_norm": 1.6715950965881348,
+      "learning_rate": 6.246070378257783e-07,
+      "loss": 0.8904,
+      "step": 16885
+    },
+    {
+      "epoch": 0.4067821102574601,
+      "grad_norm": 1.7333935499191284,
+      "learning_rate": 6.244802758340939e-07,
+      "loss": 0.9722,
+      "step": 16890
+    },
+    {
+      "epoch": 0.40690253124924736,
+      "grad_norm": 1.4930369853973389,
+      "learning_rate": 6.243535138424095e-07,
+      "loss": 0.879,
+      "step": 16895
+    },
+    {
+      "epoch": 0.40702295224103463,
+      "grad_norm": 1.5462515354156494,
+      "learning_rate": 6.24226751850725e-07,
+      "loss": 0.8793,
+      "step": 16900
+    },
+    {
+      "epoch": 0.40714337323282196,
+      "grad_norm": 1.6213178634643555,
+      "learning_rate": 6.240999898590406e-07,
+      "loss": 0.9315,
+      "step": 16905
+    },
+    {
+      "epoch": 0.40726379422460923,
+      "grad_norm": 1.5512969493865967,
+      "learning_rate": 6.239732278673562e-07,
+      "loss": 0.9116,
+      "step": 16910
+    },
+    {
+      "epoch": 0.4073842152163965,
+      "grad_norm": 1.5854614973068237,
+      "learning_rate": 6.238464658756718e-07,
+      "loss": 0.9105,
+      "step": 16915
+    },
+    {
+      "epoch": 0.40750463620818383,
+      "grad_norm": 1.4322234392166138,
+      "learning_rate": 6.237197038839875e-07,
+      "loss": 0.9013,
+      "step": 16920
+    },
+    {
+      "epoch": 0.4076250571999711,
+      "grad_norm": 1.5661823749542236,
+      "learning_rate": 6.235929418923029e-07,
+      "loss": 0.887,
+      "step": 16925
+    },
+    {
+      "epoch": 0.4077454781917584,
+      "grad_norm": 1.6415965557098389,
+      "learning_rate": 6.234661799006185e-07,
+      "loss": 0.8904,
+      "step": 16930
+    },
+    {
+      "epoch": 0.40786589918354565,
+      "grad_norm": 1.5648971796035767,
+      "learning_rate": 6.233394179089342e-07,
+      "loss": 0.9167,
+      "step": 16935
+    },
+    {
+      "epoch": 0.407986320175333,
+      "grad_norm": 1.5005626678466797,
+      "learning_rate": 6.232126559172497e-07,
+      "loss": 0.937,
+      "step": 16940
+    },
+    {
+      "epoch": 0.40810674116712026,
+      "grad_norm": 1.5273611545562744,
+      "learning_rate": 6.230858939255653e-07,
+      "loss": 0.9338,
+      "step": 16945
+    },
+    {
+      "epoch": 0.40822716215890753,
+      "grad_norm": 1.5377509593963623,
+      "learning_rate": 6.229591319338809e-07,
+      "loss": 0.9068,
+      "step": 16950
+    },
+    {
+      "epoch": 0.4083475831506948,
+      "grad_norm": 1.4508482217788696,
+      "learning_rate": 6.228323699421965e-07,
+      "loss": 0.9001,
+      "step": 16955
+    },
+    {
+      "epoch": 0.40846800414248213,
+      "grad_norm": 1.6953697204589844,
+      "learning_rate": 6.227056079505121e-07,
+      "loss": 0.9195,
+      "step": 16960
+    },
+    {
+      "epoch": 0.4085884251342694,
+      "grad_norm": 1.6012778282165527,
+      "learning_rate": 6.225788459588277e-07,
+      "loss": 0.8474,
+      "step": 16965
+    },
+    {
+      "epoch": 0.4087088461260567,
+      "grad_norm": 1.5923460721969604,
+      "learning_rate": 6.224520839671432e-07,
+      "loss": 0.9111,
+      "step": 16970
+    },
+    {
+      "epoch": 0.408829267117844,
+      "grad_norm": 1.6092580556869507,
+      "learning_rate": 6.223253219754588e-07,
+      "loss": 0.8893,
+      "step": 16975
+    },
+    {
+      "epoch": 0.4089496881096313,
+      "grad_norm": 1.7636964321136475,
+      "learning_rate": 6.221985599837745e-07,
+      "loss": 0.9521,
+      "step": 16980
+    },
+    {
+      "epoch": 0.40907010910141856,
+      "grad_norm": 1.5564132928848267,
+      "learning_rate": 6.2207179799209e-07,
+      "loss": 0.9026,
+      "step": 16985
+    },
+    {
+      "epoch": 0.40919053009320583,
+      "grad_norm": 1.6354960203170776,
+      "learning_rate": 6.219450360004056e-07,
+      "loss": 0.9423,
+      "step": 16990
+    },
+    {
+      "epoch": 0.40931095108499316,
+      "grad_norm": 1.5429902076721191,
+      "learning_rate": 6.218182740087212e-07,
+      "loss": 0.9144,
+      "step": 16995
+    },
+    {
+      "epoch": 0.40943137207678043,
+      "grad_norm": 2.275571823120117,
+      "learning_rate": 6.216915120170367e-07,
+      "loss": 0.9458,
+      "step": 17000
+    },
+    {
+      "epoch": 0.4095517930685677,
+      "grad_norm": 1.4017913341522217,
+      "learning_rate": 6.215647500253524e-07,
+      "loss": 0.9419,
+      "step": 17005
+    },
+    {
+      "epoch": 0.409672214060355,
+      "grad_norm": 1.5098503828048706,
+      "learning_rate": 6.21437988033668e-07,
+      "loss": 0.9175,
+      "step": 17010
+    },
+    {
+      "epoch": 0.4097926350521423,
+      "grad_norm": 1.4720624685287476,
+      "learning_rate": 6.213112260419834e-07,
+      "loss": 0.8629,
+      "step": 17015
+    },
+    {
+      "epoch": 0.4099130560439296,
+      "grad_norm": 1.7012861967086792,
+      "learning_rate": 6.211844640502991e-07,
+      "loss": 0.9614,
+      "step": 17020
+    },
+    {
+      "epoch": 0.41003347703571685,
+      "grad_norm": 1.732938528060913,
+      "learning_rate": 6.210577020586147e-07,
+      "loss": 0.9399,
+      "step": 17025
+    },
+    {
+      "epoch": 0.4101538980275041,
+      "grad_norm": 1.5244909524917603,
+      "learning_rate": 6.209309400669303e-07,
+      "loss": 0.8768,
+      "step": 17030
+    },
+    {
+      "epoch": 0.41027431901929146,
+      "grad_norm": 1.501592755317688,
+      "learning_rate": 6.208041780752459e-07,
+      "loss": 0.8713,
+      "step": 17035
+    },
+    {
+      "epoch": 0.41039474001107873,
+      "grad_norm": 1.55648934841156,
+      "learning_rate": 6.206774160835615e-07,
+      "loss": 0.9467,
+      "step": 17040
+    },
+    {
+      "epoch": 0.410515161002866,
+      "grad_norm": 1.4854168891906738,
+      "learning_rate": 6.20550654091877e-07,
+      "loss": 0.9206,
+      "step": 17045
+    },
+    {
+      "epoch": 0.41063558199465333,
+      "grad_norm": 1.598010540008545,
+      "learning_rate": 6.204238921001926e-07,
+      "loss": 0.892,
+      "step": 17050
+    },
+    {
+      "epoch": 0.4107560029864406,
+      "grad_norm": 1.6800447702407837,
+      "learning_rate": 6.202971301085083e-07,
+      "loss": 0.8823,
+      "step": 17055
+    },
+    {
+      "epoch": 0.4108764239782279,
+      "grad_norm": 1.541235327720642,
+      "learning_rate": 6.201703681168237e-07,
+      "loss": 0.9197,
+      "step": 17060
+    },
+    {
+      "epoch": 0.41099684497001515,
+      "grad_norm": 1.558349370956421,
+      "learning_rate": 6.200436061251394e-07,
+      "loss": 0.889,
+      "step": 17065
+    },
+    {
+      "epoch": 0.4111172659618025,
+      "grad_norm": 1.6211495399475098,
+      "learning_rate": 6.19916844133455e-07,
+      "loss": 0.9126,
+      "step": 17070
+    },
+    {
+      "epoch": 0.41123768695358975,
+      "grad_norm": 1.5009198188781738,
+      "learning_rate": 6.197900821417706e-07,
+      "loss": 0.9916,
+      "step": 17075
+    },
+    {
+      "epoch": 0.411358107945377,
+      "grad_norm": 1.7271745204925537,
+      "learning_rate": 6.196633201500862e-07,
+      "loss": 0.8982,
+      "step": 17080
+    },
+    {
+      "epoch": 0.4114785289371643,
+      "grad_norm": 1.4661312103271484,
+      "learning_rate": 6.195365581584017e-07,
+      "loss": 0.9097,
+      "step": 17085
+    },
+    {
+      "epoch": 0.41159894992895163,
+      "grad_norm": 1.656012773513794,
+      "learning_rate": 6.194097961667173e-07,
+      "loss": 0.9434,
+      "step": 17090
+    },
+    {
+      "epoch": 0.4117193709207389,
+      "grad_norm": 1.7758303880691528,
+      "learning_rate": 6.192830341750329e-07,
+      "loss": 0.9637,
+      "step": 17095
+    },
+    {
+      "epoch": 0.4118397919125262,
+      "grad_norm": 1.5642658472061157,
+      "learning_rate": 6.191562721833486e-07,
+      "loss": 0.9356,
+      "step": 17100
+    },
+    {
+      "epoch": 0.4119602129043135,
+      "grad_norm": 1.6727337837219238,
+      "learning_rate": 6.190295101916641e-07,
+      "loss": 0.958,
+      "step": 17105
+    },
+    {
+      "epoch": 0.4120806338961008,
+      "grad_norm": 1.5685449838638306,
+      "learning_rate": 6.189027481999796e-07,
+      "loss": 0.8907,
+      "step": 17110
+    },
+    {
+      "epoch": 0.41220105488788805,
+      "grad_norm": 1.5111569166183472,
+      "learning_rate": 6.187759862082953e-07,
+      "loss": 0.9169,
+      "step": 17115
+    },
+    {
+      "epoch": 0.4123214758796753,
+      "grad_norm": 1.5179859399795532,
+      "learning_rate": 6.186492242166108e-07,
+      "loss": 0.8996,
+      "step": 17120
+    },
+    {
+      "epoch": 0.41244189687146265,
+      "grad_norm": 1.6612423658370972,
+      "learning_rate": 6.185224622249265e-07,
+      "loss": 0.9161,
+      "step": 17125
+    },
+    {
+      "epoch": 0.4125623178632499,
+      "grad_norm": 1.5869808197021484,
+      "learning_rate": 6.18395700233242e-07,
+      "loss": 0.9005,
+      "step": 17130
+    },
+    {
+      "epoch": 0.4126827388550372,
+      "grad_norm": 1.490541696548462,
+      "learning_rate": 6.182689382415575e-07,
+      "loss": 0.949,
+      "step": 17135
+    },
+    {
+      "epoch": 0.4128031598468245,
+      "grad_norm": 1.481579065322876,
+      "learning_rate": 6.181421762498732e-07,
+      "loss": 0.8829,
+      "step": 17140
+    },
+    {
+      "epoch": 0.4129235808386118,
+      "grad_norm": 1.6850212812423706,
+      "learning_rate": 6.180154142581888e-07,
+      "loss": 0.8974,
+      "step": 17145
+    },
+    {
+      "epoch": 0.4130440018303991,
+      "grad_norm": 1.5239510536193848,
+      "learning_rate": 6.178886522665044e-07,
+      "loss": 0.9171,
+      "step": 17150
+    },
+    {
+      "epoch": 0.41316442282218635,
+      "grad_norm": 1.6153520345687866,
+      "learning_rate": 6.177618902748199e-07,
+      "loss": 0.9529,
+      "step": 17155
+    },
+    {
+      "epoch": 0.4132848438139737,
+      "grad_norm": 1.5189660787582397,
+      "learning_rate": 6.176351282831356e-07,
+      "loss": 0.9021,
+      "step": 17160
+    },
+    {
+      "epoch": 0.41340526480576095,
+      "grad_norm": 1.436489462852478,
+      "learning_rate": 6.175083662914511e-07,
+      "loss": 0.9482,
+      "step": 17165
+    },
+    {
+      "epoch": 0.4135256857975482,
+      "grad_norm": 1.5382148027420044,
+      "learning_rate": 6.173816042997667e-07,
+      "loss": 0.8656,
+      "step": 17170
+    },
+    {
+      "epoch": 0.4136461067893355,
+      "grad_norm": 1.4603832960128784,
+      "learning_rate": 6.172548423080824e-07,
+      "loss": 0.9581,
+      "step": 17175
+    },
+    {
+      "epoch": 0.41376652778112283,
+      "grad_norm": 1.5139708518981934,
+      "learning_rate": 6.171280803163979e-07,
+      "loss": 0.8897,
+      "step": 17180
+    },
+    {
+      "epoch": 0.4138869487729101,
+      "grad_norm": 1.4182488918304443,
+      "learning_rate": 6.170013183247135e-07,
+      "loss": 0.9158,
+      "step": 17185
+    },
+    {
+      "epoch": 0.4140073697646974,
+      "grad_norm": 1.9594801664352417,
+      "learning_rate": 6.168745563330291e-07,
+      "loss": 0.9053,
+      "step": 17190
+    },
+    {
+      "epoch": 0.41412779075648465,
+      "grad_norm": 1.588879108428955,
+      "learning_rate": 6.167477943413448e-07,
+      "loss": 0.8981,
+      "step": 17195
+    },
+    {
+      "epoch": 0.414248211748272,
+      "grad_norm": 1.7074207067489624,
+      "learning_rate": 6.166210323496602e-07,
+      "loss": 0.8977,
+      "step": 17200
+    },
+    {
+      "epoch": 0.41436863274005925,
+      "grad_norm": 1.6250323057174683,
+      "learning_rate": 6.164942703579758e-07,
+      "loss": 0.8861,
+      "step": 17205
+    },
+    {
+      "epoch": 0.4144890537318465,
+      "grad_norm": 1.850350022315979,
+      "learning_rate": 6.163675083662915e-07,
+      "loss": 0.8796,
+      "step": 17210
+    },
+    {
+      "epoch": 0.4146094747236338,
+      "grad_norm": 1.6797027587890625,
+      "learning_rate": 6.16240746374607e-07,
+      "loss": 0.8986,
+      "step": 17215
+    },
+    {
+      "epoch": 0.4147298957154211,
+      "grad_norm": 1.584898591041565,
+      "learning_rate": 6.161139843829227e-07,
+      "loss": 0.951,
+      "step": 17220
+    },
+    {
+      "epoch": 0.4148503167072084,
+      "grad_norm": 1.5470921993255615,
+      "learning_rate": 6.159872223912382e-07,
+      "loss": 0.9126,
+      "step": 17225
+    },
+    {
+      "epoch": 0.4149707376989957,
+      "grad_norm": 1.5832595825195312,
+      "learning_rate": 6.158604603995537e-07,
+      "loss": 0.9547,
+      "step": 17230
+    },
+    {
+      "epoch": 0.415091158690783,
+      "grad_norm": 1.6913233995437622,
+      "learning_rate": 6.157336984078694e-07,
+      "loss": 0.8602,
+      "step": 17235
+    },
+    {
+      "epoch": 0.4152115796825703,
+      "grad_norm": 1.8051854372024536,
+      "learning_rate": 6.15606936416185e-07,
+      "loss": 0.9028,
+      "step": 17240
+    },
+    {
+      "epoch": 0.41533200067435755,
+      "grad_norm": 1.5922828912734985,
+      "learning_rate": 6.154801744245005e-07,
+      "loss": 0.9325,
+      "step": 17245
+    },
+    {
+      "epoch": 0.4154524216661448,
+      "grad_norm": 1.8161146640777588,
+      "learning_rate": 6.153534124328161e-07,
+      "loss": 0.9126,
+      "step": 17250
+    },
+    {
+      "epoch": 0.41557284265793215,
+      "grad_norm": 1.5556491613388062,
+      "learning_rate": 6.152266504411317e-07,
+      "loss": 0.9627,
+      "step": 17255
+    },
+    {
+      "epoch": 0.4156932636497194,
+      "grad_norm": 1.4069287776947021,
+      "learning_rate": 6.150998884494473e-07,
+      "loss": 0.9157,
+      "step": 17260
+    },
+    {
+      "epoch": 0.4158136846415067,
+      "grad_norm": 1.5378355979919434,
+      "learning_rate": 6.149731264577629e-07,
+      "loss": 0.9711,
+      "step": 17265
+    },
+    {
+      "epoch": 0.41593410563329397,
+      "grad_norm": 1.4258906841278076,
+      "learning_rate": 6.148463644660785e-07,
+      "loss": 0.8693,
+      "step": 17270
+    },
+    {
+      "epoch": 0.4160545266250813,
+      "grad_norm": 1.4657279253005981,
+      "learning_rate": 6.14719602474394e-07,
+      "loss": 0.9234,
+      "step": 17275
+    },
+    {
+      "epoch": 0.4161749476168686,
+      "grad_norm": 1.731824517250061,
+      "learning_rate": 6.145928404827097e-07,
+      "loss": 0.8964,
+      "step": 17280
+    },
+    {
+      "epoch": 0.41629536860865585,
+      "grad_norm": 1.8693464994430542,
+      "learning_rate": 6.144660784910253e-07,
+      "loss": 0.9262,
+      "step": 17285
+    },
+    {
+      "epoch": 0.4164157896004432,
+      "grad_norm": 1.500510811805725,
+      "learning_rate": 6.143393164993408e-07,
+      "loss": 0.9205,
+      "step": 17290
+    },
+    {
+      "epoch": 0.41653621059223045,
+      "grad_norm": 1.625565767288208,
+      "learning_rate": 6.142125545076564e-07,
+      "loss": 0.9148,
+      "step": 17295
+    },
+    {
+      "epoch": 0.4166566315840177,
+      "grad_norm": 1.5997830629348755,
+      "learning_rate": 6.14085792515972e-07,
+      "loss": 0.9111,
+      "step": 17300
+    },
+    {
+      "epoch": 0.416777052575805,
+      "grad_norm": 1.6735154390335083,
+      "learning_rate": 6.139590305242876e-07,
+      "loss": 0.9583,
+      "step": 17305
+    },
+    {
+      "epoch": 0.4168974735675923,
+      "grad_norm": 1.5538638830184937,
+      "learning_rate": 6.138322685326032e-07,
+      "loss": 0.9632,
+      "step": 17310
+    },
+    {
+      "epoch": 0.4170178945593796,
+      "grad_norm": 1.4466016292572021,
+      "learning_rate": 6.137055065409187e-07,
+      "loss": 0.9088,
+      "step": 17315
+    },
+    {
+      "epoch": 0.41713831555116687,
+      "grad_norm": 1.5244786739349365,
+      "learning_rate": 6.135787445492343e-07,
+      "loss": 0.9279,
+      "step": 17320
+    },
+    {
+      "epoch": 0.41725873654295415,
+      "grad_norm": 1.305046796798706,
+      "learning_rate": 6.134519825575499e-07,
+      "loss": 0.9045,
+      "step": 17325
+    },
+    {
+      "epoch": 0.4173791575347415,
+      "grad_norm": 1.670288324356079,
+      "learning_rate": 6.133252205658656e-07,
+      "loss": 0.8701,
+      "step": 17330
+    },
+    {
+      "epoch": 0.41749957852652875,
+      "grad_norm": 1.6538128852844238,
+      "learning_rate": 6.131984585741811e-07,
+      "loss": 0.9362,
+      "step": 17335
+    },
+    {
+      "epoch": 0.417619999518316,
+      "grad_norm": 1.5970174074172974,
+      "learning_rate": 6.130716965824967e-07,
+      "loss": 0.9376,
+      "step": 17340
+    },
+    {
+      "epoch": 0.4177404205101033,
+      "grad_norm": 1.530699372291565,
+      "learning_rate": 6.129449345908123e-07,
+      "loss": 0.9649,
+      "step": 17345
+    },
+    {
+      "epoch": 0.4178608415018906,
+      "grad_norm": 1.6150037050247192,
+      "learning_rate": 6.128181725991278e-07,
+      "loss": 0.9158,
+      "step": 17350
+    },
+    {
+      "epoch": 0.4179812624936779,
+      "grad_norm": 1.4796768426895142,
+      "learning_rate": 6.126914106074435e-07,
+      "loss": 0.9045,
+      "step": 17355
+    },
+    {
+      "epoch": 0.41810168348546517,
+      "grad_norm": 1.542307734489441,
+      "learning_rate": 6.125646486157591e-07,
+      "loss": 0.8482,
+      "step": 17360
+    },
+    {
+      "epoch": 0.4182221044772525,
+      "grad_norm": 1.6469286680221558,
+      "learning_rate": 6.124378866240746e-07,
+      "loss": 0.929,
+      "step": 17365
+    },
+    {
+      "epoch": 0.41834252546903977,
+      "grad_norm": 1.5979429483413696,
+      "learning_rate": 6.123111246323902e-07,
+      "loss": 0.9167,
+      "step": 17370
+    },
+    {
+      "epoch": 0.41846294646082705,
+      "grad_norm": 1.4680538177490234,
+      "learning_rate": 6.121843626407058e-07,
+      "loss": 0.8667,
+      "step": 17375
+    },
+    {
+      "epoch": 0.4185833674526143,
+      "grad_norm": 1.752623200416565,
+      "learning_rate": 6.120576006490214e-07,
+      "loss": 0.9389,
+      "step": 17380
+    },
+    {
+      "epoch": 0.41870378844440165,
+      "grad_norm": 1.4785820245742798,
+      "learning_rate": 6.119308386573369e-07,
+      "loss": 0.8957,
+      "step": 17385
+    },
+    {
+      "epoch": 0.4188242094361889,
+      "grad_norm": 1.513867735862732,
+      "learning_rate": 6.118040766656526e-07,
+      "loss": 0.9167,
+      "step": 17390
+    },
+    {
+      "epoch": 0.4189446304279762,
+      "grad_norm": 1.5811598300933838,
+      "learning_rate": 6.116773146739681e-07,
+      "loss": 0.9019,
+      "step": 17395
+    },
+    {
+      "epoch": 0.41906505141976347,
+      "grad_norm": 1.643618106842041,
+      "learning_rate": 6.115505526822838e-07,
+      "loss": 0.9424,
+      "step": 17400
+    },
+    {
+      "epoch": 0.4191854724115508,
+      "grad_norm": 1.3657090663909912,
+      "learning_rate": 6.114237906905994e-07,
+      "loss": 0.9588,
+      "step": 17405
+    },
+    {
+      "epoch": 0.41930589340333807,
+      "grad_norm": 1.5041183233261108,
+      "learning_rate": 6.112970286989148e-07,
+      "loss": 0.8784,
+      "step": 17410
+    },
+    {
+      "epoch": 0.41942631439512534,
+      "grad_norm": 1.5289888381958008,
+      "learning_rate": 6.111702667072305e-07,
+      "loss": 0.8804,
+      "step": 17415
+    },
+    {
+      "epoch": 0.4195467353869127,
+      "grad_norm": 1.574815034866333,
+      "learning_rate": 6.110435047155461e-07,
+      "loss": 0.9201,
+      "step": 17420
+    },
+    {
+      "epoch": 0.41966715637869995,
+      "grad_norm": 1.6267809867858887,
+      "learning_rate": 6.109167427238617e-07,
+      "loss": 0.8833,
+      "step": 17425
+    },
+    {
+      "epoch": 0.4197875773704872,
+      "grad_norm": 1.4753605127334595,
+      "learning_rate": 6.107899807321772e-07,
+      "loss": 0.9439,
+      "step": 17430
+    },
+    {
+      "epoch": 0.4199079983622745,
+      "grad_norm": 1.5669854879379272,
+      "learning_rate": 6.106632187404928e-07,
+      "loss": 0.9109,
+      "step": 17435
+    },
+    {
+      "epoch": 0.4200284193540618,
+      "grad_norm": 1.5805938243865967,
+      "learning_rate": 6.105364567488084e-07,
+      "loss": 0.9097,
+      "step": 17440
+    },
+    {
+      "epoch": 0.4201488403458491,
+      "grad_norm": 1.501623272895813,
+      "learning_rate": 6.10409694757124e-07,
+      "loss": 0.9498,
+      "step": 17445
+    },
+    {
+      "epoch": 0.42026926133763637,
+      "grad_norm": 1.611153483390808,
+      "learning_rate": 6.102829327654397e-07,
+      "loss": 0.9448,
+      "step": 17450
+    },
+    {
+      "epoch": 0.42038968232942364,
+      "grad_norm": 1.4031232595443726,
+      "learning_rate": 6.101561707737551e-07,
+      "loss": 0.9303,
+      "step": 17455
+    },
+    {
+      "epoch": 0.42051010332121097,
+      "grad_norm": 1.5860594511032104,
+      "learning_rate": 6.100294087820707e-07,
+      "loss": 0.9008,
+      "step": 17460
+    },
+    {
+      "epoch": 0.42063052431299824,
+      "grad_norm": 1.665428876876831,
+      "learning_rate": 6.099026467903864e-07,
+      "loss": 0.9014,
+      "step": 17465
+    },
+    {
+      "epoch": 0.4207509453047855,
+      "grad_norm": 1.5833650827407837,
+      "learning_rate": 6.097758847987019e-07,
+      "loss": 0.9349,
+      "step": 17470
+    },
+    {
+      "epoch": 0.4208713662965728,
+      "grad_norm": 1.6934082508087158,
+      "learning_rate": 6.096491228070176e-07,
+      "loss": 0.9126,
+      "step": 17475
+    },
+    {
+      "epoch": 0.4209917872883601,
+      "grad_norm": 1.7227505445480347,
+      "learning_rate": 6.095223608153331e-07,
+      "loss": 1.0,
+      "step": 17480
+    },
+    {
+      "epoch": 0.4211122082801474,
+      "grad_norm": 1.4861100912094116,
+      "learning_rate": 6.093955988236487e-07,
+      "loss": 0.9235,
+      "step": 17485
+    },
+    {
+      "epoch": 0.42123262927193467,
+      "grad_norm": 1.6223926544189453,
+      "learning_rate": 6.092688368319643e-07,
+      "loss": 0.9327,
+      "step": 17490
+    },
+    {
+      "epoch": 0.421353050263722,
+      "grad_norm": 1.5731300115585327,
+      "learning_rate": 6.091420748402799e-07,
+      "loss": 0.9212,
+      "step": 17495
+    },
+    {
+      "epoch": 0.42147347125550927,
+      "grad_norm": 1.519948959350586,
+      "learning_rate": 6.090153128485954e-07,
+      "loss": 0.8683,
+      "step": 17500
+    },
+    {
+      "epoch": 0.42159389224729654,
+      "grad_norm": 1.383150339126587,
+      "learning_rate": 6.08888550856911e-07,
+      "loss": 0.9131,
+      "step": 17505
+    },
+    {
+      "epoch": 0.4217143132390838,
+      "grad_norm": 1.4093159437179565,
+      "learning_rate": 6.087617888652267e-07,
+      "loss": 0.9419,
+      "step": 17510
+    },
+    {
+      "epoch": 0.42183473423087114,
+      "grad_norm": 1.5490443706512451,
+      "learning_rate": 6.086350268735422e-07,
+      "loss": 0.9289,
+      "step": 17515
+    },
+    {
+      "epoch": 0.4219551552226584,
+      "grad_norm": 1.570052981376648,
+      "learning_rate": 6.085082648818578e-07,
+      "loss": 0.9024,
+      "step": 17520
+    },
+    {
+      "epoch": 0.4220755762144457,
+      "grad_norm": 1.583708643913269,
+      "learning_rate": 6.083815028901734e-07,
+      "loss": 0.8999,
+      "step": 17525
+    },
+    {
+      "epoch": 0.42219599720623296,
+      "grad_norm": 1.38755464553833,
+      "learning_rate": 6.082547408984889e-07,
+      "loss": 0.9087,
+      "step": 17530
+    },
+    {
+      "epoch": 0.4223164181980203,
+      "grad_norm": 1.5139927864074707,
+      "learning_rate": 6.081279789068046e-07,
+      "loss": 0.9185,
+      "step": 17535
+    },
+    {
+      "epoch": 0.42243683918980757,
+      "grad_norm": 1.3745379447937012,
+      "learning_rate": 6.080012169151202e-07,
+      "loss": 0.9573,
+      "step": 17540
+    },
+    {
+      "epoch": 0.42255726018159484,
+      "grad_norm": 1.3747448921203613,
+      "learning_rate": 6.078744549234356e-07,
+      "loss": 0.9007,
+      "step": 17545
+    },
+    {
+      "epoch": 0.42267768117338217,
+      "grad_norm": 1.657645344734192,
+      "learning_rate": 6.077476929317513e-07,
+      "loss": 0.927,
+      "step": 17550
+    },
+    {
+      "epoch": 0.42279810216516944,
+      "grad_norm": 1.5407798290252686,
+      "learning_rate": 6.076209309400669e-07,
+      "loss": 0.8979,
+      "step": 17555
+    },
+    {
+      "epoch": 0.4229185231569567,
+      "grad_norm": 1.5311154127120972,
+      "learning_rate": 6.074941689483825e-07,
+      "loss": 0.8862,
+      "step": 17560
+    },
+    {
+      "epoch": 0.423038944148744,
+      "grad_norm": 1.4764913320541382,
+      "learning_rate": 6.073674069566981e-07,
+      "loss": 0.8988,
+      "step": 17565
+    },
+    {
+      "epoch": 0.4231593651405313,
+      "grad_norm": 1.3411235809326172,
+      "learning_rate": 6.072406449650137e-07,
+      "loss": 0.9283,
+      "step": 17570
+    },
+    {
+      "epoch": 0.4232797861323186,
+      "grad_norm": 1.579624891281128,
+      "learning_rate": 6.071138829733292e-07,
+      "loss": 0.9296,
+      "step": 17575
+    },
+    {
+      "epoch": 0.42340020712410587,
+      "grad_norm": 1.8971341848373413,
+      "learning_rate": 6.069871209816448e-07,
+      "loss": 0.9012,
+      "step": 17580
+    },
+    {
+      "epoch": 0.42352062811589314,
+      "grad_norm": 1.5456421375274658,
+      "learning_rate": 6.068603589899605e-07,
+      "loss": 0.8915,
+      "step": 17585
+    },
+    {
+      "epoch": 0.42364104910768047,
+      "grad_norm": 1.4029710292816162,
+      "learning_rate": 6.067335969982759e-07,
+      "loss": 0.9787,
+      "step": 17590
+    },
+    {
+      "epoch": 0.42376147009946774,
+      "grad_norm": 1.7143163681030273,
+      "learning_rate": 6.066068350065916e-07,
+      "loss": 0.8726,
+      "step": 17595
+    },
+    {
+      "epoch": 0.423881891091255,
+      "grad_norm": 1.680330514907837,
+      "learning_rate": 6.064800730149072e-07,
+      "loss": 0.9316,
+      "step": 17600
+    },
+    {
+      "epoch": 0.42400231208304234,
+      "grad_norm": 1.5830532312393188,
+      "learning_rate": 6.063533110232228e-07,
+      "loss": 0.9317,
+      "step": 17605
+    },
+    {
+      "epoch": 0.4241227330748296,
+      "grad_norm": 1.6646406650543213,
+      "learning_rate": 6.062265490315384e-07,
+      "loss": 0.9463,
+      "step": 17610
+    },
+    {
+      "epoch": 0.4242431540666169,
+      "grad_norm": 1.6631529331207275,
+      "learning_rate": 6.060997870398539e-07,
+      "loss": 0.9289,
+      "step": 17615
+    },
+    {
+      "epoch": 0.42436357505840416,
+      "grad_norm": 1.5877068042755127,
+      "learning_rate": 6.059730250481695e-07,
+      "loss": 0.8933,
+      "step": 17620
+    },
+    {
+      "epoch": 0.4244839960501915,
+      "grad_norm": 1.5696178674697876,
+      "learning_rate": 6.058462630564851e-07,
+      "loss": 0.9762,
+      "step": 17625
+    },
+    {
+      "epoch": 0.42460441704197877,
+      "grad_norm": 2.281834363937378,
+      "learning_rate": 6.057195010648008e-07,
+      "loss": 0.8999,
+      "step": 17630
+    },
+    {
+      "epoch": 0.42472483803376604,
+      "grad_norm": 1.560080647468567,
+      "learning_rate": 6.055927390731163e-07,
+      "loss": 0.9323,
+      "step": 17635
+    },
+    {
+      "epoch": 0.4248452590255533,
+      "grad_norm": 1.710875153541565,
+      "learning_rate": 6.054659770814318e-07,
+      "loss": 0.9317,
+      "step": 17640
+    },
+    {
+      "epoch": 0.42496568001734064,
+      "grad_norm": 1.432875156402588,
+      "learning_rate": 6.053392150897475e-07,
+      "loss": 0.9254,
+      "step": 17645
+    },
+    {
+      "epoch": 0.4250861010091279,
+      "grad_norm": 1.6327581405639648,
+      "learning_rate": 6.05212453098063e-07,
+      "loss": 0.8601,
+      "step": 17650
+    },
+    {
+      "epoch": 0.4252065220009152,
+      "grad_norm": 1.535698413848877,
+      "learning_rate": 6.050856911063787e-07,
+      "loss": 0.9295,
+      "step": 17655
+    },
+    {
+      "epoch": 0.42532694299270246,
+      "grad_norm": 1.5619745254516602,
+      "learning_rate": 6.049589291146942e-07,
+      "loss": 0.9188,
+      "step": 17660
+    },
+    {
+      "epoch": 0.4254473639844898,
+      "grad_norm": 1.8153740167617798,
+      "learning_rate": 6.048321671230097e-07,
+      "loss": 0.9126,
+      "step": 17665
+    },
+    {
+      "epoch": 0.42556778497627706,
+      "grad_norm": 1.6192477941513062,
+      "learning_rate": 6.047054051313254e-07,
+      "loss": 0.8888,
+      "step": 17670
+    },
+    {
+      "epoch": 0.42568820596806434,
+      "grad_norm": 1.4542292356491089,
+      "learning_rate": 6.04578643139641e-07,
+      "loss": 0.9421,
+      "step": 17675
+    },
+    {
+      "epoch": 0.42580862695985167,
+      "grad_norm": 1.72112238407135,
+      "learning_rate": 6.044518811479566e-07,
+      "loss": 0.8955,
+      "step": 17680
+    },
+    {
+      "epoch": 0.42592904795163894,
+      "grad_norm": 1.7958778142929077,
+      "learning_rate": 6.043251191562721e-07,
+      "loss": 0.8612,
+      "step": 17685
+    },
+    {
+      "epoch": 0.4260494689434262,
+      "grad_norm": 1.5801275968551636,
+      "learning_rate": 6.041983571645878e-07,
+      "loss": 0.9085,
+      "step": 17690
+    },
+    {
+      "epoch": 0.4261698899352135,
+      "grad_norm": 1.7030131816864014,
+      "learning_rate": 6.040715951729033e-07,
+      "loss": 0.8984,
+      "step": 17695
+    },
+    {
+      "epoch": 0.4262903109270008,
+      "grad_norm": 1.5439563989639282,
+      "learning_rate": 6.039448331812189e-07,
+      "loss": 0.882,
+      "step": 17700
+    },
+    {
+      "epoch": 0.4264107319187881,
+      "grad_norm": 1.5502954721450806,
+      "learning_rate": 6.038180711895346e-07,
+      "loss": 0.8829,
+      "step": 17705
+    },
+    {
+      "epoch": 0.42653115291057536,
+      "grad_norm": 1.400022029876709,
+      "learning_rate": 6.0369130919785e-07,
+      "loss": 0.9075,
+      "step": 17710
+    },
+    {
+      "epoch": 0.42665157390236264,
+      "grad_norm": 1.654332160949707,
+      "learning_rate": 6.035645472061657e-07,
+      "loss": 0.9059,
+      "step": 17715
+    },
+    {
+      "epoch": 0.42677199489414996,
+      "grad_norm": 1.4520788192749023,
+      "learning_rate": 6.034377852144813e-07,
+      "loss": 0.8819,
+      "step": 17720
+    },
+    {
+      "epoch": 0.42689241588593724,
+      "grad_norm": 1.511816143989563,
+      "learning_rate": 6.033110232227968e-07,
+      "loss": 0.888,
+      "step": 17725
+    },
+    {
+      "epoch": 0.4270128368777245,
+      "grad_norm": 1.648866891860962,
+      "learning_rate": 6.031842612311124e-07,
+      "loss": 0.9111,
+      "step": 17730
+    },
+    {
+      "epoch": 0.42713325786951184,
+      "grad_norm": 1.6627436876296997,
+      "learning_rate": 6.03057499239428e-07,
+      "loss": 0.9159,
+      "step": 17735
+    },
+    {
+      "epoch": 0.4272536788612991,
+      "grad_norm": 1.4436696767807007,
+      "learning_rate": 6.029307372477436e-07,
+      "loss": 0.8866,
+      "step": 17740
+    },
+    {
+      "epoch": 0.4273740998530864,
+      "grad_norm": 1.6138030290603638,
+      "learning_rate": 6.028039752560592e-07,
+      "loss": 0.9277,
+      "step": 17745
+    },
+    {
+      "epoch": 0.42749452084487366,
+      "grad_norm": 1.652036190032959,
+      "learning_rate": 6.026772132643749e-07,
+      "loss": 0.9126,
+      "step": 17750
+    },
+    {
+      "epoch": 0.427614941836661,
+      "grad_norm": 1.5107982158660889,
+      "learning_rate": 6.025504512726903e-07,
+      "loss": 0.9208,
+      "step": 17755
+    },
+    {
+      "epoch": 0.42773536282844826,
+      "grad_norm": 1.5715196132659912,
+      "learning_rate": 6.024236892810059e-07,
+      "loss": 0.9569,
+      "step": 17760
+    },
+    {
+      "epoch": 0.42785578382023554,
+      "grad_norm": 1.4496690034866333,
+      "learning_rate": 6.022969272893216e-07,
+      "loss": 0.9336,
+      "step": 17765
+    },
+    {
+      "epoch": 0.4279762048120228,
+      "grad_norm": 1.6113660335540771,
+      "learning_rate": 6.021701652976371e-07,
+      "loss": 0.9136,
+      "step": 17770
+    },
+    {
+      "epoch": 0.42809662580381014,
+      "grad_norm": 1.7576104402542114,
+      "learning_rate": 6.020434033059527e-07,
+      "loss": 0.9168,
+      "step": 17775
+    },
+    {
+      "epoch": 0.4282170467955974,
+      "grad_norm": 1.6658543348312378,
+      "learning_rate": 6.019166413142683e-07,
+      "loss": 0.913,
+      "step": 17780
+    },
+    {
+      "epoch": 0.4283374677873847,
+      "grad_norm": 1.5523247718811035,
+      "learning_rate": 6.017898793225838e-07,
+      "loss": 0.9239,
+      "step": 17785
+    },
+    {
+      "epoch": 0.42845788877917196,
+      "grad_norm": 1.594376802444458,
+      "learning_rate": 6.016631173308995e-07,
+      "loss": 0.9631,
+      "step": 17790
+    },
+    {
+      "epoch": 0.4285783097709593,
+      "grad_norm": 1.3364665508270264,
+      "learning_rate": 6.015363553392151e-07,
+      "loss": 0.8864,
+      "step": 17795
+    },
+    {
+      "epoch": 0.42869873076274656,
+      "grad_norm": 1.6243319511413574,
+      "learning_rate": 6.014095933475306e-07,
+      "loss": 0.8928,
+      "step": 17800
+    },
+    {
+      "epoch": 0.42881915175453383,
+      "grad_norm": 2.047043561935425,
+      "learning_rate": 6.012828313558462e-07,
+      "loss": 0.9565,
+      "step": 17805
+    },
+    {
+      "epoch": 0.42893957274632116,
+      "grad_norm": 1.4199920892715454,
+      "learning_rate": 6.011560693641619e-07,
+      "loss": 0.8812,
+      "step": 17810
+    },
+    {
+      "epoch": 0.42905999373810844,
+      "grad_norm": 1.5130432844161987,
+      "learning_rate": 6.010293073724774e-07,
+      "loss": 0.8514,
+      "step": 17815
+    },
+    {
+      "epoch": 0.4291804147298957,
+      "grad_norm": 1.6105554103851318,
+      "learning_rate": 6.00902545380793e-07,
+      "loss": 0.8948,
+      "step": 17820
+    },
+    {
+      "epoch": 0.429300835721683,
+      "grad_norm": 1.5266790390014648,
+      "learning_rate": 6.007757833891086e-07,
+      "loss": 0.9541,
+      "step": 17825
+    },
+    {
+      "epoch": 0.4294212567134703,
+      "grad_norm": 1.5769493579864502,
+      "learning_rate": 6.006490213974241e-07,
+      "loss": 0.8985,
+      "step": 17830
+    },
+    {
+      "epoch": 0.4295416777052576,
+      "grad_norm": 1.99751615524292,
+      "learning_rate": 6.005222594057398e-07,
+      "loss": 0.9314,
+      "step": 17835
+    },
+    {
+      "epoch": 0.42966209869704486,
+      "grad_norm": 1.7150274515151978,
+      "learning_rate": 6.003954974140554e-07,
+      "loss": 0.9215,
+      "step": 17840
+    },
+    {
+      "epoch": 0.42978251968883213,
+      "grad_norm": 1.5975068807601929,
+      "learning_rate": 6.002687354223708e-07,
+      "loss": 0.908,
+      "step": 17845
+    },
+    {
+      "epoch": 0.42990294068061946,
+      "grad_norm": 1.5267257690429688,
+      "learning_rate": 6.001419734306865e-07,
+      "loss": 0.8962,
+      "step": 17850
+    },
+    {
+      "epoch": 0.43002336167240673,
+      "grad_norm": 1.561611533164978,
+      "learning_rate": 6.000152114390021e-07,
+      "loss": 0.9718,
+      "step": 17855
+    },
+    {
+      "epoch": 0.430143782664194,
+      "grad_norm": 1.575334072113037,
+      "learning_rate": 5.998884494473177e-07,
+      "loss": 0.9145,
+      "step": 17860
+    },
+    {
+      "epoch": 0.43026420365598134,
+      "grad_norm": 1.507455587387085,
+      "learning_rate": 5.997616874556333e-07,
+      "loss": 0.9244,
+      "step": 17865
+    },
+    {
+      "epoch": 0.4303846246477686,
+      "grad_norm": 1.6253801584243774,
+      "learning_rate": 5.996349254639489e-07,
+      "loss": 0.8836,
+      "step": 17870
+    },
+    {
+      "epoch": 0.4305050456395559,
+      "grad_norm": 1.6473288536071777,
+      "learning_rate": 5.995081634722644e-07,
+      "loss": 0.9189,
+      "step": 17875
+    },
+    {
+      "epoch": 0.43062546663134316,
+      "grad_norm": 1.3467274904251099,
+      "learning_rate": 5.9938140148058e-07,
+      "loss": 0.8914,
+      "step": 17880
+    },
+    {
+      "epoch": 0.4307458876231305,
+      "grad_norm": 1.6535570621490479,
+      "learning_rate": 5.992546394888957e-07,
+      "loss": 0.8545,
+      "step": 17885
+    },
+    {
+      "epoch": 0.43086630861491776,
+      "grad_norm": 1.9367197751998901,
+      "learning_rate": 5.991278774972111e-07,
+      "loss": 0.9876,
+      "step": 17890
+    },
+    {
+      "epoch": 0.43098672960670503,
+      "grad_norm": 1.5257887840270996,
+      "learning_rate": 5.990011155055268e-07,
+      "loss": 0.8616,
+      "step": 17895
+    },
+    {
+      "epoch": 0.4311071505984923,
+      "grad_norm": 1.6853076219558716,
+      "learning_rate": 5.988743535138424e-07,
+      "loss": 0.9274,
+      "step": 17900
+    },
+    {
+      "epoch": 0.43122757159027963,
+      "grad_norm": 1.6795798540115356,
+      "learning_rate": 5.987475915221579e-07,
+      "loss": 0.8983,
+      "step": 17905
+    },
+    {
+      "epoch": 0.4313479925820669,
+      "grad_norm": 1.533904790878296,
+      "learning_rate": 5.986208295304736e-07,
+      "loss": 0.8607,
+      "step": 17910
+    },
+    {
+      "epoch": 0.4314684135738542,
+      "grad_norm": 1.7242625951766968,
+      "learning_rate": 5.984940675387891e-07,
+      "loss": 0.9448,
+      "step": 17915
+    },
+    {
+      "epoch": 0.4315888345656415,
+      "grad_norm": 1.5450302362442017,
+      "learning_rate": 5.983673055471047e-07,
+      "loss": 0.9315,
+      "step": 17920
+    },
+    {
+      "epoch": 0.4317092555574288,
+      "grad_norm": 1.547723650932312,
+      "learning_rate": 5.982405435554203e-07,
+      "loss": 0.9453,
+      "step": 17925
+    },
+    {
+      "epoch": 0.43182967654921606,
+      "grad_norm": 1.5352214574813843,
+      "learning_rate": 5.98113781563736e-07,
+      "loss": 0.9061,
+      "step": 17930
+    },
+    {
+      "epoch": 0.43195009754100333,
+      "grad_norm": 1.7504684925079346,
+      "learning_rate": 5.979870195720515e-07,
+      "loss": 0.903,
+      "step": 17935
+    },
+    {
+      "epoch": 0.43207051853279066,
+      "grad_norm": 1.4188649654388428,
+      "learning_rate": 5.97860257580367e-07,
+      "loss": 0.91,
+      "step": 17940
+    },
+    {
+      "epoch": 0.43219093952457793,
+      "grad_norm": 1.8634692430496216,
+      "learning_rate": 5.977334955886827e-07,
+      "loss": 0.9519,
+      "step": 17945
+    },
+    {
+      "epoch": 0.4323113605163652,
+      "grad_norm": 1.4302127361297607,
+      "learning_rate": 5.976067335969982e-07,
+      "loss": 0.9287,
+      "step": 17950
+    },
+    {
+      "epoch": 0.4324317815081525,
+      "grad_norm": 1.4653289318084717,
+      "learning_rate": 5.974799716053139e-07,
+      "loss": 0.9013,
+      "step": 17955
+    },
+    {
+      "epoch": 0.4325522024999398,
+      "grad_norm": 1.4174020290374756,
+      "learning_rate": 5.973532096136294e-07,
+      "loss": 0.9004,
+      "step": 17960
+    },
+    {
+      "epoch": 0.4326726234917271,
+      "grad_norm": 1.780843734741211,
+      "learning_rate": 5.972264476219449e-07,
+      "loss": 0.884,
+      "step": 17965
+    },
+    {
+      "epoch": 0.43279304448351436,
+      "grad_norm": 1.7866368293762207,
+      "learning_rate": 5.970996856302606e-07,
+      "loss": 0.8893,
+      "step": 17970
+    },
+    {
+      "epoch": 0.43291346547530163,
+      "grad_norm": 1.6399894952774048,
+      "learning_rate": 5.969729236385762e-07,
+      "loss": 0.8958,
+      "step": 17975
+    },
+    {
+      "epoch": 0.43303388646708896,
+      "grad_norm": 1.3787689208984375,
+      "learning_rate": 5.968461616468918e-07,
+      "loss": 0.9251,
+      "step": 17980
+    },
+    {
+      "epoch": 0.43315430745887623,
+      "grad_norm": 1.4486820697784424,
+      "learning_rate": 5.967193996552073e-07,
+      "loss": 0.9001,
+      "step": 17985
+    },
+    {
+      "epoch": 0.4332747284506635,
+      "grad_norm": 1.5622378587722778,
+      "learning_rate": 5.96592637663523e-07,
+      "loss": 0.9648,
+      "step": 17990
+    },
+    {
+      "epoch": 0.43339514944245083,
+      "grad_norm": 1.6893956661224365,
+      "learning_rate": 5.964658756718385e-07,
+      "loss": 0.9027,
+      "step": 17995
+    },
+    {
+      "epoch": 0.4335155704342381,
+      "grad_norm": 1.5421922206878662,
+      "learning_rate": 5.963391136801541e-07,
+      "loss": 0.9626,
+      "step": 18000
+    },
+    {
+      "epoch": 0.4336359914260254,
+      "grad_norm": 1.4266217947006226,
+      "learning_rate": 5.962123516884698e-07,
+      "loss": 0.8913,
+      "step": 18005
+    },
+    {
+      "epoch": 0.43375641241781265,
+      "grad_norm": 1.4913381338119507,
+      "learning_rate": 5.960855896967852e-07,
+      "loss": 0.9139,
+      "step": 18010
+    },
+    {
+      "epoch": 0.4338768334096,
+      "grad_norm": 1.609383225440979,
+      "learning_rate": 5.959588277051009e-07,
+      "loss": 0.8989,
+      "step": 18015
+    },
+    {
+      "epoch": 0.43399725440138726,
+      "grad_norm": 1.436481237411499,
+      "learning_rate": 5.958320657134165e-07,
+      "loss": 0.9932,
+      "step": 18020
+    },
+    {
+      "epoch": 0.43411767539317453,
+      "grad_norm": 1.5978612899780273,
+      "learning_rate": 5.95705303721732e-07,
+      "loss": 0.8818,
+      "step": 18025
+    },
+    {
+      "epoch": 0.4342380963849618,
+      "grad_norm": 1.6705412864685059,
+      "learning_rate": 5.955785417300476e-07,
+      "loss": 0.8584,
+      "step": 18030
+    },
+    {
+      "epoch": 0.43435851737674913,
+      "grad_norm": 1.4328734874725342,
+      "learning_rate": 5.954517797383632e-07,
+      "loss": 0.9302,
+      "step": 18035
+    },
+    {
+      "epoch": 0.4344789383685364,
+      "grad_norm": 1.2172200679779053,
+      "learning_rate": 5.953250177466788e-07,
+      "loss": 0.9303,
+      "step": 18040
+    },
+    {
+      "epoch": 0.4345993593603237,
+      "grad_norm": 1.629767656326294,
+      "learning_rate": 5.951982557549944e-07,
+      "loss": 0.8776,
+      "step": 18045
+    },
+    {
+      "epoch": 0.434719780352111,
+      "grad_norm": 1.515710711479187,
+      "learning_rate": 5.9507149376331e-07,
+      "loss": 0.8947,
+      "step": 18050
+    },
+    {
+      "epoch": 0.4348402013438983,
+      "grad_norm": 1.75547456741333,
+      "learning_rate": 5.949447317716255e-07,
+      "loss": 0.9042,
+      "step": 18055
+    },
+    {
+      "epoch": 0.43496062233568555,
+      "grad_norm": 1.578229308128357,
+      "learning_rate": 5.948179697799411e-07,
+      "loss": 0.9092,
+      "step": 18060
+    },
+    {
+      "epoch": 0.4350810433274728,
+      "grad_norm": 1.6973870992660522,
+      "learning_rate": 5.946912077882568e-07,
+      "loss": 0.9316,
+      "step": 18065
+    },
+    {
+      "epoch": 0.43520146431926016,
+      "grad_norm": 1.6119117736816406,
+      "learning_rate": 5.945644457965723e-07,
+      "loss": 0.8674,
+      "step": 18070
+    },
+    {
+      "epoch": 0.43532188531104743,
+      "grad_norm": 1.3468626737594604,
+      "learning_rate": 5.944376838048879e-07,
+      "loss": 0.8576,
+      "step": 18075
+    },
+    {
+      "epoch": 0.4354423063028347,
+      "grad_norm": 1.451310396194458,
+      "learning_rate": 5.943109218132035e-07,
+      "loss": 0.9071,
+      "step": 18080
+    },
+    {
+      "epoch": 0.435562727294622,
+      "grad_norm": 1.6123762130737305,
+      "learning_rate": 5.94184159821519e-07,
+      "loss": 0.9343,
+      "step": 18085
+    },
+    {
+      "epoch": 0.4356831482864093,
+      "grad_norm": 1.454443335533142,
+      "learning_rate": 5.940573978298347e-07,
+      "loss": 0.9151,
+      "step": 18090
+    },
+    {
+      "epoch": 0.4358035692781966,
+      "grad_norm": 1.4933301210403442,
+      "learning_rate": 5.939306358381503e-07,
+      "loss": 0.9507,
+      "step": 18095
+    },
+    {
+      "epoch": 0.43592399026998385,
+      "grad_norm": 1.517813801765442,
+      "learning_rate": 5.938038738464658e-07,
+      "loss": 0.9592,
+      "step": 18100
+    },
+    {
+      "epoch": 0.4360444112617711,
+      "grad_norm": 1.5954922437667847,
+      "learning_rate": 5.936771118547814e-07,
+      "loss": 0.9175,
+      "step": 18105
+    },
+    {
+      "epoch": 0.43616483225355845,
+      "grad_norm": 1.4719862937927246,
+      "learning_rate": 5.93550349863097e-07,
+      "loss": 0.9304,
+      "step": 18110
+    },
+    {
+      "epoch": 0.4362852532453457,
+      "grad_norm": 1.4959850311279297,
+      "learning_rate": 5.934235878714126e-07,
+      "loss": 0.9394,
+      "step": 18115
+    },
+    {
+      "epoch": 0.436405674237133,
+      "grad_norm": 1.2718183994293213,
+      "learning_rate": 5.932968258797282e-07,
+      "loss": 0.9027,
+      "step": 18120
+    },
+    {
+      "epoch": 0.43652609522892033,
+      "grad_norm": 1.5270932912826538,
+      "learning_rate": 5.931700638880438e-07,
+      "loss": 0.9114,
+      "step": 18125
+    },
+    {
+      "epoch": 0.4366465162207076,
+      "grad_norm": 1.6102068424224854,
+      "learning_rate": 5.930433018963593e-07,
+      "loss": 0.9323,
+      "step": 18130
+    },
+    {
+      "epoch": 0.4367669372124949,
+      "grad_norm": 1.6144814491271973,
+      "learning_rate": 5.92916539904675e-07,
+      "loss": 0.9071,
+      "step": 18135
+    },
+    {
+      "epoch": 0.43688735820428215,
+      "grad_norm": 1.586299180984497,
+      "learning_rate": 5.927897779129906e-07,
+      "loss": 0.9377,
+      "step": 18140
+    },
+    {
+      "epoch": 0.4370077791960695,
+      "grad_norm": 1.690378189086914,
+      "learning_rate": 5.92663015921306e-07,
+      "loss": 0.8917,
+      "step": 18145
+    },
+    {
+      "epoch": 0.43712820018785675,
+      "grad_norm": 1.6998260021209717,
+      "learning_rate": 5.925362539296217e-07,
+      "loss": 0.9332,
+      "step": 18150
+    },
+    {
+      "epoch": 0.437248621179644,
+      "grad_norm": 1.5584924221038818,
+      "learning_rate": 5.924094919379373e-07,
+      "loss": 0.9188,
+      "step": 18155
+    },
+    {
+      "epoch": 0.4373690421714313,
+      "grad_norm": 1.5578699111938477,
+      "learning_rate": 5.922827299462529e-07,
+      "loss": 0.9049,
+      "step": 18160
+    },
+    {
+      "epoch": 0.43748946316321863,
+      "grad_norm": 1.6602588891983032,
+      "learning_rate": 5.921559679545685e-07,
+      "loss": 0.9397,
+      "step": 18165
+    },
+    {
+      "epoch": 0.4376098841550059,
+      "grad_norm": 1.7947601079940796,
+      "learning_rate": 5.92029205962884e-07,
+      "loss": 0.9206,
+      "step": 18170
+    },
+    {
+      "epoch": 0.4377303051467932,
+      "grad_norm": 1.5118459463119507,
+      "learning_rate": 5.919024439711996e-07,
+      "loss": 0.9207,
+      "step": 18175
+    },
+    {
+      "epoch": 0.4378507261385805,
+      "grad_norm": 1.6038711071014404,
+      "learning_rate": 5.917756819795152e-07,
+      "loss": 0.8964,
+      "step": 18180
+    },
+    {
+      "epoch": 0.4379711471303678,
+      "grad_norm": 1.8039369583129883,
+      "learning_rate": 5.916489199878309e-07,
+      "loss": 0.9169,
+      "step": 18185
+    },
+    {
+      "epoch": 0.43809156812215505,
+      "grad_norm": 1.5942789316177368,
+      "learning_rate": 5.915221579961463e-07,
+      "loss": 0.9819,
+      "step": 18190
+    },
+    {
+      "epoch": 0.4382119891139423,
+      "grad_norm": 1.330572485923767,
+      "learning_rate": 5.91395396004462e-07,
+      "loss": 0.9146,
+      "step": 18195
+    },
+    {
+      "epoch": 0.43833241010572965,
+      "grad_norm": 1.6067678928375244,
+      "learning_rate": 5.912686340127776e-07,
+      "loss": 0.8957,
+      "step": 18200
+    },
+    {
+      "epoch": 0.4384528310975169,
+      "grad_norm": 1.4917579889297485,
+      "learning_rate": 5.911418720210932e-07,
+      "loss": 0.9636,
+      "step": 18205
+    },
+    {
+      "epoch": 0.4385732520893042,
+      "grad_norm": 1.3807231187820435,
+      "learning_rate": 5.910151100294088e-07,
+      "loss": 0.8647,
+      "step": 18210
+    },
+    {
+      "epoch": 0.4386936730810915,
+      "grad_norm": 1.4772977828979492,
+      "learning_rate": 5.908883480377243e-07,
+      "loss": 0.9232,
+      "step": 18215
+    },
+    {
+      "epoch": 0.4388140940728788,
+      "grad_norm": 1.711417317390442,
+      "learning_rate": 5.9076158604604e-07,
+      "loss": 0.9004,
+      "step": 18220
+    },
+    {
+      "epoch": 0.4389345150646661,
+      "grad_norm": 1.4783941507339478,
+      "learning_rate": 5.906348240543555e-07,
+      "loss": 0.9263,
+      "step": 18225
+    },
+    {
+      "epoch": 0.43905493605645335,
+      "grad_norm": 1.916579008102417,
+      "learning_rate": 5.905080620626711e-07,
+      "loss": 0.8874,
+      "step": 18230
+    },
+    {
+      "epoch": 0.4391753570482406,
+      "grad_norm": 1.3718454837799072,
+      "learning_rate": 5.903813000709868e-07,
+      "loss": 0.9168,
+      "step": 18235
+    },
+    {
+      "epoch": 0.43929577804002795,
+      "grad_norm": 1.5684967041015625,
+      "learning_rate": 5.902545380793022e-07,
+      "loss": 0.8782,
+      "step": 18240
+    },
+    {
+      "epoch": 0.4394161990318152,
+      "grad_norm": 1.6263731718063354,
+      "learning_rate": 5.901277760876179e-07,
+      "loss": 0.9195,
+      "step": 18245
+    },
+    {
+      "epoch": 0.4395366200236025,
+      "grad_norm": 1.4979406595230103,
+      "learning_rate": 5.900010140959335e-07,
+      "loss": 0.9013,
+      "step": 18250
+    },
+    {
+      "epoch": 0.4396570410153898,
+      "grad_norm": 1.8347595930099487,
+      "learning_rate": 5.89874252104249e-07,
+      "loss": 0.9248,
+      "step": 18255
+    },
+    {
+      "epoch": 0.4397774620071771,
+      "grad_norm": 1.6108922958374023,
+      "learning_rate": 5.897474901125646e-07,
+      "loss": 0.8857,
+      "step": 18260
+    },
+    {
+      "epoch": 0.4398978829989644,
+      "grad_norm": 1.4353135824203491,
+      "learning_rate": 5.896207281208802e-07,
+      "loss": 0.9357,
+      "step": 18265
+    },
+    {
+      "epoch": 0.44001830399075165,
+      "grad_norm": 1.3888901472091675,
+      "learning_rate": 5.894939661291958e-07,
+      "loss": 0.9444,
+      "step": 18270
+    },
+    {
+      "epoch": 0.440138724982539,
+      "grad_norm": 1.522700548171997,
+      "learning_rate": 5.893672041375114e-07,
+      "loss": 0.8662,
+      "step": 18275
+    },
+    {
+      "epoch": 0.44025914597432625,
+      "grad_norm": 1.5019195079803467,
+      "learning_rate": 5.892404421458271e-07,
+      "loss": 0.955,
+      "step": 18280
+    },
+    {
+      "epoch": 0.4403795669661135,
+      "grad_norm": 1.7402147054672241,
+      "learning_rate": 5.891136801541425e-07,
+      "loss": 0.9712,
+      "step": 18285
+    },
+    {
+      "epoch": 0.4404999879579008,
+      "grad_norm": 1.4262351989746094,
+      "learning_rate": 5.889869181624581e-07,
+      "loss": 0.9333,
+      "step": 18290
+    },
+    {
+      "epoch": 0.4406204089496881,
+      "grad_norm": 1.7041842937469482,
+      "learning_rate": 5.888601561707738e-07,
+      "loss": 0.8435,
+      "step": 18295
+    },
+    {
+      "epoch": 0.4407408299414754,
+      "grad_norm": 1.5107301473617554,
+      "learning_rate": 5.887333941790893e-07,
+      "loss": 0.9032,
+      "step": 18300
+    },
+    {
+      "epoch": 0.44086125093326267,
+      "grad_norm": 1.543336033821106,
+      "learning_rate": 5.88606632187405e-07,
+      "loss": 0.9219,
+      "step": 18305
+    },
+    {
+      "epoch": 0.44098167192505,
+      "grad_norm": 1.5038975477218628,
+      "learning_rate": 5.884798701957205e-07,
+      "loss": 0.8895,
+      "step": 18310
+    },
+    {
+      "epoch": 0.4411020929168373,
+      "grad_norm": 1.7226147651672363,
+      "learning_rate": 5.88353108204036e-07,
+      "loss": 0.917,
+      "step": 18315
+    },
+    {
+      "epoch": 0.44122251390862455,
+      "grad_norm": 1.7164808511734009,
+      "learning_rate": 5.882263462123517e-07,
+      "loss": 0.9058,
+      "step": 18320
+    },
+    {
+      "epoch": 0.4413429349004118,
+      "grad_norm": 1.4979898929595947,
+      "learning_rate": 5.880995842206673e-07,
+      "loss": 0.9308,
+      "step": 18325
+    },
+    {
+      "epoch": 0.44146335589219915,
+      "grad_norm": 1.5202665328979492,
+      "learning_rate": 5.879728222289828e-07,
+      "loss": 0.9014,
+      "step": 18330
+    },
+    {
+      "epoch": 0.4415837768839864,
+      "grad_norm": 1.4686599969863892,
+      "learning_rate": 5.878460602372984e-07,
+      "loss": 0.9461,
+      "step": 18335
+    },
+    {
+      "epoch": 0.4417041978757737,
+      "grad_norm": 1.7539727687835693,
+      "learning_rate": 5.877192982456141e-07,
+      "loss": 0.9511,
+      "step": 18340
+    },
+    {
+      "epoch": 0.44182461886756097,
+      "grad_norm": 1.5274763107299805,
+      "learning_rate": 5.875925362539296e-07,
+      "loss": 0.9521,
+      "step": 18345
+    },
+    {
+      "epoch": 0.4419450398593483,
+      "grad_norm": 1.4581748247146606,
+      "learning_rate": 5.874657742622452e-07,
+      "loss": 0.9155,
+      "step": 18350
+    },
+    {
+      "epoch": 0.44206546085113557,
+      "grad_norm": 1.493125319480896,
+      "learning_rate": 5.873390122705608e-07,
+      "loss": 0.8899,
+      "step": 18355
+    },
+    {
+      "epoch": 0.44218588184292285,
+      "grad_norm": 1.47279953956604,
+      "learning_rate": 5.872122502788763e-07,
+      "loss": 0.8576,
+      "step": 18360
+    },
+    {
+      "epoch": 0.4423063028347102,
+      "grad_norm": 1.8338985443115234,
+      "learning_rate": 5.87085488287192e-07,
+      "loss": 0.9497,
+      "step": 18365
+    },
+    {
+      "epoch": 0.44242672382649745,
+      "grad_norm": 1.5390070676803589,
+      "learning_rate": 5.869587262955076e-07,
+      "loss": 0.9115,
+      "step": 18370
+    },
+    {
+      "epoch": 0.4425471448182847,
+      "grad_norm": 1.5464487075805664,
+      "learning_rate": 5.86831964303823e-07,
+      "loss": 0.9,
+      "step": 18375
+    },
+    {
+      "epoch": 0.442667565810072,
+      "grad_norm": 1.4387696981430054,
+      "learning_rate": 5.867052023121387e-07,
+      "loss": 0.9589,
+      "step": 18380
+    },
+    {
+      "epoch": 0.4427879868018593,
+      "grad_norm": 1.3581068515777588,
+      "learning_rate": 5.865784403204543e-07,
+      "loss": 0.8743,
+      "step": 18385
+    },
+    {
+      "epoch": 0.4429084077936466,
+      "grad_norm": 1.4792227745056152,
+      "learning_rate": 5.864516783287699e-07,
+      "loss": 0.9316,
+      "step": 18390
+    },
+    {
+      "epoch": 0.44302882878543387,
+      "grad_norm": 1.7097699642181396,
+      "learning_rate": 5.863249163370855e-07,
+      "loss": 0.8889,
+      "step": 18395
+    },
+    {
+      "epoch": 0.44314924977722114,
+      "grad_norm": 1.5005133152008057,
+      "learning_rate": 5.86198154345401e-07,
+      "loss": 0.9554,
+      "step": 18400
+    },
+    {
+      "epoch": 0.4432696707690085,
+      "grad_norm": 1.667360782623291,
+      "learning_rate": 5.860713923537166e-07,
+      "loss": 0.8946,
+      "step": 18405
+    },
+    {
+      "epoch": 0.44339009176079575,
+      "grad_norm": 1.39607572555542,
+      "learning_rate": 5.859446303620322e-07,
+      "loss": 0.9718,
+      "step": 18410
+    },
+    {
+      "epoch": 0.443510512752583,
+      "grad_norm": 1.274971842765808,
+      "learning_rate": 5.858178683703479e-07,
+      "loss": 0.9119,
+      "step": 18415
+    },
+    {
+      "epoch": 0.4436309337443703,
+      "grad_norm": 1.655137538909912,
+      "learning_rate": 5.856911063786634e-07,
+      "loss": 0.8827,
+      "step": 18420
+    },
+    {
+      "epoch": 0.4437513547361576,
+      "grad_norm": 1.6791355609893799,
+      "learning_rate": 5.85564344386979e-07,
+      "loss": 0.9148,
+      "step": 18425
+    },
+    {
+      "epoch": 0.4438717757279449,
+      "grad_norm": 1.470255732536316,
+      "learning_rate": 5.854375823952946e-07,
+      "loss": 0.9326,
+      "step": 18430
+    },
+    {
+      "epoch": 0.44399219671973217,
+      "grad_norm": 1.4369548559188843,
+      "learning_rate": 5.853108204036101e-07,
+      "loss": 0.9058,
+      "step": 18435
+    },
+    {
+      "epoch": 0.4441126177115195,
+      "grad_norm": 1.660200834274292,
+      "learning_rate": 5.851840584119258e-07,
+      "loss": 0.9008,
+      "step": 18440
+    },
+    {
+      "epoch": 0.44423303870330677,
+      "grad_norm": 1.6829956769943237,
+      "learning_rate": 5.850572964202413e-07,
+      "loss": 0.9556,
+      "step": 18445
+    },
+    {
+      "epoch": 0.44435345969509404,
+      "grad_norm": 1.4490734338760376,
+      "learning_rate": 5.849305344285569e-07,
+      "loss": 0.9025,
+      "step": 18450
+    },
+    {
+      "epoch": 0.4444738806868813,
+      "grad_norm": 1.6446168422698975,
+      "learning_rate": 5.848037724368725e-07,
+      "loss": 0.9281,
+      "step": 18455
+    },
+    {
+      "epoch": 0.44459430167866865,
+      "grad_norm": 1.7542520761489868,
+      "learning_rate": 5.846770104451882e-07,
+      "loss": 0.9345,
+      "step": 18460
+    },
+    {
+      "epoch": 0.4447147226704559,
+      "grad_norm": 1.385473608970642,
+      "learning_rate": 5.845502484535037e-07,
+      "loss": 0.86,
+      "step": 18465
+    },
+    {
+      "epoch": 0.4448351436622432,
+      "grad_norm": 1.7109313011169434,
+      "learning_rate": 5.844234864618192e-07,
+      "loss": 0.8704,
+      "step": 18470
+    },
+    {
+      "epoch": 0.44495556465403047,
+      "grad_norm": 1.594736099243164,
+      "learning_rate": 5.842967244701349e-07,
+      "loss": 0.9048,
+      "step": 18475
+    },
+    {
+      "epoch": 0.4450759856458178,
+      "grad_norm": 1.4663512706756592,
+      "learning_rate": 5.841699624784504e-07,
+      "loss": 0.9278,
+      "step": 18480
+    },
+    {
+      "epoch": 0.44519640663760507,
+      "grad_norm": 1.726747751235962,
+      "learning_rate": 5.840432004867661e-07,
+      "loss": 0.8576,
+      "step": 18485
+    },
+    {
+      "epoch": 0.44531682762939234,
+      "grad_norm": 1.5041638612747192,
+      "learning_rate": 5.839164384950817e-07,
+      "loss": 0.9079,
+      "step": 18490
+    },
+    {
+      "epoch": 0.44543724862117967,
+      "grad_norm": 1.6056970357894897,
+      "learning_rate": 5.837896765033971e-07,
+      "loss": 0.9491,
+      "step": 18495
+    },
+    {
+      "epoch": 0.44555766961296694,
+      "grad_norm": 1.565638780593872,
+      "learning_rate": 5.836629145117128e-07,
+      "loss": 0.9341,
+      "step": 18500
+    },
+    {
+      "epoch": 0.4456780906047542,
+      "grad_norm": 1.3103954792022705,
+      "learning_rate": 5.835361525200284e-07,
+      "loss": 0.9042,
+      "step": 18505
+    },
+    {
+      "epoch": 0.4457985115965415,
+      "grad_norm": 1.373991847038269,
+      "learning_rate": 5.83409390528344e-07,
+      "loss": 0.8608,
+      "step": 18510
+    },
+    {
+      "epoch": 0.4459189325883288,
+      "grad_norm": 1.3458672761917114,
+      "learning_rate": 5.832826285366595e-07,
+      "loss": 0.9019,
+      "step": 18515
+    },
+    {
+      "epoch": 0.4460393535801161,
+      "grad_norm": 1.499133586883545,
+      "learning_rate": 5.831558665449751e-07,
+      "loss": 0.902,
+      "step": 18520
+    },
+    {
+      "epoch": 0.44615977457190337,
+      "grad_norm": 1.6567200422286987,
+      "learning_rate": 5.830291045532907e-07,
+      "loss": 0.9113,
+      "step": 18525
+    },
+    {
+      "epoch": 0.44628019556369064,
+      "grad_norm": 1.7166463136672974,
+      "learning_rate": 5.829023425616063e-07,
+      "loss": 0.888,
+      "step": 18530
+    },
+    {
+      "epoch": 0.44640061655547797,
+      "grad_norm": 1.7077884674072266,
+      "learning_rate": 5.82775580569922e-07,
+      "loss": 0.9058,
+      "step": 18535
+    },
+    {
+      "epoch": 0.44652103754726524,
+      "grad_norm": 1.6900447607040405,
+      "learning_rate": 5.826488185782374e-07,
+      "loss": 0.8517,
+      "step": 18540
+    },
+    {
+      "epoch": 0.4466414585390525,
+      "grad_norm": 1.6718276739120483,
+      "learning_rate": 5.825220565865531e-07,
+      "loss": 0.8974,
+      "step": 18545
+    },
+    {
+      "epoch": 0.4467618795308398,
+      "grad_norm": 1.6190602779388428,
+      "learning_rate": 5.823952945948687e-07,
+      "loss": 0.9342,
+      "step": 18550
+    },
+    {
+      "epoch": 0.4468823005226271,
+      "grad_norm": 1.8204940557479858,
+      "learning_rate": 5.822685326031842e-07,
+      "loss": 0.8827,
+      "step": 18555
+    },
+    {
+      "epoch": 0.4470027215144144,
+      "grad_norm": 1.5450371503829956,
+      "learning_rate": 5.821417706114998e-07,
+      "loss": 0.9504,
+      "step": 18560
+    },
+    {
+      "epoch": 0.44712314250620167,
+      "grad_norm": 1.7189927101135254,
+      "learning_rate": 5.820150086198154e-07,
+      "loss": 0.921,
+      "step": 18565
+    },
+    {
+      "epoch": 0.447243563497989,
+      "grad_norm": 1.7274240255355835,
+      "learning_rate": 5.81888246628131e-07,
+      "loss": 0.9178,
+      "step": 18570
+    },
+    {
+      "epoch": 0.44736398448977627,
+      "grad_norm": 1.5280604362487793,
+      "learning_rate": 5.817614846364466e-07,
+      "loss": 0.9305,
+      "step": 18575
+    },
+    {
+      "epoch": 0.44748440548156354,
+      "grad_norm": 1.4250680208206177,
+      "learning_rate": 5.816347226447623e-07,
+      "loss": 0.8913,
+      "step": 18580
+    },
+    {
+      "epoch": 0.4476048264733508,
+      "grad_norm": 1.6027624607086182,
+      "learning_rate": 5.815079606530777e-07,
+      "loss": 0.9009,
+      "step": 18585
+    },
+    {
+      "epoch": 0.44772524746513814,
+      "grad_norm": 1.576629877090454,
+      "learning_rate": 5.813811986613933e-07,
+      "loss": 0.9222,
+      "step": 18590
+    },
+    {
+      "epoch": 0.4478456684569254,
+      "grad_norm": 1.5910590887069702,
+      "learning_rate": 5.81254436669709e-07,
+      "loss": 0.9767,
+      "step": 18595
+    },
+    {
+      "epoch": 0.4479660894487127,
+      "grad_norm": 1.580986499786377,
+      "learning_rate": 5.811276746780245e-07,
+      "loss": 0.8606,
+      "step": 18600
+    },
+    {
+      "epoch": 0.44808651044049996,
+      "grad_norm": 1.5624583959579468,
+      "learning_rate": 5.810009126863402e-07,
+      "loss": 0.8826,
+      "step": 18605
+    },
+    {
+      "epoch": 0.4482069314322873,
+      "grad_norm": 1.5432230234146118,
+      "learning_rate": 5.808741506946557e-07,
+      "loss": 0.9025,
+      "step": 18610
+    },
+    {
+      "epoch": 0.44832735242407457,
+      "grad_norm": 1.5583560466766357,
+      "learning_rate": 5.807473887029712e-07,
+      "loss": 0.8883,
+      "step": 18615
+    },
+    {
+      "epoch": 0.44844777341586184,
+      "grad_norm": 1.5831446647644043,
+      "learning_rate": 5.806206267112869e-07,
+      "loss": 0.9434,
+      "step": 18620
+    },
+    {
+      "epoch": 0.44856819440764917,
+      "grad_norm": 1.3928394317626953,
+      "learning_rate": 5.804938647196025e-07,
+      "loss": 0.8973,
+      "step": 18625
+    },
+    {
+      "epoch": 0.44868861539943644,
+      "grad_norm": 1.4622230529785156,
+      "learning_rate": 5.80367102727918e-07,
+      "loss": 0.8864,
+      "step": 18630
+    },
+    {
+      "epoch": 0.4488090363912237,
+      "grad_norm": 1.3747080564498901,
+      "learning_rate": 5.802403407362336e-07,
+      "loss": 0.8946,
+      "step": 18635
+    },
+    {
+      "epoch": 0.448929457383011,
+      "grad_norm": 1.7349574565887451,
+      "learning_rate": 5.801135787445492e-07,
+      "loss": 0.953,
+      "step": 18640
+    },
+    {
+      "epoch": 0.4490498783747983,
+      "grad_norm": 1.4037730693817139,
+      "learning_rate": 5.799868167528648e-07,
+      "loss": 0.8879,
+      "step": 18645
+    },
+    {
+      "epoch": 0.4491702993665856,
+      "grad_norm": 1.5784122943878174,
+      "learning_rate": 5.798600547611804e-07,
+      "loss": 0.9634,
+      "step": 18650
+    },
+    {
+      "epoch": 0.44929072035837286,
+      "grad_norm": 1.639102816581726,
+      "learning_rate": 5.79733292769496e-07,
+      "loss": 0.9171,
+      "step": 18655
+    },
+    {
+      "epoch": 0.44941114135016014,
+      "grad_norm": 1.5633991956710815,
+      "learning_rate": 5.796065307778115e-07,
+      "loss": 0.9123,
+      "step": 18660
+    },
+    {
+      "epoch": 0.44953156234194747,
+      "grad_norm": 1.5898741483688354,
+      "learning_rate": 5.794797687861272e-07,
+      "loss": 0.873,
+      "step": 18665
+    },
+    {
+      "epoch": 0.44965198333373474,
+      "grad_norm": 1.4367605447769165,
+      "learning_rate": 5.793530067944428e-07,
+      "loss": 0.846,
+      "step": 18670
+    },
+    {
+      "epoch": 0.449772404325522,
+      "grad_norm": 1.4805914163589478,
+      "learning_rate": 5.792262448027582e-07,
+      "loss": 0.9828,
+      "step": 18675
+    },
+    {
+      "epoch": 0.4498928253173093,
+      "grad_norm": 1.7379878759384155,
+      "learning_rate": 5.790994828110739e-07,
+      "loss": 0.9296,
+      "step": 18680
+    },
+    {
+      "epoch": 0.4500132463090966,
+      "grad_norm": 1.6005162000656128,
+      "learning_rate": 5.789727208193895e-07,
+      "loss": 0.953,
+      "step": 18685
+    },
+    {
+      "epoch": 0.4501336673008839,
+      "grad_norm": 1.7036206722259521,
+      "learning_rate": 5.788459588277051e-07,
+      "loss": 0.8372,
+      "step": 18690
+    },
+    {
+      "epoch": 0.45025408829267116,
+      "grad_norm": 1.5740010738372803,
+      "learning_rate": 5.787191968360207e-07,
+      "loss": 0.9642,
+      "step": 18695
+    },
+    {
+      "epoch": 0.4503745092844585,
+      "grad_norm": 1.5430411100387573,
+      "learning_rate": 5.785924348443362e-07,
+      "loss": 0.9323,
+      "step": 18700
+    },
+    {
+      "epoch": 0.45049493027624576,
+      "grad_norm": 1.446671724319458,
+      "learning_rate": 5.784656728526518e-07,
+      "loss": 0.9183,
+      "step": 18705
+    },
+    {
+      "epoch": 0.45061535126803304,
+      "grad_norm": 1.4905126094818115,
+      "learning_rate": 5.783389108609674e-07,
+      "loss": 0.897,
+      "step": 18710
+    },
+    {
+      "epoch": 0.4507357722598203,
+      "grad_norm": 1.5193527936935425,
+      "learning_rate": 5.782121488692831e-07,
+      "loss": 0.9159,
+      "step": 18715
+    },
+    {
+      "epoch": 0.45085619325160764,
+      "grad_norm": 1.6405127048492432,
+      "learning_rate": 5.780853868775986e-07,
+      "loss": 0.8922,
+      "step": 18720
+    },
+    {
+      "epoch": 0.4509766142433949,
+      "grad_norm": 1.4062474966049194,
+      "learning_rate": 5.779586248859141e-07,
+      "loss": 0.9168,
+      "step": 18725
+    },
+    {
+      "epoch": 0.4510970352351822,
+      "grad_norm": 1.3942046165466309,
+      "learning_rate": 5.778318628942298e-07,
+      "loss": 0.843,
+      "step": 18730
+    },
+    {
+      "epoch": 0.45121745622696946,
+      "grad_norm": 1.5190876722335815,
+      "learning_rate": 5.777051009025453e-07,
+      "loss": 0.895,
+      "step": 18735
+    },
+    {
+      "epoch": 0.4513378772187568,
+      "grad_norm": 1.42783784866333,
+      "learning_rate": 5.77578338910861e-07,
+      "loss": 0.8943,
+      "step": 18740
+    },
+    {
+      "epoch": 0.45145829821054406,
+      "grad_norm": 1.7751065492630005,
+      "learning_rate": 5.774515769191765e-07,
+      "loss": 0.8743,
+      "step": 18745
+    },
+    {
+      "epoch": 0.45157871920233134,
+      "grad_norm": 1.5588613748550415,
+      "learning_rate": 5.773248149274921e-07,
+      "loss": 0.9128,
+      "step": 18750
+    },
+    {
+      "epoch": 0.45169914019411866,
+      "grad_norm": 1.5651437044143677,
+      "learning_rate": 5.771980529358077e-07,
+      "loss": 0.8912,
+      "step": 18755
+    },
+    {
+      "epoch": 0.45181956118590594,
+      "grad_norm": 1.7240246534347534,
+      "learning_rate": 5.770712909441233e-07,
+      "loss": 0.9365,
+      "step": 18760
+    },
+    {
+      "epoch": 0.4519399821776932,
+      "grad_norm": 1.4902751445770264,
+      "learning_rate": 5.769445289524389e-07,
+      "loss": 0.8875,
+      "step": 18765
+    },
+    {
+      "epoch": 0.4520604031694805,
+      "grad_norm": 1.8112143278121948,
+      "learning_rate": 5.768177669607544e-07,
+      "loss": 0.9054,
+      "step": 18770
+    },
+    {
+      "epoch": 0.4521808241612678,
+      "grad_norm": 2.001507520675659,
+      "learning_rate": 5.766910049690701e-07,
+      "loss": 0.9174,
+      "step": 18775
+    },
+    {
+      "epoch": 0.4523012451530551,
+      "grad_norm": 1.568212628364563,
+      "learning_rate": 5.765642429773856e-07,
+      "loss": 0.9049,
+      "step": 18780
+    },
+    {
+      "epoch": 0.45242166614484236,
+      "grad_norm": 1.4313302040100098,
+      "learning_rate": 5.764374809857012e-07,
+      "loss": 0.9175,
+      "step": 18785
+    },
+    {
+      "epoch": 0.45254208713662963,
+      "grad_norm": 1.4737589359283447,
+      "learning_rate": 5.763107189940169e-07,
+      "loss": 0.8932,
+      "step": 18790
+    },
+    {
+      "epoch": 0.45266250812841696,
+      "grad_norm": 1.3767342567443848,
+      "learning_rate": 5.761839570023323e-07,
+      "loss": 0.9753,
+      "step": 18795
+    },
+    {
+      "epoch": 0.45278292912020424,
+      "grad_norm": 1.583419919013977,
+      "learning_rate": 5.76057195010648e-07,
+      "loss": 0.9356,
+      "step": 18800
+    },
+    {
+      "epoch": 0.4529033501119915,
+      "grad_norm": 1.5111521482467651,
+      "learning_rate": 5.759304330189636e-07,
+      "loss": 0.9451,
+      "step": 18805
+    },
+    {
+      "epoch": 0.45302377110377884,
+      "grad_norm": 1.6245722770690918,
+      "learning_rate": 5.758036710272792e-07,
+      "loss": 0.9008,
+      "step": 18810
+    },
+    {
+      "epoch": 0.4531441920955661,
+      "grad_norm": 1.539433240890503,
+      "learning_rate": 5.756769090355947e-07,
+      "loss": 0.9434,
+      "step": 18815
+    },
+    {
+      "epoch": 0.4532646130873534,
+      "grad_norm": 1.4418680667877197,
+      "learning_rate": 5.755501470439103e-07,
+      "loss": 0.9049,
+      "step": 18820
+    },
+    {
+      "epoch": 0.45338503407914066,
+      "grad_norm": 1.708088994026184,
+      "learning_rate": 5.754233850522259e-07,
+      "loss": 0.9196,
+      "step": 18825
+    },
+    {
+      "epoch": 0.453505455070928,
+      "grad_norm": 1.5235261917114258,
+      "learning_rate": 5.752966230605415e-07,
+      "loss": 0.9526,
+      "step": 18830
+    },
+    {
+      "epoch": 0.45362587606271526,
+      "grad_norm": 1.585042119026184,
+      "learning_rate": 5.751698610688572e-07,
+      "loss": 0.8745,
+      "step": 18835
+    },
+    {
+      "epoch": 0.45374629705450253,
+      "grad_norm": 1.689333200454712,
+      "learning_rate": 5.750430990771726e-07,
+      "loss": 0.9088,
+      "step": 18840
+    },
+    {
+      "epoch": 0.4538667180462898,
+      "grad_norm": 1.4670780897140503,
+      "learning_rate": 5.749163370854882e-07,
+      "loss": 0.9259,
+      "step": 18845
+    },
+    {
+      "epoch": 0.45398713903807714,
+      "grad_norm": 1.5117682218551636,
+      "learning_rate": 5.747895750938039e-07,
+      "loss": 0.8814,
+      "step": 18850
+    },
+    {
+      "epoch": 0.4541075600298644,
+      "grad_norm": 1.6144859790802002,
+      "learning_rate": 5.746628131021194e-07,
+      "loss": 0.9731,
+      "step": 18855
+    },
+    {
+      "epoch": 0.4542279810216517,
+      "grad_norm": 1.5023188591003418,
+      "learning_rate": 5.74536051110435e-07,
+      "loss": 0.8978,
+      "step": 18860
+    },
+    {
+      "epoch": 0.45434840201343896,
+      "grad_norm": 1.6316559314727783,
+      "learning_rate": 5.744092891187506e-07,
+      "loss": 0.9299,
+      "step": 18865
+    },
+    {
+      "epoch": 0.4544688230052263,
+      "grad_norm": 1.8118706941604614,
+      "learning_rate": 5.742825271270662e-07,
+      "loss": 0.8923,
+      "step": 18870
+    },
+    {
+      "epoch": 0.45458924399701356,
+      "grad_norm": 1.6119071245193481,
+      "learning_rate": 5.741557651353818e-07,
+      "loss": 0.8898,
+      "step": 18875
+    },
+    {
+      "epoch": 0.45470966498880083,
+      "grad_norm": 1.597812294960022,
+      "learning_rate": 5.740290031436974e-07,
+      "loss": 0.9185,
+      "step": 18880
+    },
+    {
+      "epoch": 0.45483008598058816,
+      "grad_norm": 1.587369680404663,
+      "learning_rate": 5.739022411520129e-07,
+      "loss": 0.9109,
+      "step": 18885
+    },
+    {
+      "epoch": 0.45495050697237543,
+      "grad_norm": 1.6539413928985596,
+      "learning_rate": 5.737754791603285e-07,
+      "loss": 0.8844,
+      "step": 18890
+    },
+    {
+      "epoch": 0.4550709279641627,
+      "grad_norm": 1.4326918125152588,
+      "learning_rate": 5.736487171686442e-07,
+      "loss": 0.9227,
+      "step": 18895
+    },
+    {
+      "epoch": 0.45519134895595,
+      "grad_norm": 1.6746175289154053,
+      "learning_rate": 5.735219551769597e-07,
+      "loss": 0.9478,
+      "step": 18900
+    },
+    {
+      "epoch": 0.4553117699477373,
+      "grad_norm": 1.4657456874847412,
+      "learning_rate": 5.733951931852753e-07,
+      "loss": 0.9131,
+      "step": 18905
+    },
+    {
+      "epoch": 0.4554321909395246,
+      "grad_norm": 1.613221287727356,
+      "learning_rate": 5.732684311935909e-07,
+      "loss": 0.9123,
+      "step": 18910
+    },
+    {
+      "epoch": 0.45555261193131186,
+      "grad_norm": 1.4709984064102173,
+      "learning_rate": 5.731416692019064e-07,
+      "loss": 0.856,
+      "step": 18915
+    },
+    {
+      "epoch": 0.45567303292309913,
+      "grad_norm": 1.6130681037902832,
+      "learning_rate": 5.730149072102221e-07,
+      "loss": 0.9035,
+      "step": 18920
+    },
+    {
+      "epoch": 0.45579345391488646,
+      "grad_norm": 1.5863080024719238,
+      "learning_rate": 5.728881452185377e-07,
+      "loss": 0.9249,
+      "step": 18925
+    },
+    {
+      "epoch": 0.45591387490667373,
+      "grad_norm": 1.593340516090393,
+      "learning_rate": 5.727613832268531e-07,
+      "loss": 0.9131,
+      "step": 18930
+    },
+    {
+      "epoch": 0.456034295898461,
+      "grad_norm": 1.613885521888733,
+      "learning_rate": 5.726346212351688e-07,
+      "loss": 0.914,
+      "step": 18935
+    },
+    {
+      "epoch": 0.45615471689024834,
+      "grad_norm": 1.6788082122802734,
+      "learning_rate": 5.725078592434844e-07,
+      "loss": 0.8911,
+      "step": 18940
+    },
+    {
+      "epoch": 0.4562751378820356,
+      "grad_norm": 1.7046750783920288,
+      "learning_rate": 5.723810972518e-07,
+      "loss": 0.9332,
+      "step": 18945
+    },
+    {
+      "epoch": 0.4563955588738229,
+      "grad_norm": 1.4113725423812866,
+      "learning_rate": 5.722543352601156e-07,
+      "loss": 0.9386,
+      "step": 18950
+    },
+    {
+      "epoch": 0.45651597986561016,
+      "grad_norm": 1.8018156290054321,
+      "learning_rate": 5.721275732684312e-07,
+      "loss": 0.9314,
+      "step": 18955
+    },
+    {
+      "epoch": 0.4566364008573975,
+      "grad_norm": 1.3819090127944946,
+      "learning_rate": 5.720008112767467e-07,
+      "loss": 0.9241,
+      "step": 18960
+    },
+    {
+      "epoch": 0.45675682184918476,
+      "grad_norm": 1.5696793794631958,
+      "learning_rate": 5.718740492850623e-07,
+      "loss": 0.9194,
+      "step": 18965
+    },
+    {
+      "epoch": 0.45687724284097203,
+      "grad_norm": 1.608102560043335,
+      "learning_rate": 5.71747287293378e-07,
+      "loss": 0.9204,
+      "step": 18970
+    },
+    {
+      "epoch": 0.4569976638327593,
+      "grad_norm": 1.5484436750411987,
+      "learning_rate": 5.716205253016934e-07,
+      "loss": 0.9414,
+      "step": 18975
+    },
+    {
+      "epoch": 0.45711808482454663,
+      "grad_norm": 1.551183819770813,
+      "learning_rate": 5.714937633100091e-07,
+      "loss": 0.9455,
+      "step": 18980
+    },
+    {
+      "epoch": 0.4572385058163339,
+      "grad_norm": 1.6090680360794067,
+      "learning_rate": 5.713670013183247e-07,
+      "loss": 0.9278,
+      "step": 18985
+    },
+    {
+      "epoch": 0.4573589268081212,
+      "grad_norm": 1.5775976181030273,
+      "learning_rate": 5.712402393266402e-07,
+      "loss": 0.9173,
+      "step": 18990
+    },
+    {
+      "epoch": 0.45747934779990845,
+      "grad_norm": 1.6444989442825317,
+      "learning_rate": 5.711134773349559e-07,
+      "loss": 0.9548,
+      "step": 18995
+    },
+    {
+      "epoch": 0.4575997687916958,
+      "grad_norm": 1.5032627582550049,
+      "learning_rate": 5.709867153432714e-07,
+      "loss": 0.9373,
+      "step": 19000
+    },
+    {
+      "epoch": 0.45772018978348306,
+      "grad_norm": 1.5852969884872437,
+      "learning_rate": 5.70859953351587e-07,
+      "loss": 0.926,
+      "step": 19005
+    },
+    {
+      "epoch": 0.45784061077527033,
+      "grad_norm": 1.5551520586013794,
+      "learning_rate": 5.707331913599026e-07,
+      "loss": 0.9585,
+      "step": 19010
+    },
+    {
+      "epoch": 0.45796103176705766,
+      "grad_norm": 1.562559962272644,
+      "learning_rate": 5.706064293682183e-07,
+      "loss": 0.9485,
+      "step": 19015
+    },
+    {
+      "epoch": 0.45808145275884493,
+      "grad_norm": 1.5225250720977783,
+      "learning_rate": 5.704796673765337e-07,
+      "loss": 0.8889,
+      "step": 19020
+    },
+    {
+      "epoch": 0.4582018737506322,
+      "grad_norm": 1.78077232837677,
+      "learning_rate": 5.703529053848493e-07,
+      "loss": 0.9894,
+      "step": 19025
+    },
+    {
+      "epoch": 0.4583222947424195,
+      "grad_norm": 1.660193681716919,
+      "learning_rate": 5.70226143393165e-07,
+      "loss": 0.9268,
+      "step": 19030
+    },
+    {
+      "epoch": 0.4584427157342068,
+      "grad_norm": 1.4802565574645996,
+      "learning_rate": 5.700993814014805e-07,
+      "loss": 0.9387,
+      "step": 19035
+    },
+    {
+      "epoch": 0.4585631367259941,
+      "grad_norm": 1.4176965951919556,
+      "learning_rate": 5.699726194097962e-07,
+      "loss": 0.917,
+      "step": 19040
+    },
+    {
+      "epoch": 0.45868355771778135,
+      "grad_norm": 1.7834274768829346,
+      "learning_rate": 5.698458574181117e-07,
+      "loss": 0.9135,
+      "step": 19045
+    },
+    {
+      "epoch": 0.4588039787095686,
+      "grad_norm": 1.7946574687957764,
+      "learning_rate": 5.697190954264272e-07,
+      "loss": 0.9362,
+      "step": 19050
+    },
+    {
+      "epoch": 0.45892439970135596,
+      "grad_norm": 1.5921515226364136,
+      "learning_rate": 5.695923334347429e-07,
+      "loss": 0.8743,
+      "step": 19055
+    },
+    {
+      "epoch": 0.45904482069314323,
+      "grad_norm": 1.4635932445526123,
+      "learning_rate": 5.694655714430585e-07,
+      "loss": 0.9383,
+      "step": 19060
+    },
+    {
+      "epoch": 0.4591652416849305,
+      "grad_norm": 1.5557299852371216,
+      "learning_rate": 5.693388094513741e-07,
+      "loss": 0.9065,
+      "step": 19065
+    },
+    {
+      "epoch": 0.45928566267671783,
+      "grad_norm": 1.3830097913742065,
+      "learning_rate": 5.692120474596896e-07,
+      "loss": 0.8941,
+      "step": 19070
+    },
+    {
+      "epoch": 0.4594060836685051,
+      "grad_norm": 1.644334316253662,
+      "learning_rate": 5.690852854680053e-07,
+      "loss": 0.903,
+      "step": 19075
+    },
+    {
+      "epoch": 0.4595265046602924,
+      "grad_norm": 1.4195899963378906,
+      "learning_rate": 5.689585234763208e-07,
+      "loss": 0.9194,
+      "step": 19080
+    },
+    {
+      "epoch": 0.45964692565207965,
+      "grad_norm": 1.5357506275177002,
+      "learning_rate": 5.688317614846364e-07,
+      "loss": 0.9624,
+      "step": 19085
+    },
+    {
+      "epoch": 0.459767346643867,
+      "grad_norm": 1.6892377138137817,
+      "learning_rate": 5.68704999492952e-07,
+      "loss": 0.9293,
+      "step": 19090
+    },
+    {
+      "epoch": 0.45988776763565425,
+      "grad_norm": 1.3714441061019897,
+      "learning_rate": 5.685782375012675e-07,
+      "loss": 0.8866,
+      "step": 19095
+    },
+    {
+      "epoch": 0.4600081886274415,
+      "grad_norm": 1.8560410737991333,
+      "learning_rate": 5.684514755095832e-07,
+      "loss": 0.9763,
+      "step": 19100
+    },
+    {
+      "epoch": 0.4601286096192288,
+      "grad_norm": 1.7453593015670776,
+      "learning_rate": 5.683247135178988e-07,
+      "loss": 0.914,
+      "step": 19105
+    },
+    {
+      "epoch": 0.46024903061101613,
+      "grad_norm": 1.5820306539535522,
+      "learning_rate": 5.681979515262143e-07,
+      "loss": 0.9245,
+      "step": 19110
+    },
+    {
+      "epoch": 0.4603694516028034,
+      "grad_norm": 1.48934006690979,
+      "learning_rate": 5.680711895345299e-07,
+      "loss": 0.8885,
+      "step": 19115
+    },
+    {
+      "epoch": 0.4604898725945907,
+      "grad_norm": 1.542831301689148,
+      "learning_rate": 5.679444275428455e-07,
+      "loss": 0.9049,
+      "step": 19120
+    },
+    {
+      "epoch": 0.46061029358637795,
+      "grad_norm": 1.5863815546035767,
+      "learning_rate": 5.678176655511611e-07,
+      "loss": 0.9294,
+      "step": 19125
+    },
+    {
+      "epoch": 0.4607307145781653,
+      "grad_norm": 1.705835223197937,
+      "learning_rate": 5.676909035594767e-07,
+      "loss": 0.9287,
+      "step": 19130
+    },
+    {
+      "epoch": 0.46085113556995255,
+      "grad_norm": 1.6327176094055176,
+      "learning_rate": 5.675641415677924e-07,
+      "loss": 0.9219,
+      "step": 19135
+    },
+    {
+      "epoch": 0.4609715565617398,
+      "grad_norm": 1.6016877889633179,
+      "learning_rate": 5.674373795761078e-07,
+      "loss": 0.8602,
+      "step": 19140
+    },
+    {
+      "epoch": 0.46109197755352715,
+      "grad_norm": 1.5826456546783447,
+      "learning_rate": 5.673106175844234e-07,
+      "loss": 0.9342,
+      "step": 19145
+    },
+    {
+      "epoch": 0.46121239854531443,
+      "grad_norm": 1.7967777252197266,
+      "learning_rate": 5.671838555927391e-07,
+      "loss": 0.8953,
+      "step": 19150
+    },
+    {
+      "epoch": 0.4613328195371017,
+      "grad_norm": 1.5923316478729248,
+      "learning_rate": 5.670570936010546e-07,
+      "loss": 0.9272,
+      "step": 19155
+    },
+    {
+      "epoch": 0.461453240528889,
+      "grad_norm": 1.5057430267333984,
+      "learning_rate": 5.669303316093702e-07,
+      "loss": 0.9219,
+      "step": 19160
+    },
+    {
+      "epoch": 0.4615736615206763,
+      "grad_norm": 1.4920909404754639,
+      "learning_rate": 5.668035696176858e-07,
+      "loss": 0.8777,
+      "step": 19165
+    },
+    {
+      "epoch": 0.4616940825124636,
+      "grad_norm": 1.5344513654708862,
+      "learning_rate": 5.666768076260013e-07,
+      "loss": 0.8931,
+      "step": 19170
+    },
+    {
+      "epoch": 0.46181450350425085,
+      "grad_norm": 1.4009016752243042,
+      "learning_rate": 5.66550045634317e-07,
+      "loss": 0.8792,
+      "step": 19175
+    },
+    {
+      "epoch": 0.4619349244960381,
+      "grad_norm": 1.6597226858139038,
+      "learning_rate": 5.664232836426326e-07,
+      "loss": 0.9244,
+      "step": 19180
+    },
+    {
+      "epoch": 0.46205534548782545,
+      "grad_norm": 1.7540236711502075,
+      "learning_rate": 5.662965216509481e-07,
+      "loss": 0.8937,
+      "step": 19185
+    },
+    {
+      "epoch": 0.4621757664796127,
+      "grad_norm": 1.5170707702636719,
+      "learning_rate": 5.661697596592637e-07,
+      "loss": 0.9194,
+      "step": 19190
+    },
+    {
+      "epoch": 0.4622961874714,
+      "grad_norm": 1.8222031593322754,
+      "learning_rate": 5.660429976675794e-07,
+      "loss": 0.9276,
+      "step": 19195
+    },
+    {
+      "epoch": 0.46241660846318733,
+      "grad_norm": 1.5705348253250122,
+      "learning_rate": 5.659162356758949e-07,
+      "loss": 0.9343,
+      "step": 19200
+    },
+    {
+      "epoch": 0.4625370294549746,
+      "grad_norm": 1.4817367792129517,
+      "learning_rate": 5.657894736842104e-07,
+      "loss": 0.9395,
+      "step": 19205
+    },
+    {
+      "epoch": 0.4626574504467619,
+      "grad_norm": 1.5779122114181519,
+      "learning_rate": 5.656627116925261e-07,
+      "loss": 0.8933,
+      "step": 19210
+    },
+    {
+      "epoch": 0.46277787143854915,
+      "grad_norm": 1.3455822467803955,
+      "learning_rate": 5.655359497008416e-07,
+      "loss": 0.9075,
+      "step": 19215
+    },
+    {
+      "epoch": 0.4628982924303365,
+      "grad_norm": 1.6021500825881958,
+      "learning_rate": 5.654091877091573e-07,
+      "loss": 0.9553,
+      "step": 19220
+    },
+    {
+      "epoch": 0.46301871342212375,
+      "grad_norm": 1.4690455198287964,
+      "learning_rate": 5.652824257174729e-07,
+      "loss": 0.9146,
+      "step": 19225
+    },
+    {
+      "epoch": 0.463139134413911,
+      "grad_norm": 1.4307332038879395,
+      "learning_rate": 5.651556637257883e-07,
+      "loss": 0.9641,
+      "step": 19230
+    },
+    {
+      "epoch": 0.4632595554056983,
+      "grad_norm": 1.579603672027588,
+      "learning_rate": 5.65028901734104e-07,
+      "loss": 0.8914,
+      "step": 19235
+    },
+    {
+      "epoch": 0.4633799763974856,
+      "grad_norm": 1.7392204999923706,
+      "learning_rate": 5.649021397424196e-07,
+      "loss": 0.8679,
+      "step": 19240
+    },
+    {
+      "epoch": 0.4635003973892729,
+      "grad_norm": 1.5015277862548828,
+      "learning_rate": 5.647753777507353e-07,
+      "loss": 0.9354,
+      "step": 19245
+    },
+    {
+      "epoch": 0.4636208183810602,
+      "grad_norm": 1.5179589986801147,
+      "learning_rate": 5.646486157590508e-07,
+      "loss": 0.9798,
+      "step": 19250
+    },
+    {
+      "epoch": 0.4637412393728475,
+      "grad_norm": 1.3988590240478516,
+      "learning_rate": 5.645218537673663e-07,
+      "loss": 0.9186,
+      "step": 19255
+    },
+    {
+      "epoch": 0.4638616603646348,
+      "grad_norm": 1.6146354675292969,
+      "learning_rate": 5.64395091775682e-07,
+      "loss": 0.8371,
+      "step": 19260
+    },
+    {
+      "epoch": 0.46398208135642205,
+      "grad_norm": 1.4624241590499878,
+      "learning_rate": 5.642683297839975e-07,
+      "loss": 0.9093,
+      "step": 19265
+    },
+    {
+      "epoch": 0.4641025023482093,
+      "grad_norm": 1.6220847368240356,
+      "learning_rate": 5.641415677923132e-07,
+      "loss": 0.8582,
+      "step": 19270
+    },
+    {
+      "epoch": 0.46422292333999665,
+      "grad_norm": 1.5948646068572998,
+      "learning_rate": 5.640148058006287e-07,
+      "loss": 0.937,
+      "step": 19275
+    },
+    {
+      "epoch": 0.4643433443317839,
+      "grad_norm": 1.5547462701797485,
+      "learning_rate": 5.638880438089443e-07,
+      "loss": 0.8907,
+      "step": 19280
+    },
+    {
+      "epoch": 0.4644637653235712,
+      "grad_norm": 1.6174482107162476,
+      "learning_rate": 5.637612818172599e-07,
+      "loss": 0.9369,
+      "step": 19285
+    },
+    {
+      "epoch": 0.46458418631535847,
+      "grad_norm": 1.4465388059616089,
+      "learning_rate": 5.636345198255755e-07,
+      "loss": 0.9182,
+      "step": 19290
+    },
+    {
+      "epoch": 0.4647046073071458,
+      "grad_norm": 1.8930598497390747,
+      "learning_rate": 5.635077578338911e-07,
+      "loss": 0.8905,
+      "step": 19295
+    },
+    {
+      "epoch": 0.4648250282989331,
+      "grad_norm": 1.3349248170852661,
+      "learning_rate": 5.633809958422066e-07,
+      "loss": 0.9202,
+      "step": 19300
+    },
+    {
+      "epoch": 0.46494544929072035,
+      "grad_norm": 1.38579523563385,
+      "learning_rate": 5.632542338505223e-07,
+      "loss": 0.9545,
+      "step": 19305
+    },
+    {
+      "epoch": 0.4650658702825076,
+      "grad_norm": 1.6983599662780762,
+      "learning_rate": 5.631274718588378e-07,
+      "loss": 0.9331,
+      "step": 19310
+    },
+    {
+      "epoch": 0.46518629127429495,
+      "grad_norm": 1.6366914510726929,
+      "learning_rate": 5.630007098671535e-07,
+      "loss": 0.9384,
+      "step": 19315
+    },
+    {
+      "epoch": 0.4653067122660822,
+      "grad_norm": 1.8956844806671143,
+      "learning_rate": 5.628739478754691e-07,
+      "loss": 0.8964,
+      "step": 19320
+    },
+    {
+      "epoch": 0.4654271332578695,
+      "grad_norm": 1.6888880729675293,
+      "learning_rate": 5.627471858837845e-07,
+      "loss": 0.8892,
+      "step": 19325
+    },
+    {
+      "epoch": 0.4655475542496568,
+      "grad_norm": 1.5253064632415771,
+      "learning_rate": 5.626204238921002e-07,
+      "loss": 0.8463,
+      "step": 19330
+    },
+    {
+      "epoch": 0.4656679752414441,
+      "grad_norm": 1.632535696029663,
+      "learning_rate": 5.624936619004158e-07,
+      "loss": 0.886,
+      "step": 19335
+    },
+    {
+      "epoch": 0.46578839623323137,
+      "grad_norm": 1.5638158321380615,
+      "learning_rate": 5.623668999087314e-07,
+      "loss": 0.95,
+      "step": 19340
+    },
+    {
+      "epoch": 0.46590881722501865,
+      "grad_norm": 1.78264582157135,
+      "learning_rate": 5.622401379170469e-07,
+      "loss": 0.9205,
+      "step": 19345
+    },
+    {
+      "epoch": 0.466029238216806,
+      "grad_norm": 1.6770012378692627,
+      "learning_rate": 5.621133759253625e-07,
+      "loss": 0.9276,
+      "step": 19350
+    },
+    {
+      "epoch": 0.46614965920859325,
+      "grad_norm": 1.6042697429656982,
+      "learning_rate": 5.619866139336781e-07,
+      "loss": 0.8913,
+      "step": 19355
+    },
+    {
+      "epoch": 0.4662700802003805,
+      "grad_norm": 1.4511951208114624,
+      "learning_rate": 5.618598519419937e-07,
+      "loss": 0.8983,
+      "step": 19360
+    },
+    {
+      "epoch": 0.4663905011921678,
+      "grad_norm": 1.4056437015533447,
+      "learning_rate": 5.617330899503094e-07,
+      "loss": 0.8683,
+      "step": 19365
+    },
+    {
+      "epoch": 0.4665109221839551,
+      "grad_norm": 1.493322491645813,
+      "learning_rate": 5.616063279586248e-07,
+      "loss": 0.916,
+      "step": 19370
+    },
+    {
+      "epoch": 0.4666313431757424,
+      "grad_norm": 1.4477388858795166,
+      "learning_rate": 5.614795659669404e-07,
+      "loss": 0.9469,
+      "step": 19375
+    },
+    {
+      "epoch": 0.46675176416752967,
+      "grad_norm": 1.498809576034546,
+      "learning_rate": 5.613528039752561e-07,
+      "loss": 0.9008,
+      "step": 19380
+    },
+    {
+      "epoch": 0.466872185159317,
+      "grad_norm": 1.6682100296020508,
+      "learning_rate": 5.612260419835716e-07,
+      "loss": 0.9416,
+      "step": 19385
+    },
+    {
+      "epoch": 0.4669926061511043,
+      "grad_norm": 1.6538997888565063,
+      "learning_rate": 5.610992799918872e-07,
+      "loss": 0.8469,
+      "step": 19390
+    },
+    {
+      "epoch": 0.46711302714289155,
+      "grad_norm": 1.5839837789535522,
+      "learning_rate": 5.609725180002028e-07,
+      "loss": 0.8992,
+      "step": 19395
+    },
+    {
+      "epoch": 0.4672334481346788,
+      "grad_norm": 1.6330792903900146,
+      "learning_rate": 5.608457560085184e-07,
+      "loss": 0.9026,
+      "step": 19400
+    },
+    {
+      "epoch": 0.46735386912646615,
+      "grad_norm": 1.4835282564163208,
+      "learning_rate": 5.60718994016834e-07,
+      "loss": 0.9264,
+      "step": 19405
+    },
+    {
+      "epoch": 0.4674742901182534,
+      "grad_norm": 1.6376011371612549,
+      "learning_rate": 5.605922320251496e-07,
+      "loss": 0.91,
+      "step": 19410
+    },
+    {
+      "epoch": 0.4675947111100407,
+      "grad_norm": 1.3809139728546143,
+      "learning_rate": 5.604654700334651e-07,
+      "loss": 0.908,
+      "step": 19415
+    },
+    {
+      "epoch": 0.46771513210182797,
+      "grad_norm": 1.5058355331420898,
+      "learning_rate": 5.603387080417807e-07,
+      "loss": 0.9446,
+      "step": 19420
+    },
+    {
+      "epoch": 0.4678355530936153,
+      "grad_norm": 1.6111805438995361,
+      "learning_rate": 5.602119460500964e-07,
+      "loss": 0.8859,
+      "step": 19425
+    },
+    {
+      "epoch": 0.46795597408540257,
+      "grad_norm": 1.508823037147522,
+      "learning_rate": 5.600851840584119e-07,
+      "loss": 0.9019,
+      "step": 19430
+    },
+    {
+      "epoch": 0.46807639507718984,
+      "grad_norm": 1.765920639038086,
+      "learning_rate": 5.599584220667275e-07,
+      "loss": 0.8805,
+      "step": 19435
+    },
+    {
+      "epoch": 0.4681968160689771,
+      "grad_norm": 1.5230178833007812,
+      "learning_rate": 5.598316600750431e-07,
+      "loss": 0.8594,
+      "step": 19440
+    },
+    {
+      "epoch": 0.46831723706076445,
+      "grad_norm": 1.5808008909225464,
+      "learning_rate": 5.597048980833586e-07,
+      "loss": 0.8729,
+      "step": 19445
+    },
+    {
+      "epoch": 0.4684376580525517,
+      "grad_norm": 1.6315981149673462,
+      "learning_rate": 5.595781360916743e-07,
+      "loss": 0.8992,
+      "step": 19450
+    },
+    {
+      "epoch": 0.468558079044339,
+      "grad_norm": 1.5495814085006714,
+      "learning_rate": 5.594513740999899e-07,
+      "loss": 0.9098,
+      "step": 19455
+    },
+    {
+      "epoch": 0.4686785000361263,
+      "grad_norm": 1.717214822769165,
+      "learning_rate": 5.593246121083053e-07,
+      "loss": 0.959,
+      "step": 19460
+    },
+    {
+      "epoch": 0.4687989210279136,
+      "grad_norm": 1.468532919883728,
+      "learning_rate": 5.59197850116621e-07,
+      "loss": 0.8781,
+      "step": 19465
+    },
+    {
+      "epoch": 0.46891934201970087,
+      "grad_norm": 1.4898240566253662,
+      "learning_rate": 5.590710881249366e-07,
+      "loss": 0.91,
+      "step": 19470
+    },
+    {
+      "epoch": 0.46903976301148814,
+      "grad_norm": 1.5860544443130493,
+      "learning_rate": 5.589443261332522e-07,
+      "loss": 0.8995,
+      "step": 19475
+    },
+    {
+      "epoch": 0.46916018400327547,
+      "grad_norm": 1.4536211490631104,
+      "learning_rate": 5.588175641415678e-07,
+      "loss": 0.9219,
+      "step": 19480
+    },
+    {
+      "epoch": 0.46928060499506274,
+      "grad_norm": 1.3934434652328491,
+      "learning_rate": 5.586908021498834e-07,
+      "loss": 0.8859,
+      "step": 19485
+    },
+    {
+      "epoch": 0.46940102598685,
+      "grad_norm": 1.4026354551315308,
+      "learning_rate": 5.585640401581989e-07,
+      "loss": 0.9145,
+      "step": 19490
+    },
+    {
+      "epoch": 0.4695214469786373,
+      "grad_norm": 1.5729193687438965,
+      "learning_rate": 5.584372781665145e-07,
+      "loss": 0.9152,
+      "step": 19495
+    },
+    {
+      "epoch": 0.4696418679704246,
+      "grad_norm": 1.5934133529663086,
+      "learning_rate": 5.583105161748302e-07,
+      "loss": 0.9541,
+      "step": 19500
+    },
+    {
+      "epoch": 0.4697622889622119,
+      "grad_norm": 1.3092372417449951,
+      "learning_rate": 5.581837541831456e-07,
+      "loss": 0.9333,
+      "step": 19505
+    },
+    {
+      "epoch": 0.46988270995399917,
+      "grad_norm": 1.6447803974151611,
+      "learning_rate": 5.580569921914613e-07,
+      "loss": 0.8509,
+      "step": 19510
+    },
+    {
+      "epoch": 0.4700031309457865,
+      "grad_norm": 1.5421675443649292,
+      "learning_rate": 5.579302301997769e-07,
+      "loss": 0.8606,
+      "step": 19515
+    },
+    {
+      "epoch": 0.47012355193757377,
+      "grad_norm": 1.5920692682266235,
+      "learning_rate": 5.578034682080924e-07,
+      "loss": 0.8919,
+      "step": 19520
+    },
+    {
+      "epoch": 0.47024397292936104,
+      "grad_norm": 1.4933526515960693,
+      "learning_rate": 5.576767062164081e-07,
+      "loss": 0.946,
+      "step": 19525
+    },
+    {
+      "epoch": 0.4703643939211483,
+      "grad_norm": 1.5667372941970825,
+      "learning_rate": 5.575499442247236e-07,
+      "loss": 0.8823,
+      "step": 19530
+    },
+    {
+      "epoch": 0.47048481491293564,
+      "grad_norm": 1.3437079191207886,
+      "learning_rate": 5.574231822330392e-07,
+      "loss": 0.9155,
+      "step": 19535
+    },
+    {
+      "epoch": 0.4706052359047229,
+      "grad_norm": 1.6043516397476196,
+      "learning_rate": 5.572964202413548e-07,
+      "loss": 0.9311,
+      "step": 19540
+    },
+    {
+      "epoch": 0.4707256568965102,
+      "grad_norm": 1.5555745363235474,
+      "learning_rate": 5.571696582496705e-07,
+      "loss": 0.9066,
+      "step": 19545
+    },
+    {
+      "epoch": 0.47084607788829747,
+      "grad_norm": 1.4690513610839844,
+      "learning_rate": 5.57042896257986e-07,
+      "loss": 0.915,
+      "step": 19550
+    },
+    {
+      "epoch": 0.4709664988800848,
+      "grad_norm": 1.412894368171692,
+      "learning_rate": 5.569161342663015e-07,
+      "loss": 0.9205,
+      "step": 19555
+    },
+    {
+      "epoch": 0.47108691987187207,
+      "grad_norm": 1.5198075771331787,
+      "learning_rate": 5.567893722746172e-07,
+      "loss": 0.9057,
+      "step": 19560
+    },
+    {
+      "epoch": 0.47120734086365934,
+      "grad_norm": 1.4156620502471924,
+      "learning_rate": 5.566626102829327e-07,
+      "loss": 0.9071,
+      "step": 19565
+    },
+    {
+      "epoch": 0.4713277618554466,
+      "grad_norm": 1.5950437784194946,
+      "learning_rate": 5.565358482912484e-07,
+      "loss": 0.926,
+      "step": 19570
+    },
+    {
+      "epoch": 0.47144818284723394,
+      "grad_norm": 1.900969386100769,
+      "learning_rate": 5.564090862995639e-07,
+      "loss": 0.8585,
+      "step": 19575
+    },
+    {
+      "epoch": 0.4715686038390212,
+      "grad_norm": 2.091481924057007,
+      "learning_rate": 5.562823243078794e-07,
+      "loss": 0.8834,
+      "step": 19580
+    },
+    {
+      "epoch": 0.4716890248308085,
+      "grad_norm": 1.8308582305908203,
+      "learning_rate": 5.561555623161951e-07,
+      "loss": 0.9513,
+      "step": 19585
+    },
+    {
+      "epoch": 0.4718094458225958,
+      "grad_norm": 1.667382001876831,
+      "learning_rate": 5.560288003245107e-07,
+      "loss": 0.8976,
+      "step": 19590
+    },
+    {
+      "epoch": 0.4719298668143831,
+      "grad_norm": 1.412461757659912,
+      "learning_rate": 5.559020383328263e-07,
+      "loss": 0.9328,
+      "step": 19595
+    },
+    {
+      "epoch": 0.47205028780617037,
+      "grad_norm": 1.6457659006118774,
+      "learning_rate": 5.557752763411418e-07,
+      "loss": 0.8995,
+      "step": 19600
+    },
+    {
+      "epoch": 0.47217070879795764,
+      "grad_norm": 1.5809874534606934,
+      "learning_rate": 5.556485143494575e-07,
+      "loss": 0.9108,
+      "step": 19605
+    },
+    {
+      "epoch": 0.47229112978974497,
+      "grad_norm": 1.529252290725708,
+      "learning_rate": 5.55521752357773e-07,
+      "loss": 0.8897,
+      "step": 19610
+    },
+    {
+      "epoch": 0.47241155078153224,
+      "grad_norm": 1.672222375869751,
+      "learning_rate": 5.553949903660886e-07,
+      "loss": 0.8975,
+      "step": 19615
+    },
+    {
+      "epoch": 0.4725319717733195,
+      "grad_norm": 1.5824137926101685,
+      "learning_rate": 5.552682283744043e-07,
+      "loss": 0.9066,
+      "step": 19620
+    },
+    {
+      "epoch": 0.4726523927651068,
+      "grad_norm": 1.9257863759994507,
+      "learning_rate": 5.551414663827197e-07,
+      "loss": 0.9294,
+      "step": 19625
+    },
+    {
+      "epoch": 0.4727728137568941,
+      "grad_norm": 1.6060237884521484,
+      "learning_rate": 5.550147043910354e-07,
+      "loss": 0.9267,
+      "step": 19630
+    },
+    {
+      "epoch": 0.4728932347486814,
+      "grad_norm": 1.5399199724197388,
+      "learning_rate": 5.54887942399351e-07,
+      "loss": 0.8888,
+      "step": 19635
+    },
+    {
+      "epoch": 0.47301365574046866,
+      "grad_norm": 1.5686084032058716,
+      "learning_rate": 5.547611804076665e-07,
+      "loss": 0.9176,
+      "step": 19640
+    },
+    {
+      "epoch": 0.473134076732256,
+      "grad_norm": 1.4253500699996948,
+      "learning_rate": 5.546344184159821e-07,
+      "loss": 0.9271,
+      "step": 19645
+    },
+    {
+      "epoch": 0.47325449772404327,
+      "grad_norm": 1.5969947576522827,
+      "learning_rate": 5.545076564242977e-07,
+      "loss": 0.9251,
+      "step": 19650
+    },
+    {
+      "epoch": 0.47337491871583054,
+      "grad_norm": 1.586877465248108,
+      "learning_rate": 5.543808944326133e-07,
+      "loss": 0.8853,
+      "step": 19655
+    },
+    {
+      "epoch": 0.4734953397076178,
+      "grad_norm": 1.579201340675354,
+      "learning_rate": 5.542541324409289e-07,
+      "loss": 0.8718,
+      "step": 19660
+    },
+    {
+      "epoch": 0.47361576069940514,
+      "grad_norm": 1.5453004837036133,
+      "learning_rate": 5.541273704492446e-07,
+      "loss": 0.8859,
+      "step": 19665
+    },
+    {
+      "epoch": 0.4737361816911924,
+      "grad_norm": 1.5748382806777954,
+      "learning_rate": 5.5400060845756e-07,
+      "loss": 0.8919,
+      "step": 19670
+    },
+    {
+      "epoch": 0.4738566026829797,
+      "grad_norm": 1.6768748760223389,
+      "learning_rate": 5.538738464658756e-07,
+      "loss": 0.8696,
+      "step": 19675
+    },
+    {
+      "epoch": 0.47397702367476696,
+      "grad_norm": 1.5177021026611328,
+      "learning_rate": 5.537470844741913e-07,
+      "loss": 0.8602,
+      "step": 19680
+    },
+    {
+      "epoch": 0.4740974446665543,
+      "grad_norm": 1.3657416105270386,
+      "learning_rate": 5.536203224825068e-07,
+      "loss": 0.8537,
+      "step": 19685
+    },
+    {
+      "epoch": 0.47421786565834156,
+      "grad_norm": 1.484498143196106,
+      "learning_rate": 5.534935604908224e-07,
+      "loss": 0.9626,
+      "step": 19690
+    },
+    {
+      "epoch": 0.47433828665012884,
+      "grad_norm": 1.3851714134216309,
+      "learning_rate": 5.53366798499138e-07,
+      "loss": 0.9084,
+      "step": 19695
+    },
+    {
+      "epoch": 0.47445870764191617,
+      "grad_norm": 1.6817436218261719,
+      "learning_rate": 5.532400365074535e-07,
+      "loss": 0.8925,
+      "step": 19700
+    },
+    {
+      "epoch": 0.47457912863370344,
+      "grad_norm": 1.6906062364578247,
+      "learning_rate": 5.531132745157692e-07,
+      "loss": 0.937,
+      "step": 19705
+    },
+    {
+      "epoch": 0.4746995496254907,
+      "grad_norm": 1.4443438053131104,
+      "learning_rate": 5.529865125240848e-07,
+      "loss": 0.9102,
+      "step": 19710
+    },
+    {
+      "epoch": 0.474819970617278,
+      "grad_norm": 1.527987003326416,
+      "learning_rate": 5.528597505324003e-07,
+      "loss": 0.8944,
+      "step": 19715
+    },
+    {
+      "epoch": 0.4749403916090653,
+      "grad_norm": 1.8593183755874634,
+      "learning_rate": 5.527329885407159e-07,
+      "loss": 0.8978,
+      "step": 19720
+    },
+    {
+      "epoch": 0.4750608126008526,
+      "grad_norm": 1.5174137353897095,
+      "learning_rate": 5.526062265490316e-07,
+      "loss": 0.8817,
+      "step": 19725
+    },
+    {
+      "epoch": 0.47518123359263986,
+      "grad_norm": 1.453525424003601,
+      "learning_rate": 5.524794645573471e-07,
+      "loss": 0.9118,
+      "step": 19730
+    },
+    {
+      "epoch": 0.47530165458442714,
+      "grad_norm": 1.5764050483703613,
+      "learning_rate": 5.523527025656627e-07,
+      "loss": 0.9359,
+      "step": 19735
+    },
+    {
+      "epoch": 0.47542207557621446,
+      "grad_norm": 1.7837824821472168,
+      "learning_rate": 5.522259405739783e-07,
+      "loss": 0.9578,
+      "step": 19740
+    },
+    {
+      "epoch": 0.47554249656800174,
+      "grad_norm": 1.5909477472305298,
+      "learning_rate": 5.520991785822938e-07,
+      "loss": 0.9389,
+      "step": 19745
+    },
+    {
+      "epoch": 0.475662917559789,
+      "grad_norm": 1.485216498374939,
+      "learning_rate": 5.519724165906095e-07,
+      "loss": 0.8666,
+      "step": 19750
+    },
+    {
+      "epoch": 0.4757833385515763,
+      "grad_norm": 1.3735274076461792,
+      "learning_rate": 5.518456545989251e-07,
+      "loss": 0.8216,
+      "step": 19755
+    },
+    {
+      "epoch": 0.4759037595433636,
+      "grad_norm": 1.5103178024291992,
+      "learning_rate": 5.517188926072405e-07,
+      "loss": 0.8766,
+      "step": 19760
+    },
+    {
+      "epoch": 0.4760241805351509,
+      "grad_norm": 1.46016263961792,
+      "learning_rate": 5.515921306155562e-07,
+      "loss": 0.9262,
+      "step": 19765
+    },
+    {
+      "epoch": 0.47614460152693816,
+      "grad_norm": 1.7239071130752563,
+      "learning_rate": 5.514653686238718e-07,
+      "loss": 0.9191,
+      "step": 19770
+    },
+    {
+      "epoch": 0.4762650225187255,
+      "grad_norm": 1.5743353366851807,
+      "learning_rate": 5.513386066321874e-07,
+      "loss": 0.9159,
+      "step": 19775
+    },
+    {
+      "epoch": 0.47638544351051276,
+      "grad_norm": 1.7983384132385254,
+      "learning_rate": 5.51211844640503e-07,
+      "loss": 0.925,
+      "step": 19780
+    },
+    {
+      "epoch": 0.47650586450230004,
+      "grad_norm": 1.5955098867416382,
+      "learning_rate": 5.510850826488185e-07,
+      "loss": 0.9405,
+      "step": 19785
+    },
+    {
+      "epoch": 0.4766262854940873,
+      "grad_norm": 1.530182123184204,
+      "learning_rate": 5.509583206571341e-07,
+      "loss": 0.9166,
+      "step": 19790
+    },
+    {
+      "epoch": 0.47674670648587464,
+      "grad_norm": 1.518515944480896,
+      "learning_rate": 5.508315586654497e-07,
+      "loss": 0.9451,
+      "step": 19795
+    },
+    {
+      "epoch": 0.4768671274776619,
+      "grad_norm": 1.643058180809021,
+      "learning_rate": 5.507047966737654e-07,
+      "loss": 0.8849,
+      "step": 19800
+    },
+    {
+      "epoch": 0.4769875484694492,
+      "grad_norm": 1.6911545991897583,
+      "learning_rate": 5.505780346820808e-07,
+      "loss": 0.8837,
+      "step": 19805
+    },
+    {
+      "epoch": 0.47710796946123646,
+      "grad_norm": 1.3822886943817139,
+      "learning_rate": 5.504512726903965e-07,
+      "loss": 0.8948,
+      "step": 19810
+    },
+    {
+      "epoch": 0.4772283904530238,
+      "grad_norm": 1.6200188398361206,
+      "learning_rate": 5.503245106987121e-07,
+      "loss": 0.9426,
+      "step": 19815
+    },
+    {
+      "epoch": 0.47734881144481106,
+      "grad_norm": 1.4352916479110718,
+      "learning_rate": 5.501977487070276e-07,
+      "loss": 0.8842,
+      "step": 19820
+    },
+    {
+      "epoch": 0.47746923243659833,
+      "grad_norm": 1.4828038215637207,
+      "learning_rate": 5.500709867153433e-07,
+      "loss": 0.8714,
+      "step": 19825
+    },
+    {
+      "epoch": 0.47758965342838566,
+      "grad_norm": 1.4750306606292725,
+      "learning_rate": 5.499442247236588e-07,
+      "loss": 0.9484,
+      "step": 19830
+    },
+    {
+      "epoch": 0.47771007442017294,
+      "grad_norm": 1.624162197113037,
+      "learning_rate": 5.498174627319744e-07,
+      "loss": 0.912,
+      "step": 19835
+    },
+    {
+      "epoch": 0.4778304954119602,
+      "grad_norm": 1.9018586874008179,
+      "learning_rate": 5.4969070074029e-07,
+      "loss": 0.8916,
+      "step": 19840
+    },
+    {
+      "epoch": 0.4779509164037475,
+      "grad_norm": 1.5208109617233276,
+      "learning_rate": 5.495639387486057e-07,
+      "loss": 0.9171,
+      "step": 19845
+    },
+    {
+      "epoch": 0.4780713373955348,
+      "grad_norm": 1.521644115447998,
+      "learning_rate": 5.494371767569212e-07,
+      "loss": 0.9251,
+      "step": 19850
+    },
+    {
+      "epoch": 0.4781917583873221,
+      "grad_norm": 1.4694710969924927,
+      "learning_rate": 5.493104147652367e-07,
+      "loss": 0.8663,
+      "step": 19855
+    },
+    {
+      "epoch": 0.47831217937910936,
+      "grad_norm": 1.6884132623672485,
+      "learning_rate": 5.491836527735524e-07,
+      "loss": 0.8857,
+      "step": 19860
+    },
+    {
+      "epoch": 0.47843260037089663,
+      "grad_norm": 1.7484136819839478,
+      "learning_rate": 5.490568907818679e-07,
+      "loss": 0.9107,
+      "step": 19865
+    },
+    {
+      "epoch": 0.47855302136268396,
+      "grad_norm": 1.915028691291809,
+      "learning_rate": 5.489301287901836e-07,
+      "loss": 0.8986,
+      "step": 19870
+    },
+    {
+      "epoch": 0.47867344235447123,
+      "grad_norm": 1.6282302141189575,
+      "learning_rate": 5.488033667984991e-07,
+      "loss": 0.91,
+      "step": 19875
+    },
+    {
+      "epoch": 0.4787938633462585,
+      "grad_norm": 1.9310438632965088,
+      "learning_rate": 5.486766048068146e-07,
+      "loss": 0.9392,
+      "step": 19880
+    },
+    {
+      "epoch": 0.4789142843380458,
+      "grad_norm": 1.624096393585205,
+      "learning_rate": 5.485498428151303e-07,
+      "loss": 0.8752,
+      "step": 19885
+    },
+    {
+      "epoch": 0.4790347053298331,
+      "grad_norm": 1.579325556755066,
+      "learning_rate": 5.484230808234459e-07,
+      "loss": 0.8679,
+      "step": 19890
+    },
+    {
+      "epoch": 0.4791551263216204,
+      "grad_norm": 1.6065541505813599,
+      "learning_rate": 5.482963188317615e-07,
+      "loss": 0.903,
+      "step": 19895
+    },
+    {
+      "epoch": 0.47927554731340766,
+      "grad_norm": 1.509092092514038,
+      "learning_rate": 5.48169556840077e-07,
+      "loss": 0.8833,
+      "step": 19900
+    },
+    {
+      "epoch": 0.479395968305195,
+      "grad_norm": 1.625239372253418,
+      "learning_rate": 5.480427948483926e-07,
+      "loss": 0.8796,
+      "step": 19905
+    },
+    {
+      "epoch": 0.47951638929698226,
+      "grad_norm": 1.5438594818115234,
+      "learning_rate": 5.479160328567082e-07,
+      "loss": 0.8598,
+      "step": 19910
+    },
+    {
+      "epoch": 0.47963681028876953,
+      "grad_norm": 1.5186994075775146,
+      "learning_rate": 5.477892708650238e-07,
+      "loss": 0.9471,
+      "step": 19915
+    },
+    {
+      "epoch": 0.4797572312805568,
+      "grad_norm": 1.6406010389328003,
+      "learning_rate": 5.476625088733395e-07,
+      "loss": 0.8856,
+      "step": 19920
+    },
+    {
+      "epoch": 0.47987765227234414,
+      "grad_norm": 1.6568893194198608,
+      "learning_rate": 5.475357468816549e-07,
+      "loss": 0.9892,
+      "step": 19925
+    },
+    {
+      "epoch": 0.4799980732641314,
+      "grad_norm": 1.6029942035675049,
+      "learning_rate": 5.474089848899706e-07,
+      "loss": 0.8746,
+      "step": 19930
+    },
+    {
+      "epoch": 0.4801184942559187,
+      "grad_norm": 1.5155408382415771,
+      "learning_rate": 5.472822228982862e-07,
+      "loss": 0.8756,
+      "step": 19935
+    },
+    {
+      "epoch": 0.48023891524770596,
+      "grad_norm": 1.720414638519287,
+      "learning_rate": 5.471554609066017e-07,
+      "loss": 0.8976,
+      "step": 19940
+    },
+    {
+      "epoch": 0.4803593362394933,
+      "grad_norm": 1.5202049016952515,
+      "learning_rate": 5.470286989149173e-07,
+      "loss": 0.9006,
+      "step": 19945
+    },
+    {
+      "epoch": 0.48047975723128056,
+      "grad_norm": 1.6243274211883545,
+      "learning_rate": 5.469019369232329e-07,
+      "loss": 0.9025,
+      "step": 19950
+    },
+    {
+      "epoch": 0.48060017822306783,
+      "grad_norm": 1.5468374490737915,
+      "learning_rate": 5.467751749315485e-07,
+      "loss": 0.891,
+      "step": 19955
+    },
+    {
+      "epoch": 0.48072059921485516,
+      "grad_norm": 1.7020221948623657,
+      "learning_rate": 5.466484129398641e-07,
+      "loss": 0.9049,
+      "step": 19960
+    },
+    {
+      "epoch": 0.48084102020664243,
+      "grad_norm": 1.6207133531570435,
+      "learning_rate": 5.465216509481797e-07,
+      "loss": 0.9273,
+      "step": 19965
+    },
+    {
+      "epoch": 0.4809614411984297,
+      "grad_norm": 1.7263379096984863,
+      "learning_rate": 5.463948889564952e-07,
+      "loss": 0.904,
+      "step": 19970
+    },
+    {
+      "epoch": 0.481081862190217,
+      "grad_norm": 1.5589679479599,
+      "learning_rate": 5.462681269648108e-07,
+      "loss": 0.9133,
+      "step": 19975
+    },
+    {
+      "epoch": 0.4812022831820043,
+      "grad_norm": 1.868116021156311,
+      "learning_rate": 5.461413649731265e-07,
+      "loss": 0.9224,
+      "step": 19980
+    },
+    {
+      "epoch": 0.4813227041737916,
+      "grad_norm": 1.5733035802841187,
+      "learning_rate": 5.46014602981442e-07,
+      "loss": 0.9143,
+      "step": 19985
+    },
+    {
+      "epoch": 0.48144312516557886,
+      "grad_norm": 1.7124382257461548,
+      "learning_rate": 5.458878409897575e-07,
+      "loss": 0.8175,
+      "step": 19990
+    },
+    {
+      "epoch": 0.48156354615736613,
+      "grad_norm": 1.5227895975112915,
+      "learning_rate": 5.457610789980732e-07,
+      "loss": 0.8792,
+      "step": 19995
+    },
+    {
+      "epoch": 0.48168396714915346,
+      "grad_norm": 1.7614829540252686,
+      "learning_rate": 5.456343170063887e-07,
+      "loss": 0.9365,
+      "step": 20000
+    },
+    {
+      "epoch": 0.48180438814094073,
+      "grad_norm": 1.6550016403198242,
+      "learning_rate": 5.455075550147044e-07,
+      "loss": 0.9716,
+      "step": 20005
+    },
+    {
+      "epoch": 0.481924809132728,
+      "grad_norm": 1.5123074054718018,
+      "learning_rate": 5.4538079302302e-07,
+      "loss": 0.8834,
+      "step": 20010
+    },
+    {
+      "epoch": 0.4820452301245153,
+      "grad_norm": 1.648113489151001,
+      "learning_rate": 5.452540310313355e-07,
+      "loss": 0.9213,
+      "step": 20015
+    },
+    {
+      "epoch": 0.4821656511163026,
+      "grad_norm": 1.904733657836914,
+      "learning_rate": 5.451272690396511e-07,
+      "loss": 0.8647,
+      "step": 20020
+    },
+    {
+      "epoch": 0.4822860721080899,
+      "grad_norm": 1.5189578533172607,
+      "learning_rate": 5.450005070479667e-07,
+      "loss": 0.8972,
+      "step": 20025
+    },
+    {
+      "epoch": 0.48240649309987715,
+      "grad_norm": 1.5393345355987549,
+      "learning_rate": 5.448737450562823e-07,
+      "loss": 0.8859,
+      "step": 20030
+    },
+    {
+      "epoch": 0.4825269140916645,
+      "grad_norm": 1.5878273248672485,
+      "learning_rate": 5.447469830645979e-07,
+      "loss": 0.9554,
+      "step": 20035
+    },
+    {
+      "epoch": 0.48264733508345176,
+      "grad_norm": 1.5807021856307983,
+      "learning_rate": 5.446202210729135e-07,
+      "loss": 0.9277,
+      "step": 20040
+    },
+    {
+      "epoch": 0.48276775607523903,
+      "grad_norm": 1.415024995803833,
+      "learning_rate": 5.44493459081229e-07,
+      "loss": 0.8944,
+      "step": 20045
+    },
+    {
+      "epoch": 0.4828881770670263,
+      "grad_norm": 1.553511381149292,
+      "learning_rate": 5.443666970895446e-07,
+      "loss": 0.9404,
+      "step": 20050
+    },
+    {
+      "epoch": 0.48300859805881363,
+      "grad_norm": 1.4661325216293335,
+      "learning_rate": 5.442399350978603e-07,
+      "loss": 0.8628,
+      "step": 20055
+    },
+    {
+      "epoch": 0.4831290190506009,
+      "grad_norm": 1.4524530172348022,
+      "learning_rate": 5.441131731061757e-07,
+      "loss": 0.8834,
+      "step": 20060
+    },
+    {
+      "epoch": 0.4832494400423882,
+      "grad_norm": 1.4855055809020996,
+      "learning_rate": 5.439864111144914e-07,
+      "loss": 0.899,
+      "step": 20065
+    },
+    {
+      "epoch": 0.48336986103417545,
+      "grad_norm": 1.4527695178985596,
+      "learning_rate": 5.43859649122807e-07,
+      "loss": 0.9224,
+      "step": 20070
+    },
+    {
+      "epoch": 0.4834902820259628,
+      "grad_norm": 1.781144380569458,
+      "learning_rate": 5.437328871311226e-07,
+      "loss": 0.8588,
+      "step": 20075
+    },
+    {
+      "epoch": 0.48361070301775005,
+      "grad_norm": 1.6836347579956055,
+      "learning_rate": 5.436061251394382e-07,
+      "loss": 0.9171,
+      "step": 20080
+    },
+    {
+      "epoch": 0.4837311240095373,
+      "grad_norm": 1.5334843397140503,
+      "learning_rate": 5.434793631477537e-07,
+      "loss": 0.9702,
+      "step": 20085
+    },
+    {
+      "epoch": 0.48385154500132466,
+      "grad_norm": 1.467098593711853,
+      "learning_rate": 5.433526011560693e-07,
+      "loss": 0.9252,
+      "step": 20090
+    },
+    {
+      "epoch": 0.48397196599311193,
+      "grad_norm": 1.5818877220153809,
+      "learning_rate": 5.432258391643849e-07,
+      "loss": 0.9293,
+      "step": 20095
+    },
+    {
+      "epoch": 0.4840923869848992,
+      "grad_norm": 1.555888056755066,
+      "learning_rate": 5.430990771727006e-07,
+      "loss": 0.8787,
+      "step": 20100
+    },
+    {
+      "epoch": 0.4842128079766865,
+      "grad_norm": 1.4838998317718506,
+      "learning_rate": 5.42972315181016e-07,
+      "loss": 0.9016,
+      "step": 20105
+    },
+    {
+      "epoch": 0.4843332289684738,
+      "grad_norm": 1.4996261596679688,
+      "learning_rate": 5.428455531893316e-07,
+      "loss": 0.9398,
+      "step": 20110
+    },
+    {
+      "epoch": 0.4844536499602611,
+      "grad_norm": 2.0929925441741943,
+      "learning_rate": 5.427187911976473e-07,
+      "loss": 0.9469,
+      "step": 20115
+    },
+    {
+      "epoch": 0.48457407095204835,
+      "grad_norm": 1.6425294876098633,
+      "learning_rate": 5.425920292059628e-07,
+      "loss": 0.8702,
+      "step": 20120
+    },
+    {
+      "epoch": 0.4846944919438356,
+      "grad_norm": 1.707254409790039,
+      "learning_rate": 5.424652672142785e-07,
+      "loss": 0.9226,
+      "step": 20125
+    },
+    {
+      "epoch": 0.48481491293562295,
+      "grad_norm": 1.6218721866607666,
+      "learning_rate": 5.42338505222594e-07,
+      "loss": 0.917,
+      "step": 20130
+    },
+    {
+      "epoch": 0.48493533392741023,
+      "grad_norm": 1.6185005903244019,
+      "learning_rate": 5.422117432309096e-07,
+      "loss": 0.872,
+      "step": 20135
+    },
+    {
+      "epoch": 0.4850557549191975,
+      "grad_norm": 1.5673400163650513,
+      "learning_rate": 5.420849812392252e-07,
+      "loss": 0.8663,
+      "step": 20140
+    },
+    {
+      "epoch": 0.48517617591098483,
+      "grad_norm": 1.6191776990890503,
+      "learning_rate": 5.419582192475408e-07,
+      "loss": 0.9139,
+      "step": 20145
+    },
+    {
+      "epoch": 0.4852965969027721,
+      "grad_norm": 1.4292807579040527,
+      "learning_rate": 5.418314572558564e-07,
+      "loss": 0.9326,
+      "step": 20150
+    },
+    {
+      "epoch": 0.4854170178945594,
+      "grad_norm": 2.3301453590393066,
+      "learning_rate": 5.417046952641719e-07,
+      "loss": 0.9048,
+      "step": 20155
+    },
+    {
+      "epoch": 0.48553743888634665,
+      "grad_norm": 1.5287058353424072,
+      "learning_rate": 5.415779332724876e-07,
+      "loss": 0.952,
+      "step": 20160
+    },
+    {
+      "epoch": 0.485657859878134,
+      "grad_norm": 1.7118383646011353,
+      "learning_rate": 5.414511712808031e-07,
+      "loss": 0.9114,
+      "step": 20165
+    },
+    {
+      "epoch": 0.48577828086992125,
+      "grad_norm": 1.7365562915802002,
+      "learning_rate": 5.413244092891187e-07,
+      "loss": 0.8634,
+      "step": 20170
+    },
+    {
+      "epoch": 0.4858987018617085,
+      "grad_norm": 1.5098973512649536,
+      "learning_rate": 5.411976472974343e-07,
+      "loss": 0.8725,
+      "step": 20175
+    },
+    {
+      "epoch": 0.4860191228534958,
+      "grad_norm": 2.2831051349639893,
+      "learning_rate": 5.410708853057498e-07,
+      "loss": 0.9016,
+      "step": 20180
+    },
+    {
+      "epoch": 0.48613954384528313,
+      "grad_norm": 1.7645684480667114,
+      "learning_rate": 5.409441233140655e-07,
+      "loss": 0.9091,
+      "step": 20185
+    },
+    {
+      "epoch": 0.4862599648370704,
+      "grad_norm": 1.58481764793396,
+      "learning_rate": 5.408173613223811e-07,
+      "loss": 0.9248,
+      "step": 20190
+    },
+    {
+      "epoch": 0.4863803858288577,
+      "grad_norm": 1.5158801078796387,
+      "learning_rate": 5.406905993306967e-07,
+      "loss": 0.8814,
+      "step": 20195
+    },
+    {
+      "epoch": 0.48650080682064495,
+      "grad_norm": 1.5112053155899048,
+      "learning_rate": 5.405638373390122e-07,
+      "loss": 0.9759,
+      "step": 20200
+    },
+    {
+      "epoch": 0.4866212278124323,
+      "grad_norm": 1.5802388191223145,
+      "learning_rate": 5.404370753473278e-07,
+      "loss": 0.8858,
+      "step": 20205
+    },
+    {
+      "epoch": 0.48674164880421955,
+      "grad_norm": 1.7204270362854004,
+      "learning_rate": 5.403103133556434e-07,
+      "loss": 0.9064,
+      "step": 20210
+    },
+    {
+      "epoch": 0.4868620697960068,
+      "grad_norm": 1.6286576986312866,
+      "learning_rate": 5.40183551363959e-07,
+      "loss": 0.9187,
+      "step": 20215
+    },
+    {
+      "epoch": 0.48698249078779415,
+      "grad_norm": 1.6453478336334229,
+      "learning_rate": 5.400567893722747e-07,
+      "loss": 0.9228,
+      "step": 20220
+    },
+    {
+      "epoch": 0.4871029117795814,
+      "grad_norm": 1.5130614042282104,
+      "learning_rate": 5.399300273805901e-07,
+      "loss": 0.9352,
+      "step": 20225
+    },
+    {
+      "epoch": 0.4872233327713687,
+      "grad_norm": 1.6289459466934204,
+      "learning_rate": 5.398032653889057e-07,
+      "loss": 0.8983,
+      "step": 20230
+    },
+    {
+      "epoch": 0.487343753763156,
+      "grad_norm": 1.3544524908065796,
+      "learning_rate": 5.396765033972214e-07,
+      "loss": 0.9375,
+      "step": 20235
+    },
+    {
+      "epoch": 0.4874641747549433,
+      "grad_norm": 1.5215047597885132,
+      "learning_rate": 5.395497414055369e-07,
+      "loss": 0.9086,
+      "step": 20240
+    },
+    {
+      "epoch": 0.4875845957467306,
+      "grad_norm": 1.5918240547180176,
+      "learning_rate": 5.394229794138525e-07,
+      "loss": 0.8656,
+      "step": 20245
+    },
+    {
+      "epoch": 0.48770501673851785,
+      "grad_norm": 1.523152470588684,
+      "learning_rate": 5.392962174221681e-07,
+      "loss": 0.9532,
+      "step": 20250
+    },
+    {
+      "epoch": 0.4878254377303051,
+      "grad_norm": 1.923143744468689,
+      "learning_rate": 5.391694554304836e-07,
+      "loss": 0.8834,
+      "step": 20255
+    },
+    {
+      "epoch": 0.48794585872209245,
+      "grad_norm": 1.4190473556518555,
+      "learning_rate": 5.390426934387993e-07,
+      "loss": 0.8735,
+      "step": 20260
+    },
+    {
+      "epoch": 0.4880662797138797,
+      "grad_norm": 1.439806342124939,
+      "learning_rate": 5.389159314471149e-07,
+      "loss": 0.8817,
+      "step": 20265
+    },
+    {
+      "epoch": 0.488186700705667,
+      "grad_norm": 1.634311318397522,
+      "learning_rate": 5.387891694554305e-07,
+      "loss": 0.9512,
+      "step": 20270
+    },
+    {
+      "epoch": 0.4883071216974543,
+      "grad_norm": 1.5653709173202515,
+      "learning_rate": 5.38662407463746e-07,
+      "loss": 0.9068,
+      "step": 20275
+    },
+    {
+      "epoch": 0.4884275426892416,
+      "grad_norm": 1.5077821016311646,
+      "learning_rate": 5.385356454720617e-07,
+      "loss": 0.8538,
+      "step": 20280
+    },
+    {
+      "epoch": 0.4885479636810289,
+      "grad_norm": 1.911816120147705,
+      "learning_rate": 5.384088834803773e-07,
+      "loss": 0.9354,
+      "step": 20285
+    },
+    {
+      "epoch": 0.48866838467281615,
+      "grad_norm": 1.5661500692367554,
+      "learning_rate": 5.382821214886927e-07,
+      "loss": 0.8956,
+      "step": 20290
+    },
+    {
+      "epoch": 0.4887888056646035,
+      "grad_norm": 1.4549514055252075,
+      "learning_rate": 5.381553594970084e-07,
+      "loss": 0.8628,
+      "step": 20295
+    },
+    {
+      "epoch": 0.48890922665639075,
+      "grad_norm": 1.549526572227478,
+      "learning_rate": 5.38028597505324e-07,
+      "loss": 0.9558,
+      "step": 20300
+    },
+    {
+      "epoch": 0.489029647648178,
+      "grad_norm": 1.5282357931137085,
+      "learning_rate": 5.379018355136396e-07,
+      "loss": 0.9305,
+      "step": 20305
+    },
+    {
+      "epoch": 0.4891500686399653,
+      "grad_norm": 1.5047615766525269,
+      "learning_rate": 5.377750735219552e-07,
+      "loss": 0.8253,
+      "step": 20310
+    },
+    {
+      "epoch": 0.4892704896317526,
+      "grad_norm": 1.5259456634521484,
+      "learning_rate": 5.376483115302708e-07,
+      "loss": 0.8936,
+      "step": 20315
+    },
+    {
+      "epoch": 0.4893909106235399,
+      "grad_norm": 1.5735572576522827,
+      "learning_rate": 5.375215495385863e-07,
+      "loss": 0.9149,
+      "step": 20320
+    },
+    {
+      "epoch": 0.48951133161532717,
+      "grad_norm": 1.8593881130218506,
+      "learning_rate": 5.373947875469019e-07,
+      "loss": 0.8886,
+      "step": 20325
+    },
+    {
+      "epoch": 0.48963175260711445,
+      "grad_norm": 1.5736616849899292,
+      "learning_rate": 5.372680255552176e-07,
+      "loss": 0.9432,
+      "step": 20330
+    },
+    {
+      "epoch": 0.4897521735989018,
+      "grad_norm": 1.6624977588653564,
+      "learning_rate": 5.371412635635331e-07,
+      "loss": 0.9397,
+      "step": 20335
+    },
+    {
+      "epoch": 0.48987259459068905,
+      "grad_norm": 1.5041385889053345,
+      "learning_rate": 5.370145015718487e-07,
+      "loss": 0.8978,
+      "step": 20340
+    },
+    {
+      "epoch": 0.4899930155824763,
+      "grad_norm": 1.363087773323059,
+      "learning_rate": 5.368877395801643e-07,
+      "loss": 0.8678,
+      "step": 20345
+    },
+    {
+      "epoch": 0.49011343657426365,
+      "grad_norm": 1.5875787734985352,
+      "learning_rate": 5.367609775884798e-07,
+      "loss": 0.8805,
+      "step": 20350
+    },
+    {
+      "epoch": 0.4902338575660509,
+      "grad_norm": 1.5088090896606445,
+      "learning_rate": 5.366342155967955e-07,
+      "loss": 0.913,
+      "step": 20355
+    },
+    {
+      "epoch": 0.4903542785578382,
+      "grad_norm": 1.4936821460723877,
+      "learning_rate": 5.36507453605111e-07,
+      "loss": 0.9112,
+      "step": 20360
+    },
+    {
+      "epoch": 0.49047469954962547,
+      "grad_norm": 1.4798657894134521,
+      "learning_rate": 5.363806916134266e-07,
+      "loss": 0.9366,
+      "step": 20365
+    },
+    {
+      "epoch": 0.4905951205414128,
+      "grad_norm": 1.4514061212539673,
+      "learning_rate": 5.362539296217422e-07,
+      "loss": 0.9174,
+      "step": 20370
+    },
+    {
+      "epoch": 0.4907155415332001,
+      "grad_norm": 1.5838004350662231,
+      "learning_rate": 5.361271676300579e-07,
+      "loss": 0.9367,
+      "step": 20375
+    },
+    {
+      "epoch": 0.49083596252498735,
+      "grad_norm": 1.5880588293075562,
+      "learning_rate": 5.360004056383734e-07,
+      "loss": 0.8806,
+      "step": 20380
+    },
+    {
+      "epoch": 0.4909563835167746,
+      "grad_norm": 1.36306893825531,
+      "learning_rate": 5.358736436466889e-07,
+      "loss": 0.8844,
+      "step": 20385
+    },
+    {
+      "epoch": 0.49107680450856195,
+      "grad_norm": 1.5225731134414673,
+      "learning_rate": 5.357468816550046e-07,
+      "loss": 0.9375,
+      "step": 20390
+    },
+    {
+      "epoch": 0.4911972255003492,
+      "grad_norm": 1.4811948537826538,
+      "learning_rate": 5.356201196633201e-07,
+      "loss": 0.8907,
+      "step": 20395
+    },
+    {
+      "epoch": 0.4913176464921365,
+      "grad_norm": 1.5076699256896973,
+      "learning_rate": 5.354933576716358e-07,
+      "loss": 0.9257,
+      "step": 20400
+    },
+    {
+      "epoch": 0.4914380674839238,
+      "grad_norm": 1.6767728328704834,
+      "learning_rate": 5.353665956799513e-07,
+      "loss": 0.8771,
+      "step": 20405
+    },
+    {
+      "epoch": 0.4915584884757111,
+      "grad_norm": 1.5127720832824707,
+      "learning_rate": 5.352398336882668e-07,
+      "loss": 0.9205,
+      "step": 20410
+    },
+    {
+      "epoch": 0.49167890946749837,
+      "grad_norm": 1.4718294143676758,
+      "learning_rate": 5.351130716965825e-07,
+      "loss": 0.941,
+      "step": 20415
+    },
+    {
+      "epoch": 0.49179933045928564,
+      "grad_norm": 1.371346116065979,
+      "learning_rate": 5.349863097048981e-07,
+      "loss": 0.8647,
+      "step": 20420
+    },
+    {
+      "epoch": 0.491919751451073,
+      "grad_norm": 1.5402333736419678,
+      "learning_rate": 5.348595477132137e-07,
+      "loss": 0.8228,
+      "step": 20425
+    },
+    {
+      "epoch": 0.49204017244286025,
+      "grad_norm": 1.554704189300537,
+      "learning_rate": 5.347327857215292e-07,
+      "loss": 0.8427,
+      "step": 20430
+    },
+    {
+      "epoch": 0.4921605934346475,
+      "grad_norm": 1.4905873537063599,
+      "learning_rate": 5.346060237298448e-07,
+      "loss": 0.9247,
+      "step": 20435
+    },
+    {
+      "epoch": 0.4922810144264348,
+      "grad_norm": 1.5101425647735596,
+      "learning_rate": 5.344792617381604e-07,
+      "loss": 0.8644,
+      "step": 20440
+    },
+    {
+      "epoch": 0.4924014354182221,
+      "grad_norm": 1.3385801315307617,
+      "learning_rate": 5.34352499746476e-07,
+      "loss": 0.8604,
+      "step": 20445
+    },
+    {
+      "epoch": 0.4925218564100094,
+      "grad_norm": 1.6283985376358032,
+      "learning_rate": 5.342257377547917e-07,
+      "loss": 0.8637,
+      "step": 20450
+    },
+    {
+      "epoch": 0.49264227740179667,
+      "grad_norm": 1.491113305091858,
+      "learning_rate": 5.340989757631071e-07,
+      "loss": 0.9338,
+      "step": 20455
+    },
+    {
+      "epoch": 0.49276269839358394,
+      "grad_norm": 1.8177359104156494,
+      "learning_rate": 5.339722137714228e-07,
+      "loss": 0.8777,
+      "step": 20460
+    },
+    {
+      "epoch": 0.49288311938537127,
+      "grad_norm": 1.573534607887268,
+      "learning_rate": 5.338454517797384e-07,
+      "loss": 0.9414,
+      "step": 20465
+    },
+    {
+      "epoch": 0.49300354037715854,
+      "grad_norm": 1.958213448524475,
+      "learning_rate": 5.337186897880539e-07,
+      "loss": 0.9488,
+      "step": 20470
+    },
+    {
+      "epoch": 0.4931239613689458,
+      "grad_norm": 1.4002881050109863,
+      "learning_rate": 5.335919277963695e-07,
+      "loss": 0.9283,
+      "step": 20475
+    },
+    {
+      "epoch": 0.49324438236073315,
+      "grad_norm": 1.6372034549713135,
+      "learning_rate": 5.334651658046851e-07,
+      "loss": 0.9578,
+      "step": 20480
+    },
+    {
+      "epoch": 0.4933648033525204,
+      "grad_norm": 1.5431723594665527,
+      "learning_rate": 5.333384038130007e-07,
+      "loss": 0.893,
+      "step": 20485
+    },
+    {
+      "epoch": 0.4934852243443077,
+      "grad_norm": 2.078256368637085,
+      "learning_rate": 5.332116418213163e-07,
+      "loss": 0.9235,
+      "step": 20490
+    },
+    {
+      "epoch": 0.49360564533609497,
+      "grad_norm": 1.4664630889892578,
+      "learning_rate": 5.33084879829632e-07,
+      "loss": 0.8957,
+      "step": 20495
+    },
+    {
+      "epoch": 0.4937260663278823,
+      "grad_norm": 1.4795653820037842,
+      "learning_rate": 5.329581178379474e-07,
+      "loss": 0.8877,
+      "step": 20500
+    },
+    {
+      "epoch": 0.49384648731966957,
+      "grad_norm": 1.6357628107070923,
+      "learning_rate": 5.32831355846263e-07,
+      "loss": 0.9041,
+      "step": 20505
+    },
+    {
+      "epoch": 0.49396690831145684,
+      "grad_norm": 1.5398471355438232,
+      "learning_rate": 5.327045938545787e-07,
+      "loss": 0.9487,
+      "step": 20510
+    },
+    {
+      "epoch": 0.4940873293032441,
+      "grad_norm": 1.6561620235443115,
+      "learning_rate": 5.325778318628942e-07,
+      "loss": 0.8507,
+      "step": 20515
+    },
+    {
+      "epoch": 0.49420775029503144,
+      "grad_norm": 1.5499076843261719,
+      "learning_rate": 5.324510698712097e-07,
+      "loss": 0.8637,
+      "step": 20520
+    },
+    {
+      "epoch": 0.4943281712868187,
+      "grad_norm": 1.4668582677841187,
+      "learning_rate": 5.323243078795254e-07,
+      "loss": 0.8989,
+      "step": 20525
+    },
+    {
+      "epoch": 0.494448592278606,
+      "grad_norm": 1.350682020187378,
+      "learning_rate": 5.321975458878409e-07,
+      "loss": 0.8597,
+      "step": 20530
+    },
+    {
+      "epoch": 0.4945690132703933,
+      "grad_norm": 1.4209243059158325,
+      "learning_rate": 5.320707838961566e-07,
+      "loss": 0.961,
+      "step": 20535
+    },
+    {
+      "epoch": 0.4946894342621806,
+      "grad_norm": 1.4936949014663696,
+      "learning_rate": 5.319440219044722e-07,
+      "loss": 0.8972,
+      "step": 20540
+    },
+    {
+      "epoch": 0.49480985525396787,
+      "grad_norm": 1.6538289785385132,
+      "learning_rate": 5.318172599127877e-07,
+      "loss": 0.9161,
+      "step": 20545
+    },
+    {
+      "epoch": 0.49493027624575514,
+      "grad_norm": 1.5872825384140015,
+      "learning_rate": 5.316904979211033e-07,
+      "loss": 0.856,
+      "step": 20550
+    },
+    {
+      "epoch": 0.49505069723754247,
+      "grad_norm": 1.6348376274108887,
+      "learning_rate": 5.315637359294189e-07,
+      "loss": 0.944,
+      "step": 20555
+    },
+    {
+      "epoch": 0.49517111822932974,
+      "grad_norm": 1.436761498451233,
+      "learning_rate": 5.314369739377345e-07,
+      "loss": 0.8888,
+      "step": 20560
+    },
+    {
+      "epoch": 0.495291539221117,
+      "grad_norm": 1.2941811084747314,
+      "learning_rate": 5.313102119460501e-07,
+      "loss": 0.8922,
+      "step": 20565
+    },
+    {
+      "epoch": 0.4954119602129043,
+      "grad_norm": 1.4106999635696411,
+      "learning_rate": 5.311834499543657e-07,
+      "loss": 0.9333,
+      "step": 20570
+    },
+    {
+      "epoch": 0.4955323812046916,
+      "grad_norm": 1.6659363508224487,
+      "learning_rate": 5.310566879626812e-07,
+      "loss": 0.9821,
+      "step": 20575
+    },
+    {
+      "epoch": 0.4956528021964789,
+      "grad_norm": 1.6132570505142212,
+      "learning_rate": 5.309299259709969e-07,
+      "loss": 0.8684,
+      "step": 20580
+    },
+    {
+      "epoch": 0.49577322318826617,
+      "grad_norm": 1.6882359981536865,
+      "learning_rate": 5.308031639793125e-07,
+      "loss": 0.8775,
+      "step": 20585
+    },
+    {
+      "epoch": 0.4958936441800535,
+      "grad_norm": 1.4514377117156982,
+      "learning_rate": 5.306764019876279e-07,
+      "loss": 0.95,
+      "step": 20590
+    },
+    {
+      "epoch": 0.49601406517184077,
+      "grad_norm": 1.6146961450576782,
+      "learning_rate": 5.305496399959436e-07,
+      "loss": 0.9243,
+      "step": 20595
+    },
+    {
+      "epoch": 0.49613448616362804,
+      "grad_norm": 1.6913578510284424,
+      "learning_rate": 5.304228780042592e-07,
+      "loss": 0.9379,
+      "step": 20600
+    },
+    {
+      "epoch": 0.4962549071554153,
+      "grad_norm": 1.521320104598999,
+      "learning_rate": 5.302961160125748e-07,
+      "loss": 0.8975,
+      "step": 20605
+    },
+    {
+      "epoch": 0.49637532814720264,
+      "grad_norm": 1.677592158317566,
+      "learning_rate": 5.301693540208904e-07,
+      "loss": 0.8753,
+      "step": 20610
+    },
+    {
+      "epoch": 0.4964957491389899,
+      "grad_norm": 1.498340129852295,
+      "learning_rate": 5.300425920292059e-07,
+      "loss": 0.872,
+      "step": 20615
+    },
+    {
+      "epoch": 0.4966161701307772,
+      "grad_norm": 1.684953212738037,
+      "learning_rate": 5.299158300375215e-07,
+      "loss": 0.952,
+      "step": 20620
+    },
+    {
+      "epoch": 0.49673659112256446,
+      "grad_norm": 1.5729172229766846,
+      "learning_rate": 5.297890680458371e-07,
+      "loss": 0.9443,
+      "step": 20625
+    },
+    {
+      "epoch": 0.4968570121143518,
+      "grad_norm": 1.7612193822860718,
+      "learning_rate": 5.296623060541528e-07,
+      "loss": 0.9308,
+      "step": 20630
+    },
+    {
+      "epoch": 0.49697743310613907,
+      "grad_norm": 1.7964433431625366,
+      "learning_rate": 5.295355440624682e-07,
+      "loss": 0.9247,
+      "step": 20635
+    },
+    {
+      "epoch": 0.49709785409792634,
+      "grad_norm": 1.82650625705719,
+      "learning_rate": 5.294087820707838e-07,
+      "loss": 0.95,
+      "step": 20640
+    },
+    {
+      "epoch": 0.4972182750897136,
+      "grad_norm": 1.417760968208313,
+      "learning_rate": 5.292820200790995e-07,
+      "loss": 0.9364,
+      "step": 20645
+    },
+    {
+      "epoch": 0.49733869608150094,
+      "grad_norm": 1.361054539680481,
+      "learning_rate": 5.29155258087415e-07,
+      "loss": 0.9345,
+      "step": 20650
+    },
+    {
+      "epoch": 0.4974591170732882,
+      "grad_norm": 1.4298704862594604,
+      "learning_rate": 5.290284960957307e-07,
+      "loss": 0.9849,
+      "step": 20655
+    },
+    {
+      "epoch": 0.4975795380650755,
+      "grad_norm": 1.7009376287460327,
+      "learning_rate": 5.289017341040462e-07,
+      "loss": 0.9051,
+      "step": 20660
+    },
+    {
+      "epoch": 0.4976999590568628,
+      "grad_norm": 1.4122412204742432,
+      "learning_rate": 5.287749721123618e-07,
+      "loss": 0.8759,
+      "step": 20665
+    },
+    {
+      "epoch": 0.4978203800486501,
+      "grad_norm": 1.7190061807632446,
+      "learning_rate": 5.286482101206774e-07,
+      "loss": 0.9077,
+      "step": 20670
+    },
+    {
+      "epoch": 0.49794080104043736,
+      "grad_norm": 1.5603126287460327,
+      "learning_rate": 5.28521448128993e-07,
+      "loss": 0.86,
+      "step": 20675
+    },
+    {
+      "epoch": 0.49806122203222464,
+      "grad_norm": 1.577416181564331,
+      "learning_rate": 5.283946861373086e-07,
+      "loss": 0.9753,
+      "step": 20680
+    },
+    {
+      "epoch": 0.49818164302401197,
+      "grad_norm": 1.4858267307281494,
+      "learning_rate": 5.282679241456241e-07,
+      "loss": 0.8542,
+      "step": 20685
+    },
+    {
+      "epoch": 0.49830206401579924,
+      "grad_norm": 1.597087025642395,
+      "learning_rate": 5.281411621539398e-07,
+      "loss": 0.8425,
+      "step": 20690
+    },
+    {
+      "epoch": 0.4984224850075865,
+      "grad_norm": 1.4735974073410034,
+      "learning_rate": 5.280144001622553e-07,
+      "loss": 0.9159,
+      "step": 20695
+    },
+    {
+      "epoch": 0.4985429059993738,
+      "grad_norm": 1.483817458152771,
+      "learning_rate": 5.27887638170571e-07,
+      "loss": 0.8468,
+      "step": 20700
+    },
+    {
+      "epoch": 0.4986633269911611,
+      "grad_norm": 2.0631678104400635,
+      "learning_rate": 5.277608761788865e-07,
+      "loss": 0.8393,
+      "step": 20705
+    },
+    {
+      "epoch": 0.4987837479829484,
+      "grad_norm": 1.443610668182373,
+      "learning_rate": 5.27634114187202e-07,
+      "loss": 0.9125,
+      "step": 20710
+    },
+    {
+      "epoch": 0.49890416897473566,
+      "grad_norm": 1.6772816181182861,
+      "learning_rate": 5.275073521955177e-07,
+      "loss": 0.8757,
+      "step": 20715
+    },
+    {
+      "epoch": 0.499024589966523,
+      "grad_norm": 1.4628548622131348,
+      "learning_rate": 5.273805902038333e-07,
+      "loss": 0.9609,
+      "step": 20720
+    },
+    {
+      "epoch": 0.49914501095831026,
+      "grad_norm": 1.4909286499023438,
+      "learning_rate": 5.272538282121489e-07,
+      "loss": 0.9024,
+      "step": 20725
+    },
+    {
+      "epoch": 0.49926543195009754,
+      "grad_norm": 1.5248208045959473,
+      "learning_rate": 5.271270662204644e-07,
+      "loss": 0.8609,
+      "step": 20730
+    },
+    {
+      "epoch": 0.4993858529418848,
+      "grad_norm": 1.7877882719039917,
+      "learning_rate": 5.2700030422878e-07,
+      "loss": 0.8867,
+      "step": 20735
+    },
+    {
+      "epoch": 0.49950627393367214,
+      "grad_norm": 1.4456939697265625,
+      "learning_rate": 5.268735422370956e-07,
+      "loss": 0.8803,
+      "step": 20740
+    },
+    {
+      "epoch": 0.4996266949254594,
+      "grad_norm": 1.5736900568008423,
+      "learning_rate": 5.267467802454112e-07,
+      "loss": 0.864,
+      "step": 20745
+    },
+    {
+      "epoch": 0.4997471159172467,
+      "grad_norm": 1.497660756111145,
+      "learning_rate": 5.266200182537269e-07,
+      "loss": 0.9403,
+      "step": 20750
+    },
+    {
+      "epoch": 0.49986753690903396,
+      "grad_norm": 1.4500926733016968,
+      "learning_rate": 5.264932562620423e-07,
+      "loss": 0.9268,
+      "step": 20755
+    },
+    {
+      "epoch": 0.4999879579008213,
+      "grad_norm": 1.5371147394180298,
+      "learning_rate": 5.263664942703579e-07,
+      "loss": 0.9218,
+      "step": 20760
+    },
+    {
+      "epoch": 0.5001083788926085,
+      "grad_norm": 1.5035649538040161,
+      "learning_rate": 5.262397322786736e-07,
+      "loss": 0.8771,
+      "step": 20765
+    },
+    {
+      "epoch": 0.5002287998843958,
+      "grad_norm": 1.7202868461608887,
+      "learning_rate": 5.261129702869891e-07,
+      "loss": 0.9059,
+      "step": 20770
+    },
+    {
+      "epoch": 0.5003492208761832,
+      "grad_norm": 1.4313465356826782,
+      "learning_rate": 5.259862082953047e-07,
+      "loss": 0.9434,
+      "step": 20775
+    },
+    {
+      "epoch": 0.5004696418679704,
+      "grad_norm": 1.7266535758972168,
+      "learning_rate": 5.258594463036203e-07,
+      "loss": 0.9265,
+      "step": 20780
+    },
+    {
+      "epoch": 0.5005900628597577,
+      "grad_norm": 1.4797693490982056,
+      "learning_rate": 5.257326843119358e-07,
+      "loss": 0.9025,
+      "step": 20785
+    },
+    {
+      "epoch": 0.500710483851545,
+      "grad_norm": 1.5850337743759155,
+      "learning_rate": 5.256059223202515e-07,
+      "loss": 0.8846,
+      "step": 20790
+    },
+    {
+      "epoch": 0.5008309048433323,
+      "grad_norm": 1.5374903678894043,
+      "learning_rate": 5.254791603285671e-07,
+      "loss": 0.9484,
+      "step": 20795
+    },
+    {
+      "epoch": 0.5009513258351196,
+      "grad_norm": 1.6088073253631592,
+      "learning_rate": 5.253523983368826e-07,
+      "loss": 0.8937,
+      "step": 20800
+    },
+    {
+      "epoch": 0.5010717468269069,
+      "grad_norm": 1.4605728387832642,
+      "learning_rate": 5.252256363451982e-07,
+      "loss": 0.8833,
+      "step": 20805
+    },
+    {
+      "epoch": 0.5011921678186941,
+      "grad_norm": 1.5764379501342773,
+      "learning_rate": 5.250988743535139e-07,
+      "loss": 0.8859,
+      "step": 20810
+    },
+    {
+      "epoch": 0.5013125888104815,
+      "grad_norm": 1.687082290649414,
+      "learning_rate": 5.249721123618294e-07,
+      "loss": 0.9462,
+      "step": 20815
+    },
+    {
+      "epoch": 0.5014330098022687,
+      "grad_norm": 1.431099534034729,
+      "learning_rate": 5.248453503701449e-07,
+      "loss": 0.8679,
+      "step": 20820
+    },
+    {
+      "epoch": 0.501553430794056,
+      "grad_norm": 1.5310856103897095,
+      "learning_rate": 5.247185883784606e-07,
+      "loss": 0.929,
+      "step": 20825
+    },
+    {
+      "epoch": 0.5016738517858433,
+      "grad_norm": 1.4208393096923828,
+      "learning_rate": 5.245918263867761e-07,
+      "loss": 0.8908,
+      "step": 20830
+    },
+    {
+      "epoch": 0.5017942727776306,
+      "grad_norm": 1.3968501091003418,
+      "learning_rate": 5.244650643950918e-07,
+      "loss": 0.8749,
+      "step": 20835
+    },
+    {
+      "epoch": 0.5019146937694179,
+      "grad_norm": 1.4458131790161133,
+      "learning_rate": 5.243383024034074e-07,
+      "loss": 0.9871,
+      "step": 20840
+    },
+    {
+      "epoch": 0.5020351147612052,
+      "grad_norm": 1.638501763343811,
+      "learning_rate": 5.242115404117228e-07,
+      "loss": 0.9494,
+      "step": 20845
+    },
+    {
+      "epoch": 0.5021555357529924,
+      "grad_norm": 2.0048091411590576,
+      "learning_rate": 5.240847784200385e-07,
+      "loss": 0.8665,
+      "step": 20850
+    },
+    {
+      "epoch": 0.5022759567447798,
+      "grad_norm": 1.4994688034057617,
+      "learning_rate": 5.239580164283541e-07,
+      "loss": 0.9034,
+      "step": 20855
+    },
+    {
+      "epoch": 0.5023963777365671,
+      "grad_norm": 1.4207297563552856,
+      "learning_rate": 5.238312544366697e-07,
+      "loss": 0.9173,
+      "step": 20860
+    },
+    {
+      "epoch": 0.5025167987283543,
+      "grad_norm": 1.6375930309295654,
+      "learning_rate": 5.237044924449853e-07,
+      "loss": 0.9082,
+      "step": 20865
+    },
+    {
+      "epoch": 0.5026372197201416,
+      "grad_norm": 1.6961311101913452,
+      "learning_rate": 5.235777304533009e-07,
+      "loss": 0.9142,
+      "step": 20870
+    },
+    {
+      "epoch": 0.5027576407119289,
+      "grad_norm": 1.500553011894226,
+      "learning_rate": 5.234509684616164e-07,
+      "loss": 0.8992,
+      "step": 20875
+    },
+    {
+      "epoch": 0.5028780617037162,
+      "grad_norm": 1.6316986083984375,
+      "learning_rate": 5.23324206469932e-07,
+      "loss": 0.8745,
+      "step": 20880
+    },
+    {
+      "epoch": 0.5029984826955035,
+      "grad_norm": 1.5603423118591309,
+      "learning_rate": 5.231974444782477e-07,
+      "loss": 0.913,
+      "step": 20885
+    },
+    {
+      "epoch": 0.5031189036872907,
+      "grad_norm": 1.532484531402588,
+      "learning_rate": 5.230706824865631e-07,
+      "loss": 0.8834,
+      "step": 20890
+    },
+    {
+      "epoch": 0.5032393246790781,
+      "grad_norm": 1.8932911157608032,
+      "learning_rate": 5.229439204948788e-07,
+      "loss": 0.904,
+      "step": 20895
+    },
+    {
+      "epoch": 0.5033597456708654,
+      "grad_norm": 1.5125609636306763,
+      "learning_rate": 5.228171585031944e-07,
+      "loss": 0.8856,
+      "step": 20900
+    },
+    {
+      "epoch": 0.5034801666626526,
+      "grad_norm": 1.4216543436050415,
+      "learning_rate": 5.226903965115099e-07,
+      "loss": 0.9319,
+      "step": 20905
+    },
+    {
+      "epoch": 0.5036005876544399,
+      "grad_norm": 1.739990234375,
+      "learning_rate": 5.225636345198256e-07,
+      "loss": 0.8709,
+      "step": 20910
+    },
+    {
+      "epoch": 0.5037210086462273,
+      "grad_norm": 1.427313208580017,
+      "learning_rate": 5.224368725281411e-07,
+      "loss": 0.8995,
+      "step": 20915
+    },
+    {
+      "epoch": 0.5038414296380145,
+      "grad_norm": 1.5396627187728882,
+      "learning_rate": 5.223101105364567e-07,
+      "loss": 0.9248,
+      "step": 20920
+    },
+    {
+      "epoch": 0.5039618506298018,
+      "grad_norm": 1.3887165784835815,
+      "learning_rate": 5.221833485447723e-07,
+      "loss": 0.895,
+      "step": 20925
+    },
+    {
+      "epoch": 0.504082271621589,
+      "grad_norm": 1.5276862382888794,
+      "learning_rate": 5.22056586553088e-07,
+      "loss": 0.8515,
+      "step": 20930
+    },
+    {
+      "epoch": 0.5042026926133764,
+      "grad_norm": 1.6163544654846191,
+      "learning_rate": 5.219298245614034e-07,
+      "loss": 0.8576,
+      "step": 20935
+    },
+    {
+      "epoch": 0.5043231136051637,
+      "grad_norm": 1.6514025926589966,
+      "learning_rate": 5.21803062569719e-07,
+      "loss": 0.9141,
+      "step": 20940
+    },
+    {
+      "epoch": 0.5044435345969509,
+      "grad_norm": 1.6270771026611328,
+      "learning_rate": 5.216763005780347e-07,
+      "loss": 0.8765,
+      "step": 20945
+    },
+    {
+      "epoch": 0.5045639555887382,
+      "grad_norm": 1.5601426362991333,
+      "learning_rate": 5.215495385863502e-07,
+      "loss": 0.8455,
+      "step": 20950
+    },
+    {
+      "epoch": 0.5046843765805256,
+      "grad_norm": 1.6472203731536865,
+      "learning_rate": 5.214227765946659e-07,
+      "loss": 0.8874,
+      "step": 20955
+    },
+    {
+      "epoch": 0.5048047975723128,
+      "grad_norm": 1.4407867193222046,
+      "learning_rate": 5.212960146029814e-07,
+      "loss": 0.902,
+      "step": 20960
+    },
+    {
+      "epoch": 0.5049252185641001,
+      "grad_norm": 1.5756430625915527,
+      "learning_rate": 5.211692526112969e-07,
+      "loss": 0.9313,
+      "step": 20965
+    },
+    {
+      "epoch": 0.5050456395558873,
+      "grad_norm": 1.4176348447799683,
+      "learning_rate": 5.210424906196126e-07,
+      "loss": 0.9203,
+      "step": 20970
+    },
+    {
+      "epoch": 0.5051660605476747,
+      "grad_norm": 1.592527151107788,
+      "learning_rate": 5.209157286279282e-07,
+      "loss": 0.9171,
+      "step": 20975
+    },
+    {
+      "epoch": 0.505286481539462,
+      "grad_norm": 1.5845403671264648,
+      "learning_rate": 5.207889666362438e-07,
+      "loss": 0.9015,
+      "step": 20980
+    },
+    {
+      "epoch": 0.5054069025312492,
+      "grad_norm": 1.458234190940857,
+      "learning_rate": 5.206622046445593e-07,
+      "loss": 0.8887,
+      "step": 20985
+    },
+    {
+      "epoch": 0.5055273235230365,
+      "grad_norm": 1.497634768486023,
+      "learning_rate": 5.20535442652875e-07,
+      "loss": 0.8902,
+      "step": 20990
+    },
+    {
+      "epoch": 0.5056477445148239,
+      "grad_norm": 1.5629130601882935,
+      "learning_rate": 5.204086806611905e-07,
+      "loss": 0.9415,
+      "step": 20995
+    },
+    {
+      "epoch": 0.5057681655066111,
+      "grad_norm": 1.4125767946243286,
+      "learning_rate": 5.202819186695061e-07,
+      "loss": 0.892,
+      "step": 21000
+    },
+    {
+      "epoch": 0.5058885864983984,
+      "grad_norm": 1.7816734313964844,
+      "learning_rate": 5.201551566778217e-07,
+      "loss": 0.8771,
+      "step": 21005
+    },
+    {
+      "epoch": 0.5060090074901857,
+      "grad_norm": 1.4958962202072144,
+      "learning_rate": 5.200283946861372e-07,
+      "loss": 0.9123,
+      "step": 21010
+    },
+    {
+      "epoch": 0.506129428481973,
+      "grad_norm": 1.4297374486923218,
+      "learning_rate": 5.199016326944529e-07,
+      "loss": 0.9241,
+      "step": 21015
+    },
+    {
+      "epoch": 0.5062498494737603,
+      "grad_norm": 1.6796749830245972,
+      "learning_rate": 5.197748707027685e-07,
+      "loss": 0.9057,
+      "step": 21020
+    },
+    {
+      "epoch": 0.5063702704655475,
+      "grad_norm": 1.3399157524108887,
+      "learning_rate": 5.19648108711084e-07,
+      "loss": 0.8854,
+      "step": 21025
+    },
+    {
+      "epoch": 0.5064906914573348,
+      "grad_norm": 1.734200358390808,
+      "learning_rate": 5.195213467193996e-07,
+      "loss": 0.9212,
+      "step": 21030
+    },
+    {
+      "epoch": 0.5066111124491222,
+      "grad_norm": 1.377709150314331,
+      "learning_rate": 5.193945847277152e-07,
+      "loss": 0.8889,
+      "step": 21035
+    },
+    {
+      "epoch": 0.5067315334409094,
+      "grad_norm": 1.5945323705673218,
+      "learning_rate": 5.192678227360308e-07,
+      "loss": 0.9097,
+      "step": 21040
+    },
+    {
+      "epoch": 0.5068519544326967,
+      "grad_norm": 1.528894305229187,
+      "learning_rate": 5.191410607443464e-07,
+      "loss": 0.9147,
+      "step": 21045
+    },
+    {
+      "epoch": 0.506972375424484,
+      "grad_norm": 1.6350053548812866,
+      "learning_rate": 5.190142987526621e-07,
+      "loss": 0.9181,
+      "step": 21050
+    },
+    {
+      "epoch": 0.5070927964162713,
+      "grad_norm": 1.4047951698303223,
+      "learning_rate": 5.188875367609775e-07,
+      "loss": 0.9403,
+      "step": 21055
+    },
+    {
+      "epoch": 0.5072132174080586,
+      "grad_norm": 1.8590868711471558,
+      "learning_rate": 5.187607747692931e-07,
+      "loss": 0.8864,
+      "step": 21060
+    },
+    {
+      "epoch": 0.5073336383998459,
+      "grad_norm": 1.7068983316421509,
+      "learning_rate": 5.186340127776088e-07,
+      "loss": 0.9501,
+      "step": 21065
+    },
+    {
+      "epoch": 0.5074540593916331,
+      "grad_norm": 1.4411805868148804,
+      "learning_rate": 5.185072507859243e-07,
+      "loss": 0.9195,
+      "step": 21070
+    },
+    {
+      "epoch": 0.5075744803834205,
+      "grad_norm": 1.4354585409164429,
+      "learning_rate": 5.183804887942399e-07,
+      "loss": 0.8736,
+      "step": 21075
+    },
+    {
+      "epoch": 0.5076949013752077,
+      "grad_norm": 1.5002541542053223,
+      "learning_rate": 5.182537268025555e-07,
+      "loss": 0.8807,
+      "step": 21080
+    },
+    {
+      "epoch": 0.507815322366995,
+      "grad_norm": 1.4930588006973267,
+      "learning_rate": 5.18126964810871e-07,
+      "loss": 0.9046,
+      "step": 21085
+    },
+    {
+      "epoch": 0.5079357433587823,
+      "grad_norm": 1.43183171749115,
+      "learning_rate": 5.180002028191867e-07,
+      "loss": 0.9073,
+      "step": 21090
+    },
+    {
+      "epoch": 0.5080561643505696,
+      "grad_norm": 1.4301148653030396,
+      "learning_rate": 5.178734408275023e-07,
+      "loss": 0.9289,
+      "step": 21095
+    },
+    {
+      "epoch": 0.5081765853423569,
+      "grad_norm": 1.5547271966934204,
+      "learning_rate": 5.177466788358178e-07,
+      "loss": 0.917,
+      "step": 21100
+    },
+    {
+      "epoch": 0.5082970063341442,
+      "grad_norm": 1.5416436195373535,
+      "learning_rate": 5.176199168441334e-07,
+      "loss": 0.9489,
+      "step": 21105
+    },
+    {
+      "epoch": 0.5084174273259314,
+      "grad_norm": 1.5894789695739746,
+      "learning_rate": 5.17493154852449e-07,
+      "loss": 0.8787,
+      "step": 21110
+    },
+    {
+      "epoch": 0.5085378483177188,
+      "grad_norm": 1.686223030090332,
+      "learning_rate": 5.173663928607646e-07,
+      "loss": 0.912,
+      "step": 21115
+    },
+    {
+      "epoch": 0.5086582693095061,
+      "grad_norm": 1.5424811840057373,
+      "learning_rate": 5.172396308690801e-07,
+      "loss": 0.8377,
+      "step": 21120
+    },
+    {
+      "epoch": 0.5087786903012933,
+      "grad_norm": 1.7151985168457031,
+      "learning_rate": 5.171128688773958e-07,
+      "loss": 0.9181,
+      "step": 21125
+    },
+    {
+      "epoch": 0.5088991112930806,
+      "grad_norm": 3.097501277923584,
+      "learning_rate": 5.169861068857113e-07,
+      "loss": 0.8949,
+      "step": 21130
+    },
+    {
+      "epoch": 0.5090195322848678,
+      "grad_norm": 1.4970147609710693,
+      "learning_rate": 5.16859344894027e-07,
+      "loss": 0.8966,
+      "step": 21135
+    },
+    {
+      "epoch": 0.5091399532766552,
+      "grad_norm": 1.6402392387390137,
+      "learning_rate": 5.167325829023426e-07,
+      "loss": 0.9161,
+      "step": 21140
+    },
+    {
+      "epoch": 0.5092603742684425,
+      "grad_norm": 1.5018306970596313,
+      "learning_rate": 5.16605820910658e-07,
+      "loss": 0.8831,
+      "step": 21145
+    },
+    {
+      "epoch": 0.5093807952602297,
+      "grad_norm": 1.8322926759719849,
+      "learning_rate": 5.164790589189737e-07,
+      "loss": 0.8392,
+      "step": 21150
+    },
+    {
+      "epoch": 0.509501216252017,
+      "grad_norm": 1.645729899406433,
+      "learning_rate": 5.163522969272893e-07,
+      "loss": 0.901,
+      "step": 21155
+    },
+    {
+      "epoch": 0.5096216372438044,
+      "grad_norm": 1.7000799179077148,
+      "learning_rate": 5.162255349356049e-07,
+      "loss": 0.8458,
+      "step": 21160
+    },
+    {
+      "epoch": 0.5097420582355916,
+      "grad_norm": 1.3836746215820312,
+      "learning_rate": 5.160987729439205e-07,
+      "loss": 0.9419,
+      "step": 21165
+    },
+    {
+      "epoch": 0.5098624792273789,
+      "grad_norm": 1.404828429222107,
+      "learning_rate": 5.15972010952236e-07,
+      "loss": 0.9054,
+      "step": 21170
+    },
+    {
+      "epoch": 0.5099829002191663,
+      "grad_norm": 1.502503514289856,
+      "learning_rate": 5.158452489605516e-07,
+      "loss": 0.9748,
+      "step": 21175
+    },
+    {
+      "epoch": 0.5101033212109535,
+      "grad_norm": 1.533842921257019,
+      "learning_rate": 5.157184869688672e-07,
+      "loss": 0.9296,
+      "step": 21180
+    },
+    {
+      "epoch": 0.5102237422027408,
+      "grad_norm": 1.3932472467422485,
+      "learning_rate": 5.155917249771829e-07,
+      "loss": 0.9343,
+      "step": 21185
+    },
+    {
+      "epoch": 0.510344163194528,
+      "grad_norm": 1.562790870666504,
+      "learning_rate": 5.154649629854983e-07,
+      "loss": 0.9493,
+      "step": 21190
+    },
+    {
+      "epoch": 0.5104645841863154,
+      "grad_norm": 1.5711618661880493,
+      "learning_rate": 5.15338200993814e-07,
+      "loss": 0.9587,
+      "step": 21195
+    },
+    {
+      "epoch": 0.5105850051781027,
+      "grad_norm": 1.5852607488632202,
+      "learning_rate": 5.152114390021296e-07,
+      "loss": 0.9167,
+      "step": 21200
+    },
+    {
+      "epoch": 0.5107054261698899,
+      "grad_norm": 1.6820813417434692,
+      "learning_rate": 5.150846770104451e-07,
+      "loss": 0.8955,
+      "step": 21205
+    },
+    {
+      "epoch": 0.5108258471616772,
+      "grad_norm": 1.5433590412139893,
+      "learning_rate": 5.149579150187608e-07,
+      "loss": 0.9351,
+      "step": 21210
+    },
+    {
+      "epoch": 0.5109462681534646,
+      "grad_norm": 1.5106788873672485,
+      "learning_rate": 5.148311530270763e-07,
+      "loss": 0.8874,
+      "step": 21215
+    },
+    {
+      "epoch": 0.5110666891452518,
+      "grad_norm": 1.558750867843628,
+      "learning_rate": 5.147043910353919e-07,
+      "loss": 0.9364,
+      "step": 21220
+    },
+    {
+      "epoch": 0.5111871101370391,
+      "grad_norm": 1.4906758069992065,
+      "learning_rate": 5.145776290437075e-07,
+      "loss": 0.868,
+      "step": 21225
+    },
+    {
+      "epoch": 0.5113075311288264,
+      "grad_norm": 1.4711549282073975,
+      "learning_rate": 5.144508670520231e-07,
+      "loss": 0.9563,
+      "step": 21230
+    },
+    {
+      "epoch": 0.5114279521206136,
+      "grad_norm": 1.7329277992248535,
+      "learning_rate": 5.143241050603386e-07,
+      "loss": 0.9388,
+      "step": 21235
+    },
+    {
+      "epoch": 0.511548373112401,
+      "grad_norm": 1.5840879678726196,
+      "learning_rate": 5.141973430686542e-07,
+      "loss": 0.924,
+      "step": 21240
+    },
+    {
+      "epoch": 0.5116687941041882,
+      "grad_norm": 1.4744466543197632,
+      "learning_rate": 5.140705810769699e-07,
+      "loss": 0.9193,
+      "step": 21245
+    },
+    {
+      "epoch": 0.5117892150959755,
+      "grad_norm": 1.7825807332992554,
+      "learning_rate": 5.139438190852854e-07,
+      "loss": 0.9507,
+      "step": 21250
+    },
+    {
+      "epoch": 0.5119096360877629,
+      "grad_norm": 1.3288898468017578,
+      "learning_rate": 5.138170570936011e-07,
+      "loss": 0.9705,
+      "step": 21255
+    },
+    {
+      "epoch": 0.5120300570795501,
+      "grad_norm": 1.7006186246871948,
+      "learning_rate": 5.136902951019166e-07,
+      "loss": 0.9218,
+      "step": 21260
+    },
+    {
+      "epoch": 0.5121504780713374,
+      "grad_norm": 1.5173015594482422,
+      "learning_rate": 5.135635331102321e-07,
+      "loss": 0.9247,
+      "step": 21265
+    },
+    {
+      "epoch": 0.5122708990631247,
+      "grad_norm": 1.546424150466919,
+      "learning_rate": 5.134367711185478e-07,
+      "loss": 0.9388,
+      "step": 21270
+    },
+    {
+      "epoch": 0.512391320054912,
+      "grad_norm": 1.7654190063476562,
+      "learning_rate": 5.133100091268634e-07,
+      "loss": 0.941,
+      "step": 21275
+    },
+    {
+      "epoch": 0.5125117410466993,
+      "grad_norm": 1.3392126560211182,
+      "learning_rate": 5.13183247135179e-07,
+      "loss": 0.8702,
+      "step": 21280
+    },
+    {
+      "epoch": 0.5126321620384865,
+      "grad_norm": 2.080995798110962,
+      "learning_rate": 5.130564851434945e-07,
+      "loss": 0.9056,
+      "step": 21285
+    },
+    {
+      "epoch": 0.5127525830302738,
+      "grad_norm": 1.460829257965088,
+      "learning_rate": 5.129297231518101e-07,
+      "loss": 0.8453,
+      "step": 21290
+    },
+    {
+      "epoch": 0.5128730040220612,
+      "grad_norm": 1.692646861076355,
+      "learning_rate": 5.128029611601258e-07,
+      "loss": 0.9056,
+      "step": 21295
+    },
+    {
+      "epoch": 0.5129934250138484,
+      "grad_norm": 1.5984737873077393,
+      "learning_rate": 5.126761991684413e-07,
+      "loss": 0.8921,
+      "step": 21300
+    },
+    {
+      "epoch": 0.5131138460056357,
+      "grad_norm": 1.5816010236740112,
+      "learning_rate": 5.125494371767569e-07,
+      "loss": 0.9293,
+      "step": 21305
+    },
+    {
+      "epoch": 0.513234266997423,
+      "grad_norm": 1.5962992906570435,
+      "learning_rate": 5.124226751850725e-07,
+      "loss": 0.8892,
+      "step": 21310
+    },
+    {
+      "epoch": 0.5133546879892102,
+      "grad_norm": 1.524151086807251,
+      "learning_rate": 5.12295913193388e-07,
+      "loss": 0.89,
+      "step": 21315
+    },
+    {
+      "epoch": 0.5134751089809976,
+      "grad_norm": 1.9870415925979614,
+      "learning_rate": 5.121691512017037e-07,
+      "loss": 0.9538,
+      "step": 21320
+    },
+    {
+      "epoch": 0.5135955299727849,
+      "grad_norm": 1.774303674697876,
+      "learning_rate": 5.120423892100193e-07,
+      "loss": 0.8695,
+      "step": 21325
+    },
+    {
+      "epoch": 0.5137159509645721,
+      "grad_norm": 1.5197960138320923,
+      "learning_rate": 5.119156272183348e-07,
+      "loss": 0.859,
+      "step": 21330
+    },
+    {
+      "epoch": 0.5138363719563595,
+      "grad_norm": 1.4730110168457031,
+      "learning_rate": 5.117888652266504e-07,
+      "loss": 0.8801,
+      "step": 21335
+    },
+    {
+      "epoch": 0.5139567929481467,
+      "grad_norm": 1.5757942199707031,
+      "learning_rate": 5.116621032349661e-07,
+      "loss": 0.8986,
+      "step": 21340
+    },
+    {
+      "epoch": 0.514077213939934,
+      "grad_norm": 1.787201166152954,
+      "learning_rate": 5.115353412432816e-07,
+      "loss": 0.9594,
+      "step": 21345
+    },
+    {
+      "epoch": 0.5141976349317213,
+      "grad_norm": 1.6020995378494263,
+      "learning_rate": 5.114085792515972e-07,
+      "loss": 0.8921,
+      "step": 21350
+    },
+    {
+      "epoch": 0.5143180559235085,
+      "grad_norm": 1.7841721773147583,
+      "learning_rate": 5.112818172599128e-07,
+      "loss": 0.8728,
+      "step": 21355
+    },
+    {
+      "epoch": 0.5144384769152959,
+      "grad_norm": 1.4041569232940674,
+      "learning_rate": 5.111550552682283e-07,
+      "loss": 0.86,
+      "step": 21360
+    },
+    {
+      "epoch": 0.5145588979070832,
+      "grad_norm": 1.733370065689087,
+      "learning_rate": 5.11028293276544e-07,
+      "loss": 0.9056,
+      "step": 21365
+    },
+    {
+      "epoch": 0.5146793188988704,
+      "grad_norm": 1.4316951036453247,
+      "learning_rate": 5.109015312848596e-07,
+      "loss": 0.9368,
+      "step": 21370
+    },
+    {
+      "epoch": 0.5147997398906577,
+      "grad_norm": 1.5830496549606323,
+      "learning_rate": 5.10774769293175e-07,
+      "loss": 0.9564,
+      "step": 21375
+    },
+    {
+      "epoch": 0.5149201608824451,
+      "grad_norm": 1.5182487964630127,
+      "learning_rate": 5.106480073014907e-07,
+      "loss": 0.9204,
+      "step": 21380
+    },
+    {
+      "epoch": 0.5150405818742323,
+      "grad_norm": 1.5096020698547363,
+      "learning_rate": 5.105212453098063e-07,
+      "loss": 0.8937,
+      "step": 21385
+    },
+    {
+      "epoch": 0.5151610028660196,
+      "grad_norm": 1.4812935590744019,
+      "learning_rate": 5.103944833181219e-07,
+      "loss": 0.9174,
+      "step": 21390
+    },
+    {
+      "epoch": 0.5152814238578068,
+      "grad_norm": 1.6584185361862183,
+      "learning_rate": 5.102677213264375e-07,
+      "loss": 0.9187,
+      "step": 21395
+    },
+    {
+      "epoch": 0.5154018448495942,
+      "grad_norm": 1.5996214151382446,
+      "learning_rate": 5.101409593347531e-07,
+      "loss": 0.9131,
+      "step": 21400
+    },
+    {
+      "epoch": 0.5155222658413815,
+      "grad_norm": 1.6251389980316162,
+      "learning_rate": 5.100141973430686e-07,
+      "loss": 0.8343,
+      "step": 21405
+    },
+    {
+      "epoch": 0.5156426868331687,
+      "grad_norm": 1.4660563468933105,
+      "learning_rate": 5.098874353513842e-07,
+      "loss": 0.927,
+      "step": 21410
+    },
+    {
+      "epoch": 0.515763107824956,
+      "grad_norm": 1.6889156103134155,
+      "learning_rate": 5.097606733596999e-07,
+      "loss": 0.8799,
+      "step": 21415
+    },
+    {
+      "epoch": 0.5158835288167434,
+      "grad_norm": 1.6388341188430786,
+      "learning_rate": 5.096339113680153e-07,
+      "loss": 0.9416,
+      "step": 21420
+    },
+    {
+      "epoch": 0.5160039498085306,
+      "grad_norm": 1.6233501434326172,
+      "learning_rate": 5.09507149376331e-07,
+      "loss": 0.8411,
+      "step": 21425
+    },
+    {
+      "epoch": 0.5161243708003179,
+      "grad_norm": 1.8281123638153076,
+      "learning_rate": 5.093803873846466e-07,
+      "loss": 0.9372,
+      "step": 21430
+    },
+    {
+      "epoch": 0.5162447917921053,
+      "grad_norm": 1.526695728302002,
+      "learning_rate": 5.092536253929621e-07,
+      "loss": 0.9745,
+      "step": 21435
+    },
+    {
+      "epoch": 0.5163652127838925,
+      "grad_norm": 1.6644231081008911,
+      "learning_rate": 5.091268634012778e-07,
+      "loss": 0.9452,
+      "step": 21440
+    },
+    {
+      "epoch": 0.5164856337756798,
+      "grad_norm": 1.6076133251190186,
+      "learning_rate": 5.090001014095933e-07,
+      "loss": 0.8995,
+      "step": 21445
+    },
+    {
+      "epoch": 0.516606054767467,
+      "grad_norm": 1.7203131914138794,
+      "learning_rate": 5.088733394179089e-07,
+      "loss": 0.9067,
+      "step": 21450
+    },
+    {
+      "epoch": 0.5167264757592543,
+      "grad_norm": 1.5742437839508057,
+      "learning_rate": 5.087465774262245e-07,
+      "loss": 0.87,
+      "step": 21455
+    },
+    {
+      "epoch": 0.5168468967510417,
+      "grad_norm": 1.6953926086425781,
+      "learning_rate": 5.086198154345402e-07,
+      "loss": 0.8837,
+      "step": 21460
+    },
+    {
+      "epoch": 0.5169673177428289,
+      "grad_norm": 1.421783208847046,
+      "learning_rate": 5.084930534428557e-07,
+      "loss": 0.8819,
+      "step": 21465
+    },
+    {
+      "epoch": 0.5170877387346162,
+      "grad_norm": 1.4581353664398193,
+      "learning_rate": 5.083662914511712e-07,
+      "loss": 0.938,
+      "step": 21470
+    },
+    {
+      "epoch": 0.5172081597264035,
+      "grad_norm": 1.642289638519287,
+      "learning_rate": 5.082395294594869e-07,
+      "loss": 0.9865,
+      "step": 21475
+    },
+    {
+      "epoch": 0.5173285807181908,
+      "grad_norm": 1.6251811981201172,
+      "learning_rate": 5.081127674678024e-07,
+      "loss": 0.924,
+      "step": 21480
+    },
+    {
+      "epoch": 0.5174490017099781,
+      "grad_norm": 1.596927285194397,
+      "learning_rate": 5.079860054761181e-07,
+      "loss": 0.9044,
+      "step": 21485
+    },
+    {
+      "epoch": 0.5175694227017654,
+      "grad_norm": 1.4459487199783325,
+      "learning_rate": 5.078592434844336e-07,
+      "loss": 0.9021,
+      "step": 21490
+    },
+    {
+      "epoch": 0.5176898436935526,
+      "grad_norm": 1.5127309560775757,
+      "learning_rate": 5.077324814927491e-07,
+      "loss": 0.9085,
+      "step": 21495
+    },
+    {
+      "epoch": 0.51781026468534,
+      "grad_norm": 1.7187678813934326,
+      "learning_rate": 5.076057195010648e-07,
+      "loss": 0.9297,
+      "step": 21500
+    },
+    {
+      "epoch": 0.5179306856771272,
+      "grad_norm": 1.4774119853973389,
+      "learning_rate": 5.074789575093804e-07,
+      "loss": 0.9093,
+      "step": 21505
+    },
+    {
+      "epoch": 0.5180511066689145,
+      "grad_norm": 1.88589608669281,
+      "learning_rate": 5.07352195517696e-07,
+      "loss": 0.8903,
+      "step": 21510
+    },
+    {
+      "epoch": 0.5181715276607018,
+      "grad_norm": 1.450732946395874,
+      "learning_rate": 5.072254335260115e-07,
+      "loss": 0.8562,
+      "step": 21515
+    },
+    {
+      "epoch": 0.5182919486524891,
+      "grad_norm": 1.2906776666641235,
+      "learning_rate": 5.070986715343272e-07,
+      "loss": 0.8162,
+      "step": 21520
+    },
+    {
+      "epoch": 0.5184123696442764,
+      "grad_norm": 1.682756781578064,
+      "learning_rate": 5.069719095426427e-07,
+      "loss": 0.9344,
+      "step": 21525
+    },
+    {
+      "epoch": 0.5185327906360637,
+      "grad_norm": 1.6122827529907227,
+      "learning_rate": 5.068451475509583e-07,
+      "loss": 0.9121,
+      "step": 21530
+    },
+    {
+      "epoch": 0.5186532116278509,
+      "grad_norm": 1.6736096143722534,
+      "learning_rate": 5.06718385559274e-07,
+      "loss": 0.9253,
+      "step": 21535
+    },
+    {
+      "epoch": 0.5187736326196383,
+      "grad_norm": 1.7354846000671387,
+      "learning_rate": 5.065916235675894e-07,
+      "loss": 0.9529,
+      "step": 21540
+    },
+    {
+      "epoch": 0.5188940536114256,
+      "grad_norm": 1.8778737783432007,
+      "learning_rate": 5.064648615759051e-07,
+      "loss": 0.859,
+      "step": 21545
+    },
+    {
+      "epoch": 0.5190144746032128,
+      "grad_norm": 1.4433398246765137,
+      "learning_rate": 5.063380995842207e-07,
+      "loss": 0.8193,
+      "step": 21550
+    },
+    {
+      "epoch": 0.5191348955950001,
+      "grad_norm": 1.7733087539672852,
+      "learning_rate": 5.062113375925362e-07,
+      "loss": 0.8426,
+      "step": 21555
+    },
+    {
+      "epoch": 0.5192553165867874,
+      "grad_norm": 1.736220359802246,
+      "learning_rate": 5.060845756008518e-07,
+      "loss": 0.9707,
+      "step": 21560
+    },
+    {
+      "epoch": 0.5193757375785747,
+      "grad_norm": 1.5831776857376099,
+      "learning_rate": 5.059578136091674e-07,
+      "loss": 0.956,
+      "step": 21565
+    },
+    {
+      "epoch": 0.519496158570362,
+      "grad_norm": 1.6207410097122192,
+      "learning_rate": 5.05831051617483e-07,
+      "loss": 0.87,
+      "step": 21570
+    },
+    {
+      "epoch": 0.5196165795621492,
+      "grad_norm": 1.3546632528305054,
+      "learning_rate": 5.057042896257986e-07,
+      "loss": 0.9305,
+      "step": 21575
+    },
+    {
+      "epoch": 0.5197370005539366,
+      "grad_norm": 1.4451007843017578,
+      "learning_rate": 5.055775276341143e-07,
+      "loss": 0.9327,
+      "step": 21580
+    },
+    {
+      "epoch": 0.5198574215457239,
+      "grad_norm": 1.677097201347351,
+      "learning_rate": 5.054507656424297e-07,
+      "loss": 0.8981,
+      "step": 21585
+    },
+    {
+      "epoch": 0.5199778425375111,
+      "grad_norm": 1.4165575504302979,
+      "learning_rate": 5.053240036507453e-07,
+      "loss": 0.9142,
+      "step": 21590
+    },
+    {
+      "epoch": 0.5200982635292984,
+      "grad_norm": 1.5407862663269043,
+      "learning_rate": 5.05197241659061e-07,
+      "loss": 0.8778,
+      "step": 21595
+    },
+    {
+      "epoch": 0.5202186845210857,
+      "grad_norm": 1.4036144018173218,
+      "learning_rate": 5.050704796673765e-07,
+      "loss": 0.9203,
+      "step": 21600
+    },
+    {
+      "epoch": 0.520339105512873,
+      "grad_norm": 1.6382545232772827,
+      "learning_rate": 5.049437176756921e-07,
+      "loss": 0.9628,
+      "step": 21605
+    },
+    {
+      "epoch": 0.5204595265046603,
+      "grad_norm": 1.9323283433914185,
+      "learning_rate": 5.048169556840077e-07,
+      "loss": 0.9123,
+      "step": 21610
+    },
+    {
+      "epoch": 0.5205799474964475,
+      "grad_norm": 1.5797611474990845,
+      "learning_rate": 5.046901936923232e-07,
+      "loss": 0.9022,
+      "step": 21615
+    },
+    {
+      "epoch": 0.5207003684882349,
+      "grad_norm": 1.5481973886489868,
+      "learning_rate": 5.045634317006389e-07,
+      "loss": 0.9458,
+      "step": 21620
+    },
+    {
+      "epoch": 0.5208207894800222,
+      "grad_norm": 1.4047622680664062,
+      "learning_rate": 5.044366697089545e-07,
+      "loss": 0.8682,
+      "step": 21625
+    },
+    {
+      "epoch": 0.5209412104718094,
+      "grad_norm": 1.4203743934631348,
+      "learning_rate": 5.0430990771727e-07,
+      "loss": 0.8861,
+      "step": 21630
+    },
+    {
+      "epoch": 0.5210616314635967,
+      "grad_norm": 1.6246775388717651,
+      "learning_rate": 5.041831457255856e-07,
+      "loss": 0.8425,
+      "step": 21635
+    },
+    {
+      "epoch": 0.5211820524553841,
+      "grad_norm": 1.6451424360275269,
+      "learning_rate": 5.040563837339013e-07,
+      "loss": 0.9304,
+      "step": 21640
+    },
+    {
+      "epoch": 0.5213024734471713,
+      "grad_norm": 1.7852120399475098,
+      "learning_rate": 5.039296217422168e-07,
+      "loss": 0.8965,
+      "step": 21645
+    },
+    {
+      "epoch": 0.5214228944389586,
+      "grad_norm": 1.5709748268127441,
+      "learning_rate": 5.038028597505324e-07,
+      "loss": 0.9198,
+      "step": 21650
+    },
+    {
+      "epoch": 0.5215433154307458,
+      "grad_norm": 2.4693374633789062,
+      "learning_rate": 5.03676097758848e-07,
+      "loss": 0.8776,
+      "step": 21655
+    },
+    {
+      "epoch": 0.5216637364225332,
+      "grad_norm": 1.5472397804260254,
+      "learning_rate": 5.035493357671635e-07,
+      "loss": 0.8557,
+      "step": 21660
+    },
+    {
+      "epoch": 0.5217841574143205,
+      "grad_norm": 1.5521728992462158,
+      "learning_rate": 5.034225737754792e-07,
+      "loss": 0.9405,
+      "step": 21665
+    },
+    {
+      "epoch": 0.5219045784061077,
+      "grad_norm": 1.5593839883804321,
+      "learning_rate": 5.032958117837948e-07,
+      "loss": 0.8569,
+      "step": 21670
+    },
+    {
+      "epoch": 0.522024999397895,
+      "grad_norm": 1.4785000085830688,
+      "learning_rate": 5.031690497921102e-07,
+      "loss": 0.8832,
+      "step": 21675
+    },
+    {
+      "epoch": 0.5221454203896824,
+      "grad_norm": 1.4843261241912842,
+      "learning_rate": 5.030422878004259e-07,
+      "loss": 0.8979,
+      "step": 21680
+    },
+    {
+      "epoch": 0.5222658413814696,
+      "grad_norm": 1.536954641342163,
+      "learning_rate": 5.029155258087415e-07,
+      "loss": 0.9145,
+      "step": 21685
+    },
+    {
+      "epoch": 0.5223862623732569,
+      "grad_norm": 1.4595369100570679,
+      "learning_rate": 5.027887638170571e-07,
+      "loss": 0.8905,
+      "step": 21690
+    },
+    {
+      "epoch": 0.5225066833650442,
+      "grad_norm": 1.3696085214614868,
+      "learning_rate": 5.026620018253727e-07,
+      "loss": 0.9039,
+      "step": 21695
+    },
+    {
+      "epoch": 0.5226271043568315,
+      "grad_norm": 1.6402760744094849,
+      "learning_rate": 5.025352398336882e-07,
+      "loss": 0.8909,
+      "step": 21700
+    },
+    {
+      "epoch": 0.5227475253486188,
+      "grad_norm": 1.538196325302124,
+      "learning_rate": 5.024084778420038e-07,
+      "loss": 0.9148,
+      "step": 21705
+    },
+    {
+      "epoch": 0.522867946340406,
+      "grad_norm": 1.3932260274887085,
+      "learning_rate": 5.022817158503194e-07,
+      "loss": 0.8892,
+      "step": 21710
+    },
+    {
+      "epoch": 0.5229883673321933,
+      "grad_norm": 1.6184591054916382,
+      "learning_rate": 5.021549538586351e-07,
+      "loss": 0.8681,
+      "step": 21715
+    },
+    {
+      "epoch": 0.5231087883239807,
+      "grad_norm": 1.3606884479522705,
+      "learning_rate": 5.020281918669505e-07,
+      "loss": 0.8935,
+      "step": 21720
+    },
+    {
+      "epoch": 0.5232292093157679,
+      "grad_norm": 1.7561999559402466,
+      "learning_rate": 5.019014298752662e-07,
+      "loss": 0.8992,
+      "step": 21725
+    },
+    {
+      "epoch": 0.5233496303075552,
+      "grad_norm": 1.3053646087646484,
+      "learning_rate": 5.017746678835818e-07,
+      "loss": 0.9059,
+      "step": 21730
+    },
+    {
+      "epoch": 0.5234700512993425,
+      "grad_norm": 1.863433599472046,
+      "learning_rate": 5.016479058918973e-07,
+      "loss": 0.9344,
+      "step": 21735
+    },
+    {
+      "epoch": 0.5235904722911298,
+      "grad_norm": 1.3806565999984741,
+      "learning_rate": 5.01521143900213e-07,
+      "loss": 0.9454,
+      "step": 21740
+    },
+    {
+      "epoch": 0.5237108932829171,
+      "grad_norm": 2.0067007541656494,
+      "learning_rate": 5.013943819085285e-07,
+      "loss": 0.8896,
+      "step": 21745
+    },
+    {
+      "epoch": 0.5238313142747044,
+      "grad_norm": 1.4963665008544922,
+      "learning_rate": 5.012676199168441e-07,
+      "loss": 0.8821,
+      "step": 21750
+    },
+    {
+      "epoch": 0.5239517352664916,
+      "grad_norm": 1.5632950067520142,
+      "learning_rate": 5.011408579251597e-07,
+      "loss": 0.9355,
+      "step": 21755
+    },
+    {
+      "epoch": 0.524072156258279,
+      "grad_norm": 1.5998642444610596,
+      "learning_rate": 5.010140959334753e-07,
+      "loss": 0.9616,
+      "step": 21760
+    },
+    {
+      "epoch": 0.5241925772500662,
+      "grad_norm": 1.5908117294311523,
+      "learning_rate": 5.008873339417909e-07,
+      "loss": 0.8835,
+      "step": 21765
+    },
+    {
+      "epoch": 0.5243129982418535,
+      "grad_norm": 1.4726042747497559,
+      "learning_rate": 5.007605719501064e-07,
+      "loss": 0.8774,
+      "step": 21770
+    },
+    {
+      "epoch": 0.5244334192336408,
+      "grad_norm": 1.5189387798309326,
+      "learning_rate": 5.006338099584221e-07,
+      "loss": 0.9001,
+      "step": 21775
+    },
+    {
+      "epoch": 0.5245538402254281,
+      "grad_norm": 1.428139090538025,
+      "learning_rate": 5.005070479667376e-07,
+      "loss": 0.9265,
+      "step": 21780
+    },
+    {
+      "epoch": 0.5246742612172154,
+      "grad_norm": 1.4064525365829468,
+      "learning_rate": 5.003802859750533e-07,
+      "loss": 0.8685,
+      "step": 21785
+    },
+    {
+      "epoch": 0.5247946822090027,
+      "grad_norm": 1.557692289352417,
+      "learning_rate": 5.002535239833688e-07,
+      "loss": 0.9168,
+      "step": 21790
+    },
+    {
+      "epoch": 0.5249151032007899,
+      "grad_norm": 1.5677871704101562,
+      "learning_rate": 5.001267619916843e-07,
+      "loss": 0.9341,
+      "step": 21795
+    },
+    {
+      "epoch": 0.5250355241925773,
+      "grad_norm": 1.5455026626586914,
+      "learning_rate": 5e-07,
+      "loss": 0.8812,
+      "step": 21800
+    },
+    {
+      "epoch": 0.5251559451843646,
+      "grad_norm": 1.6961249113082886,
+      "learning_rate": 4.998732380083155e-07,
+      "loss": 0.9275,
+      "step": 21805
+    },
+    {
+      "epoch": 0.5252763661761518,
+      "grad_norm": 1.4484466314315796,
+      "learning_rate": 4.997464760166312e-07,
+      "loss": 0.8577,
+      "step": 21810
+    },
+    {
+      "epoch": 0.5253967871679391,
+      "grad_norm": 1.5019750595092773,
+      "learning_rate": 4.996197140249467e-07,
+      "loss": 0.9201,
+      "step": 21815
+    },
+    {
+      "epoch": 0.5255172081597264,
+      "grad_norm": 1.5143229961395264,
+      "learning_rate": 4.994929520332623e-07,
+      "loss": 0.9347,
+      "step": 21820
+    },
+    {
+      "epoch": 0.5256376291515137,
+      "grad_norm": 1.7485090494155884,
+      "learning_rate": 4.993661900415779e-07,
+      "loss": 0.9222,
+      "step": 21825
+    },
+    {
+      "epoch": 0.525758050143301,
+      "grad_norm": 1.5002058744430542,
+      "learning_rate": 4.992394280498935e-07,
+      "loss": 0.882,
+      "step": 21830
+    },
+    {
+      "epoch": 0.5258784711350882,
+      "grad_norm": 2.1091864109039307,
+      "learning_rate": 4.991126660582091e-07,
+      "loss": 0.9634,
+      "step": 21835
+    },
+    {
+      "epoch": 0.5259988921268756,
+      "grad_norm": 1.8817534446716309,
+      "learning_rate": 4.989859040665247e-07,
+      "loss": 0.9068,
+      "step": 21840
+    },
+    {
+      "epoch": 0.5261193131186629,
+      "grad_norm": 1.529942274093628,
+      "learning_rate": 4.988591420748403e-07,
+      "loss": 0.8842,
+      "step": 21845
+    },
+    {
+      "epoch": 0.5262397341104501,
+      "grad_norm": 1.7217942476272583,
+      "learning_rate": 4.987323800831558e-07,
+      "loss": 0.9049,
+      "step": 21850
+    },
+    {
+      "epoch": 0.5263601551022374,
+      "grad_norm": 1.5230528116226196,
+      "learning_rate": 4.986056180914714e-07,
+      "loss": 0.9461,
+      "step": 21855
+    },
+    {
+      "epoch": 0.5264805760940248,
+      "grad_norm": 1.4534242153167725,
+      "learning_rate": 4.98478856099787e-07,
+      "loss": 0.8935,
+      "step": 21860
+    },
+    {
+      "epoch": 0.526600997085812,
+      "grad_norm": 1.3972665071487427,
+      "learning_rate": 4.983520941081026e-07,
+      "loss": 0.8875,
+      "step": 21865
+    },
+    {
+      "epoch": 0.5267214180775993,
+      "grad_norm": 1.5731639862060547,
+      "learning_rate": 4.982253321164182e-07,
+      "loss": 0.8769,
+      "step": 21870
+    },
+    {
+      "epoch": 0.5268418390693865,
+      "grad_norm": 1.5789008140563965,
+      "learning_rate": 4.980985701247338e-07,
+      "loss": 0.9617,
+      "step": 21875
+    },
+    {
+      "epoch": 0.5269622600611739,
+      "grad_norm": 1.4749755859375,
+      "learning_rate": 4.979718081330493e-07,
+      "loss": 0.8825,
+      "step": 21880
+    },
+    {
+      "epoch": 0.5270826810529612,
+      "grad_norm": 2.0449745655059814,
+      "learning_rate": 4.97845046141365e-07,
+      "loss": 0.9582,
+      "step": 21885
+    },
+    {
+      "epoch": 0.5272031020447484,
+      "grad_norm": 1.6935410499572754,
+      "learning_rate": 4.977182841496805e-07,
+      "loss": 0.9129,
+      "step": 21890
+    },
+    {
+      "epoch": 0.5273235230365357,
+      "grad_norm": 1.5887279510498047,
+      "learning_rate": 4.975915221579961e-07,
+      "loss": 0.8921,
+      "step": 21895
+    },
+    {
+      "epoch": 0.5274439440283231,
+      "grad_norm": 1.8212227821350098,
+      "learning_rate": 4.974647601663117e-07,
+      "loss": 0.9361,
+      "step": 21900
+    },
+    {
+      "epoch": 0.5275643650201103,
+      "grad_norm": 1.4673494100570679,
+      "learning_rate": 4.973379981746272e-07,
+      "loss": 0.919,
+      "step": 21905
+    },
+    {
+      "epoch": 0.5276847860118976,
+      "grad_norm": 1.492056131362915,
+      "learning_rate": 4.972112361829429e-07,
+      "loss": 0.9292,
+      "step": 21910
+    },
+    {
+      "epoch": 0.5278052070036848,
+      "grad_norm": 1.6951390504837036,
+      "learning_rate": 4.970844741912585e-07,
+      "loss": 0.9186,
+      "step": 21915
+    },
+    {
+      "epoch": 0.5279256279954722,
+      "grad_norm": 1.5264896154403687,
+      "learning_rate": 4.969577121995741e-07,
+      "loss": 0.8546,
+      "step": 21920
+    },
+    {
+      "epoch": 0.5280460489872595,
+      "grad_norm": 1.3826159238815308,
+      "learning_rate": 4.968309502078896e-07,
+      "loss": 0.9198,
+      "step": 21925
+    },
+    {
+      "epoch": 0.5281664699790467,
+      "grad_norm": 1.53154718875885,
+      "learning_rate": 4.967041882162053e-07,
+      "loss": 0.8998,
+      "step": 21930
+    },
+    {
+      "epoch": 0.528286890970834,
+      "grad_norm": 1.5544826984405518,
+      "learning_rate": 4.965774262245208e-07,
+      "loss": 0.8952,
+      "step": 21935
+    },
+    {
+      "epoch": 0.5284073119626214,
+      "grad_norm": 1.7008835077285767,
+      "learning_rate": 4.964506642328363e-07,
+      "loss": 0.9436,
+      "step": 21940
+    },
+    {
+      "epoch": 0.5285277329544086,
+      "grad_norm": 1.7531218528747559,
+      "learning_rate": 4.96323902241152e-07,
+      "loss": 0.9292,
+      "step": 21945
+    },
+    {
+      "epoch": 0.5286481539461959,
+      "grad_norm": 1.6096489429473877,
+      "learning_rate": 4.961971402494675e-07,
+      "loss": 0.9339,
+      "step": 21950
+    },
+    {
+      "epoch": 0.5287685749379832,
+      "grad_norm": 1.3603724241256714,
+      "learning_rate": 4.960703782577832e-07,
+      "loss": 0.8547,
+      "step": 21955
+    },
+    {
+      "epoch": 0.5288889959297705,
+      "grad_norm": 1.8208463191986084,
+      "learning_rate": 4.959436162660988e-07,
+      "loss": 0.9059,
+      "step": 21960
+    },
+    {
+      "epoch": 0.5290094169215578,
+      "grad_norm": 1.6965110301971436,
+      "learning_rate": 4.958168542744143e-07,
+      "loss": 0.8995,
+      "step": 21965
+    },
+    {
+      "epoch": 0.529129837913345,
+      "grad_norm": 1.5166840553283691,
+      "learning_rate": 4.956900922827299e-07,
+      "loss": 0.8476,
+      "step": 21970
+    },
+    {
+      "epoch": 0.5292502589051323,
+      "grad_norm": 1.7423069477081299,
+      "learning_rate": 4.955633302910455e-07,
+      "loss": 0.9335,
+      "step": 21975
+    },
+    {
+      "epoch": 0.5293706798969197,
+      "grad_norm": 1.5373573303222656,
+      "learning_rate": 4.954365682993611e-07,
+      "loss": 0.9236,
+      "step": 21980
+    },
+    {
+      "epoch": 0.5294911008887069,
+      "grad_norm": 1.5119625329971313,
+      "learning_rate": 4.953098063076766e-07,
+      "loss": 0.8502,
+      "step": 21985
+    },
+    {
+      "epoch": 0.5296115218804942,
+      "grad_norm": 1.5202735662460327,
+      "learning_rate": 4.951830443159923e-07,
+      "loss": 0.8701,
+      "step": 21990
+    },
+    {
+      "epoch": 0.5297319428722815,
+      "grad_norm": 1.4449344873428345,
+      "learning_rate": 4.950562823243079e-07,
+      "loss": 0.9292,
+      "step": 21995
+    },
+    {
+      "epoch": 0.5298523638640688,
+      "grad_norm": 1.5788203477859497,
+      "learning_rate": 4.949295203326234e-07,
+      "loss": 0.889,
+      "step": 22000
+    },
+    {
+      "epoch": 0.5299727848558561,
+      "grad_norm": 1.403855562210083,
+      "learning_rate": 4.948027583409391e-07,
+      "loss": 0.8862,
+      "step": 22005
+    },
+    {
+      "epoch": 0.5300932058476434,
+      "grad_norm": 1.394984245300293,
+      "learning_rate": 4.946759963492546e-07,
+      "loss": 0.9096,
+      "step": 22010
+    },
+    {
+      "epoch": 0.5302136268394306,
+      "grad_norm": 1.433779239654541,
+      "learning_rate": 4.945492343575702e-07,
+      "loss": 0.8892,
+      "step": 22015
+    },
+    {
+      "epoch": 0.530334047831218,
+      "grad_norm": 1.5095500946044922,
+      "learning_rate": 4.944224723658858e-07,
+      "loss": 0.9021,
+      "step": 22020
+    },
+    {
+      "epoch": 0.5304544688230052,
+      "grad_norm": 1.5243232250213623,
+      "learning_rate": 4.942957103742013e-07,
+      "loss": 0.9025,
+      "step": 22025
+    },
+    {
+      "epoch": 0.5305748898147925,
+      "grad_norm": 1.395520806312561,
+      "learning_rate": 4.94168948382517e-07,
+      "loss": 0.914,
+      "step": 22030
+    },
+    {
+      "epoch": 0.5306953108065798,
+      "grad_norm": 1.6437811851501465,
+      "learning_rate": 4.940421863908325e-07,
+      "loss": 0.9334,
+      "step": 22035
+    },
+    {
+      "epoch": 0.530815731798367,
+      "grad_norm": 1.5893141031265259,
+      "learning_rate": 4.939154243991482e-07,
+      "loss": 0.8897,
+      "step": 22040
+    },
+    {
+      "epoch": 0.5309361527901544,
+      "grad_norm": 1.482459545135498,
+      "learning_rate": 4.937886624074637e-07,
+      "loss": 0.866,
+      "step": 22045
+    },
+    {
+      "epoch": 0.5310565737819417,
+      "grad_norm": 1.573954701423645,
+      "learning_rate": 4.936619004157794e-07,
+      "loss": 0.8911,
+      "step": 22050
+    },
+    {
+      "epoch": 0.5311769947737289,
+      "grad_norm": 1.6320579051971436,
+      "learning_rate": 4.935351384240949e-07,
+      "loss": 0.9018,
+      "step": 22055
+    },
+    {
+      "epoch": 0.5312974157655163,
+      "grad_norm": 1.6011643409729004,
+      "learning_rate": 4.934083764324105e-07,
+      "loss": 0.9071,
+      "step": 22060
+    },
+    {
+      "epoch": 0.5314178367573036,
+      "grad_norm": 2.10659122467041,
+      "learning_rate": 4.932816144407261e-07,
+      "loss": 0.9777,
+      "step": 22065
+    },
+    {
+      "epoch": 0.5315382577490908,
+      "grad_norm": 1.4913204908370972,
+      "learning_rate": 4.931548524490416e-07,
+      "loss": 0.9108,
+      "step": 22070
+    },
+    {
+      "epoch": 0.5316586787408781,
+      "grad_norm": 1.5963008403778076,
+      "learning_rate": 4.930280904573573e-07,
+      "loss": 0.8917,
+      "step": 22075
+    },
+    {
+      "epoch": 0.5317790997326654,
+      "grad_norm": 1.4318112134933472,
+      "learning_rate": 4.929013284656728e-07,
+      "loss": 0.8998,
+      "step": 22080
+    },
+    {
+      "epoch": 0.5318995207244527,
+      "grad_norm": 1.4206163883209229,
+      "learning_rate": 4.927745664739884e-07,
+      "loss": 0.917,
+      "step": 22085
+    },
+    {
+      "epoch": 0.53201994171624,
+      "grad_norm": 1.5320185422897339,
+      "learning_rate": 4.92647804482304e-07,
+      "loss": 0.9307,
+      "step": 22090
+    },
+    {
+      "epoch": 0.5321403627080272,
+      "grad_norm": 1.8943982124328613,
+      "learning_rate": 4.925210424906196e-07,
+      "loss": 0.8849,
+      "step": 22095
+    },
+    {
+      "epoch": 0.5322607836998146,
+      "grad_norm": 1.48078453540802,
+      "learning_rate": 4.923942804989352e-07,
+      "loss": 0.8307,
+      "step": 22100
+    },
+    {
+      "epoch": 0.5323812046916019,
+      "grad_norm": 1.756567120552063,
+      "learning_rate": 4.922675185072508e-07,
+      "loss": 0.8657,
+      "step": 22105
+    },
+    {
+      "epoch": 0.5325016256833891,
+      "grad_norm": 1.6934877634048462,
+      "learning_rate": 4.921407565155664e-07,
+      "loss": 0.9453,
+      "step": 22110
+    },
+    {
+      "epoch": 0.5326220466751764,
+      "grad_norm": 1.4306222200393677,
+      "learning_rate": 4.920139945238819e-07,
+      "loss": 0.9266,
+      "step": 22115
+    },
+    {
+      "epoch": 0.5327424676669638,
+      "grad_norm": 1.571732759475708,
+      "learning_rate": 4.918872325321975e-07,
+      "loss": 0.8887,
+      "step": 22120
+    },
+    {
+      "epoch": 0.532862888658751,
+      "grad_norm": 1.6166703701019287,
+      "learning_rate": 4.917604705405131e-07,
+      "loss": 0.9261,
+      "step": 22125
+    },
+    {
+      "epoch": 0.5329833096505383,
+      "grad_norm": 1.8382987976074219,
+      "learning_rate": 4.916337085488287e-07,
+      "loss": 0.8377,
+      "step": 22130
+    },
+    {
+      "epoch": 0.5331037306423255,
+      "grad_norm": 2.1283352375030518,
+      "learning_rate": 4.915069465571443e-07,
+      "loss": 0.9282,
+      "step": 22135
+    },
+    {
+      "epoch": 0.5332241516341129,
+      "grad_norm": 1.431143879890442,
+      "learning_rate": 4.913801845654599e-07,
+      "loss": 0.8822,
+      "step": 22140
+    },
+    {
+      "epoch": 0.5333445726259002,
+      "grad_norm": 1.5645533800125122,
+      "learning_rate": 4.912534225737754e-07,
+      "loss": 0.8755,
+      "step": 22145
+    },
+    {
+      "epoch": 0.5334649936176874,
+      "grad_norm": 1.641961693763733,
+      "learning_rate": 4.911266605820911e-07,
+      "loss": 0.9234,
+      "step": 22150
+    },
+    {
+      "epoch": 0.5335854146094747,
+      "grad_norm": 1.3797447681427002,
+      "learning_rate": 4.909998985904066e-07,
+      "loss": 0.964,
+      "step": 22155
+    },
+    {
+      "epoch": 0.5337058356012621,
+      "grad_norm": 1.3773528337478638,
+      "learning_rate": 4.908731365987222e-07,
+      "loss": 0.8885,
+      "step": 22160
+    },
+    {
+      "epoch": 0.5338262565930493,
+      "grad_norm": 1.5127993822097778,
+      "learning_rate": 4.907463746070378e-07,
+      "loss": 0.9015,
+      "step": 22165
+    },
+    {
+      "epoch": 0.5339466775848366,
+      "grad_norm": 1.6455553770065308,
+      "learning_rate": 4.906196126153533e-07,
+      "loss": 0.8727,
+      "step": 22170
+    },
+    {
+      "epoch": 0.5340670985766238,
+      "grad_norm": 1.4866843223571777,
+      "learning_rate": 4.90492850623669e-07,
+      "loss": 0.9068,
+      "step": 22175
+    },
+    {
+      "epoch": 0.5341875195684112,
+      "grad_norm": 1.6697514057159424,
+      "learning_rate": 4.903660886319846e-07,
+      "loss": 0.9244,
+      "step": 22180
+    },
+    {
+      "epoch": 0.5343079405601985,
+      "grad_norm": 1.9051423072814941,
+      "learning_rate": 4.902393266403002e-07,
+      "loss": 0.9537,
+      "step": 22185
+    },
+    {
+      "epoch": 0.5344283615519857,
+      "grad_norm": 1.4561595916748047,
+      "learning_rate": 4.901125646486157e-07,
+      "loss": 0.8634,
+      "step": 22190
+    },
+    {
+      "epoch": 0.534548782543773,
+      "grad_norm": 1.6183537244796753,
+      "learning_rate": 4.899858026569314e-07,
+      "loss": 0.8765,
+      "step": 22195
+    },
+    {
+      "epoch": 0.5346692035355604,
+      "grad_norm": 1.6155229806900024,
+      "learning_rate": 4.898590406652469e-07,
+      "loss": 0.9336,
+      "step": 22200
+    },
+    {
+      "epoch": 0.5347896245273476,
+      "grad_norm": 1.5258883237838745,
+      "learning_rate": 4.897322786735624e-07,
+      "loss": 0.9099,
+      "step": 22205
+    },
+    {
+      "epoch": 0.5349100455191349,
+      "grad_norm": 1.4723219871520996,
+      "learning_rate": 4.896055166818781e-07,
+      "loss": 0.8694,
+      "step": 22210
+    },
+    {
+      "epoch": 0.5350304665109222,
+      "grad_norm": 1.4346950054168701,
+      "learning_rate": 4.894787546901937e-07,
+      "loss": 0.914,
+      "step": 22215
+    },
+    {
+      "epoch": 0.5351508875027094,
+      "grad_norm": 1.5123918056488037,
+      "learning_rate": 4.893519926985093e-07,
+      "loss": 0.8903,
+      "step": 22220
+    },
+    {
+      "epoch": 0.5352713084944968,
+      "grad_norm": 1.8437974452972412,
+      "learning_rate": 4.892252307068249e-07,
+      "loss": 0.9109,
+      "step": 22225
+    },
+    {
+      "epoch": 0.535391729486284,
+      "grad_norm": 1.5314289331436157,
+      "learning_rate": 4.890984687151404e-07,
+      "loss": 0.8496,
+      "step": 22230
+    },
+    {
+      "epoch": 0.5355121504780713,
+      "grad_norm": 1.5800902843475342,
+      "learning_rate": 4.88971706723456e-07,
+      "loss": 0.9294,
+      "step": 22235
+    },
+    {
+      "epoch": 0.5356325714698587,
+      "grad_norm": 1.6109490394592285,
+      "learning_rate": 4.888449447317716e-07,
+      "loss": 0.8759,
+      "step": 22240
+    },
+    {
+      "epoch": 0.5357529924616459,
+      "grad_norm": 1.414548397064209,
+      "learning_rate": 4.887181827400872e-07,
+      "loss": 0.9122,
+      "step": 22245
+    },
+    {
+      "epoch": 0.5358734134534332,
+      "grad_norm": 1.4882311820983887,
+      "learning_rate": 4.885914207484027e-07,
+      "loss": 0.8801,
+      "step": 22250
+    },
+    {
+      "epoch": 0.5359938344452205,
+      "grad_norm": 1.5739588737487793,
+      "learning_rate": 4.884646587567184e-07,
+      "loss": 0.8628,
+      "step": 22255
+    },
+    {
+      "epoch": 0.5361142554370077,
+      "grad_norm": 1.5703134536743164,
+      "learning_rate": 4.88337896765034e-07,
+      "loss": 0.8883,
+      "step": 22260
+    },
+    {
+      "epoch": 0.5362346764287951,
+      "grad_norm": 1.537899374961853,
+      "learning_rate": 4.882111347733495e-07,
+      "loss": 0.938,
+      "step": 22265
+    },
+    {
+      "epoch": 0.5363550974205824,
+      "grad_norm": 1.603467345237732,
+      "learning_rate": 4.880843727816652e-07,
+      "loss": 0.907,
+      "step": 22270
+    },
+    {
+      "epoch": 0.5364755184123696,
+      "grad_norm": 1.553680419921875,
+      "learning_rate": 4.879576107899807e-07,
+      "loss": 0.9072,
+      "step": 22275
+    },
+    {
+      "epoch": 0.536595939404157,
+      "grad_norm": 1.4864248037338257,
+      "learning_rate": 4.878308487982963e-07,
+      "loss": 0.8833,
+      "step": 22280
+    },
+    {
+      "epoch": 0.5367163603959442,
+      "grad_norm": 1.5189486742019653,
+      "learning_rate": 4.877040868066119e-07,
+      "loss": 0.9643,
+      "step": 22285
+    },
+    {
+      "epoch": 0.5368367813877315,
+      "grad_norm": 1.6031887531280518,
+      "learning_rate": 4.875773248149274e-07,
+      "loss": 0.9135,
+      "step": 22290
+    },
+    {
+      "epoch": 0.5369572023795188,
+      "grad_norm": 1.531436562538147,
+      "learning_rate": 4.874505628232431e-07,
+      "loss": 0.9054,
+      "step": 22295
+    },
+    {
+      "epoch": 0.537077623371306,
+      "grad_norm": 1.7289807796478271,
+      "learning_rate": 4.873238008315586e-07,
+      "loss": 0.9326,
+      "step": 22300
+    },
+    {
+      "epoch": 0.5371980443630934,
+      "grad_norm": 1.4896320104599,
+      "learning_rate": 4.871970388398743e-07,
+      "loss": 0.8859,
+      "step": 22305
+    },
+    {
+      "epoch": 0.5373184653548807,
+      "grad_norm": 1.5204687118530273,
+      "learning_rate": 4.870702768481898e-07,
+      "loss": 0.8812,
+      "step": 22310
+    },
+    {
+      "epoch": 0.5374388863466679,
+      "grad_norm": 1.5427957773208618,
+      "learning_rate": 4.869435148565055e-07,
+      "loss": 0.9262,
+      "step": 22315
+    },
+    {
+      "epoch": 0.5375593073384553,
+      "grad_norm": 1.3800618648529053,
+      "learning_rate": 4.86816752864821e-07,
+      "loss": 0.8908,
+      "step": 22320
+    },
+    {
+      "epoch": 0.5376797283302426,
+      "grad_norm": 1.6934906244277954,
+      "learning_rate": 4.866899908731365e-07,
+      "loss": 0.9171,
+      "step": 22325
+    },
+    {
+      "epoch": 0.5378001493220298,
+      "grad_norm": 1.7136369943618774,
+      "learning_rate": 4.865632288814522e-07,
+      "loss": 0.9061,
+      "step": 22330
+    },
+    {
+      "epoch": 0.5379205703138171,
+      "grad_norm": 2.0289530754089355,
+      "learning_rate": 4.864364668897677e-07,
+      "loss": 0.8473,
+      "step": 22335
+    },
+    {
+      "epoch": 0.5380409913056043,
+      "grad_norm": 1.7243914604187012,
+      "learning_rate": 4.863097048980834e-07,
+      "loss": 0.9861,
+      "step": 22340
+    },
+    {
+      "epoch": 0.5381614122973917,
+      "grad_norm": 1.466810941696167,
+      "learning_rate": 4.861829429063989e-07,
+      "loss": 0.9319,
+      "step": 22345
+    },
+    {
+      "epoch": 0.538281833289179,
+      "grad_norm": 1.6713786125183105,
+      "learning_rate": 4.860561809147145e-07,
+      "loss": 0.9378,
+      "step": 22350
+    },
+    {
+      "epoch": 0.5384022542809662,
+      "grad_norm": 1.7098300457000732,
+      "learning_rate": 4.859294189230301e-07,
+      "loss": 0.9032,
+      "step": 22355
+    },
+    {
+      "epoch": 0.5385226752727535,
+      "grad_norm": 1.8780438899993896,
+      "learning_rate": 4.858026569313457e-07,
+      "loss": 0.8889,
+      "step": 22360
+    },
+    {
+      "epoch": 0.5386430962645409,
+      "grad_norm": 1.4092459678649902,
+      "learning_rate": 4.856758949396613e-07,
+      "loss": 0.9362,
+      "step": 22365
+    },
+    {
+      "epoch": 0.5387635172563281,
+      "grad_norm": 1.4840173721313477,
+      "learning_rate": 4.855491329479768e-07,
+      "loss": 0.9139,
+      "step": 22370
+    },
+    {
+      "epoch": 0.5388839382481154,
+      "grad_norm": 1.5067224502563477,
+      "learning_rate": 4.854223709562925e-07,
+      "loss": 0.9317,
+      "step": 22375
+    },
+    {
+      "epoch": 0.5390043592399028,
+      "grad_norm": 1.5796935558319092,
+      "learning_rate": 4.85295608964608e-07,
+      "loss": 0.8784,
+      "step": 22380
+    },
+    {
+      "epoch": 0.53912478023169,
+      "grad_norm": 2.1127724647521973,
+      "learning_rate": 4.851688469729236e-07,
+      "loss": 0.852,
+      "step": 22385
+    },
+    {
+      "epoch": 0.5392452012234773,
+      "grad_norm": 1.3739427328109741,
+      "learning_rate": 4.850420849812392e-07,
+      "loss": 0.9154,
+      "step": 22390
+    },
+    {
+      "epoch": 0.5393656222152645,
+      "grad_norm": 1.535537600517273,
+      "learning_rate": 4.849153229895548e-07,
+      "loss": 0.9482,
+      "step": 22395
+    },
+    {
+      "epoch": 0.5394860432070518,
+      "grad_norm": 1.9030694961547852,
+      "learning_rate": 4.847885609978704e-07,
+      "loss": 0.9118,
+      "step": 22400
+    },
+    {
+      "epoch": 0.5396064641988392,
+      "grad_norm": 1.6682997941970825,
+      "learning_rate": 4.84661799006186e-07,
+      "loss": 0.909,
+      "step": 22405
+    },
+    {
+      "epoch": 0.5397268851906264,
+      "grad_norm": 1.4649995565414429,
+      "learning_rate": 4.845350370145015e-07,
+      "loss": 0.9428,
+      "step": 22410
+    },
+    {
+      "epoch": 0.5398473061824137,
+      "grad_norm": 1.4420379400253296,
+      "learning_rate": 4.844082750228171e-07,
+      "loss": 0.892,
+      "step": 22415
+    },
+    {
+      "epoch": 0.539967727174201,
+      "grad_norm": 1.4790561199188232,
+      "learning_rate": 4.842815130311327e-07,
+      "loss": 0.9114,
+      "step": 22420
+    },
+    {
+      "epoch": 0.5400881481659883,
+      "grad_norm": 1.670272946357727,
+      "learning_rate": 4.841547510394483e-07,
+      "loss": 0.92,
+      "step": 22425
+    },
+    {
+      "epoch": 0.5402085691577756,
+      "grad_norm": 1.5772608518600464,
+      "learning_rate": 4.840279890477639e-07,
+      "loss": 0.9178,
+      "step": 22430
+    },
+    {
+      "epoch": 0.5403289901495629,
+      "grad_norm": 1.4250884056091309,
+      "learning_rate": 4.839012270560794e-07,
+      "loss": 0.9318,
+      "step": 22435
+    },
+    {
+      "epoch": 0.5404494111413501,
+      "grad_norm": 1.5450215339660645,
+      "learning_rate": 4.837744650643951e-07,
+      "loss": 0.904,
+      "step": 22440
+    },
+    {
+      "epoch": 0.5405698321331375,
+      "grad_norm": 1.6371772289276123,
+      "learning_rate": 4.836477030727106e-07,
+      "loss": 0.9654,
+      "step": 22445
+    },
+    {
+      "epoch": 0.5406902531249247,
+      "grad_norm": 1.6197057962417603,
+      "learning_rate": 4.835209410810263e-07,
+      "loss": 0.8841,
+      "step": 22450
+    },
+    {
+      "epoch": 0.540810674116712,
+      "grad_norm": 1.5857040882110596,
+      "learning_rate": 4.833941790893418e-07,
+      "loss": 0.8552,
+      "step": 22455
+    },
+    {
+      "epoch": 0.5409310951084993,
+      "grad_norm": 1.353580355644226,
+      "learning_rate": 4.832674170976574e-07,
+      "loss": 0.9054,
+      "step": 22460
+    },
+    {
+      "epoch": 0.5410515161002866,
+      "grad_norm": 1.7279608249664307,
+      "learning_rate": 4.83140655105973e-07,
+      "loss": 0.9183,
+      "step": 22465
+    },
+    {
+      "epoch": 0.5411719370920739,
+      "grad_norm": 1.3793567419052124,
+      "learning_rate": 4.830138931142885e-07,
+      "loss": 0.931,
+      "step": 22470
+    },
+    {
+      "epoch": 0.5412923580838612,
+      "grad_norm": 1.5536220073699951,
+      "learning_rate": 4.828871311226042e-07,
+      "loss": 0.9399,
+      "step": 22475
+    },
+    {
+      "epoch": 0.5414127790756484,
+      "grad_norm": 1.6547752618789673,
+      "learning_rate": 4.827603691309198e-07,
+      "loss": 0.9144,
+      "step": 22480
+    },
+    {
+      "epoch": 0.5415332000674358,
+      "grad_norm": 1.457759976387024,
+      "learning_rate": 4.826336071392354e-07,
+      "loss": 0.9018,
+      "step": 22485
+    },
+    {
+      "epoch": 0.541653621059223,
+      "grad_norm": 1.404378056526184,
+      "learning_rate": 4.825068451475509e-07,
+      "loss": 0.8929,
+      "step": 22490
+    },
+    {
+      "epoch": 0.5417740420510103,
+      "grad_norm": 1.51789391040802,
+      "learning_rate": 4.823800831558665e-07,
+      "loss": 0.8866,
+      "step": 22495
+    },
+    {
+      "epoch": 0.5418944630427976,
+      "grad_norm": 1.4357305765151978,
+      "learning_rate": 4.822533211641821e-07,
+      "loss": 0.9547,
+      "step": 22500
+    },
+    {
+      "epoch": 0.5420148840345849,
+      "grad_norm": 1.4671056270599365,
+      "learning_rate": 4.821265591724976e-07,
+      "loss": 0.8463,
+      "step": 22505
+    },
+    {
+      "epoch": 0.5421353050263722,
+      "grad_norm": 1.8944841623306274,
+      "learning_rate": 4.819997971808133e-07,
+      "loss": 0.9059,
+      "step": 22510
+    },
+    {
+      "epoch": 0.5422557260181595,
+      "grad_norm": 1.6028144359588623,
+      "learning_rate": 4.818730351891288e-07,
+      "loss": 0.8956,
+      "step": 22515
+    },
+    {
+      "epoch": 0.5423761470099467,
+      "grad_norm": 1.5150758028030396,
+      "learning_rate": 4.817462731974445e-07,
+      "loss": 0.9154,
+      "step": 22520
+    },
+    {
+      "epoch": 0.5424965680017341,
+      "grad_norm": 1.5906963348388672,
+      "learning_rate": 4.816195112057601e-07,
+      "loss": 0.883,
+      "step": 22525
+    },
+    {
+      "epoch": 0.5426169889935214,
+      "grad_norm": 1.5873745679855347,
+      "learning_rate": 4.814927492140756e-07,
+      "loss": 0.9274,
+      "step": 22530
+    },
+    {
+      "epoch": 0.5427374099853086,
+      "grad_norm": 1.822314739227295,
+      "learning_rate": 4.813659872223912e-07,
+      "loss": 0.9055,
+      "step": 22535
+    },
+    {
+      "epoch": 0.542857830977096,
+      "grad_norm": 1.9757715463638306,
+      "learning_rate": 4.812392252307068e-07,
+      "loss": 0.9248,
+      "step": 22540
+    },
+    {
+      "epoch": 0.5429782519688832,
+      "grad_norm": 1.5905802249908447,
+      "learning_rate": 4.811124632390224e-07,
+      "loss": 0.861,
+      "step": 22545
+    },
+    {
+      "epoch": 0.5430986729606705,
+      "grad_norm": 1.6692482233047485,
+      "learning_rate": 4.809857012473379e-07,
+      "loss": 0.9657,
+      "step": 22550
+    },
+    {
+      "epoch": 0.5432190939524578,
+      "grad_norm": 1.6076393127441406,
+      "learning_rate": 4.808589392556535e-07,
+      "loss": 0.957,
+      "step": 22555
+    },
+    {
+      "epoch": 0.543339514944245,
+      "grad_norm": 1.3963638544082642,
+      "learning_rate": 4.807321772639692e-07,
+      "loss": 0.8874,
+      "step": 22560
+    },
+    {
+      "epoch": 0.5434599359360324,
+      "grad_norm": 1.4064629077911377,
+      "learning_rate": 4.806054152722847e-07,
+      "loss": 0.8713,
+      "step": 22565
+    },
+    {
+      "epoch": 0.5435803569278197,
+      "grad_norm": 1.4674707651138306,
+      "learning_rate": 4.804786532806004e-07,
+      "loss": 0.92,
+      "step": 22570
+    },
+    {
+      "epoch": 0.5437007779196069,
+      "grad_norm": 1.6980865001678467,
+      "learning_rate": 4.803518912889159e-07,
+      "loss": 0.8332,
+      "step": 22575
+    },
+    {
+      "epoch": 0.5438211989113942,
+      "grad_norm": 1.4662402868270874,
+      "learning_rate": 4.802251292972316e-07,
+      "loss": 0.8667,
+      "step": 22580
+    },
+    {
+      "epoch": 0.5439416199031816,
+      "grad_norm": 1.815271019935608,
+      "learning_rate": 4.800983673055471e-07,
+      "loss": 0.9332,
+      "step": 22585
+    },
+    {
+      "epoch": 0.5440620408949688,
+      "grad_norm": 1.5627825260162354,
+      "learning_rate": 4.799716053138626e-07,
+      "loss": 0.8746,
+      "step": 22590
+    },
+    {
+      "epoch": 0.5441824618867561,
+      "grad_norm": 1.4710911512374878,
+      "learning_rate": 4.798448433221783e-07,
+      "loss": 0.9136,
+      "step": 22595
+    },
+    {
+      "epoch": 0.5443028828785433,
+      "grad_norm": 1.5587544441223145,
+      "learning_rate": 4.797180813304938e-07,
+      "loss": 0.9304,
+      "step": 22600
+    },
+    {
+      "epoch": 0.5444233038703307,
+      "grad_norm": 1.7592480182647705,
+      "learning_rate": 4.795913193388095e-07,
+      "loss": 0.8881,
+      "step": 22605
+    },
+    {
+      "epoch": 0.544543724862118,
+      "grad_norm": 1.6748290061950684,
+      "learning_rate": 4.79464557347125e-07,
+      "loss": 0.9363,
+      "step": 22610
+    },
+    {
+      "epoch": 0.5446641458539052,
+      "grad_norm": 1.4647150039672852,
+      "learning_rate": 4.793377953554406e-07,
+      "loss": 0.8691,
+      "step": 22615
+    },
+    {
+      "epoch": 0.5447845668456925,
+      "grad_norm": 1.3892542123794556,
+      "learning_rate": 4.792110333637562e-07,
+      "loss": 0.8757,
+      "step": 22620
+    },
+    {
+      "epoch": 0.5449049878374799,
+      "grad_norm": 1.6088064908981323,
+      "learning_rate": 4.790842713720718e-07,
+      "loss": 0.8947,
+      "step": 22625
+    },
+    {
+      "epoch": 0.5450254088292671,
+      "grad_norm": 1.6441105604171753,
+      "learning_rate": 4.789575093803874e-07,
+      "loss": 0.9049,
+      "step": 22630
+    },
+    {
+      "epoch": 0.5451458298210544,
+      "grad_norm": 1.6489001512527466,
+      "learning_rate": 4.788307473887029e-07,
+      "loss": 0.8746,
+      "step": 22635
+    },
+    {
+      "epoch": 0.5452662508128417,
+      "grad_norm": 1.6179732084274292,
+      "learning_rate": 4.787039853970186e-07,
+      "loss": 0.931,
+      "step": 22640
+    },
+    {
+      "epoch": 0.545386671804629,
+      "grad_norm": 1.3718429803848267,
+      "learning_rate": 4.785772234053341e-07,
+      "loss": 0.8971,
+      "step": 22645
+    },
+    {
+      "epoch": 0.5455070927964163,
+      "grad_norm": 1.6162413358688354,
+      "learning_rate": 4.784504614136497e-07,
+      "loss": 0.9371,
+      "step": 22650
+    },
+    {
+      "epoch": 0.5456275137882035,
+      "grad_norm": 1.4348382949829102,
+      "learning_rate": 4.783236994219653e-07,
+      "loss": 0.9229,
+      "step": 22655
+    },
+    {
+      "epoch": 0.5457479347799908,
+      "grad_norm": 1.6404297351837158,
+      "learning_rate": 4.781969374302809e-07,
+      "loss": 0.8652,
+      "step": 22660
+    },
+    {
+      "epoch": 0.5458683557717782,
+      "grad_norm": 1.6286433935165405,
+      "learning_rate": 4.780701754385965e-07,
+      "loss": 0.8742,
+      "step": 22665
+    },
+    {
+      "epoch": 0.5459887767635654,
+      "grad_norm": 1.5594221353530884,
+      "learning_rate": 4.779434134469121e-07,
+      "loss": 0.9247,
+      "step": 22670
+    },
+    {
+      "epoch": 0.5461091977553527,
+      "grad_norm": 1.4674808979034424,
+      "learning_rate": 4.778166514552276e-07,
+      "loss": 0.8623,
+      "step": 22675
+    },
+    {
+      "epoch": 0.54622961874714,
+      "grad_norm": 1.5083880424499512,
+      "learning_rate": 4.776898894635432e-07,
+      "loss": 0.9244,
+      "step": 22680
+    },
+    {
+      "epoch": 0.5463500397389273,
+      "grad_norm": 1.5525587797164917,
+      "learning_rate": 4.775631274718588e-07,
+      "loss": 0.9286,
+      "step": 22685
+    },
+    {
+      "epoch": 0.5464704607307146,
+      "grad_norm": 1.4932695627212524,
+      "learning_rate": 4.774363654801744e-07,
+      "loss": 0.9448,
+      "step": 22690
+    },
+    {
+      "epoch": 0.5465908817225019,
+      "grad_norm": 1.7255370616912842,
+      "learning_rate": 4.7730960348849e-07,
+      "loss": 0.9156,
+      "step": 22695
+    },
+    {
+      "epoch": 0.5467113027142891,
+      "grad_norm": 1.6489670276641846,
+      "learning_rate": 4.771828414968055e-07,
+      "loss": 0.8368,
+      "step": 22700
+    },
+    {
+      "epoch": 0.5468317237060765,
+      "grad_norm": 1.7103477716445923,
+      "learning_rate": 4.770560795051212e-07,
+      "loss": 0.9328,
+      "step": 22705
+    },
+    {
+      "epoch": 0.5469521446978637,
+      "grad_norm": 1.6121007204055786,
+      "learning_rate": 4.769293175134367e-07,
+      "loss": 0.877,
+      "step": 22710
+    },
+    {
+      "epoch": 0.547072565689651,
+      "grad_norm": 1.7308216094970703,
+      "learning_rate": 4.768025555217524e-07,
+      "loss": 0.9461,
+      "step": 22715
+    },
+    {
+      "epoch": 0.5471929866814383,
+      "grad_norm": 1.5269876718521118,
+      "learning_rate": 4.766757935300679e-07,
+      "loss": 0.8507,
+      "step": 22720
+    },
+    {
+      "epoch": 0.5473134076732256,
+      "grad_norm": 1.4367316961288452,
+      "learning_rate": 4.765490315383835e-07,
+      "loss": 0.9051,
+      "step": 22725
+    },
+    {
+      "epoch": 0.5474338286650129,
+      "grad_norm": 1.6838974952697754,
+      "learning_rate": 4.764222695466991e-07,
+      "loss": 0.9335,
+      "step": 22730
+    },
+    {
+      "epoch": 0.5475542496568002,
+      "grad_norm": 1.4725620746612549,
+      "learning_rate": 4.762955075550147e-07,
+      "loss": 0.9563,
+      "step": 22735
+    },
+    {
+      "epoch": 0.5476746706485874,
+      "grad_norm": 1.7399842739105225,
+      "learning_rate": 4.7616874556333023e-07,
+      "loss": 0.918,
+      "step": 22740
+    },
+    {
+      "epoch": 0.5477950916403748,
+      "grad_norm": 1.464184045791626,
+      "learning_rate": 4.7604198357164587e-07,
+      "loss": 0.9001,
+      "step": 22745
+    },
+    {
+      "epoch": 0.5479155126321621,
+      "grad_norm": 1.5431313514709473,
+      "learning_rate": 4.7591522157996147e-07,
+      "loss": 0.9322,
+      "step": 22750
+    },
+    {
+      "epoch": 0.5480359336239493,
+      "grad_norm": 1.7318702936172485,
+      "learning_rate": 4.75788459588277e-07,
+      "loss": 0.9075,
+      "step": 22755
+    },
+    {
+      "epoch": 0.5481563546157366,
+      "grad_norm": 1.4880088567733765,
+      "learning_rate": 4.7566169759659265e-07,
+      "loss": 0.8943,
+      "step": 22760
+    },
+    {
+      "epoch": 0.5482767756075239,
+      "grad_norm": 1.956777811050415,
+      "learning_rate": 4.755349356049082e-07,
+      "loss": 0.8973,
+      "step": 22765
+    },
+    {
+      "epoch": 0.5483971965993112,
+      "grad_norm": 1.4007956981658936,
+      "learning_rate": 4.754081736132238e-07,
+      "loss": 0.9259,
+      "step": 22770
+    },
+    {
+      "epoch": 0.5485176175910985,
+      "grad_norm": 1.6305876970291138,
+      "learning_rate": 4.7528141162153937e-07,
+      "loss": 0.9025,
+      "step": 22775
+    },
+    {
+      "epoch": 0.5486380385828857,
+      "grad_norm": 1.549984097480774,
+      "learning_rate": 4.7515464962985496e-07,
+      "loss": 0.9122,
+      "step": 22780
+    },
+    {
+      "epoch": 0.5487584595746731,
+      "grad_norm": 1.472762107849121,
+      "learning_rate": 4.7502788763817055e-07,
+      "loss": 0.8993,
+      "step": 22785
+    },
+    {
+      "epoch": 0.5488788805664604,
+      "grad_norm": 1.4677027463912964,
+      "learning_rate": 4.7490112564648615e-07,
+      "loss": 0.9269,
+      "step": 22790
+    },
+    {
+      "epoch": 0.5489993015582476,
+      "grad_norm": 1.7056957483291626,
+      "learning_rate": 4.7477436365480174e-07,
+      "loss": 0.8521,
+      "step": 22795
+    },
+    {
+      "epoch": 0.5491197225500349,
+      "grad_norm": 1.5326414108276367,
+      "learning_rate": 4.746476016631173e-07,
+      "loss": 0.9362,
+      "step": 22800
+    },
+    {
+      "epoch": 0.5492401435418222,
+      "grad_norm": 1.6755796670913696,
+      "learning_rate": 4.745208396714329e-07,
+      "loss": 0.9232,
+      "step": 22805
+    },
+    {
+      "epoch": 0.5493605645336095,
+      "grad_norm": 1.5694706439971924,
+      "learning_rate": 4.7439407767974846e-07,
+      "loss": 0.9571,
+      "step": 22810
+    },
+    {
+      "epoch": 0.5494809855253968,
+      "grad_norm": 1.4975767135620117,
+      "learning_rate": 4.7426731568806405e-07,
+      "loss": 0.8897,
+      "step": 22815
+    },
+    {
+      "epoch": 0.549601406517184,
+      "grad_norm": 1.4253288507461548,
+      "learning_rate": 4.741405536963797e-07,
+      "loss": 0.9265,
+      "step": 22820
+    },
+    {
+      "epoch": 0.5497218275089714,
+      "grad_norm": 1.4730470180511475,
+      "learning_rate": 4.7401379170469524e-07,
+      "loss": 0.9216,
+      "step": 22825
+    },
+    {
+      "epoch": 0.5498422485007587,
+      "grad_norm": 1.5416759252548218,
+      "learning_rate": 4.7388702971301083e-07,
+      "loss": 0.934,
+      "step": 22830
+    },
+    {
+      "epoch": 0.5499626694925459,
+      "grad_norm": 1.5281325578689575,
+      "learning_rate": 4.737602677213264e-07,
+      "loss": 0.877,
+      "step": 22835
+    },
+    {
+      "epoch": 0.5500830904843332,
+      "grad_norm": 1.6388013362884521,
+      "learning_rate": 4.73633505729642e-07,
+      "loss": 0.9129,
+      "step": 22840
+    },
+    {
+      "epoch": 0.5502035114761206,
+      "grad_norm": 1.5889734029769897,
+      "learning_rate": 4.7350674373795755e-07,
+      "loss": 0.9524,
+      "step": 22845
+    },
+    {
+      "epoch": 0.5503239324679078,
+      "grad_norm": 1.3709381818771362,
+      "learning_rate": 4.733799817462732e-07,
+      "loss": 0.8604,
+      "step": 22850
+    },
+    {
+      "epoch": 0.5504443534596951,
+      "grad_norm": 1.5920021533966064,
+      "learning_rate": 4.732532197545888e-07,
+      "loss": 0.9046,
+      "step": 22855
+    },
+    {
+      "epoch": 0.5505647744514823,
+      "grad_norm": 1.5857954025268555,
+      "learning_rate": 4.731264577629043e-07,
+      "loss": 0.9432,
+      "step": 22860
+    },
+    {
+      "epoch": 0.5506851954432697,
+      "grad_norm": 1.6190216541290283,
+      "learning_rate": 4.7299969577121997e-07,
+      "loss": 0.9056,
+      "step": 22865
+    },
+    {
+      "epoch": 0.550805616435057,
+      "grad_norm": 1.6539288759231567,
+      "learning_rate": 4.728729337795355e-07,
+      "loss": 0.8823,
+      "step": 22870
+    },
+    {
+      "epoch": 0.5509260374268442,
+      "grad_norm": 1.4087973833084106,
+      "learning_rate": 4.727461717878511e-07,
+      "loss": 0.8828,
+      "step": 22875
+    },
+    {
+      "epoch": 0.5510464584186315,
+      "grad_norm": 1.6573415994644165,
+      "learning_rate": 4.726194097961667e-07,
+      "loss": 0.8531,
+      "step": 22880
+    },
+    {
+      "epoch": 0.5511668794104189,
+      "grad_norm": 1.5957608222961426,
+      "learning_rate": 4.724926478044823e-07,
+      "loss": 0.9273,
+      "step": 22885
+    },
+    {
+      "epoch": 0.5512873004022061,
+      "grad_norm": 1.4989583492279053,
+      "learning_rate": 4.723658858127978e-07,
+      "loss": 0.9151,
+      "step": 22890
+    },
+    {
+      "epoch": 0.5514077213939934,
+      "grad_norm": 1.7770851850509644,
+      "learning_rate": 4.7223912382111347e-07,
+      "loss": 0.9288,
+      "step": 22895
+    },
+    {
+      "epoch": 0.5515281423857807,
+      "grad_norm": 1.4110249280929565,
+      "learning_rate": 4.7211236182942906e-07,
+      "loss": 0.9065,
+      "step": 22900
+    },
+    {
+      "epoch": 0.551648563377568,
+      "grad_norm": 1.5001226663589478,
+      "learning_rate": 4.719855998377446e-07,
+      "loss": 0.917,
+      "step": 22905
+    },
+    {
+      "epoch": 0.5517689843693553,
+      "grad_norm": 1.3509827852249146,
+      "learning_rate": 4.7185883784606024e-07,
+      "loss": 0.9521,
+      "step": 22910
+    },
+    {
+      "epoch": 0.5518894053611425,
+      "grad_norm": 1.6468544006347656,
+      "learning_rate": 4.717320758543758e-07,
+      "loss": 0.8757,
+      "step": 22915
+    },
+    {
+      "epoch": 0.5520098263529298,
+      "grad_norm": 1.6438676118850708,
+      "learning_rate": 4.716053138626914e-07,
+      "loss": 0.8843,
+      "step": 22920
+    },
+    {
+      "epoch": 0.5521302473447172,
+      "grad_norm": 1.6613961458206177,
+      "learning_rate": 4.7147855187100697e-07,
+      "loss": 0.946,
+      "step": 22925
+    },
+    {
+      "epoch": 0.5522506683365044,
+      "grad_norm": 1.5499963760375977,
+      "learning_rate": 4.7135178987932256e-07,
+      "loss": 0.89,
+      "step": 22930
+    },
+    {
+      "epoch": 0.5523710893282917,
+      "grad_norm": 1.6481000185012817,
+      "learning_rate": 4.7122502788763815e-07,
+      "loss": 0.926,
+      "step": 22935
+    },
+    {
+      "epoch": 0.552491510320079,
+      "grad_norm": 1.503867506980896,
+      "learning_rate": 4.7109826589595374e-07,
+      "loss": 0.9033,
+      "step": 22940
+    },
+    {
+      "epoch": 0.5526119313118663,
+      "grad_norm": 1.8247016668319702,
+      "learning_rate": 4.7097150390426933e-07,
+      "loss": 0.8992,
+      "step": 22945
+    },
+    {
+      "epoch": 0.5527323523036536,
+      "grad_norm": 1.4254915714263916,
+      "learning_rate": 4.7084474191258487e-07,
+      "loss": 0.8789,
+      "step": 22950
+    },
+    {
+      "epoch": 0.5528527732954409,
+      "grad_norm": 1.5164680480957031,
+      "learning_rate": 4.707179799209005e-07,
+      "loss": 0.9925,
+      "step": 22955
+    },
+    {
+      "epoch": 0.5529731942872281,
+      "grad_norm": 1.5180974006652832,
+      "learning_rate": 4.7059121792921606e-07,
+      "loss": 0.9097,
+      "step": 22960
+    },
+    {
+      "epoch": 0.5530936152790155,
+      "grad_norm": 1.6972721815109253,
+      "learning_rate": 4.7046445593753165e-07,
+      "loss": 0.855,
+      "step": 22965
+    },
+    {
+      "epoch": 0.5532140362708027,
+      "grad_norm": 1.5323702096939087,
+      "learning_rate": 4.703376939458473e-07,
+      "loss": 0.931,
+      "step": 22970
+    },
+    {
+      "epoch": 0.55333445726259,
+      "grad_norm": 1.7028616666793823,
+      "learning_rate": 4.7021093195416283e-07,
+      "loss": 0.8891,
+      "step": 22975
+    },
+    {
+      "epoch": 0.5534548782543773,
+      "grad_norm": 1.4747989177703857,
+      "learning_rate": 4.700841699624784e-07,
+      "loss": 0.9096,
+      "step": 22980
+    },
+    {
+      "epoch": 0.5535752992461646,
+      "grad_norm": 1.4280892610549927,
+      "learning_rate": 4.69957407970794e-07,
+      "loss": 0.8631,
+      "step": 22985
+    },
+    {
+      "epoch": 0.5536957202379519,
+      "grad_norm": 1.6729850769042969,
+      "learning_rate": 4.698306459791096e-07,
+      "loss": 0.9157,
+      "step": 22990
+    },
+    {
+      "epoch": 0.5538161412297392,
+      "grad_norm": 1.6169942617416382,
+      "learning_rate": 4.6970388398742515e-07,
+      "loss": 0.8625,
+      "step": 22995
+    },
+    {
+      "epoch": 0.5539365622215264,
+      "grad_norm": 1.4130114316940308,
+      "learning_rate": 4.695771219957408e-07,
+      "loss": 0.9,
+      "step": 23000
+    },
+    {
+      "epoch": 0.5540569832133138,
+      "grad_norm": 1.5743272304534912,
+      "learning_rate": 4.694503600040564e-07,
+      "loss": 0.8981,
+      "step": 23005
+    },
+    {
+      "epoch": 0.5541774042051011,
+      "grad_norm": 1.4807997941970825,
+      "learning_rate": 4.693235980123719e-07,
+      "loss": 0.874,
+      "step": 23010
+    },
+    {
+      "epoch": 0.5542978251968883,
+      "grad_norm": 1.649823784828186,
+      "learning_rate": 4.6919683602068757e-07,
+      "loss": 0.8946,
+      "step": 23015
+    },
+    {
+      "epoch": 0.5544182461886756,
+      "grad_norm": 1.410031795501709,
+      "learning_rate": 4.690700740290031e-07,
+      "loss": 0.9026,
+      "step": 23020
+    },
+    {
+      "epoch": 0.5545386671804629,
+      "grad_norm": 1.7367534637451172,
+      "learning_rate": 4.689433120373187e-07,
+      "loss": 0.9166,
+      "step": 23025
+    },
+    {
+      "epoch": 0.5546590881722502,
+      "grad_norm": 1.5176085233688354,
+      "learning_rate": 4.688165500456343e-07,
+      "loss": 0.9129,
+      "step": 23030
+    },
+    {
+      "epoch": 0.5547795091640375,
+      "grad_norm": 1.437880039215088,
+      "learning_rate": 4.686897880539499e-07,
+      "loss": 0.9501,
+      "step": 23035
+    },
+    {
+      "epoch": 0.5548999301558247,
+      "grad_norm": 1.7201894521713257,
+      "learning_rate": 4.685630260622654e-07,
+      "loss": 0.8532,
+      "step": 23040
+    },
+    {
+      "epoch": 0.5550203511476121,
+      "grad_norm": 1.4754242897033691,
+      "learning_rate": 4.6843626407058106e-07,
+      "loss": 0.8962,
+      "step": 23045
+    },
+    {
+      "epoch": 0.5551407721393994,
+      "grad_norm": 1.5195088386535645,
+      "learning_rate": 4.6830950207889665e-07,
+      "loss": 0.8434,
+      "step": 23050
+    },
+    {
+      "epoch": 0.5552611931311866,
+      "grad_norm": 1.5234872102737427,
+      "learning_rate": 4.681827400872122e-07,
+      "loss": 0.9074,
+      "step": 23055
+    },
+    {
+      "epoch": 0.5553816141229739,
+      "grad_norm": 1.3597567081451416,
+      "learning_rate": 4.6805597809552784e-07,
+      "loss": 0.8834,
+      "step": 23060
+    },
+    {
+      "epoch": 0.5555020351147612,
+      "grad_norm": 1.5739555358886719,
+      "learning_rate": 4.679292161038434e-07,
+      "loss": 0.8921,
+      "step": 23065
+    },
+    {
+      "epoch": 0.5556224561065485,
+      "grad_norm": 1.5108563899993896,
+      "learning_rate": 4.6780245411215897e-07,
+      "loss": 0.9429,
+      "step": 23070
+    },
+    {
+      "epoch": 0.5557428770983358,
+      "grad_norm": 1.4870704412460327,
+      "learning_rate": 4.6767569212047456e-07,
+      "loss": 0.8813,
+      "step": 23075
+    },
+    {
+      "epoch": 0.555863298090123,
+      "grad_norm": 1.5405627489089966,
+      "learning_rate": 4.6754893012879015e-07,
+      "loss": 0.9435,
+      "step": 23080
+    },
+    {
+      "epoch": 0.5559837190819104,
+      "grad_norm": 1.615705966949463,
+      "learning_rate": 4.674221681371058e-07,
+      "loss": 0.9322,
+      "step": 23085
+    },
+    {
+      "epoch": 0.5561041400736977,
+      "grad_norm": 1.6554853916168213,
+      "learning_rate": 4.6729540614542134e-07,
+      "loss": 0.8981,
+      "step": 23090
+    },
+    {
+      "epoch": 0.5562245610654849,
+      "grad_norm": 1.5215091705322266,
+      "learning_rate": 4.6716864415373693e-07,
+      "loss": 0.8765,
+      "step": 23095
+    },
+    {
+      "epoch": 0.5563449820572722,
+      "grad_norm": 2.1298627853393555,
+      "learning_rate": 4.670418821620525e-07,
+      "loss": 0.877,
+      "step": 23100
+    },
+    {
+      "epoch": 0.5564654030490596,
+      "grad_norm": 1.5979092121124268,
+      "learning_rate": 4.669151201703681e-07,
+      "loss": 0.9118,
+      "step": 23105
+    },
+    {
+      "epoch": 0.5565858240408468,
+      "grad_norm": 1.449164867401123,
+      "learning_rate": 4.6678835817868365e-07,
+      "loss": 0.8959,
+      "step": 23110
+    },
+    {
+      "epoch": 0.5567062450326341,
+      "grad_norm": 1.5024070739746094,
+      "learning_rate": 4.666615961869993e-07,
+      "loss": 0.9059,
+      "step": 23115
+    },
+    {
+      "epoch": 0.5568266660244213,
+      "grad_norm": 1.5619574785232544,
+      "learning_rate": 4.665348341953149e-07,
+      "loss": 0.9053,
+      "step": 23120
+    },
+    {
+      "epoch": 0.5569470870162087,
+      "grad_norm": 1.5824140310287476,
+      "learning_rate": 4.664080722036304e-07,
+      "loss": 0.8958,
+      "step": 23125
+    },
+    {
+      "epoch": 0.557067508007996,
+      "grad_norm": 1.613577127456665,
+      "learning_rate": 4.6628131021194607e-07,
+      "loss": 0.8602,
+      "step": 23130
+    },
+    {
+      "epoch": 0.5571879289997832,
+      "grad_norm": 1.4480857849121094,
+      "learning_rate": 4.661545482202616e-07,
+      "loss": 0.8349,
+      "step": 23135
+    },
+    {
+      "epoch": 0.5573083499915705,
+      "grad_norm": 1.4946420192718506,
+      "learning_rate": 4.660277862285772e-07,
+      "loss": 0.8484,
+      "step": 23140
+    },
+    {
+      "epoch": 0.5574287709833579,
+      "grad_norm": 1.4564869403839111,
+      "learning_rate": 4.659010242368928e-07,
+      "loss": 0.8754,
+      "step": 23145
+    },
+    {
+      "epoch": 0.5575491919751451,
+      "grad_norm": 1.4581509828567505,
+      "learning_rate": 4.657742622452084e-07,
+      "loss": 0.8587,
+      "step": 23150
+    },
+    {
+      "epoch": 0.5576696129669324,
+      "grad_norm": 1.5620570182800293,
+      "learning_rate": 4.65647500253524e-07,
+      "loss": 0.9003,
+      "step": 23155
+    },
+    {
+      "epoch": 0.5577900339587197,
+      "grad_norm": 1.7160062789916992,
+      "learning_rate": 4.6552073826183957e-07,
+      "loss": 0.9365,
+      "step": 23160
+    },
+    {
+      "epoch": 0.557910454950507,
+      "grad_norm": 1.7763422727584839,
+      "learning_rate": 4.6539397627015516e-07,
+      "loss": 0.9056,
+      "step": 23165
+    },
+    {
+      "epoch": 0.5580308759422943,
+      "grad_norm": 1.4742810726165771,
+      "learning_rate": 4.652672142784707e-07,
+      "loss": 0.915,
+      "step": 23170
+    },
+    {
+      "epoch": 0.5581512969340815,
+      "grad_norm": 1.7074137926101685,
+      "learning_rate": 4.6514045228678634e-07,
+      "loss": 0.9176,
+      "step": 23175
+    },
+    {
+      "epoch": 0.5582717179258688,
+      "grad_norm": 1.6095412969589233,
+      "learning_rate": 4.650136902951019e-07,
+      "loss": 0.9024,
+      "step": 23180
+    },
+    {
+      "epoch": 0.5583921389176562,
+      "grad_norm": 1.5504175424575806,
+      "learning_rate": 4.648869283034175e-07,
+      "loss": 0.848,
+      "step": 23185
+    },
+    {
+      "epoch": 0.5585125599094434,
+      "grad_norm": 1.5489128828048706,
+      "learning_rate": 4.647601663117331e-07,
+      "loss": 0.928,
+      "step": 23190
+    },
+    {
+      "epoch": 0.5586329809012307,
+      "grad_norm": 1.507321834564209,
+      "learning_rate": 4.6463340432004866e-07,
+      "loss": 0.8686,
+      "step": 23195
+    },
+    {
+      "epoch": 0.558753401893018,
+      "grad_norm": 1.7442164421081543,
+      "learning_rate": 4.6450664232836425e-07,
+      "loss": 0.9243,
+      "step": 23200
+    },
+    {
+      "epoch": 0.5588738228848052,
+      "grad_norm": 1.522963047027588,
+      "learning_rate": 4.6437988033667984e-07,
+      "loss": 0.9081,
+      "step": 23205
+    },
+    {
+      "epoch": 0.5589942438765926,
+      "grad_norm": 1.4842394590377808,
+      "learning_rate": 4.6425311834499543e-07,
+      "loss": 0.9395,
+      "step": 23210
+    },
+    {
+      "epoch": 0.5591146648683799,
+      "grad_norm": 1.8214404582977295,
+      "learning_rate": 4.6412635635331097e-07,
+      "loss": 0.9317,
+      "step": 23215
+    },
+    {
+      "epoch": 0.5592350858601671,
+      "grad_norm": 1.510892629623413,
+      "learning_rate": 4.639995943616266e-07,
+      "loss": 0.8752,
+      "step": 23220
+    },
+    {
+      "epoch": 0.5593555068519545,
+      "grad_norm": 1.4507983922958374,
+      "learning_rate": 4.6387283236994216e-07,
+      "loss": 0.8842,
+      "step": 23225
+    },
+    {
+      "epoch": 0.5594759278437417,
+      "grad_norm": 1.5228618383407593,
+      "learning_rate": 4.6374607037825775e-07,
+      "loss": 0.8702,
+      "step": 23230
+    },
+    {
+      "epoch": 0.559596348835529,
+      "grad_norm": 1.431118130683899,
+      "learning_rate": 4.636193083865734e-07,
+      "loss": 0.913,
+      "step": 23235
+    },
+    {
+      "epoch": 0.5597167698273163,
+      "grad_norm": 1.5003418922424316,
+      "learning_rate": 4.6349254639488893e-07,
+      "loss": 0.9081,
+      "step": 23240
+    },
+    {
+      "epoch": 0.5598371908191035,
+      "grad_norm": 1.5814661979675293,
+      "learning_rate": 4.633657844032045e-07,
+      "loss": 0.9273,
+      "step": 23245
+    },
+    {
+      "epoch": 0.5599576118108909,
+      "grad_norm": 1.5471625328063965,
+      "learning_rate": 4.632390224115201e-07,
+      "loss": 0.8741,
+      "step": 23250
+    },
+    {
+      "epoch": 0.5600780328026782,
+      "grad_norm": 1.619927167892456,
+      "learning_rate": 4.631122604198357e-07,
+      "loss": 0.893,
+      "step": 23255
+    },
+    {
+      "epoch": 0.5601984537944654,
+      "grad_norm": 1.6046520471572876,
+      "learning_rate": 4.6298549842815125e-07,
+      "loss": 0.9589,
+      "step": 23260
+    },
+    {
+      "epoch": 0.5603188747862528,
+      "grad_norm": 1.6255428791046143,
+      "learning_rate": 4.628587364364669e-07,
+      "loss": 0.8499,
+      "step": 23265
+    },
+    {
+      "epoch": 0.5604392957780401,
+      "grad_norm": 1.6193327903747559,
+      "learning_rate": 4.627319744447825e-07,
+      "loss": 0.8742,
+      "step": 23270
+    },
+    {
+      "epoch": 0.5605597167698273,
+      "grad_norm": 1.8600654602050781,
+      "learning_rate": 4.62605212453098e-07,
+      "loss": 0.8795,
+      "step": 23275
+    },
+    {
+      "epoch": 0.5606801377616146,
+      "grad_norm": 1.4893161058425903,
+      "learning_rate": 4.6247845046141367e-07,
+      "loss": 0.9228,
+      "step": 23280
+    },
+    {
+      "epoch": 0.5608005587534018,
+      "grad_norm": 1.4823611974716187,
+      "learning_rate": 4.623516884697292e-07,
+      "loss": 0.8934,
+      "step": 23285
+    },
+    {
+      "epoch": 0.5609209797451892,
+      "grad_norm": 1.4936082363128662,
+      "learning_rate": 4.622249264780448e-07,
+      "loss": 0.8465,
+      "step": 23290
+    },
+    {
+      "epoch": 0.5610414007369765,
+      "grad_norm": 1.565303087234497,
+      "learning_rate": 4.620981644863604e-07,
+      "loss": 0.8208,
+      "step": 23295
+    },
+    {
+      "epoch": 0.5611618217287637,
+      "grad_norm": 1.5002118349075317,
+      "learning_rate": 4.61971402494676e-07,
+      "loss": 0.8306,
+      "step": 23300
+    },
+    {
+      "epoch": 0.561282242720551,
+      "grad_norm": 1.7962204217910767,
+      "learning_rate": 4.618446405029915e-07,
+      "loss": 0.9261,
+      "step": 23305
+    },
+    {
+      "epoch": 0.5614026637123384,
+      "grad_norm": 1.486973524093628,
+      "learning_rate": 4.6171787851130716e-07,
+      "loss": 0.8782,
+      "step": 23310
+    },
+    {
+      "epoch": 0.5615230847041256,
+      "grad_norm": 1.7047168016433716,
+      "learning_rate": 4.6159111651962276e-07,
+      "loss": 0.937,
+      "step": 23315
+    },
+    {
+      "epoch": 0.5616435056959129,
+      "grad_norm": 1.7348361015319824,
+      "learning_rate": 4.614643545279383e-07,
+      "loss": 0.9046,
+      "step": 23320
+    },
+    {
+      "epoch": 0.5617639266877003,
+      "grad_norm": 1.6398698091506958,
+      "learning_rate": 4.6133759253625394e-07,
+      "loss": 0.9449,
+      "step": 23325
+    },
+    {
+      "epoch": 0.5618843476794875,
+      "grad_norm": 1.6072243452072144,
+      "learning_rate": 4.612108305445695e-07,
+      "loss": 0.8987,
+      "step": 23330
+    },
+    {
+      "epoch": 0.5620047686712748,
+      "grad_norm": 1.5447580814361572,
+      "learning_rate": 4.6108406855288507e-07,
+      "loss": 0.8998,
+      "step": 23335
+    },
+    {
+      "epoch": 0.562125189663062,
+      "grad_norm": 1.4506652355194092,
+      "learning_rate": 4.6095730656120066e-07,
+      "loss": 0.9006,
+      "step": 23340
+    },
+    {
+      "epoch": 0.5622456106548493,
+      "grad_norm": 1.8912750482559204,
+      "learning_rate": 4.6083054456951625e-07,
+      "loss": 0.9356,
+      "step": 23345
+    },
+    {
+      "epoch": 0.5623660316466367,
+      "grad_norm": 1.5937507152557373,
+      "learning_rate": 4.6070378257783184e-07,
+      "loss": 0.8739,
+      "step": 23350
+    },
+    {
+      "epoch": 0.5624864526384239,
+      "grad_norm": 1.5998436212539673,
+      "learning_rate": 4.6057702058614744e-07,
+      "loss": 0.9358,
+      "step": 23355
+    },
+    {
+      "epoch": 0.5626068736302112,
+      "grad_norm": 1.690090537071228,
+      "learning_rate": 4.6045025859446303e-07,
+      "loss": 0.9209,
+      "step": 23360
+    },
+    {
+      "epoch": 0.5627272946219986,
+      "grad_norm": 1.6583771705627441,
+      "learning_rate": 4.6032349660277857e-07,
+      "loss": 0.9291,
+      "step": 23365
+    },
+    {
+      "epoch": 0.5628477156137858,
+      "grad_norm": 1.3349385261535645,
+      "learning_rate": 4.601967346110942e-07,
+      "loss": 0.865,
+      "step": 23370
+    },
+    {
+      "epoch": 0.5629681366055731,
+      "grad_norm": 1.6301311254501343,
+      "learning_rate": 4.6006997261940975e-07,
+      "loss": 0.8853,
+      "step": 23375
+    },
+    {
+      "epoch": 0.5630885575973603,
+      "grad_norm": 1.294317603111267,
+      "learning_rate": 4.5994321062772534e-07,
+      "loss": 0.8765,
+      "step": 23380
+    },
+    {
+      "epoch": 0.5632089785891476,
+      "grad_norm": 1.6579155921936035,
+      "learning_rate": 4.59816448636041e-07,
+      "loss": 0.8699,
+      "step": 23385
+    },
+    {
+      "epoch": 0.563329399580935,
+      "grad_norm": 1.5684454441070557,
+      "learning_rate": 4.596896866443565e-07,
+      "loss": 0.9242,
+      "step": 23390
+    },
+    {
+      "epoch": 0.5634498205727222,
+      "grad_norm": 1.7211233377456665,
+      "learning_rate": 4.595629246526721e-07,
+      "loss": 0.9232,
+      "step": 23395
+    },
+    {
+      "epoch": 0.5635702415645095,
+      "grad_norm": 1.4012408256530762,
+      "learning_rate": 4.594361626609877e-07,
+      "loss": 0.9368,
+      "step": 23400
+    },
+    {
+      "epoch": 0.5636906625562969,
+      "grad_norm": 1.6189051866531372,
+      "learning_rate": 4.593094006693033e-07,
+      "loss": 0.8924,
+      "step": 23405
+    },
+    {
+      "epoch": 0.5638110835480841,
+      "grad_norm": 1.5496536493301392,
+      "learning_rate": 4.5918263867761884e-07,
+      "loss": 0.9362,
+      "step": 23410
+    },
+    {
+      "epoch": 0.5639315045398714,
+      "grad_norm": 1.6953816413879395,
+      "learning_rate": 4.590558766859345e-07,
+      "loss": 0.91,
+      "step": 23415
+    },
+    {
+      "epoch": 0.5640519255316587,
+      "grad_norm": 1.9266865253448486,
+      "learning_rate": 4.589291146942501e-07,
+      "loss": 0.8785,
+      "step": 23420
+    },
+    {
+      "epoch": 0.5641723465234459,
+      "grad_norm": 1.821556568145752,
+      "learning_rate": 4.588023527025656e-07,
+      "loss": 0.889,
+      "step": 23425
+    },
+    {
+      "epoch": 0.5642927675152333,
+      "grad_norm": 1.4360072612762451,
+      "learning_rate": 4.5867559071088126e-07,
+      "loss": 0.9084,
+      "step": 23430
+    },
+    {
+      "epoch": 0.5644131885070205,
+      "grad_norm": 1.4635649919509888,
+      "learning_rate": 4.585488287191968e-07,
+      "loss": 0.9179,
+      "step": 23435
+    },
+    {
+      "epoch": 0.5645336094988078,
+      "grad_norm": 1.7907929420471191,
+      "learning_rate": 4.584220667275124e-07,
+      "loss": 0.8953,
+      "step": 23440
+    },
+    {
+      "epoch": 0.5646540304905951,
+      "grad_norm": 1.54899263381958,
+      "learning_rate": 4.58295304735828e-07,
+      "loss": 0.9073,
+      "step": 23445
+    },
+    {
+      "epoch": 0.5647744514823824,
+      "grad_norm": 1.6154134273529053,
+      "learning_rate": 4.581685427441436e-07,
+      "loss": 0.8705,
+      "step": 23450
+    },
+    {
+      "epoch": 0.5648948724741697,
+      "grad_norm": 1.8652158975601196,
+      "learning_rate": 4.580417807524591e-07,
+      "loss": 0.9056,
+      "step": 23455
+    },
+    {
+      "epoch": 0.565015293465957,
+      "grad_norm": 1.623321533203125,
+      "learning_rate": 4.5791501876077476e-07,
+      "loss": 0.8701,
+      "step": 23460
+    },
+    {
+      "epoch": 0.5651357144577442,
+      "grad_norm": 1.4679025411605835,
+      "learning_rate": 4.5778825676909035e-07,
+      "loss": 0.9068,
+      "step": 23465
+    },
+    {
+      "epoch": 0.5652561354495316,
+      "grad_norm": 1.803037166595459,
+      "learning_rate": 4.576614947774059e-07,
+      "loss": 0.9303,
+      "step": 23470
+    },
+    {
+      "epoch": 0.5653765564413189,
+      "grad_norm": 1.5411866903305054,
+      "learning_rate": 4.5753473278572153e-07,
+      "loss": 0.9731,
+      "step": 23475
+    },
+    {
+      "epoch": 0.5654969774331061,
+      "grad_norm": 1.4283922910690308,
+      "learning_rate": 4.5740797079403707e-07,
+      "loss": 0.8631,
+      "step": 23480
+    },
+    {
+      "epoch": 0.5656173984248934,
+      "grad_norm": 1.6044667959213257,
+      "learning_rate": 4.5728120880235266e-07,
+      "loss": 0.8486,
+      "step": 23485
+    },
+    {
+      "epoch": 0.5657378194166807,
+      "grad_norm": 1.5341675281524658,
+      "learning_rate": 4.5715444681066826e-07,
+      "loss": 0.9033,
+      "step": 23490
+    },
+    {
+      "epoch": 0.565858240408468,
+      "grad_norm": 1.3251118659973145,
+      "learning_rate": 4.5702768481898385e-07,
+      "loss": 0.9178,
+      "step": 23495
+    },
+    {
+      "epoch": 0.5659786614002553,
+      "grad_norm": 1.699512004852295,
+      "learning_rate": 4.5690092282729944e-07,
+      "loss": 0.8354,
+      "step": 23500
+    },
+    {
+      "epoch": 0.5660990823920425,
+      "grad_norm": 1.61118483543396,
+      "learning_rate": 4.5677416083561503e-07,
+      "loss": 0.8689,
+      "step": 23505
+    },
+    {
+      "epoch": 0.5662195033838299,
+      "grad_norm": 1.5231705904006958,
+      "learning_rate": 4.566473988439306e-07,
+      "loss": 0.9298,
+      "step": 23510
+    },
+    {
+      "epoch": 0.5663399243756172,
+      "grad_norm": 1.595055103302002,
+      "learning_rate": 4.5652063685224616e-07,
+      "loss": 0.8918,
+      "step": 23515
+    },
+    {
+      "epoch": 0.5664603453674044,
+      "grad_norm": 1.7120904922485352,
+      "learning_rate": 4.563938748605618e-07,
+      "loss": 0.8982,
+      "step": 23520
+    },
+    {
+      "epoch": 0.5665807663591917,
+      "grad_norm": 1.5685055255889893,
+      "learning_rate": 4.5626711286887735e-07,
+      "loss": 0.9227,
+      "step": 23525
+    },
+    {
+      "epoch": 0.5667011873509791,
+      "grad_norm": 1.5623687505722046,
+      "learning_rate": 4.5614035087719294e-07,
+      "loss": 0.9282,
+      "step": 23530
+    },
+    {
+      "epoch": 0.5668216083427663,
+      "grad_norm": 1.6257749795913696,
+      "learning_rate": 4.560135888855086e-07,
+      "loss": 0.9173,
+      "step": 23535
+    },
+    {
+      "epoch": 0.5669420293345536,
+      "grad_norm": 1.4779962301254272,
+      "learning_rate": 4.558868268938241e-07,
+      "loss": 0.9495,
+      "step": 23540
+    },
+    {
+      "epoch": 0.5670624503263408,
+      "grad_norm": 1.5573982000350952,
+      "learning_rate": 4.557600649021397e-07,
+      "loss": 0.936,
+      "step": 23545
+    },
+    {
+      "epoch": 0.5671828713181282,
+      "grad_norm": 1.638119101524353,
+      "learning_rate": 4.556333029104553e-07,
+      "loss": 0.9203,
+      "step": 23550
+    },
+    {
+      "epoch": 0.5673032923099155,
+      "grad_norm": 1.526727318763733,
+      "learning_rate": 4.555065409187709e-07,
+      "loss": 0.9222,
+      "step": 23555
+    },
+    {
+      "epoch": 0.5674237133017027,
+      "grad_norm": 1.6723929643630981,
+      "learning_rate": 4.5537977892708644e-07,
+      "loss": 0.9457,
+      "step": 23560
+    },
+    {
+      "epoch": 0.56754413429349,
+      "grad_norm": 1.4846093654632568,
+      "learning_rate": 4.552530169354021e-07,
+      "loss": 0.9399,
+      "step": 23565
+    },
+    {
+      "epoch": 0.5676645552852774,
+      "grad_norm": 1.665597677230835,
+      "learning_rate": 4.5512625494371767e-07,
+      "loss": 0.9339,
+      "step": 23570
+    },
+    {
+      "epoch": 0.5677849762770646,
+      "grad_norm": 1.617017149925232,
+      "learning_rate": 4.549994929520332e-07,
+      "loss": 0.9112,
+      "step": 23575
+    },
+    {
+      "epoch": 0.5679053972688519,
+      "grad_norm": 1.481245756149292,
+      "learning_rate": 4.5487273096034886e-07,
+      "loss": 0.9376,
+      "step": 23580
+    },
+    {
+      "epoch": 0.5680258182606392,
+      "grad_norm": 1.5374301671981812,
+      "learning_rate": 4.547459689686644e-07,
+      "loss": 0.9231,
+      "step": 23585
+    },
+    {
+      "epoch": 0.5681462392524265,
+      "grad_norm": 1.8817466497421265,
+      "learning_rate": 4.5461920697698e-07,
+      "loss": 0.9229,
+      "step": 23590
+    },
+    {
+      "epoch": 0.5682666602442138,
+      "grad_norm": 1.4146652221679688,
+      "learning_rate": 4.544924449852956e-07,
+      "loss": 0.8861,
+      "step": 23595
+    },
+    {
+      "epoch": 0.568387081236001,
+      "grad_norm": 1.700593113899231,
+      "learning_rate": 4.5436568299361117e-07,
+      "loss": 0.8761,
+      "step": 23600
+    },
+    {
+      "epoch": 0.5685075022277883,
+      "grad_norm": 1.4120657444000244,
+      "learning_rate": 4.542389210019268e-07,
+      "loss": 0.963,
+      "step": 23605
+    },
+    {
+      "epoch": 0.5686279232195757,
+      "grad_norm": 1.45216703414917,
+      "learning_rate": 4.5411215901024235e-07,
+      "loss": 0.913,
+      "step": 23610
+    },
+    {
+      "epoch": 0.5687483442113629,
+      "grad_norm": 1.4401744604110718,
+      "learning_rate": 4.5398539701855795e-07,
+      "loss": 0.9264,
+      "step": 23615
+    },
+    {
+      "epoch": 0.5688687652031502,
+      "grad_norm": 1.587065577507019,
+      "learning_rate": 4.5385863502687354e-07,
+      "loss": 0.8823,
+      "step": 23620
+    },
+    {
+      "epoch": 0.5689891861949375,
+      "grad_norm": 1.7187548875808716,
+      "learning_rate": 4.5373187303518913e-07,
+      "loss": 0.8663,
+      "step": 23625
+    },
+    {
+      "epoch": 0.5691096071867248,
+      "grad_norm": 1.6272677183151245,
+      "learning_rate": 4.5360511104350467e-07,
+      "loss": 0.8952,
+      "step": 23630
+    },
+    {
+      "epoch": 0.5692300281785121,
+      "grad_norm": 1.7676702737808228,
+      "learning_rate": 4.534783490518203e-07,
+      "loss": 0.94,
+      "step": 23635
+    },
+    {
+      "epoch": 0.5693504491702994,
+      "grad_norm": 1.674362301826477,
+      "learning_rate": 4.5335158706013585e-07,
+      "loss": 0.9102,
+      "step": 23640
+    },
+    {
+      "epoch": 0.5694708701620866,
+      "grad_norm": 1.3730660676956177,
+      "learning_rate": 4.5322482506845144e-07,
+      "loss": 0.808,
+      "step": 23645
+    },
+    {
+      "epoch": 0.569591291153874,
+      "grad_norm": 1.5507595539093018,
+      "learning_rate": 4.530980630767671e-07,
+      "loss": 0.8651,
+      "step": 23650
+    },
+    {
+      "epoch": 0.5697117121456612,
+      "grad_norm": 1.7860714197158813,
+      "learning_rate": 4.5297130108508263e-07,
+      "loss": 0.9472,
+      "step": 23655
+    },
+    {
+      "epoch": 0.5698321331374485,
+      "grad_norm": 1.460296869277954,
+      "learning_rate": 4.528445390933982e-07,
+      "loss": 0.8818,
+      "step": 23660
+    },
+    {
+      "epoch": 0.5699525541292358,
+      "grad_norm": 1.6900402307510376,
+      "learning_rate": 4.527177771017138e-07,
+      "loss": 1.0138,
+      "step": 23665
+    },
+    {
+      "epoch": 0.5700729751210231,
+      "grad_norm": 1.537196159362793,
+      "learning_rate": 4.525910151100294e-07,
+      "loss": 0.8917,
+      "step": 23670
+    },
+    {
+      "epoch": 0.5701933961128104,
+      "grad_norm": 1.4395272731781006,
+      "learning_rate": 4.5246425311834494e-07,
+      "loss": 0.9106,
+      "step": 23675
+    },
+    {
+      "epoch": 0.5703138171045977,
+      "grad_norm": 1.5833889245986938,
+      "learning_rate": 4.523374911266606e-07,
+      "loss": 0.8752,
+      "step": 23680
+    },
+    {
+      "epoch": 0.5704342380963849,
+      "grad_norm": 1.5276377201080322,
+      "learning_rate": 4.522107291349762e-07,
+      "loss": 0.9327,
+      "step": 23685
+    },
+    {
+      "epoch": 0.5705546590881723,
+      "grad_norm": 1.5240648984909058,
+      "learning_rate": 4.520839671432917e-07,
+      "loss": 0.9534,
+      "step": 23690
+    },
+    {
+      "epoch": 0.5706750800799595,
+      "grad_norm": 1.4590954780578613,
+      "learning_rate": 4.5195720515160736e-07,
+      "loss": 0.8412,
+      "step": 23695
+    },
+    {
+      "epoch": 0.5707955010717468,
+      "grad_norm": 1.6685837507247925,
+      "learning_rate": 4.518304431599229e-07,
+      "loss": 0.9154,
+      "step": 23700
+    },
+    {
+      "epoch": 0.5709159220635341,
+      "grad_norm": 1.4996044635772705,
+      "learning_rate": 4.517036811682385e-07,
+      "loss": 0.9211,
+      "step": 23705
+    },
+    {
+      "epoch": 0.5710363430553214,
+      "grad_norm": 1.4348300695419312,
+      "learning_rate": 4.515769191765541e-07,
+      "loss": 0.8831,
+      "step": 23710
+    },
+    {
+      "epoch": 0.5711567640471087,
+      "grad_norm": 1.3542038202285767,
+      "learning_rate": 4.514501571848697e-07,
+      "loss": 0.87,
+      "step": 23715
+    },
+    {
+      "epoch": 0.571277185038896,
+      "grad_norm": 1.7364726066589355,
+      "learning_rate": 4.5132339519318527e-07,
+      "loss": 0.9395,
+      "step": 23720
+    },
+    {
+      "epoch": 0.5713976060306832,
+      "grad_norm": 1.9224735498428345,
+      "learning_rate": 4.5119663320150086e-07,
+      "loss": 0.8701,
+      "step": 23725
+    },
+    {
+      "epoch": 0.5715180270224706,
+      "grad_norm": 1.6598951816558838,
+      "learning_rate": 4.5106987120981645e-07,
+      "loss": 0.8552,
+      "step": 23730
+    },
+    {
+      "epoch": 0.5716384480142579,
+      "grad_norm": 1.522518515586853,
+      "learning_rate": 4.50943109218132e-07,
+      "loss": 0.8725,
+      "step": 23735
+    },
+    {
+      "epoch": 0.5717588690060451,
+      "grad_norm": 1.536367654800415,
+      "learning_rate": 4.5081634722644763e-07,
+      "loss": 0.9576,
+      "step": 23740
+    },
+    {
+      "epoch": 0.5718792899978324,
+      "grad_norm": 1.6370139122009277,
+      "learning_rate": 4.5068958523476317e-07,
+      "loss": 0.8727,
+      "step": 23745
+    },
+    {
+      "epoch": 0.5719997109896197,
+      "grad_norm": 1.4542649984359741,
+      "learning_rate": 4.5056282324307876e-07,
+      "loss": 0.89,
+      "step": 23750
+    },
+    {
+      "epoch": 0.572120131981407,
+      "grad_norm": 1.434571385383606,
+      "learning_rate": 4.504360612513944e-07,
+      "loss": 0.9192,
+      "step": 23755
+    },
+    {
+      "epoch": 0.5722405529731943,
+      "grad_norm": 1.6002957820892334,
+      "learning_rate": 4.5030929925970995e-07,
+      "loss": 0.885,
+      "step": 23760
+    },
+    {
+      "epoch": 0.5723609739649815,
+      "grad_norm": 1.706491470336914,
+      "learning_rate": 4.5018253726802554e-07,
+      "loss": 0.9299,
+      "step": 23765
+    },
+    {
+      "epoch": 0.5724813949567689,
+      "grad_norm": 1.5384788513183594,
+      "learning_rate": 4.5005577527634113e-07,
+      "loss": 0.9399,
+      "step": 23770
+    },
+    {
+      "epoch": 0.5726018159485562,
+      "grad_norm": 1.4597822427749634,
+      "learning_rate": 4.499290132846567e-07,
+      "loss": 0.905,
+      "step": 23775
+    },
+    {
+      "epoch": 0.5727222369403434,
+      "grad_norm": 1.4537782669067383,
+      "learning_rate": 4.4980225129297226e-07,
+      "loss": 0.9231,
+      "step": 23780
+    },
+    {
+      "epoch": 0.5728426579321307,
+      "grad_norm": 1.4657148122787476,
+      "learning_rate": 4.496754893012879e-07,
+      "loss": 0.9407,
+      "step": 23785
+    },
+    {
+      "epoch": 0.5729630789239181,
+      "grad_norm": 1.6405856609344482,
+      "learning_rate": 4.4954872730960345e-07,
+      "loss": 0.8758,
+      "step": 23790
+    },
+    {
+      "epoch": 0.5730834999157053,
+      "grad_norm": 1.7731573581695557,
+      "learning_rate": 4.4942196531791904e-07,
+      "loss": 0.9526,
+      "step": 23795
+    },
+    {
+      "epoch": 0.5732039209074926,
+      "grad_norm": 1.3865348100662231,
+      "learning_rate": 4.492952033262347e-07,
+      "loss": 0.9243,
+      "step": 23800
+    },
+    {
+      "epoch": 0.5733243418992798,
+      "grad_norm": 1.372140645980835,
+      "learning_rate": 4.491684413345502e-07,
+      "loss": 0.8891,
+      "step": 23805
+    },
+    {
+      "epoch": 0.5734447628910672,
+      "grad_norm": 1.4230656623840332,
+      "learning_rate": 4.490416793428658e-07,
+      "loss": 0.8814,
+      "step": 23810
+    },
+    {
+      "epoch": 0.5735651838828545,
+      "grad_norm": 1.5180834531784058,
+      "learning_rate": 4.489149173511814e-07,
+      "loss": 0.9059,
+      "step": 23815
+    },
+    {
+      "epoch": 0.5736856048746417,
+      "grad_norm": 1.4684674739837646,
+      "learning_rate": 4.48788155359497e-07,
+      "loss": 0.9095,
+      "step": 23820
+    },
+    {
+      "epoch": 0.573806025866429,
+      "grad_norm": 1.4561958312988281,
+      "learning_rate": 4.4866139336781254e-07,
+      "loss": 0.9206,
+      "step": 23825
+    },
+    {
+      "epoch": 0.5739264468582164,
+      "grad_norm": 1.6689949035644531,
+      "learning_rate": 4.485346313761282e-07,
+      "loss": 0.9057,
+      "step": 23830
+    },
+    {
+      "epoch": 0.5740468678500036,
+      "grad_norm": 1.494704246520996,
+      "learning_rate": 4.4840786938444377e-07,
+      "loss": 0.8743,
+      "step": 23835
+    },
+    {
+      "epoch": 0.5741672888417909,
+      "grad_norm": 1.4397083520889282,
+      "learning_rate": 4.482811073927593e-07,
+      "loss": 0.9231,
+      "step": 23840
+    },
+    {
+      "epoch": 0.5742877098335782,
+      "grad_norm": 1.6924148797988892,
+      "learning_rate": 4.4815434540107496e-07,
+      "loss": 0.9065,
+      "step": 23845
+    },
+    {
+      "epoch": 0.5744081308253655,
+      "grad_norm": 1.6249243021011353,
+      "learning_rate": 4.480275834093905e-07,
+      "loss": 0.9242,
+      "step": 23850
+    },
+    {
+      "epoch": 0.5745285518171528,
+      "grad_norm": 1.471760869026184,
+      "learning_rate": 4.479008214177061e-07,
+      "loss": 0.865,
+      "step": 23855
+    },
+    {
+      "epoch": 0.57464897280894,
+      "grad_norm": 1.5433714389801025,
+      "learning_rate": 4.477740594260217e-07,
+      "loss": 0.866,
+      "step": 23860
+    },
+    {
+      "epoch": 0.5747693938007273,
+      "grad_norm": 1.5583155155181885,
+      "learning_rate": 4.4764729743433727e-07,
+      "loss": 0.913,
+      "step": 23865
+    },
+    {
+      "epoch": 0.5748898147925147,
+      "grad_norm": 1.5120084285736084,
+      "learning_rate": 4.4752053544265286e-07,
+      "loss": 0.9512,
+      "step": 23870
+    },
+    {
+      "epoch": 0.5750102357843019,
+      "grad_norm": 1.3898836374282837,
+      "learning_rate": 4.4739377345096845e-07,
+      "loss": 0.9161,
+      "step": 23875
+    },
+    {
+      "epoch": 0.5751306567760892,
+      "grad_norm": 1.754967212677002,
+      "learning_rate": 4.4726701145928405e-07,
+      "loss": 0.9644,
+      "step": 23880
+    },
+    {
+      "epoch": 0.5752510777678765,
+      "grad_norm": 1.5690884590148926,
+      "learning_rate": 4.471402494675996e-07,
+      "loss": 0.9398,
+      "step": 23885
+    },
+    {
+      "epoch": 0.5753714987596638,
+      "grad_norm": 1.4521434307098389,
+      "learning_rate": 4.4701348747591523e-07,
+      "loss": 0.9002,
+      "step": 23890
+    },
+    {
+      "epoch": 0.5754919197514511,
+      "grad_norm": 1.5135856866836548,
+      "learning_rate": 4.4688672548423077e-07,
+      "loss": 0.846,
+      "step": 23895
+    },
+    {
+      "epoch": 0.5756123407432384,
+      "grad_norm": 1.5185608863830566,
+      "learning_rate": 4.4675996349254636e-07,
+      "loss": 0.8862,
+      "step": 23900
+    },
+    {
+      "epoch": 0.5757327617350256,
+      "grad_norm": 1.4188989400863647,
+      "learning_rate": 4.46633201500862e-07,
+      "loss": 0.8444,
+      "step": 23905
+    },
+    {
+      "epoch": 0.575853182726813,
+      "grad_norm": 1.6794841289520264,
+      "learning_rate": 4.4650643950917754e-07,
+      "loss": 0.9201,
+      "step": 23910
+    },
+    {
+      "epoch": 0.5759736037186002,
+      "grad_norm": 1.6270520687103271,
+      "learning_rate": 4.4637967751749314e-07,
+      "loss": 0.9318,
+      "step": 23915
+    },
+    {
+      "epoch": 0.5760940247103875,
+      "grad_norm": 1.71327543258667,
+      "learning_rate": 4.4625291552580873e-07,
+      "loss": 0.9029,
+      "step": 23920
+    },
+    {
+      "epoch": 0.5762144457021748,
+      "grad_norm": 1.8402167558670044,
+      "learning_rate": 4.461261535341243e-07,
+      "loss": 0.9023,
+      "step": 23925
+    },
+    {
+      "epoch": 0.5763348666939621,
+      "grad_norm": 1.5386940240859985,
+      "learning_rate": 4.4599939154243986e-07,
+      "loss": 0.8126,
+      "step": 23930
+    },
+    {
+      "epoch": 0.5764552876857494,
+      "grad_norm": 1.6440932750701904,
+      "learning_rate": 4.458726295507555e-07,
+      "loss": 0.8952,
+      "step": 23935
+    },
+    {
+      "epoch": 0.5765757086775367,
+      "grad_norm": 1.5482745170593262,
+      "learning_rate": 4.4574586755907104e-07,
+      "loss": 0.9059,
+      "step": 23940
+    },
+    {
+      "epoch": 0.5766961296693239,
+      "grad_norm": 2.264068603515625,
+      "learning_rate": 4.4561910556738663e-07,
+      "loss": 0.876,
+      "step": 23945
+    },
+    {
+      "epoch": 0.5768165506611113,
+      "grad_norm": 1.9847251176834106,
+      "learning_rate": 4.454923435757023e-07,
+      "loss": 0.933,
+      "step": 23950
+    },
+    {
+      "epoch": 0.5769369716528985,
+      "grad_norm": 1.469585657119751,
+      "learning_rate": 4.453655815840178e-07,
+      "loss": 0.8767,
+      "step": 23955
+    },
+    {
+      "epoch": 0.5770573926446858,
+      "grad_norm": 1.5730228424072266,
+      "learning_rate": 4.452388195923334e-07,
+      "loss": 0.8507,
+      "step": 23960
+    },
+    {
+      "epoch": 0.5771778136364731,
+      "grad_norm": 1.466878890991211,
+      "learning_rate": 4.45112057600649e-07,
+      "loss": 0.8549,
+      "step": 23965
+    },
+    {
+      "epoch": 0.5772982346282604,
+      "grad_norm": 1.378785252571106,
+      "learning_rate": 4.449852956089646e-07,
+      "loss": 0.9119,
+      "step": 23970
+    },
+    {
+      "epoch": 0.5774186556200477,
+      "grad_norm": 1.5784924030303955,
+      "learning_rate": 4.4485853361728013e-07,
+      "loss": 0.912,
+      "step": 23975
+    },
+    {
+      "epoch": 0.577539076611835,
+      "grad_norm": 1.2947007417678833,
+      "learning_rate": 4.447317716255958e-07,
+      "loss": 0.8894,
+      "step": 23980
+    },
+    {
+      "epoch": 0.5776594976036222,
+      "grad_norm": 1.5826587677001953,
+      "learning_rate": 4.4460500963391137e-07,
+      "loss": 0.9001,
+      "step": 23985
+    },
+    {
+      "epoch": 0.5777799185954096,
+      "grad_norm": 1.6320418119430542,
+      "learning_rate": 4.444782476422269e-07,
+      "loss": 0.9695,
+      "step": 23990
+    },
+    {
+      "epoch": 0.5779003395871969,
+      "grad_norm": 1.3234777450561523,
+      "learning_rate": 4.4435148565054255e-07,
+      "loss": 0.931,
+      "step": 23995
+    },
+    {
+      "epoch": 0.5780207605789841,
+      "grad_norm": 1.4822025299072266,
+      "learning_rate": 4.442247236588581e-07,
+      "loss": 0.8501,
+      "step": 24000
+    },
+    {
+      "epoch": 0.5781411815707714,
+      "grad_norm": 1.6573212146759033,
+      "learning_rate": 4.440979616671737e-07,
+      "loss": 0.9256,
+      "step": 24005
+    },
+    {
+      "epoch": 0.5782616025625587,
+      "grad_norm": 1.4438883066177368,
+      "learning_rate": 4.4397119967548927e-07,
+      "loss": 0.8933,
+      "step": 24010
+    },
+    {
+      "epoch": 0.578382023554346,
+      "grad_norm": 1.6382768154144287,
+      "learning_rate": 4.4384443768380487e-07,
+      "loss": 0.91,
+      "step": 24015
+    },
+    {
+      "epoch": 0.5785024445461333,
+      "grad_norm": 1.5180890560150146,
+      "learning_rate": 4.437176756921204e-07,
+      "loss": 0.8616,
+      "step": 24020
+    },
+    {
+      "epoch": 0.5786228655379205,
+      "grad_norm": 1.5098698139190674,
+      "learning_rate": 4.4359091370043605e-07,
+      "loss": 0.8957,
+      "step": 24025
+    },
+    {
+      "epoch": 0.5787432865297079,
+      "grad_norm": 1.3582013845443726,
+      "learning_rate": 4.4346415170875164e-07,
+      "loss": 0.8859,
+      "step": 24030
+    },
+    {
+      "epoch": 0.5788637075214952,
+      "grad_norm": 1.6167140007019043,
+      "learning_rate": 4.433373897170672e-07,
+      "loss": 0.8977,
+      "step": 24035
+    },
+    {
+      "epoch": 0.5789841285132824,
+      "grad_norm": 1.5289353132247925,
+      "learning_rate": 4.432106277253828e-07,
+      "loss": 0.8622,
+      "step": 24040
+    },
+    {
+      "epoch": 0.5791045495050697,
+      "grad_norm": 1.5419551134109497,
+      "learning_rate": 4.4308386573369836e-07,
+      "loss": 0.9116,
+      "step": 24045
+    },
+    {
+      "epoch": 0.5792249704968571,
+      "grad_norm": 1.5560519695281982,
+      "learning_rate": 4.4295710374201395e-07,
+      "loss": 0.9161,
+      "step": 24050
+    },
+    {
+      "epoch": 0.5793453914886443,
+      "grad_norm": 1.4875389337539673,
+      "learning_rate": 4.4283034175032955e-07,
+      "loss": 0.912,
+      "step": 24055
+    },
+    {
+      "epoch": 0.5794658124804316,
+      "grad_norm": 1.446847915649414,
+      "learning_rate": 4.4270357975864514e-07,
+      "loss": 0.8848,
+      "step": 24060
+    },
+    {
+      "epoch": 0.5795862334722188,
+      "grad_norm": 1.881361484527588,
+      "learning_rate": 4.4257681776696073e-07,
+      "loss": 0.9394,
+      "step": 24065
+    },
+    {
+      "epoch": 0.5797066544640062,
+      "grad_norm": 1.6584711074829102,
+      "learning_rate": 4.424500557752763e-07,
+      "loss": 0.9249,
+      "step": 24070
+    },
+    {
+      "epoch": 0.5798270754557935,
+      "grad_norm": 1.4165360927581787,
+      "learning_rate": 4.423232937835919e-07,
+      "loss": 0.9157,
+      "step": 24075
+    },
+    {
+      "epoch": 0.5799474964475807,
+      "grad_norm": 1.5634981393814087,
+      "learning_rate": 4.4219653179190745e-07,
+      "loss": 0.9077,
+      "step": 24080
+    },
+    {
+      "epoch": 0.580067917439368,
+      "grad_norm": 1.435659408569336,
+      "learning_rate": 4.420697698002231e-07,
+      "loss": 0.9554,
+      "step": 24085
+    },
+    {
+      "epoch": 0.5801883384311554,
+      "grad_norm": 1.7094619274139404,
+      "learning_rate": 4.4194300780853864e-07,
+      "loss": 0.8962,
+      "step": 24090
+    },
+    {
+      "epoch": 0.5803087594229426,
+      "grad_norm": 1.600085973739624,
+      "learning_rate": 4.4181624581685423e-07,
+      "loss": 0.9115,
+      "step": 24095
+    },
+    {
+      "epoch": 0.5804291804147299,
+      "grad_norm": 1.5288809537887573,
+      "learning_rate": 4.4168948382516987e-07,
+      "loss": 0.9156,
+      "step": 24100
+    },
+    {
+      "epoch": 0.5805496014065172,
+      "grad_norm": 1.5837684869766235,
+      "learning_rate": 4.415627218334854e-07,
+      "loss": 0.8953,
+      "step": 24105
+    },
+    {
+      "epoch": 0.5806700223983045,
+      "grad_norm": 1.609930396080017,
+      "learning_rate": 4.41435959841801e-07,
+      "loss": 0.9261,
+      "step": 24110
+    },
+    {
+      "epoch": 0.5807904433900918,
+      "grad_norm": 1.6778355836868286,
+      "learning_rate": 4.413091978501166e-07,
+      "loss": 0.9514,
+      "step": 24115
+    },
+    {
+      "epoch": 0.580910864381879,
+      "grad_norm": 1.585976004600525,
+      "learning_rate": 4.411824358584322e-07,
+      "loss": 0.9506,
+      "step": 24120
+    },
+    {
+      "epoch": 0.5810312853736663,
+      "grad_norm": 1.465681552886963,
+      "learning_rate": 4.410556738667478e-07,
+      "loss": 0.8849,
+      "step": 24125
+    },
+    {
+      "epoch": 0.5811517063654537,
+      "grad_norm": 1.6639018058776855,
+      "learning_rate": 4.4092891187506337e-07,
+      "loss": 0.8529,
+      "step": 24130
+    },
+    {
+      "epoch": 0.5812721273572409,
+      "grad_norm": 1.3908381462097168,
+      "learning_rate": 4.4080214988337896e-07,
+      "loss": 0.9167,
+      "step": 24135
+    },
+    {
+      "epoch": 0.5813925483490282,
+      "grad_norm": 1.605887770652771,
+      "learning_rate": 4.4067538789169455e-07,
+      "loss": 0.9068,
+      "step": 24140
+    },
+    {
+      "epoch": 0.5815129693408155,
+      "grad_norm": 1.6921488046646118,
+      "learning_rate": 4.4054862590001015e-07,
+      "loss": 0.9202,
+      "step": 24145
+    },
+    {
+      "epoch": 0.5816333903326028,
+      "grad_norm": 1.6987926959991455,
+      "learning_rate": 4.404218639083257e-07,
+      "loss": 0.8765,
+      "step": 24150
+    },
+    {
+      "epoch": 0.5817538113243901,
+      "grad_norm": 1.5516892671585083,
+      "learning_rate": 4.4029510191664133e-07,
+      "loss": 0.8981,
+      "step": 24155
+    },
+    {
+      "epoch": 0.5818742323161774,
+      "grad_norm": 1.5336050987243652,
+      "learning_rate": 4.4016833992495687e-07,
+      "loss": 0.8836,
+      "step": 24160
+    },
+    {
+      "epoch": 0.5819946533079646,
+      "grad_norm": 1.3349018096923828,
+      "learning_rate": 4.4004157793327246e-07,
+      "loss": 0.8669,
+      "step": 24165
+    },
+    {
+      "epoch": 0.582115074299752,
+      "grad_norm": 2.0041520595550537,
+      "learning_rate": 4.399148159415881e-07,
+      "loss": 0.9013,
+      "step": 24170
+    },
+    {
+      "epoch": 0.5822354952915392,
+      "grad_norm": 1.5998023748397827,
+      "learning_rate": 4.3978805394990364e-07,
+      "loss": 0.9487,
+      "step": 24175
+    },
+    {
+      "epoch": 0.5823559162833265,
+      "grad_norm": 2.7034430503845215,
+      "learning_rate": 4.3966129195821924e-07,
+      "loss": 0.8808,
+      "step": 24180
+    },
+    {
+      "epoch": 0.5824763372751138,
+      "grad_norm": 1.5282458066940308,
+      "learning_rate": 4.3953452996653483e-07,
+      "loss": 0.8657,
+      "step": 24185
+    },
+    {
+      "epoch": 0.582596758266901,
+      "grad_norm": 1.5360181331634521,
+      "learning_rate": 4.394077679748504e-07,
+      "loss": 0.8707,
+      "step": 24190
+    },
+    {
+      "epoch": 0.5827171792586884,
+      "grad_norm": 1.6115931272506714,
+      "learning_rate": 4.3928100598316596e-07,
+      "loss": 0.8551,
+      "step": 24195
+    },
+    {
+      "epoch": 0.5828376002504757,
+      "grad_norm": 1.478049635887146,
+      "learning_rate": 4.391542439914816e-07,
+      "loss": 0.8887,
+      "step": 24200
+    },
+    {
+      "epoch": 0.5829580212422629,
+      "grad_norm": 1.7876139879226685,
+      "learning_rate": 4.3902748199979714e-07,
+      "loss": 0.9005,
+      "step": 24205
+    },
+    {
+      "epoch": 0.5830784422340503,
+      "grad_norm": 1.5490608215332031,
+      "learning_rate": 4.3890072000811273e-07,
+      "loss": 0.9169,
+      "step": 24210
+    },
+    {
+      "epoch": 0.5831988632258376,
+      "grad_norm": 1.4674124717712402,
+      "learning_rate": 4.387739580164284e-07,
+      "loss": 0.9054,
+      "step": 24215
+    },
+    {
+      "epoch": 0.5833192842176248,
+      "grad_norm": 1.3796958923339844,
+      "learning_rate": 4.386471960247439e-07,
+      "loss": 0.8739,
+      "step": 24220
+    },
+    {
+      "epoch": 0.5834397052094121,
+      "grad_norm": 1.6711292266845703,
+      "learning_rate": 4.385204340330595e-07,
+      "loss": 0.8855,
+      "step": 24225
+    },
+    {
+      "epoch": 0.5835601262011993,
+      "grad_norm": 1.5073076486587524,
+      "learning_rate": 4.383936720413751e-07,
+      "loss": 0.9541,
+      "step": 24230
+    },
+    {
+      "epoch": 0.5836805471929867,
+      "grad_norm": 1.5467579364776611,
+      "learning_rate": 4.382669100496907e-07,
+      "loss": 0.9097,
+      "step": 24235
+    },
+    {
+      "epoch": 0.583800968184774,
+      "grad_norm": 1.4945591688156128,
+      "learning_rate": 4.3814014805800623e-07,
+      "loss": 0.8685,
+      "step": 24240
+    },
+    {
+      "epoch": 0.5839213891765612,
+      "grad_norm": 1.4106559753417969,
+      "learning_rate": 4.380133860663219e-07,
+      "loss": 0.8636,
+      "step": 24245
+    },
+    {
+      "epoch": 0.5840418101683486,
+      "grad_norm": 1.6017537117004395,
+      "learning_rate": 4.3788662407463747e-07,
+      "loss": 0.9221,
+      "step": 24250
+    },
+    {
+      "epoch": 0.5841622311601359,
+      "grad_norm": 1.5603747367858887,
+      "learning_rate": 4.37759862082953e-07,
+      "loss": 0.8357,
+      "step": 24255
+    },
+    {
+      "epoch": 0.5842826521519231,
+      "grad_norm": 1.6029847860336304,
+      "learning_rate": 4.3763310009126865e-07,
+      "loss": 0.9294,
+      "step": 24260
+    },
+    {
+      "epoch": 0.5844030731437104,
+      "grad_norm": 1.52727472782135,
+      "learning_rate": 4.375063380995842e-07,
+      "loss": 0.8984,
+      "step": 24265
+    },
+    {
+      "epoch": 0.5845234941354976,
+      "grad_norm": 1.6402517557144165,
+      "learning_rate": 4.373795761078998e-07,
+      "loss": 0.9071,
+      "step": 24270
+    },
+    {
+      "epoch": 0.584643915127285,
+      "grad_norm": 1.5726128816604614,
+      "learning_rate": 4.372528141162154e-07,
+      "loss": 0.8772,
+      "step": 24275
+    },
+    {
+      "epoch": 0.5847643361190723,
+      "grad_norm": 1.6014689207077026,
+      "learning_rate": 4.3712605212453097e-07,
+      "loss": 0.9104,
+      "step": 24280
+    },
+    {
+      "epoch": 0.5848847571108595,
+      "grad_norm": 1.4787907600402832,
+      "learning_rate": 4.3699929013284656e-07,
+      "loss": 0.8989,
+      "step": 24285
+    },
+    {
+      "epoch": 0.5850051781026469,
+      "grad_norm": 1.6082152128219604,
+      "learning_rate": 4.3687252814116215e-07,
+      "loss": 0.9106,
+      "step": 24290
+    },
+    {
+      "epoch": 0.5851255990944342,
+      "grad_norm": 1.7320561408996582,
+      "learning_rate": 4.3674576614947774e-07,
+      "loss": 0.9092,
+      "step": 24295
+    },
+    {
+      "epoch": 0.5852460200862214,
+      "grad_norm": 1.379388689994812,
+      "learning_rate": 4.366190041577933e-07,
+      "loss": 0.9151,
+      "step": 24300
+    },
+    {
+      "epoch": 0.5853664410780087,
+      "grad_norm": 1.8676650524139404,
+      "learning_rate": 4.364922421661089e-07,
+      "loss": 0.9412,
+      "step": 24305
+    },
+    {
+      "epoch": 0.585486862069796,
+      "grad_norm": 1.5080041885375977,
+      "learning_rate": 4.3636548017442446e-07,
+      "loss": 0.9029,
+      "step": 24310
+    },
+    {
+      "epoch": 0.5856072830615833,
+      "grad_norm": 1.5643041133880615,
+      "learning_rate": 4.3623871818274006e-07,
+      "loss": 0.9524,
+      "step": 24315
+    },
+    {
+      "epoch": 0.5857277040533706,
+      "grad_norm": 1.6688721179962158,
+      "learning_rate": 4.361119561910557e-07,
+      "loss": 0.8733,
+      "step": 24320
+    },
+    {
+      "epoch": 0.5858481250451578,
+      "grad_norm": 1.5123419761657715,
+      "learning_rate": 4.3598519419937124e-07,
+      "loss": 0.8769,
+      "step": 24325
+    },
+    {
+      "epoch": 0.5859685460369451,
+      "grad_norm": 1.4766058921813965,
+      "learning_rate": 4.3585843220768683e-07,
+      "loss": 0.9166,
+      "step": 24330
+    },
+    {
+      "epoch": 0.5860889670287325,
+      "grad_norm": 1.4903916120529175,
+      "learning_rate": 4.357316702160024e-07,
+      "loss": 0.8974,
+      "step": 24335
+    },
+    {
+      "epoch": 0.5862093880205197,
+      "grad_norm": 1.6494252681732178,
+      "learning_rate": 4.35604908224318e-07,
+      "loss": 0.9333,
+      "step": 24340
+    },
+    {
+      "epoch": 0.586329809012307,
+      "grad_norm": 1.5249848365783691,
+      "learning_rate": 4.3547814623263355e-07,
+      "loss": 0.9179,
+      "step": 24345
+    },
+    {
+      "epoch": 0.5864502300040944,
+      "grad_norm": 1.384194016456604,
+      "learning_rate": 4.353513842409492e-07,
+      "loss": 0.9111,
+      "step": 24350
+    },
+    {
+      "epoch": 0.5865706509958816,
+      "grad_norm": 1.5499690771102905,
+      "learning_rate": 4.3522462224926474e-07,
+      "loss": 0.9198,
+      "step": 24355
+    },
+    {
+      "epoch": 0.5866910719876689,
+      "grad_norm": 1.390620470046997,
+      "learning_rate": 4.3509786025758033e-07,
+      "loss": 0.8384,
+      "step": 24360
+    },
+    {
+      "epoch": 0.5868114929794562,
+      "grad_norm": 1.4609301090240479,
+      "learning_rate": 4.3497109826589597e-07,
+      "loss": 0.8811,
+      "step": 24365
+    },
+    {
+      "epoch": 0.5869319139712434,
+      "grad_norm": 1.515616774559021,
+      "learning_rate": 4.348443362742115e-07,
+      "loss": 0.8662,
+      "step": 24370
+    },
+    {
+      "epoch": 0.5870523349630308,
+      "grad_norm": 1.6101609468460083,
+      "learning_rate": 4.347175742825271e-07,
+      "loss": 0.9414,
+      "step": 24375
+    },
+    {
+      "epoch": 0.587172755954818,
+      "grad_norm": 1.5424721240997314,
+      "learning_rate": 4.345908122908427e-07,
+      "loss": 0.9024,
+      "step": 24380
+    },
+    {
+      "epoch": 0.5872931769466053,
+      "grad_norm": 1.7643179893493652,
+      "learning_rate": 4.344640502991583e-07,
+      "loss": 0.8595,
+      "step": 24385
+    },
+    {
+      "epoch": 0.5874135979383927,
+      "grad_norm": 1.6247128248214722,
+      "learning_rate": 4.343372883074738e-07,
+      "loss": 0.9067,
+      "step": 24390
+    },
+    {
+      "epoch": 0.5875340189301799,
+      "grad_norm": 1.526525855064392,
+      "learning_rate": 4.3421052631578947e-07,
+      "loss": 0.9579,
+      "step": 24395
+    },
+    {
+      "epoch": 0.5876544399219672,
+      "grad_norm": 1.9331629276275635,
+      "learning_rate": 4.3408376432410506e-07,
+      "loss": 0.8746,
+      "step": 24400
+    },
+    {
+      "epoch": 0.5877748609137545,
+      "grad_norm": 1.432662844657898,
+      "learning_rate": 4.339570023324206e-07,
+      "loss": 0.9269,
+      "step": 24405
+    },
+    {
+      "epoch": 0.5878952819055417,
+      "grad_norm": 1.6652055978775024,
+      "learning_rate": 4.3383024034073625e-07,
+      "loss": 0.922,
+      "step": 24410
+    },
+    {
+      "epoch": 0.5880157028973291,
+      "grad_norm": 1.549464225769043,
+      "learning_rate": 4.337034783490518e-07,
+      "loss": 0.9436,
+      "step": 24415
+    },
+    {
+      "epoch": 0.5881361238891164,
+      "grad_norm": 1.5567026138305664,
+      "learning_rate": 4.335767163573674e-07,
+      "loss": 0.8839,
+      "step": 24420
+    },
+    {
+      "epoch": 0.5882565448809036,
+      "grad_norm": 1.5347050428390503,
+      "learning_rate": 4.3344995436568297e-07,
+      "loss": 0.8827,
+      "step": 24425
+    },
+    {
+      "epoch": 0.588376965872691,
+      "grad_norm": 1.6135860681533813,
+      "learning_rate": 4.3332319237399856e-07,
+      "loss": 0.8946,
+      "step": 24430
+    },
+    {
+      "epoch": 0.5884973868644782,
+      "grad_norm": 1.5451579093933105,
+      "learning_rate": 4.3319643038231415e-07,
+      "loss": 0.9577,
+      "step": 24435
+    },
+    {
+      "epoch": 0.5886178078562655,
+      "grad_norm": 1.5067342519760132,
+      "learning_rate": 4.3306966839062974e-07,
+      "loss": 0.8841,
+      "step": 24440
+    },
+    {
+      "epoch": 0.5887382288480528,
+      "grad_norm": 1.3985469341278076,
+      "learning_rate": 4.3294290639894534e-07,
+      "loss": 0.8999,
+      "step": 24445
+    },
+    {
+      "epoch": 0.58885864983984,
+      "grad_norm": 1.6078155040740967,
+      "learning_rate": 4.328161444072609e-07,
+      "loss": 0.939,
+      "step": 24450
+    },
+    {
+      "epoch": 0.5889790708316274,
+      "grad_norm": 1.5124714374542236,
+      "learning_rate": 4.326893824155765e-07,
+      "loss": 0.9505,
+      "step": 24455
+    },
+    {
+      "epoch": 0.5890994918234147,
+      "grad_norm": 1.4121074676513672,
+      "learning_rate": 4.3256262042389206e-07,
+      "loss": 0.9179,
+      "step": 24460
+    },
+    {
+      "epoch": 0.5892199128152019,
+      "grad_norm": 1.6301769018173218,
+      "learning_rate": 4.3243585843220765e-07,
+      "loss": 0.9344,
+      "step": 24465
+    },
+    {
+      "epoch": 0.5893403338069892,
+      "grad_norm": 1.4103622436523438,
+      "learning_rate": 4.323090964405233e-07,
+      "loss": 0.9356,
+      "step": 24470
+    },
+    {
+      "epoch": 0.5894607547987766,
+      "grad_norm": 1.471883773803711,
+      "learning_rate": 4.3218233444883883e-07,
+      "loss": 0.9248,
+      "step": 24475
+    },
+    {
+      "epoch": 0.5895811757905638,
+      "grad_norm": 1.5216230154037476,
+      "learning_rate": 4.320555724571544e-07,
+      "loss": 0.8915,
+      "step": 24480
+    },
+    {
+      "epoch": 0.5897015967823511,
+      "grad_norm": 1.7749937772750854,
+      "learning_rate": 4.3192881046547e-07,
+      "loss": 0.8772,
+      "step": 24485
+    },
+    {
+      "epoch": 0.5898220177741383,
+      "grad_norm": 1.7109736204147339,
+      "learning_rate": 4.318020484737856e-07,
+      "loss": 0.9155,
+      "step": 24490
+    },
+    {
+      "epoch": 0.5899424387659257,
+      "grad_norm": 1.5612523555755615,
+      "learning_rate": 4.3167528648210115e-07,
+      "loss": 0.9747,
+      "step": 24495
+    },
+    {
+      "epoch": 0.590062859757713,
+      "grad_norm": 1.8180674314498901,
+      "learning_rate": 4.315485244904168e-07,
+      "loss": 0.9173,
+      "step": 24500
+    },
+    {
+      "epoch": 0.5901832807495002,
+      "grad_norm": 1.5764421224594116,
+      "learning_rate": 4.3142176249873233e-07,
+      "loss": 0.9285,
+      "step": 24505
+    },
+    {
+      "epoch": 0.5903037017412875,
+      "grad_norm": 1.7687420845031738,
+      "learning_rate": 4.312950005070479e-07,
+      "loss": 0.9144,
+      "step": 24510
+    },
+    {
+      "epoch": 0.5904241227330749,
+      "grad_norm": 1.494286298751831,
+      "learning_rate": 4.3116823851536357e-07,
+      "loss": 0.8958,
+      "step": 24515
+    },
+    {
+      "epoch": 0.5905445437248621,
+      "grad_norm": 1.4615665674209595,
+      "learning_rate": 4.310414765236791e-07,
+      "loss": 0.889,
+      "step": 24520
+    },
+    {
+      "epoch": 0.5906649647166494,
+      "grad_norm": 1.6046537160873413,
+      "learning_rate": 4.309147145319947e-07,
+      "loss": 0.9077,
+      "step": 24525
+    },
+    {
+      "epoch": 0.5907853857084367,
+      "grad_norm": 1.524892807006836,
+      "learning_rate": 4.307879525403103e-07,
+      "loss": 0.91,
+      "step": 24530
+    },
+    {
+      "epoch": 0.590905806700224,
+      "grad_norm": 1.6162455081939697,
+      "learning_rate": 4.306611905486259e-07,
+      "loss": 0.9017,
+      "step": 24535
+    },
+    {
+      "epoch": 0.5910262276920113,
+      "grad_norm": 1.5659140348434448,
+      "learning_rate": 4.305344285569414e-07,
+      "loss": 0.9378,
+      "step": 24540
+    },
+    {
+      "epoch": 0.5911466486837985,
+      "grad_norm": 1.538069248199463,
+      "learning_rate": 4.3040766656525707e-07,
+      "loss": 0.9146,
+      "step": 24545
+    },
+    {
+      "epoch": 0.5912670696755858,
+      "grad_norm": 1.7028398513793945,
+      "learning_rate": 4.3028090457357266e-07,
+      "loss": 0.9234,
+      "step": 24550
+    },
+    {
+      "epoch": 0.5913874906673732,
+      "grad_norm": 1.4257739782333374,
+      "learning_rate": 4.301541425818882e-07,
+      "loss": 0.9103,
+      "step": 24555
+    },
+    {
+      "epoch": 0.5915079116591604,
+      "grad_norm": 1.5835907459259033,
+      "learning_rate": 4.3002738059020384e-07,
+      "loss": 0.9112,
+      "step": 24560
+    },
+    {
+      "epoch": 0.5916283326509477,
+      "grad_norm": 1.4983320236206055,
+      "learning_rate": 4.299006185985194e-07,
+      "loss": 0.859,
+      "step": 24565
+    },
+    {
+      "epoch": 0.591748753642735,
+      "grad_norm": 1.4904733896255493,
+      "learning_rate": 4.2977385660683497e-07,
+      "loss": 0.9099,
+      "step": 24570
+    },
+    {
+      "epoch": 0.5918691746345223,
+      "grad_norm": 1.565053105354309,
+      "learning_rate": 4.2964709461515056e-07,
+      "loss": 0.9727,
+      "step": 24575
+    },
+    {
+      "epoch": 0.5919895956263096,
+      "grad_norm": 1.3869820833206177,
+      "learning_rate": 4.2952033262346616e-07,
+      "loss": 0.8663,
+      "step": 24580
+    },
+    {
+      "epoch": 0.5921100166180968,
+      "grad_norm": 1.8419201374053955,
+      "learning_rate": 4.2939357063178175e-07,
+      "loss": 0.9035,
+      "step": 24585
+    },
+    {
+      "epoch": 0.5922304376098841,
+      "grad_norm": 1.458988904953003,
+      "learning_rate": 4.2926680864009734e-07,
+      "loss": 0.9035,
+      "step": 24590
+    },
+    {
+      "epoch": 0.5923508586016715,
+      "grad_norm": 1.638648509979248,
+      "learning_rate": 4.2914004664841293e-07,
+      "loss": 0.8944,
+      "step": 24595
+    },
+    {
+      "epoch": 0.5924712795934587,
+      "grad_norm": 1.8352802991867065,
+      "learning_rate": 4.2901328465672847e-07,
+      "loss": 0.8325,
+      "step": 24600
+    },
+    {
+      "epoch": 0.592591700585246,
+      "grad_norm": 1.7183581590652466,
+      "learning_rate": 4.288865226650441e-07,
+      "loss": 0.9469,
+      "step": 24605
+    },
+    {
+      "epoch": 0.5927121215770333,
+      "grad_norm": 1.4810676574707031,
+      "learning_rate": 4.2875976067335965e-07,
+      "loss": 0.8973,
+      "step": 24610
+    },
+    {
+      "epoch": 0.5928325425688206,
+      "grad_norm": 1.4876949787139893,
+      "learning_rate": 4.2863299868167524e-07,
+      "loss": 0.8831,
+      "step": 24615
+    },
+    {
+      "epoch": 0.5929529635606079,
+      "grad_norm": 1.4875684976577759,
+      "learning_rate": 4.2850623668999084e-07,
+      "loss": 0.9086,
+      "step": 24620
+    },
+    {
+      "epoch": 0.5930733845523952,
+      "grad_norm": 1.4098703861236572,
+      "learning_rate": 4.2837947469830643e-07,
+      "loss": 0.9037,
+      "step": 24625
+    },
+    {
+      "epoch": 0.5931938055441824,
+      "grad_norm": 1.6200228929519653,
+      "learning_rate": 4.2825271270662207e-07,
+      "loss": 0.8808,
+      "step": 24630
+    },
+    {
+      "epoch": 0.5933142265359698,
+      "grad_norm": 1.5252360105514526,
+      "learning_rate": 4.281259507149376e-07,
+      "loss": 0.9197,
+      "step": 24635
+    },
+    {
+      "epoch": 0.593434647527757,
+      "grad_norm": 1.503819465637207,
+      "learning_rate": 4.279991887232532e-07,
+      "loss": 0.9064,
+      "step": 24640
+    },
+    {
+      "epoch": 0.5935550685195443,
+      "grad_norm": 1.649817705154419,
+      "learning_rate": 4.278724267315688e-07,
+      "loss": 0.9156,
+      "step": 24645
+    },
+    {
+      "epoch": 0.5936754895113316,
+      "grad_norm": 1.7362719774246216,
+      "learning_rate": 4.277456647398844e-07,
+      "loss": 0.9057,
+      "step": 24650
+    },
+    {
+      "epoch": 0.5937959105031189,
+      "grad_norm": 1.6223540306091309,
+      "learning_rate": 4.276189027481999e-07,
+      "loss": 0.9153,
+      "step": 24655
+    },
+    {
+      "epoch": 0.5939163314949062,
+      "grad_norm": 1.979633092880249,
+      "learning_rate": 4.2749214075651557e-07,
+      "loss": 0.92,
+      "step": 24660
+    },
+    {
+      "epoch": 0.5940367524866935,
+      "grad_norm": 1.5584977865219116,
+      "learning_rate": 4.2736537876483116e-07,
+      "loss": 0.9124,
+      "step": 24665
+    },
+    {
+      "epoch": 0.5941571734784807,
+      "grad_norm": 1.7202004194259644,
+      "learning_rate": 4.272386167731467e-07,
+      "loss": 0.8668,
+      "step": 24670
+    },
+    {
+      "epoch": 0.5942775944702681,
+      "grad_norm": 1.3186968564987183,
+      "learning_rate": 4.2711185478146235e-07,
+      "loss": 0.895,
+      "step": 24675
+    },
+    {
+      "epoch": 0.5943980154620554,
+      "grad_norm": 1.4132670164108276,
+      "learning_rate": 4.269850927897779e-07,
+      "loss": 0.8781,
+      "step": 24680
+    },
+    {
+      "epoch": 0.5945184364538426,
+      "grad_norm": 1.6592215299606323,
+      "learning_rate": 4.268583307980935e-07,
+      "loss": 0.9732,
+      "step": 24685
+    },
+    {
+      "epoch": 0.5946388574456299,
+      "grad_norm": 1.5422641038894653,
+      "learning_rate": 4.2673156880640907e-07,
+      "loss": 0.8949,
+      "step": 24690
+    },
+    {
+      "epoch": 0.5947592784374172,
+      "grad_norm": 1.4672731161117554,
+      "learning_rate": 4.2660480681472466e-07,
+      "loss": 0.8768,
+      "step": 24695
+    },
+    {
+      "epoch": 0.5948796994292045,
+      "grad_norm": 1.4263854026794434,
+      "learning_rate": 4.2647804482304025e-07,
+      "loss": 0.8618,
+      "step": 24700
+    },
+    {
+      "epoch": 0.5950001204209918,
+      "grad_norm": 1.4597587585449219,
+      "learning_rate": 4.2635128283135584e-07,
+      "loss": 0.9404,
+      "step": 24705
+    },
+    {
+      "epoch": 0.595120541412779,
+      "grad_norm": 1.6408240795135498,
+      "learning_rate": 4.2622452083967144e-07,
+      "loss": 0.9067,
+      "step": 24710
+    },
+    {
+      "epoch": 0.5952409624045664,
+      "grad_norm": 1.608472228050232,
+      "learning_rate": 4.26097758847987e-07,
+      "loss": 0.8881,
+      "step": 24715
+    },
+    {
+      "epoch": 0.5953613833963537,
+      "grad_norm": 1.5572516918182373,
+      "learning_rate": 4.259709968563026e-07,
+      "loss": 0.9285,
+      "step": 24720
+    },
+    {
+      "epoch": 0.5954818043881409,
+      "grad_norm": 1.5052424669265747,
+      "learning_rate": 4.2584423486461816e-07,
+      "loss": 0.8771,
+      "step": 24725
+    },
+    {
+      "epoch": 0.5956022253799282,
+      "grad_norm": 1.4403895139694214,
+      "learning_rate": 4.2571747287293375e-07,
+      "loss": 0.9194,
+      "step": 24730
+    },
+    {
+      "epoch": 0.5957226463717156,
+      "grad_norm": 1.5183978080749512,
+      "learning_rate": 4.255907108812494e-07,
+      "loss": 0.8902,
+      "step": 24735
+    },
+    {
+      "epoch": 0.5958430673635028,
+      "grad_norm": 1.5620477199554443,
+      "learning_rate": 4.2546394888956493e-07,
+      "loss": 0.8974,
+      "step": 24740
+    },
+    {
+      "epoch": 0.5959634883552901,
+      "grad_norm": 1.5473697185516357,
+      "learning_rate": 4.253371868978805e-07,
+      "loss": 0.9218,
+      "step": 24745
+    },
+    {
+      "epoch": 0.5960839093470773,
+      "grad_norm": 1.8315085172653198,
+      "learning_rate": 4.252104249061961e-07,
+      "loss": 0.8657,
+      "step": 24750
+    },
+    {
+      "epoch": 0.5962043303388647,
+      "grad_norm": 1.4640153646469116,
+      "learning_rate": 4.250836629145117e-07,
+      "loss": 0.8704,
+      "step": 24755
+    },
+    {
+      "epoch": 0.596324751330652,
+      "grad_norm": 1.628290057182312,
+      "learning_rate": 4.2495690092282725e-07,
+      "loss": 0.9884,
+      "step": 24760
+    },
+    {
+      "epoch": 0.5964451723224392,
+      "grad_norm": 1.4359389543533325,
+      "learning_rate": 4.248301389311429e-07,
+      "loss": 0.8803,
+      "step": 24765
+    },
+    {
+      "epoch": 0.5965655933142265,
+      "grad_norm": 1.5545084476470947,
+      "learning_rate": 4.2470337693945843e-07,
+      "loss": 0.8864,
+      "step": 24770
+    },
+    {
+      "epoch": 0.5966860143060139,
+      "grad_norm": 1.6062496900558472,
+      "learning_rate": 4.24576614947774e-07,
+      "loss": 0.9378,
+      "step": 24775
+    },
+    {
+      "epoch": 0.5968064352978011,
+      "grad_norm": 1.5714846849441528,
+      "learning_rate": 4.2444985295608967e-07,
+      "loss": 0.9166,
+      "step": 24780
+    },
+    {
+      "epoch": 0.5969268562895884,
+      "grad_norm": 1.3731664419174194,
+      "learning_rate": 4.243230909644052e-07,
+      "loss": 0.9121,
+      "step": 24785
+    },
+    {
+      "epoch": 0.5970472772813757,
+      "grad_norm": 1.2442537546157837,
+      "learning_rate": 4.241963289727208e-07,
+      "loss": 0.8615,
+      "step": 24790
+    },
+    {
+      "epoch": 0.597167698273163,
+      "grad_norm": 1.513586163520813,
+      "learning_rate": 4.240695669810364e-07,
+      "loss": 0.8859,
+      "step": 24795
+    },
+    {
+      "epoch": 0.5972881192649503,
+      "grad_norm": 1.4235986471176147,
+      "learning_rate": 4.23942804989352e-07,
+      "loss": 0.8807,
+      "step": 24800
+    },
+    {
+      "epoch": 0.5974085402567375,
+      "grad_norm": 1.383156657218933,
+      "learning_rate": 4.238160429976675e-07,
+      "loss": 0.911,
+      "step": 24805
+    },
+    {
+      "epoch": 0.5975289612485248,
+      "grad_norm": 1.4993770122528076,
+      "learning_rate": 4.2368928100598317e-07,
+      "loss": 0.8858,
+      "step": 24810
+    },
+    {
+      "epoch": 0.5976493822403122,
+      "grad_norm": 1.6158260107040405,
+      "learning_rate": 4.2356251901429876e-07,
+      "loss": 0.9184,
+      "step": 24815
+    },
+    {
+      "epoch": 0.5977698032320994,
+      "grad_norm": 1.532894253730774,
+      "learning_rate": 4.234357570226143e-07,
+      "loss": 0.9393,
+      "step": 24820
+    },
+    {
+      "epoch": 0.5978902242238867,
+      "grad_norm": 1.5125641822814941,
+      "learning_rate": 4.2330899503092994e-07,
+      "loss": 0.9201,
+      "step": 24825
+    },
+    {
+      "epoch": 0.598010645215674,
+      "grad_norm": 1.6651484966278076,
+      "learning_rate": 4.231822330392455e-07,
+      "loss": 0.8663,
+      "step": 24830
+    },
+    {
+      "epoch": 0.5981310662074613,
+      "grad_norm": 1.5242916345596313,
+      "learning_rate": 4.2305547104756107e-07,
+      "loss": 0.9125,
+      "step": 24835
+    },
+    {
+      "epoch": 0.5982514871992486,
+      "grad_norm": 1.4305624961853027,
+      "learning_rate": 4.2292870905587666e-07,
+      "loss": 0.926,
+      "step": 24840
+    },
+    {
+      "epoch": 0.5983719081910358,
+      "grad_norm": 1.5143855810165405,
+      "learning_rate": 4.2280194706419226e-07,
+      "loss": 0.9227,
+      "step": 24845
+    },
+    {
+      "epoch": 0.5984923291828231,
+      "grad_norm": 1.9623805284500122,
+      "learning_rate": 4.2267518507250785e-07,
+      "loss": 0.9232,
+      "step": 24850
+    },
+    {
+      "epoch": 0.5986127501746105,
+      "grad_norm": 1.474603295326233,
+      "learning_rate": 4.2254842308082344e-07,
+      "loss": 0.9077,
+      "step": 24855
+    },
+    {
+      "epoch": 0.5987331711663977,
+      "grad_norm": 1.7623028755187988,
+      "learning_rate": 4.2242166108913903e-07,
+      "loss": 0.8742,
+      "step": 24860
+    },
+    {
+      "epoch": 0.598853592158185,
+      "grad_norm": 1.4490970373153687,
+      "learning_rate": 4.2229489909745457e-07,
+      "loss": 0.9129,
+      "step": 24865
+    },
+    {
+      "epoch": 0.5989740131499723,
+      "grad_norm": 1.5933712720870972,
+      "learning_rate": 4.221681371057702e-07,
+      "loss": 0.892,
+      "step": 24870
+    },
+    {
+      "epoch": 0.5990944341417596,
+      "grad_norm": 1.4571937322616577,
+      "learning_rate": 4.2204137511408575e-07,
+      "loss": 0.898,
+      "step": 24875
+    },
+    {
+      "epoch": 0.5992148551335469,
+      "grad_norm": 1.9972020387649536,
+      "learning_rate": 4.2191461312240135e-07,
+      "loss": 0.8579,
+      "step": 24880
+    },
+    {
+      "epoch": 0.5993352761253342,
+      "grad_norm": 1.5374258756637573,
+      "learning_rate": 4.21787851130717e-07,
+      "loss": 0.8947,
+      "step": 24885
+    },
+    {
+      "epoch": 0.5994556971171214,
+      "grad_norm": 1.6821030378341675,
+      "learning_rate": 4.2166108913903253e-07,
+      "loss": 0.965,
+      "step": 24890
+    },
+    {
+      "epoch": 0.5995761181089088,
+      "grad_norm": 1.6769107580184937,
+      "learning_rate": 4.215343271473481e-07,
+      "loss": 0.8754,
+      "step": 24895
+    },
+    {
+      "epoch": 0.599696539100696,
+      "grad_norm": 1.5854194164276123,
+      "learning_rate": 4.214075651556637e-07,
+      "loss": 0.9242,
+      "step": 24900
+    },
+    {
+      "epoch": 0.5998169600924833,
+      "grad_norm": 1.4867968559265137,
+      "learning_rate": 4.212808031639793e-07,
+      "loss": 0.9153,
+      "step": 24905
+    },
+    {
+      "epoch": 0.5999373810842706,
+      "grad_norm": 1.6546216011047363,
+      "learning_rate": 4.2115404117229484e-07,
+      "loss": 0.8559,
+      "step": 24910
+    },
+    {
+      "epoch": 0.6000578020760579,
+      "grad_norm": 1.4881573915481567,
+      "learning_rate": 4.210272791806105e-07,
+      "loss": 0.9238,
+      "step": 24915
+    },
+    {
+      "epoch": 0.6001782230678452,
+      "grad_norm": 1.7282466888427734,
+      "learning_rate": 4.2090051718892603e-07,
+      "loss": 0.9407,
+      "step": 24920
+    },
+    {
+      "epoch": 0.6002986440596325,
+      "grad_norm": 1.6105860471725464,
+      "learning_rate": 4.207737551972416e-07,
+      "loss": 0.9407,
+      "step": 24925
+    },
+    {
+      "epoch": 0.6004190650514197,
+      "grad_norm": 1.537200689315796,
+      "learning_rate": 4.2064699320555726e-07,
+      "loss": 0.8998,
+      "step": 24930
+    },
+    {
+      "epoch": 0.6005394860432071,
+      "grad_norm": 1.5517336130142212,
+      "learning_rate": 4.205202312138728e-07,
+      "loss": 0.9177,
+      "step": 24935
+    },
+    {
+      "epoch": 0.6006599070349944,
+      "grad_norm": 1.5445387363433838,
+      "learning_rate": 4.203934692221884e-07,
+      "loss": 0.9164,
+      "step": 24940
+    },
+    {
+      "epoch": 0.6007803280267816,
+      "grad_norm": 1.5406125783920288,
+      "learning_rate": 4.20266707230504e-07,
+      "loss": 0.9197,
+      "step": 24945
+    },
+    {
+      "epoch": 0.6009007490185689,
+      "grad_norm": 1.7499818801879883,
+      "learning_rate": 4.201399452388196e-07,
+      "loss": 0.8886,
+      "step": 24950
+    },
+    {
+      "epoch": 0.6010211700103562,
+      "grad_norm": 1.4955873489379883,
+      "learning_rate": 4.200131832471351e-07,
+      "loss": 0.9468,
+      "step": 24955
+    },
+    {
+      "epoch": 0.6011415910021435,
+      "grad_norm": 1.6411209106445312,
+      "learning_rate": 4.1988642125545076e-07,
+      "loss": 0.8876,
+      "step": 24960
+    },
+    {
+      "epoch": 0.6012620119939308,
+      "grad_norm": 1.583150863647461,
+      "learning_rate": 4.1975965926376635e-07,
+      "loss": 0.8628,
+      "step": 24965
+    },
+    {
+      "epoch": 0.601382432985718,
+      "grad_norm": 1.4690210819244385,
+      "learning_rate": 4.196328972720819e-07,
+      "loss": 0.9331,
+      "step": 24970
+    },
+    {
+      "epoch": 0.6015028539775054,
+      "grad_norm": 1.6037509441375732,
+      "learning_rate": 4.1950613528039754e-07,
+      "loss": 0.9168,
+      "step": 24975
+    },
+    {
+      "epoch": 0.6016232749692927,
+      "grad_norm": 1.454247236251831,
+      "learning_rate": 4.193793732887131e-07,
+      "loss": 0.9042,
+      "step": 24980
+    },
+    {
+      "epoch": 0.6017436959610799,
+      "grad_norm": 1.6040294170379639,
+      "learning_rate": 4.1925261129702867e-07,
+      "loss": 0.8736,
+      "step": 24985
+    },
+    {
+      "epoch": 0.6018641169528672,
+      "grad_norm": 1.4939274787902832,
+      "learning_rate": 4.1912584930534426e-07,
+      "loss": 0.8957,
+      "step": 24990
+    },
+    {
+      "epoch": 0.6019845379446546,
+      "grad_norm": 1.7517776489257812,
+      "learning_rate": 4.1899908731365985e-07,
+      "loss": 0.9518,
+      "step": 24995
+    },
+    {
+      "epoch": 0.6021049589364418,
+      "grad_norm": 1.6388598680496216,
+      "learning_rate": 4.1887232532197544e-07,
+      "loss": 0.8468,
+      "step": 25000
+    },
+    {
+      "epoch": 0.6022253799282291,
+      "grad_norm": 1.7947436571121216,
+      "learning_rate": 4.1874556333029103e-07,
+      "loss": 0.9134,
+      "step": 25005
+    },
+    {
+      "epoch": 0.6023458009200163,
+      "grad_norm": 1.5894259214401245,
+      "learning_rate": 4.186188013386066e-07,
+      "loss": 0.8828,
+      "step": 25010
+    },
+    {
+      "epoch": 0.6024662219118037,
+      "grad_norm": 1.5322158336639404,
+      "learning_rate": 4.1849203934692216e-07,
+      "loss": 0.8498,
+      "step": 25015
+    },
+    {
+      "epoch": 0.602586642903591,
+      "grad_norm": 1.5399775505065918,
+      "learning_rate": 4.183652773552378e-07,
+      "loss": 0.8729,
+      "step": 25020
+    },
+    {
+      "epoch": 0.6027070638953782,
+      "grad_norm": 1.4279295206069946,
+      "learning_rate": 4.1823851536355335e-07,
+      "loss": 0.8754,
+      "step": 25025
+    },
+    {
+      "epoch": 0.6028274848871655,
+      "grad_norm": 1.558327555656433,
+      "learning_rate": 4.1811175337186894e-07,
+      "loss": 0.9001,
+      "step": 25030
+    },
+    {
+      "epoch": 0.6029479058789529,
+      "grad_norm": 1.5234421491622925,
+      "learning_rate": 4.179849913801846e-07,
+      "loss": 0.9507,
+      "step": 25035
+    },
+    {
+      "epoch": 0.6030683268707401,
+      "grad_norm": 1.3733103275299072,
+      "learning_rate": 4.178582293885001e-07,
+      "loss": 0.8795,
+      "step": 25040
+    },
+    {
+      "epoch": 0.6031887478625274,
+      "grad_norm": 1.5034695863723755,
+      "learning_rate": 4.177314673968157e-07,
+      "loss": 0.8785,
+      "step": 25045
+    },
+    {
+      "epoch": 0.6033091688543147,
+      "grad_norm": 1.475675344467163,
+      "learning_rate": 4.176047054051313e-07,
+      "loss": 0.8879,
+      "step": 25050
+    },
+    {
+      "epoch": 0.603429589846102,
+      "grad_norm": 1.5570924282073975,
+      "learning_rate": 4.174779434134469e-07,
+      "loss": 0.8826,
+      "step": 25055
+    },
+    {
+      "epoch": 0.6035500108378893,
+      "grad_norm": 1.369581699371338,
+      "learning_rate": 4.1735118142176244e-07,
+      "loss": 0.9142,
+      "step": 25060
+    },
+    {
+      "epoch": 0.6036704318296765,
+      "grad_norm": 1.528016448020935,
+      "learning_rate": 4.172244194300781e-07,
+      "loss": 0.9121,
+      "step": 25065
+    },
+    {
+      "epoch": 0.6037908528214638,
+      "grad_norm": 1.5586113929748535,
+      "learning_rate": 4.170976574383936e-07,
+      "loss": 0.8676,
+      "step": 25070
+    },
+    {
+      "epoch": 0.6039112738132512,
+      "grad_norm": 1.7165907621383667,
+      "learning_rate": 4.169708954467092e-07,
+      "loss": 0.9142,
+      "step": 25075
+    },
+    {
+      "epoch": 0.6040316948050384,
+      "grad_norm": 1.5208547115325928,
+      "learning_rate": 4.1684413345502486e-07,
+      "loss": 0.8691,
+      "step": 25080
+    },
+    {
+      "epoch": 0.6041521157968257,
+      "grad_norm": 1.4242299795150757,
+      "learning_rate": 4.167173714633404e-07,
+      "loss": 0.8897,
+      "step": 25085
+    },
+    {
+      "epoch": 0.604272536788613,
+      "grad_norm": 1.5583271980285645,
+      "learning_rate": 4.16590609471656e-07,
+      "loss": 0.9203,
+      "step": 25090
+    },
+    {
+      "epoch": 0.6043929577804003,
+      "grad_norm": 1.7204725742340088,
+      "learning_rate": 4.164638474799716e-07,
+      "loss": 0.8859,
+      "step": 25095
+    },
+    {
+      "epoch": 0.6045133787721876,
+      "grad_norm": 1.7835016250610352,
+      "learning_rate": 4.1633708548828717e-07,
+      "loss": 0.9605,
+      "step": 25100
+    },
+    {
+      "epoch": 0.6046337997639749,
+      "grad_norm": 1.520129919052124,
+      "learning_rate": 4.162103234966027e-07,
+      "loss": 0.8718,
+      "step": 25105
+    },
+    {
+      "epoch": 0.6047542207557621,
+      "grad_norm": 1.5081130266189575,
+      "learning_rate": 4.1608356150491836e-07,
+      "loss": 0.9175,
+      "step": 25110
+    },
+    {
+      "epoch": 0.6048746417475495,
+      "grad_norm": 1.61761474609375,
+      "learning_rate": 4.1595679951323395e-07,
+      "loss": 0.885,
+      "step": 25115
+    },
+    {
+      "epoch": 0.6049950627393367,
+      "grad_norm": 1.576309323310852,
+      "learning_rate": 4.158300375215495e-07,
+      "loss": 0.8932,
+      "step": 25120
+    },
+    {
+      "epoch": 0.605115483731124,
+      "grad_norm": 1.5259867906570435,
+      "learning_rate": 4.1570327552986513e-07,
+      "loss": 0.8986,
+      "step": 25125
+    },
+    {
+      "epoch": 0.6052359047229113,
+      "grad_norm": 1.5326902866363525,
+      "learning_rate": 4.1557651353818067e-07,
+      "loss": 0.9337,
+      "step": 25130
+    },
+    {
+      "epoch": 0.6053563257146986,
+      "grad_norm": 1.5314621925354004,
+      "learning_rate": 4.1544975154649626e-07,
+      "loss": 0.8312,
+      "step": 25135
+    },
+    {
+      "epoch": 0.6054767467064859,
+      "grad_norm": 1.5467171669006348,
+      "learning_rate": 4.1532298955481185e-07,
+      "loss": 0.9158,
+      "step": 25140
+    },
+    {
+      "epoch": 0.6055971676982732,
+      "grad_norm": 1.5144315958023071,
+      "learning_rate": 4.1519622756312745e-07,
+      "loss": 0.8733,
+      "step": 25145
+    },
+    {
+      "epoch": 0.6057175886900604,
+      "grad_norm": 1.8202297687530518,
+      "learning_rate": 4.150694655714431e-07,
+      "loss": 0.9066,
+      "step": 25150
+    },
+    {
+      "epoch": 0.6058380096818478,
+      "grad_norm": 1.5166350603103638,
+      "learning_rate": 4.1494270357975863e-07,
+      "loss": 0.8701,
+      "step": 25155
+    },
+    {
+      "epoch": 0.605958430673635,
+      "grad_norm": 1.5550726652145386,
+      "learning_rate": 4.148159415880742e-07,
+      "loss": 0.9135,
+      "step": 25160
+    },
+    {
+      "epoch": 0.6060788516654223,
+      "grad_norm": 1.5523972511291504,
+      "learning_rate": 4.146891795963898e-07,
+      "loss": 0.9192,
+      "step": 25165
+    },
+    {
+      "epoch": 0.6061992726572096,
+      "grad_norm": 2.545870065689087,
+      "learning_rate": 4.145624176047054e-07,
+      "loss": 0.9183,
+      "step": 25170
+    },
+    {
+      "epoch": 0.6063196936489968,
+      "grad_norm": 1.58835768699646,
+      "learning_rate": 4.1443565561302094e-07,
+      "loss": 0.9499,
+      "step": 25175
+    },
+    {
+      "epoch": 0.6064401146407842,
+      "grad_norm": 1.5333999395370483,
+      "learning_rate": 4.143088936213366e-07,
+      "loss": 0.9539,
+      "step": 25180
+    },
+    {
+      "epoch": 0.6065605356325715,
+      "grad_norm": 1.4999562501907349,
+      "learning_rate": 4.141821316296522e-07,
+      "loss": 0.9189,
+      "step": 25185
+    },
+    {
+      "epoch": 0.6066809566243587,
+      "grad_norm": 1.7383499145507812,
+      "learning_rate": 4.140553696379677e-07,
+      "loss": 0.9103,
+      "step": 25190
+    },
+    {
+      "epoch": 0.606801377616146,
+      "grad_norm": 1.4625837802886963,
+      "learning_rate": 4.1392860764628336e-07,
+      "loss": 0.9116,
+      "step": 25195
+    },
+    {
+      "epoch": 0.6069217986079334,
+      "grad_norm": 1.3118094205856323,
+      "learning_rate": 4.138018456545989e-07,
+      "loss": 0.8632,
+      "step": 25200
+    },
+    {
+      "epoch": 0.6070422195997206,
+      "grad_norm": 1.6964633464813232,
+      "learning_rate": 4.136750836629145e-07,
+      "loss": 0.9179,
+      "step": 25205
+    },
+    {
+      "epoch": 0.6071626405915079,
+      "grad_norm": 1.8581135272979736,
+      "learning_rate": 4.135483216712301e-07,
+      "loss": 0.922,
+      "step": 25210
+    },
+    {
+      "epoch": 0.6072830615832951,
+      "grad_norm": 1.6567142009735107,
+      "learning_rate": 4.134215596795457e-07,
+      "loss": 0.9666,
+      "step": 25215
+    },
+    {
+      "epoch": 0.6074034825750825,
+      "grad_norm": 1.8037158250808716,
+      "learning_rate": 4.132947976878612e-07,
+      "loss": 0.9274,
+      "step": 25220
+    },
+    {
+      "epoch": 0.6075239035668698,
+      "grad_norm": 1.6046236753463745,
+      "learning_rate": 4.1316803569617686e-07,
+      "loss": 0.8757,
+      "step": 25225
+    },
+    {
+      "epoch": 0.607644324558657,
+      "grad_norm": 1.5918865203857422,
+      "learning_rate": 4.1304127370449245e-07,
+      "loss": 0.8372,
+      "step": 25230
+    },
+    {
+      "epoch": 0.6077647455504444,
+      "grad_norm": 1.3744317293167114,
+      "learning_rate": 4.12914511712808e-07,
+      "loss": 0.914,
+      "step": 25235
+    },
+    {
+      "epoch": 0.6078851665422317,
+      "grad_norm": 1.3174458742141724,
+      "learning_rate": 4.1278774972112364e-07,
+      "loss": 0.8677,
+      "step": 25240
+    },
+    {
+      "epoch": 0.6080055875340189,
+      "grad_norm": 1.6742388010025024,
+      "learning_rate": 4.126609877294392e-07,
+      "loss": 0.9147,
+      "step": 25245
+    },
+    {
+      "epoch": 0.6081260085258062,
+      "grad_norm": 1.6274393796920776,
+      "learning_rate": 4.1253422573775477e-07,
+      "loss": 0.9061,
+      "step": 25250
+    },
+    {
+      "epoch": 0.6082464295175936,
+      "grad_norm": 1.349190592765808,
+      "learning_rate": 4.1240746374607036e-07,
+      "loss": 0.8988,
+      "step": 25255
+    },
+    {
+      "epoch": 0.6083668505093808,
+      "grad_norm": 1.382152795791626,
+      "learning_rate": 4.1228070175438595e-07,
+      "loss": 0.9355,
+      "step": 25260
+    },
+    {
+      "epoch": 0.6084872715011681,
+      "grad_norm": 1.608886957168579,
+      "learning_rate": 4.1215393976270154e-07,
+      "loss": 0.9753,
+      "step": 25265
+    },
+    {
+      "epoch": 0.6086076924929553,
+      "grad_norm": 1.4702397584915161,
+      "learning_rate": 4.1202717777101713e-07,
+      "loss": 0.9265,
+      "step": 25270
+    },
+    {
+      "epoch": 0.6087281134847427,
+      "grad_norm": 1.6230460405349731,
+      "learning_rate": 4.119004157793327e-07,
+      "loss": 0.9075,
+      "step": 25275
+    },
+    {
+      "epoch": 0.60884853447653,
+      "grad_norm": 1.5647751092910767,
+      "learning_rate": 4.1177365378764827e-07,
+      "loss": 0.896,
+      "step": 25280
+    },
+    {
+      "epoch": 0.6089689554683172,
+      "grad_norm": 1.497682809829712,
+      "learning_rate": 4.116468917959639e-07,
+      "loss": 0.8999,
+      "step": 25285
+    },
+    {
+      "epoch": 0.6090893764601045,
+      "grad_norm": 1.3031469583511353,
+      "learning_rate": 4.1152012980427945e-07,
+      "loss": 0.8519,
+      "step": 25290
+    },
+    {
+      "epoch": 0.6092097974518919,
+      "grad_norm": 1.5770944356918335,
+      "learning_rate": 4.1139336781259504e-07,
+      "loss": 0.8997,
+      "step": 25295
+    },
+    {
+      "epoch": 0.6093302184436791,
+      "grad_norm": 1.457399606704712,
+      "learning_rate": 4.112666058209107e-07,
+      "loss": 0.8614,
+      "step": 25300
+    },
+    {
+      "epoch": 0.6094506394354664,
+      "grad_norm": 1.5790886878967285,
+      "learning_rate": 4.111398438292262e-07,
+      "loss": 0.8802,
+      "step": 25305
+    },
+    {
+      "epoch": 0.6095710604272537,
+      "grad_norm": 1.544673204421997,
+      "learning_rate": 4.110130818375418e-07,
+      "loss": 0.8845,
+      "step": 25310
+    },
+    {
+      "epoch": 0.609691481419041,
+      "grad_norm": 1.6734991073608398,
+      "learning_rate": 4.108863198458574e-07,
+      "loss": 0.9178,
+      "step": 25315
+    },
+    {
+      "epoch": 0.6098119024108283,
+      "grad_norm": 1.4224692583084106,
+      "learning_rate": 4.10759557854173e-07,
+      "loss": 0.8882,
+      "step": 25320
+    },
+    {
+      "epoch": 0.6099323234026155,
+      "grad_norm": 1.4789096117019653,
+      "learning_rate": 4.1063279586248854e-07,
+      "loss": 0.9298,
+      "step": 25325
+    },
+    {
+      "epoch": 0.6100527443944028,
+      "grad_norm": 1.5312858819961548,
+      "learning_rate": 4.105060338708042e-07,
+      "loss": 0.8349,
+      "step": 25330
+    },
+    {
+      "epoch": 0.6101731653861902,
+      "grad_norm": 1.6153223514556885,
+      "learning_rate": 4.103792718791197e-07,
+      "loss": 0.948,
+      "step": 25335
+    },
+    {
+      "epoch": 0.6102935863779774,
+      "grad_norm": 1.6232482194900513,
+      "learning_rate": 4.102525098874353e-07,
+      "loss": 0.9349,
+      "step": 25340
+    },
+    {
+      "epoch": 0.6104140073697647,
+      "grad_norm": 1.5309494733810425,
+      "learning_rate": 4.1012574789575096e-07,
+      "loss": 0.8721,
+      "step": 25345
+    },
+    {
+      "epoch": 0.610534428361552,
+      "grad_norm": 1.5360968112945557,
+      "learning_rate": 4.099989859040665e-07,
+      "loss": 0.8604,
+      "step": 25350
+    },
+    {
+      "epoch": 0.6106548493533392,
+      "grad_norm": 1.5799692869186401,
+      "learning_rate": 4.098722239123821e-07,
+      "loss": 0.9158,
+      "step": 25355
+    },
+    {
+      "epoch": 0.6107752703451266,
+      "grad_norm": 1.5045326948165894,
+      "learning_rate": 4.097454619206977e-07,
+      "loss": 0.9081,
+      "step": 25360
+    },
+    {
+      "epoch": 0.6108956913369139,
+      "grad_norm": 1.4619923830032349,
+      "learning_rate": 4.0961869992901327e-07,
+      "loss": 0.9109,
+      "step": 25365
+    },
+    {
+      "epoch": 0.6110161123287011,
+      "grad_norm": 1.5276981592178345,
+      "learning_rate": 4.094919379373288e-07,
+      "loss": 0.9295,
+      "step": 25370
+    },
+    {
+      "epoch": 0.6111365333204885,
+      "grad_norm": 1.6464943885803223,
+      "learning_rate": 4.0936517594564446e-07,
+      "loss": 0.8637,
+      "step": 25375
+    },
+    {
+      "epoch": 0.6112569543122757,
+      "grad_norm": 1.5115057229995728,
+      "learning_rate": 4.0923841395396005e-07,
+      "loss": 0.9234,
+      "step": 25380
+    },
+    {
+      "epoch": 0.611377375304063,
+      "grad_norm": 1.4718031883239746,
+      "learning_rate": 4.091116519622756e-07,
+      "loss": 0.8478,
+      "step": 25385
+    },
+    {
+      "epoch": 0.6114977962958503,
+      "grad_norm": 1.651680827140808,
+      "learning_rate": 4.0898488997059123e-07,
+      "loss": 0.95,
+      "step": 25390
+    },
+    {
+      "epoch": 0.6116182172876375,
+      "grad_norm": 1.441233515739441,
+      "learning_rate": 4.0885812797890677e-07,
+      "loss": 0.8692,
+      "step": 25395
+    },
+    {
+      "epoch": 0.6117386382794249,
+      "grad_norm": 1.785135269165039,
+      "learning_rate": 4.0873136598722236e-07,
+      "loss": 0.9518,
+      "step": 25400
+    },
+    {
+      "epoch": 0.6118590592712122,
+      "grad_norm": 1.4425971508026123,
+      "learning_rate": 4.0860460399553795e-07,
+      "loss": 0.8859,
+      "step": 25405
+    },
+    {
+      "epoch": 0.6119794802629994,
+      "grad_norm": 1.6750375032424927,
+      "learning_rate": 4.0847784200385355e-07,
+      "loss": 0.9056,
+      "step": 25410
+    },
+    {
+      "epoch": 0.6120999012547867,
+      "grad_norm": 1.4933054447174072,
+      "learning_rate": 4.0835108001216914e-07,
+      "loss": 0.8847,
+      "step": 25415
+    },
+    {
+      "epoch": 0.6122203222465741,
+      "grad_norm": 1.4828130006790161,
+      "learning_rate": 4.0822431802048473e-07,
+      "loss": 0.916,
+      "step": 25420
+    },
+    {
+      "epoch": 0.6123407432383613,
+      "grad_norm": 1.4448007345199585,
+      "learning_rate": 4.080975560288003e-07,
+      "loss": 0.8932,
+      "step": 25425
+    },
+    {
+      "epoch": 0.6124611642301486,
+      "grad_norm": 1.5228817462921143,
+      "learning_rate": 4.0797079403711586e-07,
+      "loss": 0.8805,
+      "step": 25430
+    },
+    {
+      "epoch": 0.6125815852219358,
+      "grad_norm": 1.779428482055664,
+      "learning_rate": 4.078440320454315e-07,
+      "loss": 0.9489,
+      "step": 25435
+    },
+    {
+      "epoch": 0.6127020062137232,
+      "grad_norm": 1.5332070589065552,
+      "learning_rate": 4.0771727005374704e-07,
+      "loss": 0.8749,
+      "step": 25440
+    },
+    {
+      "epoch": 0.6128224272055105,
+      "grad_norm": 1.5100640058517456,
+      "learning_rate": 4.0759050806206264e-07,
+      "loss": 0.8826,
+      "step": 25445
+    },
+    {
+      "epoch": 0.6129428481972977,
+      "grad_norm": 1.4698293209075928,
+      "learning_rate": 4.074637460703783e-07,
+      "loss": 0.9082,
+      "step": 25450
+    },
+    {
+      "epoch": 0.613063269189085,
+      "grad_norm": 1.4528988599777222,
+      "learning_rate": 4.073369840786938e-07,
+      "loss": 0.9264,
+      "step": 25455
+    },
+    {
+      "epoch": 0.6131836901808724,
+      "grad_norm": 1.6008024215698242,
+      "learning_rate": 4.072102220870094e-07,
+      "loss": 0.8836,
+      "step": 25460
+    },
+    {
+      "epoch": 0.6133041111726596,
+      "grad_norm": 1.3941173553466797,
+      "learning_rate": 4.07083460095325e-07,
+      "loss": 0.8613,
+      "step": 25465
+    },
+    {
+      "epoch": 0.6134245321644469,
+      "grad_norm": 1.6055837869644165,
+      "learning_rate": 4.069566981036406e-07,
+      "loss": 0.9142,
+      "step": 25470
+    },
+    {
+      "epoch": 0.6135449531562341,
+      "grad_norm": 1.5556641817092896,
+      "learning_rate": 4.0682993611195613e-07,
+      "loss": 0.8544,
+      "step": 25475
+    },
+    {
+      "epoch": 0.6136653741480215,
+      "grad_norm": 1.5129414796829224,
+      "learning_rate": 4.067031741202718e-07,
+      "loss": 0.9043,
+      "step": 25480
+    },
+    {
+      "epoch": 0.6137857951398088,
+      "grad_norm": 1.6764026880264282,
+      "learning_rate": 4.065764121285873e-07,
+      "loss": 0.8123,
+      "step": 25485
+    },
+    {
+      "epoch": 0.613906216131596,
+      "grad_norm": 1.4478250741958618,
+      "learning_rate": 4.064496501369029e-07,
+      "loss": 0.8387,
+      "step": 25490
+    },
+    {
+      "epoch": 0.6140266371233833,
+      "grad_norm": 1.7998301982879639,
+      "learning_rate": 4.0632288814521855e-07,
+      "loss": 0.8784,
+      "step": 25495
+    },
+    {
+      "epoch": 0.6141470581151707,
+      "grad_norm": 1.9295172691345215,
+      "learning_rate": 4.061961261535341e-07,
+      "loss": 0.9652,
+      "step": 25500
+    },
+    {
+      "epoch": 0.6142674791069579,
+      "grad_norm": 1.382350206375122,
+      "learning_rate": 4.060693641618497e-07,
+      "loss": 0.8637,
+      "step": 25505
+    },
+    {
+      "epoch": 0.6143879000987452,
+      "grad_norm": 1.5238209962844849,
+      "learning_rate": 4.059426021701653e-07,
+      "loss": 0.892,
+      "step": 25510
+    },
+    {
+      "epoch": 0.6145083210905325,
+      "grad_norm": 1.3008952140808105,
+      "learning_rate": 4.0581584017848087e-07,
+      "loss": 0.8638,
+      "step": 25515
+    },
+    {
+      "epoch": 0.6146287420823198,
+      "grad_norm": 1.6505320072174072,
+      "learning_rate": 4.056890781867964e-07,
+      "loss": 0.9282,
+      "step": 25520
+    },
+    {
+      "epoch": 0.6147491630741071,
+      "grad_norm": 1.905221700668335,
+      "learning_rate": 4.0556231619511205e-07,
+      "loss": 0.8916,
+      "step": 25525
+    },
+    {
+      "epoch": 0.6148695840658943,
+      "grad_norm": 1.665234923362732,
+      "learning_rate": 4.0543555420342764e-07,
+      "loss": 0.911,
+      "step": 25530
+    },
+    {
+      "epoch": 0.6149900050576816,
+      "grad_norm": 1.7937983274459839,
+      "learning_rate": 4.053087922117432e-07,
+      "loss": 0.9336,
+      "step": 25535
+    },
+    {
+      "epoch": 0.615110426049469,
+      "grad_norm": 1.494798183441162,
+      "learning_rate": 4.0518203022005883e-07,
+      "loss": 0.8983,
+      "step": 25540
+    },
+    {
+      "epoch": 0.6152308470412562,
+      "grad_norm": 1.6080095767974854,
+      "learning_rate": 4.0505526822837437e-07,
+      "loss": 0.8837,
+      "step": 25545
+    },
+    {
+      "epoch": 0.6153512680330435,
+      "grad_norm": 1.746788740158081,
+      "learning_rate": 4.0492850623668996e-07,
+      "loss": 0.8431,
+      "step": 25550
+    },
+    {
+      "epoch": 0.6154716890248308,
+      "grad_norm": 1.5956007242202759,
+      "learning_rate": 4.0480174424500555e-07,
+      "loss": 0.8483,
+      "step": 25555
+    },
+    {
+      "epoch": 0.6155921100166181,
+      "grad_norm": 1.6954641342163086,
+      "learning_rate": 4.0467498225332114e-07,
+      "loss": 0.8781,
+      "step": 25560
+    },
+    {
+      "epoch": 0.6157125310084054,
+      "grad_norm": 1.6233452558517456,
+      "learning_rate": 4.0454822026163673e-07,
+      "loss": 0.9598,
+      "step": 25565
+    },
+    {
+      "epoch": 0.6158329520001927,
+      "grad_norm": 1.3930495977401733,
+      "learning_rate": 4.044214582699523e-07,
+      "loss": 0.8972,
+      "step": 25570
+    },
+    {
+      "epoch": 0.6159533729919799,
+      "grad_norm": 1.6070493459701538,
+      "learning_rate": 4.042946962782679e-07,
+      "loss": 0.896,
+      "step": 25575
+    },
+    {
+      "epoch": 0.6160737939837673,
+      "grad_norm": 1.5206298828125,
+      "learning_rate": 4.0416793428658346e-07,
+      "loss": 0.8612,
+      "step": 25580
+    },
+    {
+      "epoch": 0.6161942149755545,
+      "grad_norm": 1.558423399925232,
+      "learning_rate": 4.040411722948991e-07,
+      "loss": 0.8981,
+      "step": 25585
+    },
+    {
+      "epoch": 0.6163146359673418,
+      "grad_norm": 1.5652716159820557,
+      "learning_rate": 4.0391441030321464e-07,
+      "loss": 0.8786,
+      "step": 25590
+    },
+    {
+      "epoch": 0.6164350569591291,
+      "grad_norm": 1.5595407485961914,
+      "learning_rate": 4.0378764831153023e-07,
+      "loss": 0.934,
+      "step": 25595
+    },
+    {
+      "epoch": 0.6165554779509164,
+      "grad_norm": 1.6016736030578613,
+      "learning_rate": 4.036608863198459e-07,
+      "loss": 0.9026,
+      "step": 25600
+    },
+    {
+      "epoch": 0.6166758989427037,
+      "grad_norm": 1.8003485202789307,
+      "learning_rate": 4.035341243281614e-07,
+      "loss": 0.8784,
+      "step": 25605
+    },
+    {
+      "epoch": 0.616796319934491,
+      "grad_norm": 1.4704478979110718,
+      "learning_rate": 4.03407362336477e-07,
+      "loss": 0.8715,
+      "step": 25610
+    },
+    {
+      "epoch": 0.6169167409262782,
+      "grad_norm": 1.592137098312378,
+      "learning_rate": 4.032806003447926e-07,
+      "loss": 0.8437,
+      "step": 25615
+    },
+    {
+      "epoch": 0.6170371619180656,
+      "grad_norm": 1.5061300992965698,
+      "learning_rate": 4.031538383531082e-07,
+      "loss": 0.9236,
+      "step": 25620
+    },
+    {
+      "epoch": 0.6171575829098529,
+      "grad_norm": 1.4585505723953247,
+      "learning_rate": 4.0302707636142373e-07,
+      "loss": 0.9172,
+      "step": 25625
+    },
+    {
+      "epoch": 0.6172780039016401,
+      "grad_norm": 1.784454584121704,
+      "learning_rate": 4.0290031436973937e-07,
+      "loss": 0.9294,
+      "step": 25630
+    },
+    {
+      "epoch": 0.6173984248934274,
+      "grad_norm": 1.5327225923538208,
+      "learning_rate": 4.027735523780549e-07,
+      "loss": 0.894,
+      "step": 25635
+    },
+    {
+      "epoch": 0.6175188458852147,
+      "grad_norm": 1.559232473373413,
+      "learning_rate": 4.026467903863705e-07,
+      "loss": 0.898,
+      "step": 25640
+    },
+    {
+      "epoch": 0.617639266877002,
+      "grad_norm": 1.4772984981536865,
+      "learning_rate": 4.0252002839468615e-07,
+      "loss": 0.9417,
+      "step": 25645
+    },
+    {
+      "epoch": 0.6177596878687893,
+      "grad_norm": 1.6661555767059326,
+      "learning_rate": 4.023932664030017e-07,
+      "loss": 0.8945,
+      "step": 25650
+    },
+    {
+      "epoch": 0.6178801088605765,
+      "grad_norm": 1.5550674200057983,
+      "learning_rate": 4.0226650441131733e-07,
+      "loss": 0.8786,
+      "step": 25655
+    },
+    {
+      "epoch": 0.6180005298523639,
+      "grad_norm": 1.5687839984893799,
+      "learning_rate": 4.0213974241963287e-07,
+      "loss": 0.9566,
+      "step": 25660
+    },
+    {
+      "epoch": 0.6181209508441512,
+      "grad_norm": 1.4756946563720703,
+      "learning_rate": 4.0201298042794846e-07,
+      "loss": 0.89,
+      "step": 25665
+    },
+    {
+      "epoch": 0.6182413718359384,
+      "grad_norm": 1.5157841444015503,
+      "learning_rate": 4.0188621843626405e-07,
+      "loss": 0.9546,
+      "step": 25670
+    },
+    {
+      "epoch": 0.6183617928277257,
+      "grad_norm": 1.4619146585464478,
+      "learning_rate": 4.0175945644457965e-07,
+      "loss": 0.9052,
+      "step": 25675
+    },
+    {
+      "epoch": 0.6184822138195131,
+      "grad_norm": 1.655191421508789,
+      "learning_rate": 4.0163269445289524e-07,
+      "loss": 0.8417,
+      "step": 25680
+    },
+    {
+      "epoch": 0.6186026348113003,
+      "grad_norm": 1.5669910907745361,
+      "learning_rate": 4.0150593246121083e-07,
+      "loss": 0.8885,
+      "step": 25685
+    },
+    {
+      "epoch": 0.6187230558030876,
+      "grad_norm": 1.7892998456954956,
+      "learning_rate": 4.013791704695264e-07,
+      "loss": 0.8722,
+      "step": 25690
+    },
+    {
+      "epoch": 0.6188434767948748,
+      "grad_norm": 1.5242003202438354,
+      "learning_rate": 4.0125240847784196e-07,
+      "loss": 0.8916,
+      "step": 25695
+    },
+    {
+      "epoch": 0.6189638977866622,
+      "grad_norm": 1.4352613687515259,
+      "learning_rate": 4.011256464861576e-07,
+      "loss": 0.9305,
+      "step": 25700
+    },
+    {
+      "epoch": 0.6190843187784495,
+      "grad_norm": 1.5476058721542358,
+      "learning_rate": 4.0099888449447314e-07,
+      "loss": 0.9093,
+      "step": 25705
+    },
+    {
+      "epoch": 0.6192047397702367,
+      "grad_norm": 1.5854041576385498,
+      "learning_rate": 4.0087212250278874e-07,
+      "loss": 0.8884,
+      "step": 25710
+    },
+    {
+      "epoch": 0.619325160762024,
+      "grad_norm": 1.4123213291168213,
+      "learning_rate": 4.007453605111044e-07,
+      "loss": 0.918,
+      "step": 25715
+    },
+    {
+      "epoch": 0.6194455817538114,
+      "grad_norm": 1.6956759691238403,
+      "learning_rate": 4.006185985194199e-07,
+      "loss": 0.8959,
+      "step": 25720
+    },
+    {
+      "epoch": 0.6195660027455986,
+      "grad_norm": 1.5501787662506104,
+      "learning_rate": 4.004918365277355e-07,
+      "loss": 0.8972,
+      "step": 25725
+    },
+    {
+      "epoch": 0.6196864237373859,
+      "grad_norm": 1.5617278814315796,
+      "learning_rate": 4.003650745360511e-07,
+      "loss": 0.885,
+      "step": 25730
+    },
+    {
+      "epoch": 0.6198068447291731,
+      "grad_norm": 1.554679274559021,
+      "learning_rate": 4.002383125443667e-07,
+      "loss": 0.9297,
+      "step": 25735
+    },
+    {
+      "epoch": 0.6199272657209605,
+      "grad_norm": 1.5556864738464355,
+      "learning_rate": 4.0011155055268223e-07,
+      "loss": 0.9393,
+      "step": 25740
+    },
+    {
+      "epoch": 0.6200476867127478,
+      "grad_norm": 1.5048941373825073,
+      "learning_rate": 3.999847885609979e-07,
+      "loss": 0.882,
+      "step": 25745
+    },
+    {
+      "epoch": 0.620168107704535,
+      "grad_norm": 1.519623041152954,
+      "learning_rate": 3.9985802656931347e-07,
+      "loss": 0.884,
+      "step": 25750
+    },
+    {
+      "epoch": 0.6202885286963223,
+      "grad_norm": 1.8814226388931274,
+      "learning_rate": 3.99731264577629e-07,
+      "loss": 0.907,
+      "step": 25755
+    },
+    {
+      "epoch": 0.6204089496881097,
+      "grad_norm": 1.719388723373413,
+      "learning_rate": 3.9960450258594465e-07,
+      "loss": 0.8555,
+      "step": 25760
+    },
+    {
+      "epoch": 0.6205293706798969,
+      "grad_norm": 1.635589361190796,
+      "learning_rate": 3.994777405942602e-07,
+      "loss": 0.9124,
+      "step": 25765
+    },
+    {
+      "epoch": 0.6206497916716842,
+      "grad_norm": 1.462107539176941,
+      "learning_rate": 3.993509786025758e-07,
+      "loss": 0.8733,
+      "step": 25770
+    },
+    {
+      "epoch": 0.6207702126634715,
+      "grad_norm": 1.593778371810913,
+      "learning_rate": 3.992242166108914e-07,
+      "loss": 0.8644,
+      "step": 25775
+    },
+    {
+      "epoch": 0.6208906336552588,
+      "grad_norm": 1.617329716682434,
+      "learning_rate": 3.9909745461920697e-07,
+      "loss": 0.8413,
+      "step": 25780
+    },
+    {
+      "epoch": 0.6210110546470461,
+      "grad_norm": 1.8050616979599,
+      "learning_rate": 3.989706926275225e-07,
+      "loss": 0.904,
+      "step": 25785
+    },
+    {
+      "epoch": 0.6211314756388333,
+      "grad_norm": 1.5802415609359741,
+      "learning_rate": 3.9884393063583815e-07,
+      "loss": 0.859,
+      "step": 25790
+    },
+    {
+      "epoch": 0.6212518966306206,
+      "grad_norm": 1.4133882522583008,
+      "learning_rate": 3.9871716864415374e-07,
+      "loss": 0.8923,
+      "step": 25795
+    },
+    {
+      "epoch": 0.621372317622408,
+      "grad_norm": 1.67210054397583,
+      "learning_rate": 3.985904066524693e-07,
+      "loss": 0.8528,
+      "step": 25800
+    },
+    {
+      "epoch": 0.6214927386141952,
+      "grad_norm": 1.5611308813095093,
+      "learning_rate": 3.9846364466078493e-07,
+      "loss": 0.8864,
+      "step": 25805
+    },
+    {
+      "epoch": 0.6216131596059825,
+      "grad_norm": 1.5887681245803833,
+      "learning_rate": 3.9833688266910047e-07,
+      "loss": 0.8617,
+      "step": 25810
+    },
+    {
+      "epoch": 0.6217335805977698,
+      "grad_norm": 1.3824416399002075,
+      "learning_rate": 3.9821012067741606e-07,
+      "loss": 0.8854,
+      "step": 25815
+    },
+    {
+      "epoch": 0.6218540015895571,
+      "grad_norm": 1.5376073122024536,
+      "learning_rate": 3.9808335868573165e-07,
+      "loss": 0.9021,
+      "step": 25820
+    },
+    {
+      "epoch": 0.6219744225813444,
+      "grad_norm": 1.580991506576538,
+      "learning_rate": 3.9795659669404724e-07,
+      "loss": 0.9195,
+      "step": 25825
+    },
+    {
+      "epoch": 0.6220948435731317,
+      "grad_norm": 1.452588677406311,
+      "learning_rate": 3.9782983470236283e-07,
+      "loss": 0.9412,
+      "step": 25830
+    },
+    {
+      "epoch": 0.6222152645649189,
+      "grad_norm": 1.4612324237823486,
+      "learning_rate": 3.977030727106784e-07,
+      "loss": 0.8796,
+      "step": 25835
+    },
+    {
+      "epoch": 0.6223356855567063,
+      "grad_norm": 1.5148411989212036,
+      "learning_rate": 3.97576310718994e-07,
+      "loss": 0.8432,
+      "step": 25840
+    },
+    {
+      "epoch": 0.6224561065484935,
+      "grad_norm": 1.6802312135696411,
+      "learning_rate": 3.9744954872730956e-07,
+      "loss": 0.8767,
+      "step": 25845
+    },
+    {
+      "epoch": 0.6225765275402808,
+      "grad_norm": 1.4846410751342773,
+      "learning_rate": 3.973227867356252e-07,
+      "loss": 0.8864,
+      "step": 25850
+    },
+    {
+      "epoch": 0.6226969485320681,
+      "grad_norm": 1.74970543384552,
+      "learning_rate": 3.9719602474394074e-07,
+      "loss": 0.9246,
+      "step": 25855
+    },
+    {
+      "epoch": 0.6228173695238554,
+      "grad_norm": 1.4260430335998535,
+      "learning_rate": 3.9706926275225633e-07,
+      "loss": 0.8989,
+      "step": 25860
+    },
+    {
+      "epoch": 0.6229377905156427,
+      "grad_norm": 1.3911371231079102,
+      "learning_rate": 3.96942500760572e-07,
+      "loss": 0.8305,
+      "step": 25865
+    },
+    {
+      "epoch": 0.62305821150743,
+      "grad_norm": 1.4906085729599,
+      "learning_rate": 3.968157387688875e-07,
+      "loss": 0.9143,
+      "step": 25870
+    },
+    {
+      "epoch": 0.6231786324992172,
+      "grad_norm": 1.4780354499816895,
+      "learning_rate": 3.966889767772031e-07,
+      "loss": 0.8974,
+      "step": 25875
+    },
+    {
+      "epoch": 0.6232990534910046,
+      "grad_norm": 1.6314855813980103,
+      "learning_rate": 3.965622147855187e-07,
+      "loss": 0.9217,
+      "step": 25880
+    },
+    {
+      "epoch": 0.6234194744827919,
+      "grad_norm": 1.570934772491455,
+      "learning_rate": 3.964354527938343e-07,
+      "loss": 0.9263,
+      "step": 25885
+    },
+    {
+      "epoch": 0.6235398954745791,
+      "grad_norm": 1.472892165184021,
+      "learning_rate": 3.9630869080214983e-07,
+      "loss": 0.9212,
+      "step": 25890
+    },
+    {
+      "epoch": 0.6236603164663664,
+      "grad_norm": 1.8066248893737793,
+      "learning_rate": 3.9618192881046547e-07,
+      "loss": 0.8877,
+      "step": 25895
+    },
+    {
+      "epoch": 0.6237807374581537,
+      "grad_norm": 1.555860161781311,
+      "learning_rate": 3.9605516681878107e-07,
+      "loss": 0.8556,
+      "step": 25900
+    },
+    {
+      "epoch": 0.623901158449941,
+      "grad_norm": 1.4210270643234253,
+      "learning_rate": 3.959284048270966e-07,
+      "loss": 0.874,
+      "step": 25905
+    },
+    {
+      "epoch": 0.6240215794417283,
+      "grad_norm": 1.781644344329834,
+      "learning_rate": 3.9580164283541225e-07,
+      "loss": 0.8762,
+      "step": 25910
+    },
+    {
+      "epoch": 0.6241420004335155,
+      "grad_norm": 1.4464399814605713,
+      "learning_rate": 3.956748808437278e-07,
+      "loss": 0.8866,
+      "step": 25915
+    },
+    {
+      "epoch": 0.6242624214253029,
+      "grad_norm": 1.4830899238586426,
+      "learning_rate": 3.955481188520434e-07,
+      "loss": 0.9026,
+      "step": 25920
+    },
+    {
+      "epoch": 0.6243828424170902,
+      "grad_norm": 1.7051576375961304,
+      "learning_rate": 3.9542135686035897e-07,
+      "loss": 0.8593,
+      "step": 25925
+    },
+    {
+      "epoch": 0.6245032634088774,
+      "grad_norm": 1.7752699851989746,
+      "learning_rate": 3.9529459486867456e-07,
+      "loss": 0.8789,
+      "step": 25930
+    },
+    {
+      "epoch": 0.6246236844006647,
+      "grad_norm": 1.7253508567810059,
+      "learning_rate": 3.951678328769901e-07,
+      "loss": 0.897,
+      "step": 25935
+    },
+    {
+      "epoch": 0.6247441053924521,
+      "grad_norm": 1.6633070707321167,
+      "learning_rate": 3.9504107088530575e-07,
+      "loss": 0.8794,
+      "step": 25940
+    },
+    {
+      "epoch": 0.6248645263842393,
+      "grad_norm": 1.5716090202331543,
+      "learning_rate": 3.9491430889362134e-07,
+      "loss": 0.8672,
+      "step": 25945
+    },
+    {
+      "epoch": 0.6249849473760266,
+      "grad_norm": 1.3929953575134277,
+      "learning_rate": 3.947875469019369e-07,
+      "loss": 0.8398,
+      "step": 25950
+    },
+    {
+      "epoch": 0.6251053683678138,
+      "grad_norm": 1.645712971687317,
+      "learning_rate": 3.946607849102525e-07,
+      "loss": 0.9471,
+      "step": 25955
+    },
+    {
+      "epoch": 0.6252257893596012,
+      "grad_norm": 1.4748599529266357,
+      "learning_rate": 3.9453402291856806e-07,
+      "loss": 0.9099,
+      "step": 25960
+    },
+    {
+      "epoch": 0.6253462103513885,
+      "grad_norm": 1.6290721893310547,
+      "learning_rate": 3.9440726092688365e-07,
+      "loss": 0.9081,
+      "step": 25965
+    },
+    {
+      "epoch": 0.6254666313431757,
+      "grad_norm": 1.4641081094741821,
+      "learning_rate": 3.9428049893519924e-07,
+      "loss": 0.884,
+      "step": 25970
+    },
+    {
+      "epoch": 0.625587052334963,
+      "grad_norm": 1.4148682355880737,
+      "learning_rate": 3.9415373694351484e-07,
+      "loss": 0.921,
+      "step": 25975
+    },
+    {
+      "epoch": 0.6257074733267504,
+      "grad_norm": 1.4865994453430176,
+      "learning_rate": 3.9402697495183043e-07,
+      "loss": 0.9002,
+      "step": 25980
+    },
+    {
+      "epoch": 0.6258278943185376,
+      "grad_norm": 2.53155779838562,
+      "learning_rate": 3.93900212960146e-07,
+      "loss": 0.8965,
+      "step": 25985
+    },
+    {
+      "epoch": 0.6259483153103249,
+      "grad_norm": 1.5487921237945557,
+      "learning_rate": 3.937734509684616e-07,
+      "loss": 0.9097,
+      "step": 25990
+    },
+    {
+      "epoch": 0.6260687363021122,
+      "grad_norm": 1.5251574516296387,
+      "learning_rate": 3.9364668897677715e-07,
+      "loss": 0.8939,
+      "step": 25995
+    },
+    {
+      "epoch": 0.6261891572938995,
+      "grad_norm": 1.7059253454208374,
+      "learning_rate": 3.935199269850928e-07,
+      "loss": 0.9429,
+      "step": 26000
+    },
+    {
+      "epoch": 0.6263095782856868,
+      "grad_norm": 1.45933997631073,
+      "learning_rate": 3.9339316499340833e-07,
+      "loss": 0.9006,
+      "step": 26005
+    },
+    {
+      "epoch": 0.626429999277474,
+      "grad_norm": 1.5664831399917603,
+      "learning_rate": 3.932664030017239e-07,
+      "loss": 0.8552,
+      "step": 26010
+    },
+    {
+      "epoch": 0.6265504202692613,
+      "grad_norm": 1.6096681356430054,
+      "learning_rate": 3.9313964101003957e-07,
+      "loss": 0.9251,
+      "step": 26015
+    },
+    {
+      "epoch": 0.6266708412610487,
+      "grad_norm": 1.4551560878753662,
+      "learning_rate": 3.930128790183551e-07,
+      "loss": 0.9036,
+      "step": 26020
+    },
+    {
+      "epoch": 0.6267912622528359,
+      "grad_norm": 1.4428728818893433,
+      "learning_rate": 3.928861170266707e-07,
+      "loss": 0.8899,
+      "step": 26025
+    },
+    {
+      "epoch": 0.6269116832446232,
+      "grad_norm": 1.5575615167617798,
+      "learning_rate": 3.927593550349863e-07,
+      "loss": 0.9135,
+      "step": 26030
+    },
+    {
+      "epoch": 0.6270321042364105,
+      "grad_norm": 1.4295635223388672,
+      "learning_rate": 3.926325930433019e-07,
+      "loss": 0.8989,
+      "step": 26035
+    },
+    {
+      "epoch": 0.6271525252281978,
+      "grad_norm": 1.672738790512085,
+      "learning_rate": 3.925058310516174e-07,
+      "loss": 0.9512,
+      "step": 26040
+    },
+    {
+      "epoch": 0.6272729462199851,
+      "grad_norm": 1.6317218542099,
+      "learning_rate": 3.9237906905993307e-07,
+      "loss": 0.9122,
+      "step": 26045
+    },
+    {
+      "epoch": 0.6273933672117723,
+      "grad_norm": 1.7149540185928345,
+      "learning_rate": 3.922523070682486e-07,
+      "loss": 0.8781,
+      "step": 26050
+    },
+    {
+      "epoch": 0.6275137882035596,
+      "grad_norm": 1.5915231704711914,
+      "learning_rate": 3.921255450765642e-07,
+      "loss": 0.916,
+      "step": 26055
+    },
+    {
+      "epoch": 0.627634209195347,
+      "grad_norm": 1.4769103527069092,
+      "learning_rate": 3.9199878308487984e-07,
+      "loss": 0.8903,
+      "step": 26060
+    },
+    {
+      "epoch": 0.6277546301871342,
+      "grad_norm": 1.4845454692840576,
+      "learning_rate": 3.918720210931954e-07,
+      "loss": 0.8652,
+      "step": 26065
+    },
+    {
+      "epoch": 0.6278750511789215,
+      "grad_norm": 1.4999992847442627,
+      "learning_rate": 3.91745259101511e-07,
+      "loss": 0.9442,
+      "step": 26070
+    },
+    {
+      "epoch": 0.6279954721707088,
+      "grad_norm": 1.6089973449707031,
+      "learning_rate": 3.9161849710982657e-07,
+      "loss": 0.9395,
+      "step": 26075
+    },
+    {
+      "epoch": 0.628115893162496,
+      "grad_norm": 1.5661829710006714,
+      "learning_rate": 3.9149173511814216e-07,
+      "loss": 0.8413,
+      "step": 26080
+    },
+    {
+      "epoch": 0.6282363141542834,
+      "grad_norm": 1.571110725402832,
+      "learning_rate": 3.913649731264577e-07,
+      "loss": 0.8922,
+      "step": 26085
+    },
+    {
+      "epoch": 0.6283567351460707,
+      "grad_norm": 1.6855576038360596,
+      "learning_rate": 3.9123821113477334e-07,
+      "loss": 0.9262,
+      "step": 26090
+    },
+    {
+      "epoch": 0.6284771561378579,
+      "grad_norm": 1.6715831756591797,
+      "learning_rate": 3.9111144914308893e-07,
+      "loss": 0.8839,
+      "step": 26095
+    },
+    {
+      "epoch": 0.6285975771296453,
+      "grad_norm": 1.7230852842330933,
+      "learning_rate": 3.9098468715140447e-07,
+      "loss": 0.8953,
+      "step": 26100
+    },
+    {
+      "epoch": 0.6287179981214325,
+      "grad_norm": 1.8735923767089844,
+      "learning_rate": 3.908579251597201e-07,
+      "loss": 0.8738,
+      "step": 26105
+    },
+    {
+      "epoch": 0.6288384191132198,
+      "grad_norm": 1.372200608253479,
+      "learning_rate": 3.9073116316803566e-07,
+      "loss": 0.9413,
+      "step": 26110
+    },
+    {
+      "epoch": 0.6289588401050071,
+      "grad_norm": 1.5144858360290527,
+      "learning_rate": 3.9060440117635125e-07,
+      "loss": 0.9318,
+      "step": 26115
+    },
+    {
+      "epoch": 0.6290792610967944,
+      "grad_norm": 1.5318498611450195,
+      "learning_rate": 3.9047763918466684e-07,
+      "loss": 0.9534,
+      "step": 26120
+    },
+    {
+      "epoch": 0.6291996820885817,
+      "grad_norm": 1.483949065208435,
+      "learning_rate": 3.9035087719298243e-07,
+      "loss": 0.9023,
+      "step": 26125
+    },
+    {
+      "epoch": 0.629320103080369,
+      "grad_norm": 1.7388157844543457,
+      "learning_rate": 3.90224115201298e-07,
+      "loss": 0.9406,
+      "step": 26130
+    },
+    {
+      "epoch": 0.6294405240721562,
+      "grad_norm": 1.5275570154190063,
+      "learning_rate": 3.900973532096136e-07,
+      "loss": 0.8946,
+      "step": 26135
+    },
+    {
+      "epoch": 0.6295609450639436,
+      "grad_norm": 1.5564861297607422,
+      "learning_rate": 3.899705912179292e-07,
+      "loss": 0.8627,
+      "step": 26140
+    },
+    {
+      "epoch": 0.6296813660557309,
+      "grad_norm": 1.4270212650299072,
+      "learning_rate": 3.8984382922624475e-07,
+      "loss": 0.9366,
+      "step": 26145
+    },
+    {
+      "epoch": 0.6298017870475181,
+      "grad_norm": 1.545400619506836,
+      "learning_rate": 3.897170672345604e-07,
+      "loss": 0.9127,
+      "step": 26150
+    },
+    {
+      "epoch": 0.6299222080393054,
+      "grad_norm": 1.3926692008972168,
+      "learning_rate": 3.8959030524287593e-07,
+      "loss": 0.9545,
+      "step": 26155
+    },
+    {
+      "epoch": 0.6300426290310926,
+      "grad_norm": 1.4838570356369019,
+      "learning_rate": 3.894635432511915e-07,
+      "loss": 0.8884,
+      "step": 26160
+    },
+    {
+      "epoch": 0.63016305002288,
+      "grad_norm": 1.6471329927444458,
+      "learning_rate": 3.8933678125950717e-07,
+      "loss": 0.8492,
+      "step": 26165
+    },
+    {
+      "epoch": 0.6302834710146673,
+      "grad_norm": 1.645909070968628,
+      "learning_rate": 3.892100192678227e-07,
+      "loss": 0.9108,
+      "step": 26170
+    },
+    {
+      "epoch": 0.6304038920064545,
+      "grad_norm": 1.407267689704895,
+      "learning_rate": 3.8908325727613835e-07,
+      "loss": 0.9539,
+      "step": 26175
+    },
+    {
+      "epoch": 0.6305243129982419,
+      "grad_norm": 1.7201238870620728,
+      "learning_rate": 3.889564952844539e-07,
+      "loss": 0.8818,
+      "step": 26180
+    },
+    {
+      "epoch": 0.6306447339900292,
+      "grad_norm": 1.3268872499465942,
+      "learning_rate": 3.888297332927695e-07,
+      "loss": 0.9351,
+      "step": 26185
+    },
+    {
+      "epoch": 0.6307651549818164,
+      "grad_norm": 1.6392351388931274,
+      "learning_rate": 3.8870297130108507e-07,
+      "loss": 0.985,
+      "step": 26190
+    },
+    {
+      "epoch": 0.6308855759736037,
+      "grad_norm": 1.5395928621292114,
+      "learning_rate": 3.8857620930940066e-07,
+      "loss": 0.9171,
+      "step": 26195
+    },
+    {
+      "epoch": 0.6310059969653911,
+      "grad_norm": 1.3914449214935303,
+      "learning_rate": 3.884494473177162e-07,
+      "loss": 0.8717,
+      "step": 26200
+    },
+    {
+      "epoch": 0.6311264179571783,
+      "grad_norm": 1.6061489582061768,
+      "learning_rate": 3.8832268532603185e-07,
+      "loss": 0.8801,
+      "step": 26205
+    },
+    {
+      "epoch": 0.6312468389489656,
+      "grad_norm": 1.565010666847229,
+      "learning_rate": 3.8819592333434744e-07,
+      "loss": 0.8897,
+      "step": 26210
+    },
+    {
+      "epoch": 0.6313672599407528,
+      "grad_norm": 1.7284237146377563,
+      "learning_rate": 3.88069161342663e-07,
+      "loss": 0.9043,
+      "step": 26215
+    },
+    {
+      "epoch": 0.6314876809325402,
+      "grad_norm": 1.5633896589279175,
+      "learning_rate": 3.879423993509786e-07,
+      "loss": 0.889,
+      "step": 26220
+    },
+    {
+      "epoch": 0.6316081019243275,
+      "grad_norm": 1.578115701675415,
+      "learning_rate": 3.8781563735929416e-07,
+      "loss": 0.9369,
+      "step": 26225
+    },
+    {
+      "epoch": 0.6317285229161147,
+      "grad_norm": 1.5071097612380981,
+      "learning_rate": 3.8768887536760975e-07,
+      "loss": 0.9064,
+      "step": 26230
+    },
+    {
+      "epoch": 0.631848943907902,
+      "grad_norm": 1.3944430351257324,
+      "learning_rate": 3.8756211337592534e-07,
+      "loss": 0.9425,
+      "step": 26235
+    },
+    {
+      "epoch": 0.6319693648996894,
+      "grad_norm": 1.3645925521850586,
+      "learning_rate": 3.8743535138424094e-07,
+      "loss": 0.9139,
+      "step": 26240
+    },
+    {
+      "epoch": 0.6320897858914766,
+      "grad_norm": 1.6590458154678345,
+      "learning_rate": 3.8730858939255653e-07,
+      "loss": 0.9541,
+      "step": 26245
+    },
+    {
+      "epoch": 0.6322102068832639,
+      "grad_norm": 1.7004053592681885,
+      "learning_rate": 3.871818274008721e-07,
+      "loss": 0.8755,
+      "step": 26250
+    },
+    {
+      "epoch": 0.6323306278750512,
+      "grad_norm": 1.7825452089309692,
+      "learning_rate": 3.870550654091877e-07,
+      "loss": 0.8939,
+      "step": 26255
+    },
+    {
+      "epoch": 0.6324510488668385,
+      "grad_norm": 1.6429630517959595,
+      "learning_rate": 3.8692830341750325e-07,
+      "loss": 0.8728,
+      "step": 26260
+    },
+    {
+      "epoch": 0.6325714698586258,
+      "grad_norm": 1.5073269605636597,
+      "learning_rate": 3.868015414258189e-07,
+      "loss": 0.8906,
+      "step": 26265
+    },
+    {
+      "epoch": 0.632691890850413,
+      "grad_norm": 1.5260565280914307,
+      "learning_rate": 3.8667477943413443e-07,
+      "loss": 0.8797,
+      "step": 26270
+    },
+    {
+      "epoch": 0.6328123118422003,
+      "grad_norm": 1.4282317161560059,
+      "learning_rate": 3.8654801744245e-07,
+      "loss": 0.8967,
+      "step": 26275
+    },
+    {
+      "epoch": 0.6329327328339877,
+      "grad_norm": 1.5447337627410889,
+      "learning_rate": 3.8642125545076567e-07,
+      "loss": 0.8773,
+      "step": 26280
+    },
+    {
+      "epoch": 0.6330531538257749,
+      "grad_norm": 1.4414620399475098,
+      "learning_rate": 3.862944934590812e-07,
+      "loss": 0.9458,
+      "step": 26285
+    },
+    {
+      "epoch": 0.6331735748175622,
+      "grad_norm": 1.4325686693191528,
+      "learning_rate": 3.861677314673968e-07,
+      "loss": 0.8641,
+      "step": 26290
+    },
+    {
+      "epoch": 0.6332939958093495,
+      "grad_norm": 1.695372223854065,
+      "learning_rate": 3.860409694757124e-07,
+      "loss": 0.9215,
+      "step": 26295
+    },
+    {
+      "epoch": 0.6334144168011367,
+      "grad_norm": 1.6153851747512817,
+      "learning_rate": 3.85914207484028e-07,
+      "loss": 0.9339,
+      "step": 26300
+    },
+    {
+      "epoch": 0.6335348377929241,
+      "grad_norm": 1.4820550680160522,
+      "learning_rate": 3.857874454923435e-07,
+      "loss": 0.927,
+      "step": 26305
+    },
+    {
+      "epoch": 0.6336552587847114,
+      "grad_norm": 1.3766353130340576,
+      "learning_rate": 3.8566068350065917e-07,
+      "loss": 0.8868,
+      "step": 26310
+    },
+    {
+      "epoch": 0.6337756797764986,
+      "grad_norm": 1.4582586288452148,
+      "learning_rate": 3.8553392150897476e-07,
+      "loss": 0.8952,
+      "step": 26315
+    },
+    {
+      "epoch": 0.633896100768286,
+      "grad_norm": 1.5797805786132812,
+      "learning_rate": 3.854071595172903e-07,
+      "loss": 0.9397,
+      "step": 26320
+    },
+    {
+      "epoch": 0.6340165217600732,
+      "grad_norm": 1.3752812147140503,
+      "learning_rate": 3.8528039752560594e-07,
+      "loss": 0.9331,
+      "step": 26325
+    },
+    {
+      "epoch": 0.6341369427518605,
+      "grad_norm": 1.3963216543197632,
+      "learning_rate": 3.851536355339215e-07,
+      "loss": 0.9193,
+      "step": 26330
+    },
+    {
+      "epoch": 0.6342573637436478,
+      "grad_norm": 1.6616880893707275,
+      "learning_rate": 3.850268735422371e-07,
+      "loss": 0.9022,
+      "step": 26335
+    },
+    {
+      "epoch": 0.634377784735435,
+      "grad_norm": 1.4412713050842285,
+      "learning_rate": 3.8490011155055267e-07,
+      "loss": 0.9221,
+      "step": 26340
+    },
+    {
+      "epoch": 0.6344982057272224,
+      "grad_norm": 1.553244709968567,
+      "learning_rate": 3.8477334955886826e-07,
+      "loss": 0.9253,
+      "step": 26345
+    },
+    {
+      "epoch": 0.6346186267190097,
+      "grad_norm": 1.510672688484192,
+      "learning_rate": 3.846465875671838e-07,
+      "loss": 0.9447,
+      "step": 26350
+    },
+    {
+      "epoch": 0.6347390477107969,
+      "grad_norm": 1.4998446702957153,
+      "learning_rate": 3.8451982557549944e-07,
+      "loss": 0.8506,
+      "step": 26355
+    },
+    {
+      "epoch": 0.6348594687025843,
+      "grad_norm": 1.5620739459991455,
+      "learning_rate": 3.8439306358381503e-07,
+      "loss": 0.9012,
+      "step": 26360
+    },
+    {
+      "epoch": 0.6349798896943715,
+      "grad_norm": 1.568901777267456,
+      "learning_rate": 3.8426630159213057e-07,
+      "loss": 0.8616,
+      "step": 26365
+    },
+    {
+      "epoch": 0.6351003106861588,
+      "grad_norm": 1.428557276725769,
+      "learning_rate": 3.841395396004462e-07,
+      "loss": 0.8568,
+      "step": 26370
+    },
+    {
+      "epoch": 0.6352207316779461,
+      "grad_norm": 1.4728060960769653,
+      "learning_rate": 3.8401277760876176e-07,
+      "loss": 0.9468,
+      "step": 26375
+    },
+    {
+      "epoch": 0.6353411526697333,
+      "grad_norm": 1.4308539628982544,
+      "learning_rate": 3.8388601561707735e-07,
+      "loss": 0.9365,
+      "step": 26380
+    },
+    {
+      "epoch": 0.6354615736615207,
+      "grad_norm": 1.3621821403503418,
+      "learning_rate": 3.8375925362539294e-07,
+      "loss": 0.8578,
+      "step": 26385
+    },
+    {
+      "epoch": 0.635581994653308,
+      "grad_norm": 1.6784234046936035,
+      "learning_rate": 3.8363249163370853e-07,
+      "loss": 0.918,
+      "step": 26390
+    },
+    {
+      "epoch": 0.6357024156450952,
+      "grad_norm": 1.464553952217102,
+      "learning_rate": 3.835057296420241e-07,
+      "loss": 0.8666,
+      "step": 26395
+    },
+    {
+      "epoch": 0.6358228366368825,
+      "grad_norm": 1.692110538482666,
+      "learning_rate": 3.833789676503397e-07,
+      "loss": 0.8731,
+      "step": 26400
+    },
+    {
+      "epoch": 0.6359432576286699,
+      "grad_norm": 1.6025325059890747,
+      "learning_rate": 3.832522056586553e-07,
+      "loss": 0.8945,
+      "step": 26405
+    },
+    {
+      "epoch": 0.6360636786204571,
+      "grad_norm": 1.3949683904647827,
+      "learning_rate": 3.8312544366697085e-07,
+      "loss": 0.9353,
+      "step": 26410
+    },
+    {
+      "epoch": 0.6361840996122444,
+      "grad_norm": 2.3893911838531494,
+      "learning_rate": 3.829986816752865e-07,
+      "loss": 0.9459,
+      "step": 26415
+    },
+    {
+      "epoch": 0.6363045206040316,
+      "grad_norm": 1.7152431011199951,
+      "learning_rate": 3.8287191968360203e-07,
+      "loss": 0.8687,
+      "step": 26420
+    },
+    {
+      "epoch": 0.636424941595819,
+      "grad_norm": 1.7054212093353271,
+      "learning_rate": 3.827451576919176e-07,
+      "loss": 0.9631,
+      "step": 26425
+    },
+    {
+      "epoch": 0.6365453625876063,
+      "grad_norm": 1.4074687957763672,
+      "learning_rate": 3.8261839570023327e-07,
+      "loss": 0.9171,
+      "step": 26430
+    },
+    {
+      "epoch": 0.6366657835793935,
+      "grad_norm": 1.5717144012451172,
+      "learning_rate": 3.824916337085488e-07,
+      "loss": 0.8751,
+      "step": 26435
+    },
+    {
+      "epoch": 0.6367862045711808,
+      "grad_norm": 1.6245671510696411,
+      "learning_rate": 3.823648717168644e-07,
+      "loss": 0.8708,
+      "step": 26440
+    },
+    {
+      "epoch": 0.6369066255629682,
+      "grad_norm": 1.5715320110321045,
+      "learning_rate": 3.8223810972518e-07,
+      "loss": 0.9241,
+      "step": 26445
+    },
+    {
+      "epoch": 0.6370270465547554,
+      "grad_norm": 1.4957575798034668,
+      "learning_rate": 3.821113477334956e-07,
+      "loss": 0.8936,
+      "step": 26450
+    },
+    {
+      "epoch": 0.6371474675465427,
+      "grad_norm": 1.4049856662750244,
+      "learning_rate": 3.819845857418111e-07,
+      "loss": 0.8851,
+      "step": 26455
+    },
+    {
+      "epoch": 0.63726788853833,
+      "grad_norm": 1.5102617740631104,
+      "learning_rate": 3.8185782375012676e-07,
+      "loss": 0.9093,
+      "step": 26460
+    },
+    {
+      "epoch": 0.6373883095301173,
+      "grad_norm": 1.5172317028045654,
+      "learning_rate": 3.8173106175844236e-07,
+      "loss": 0.8992,
+      "step": 26465
+    },
+    {
+      "epoch": 0.6375087305219046,
+      "grad_norm": 1.799931526184082,
+      "learning_rate": 3.816042997667579e-07,
+      "loss": 0.8988,
+      "step": 26470
+    },
+    {
+      "epoch": 0.6376291515136918,
+      "grad_norm": 1.5349489450454712,
+      "learning_rate": 3.8147753777507354e-07,
+      "loss": 0.8534,
+      "step": 26475
+    },
+    {
+      "epoch": 0.6377495725054791,
+      "grad_norm": 1.5610055923461914,
+      "learning_rate": 3.813507757833891e-07,
+      "loss": 0.8847,
+      "step": 26480
+    },
+    {
+      "epoch": 0.6378699934972665,
+      "grad_norm": 1.4507337808609009,
+      "learning_rate": 3.8122401379170467e-07,
+      "loss": 0.8905,
+      "step": 26485
+    },
+    {
+      "epoch": 0.6379904144890537,
+      "grad_norm": 1.7067121267318726,
+      "learning_rate": 3.8109725180002026e-07,
+      "loss": 0.8915,
+      "step": 26490
+    },
+    {
+      "epoch": 0.638110835480841,
+      "grad_norm": 1.573899745941162,
+      "learning_rate": 3.8097048980833585e-07,
+      "loss": 0.8917,
+      "step": 26495
+    },
+    {
+      "epoch": 0.6382312564726283,
+      "grad_norm": 1.5376611948013306,
+      "learning_rate": 3.808437278166514e-07,
+      "loss": 0.895,
+      "step": 26500
+    },
+    {
+      "epoch": 0.6383516774644156,
+      "grad_norm": 1.6120070219039917,
+      "learning_rate": 3.8071696582496704e-07,
+      "loss": 0.8908,
+      "step": 26505
+    },
+    {
+      "epoch": 0.6384720984562029,
+      "grad_norm": 1.5504016876220703,
+      "learning_rate": 3.8059020383328263e-07,
+      "loss": 0.9152,
+      "step": 26510
+    },
+    {
+      "epoch": 0.6385925194479902,
+      "grad_norm": 1.4211766719818115,
+      "learning_rate": 3.8046344184159817e-07,
+      "loss": 0.8928,
+      "step": 26515
+    },
+    {
+      "epoch": 0.6387129404397774,
+      "grad_norm": 1.503017544746399,
+      "learning_rate": 3.803366798499138e-07,
+      "loss": 0.9085,
+      "step": 26520
+    },
+    {
+      "epoch": 0.6388333614315648,
+      "grad_norm": 1.5840296745300293,
+      "learning_rate": 3.8020991785822935e-07,
+      "loss": 0.948,
+      "step": 26525
+    },
+    {
+      "epoch": 0.638953782423352,
+      "grad_norm": 1.7545835971832275,
+      "learning_rate": 3.8008315586654494e-07,
+      "loss": 0.8774,
+      "step": 26530
+    },
+    {
+      "epoch": 0.6390742034151393,
+      "grad_norm": 1.5773112773895264,
+      "learning_rate": 3.7995639387486053e-07,
+      "loss": 0.9347,
+      "step": 26535
+    },
+    {
+      "epoch": 0.6391946244069266,
+      "grad_norm": 1.4779413938522339,
+      "learning_rate": 3.798296318831761e-07,
+      "loss": 0.8947,
+      "step": 26540
+    },
+    {
+      "epoch": 0.6393150453987139,
+      "grad_norm": 1.5355712175369263,
+      "learning_rate": 3.797028698914917e-07,
+      "loss": 0.922,
+      "step": 26545
+    },
+    {
+      "epoch": 0.6394354663905012,
+      "grad_norm": 1.4067556858062744,
+      "learning_rate": 3.795761078998073e-07,
+      "loss": 0.9262,
+      "step": 26550
+    },
+    {
+      "epoch": 0.6395558873822885,
+      "grad_norm": 1.55918550491333,
+      "learning_rate": 3.794493459081229e-07,
+      "loss": 0.969,
+      "step": 26555
+    },
+    {
+      "epoch": 0.6396763083740757,
+      "grad_norm": 1.6701033115386963,
+      "learning_rate": 3.7932258391643844e-07,
+      "loss": 0.9076,
+      "step": 26560
+    },
+    {
+      "epoch": 0.6397967293658631,
+      "grad_norm": 1.5663307905197144,
+      "learning_rate": 3.791958219247541e-07,
+      "loss": 0.889,
+      "step": 26565
+    },
+    {
+      "epoch": 0.6399171503576504,
+      "grad_norm": 1.6836600303649902,
+      "learning_rate": 3.790690599330696e-07,
+      "loss": 0.893,
+      "step": 26570
+    },
+    {
+      "epoch": 0.6400375713494376,
+      "grad_norm": 1.670732021331787,
+      "learning_rate": 3.789422979413852e-07,
+      "loss": 0.96,
+      "step": 26575
+    },
+    {
+      "epoch": 0.640157992341225,
+      "grad_norm": 1.6099560260772705,
+      "learning_rate": 3.7881553594970086e-07,
+      "loss": 0.8397,
+      "step": 26580
+    },
+    {
+      "epoch": 0.6402784133330122,
+      "grad_norm": 1.6978719234466553,
+      "learning_rate": 3.786887739580164e-07,
+      "loss": 0.8757,
+      "step": 26585
+    },
+    {
+      "epoch": 0.6403988343247995,
+      "grad_norm": 1.590781569480896,
+      "learning_rate": 3.78562011966332e-07,
+      "loss": 0.887,
+      "step": 26590
+    },
+    {
+      "epoch": 0.6405192553165868,
+      "grad_norm": 1.5266937017440796,
+      "learning_rate": 3.784352499746476e-07,
+      "loss": 0.9099,
+      "step": 26595
+    },
+    {
+      "epoch": 0.640639676308374,
+      "grad_norm": 1.675328254699707,
+      "learning_rate": 3.783084879829632e-07,
+      "loss": 0.8897,
+      "step": 26600
+    },
+    {
+      "epoch": 0.6407600973001614,
+      "grad_norm": 1.6646935939788818,
+      "learning_rate": 3.781817259912787e-07,
+      "loss": 0.8669,
+      "step": 26605
+    },
+    {
+      "epoch": 0.6408805182919487,
+      "grad_norm": 1.7408345937728882,
+      "learning_rate": 3.7805496399959436e-07,
+      "loss": 0.8912,
+      "step": 26610
+    },
+    {
+      "epoch": 0.6410009392837359,
+      "grad_norm": 1.5620888471603394,
+      "learning_rate": 3.7792820200790995e-07,
+      "loss": 0.8979,
+      "step": 26615
+    },
+    {
+      "epoch": 0.6411213602755232,
+      "grad_norm": 1.50484299659729,
+      "learning_rate": 3.778014400162255e-07,
+      "loss": 0.9288,
+      "step": 26620
+    },
+    {
+      "epoch": 0.6412417812673105,
+      "grad_norm": 1.5604791641235352,
+      "learning_rate": 3.7767467802454113e-07,
+      "loss": 0.9343,
+      "step": 26625
+    },
+    {
+      "epoch": 0.6413622022590978,
+      "grad_norm": 1.6399976015090942,
+      "learning_rate": 3.7754791603285667e-07,
+      "loss": 0.9413,
+      "step": 26630
+    },
+    {
+      "epoch": 0.6414826232508851,
+      "grad_norm": 1.6691919565200806,
+      "learning_rate": 3.7742115404117226e-07,
+      "loss": 0.9217,
+      "step": 26635
+    },
+    {
+      "epoch": 0.6416030442426723,
+      "grad_norm": 1.4400848150253296,
+      "learning_rate": 3.7729439204948786e-07,
+      "loss": 0.8995,
+      "step": 26640
+    },
+    {
+      "epoch": 0.6417234652344597,
+      "grad_norm": 1.523863673210144,
+      "learning_rate": 3.7716763005780345e-07,
+      "loss": 0.8833,
+      "step": 26645
+    },
+    {
+      "epoch": 0.641843886226247,
+      "grad_norm": 1.4621446132659912,
+      "learning_rate": 3.77040868066119e-07,
+      "loss": 0.923,
+      "step": 26650
+    },
+    {
+      "epoch": 0.6419643072180342,
+      "grad_norm": 1.5442419052124023,
+      "learning_rate": 3.7691410607443463e-07,
+      "loss": 0.9418,
+      "step": 26655
+    },
+    {
+      "epoch": 0.6420847282098215,
+      "grad_norm": 1.7486423254013062,
+      "learning_rate": 3.767873440827502e-07,
+      "loss": 0.8763,
+      "step": 26660
+    },
+    {
+      "epoch": 0.6422051492016089,
+      "grad_norm": 1.7192951440811157,
+      "learning_rate": 3.7666058209106576e-07,
+      "loss": 0.8825,
+      "step": 26665
+    },
+    {
+      "epoch": 0.6423255701933961,
+      "grad_norm": 1.8942385911941528,
+      "learning_rate": 3.765338200993814e-07,
+      "loss": 0.9291,
+      "step": 26670
+    },
+    {
+      "epoch": 0.6424459911851834,
+      "grad_norm": 1.748542070388794,
+      "learning_rate": 3.7640705810769695e-07,
+      "loss": 0.9082,
+      "step": 26675
+    },
+    {
+      "epoch": 0.6425664121769706,
+      "grad_norm": 1.3566148281097412,
+      "learning_rate": 3.762802961160126e-07,
+      "loss": 0.8808,
+      "step": 26680
+    },
+    {
+      "epoch": 0.642686833168758,
+      "grad_norm": 1.5772252082824707,
+      "learning_rate": 3.7615353412432813e-07,
+      "loss": 0.8409,
+      "step": 26685
+    },
+    {
+      "epoch": 0.6428072541605453,
+      "grad_norm": 1.758379340171814,
+      "learning_rate": 3.760267721326437e-07,
+      "loss": 0.9144,
+      "step": 26690
+    },
+    {
+      "epoch": 0.6429276751523325,
+      "grad_norm": 1.517451524734497,
+      "learning_rate": 3.7590001014095937e-07,
+      "loss": 0.9186,
+      "step": 26695
+    },
+    {
+      "epoch": 0.6430480961441198,
+      "grad_norm": 1.5986350774765015,
+      "learning_rate": 3.757732481492749e-07,
+      "loss": 0.9321,
+      "step": 26700
+    },
+    {
+      "epoch": 0.6431685171359072,
+      "grad_norm": 1.535396695137024,
+      "learning_rate": 3.756464861575905e-07,
+      "loss": 0.9007,
+      "step": 26705
+    },
+    {
+      "epoch": 0.6432889381276944,
+      "grad_norm": 1.7621718645095825,
+      "learning_rate": 3.755197241659061e-07,
+      "loss": 0.9153,
+      "step": 26710
+    },
+    {
+      "epoch": 0.6434093591194817,
+      "grad_norm": 1.6428961753845215,
+      "learning_rate": 3.753929621742217e-07,
+      "loss": 0.9142,
+      "step": 26715
+    },
+    {
+      "epoch": 0.643529780111269,
+      "grad_norm": 1.5684250593185425,
+      "learning_rate": 3.752662001825372e-07,
+      "loss": 0.9805,
+      "step": 26720
+    },
+    {
+      "epoch": 0.6436502011030563,
+      "grad_norm": 1.5175715684890747,
+      "learning_rate": 3.7513943819085286e-07,
+      "loss": 0.868,
+      "step": 26725
+    },
+    {
+      "epoch": 0.6437706220948436,
+      "grad_norm": 1.5524253845214844,
+      "learning_rate": 3.7501267619916846e-07,
+      "loss": 0.869,
+      "step": 26730
+    },
+    {
+      "epoch": 0.6438910430866308,
+      "grad_norm": 1.3661057949066162,
+      "learning_rate": 3.74885914207484e-07,
+      "loss": 0.9048,
+      "step": 26735
+    },
+    {
+      "epoch": 0.6440114640784181,
+      "grad_norm": 1.640135407447815,
+      "learning_rate": 3.7475915221579964e-07,
+      "loss": 0.9211,
+      "step": 26740
+    },
+    {
+      "epoch": 0.6441318850702055,
+      "grad_norm": 1.5012540817260742,
+      "learning_rate": 3.746323902241152e-07,
+      "loss": 0.883,
+      "step": 26745
+    },
+    {
+      "epoch": 0.6442523060619927,
+      "grad_norm": 2.04201078414917,
+      "learning_rate": 3.7450562823243077e-07,
+      "loss": 0.9226,
+      "step": 26750
+    },
+    {
+      "epoch": 0.64437272705378,
+      "grad_norm": 1.4492322206497192,
+      "learning_rate": 3.7437886624074636e-07,
+      "loss": 0.8431,
+      "step": 26755
+    },
+    {
+      "epoch": 0.6444931480455673,
+      "grad_norm": 1.5902260541915894,
+      "learning_rate": 3.7425210424906195e-07,
+      "loss": 0.8626,
+      "step": 26760
+    },
+    {
+      "epoch": 0.6446135690373546,
+      "grad_norm": 1.510667324066162,
+      "learning_rate": 3.741253422573775e-07,
+      "loss": 0.8645,
+      "step": 26765
+    },
+    {
+      "epoch": 0.6447339900291419,
+      "grad_norm": 1.7094923257827759,
+      "learning_rate": 3.7399858026569314e-07,
+      "loss": 0.9253,
+      "step": 26770
+    },
+    {
+      "epoch": 0.6448544110209292,
+      "grad_norm": 1.5884010791778564,
+      "learning_rate": 3.7387181827400873e-07,
+      "loss": 0.9593,
+      "step": 26775
+    },
+    {
+      "epoch": 0.6449748320127164,
+      "grad_norm": 1.6673965454101562,
+      "learning_rate": 3.7374505628232427e-07,
+      "loss": 0.8712,
+      "step": 26780
+    },
+    {
+      "epoch": 0.6450952530045038,
+      "grad_norm": 1.4968947172164917,
+      "learning_rate": 3.736182942906399e-07,
+      "loss": 0.8972,
+      "step": 26785
+    },
+    {
+      "epoch": 0.645215673996291,
+      "grad_norm": 1.5666877031326294,
+      "learning_rate": 3.7349153229895545e-07,
+      "loss": 0.9499,
+      "step": 26790
+    },
+    {
+      "epoch": 0.6453360949880783,
+      "grad_norm": 1.5352009534835815,
+      "learning_rate": 3.7336477030727104e-07,
+      "loss": 0.8903,
+      "step": 26795
+    },
+    {
+      "epoch": 0.6454565159798656,
+      "grad_norm": 1.6426624059677124,
+      "learning_rate": 3.7323800831558663e-07,
+      "loss": 0.8634,
+      "step": 26800
+    },
+    {
+      "epoch": 0.6455769369716529,
+      "grad_norm": 1.7372010946273804,
+      "learning_rate": 3.7311124632390223e-07,
+      "loss": 0.9357,
+      "step": 26805
+    },
+    {
+      "epoch": 0.6456973579634402,
+      "grad_norm": 1.5004019737243652,
+      "learning_rate": 3.729844843322178e-07,
+      "loss": 0.9616,
+      "step": 26810
+    },
+    {
+      "epoch": 0.6458177789552275,
+      "grad_norm": 1.5163060426712036,
+      "learning_rate": 3.728577223405334e-07,
+      "loss": 0.9188,
+      "step": 26815
+    },
+    {
+      "epoch": 0.6459381999470147,
+      "grad_norm": 1.4289931058883667,
+      "learning_rate": 3.72730960348849e-07,
+      "loss": 0.8999,
+      "step": 26820
+    },
+    {
+      "epoch": 0.6460586209388021,
+      "grad_norm": 1.5710238218307495,
+      "learning_rate": 3.7260419835716454e-07,
+      "loss": 0.9245,
+      "step": 26825
+    },
+    {
+      "epoch": 0.6461790419305894,
+      "grad_norm": 1.4990209341049194,
+      "learning_rate": 3.724774363654802e-07,
+      "loss": 0.9087,
+      "step": 26830
+    },
+    {
+      "epoch": 0.6462994629223766,
+      "grad_norm": 1.4621906280517578,
+      "learning_rate": 3.723506743737957e-07,
+      "loss": 0.9007,
+      "step": 26835
+    },
+    {
+      "epoch": 0.6464198839141639,
+      "grad_norm": 1.655481219291687,
+      "learning_rate": 3.722239123821113e-07,
+      "loss": 0.9313,
+      "step": 26840
+    },
+    {
+      "epoch": 0.6465403049059512,
+      "grad_norm": 1.8018776178359985,
+      "learning_rate": 3.7209715039042696e-07,
+      "loss": 0.9167,
+      "step": 26845
+    },
+    {
+      "epoch": 0.6466607258977385,
+      "grad_norm": 1.529486894607544,
+      "learning_rate": 3.719703883987425e-07,
+      "loss": 0.9216,
+      "step": 26850
+    },
+    {
+      "epoch": 0.6467811468895258,
+      "grad_norm": 1.5701395273208618,
+      "learning_rate": 3.718436264070581e-07,
+      "loss": 0.8683,
+      "step": 26855
+    },
+    {
+      "epoch": 0.646901567881313,
+      "grad_norm": 1.528743863105774,
+      "learning_rate": 3.717168644153737e-07,
+      "loss": 0.9513,
+      "step": 26860
+    },
+    {
+      "epoch": 0.6470219888731004,
+      "grad_norm": 1.7234257459640503,
+      "learning_rate": 3.715901024236893e-07,
+      "loss": 0.9194,
+      "step": 26865
+    },
+    {
+      "epoch": 0.6471424098648877,
+      "grad_norm": 1.337449550628662,
+      "learning_rate": 3.714633404320048e-07,
+      "loss": 0.8864,
+      "step": 26870
+    },
+    {
+      "epoch": 0.6472628308566749,
+      "grad_norm": 1.568869948387146,
+      "learning_rate": 3.7133657844032046e-07,
+      "loss": 0.8847,
+      "step": 26875
+    },
+    {
+      "epoch": 0.6473832518484622,
+      "grad_norm": 1.664362907409668,
+      "learning_rate": 3.7120981644863605e-07,
+      "loss": 0.8685,
+      "step": 26880
+    },
+    {
+      "epoch": 0.6475036728402496,
+      "grad_norm": 1.5255695581436157,
+      "learning_rate": 3.710830544569516e-07,
+      "loss": 0.943,
+      "step": 26885
+    },
+    {
+      "epoch": 0.6476240938320368,
+      "grad_norm": 1.3643174171447754,
+      "learning_rate": 3.7095629246526723e-07,
+      "loss": 0.8955,
+      "step": 26890
+    },
+    {
+      "epoch": 0.6477445148238241,
+      "grad_norm": 1.5007296800613403,
+      "learning_rate": 3.7082953047358277e-07,
+      "loss": 0.924,
+      "step": 26895
+    },
+    {
+      "epoch": 0.6478649358156113,
+      "grad_norm": 1.6384427547454834,
+      "learning_rate": 3.7070276848189836e-07,
+      "loss": 0.9658,
+      "step": 26900
+    },
+    {
+      "epoch": 0.6479853568073987,
+      "grad_norm": 1.6323484182357788,
+      "learning_rate": 3.7057600649021396e-07,
+      "loss": 0.9548,
+      "step": 26905
+    },
+    {
+      "epoch": 0.648105777799186,
+      "grad_norm": 1.6146575212478638,
+      "learning_rate": 3.7044924449852955e-07,
+      "loss": 0.8944,
+      "step": 26910
+    },
+    {
+      "epoch": 0.6482261987909732,
+      "grad_norm": 1.5658189058303833,
+      "learning_rate": 3.703224825068451e-07,
+      "loss": 0.9234,
+      "step": 26915
+    },
+    {
+      "epoch": 0.6483466197827605,
+      "grad_norm": 1.5981191396713257,
+      "learning_rate": 3.7019572051516073e-07,
+      "loss": 0.8716,
+      "step": 26920
+    },
+    {
+      "epoch": 0.6484670407745479,
+      "grad_norm": 1.6653273105621338,
+      "learning_rate": 3.700689585234763e-07,
+      "loss": 0.8921,
+      "step": 26925
+    },
+    {
+      "epoch": 0.6485874617663351,
+      "grad_norm": 1.453226089477539,
+      "learning_rate": 3.6994219653179186e-07,
+      "loss": 0.8319,
+      "step": 26930
+    },
+    {
+      "epoch": 0.6487078827581224,
+      "grad_norm": 1.4281163215637207,
+      "learning_rate": 3.698154345401075e-07,
+      "loss": 0.8778,
+      "step": 26935
+    },
+    {
+      "epoch": 0.6488283037499096,
+      "grad_norm": 1.6727575063705444,
+      "learning_rate": 3.6968867254842305e-07,
+      "loss": 0.8621,
+      "step": 26940
+    },
+    {
+      "epoch": 0.648948724741697,
+      "grad_norm": 1.6255943775177002,
+      "learning_rate": 3.6956191055673864e-07,
+      "loss": 0.8967,
+      "step": 26945
+    },
+    {
+      "epoch": 0.6490691457334843,
+      "grad_norm": 1.9116973876953125,
+      "learning_rate": 3.6943514856505423e-07,
+      "loss": 0.886,
+      "step": 26950
+    },
+    {
+      "epoch": 0.6491895667252715,
+      "grad_norm": 1.4924372434616089,
+      "learning_rate": 3.693083865733698e-07,
+      "loss": 0.9099,
+      "step": 26955
+    },
+    {
+      "epoch": 0.6493099877170588,
+      "grad_norm": 1.5505237579345703,
+      "learning_rate": 3.691816245816854e-07,
+      "loss": 0.9415,
+      "step": 26960
+    },
+    {
+      "epoch": 0.6494304087088462,
+      "grad_norm": 1.570006251335144,
+      "learning_rate": 3.69054862590001e-07,
+      "loss": 0.9139,
+      "step": 26965
+    },
+    {
+      "epoch": 0.6495508297006334,
+      "grad_norm": 1.5226932764053345,
+      "learning_rate": 3.689281005983166e-07,
+      "loss": 0.8751,
+      "step": 26970
+    },
+    {
+      "epoch": 0.6496712506924207,
+      "grad_norm": 1.5911332368850708,
+      "learning_rate": 3.6880133860663214e-07,
+      "loss": 0.8287,
+      "step": 26975
+    },
+    {
+      "epoch": 0.649791671684208,
+      "grad_norm": 1.4197216033935547,
+      "learning_rate": 3.686745766149478e-07,
+      "loss": 0.9192,
+      "step": 26980
+    },
+    {
+      "epoch": 0.6499120926759953,
+      "grad_norm": 1.5569384098052979,
+      "learning_rate": 3.685478146232633e-07,
+      "loss": 0.9194,
+      "step": 26985
+    },
+    {
+      "epoch": 0.6500325136677826,
+      "grad_norm": 1.6305328607559204,
+      "learning_rate": 3.684210526315789e-07,
+      "loss": 0.8704,
+      "step": 26990
+    },
+    {
+      "epoch": 0.6501529346595698,
+      "grad_norm": 1.5222176313400269,
+      "learning_rate": 3.6829429063989456e-07,
+      "loss": 0.9204,
+      "step": 26995
+    },
+    {
+      "epoch": 0.6502733556513571,
+      "grad_norm": 1.4481796026229858,
+      "learning_rate": 3.681675286482101e-07,
+      "loss": 0.8853,
+      "step": 27000
+    },
+    {
+      "epoch": 0.6503937766431445,
+      "grad_norm": 1.665404200553894,
+      "learning_rate": 3.680407666565257e-07,
+      "loss": 0.9072,
+      "step": 27005
+    },
+    {
+      "epoch": 0.6505141976349317,
+      "grad_norm": 1.5380555391311646,
+      "learning_rate": 3.679140046648413e-07,
+      "loss": 0.8119,
+      "step": 27010
+    },
+    {
+      "epoch": 0.650634618626719,
+      "grad_norm": 1.5314487218856812,
+      "learning_rate": 3.6778724267315687e-07,
+      "loss": 0.8875,
+      "step": 27015
+    },
+    {
+      "epoch": 0.6507550396185063,
+      "grad_norm": 1.5317686796188354,
+      "learning_rate": 3.676604806814724e-07,
+      "loss": 0.8951,
+      "step": 27020
+    },
+    {
+      "epoch": 0.6508754606102936,
+      "grad_norm": 1.418816089630127,
+      "learning_rate": 3.6753371868978805e-07,
+      "loss": 0.8695,
+      "step": 27025
+    },
+    {
+      "epoch": 0.6509958816020809,
+      "grad_norm": 1.6045305728912354,
+      "learning_rate": 3.6740695669810365e-07,
+      "loss": 0.8818,
+      "step": 27030
+    },
+    {
+      "epoch": 0.6511163025938682,
+      "grad_norm": 1.4292452335357666,
+      "learning_rate": 3.672801947064192e-07,
+      "loss": 0.9099,
+      "step": 27035
+    },
+    {
+      "epoch": 0.6512367235856554,
+      "grad_norm": 1.8049204349517822,
+      "learning_rate": 3.6715343271473483e-07,
+      "loss": 0.9229,
+      "step": 27040
+    },
+    {
+      "epoch": 0.6513571445774428,
+      "grad_norm": 1.461598515510559,
+      "learning_rate": 3.6702667072305037e-07,
+      "loss": 0.912,
+      "step": 27045
+    },
+    {
+      "epoch": 0.65147756556923,
+      "grad_norm": 1.686318039894104,
+      "learning_rate": 3.6689990873136596e-07,
+      "loss": 0.8798,
+      "step": 27050
+    },
+    {
+      "epoch": 0.6515979865610173,
+      "grad_norm": 1.6203144788742065,
+      "learning_rate": 3.6677314673968155e-07,
+      "loss": 0.8777,
+      "step": 27055
+    },
+    {
+      "epoch": 0.6517184075528046,
+      "grad_norm": 1.5647941827774048,
+      "learning_rate": 3.6664638474799714e-07,
+      "loss": 0.9458,
+      "step": 27060
+    },
+    {
+      "epoch": 0.6518388285445919,
+      "grad_norm": 1.4100810289382935,
+      "learning_rate": 3.665196227563127e-07,
+      "loss": 0.869,
+      "step": 27065
+    },
+    {
+      "epoch": 0.6519592495363792,
+      "grad_norm": 1.5744898319244385,
+      "learning_rate": 3.6639286076462833e-07,
+      "loss": 0.8851,
+      "step": 27070
+    },
+    {
+      "epoch": 0.6520796705281665,
+      "grad_norm": 1.6459345817565918,
+      "learning_rate": 3.662660987729439e-07,
+      "loss": 0.8938,
+      "step": 27075
+    },
+    {
+      "epoch": 0.6522000915199537,
+      "grad_norm": 1.555381178855896,
+      "learning_rate": 3.6613933678125946e-07,
+      "loss": 0.9509,
+      "step": 27080
+    },
+    {
+      "epoch": 0.6523205125117411,
+      "grad_norm": 1.4916878938674927,
+      "learning_rate": 3.660125747895751e-07,
+      "loss": 0.8233,
+      "step": 27085
+    },
+    {
+      "epoch": 0.6524409335035284,
+      "grad_norm": 1.42539381980896,
+      "learning_rate": 3.6588581279789064e-07,
+      "loss": 0.9541,
+      "step": 27090
+    },
+    {
+      "epoch": 0.6525613544953156,
+      "grad_norm": 1.3812246322631836,
+      "learning_rate": 3.6575905080620623e-07,
+      "loss": 0.9144,
+      "step": 27095
+    },
+    {
+      "epoch": 0.6526817754871029,
+      "grad_norm": 1.936347484588623,
+      "learning_rate": 3.656322888145218e-07,
+      "loss": 0.8679,
+      "step": 27100
+    },
+    {
+      "epoch": 0.6528021964788902,
+      "grad_norm": 1.5986363887786865,
+      "learning_rate": 3.655055268228374e-07,
+      "loss": 0.8995,
+      "step": 27105
+    },
+    {
+      "epoch": 0.6529226174706775,
+      "grad_norm": 1.5964977741241455,
+      "learning_rate": 3.65378764831153e-07,
+      "loss": 0.8813,
+      "step": 27110
+    },
+    {
+      "epoch": 0.6530430384624648,
+      "grad_norm": 1.3528989553451538,
+      "learning_rate": 3.652520028394686e-07,
+      "loss": 0.8285,
+      "step": 27115
+    },
+    {
+      "epoch": 0.653163459454252,
+      "grad_norm": 1.4709807634353638,
+      "learning_rate": 3.651252408477842e-07,
+      "loss": 0.9106,
+      "step": 27120
+    },
+    {
+      "epoch": 0.6532838804460394,
+      "grad_norm": 1.7439950704574585,
+      "learning_rate": 3.6499847885609973e-07,
+      "loss": 0.9014,
+      "step": 27125
+    },
+    {
+      "epoch": 0.6534043014378267,
+      "grad_norm": 1.5640136003494263,
+      "learning_rate": 3.648717168644154e-07,
+      "loss": 0.8836,
+      "step": 27130
+    },
+    {
+      "epoch": 0.6535247224296139,
+      "grad_norm": 1.639180064201355,
+      "learning_rate": 3.647449548727309e-07,
+      "loss": 0.8945,
+      "step": 27135
+    },
+    {
+      "epoch": 0.6536451434214012,
+      "grad_norm": 1.587928056716919,
+      "learning_rate": 3.646181928810465e-07,
+      "loss": 0.8792,
+      "step": 27140
+    },
+    {
+      "epoch": 0.6537655644131886,
+      "grad_norm": 1.5554908514022827,
+      "learning_rate": 3.6449143088936215e-07,
+      "loss": 0.9404,
+      "step": 27145
+    },
+    {
+      "epoch": 0.6538859854049758,
+      "grad_norm": 1.4645404815673828,
+      "learning_rate": 3.643646688976777e-07,
+      "loss": 0.8887,
+      "step": 27150
+    },
+    {
+      "epoch": 0.6540064063967631,
+      "grad_norm": 1.6191918849945068,
+      "learning_rate": 3.642379069059933e-07,
+      "loss": 0.8678,
+      "step": 27155
+    },
+    {
+      "epoch": 0.6541268273885503,
+      "grad_norm": 1.373336672782898,
+      "learning_rate": 3.6411114491430887e-07,
+      "loss": 0.9033,
+      "step": 27160
+    },
+    {
+      "epoch": 0.6542472483803377,
+      "grad_norm": 1.5440839529037476,
+      "learning_rate": 3.6398438292262447e-07,
+      "loss": 0.8413,
+      "step": 27165
+    },
+    {
+      "epoch": 0.654367669372125,
+      "grad_norm": 1.6653863191604614,
+      "learning_rate": 3.6385762093094e-07,
+      "loss": 0.926,
+      "step": 27170
+    },
+    {
+      "epoch": 0.6544880903639122,
+      "grad_norm": 1.4619879722595215,
+      "learning_rate": 3.6373085893925565e-07,
+      "loss": 0.919,
+      "step": 27175
+    },
+    {
+      "epoch": 0.6546085113556995,
+      "grad_norm": 1.5410512685775757,
+      "learning_rate": 3.6360409694757124e-07,
+      "loss": 0.8741,
+      "step": 27180
+    },
+    {
+      "epoch": 0.6547289323474869,
+      "grad_norm": 1.6692581176757812,
+      "learning_rate": 3.634773349558868e-07,
+      "loss": 0.9221,
+      "step": 27185
+    },
+    {
+      "epoch": 0.6548493533392741,
+      "grad_norm": 1.6221915483474731,
+      "learning_rate": 3.633505729642024e-07,
+      "loss": 0.9062,
+      "step": 27190
+    },
+    {
+      "epoch": 0.6549697743310614,
+      "grad_norm": 1.5964298248291016,
+      "learning_rate": 3.6322381097251796e-07,
+      "loss": 0.8578,
+      "step": 27195
+    },
+    {
+      "epoch": 0.6550901953228487,
+      "grad_norm": 1.405503511428833,
+      "learning_rate": 3.630970489808336e-07,
+      "loss": 0.9153,
+      "step": 27200
+    },
+    {
+      "epoch": 0.655210616314636,
+      "grad_norm": 1.5333468914031982,
+      "learning_rate": 3.6297028698914915e-07,
+      "loss": 0.862,
+      "step": 27205
+    },
+    {
+      "epoch": 0.6553310373064233,
+      "grad_norm": 1.7078754901885986,
+      "learning_rate": 3.6284352499746474e-07,
+      "loss": 0.861,
+      "step": 27210
+    },
+    {
+      "epoch": 0.6554514582982105,
+      "grad_norm": 1.62969172000885,
+      "learning_rate": 3.627167630057804e-07,
+      "loss": 0.8841,
+      "step": 27215
+    },
+    {
+      "epoch": 0.6555718792899978,
+      "grad_norm": 1.381977915763855,
+      "learning_rate": 3.625900010140959e-07,
+      "loss": 0.9321,
+      "step": 27220
+    },
+    {
+      "epoch": 0.6556923002817852,
+      "grad_norm": 1.4872454404830933,
+      "learning_rate": 3.624632390224115e-07,
+      "loss": 0.8616,
+      "step": 27225
+    },
+    {
+      "epoch": 0.6558127212735724,
+      "grad_norm": 1.4531726837158203,
+      "learning_rate": 3.623364770307271e-07,
+      "loss": 0.9679,
+      "step": 27230
+    },
+    {
+      "epoch": 0.6559331422653597,
+      "grad_norm": 1.531635046005249,
+      "learning_rate": 3.622097150390427e-07,
+      "loss": 0.9636,
+      "step": 27235
+    },
+    {
+      "epoch": 0.656053563257147,
+      "grad_norm": 1.585182785987854,
+      "learning_rate": 3.6208295304735824e-07,
+      "loss": 0.867,
+      "step": 27240
+    },
+    {
+      "epoch": 0.6561739842489343,
+      "grad_norm": 1.6876927614212036,
+      "learning_rate": 3.619561910556739e-07,
+      "loss": 0.9311,
+      "step": 27245
+    },
+    {
+      "epoch": 0.6562944052407216,
+      "grad_norm": 1.6832166910171509,
+      "learning_rate": 3.618294290639894e-07,
+      "loss": 0.9061,
+      "step": 27250
+    },
+    {
+      "epoch": 0.6564148262325088,
+      "grad_norm": 1.585931658744812,
+      "learning_rate": 3.61702667072305e-07,
+      "loss": 0.8733,
+      "step": 27255
+    },
+    {
+      "epoch": 0.6565352472242961,
+      "grad_norm": 1.6085927486419678,
+      "learning_rate": 3.6157590508062066e-07,
+      "loss": 0.8653,
+      "step": 27260
+    },
+    {
+      "epoch": 0.6566556682160835,
+      "grad_norm": 1.4381110668182373,
+      "learning_rate": 3.614491430889362e-07,
+      "loss": 0.884,
+      "step": 27265
+    },
+    {
+      "epoch": 0.6567760892078707,
+      "grad_norm": 1.5061804056167603,
+      "learning_rate": 3.613223810972518e-07,
+      "loss": 0.912,
+      "step": 27270
+    },
+    {
+      "epoch": 0.656896510199658,
+      "grad_norm": 1.5456966161727905,
+      "learning_rate": 3.611956191055674e-07,
+      "loss": 0.873,
+      "step": 27275
+    },
+    {
+      "epoch": 0.6570169311914453,
+      "grad_norm": 1.3770190477371216,
+      "learning_rate": 3.6106885711388297e-07,
+      "loss": 0.9152,
+      "step": 27280
+    },
+    {
+      "epoch": 0.6571373521832325,
+      "grad_norm": 1.554132342338562,
+      "learning_rate": 3.609420951221985e-07,
+      "loss": 0.9046,
+      "step": 27285
+    },
+    {
+      "epoch": 0.6572577731750199,
+      "grad_norm": 1.5005851984024048,
+      "learning_rate": 3.6081533313051415e-07,
+      "loss": 0.8832,
+      "step": 27290
+    },
+    {
+      "epoch": 0.6573781941668072,
+      "grad_norm": 1.5317542552947998,
+      "learning_rate": 3.6068857113882975e-07,
+      "loss": 0.8734,
+      "step": 27295
+    },
+    {
+      "epoch": 0.6574986151585944,
+      "grad_norm": 1.7083379030227661,
+      "learning_rate": 3.605618091471453e-07,
+      "loss": 0.9485,
+      "step": 27300
+    },
+    {
+      "epoch": 0.6576190361503818,
+      "grad_norm": 1.5612186193466187,
+      "learning_rate": 3.6043504715546093e-07,
+      "loss": 0.9064,
+      "step": 27305
+    },
+    {
+      "epoch": 0.657739457142169,
+      "grad_norm": 1.361822247505188,
+      "learning_rate": 3.6030828516377647e-07,
+      "loss": 0.9019,
+      "step": 27310
+    },
+    {
+      "epoch": 0.6578598781339563,
+      "grad_norm": 2.116149663925171,
+      "learning_rate": 3.6018152317209206e-07,
+      "loss": 0.869,
+      "step": 27315
+    },
+    {
+      "epoch": 0.6579802991257436,
+      "grad_norm": 1.8238283395767212,
+      "learning_rate": 3.6005476118040765e-07,
+      "loss": 0.8696,
+      "step": 27320
+    },
+    {
+      "epoch": 0.6581007201175308,
+      "grad_norm": 1.6186177730560303,
+      "learning_rate": 3.5992799918872324e-07,
+      "loss": 0.8781,
+      "step": 27325
+    },
+    {
+      "epoch": 0.6582211411093182,
+      "grad_norm": 1.4667737483978271,
+      "learning_rate": 3.5980123719703884e-07,
+      "loss": 0.8702,
+      "step": 27330
+    },
+    {
+      "epoch": 0.6583415621011055,
+      "grad_norm": 1.6477582454681396,
+      "learning_rate": 3.5967447520535443e-07,
+      "loss": 0.8977,
+      "step": 27335
+    },
+    {
+      "epoch": 0.6584619830928927,
+      "grad_norm": 1.6902414560317993,
+      "learning_rate": 3.5954771321367e-07,
+      "loss": 0.9111,
+      "step": 27340
+    },
+    {
+      "epoch": 0.65858240408468,
+      "grad_norm": 1.6111053228378296,
+      "learning_rate": 3.5942095122198556e-07,
+      "loss": 0.9084,
+      "step": 27345
+    },
+    {
+      "epoch": 0.6587028250764674,
+      "grad_norm": 1.544984221458435,
+      "learning_rate": 3.592941892303012e-07,
+      "loss": 0.9529,
+      "step": 27350
+    },
+    {
+      "epoch": 0.6588232460682546,
+      "grad_norm": 1.3802094459533691,
+      "learning_rate": 3.5916742723861674e-07,
+      "loss": 0.8947,
+      "step": 27355
+    },
+    {
+      "epoch": 0.6589436670600419,
+      "grad_norm": 1.6167811155319214,
+      "learning_rate": 3.5904066524693233e-07,
+      "loss": 0.8962,
+      "step": 27360
+    },
+    {
+      "epoch": 0.6590640880518291,
+      "grad_norm": 1.3995361328125,
+      "learning_rate": 3.58913903255248e-07,
+      "loss": 0.8824,
+      "step": 27365
+    },
+    {
+      "epoch": 0.6591845090436165,
+      "grad_norm": 1.7289831638336182,
+      "learning_rate": 3.587871412635635e-07,
+      "loss": 0.9314,
+      "step": 27370
+    },
+    {
+      "epoch": 0.6593049300354038,
+      "grad_norm": 1.5664397478103638,
+      "learning_rate": 3.586603792718791e-07,
+      "loss": 0.8917,
+      "step": 27375
+    },
+    {
+      "epoch": 0.659425351027191,
+      "grad_norm": 2.464617967605591,
+      "learning_rate": 3.585336172801947e-07,
+      "loss": 0.8862,
+      "step": 27380
+    },
+    {
+      "epoch": 0.6595457720189783,
+      "grad_norm": 1.5785813331604004,
+      "learning_rate": 3.584068552885103e-07,
+      "loss": 0.8589,
+      "step": 27385
+    },
+    {
+      "epoch": 0.6596661930107657,
+      "grad_norm": 1.6598893404006958,
+      "learning_rate": 3.5828009329682583e-07,
+      "loss": 0.8989,
+      "step": 27390
+    },
+    {
+      "epoch": 0.6597866140025529,
+      "grad_norm": 1.4237953424453735,
+      "learning_rate": 3.581533313051415e-07,
+      "loss": 0.9035,
+      "step": 27395
+    },
+    {
+      "epoch": 0.6599070349943402,
+      "grad_norm": 1.4129544496536255,
+      "learning_rate": 3.58026569313457e-07,
+      "loss": 0.8987,
+      "step": 27400
+    },
+    {
+      "epoch": 0.6600274559861276,
+      "grad_norm": 1.421212911605835,
+      "learning_rate": 3.578998073217726e-07,
+      "loss": 0.9319,
+      "step": 27405
+    },
+    {
+      "epoch": 0.6601478769779148,
+      "grad_norm": 1.5244791507720947,
+      "learning_rate": 3.5777304533008825e-07,
+      "loss": 0.903,
+      "step": 27410
+    },
+    {
+      "epoch": 0.6602682979697021,
+      "grad_norm": 1.5054500102996826,
+      "learning_rate": 3.576462833384038e-07,
+      "loss": 0.9038,
+      "step": 27415
+    },
+    {
+      "epoch": 0.6603887189614893,
+      "grad_norm": 1.5401767492294312,
+      "learning_rate": 3.575195213467194e-07,
+      "loss": 0.9308,
+      "step": 27420
+    },
+    {
+      "epoch": 0.6605091399532766,
+      "grad_norm": 1.5578993558883667,
+      "learning_rate": 3.57392759355035e-07,
+      "loss": 0.8465,
+      "step": 27425
+    },
+    {
+      "epoch": 0.660629560945064,
+      "grad_norm": 1.6200608015060425,
+      "learning_rate": 3.5726599736335057e-07,
+      "loss": 0.9103,
+      "step": 27430
+    },
+    {
+      "epoch": 0.6607499819368512,
+      "grad_norm": 1.6002734899520874,
+      "learning_rate": 3.571392353716661e-07,
+      "loss": 0.8838,
+      "step": 27435
+    },
+    {
+      "epoch": 0.6608704029286385,
+      "grad_norm": 1.7070163488388062,
+      "learning_rate": 3.5701247337998175e-07,
+      "loss": 0.8847,
+      "step": 27440
+    },
+    {
+      "epoch": 0.6609908239204259,
+      "grad_norm": 1.6092935800552368,
+      "learning_rate": 3.5688571138829734e-07,
+      "loss": 0.8689,
+      "step": 27445
+    },
+    {
+      "epoch": 0.6611112449122131,
+      "grad_norm": 1.691953420639038,
+      "learning_rate": 3.567589493966129e-07,
+      "loss": 0.8966,
+      "step": 27450
+    },
+    {
+      "epoch": 0.6612316659040004,
+      "grad_norm": 1.625665545463562,
+      "learning_rate": 3.566321874049285e-07,
+      "loss": 0.9179,
+      "step": 27455
+    },
+    {
+      "epoch": 0.6613520868957877,
+      "grad_norm": 1.5149598121643066,
+      "learning_rate": 3.5650542541324406e-07,
+      "loss": 0.9011,
+      "step": 27460
+    },
+    {
+      "epoch": 0.661472507887575,
+      "grad_norm": 1.4930402040481567,
+      "learning_rate": 3.5637866342155966e-07,
+      "loss": 0.9046,
+      "step": 27465
+    },
+    {
+      "epoch": 0.6615929288793623,
+      "grad_norm": 1.608102560043335,
+      "learning_rate": 3.5625190142987525e-07,
+      "loss": 0.8862,
+      "step": 27470
+    },
+    {
+      "epoch": 0.6617133498711495,
+      "grad_norm": 1.6462239027023315,
+      "learning_rate": 3.5612513943819084e-07,
+      "loss": 0.9726,
+      "step": 27475
+    },
+    {
+      "epoch": 0.6618337708629368,
+      "grad_norm": 1.5935025215148926,
+      "learning_rate": 3.559983774465064e-07,
+      "loss": 0.9017,
+      "step": 27480
+    },
+    {
+      "epoch": 0.6619541918547241,
+      "grad_norm": 1.6776366233825684,
+      "learning_rate": 3.55871615454822e-07,
+      "loss": 0.8751,
+      "step": 27485
+    },
+    {
+      "epoch": 0.6620746128465114,
+      "grad_norm": 1.3133662939071655,
+      "learning_rate": 3.557448534631376e-07,
+      "loss": 0.8528,
+      "step": 27490
+    },
+    {
+      "epoch": 0.6621950338382987,
+      "grad_norm": 1.6722146272659302,
+      "learning_rate": 3.5561809147145315e-07,
+      "loss": 0.9213,
+      "step": 27495
+    },
+    {
+      "epoch": 0.662315454830086,
+      "grad_norm": 1.365938663482666,
+      "learning_rate": 3.554913294797688e-07,
+      "loss": 0.8883,
+      "step": 27500
+    },
+    {
+      "epoch": 0.6624358758218732,
+      "grad_norm": 1.6689952611923218,
+      "learning_rate": 3.5536456748808434e-07,
+      "loss": 0.8746,
+      "step": 27505
+    },
+    {
+      "epoch": 0.6625562968136606,
+      "grad_norm": 1.7636252641677856,
+      "learning_rate": 3.5523780549639993e-07,
+      "loss": 0.9094,
+      "step": 27510
+    },
+    {
+      "epoch": 0.6626767178054478,
+      "grad_norm": 1.6054761409759521,
+      "learning_rate": 3.551110435047155e-07,
+      "loss": 0.9026,
+      "step": 27515
+    },
+    {
+      "epoch": 0.6627971387972351,
+      "grad_norm": 1.563241720199585,
+      "learning_rate": 3.549842815130311e-07,
+      "loss": 0.9092,
+      "step": 27520
+    },
+    {
+      "epoch": 0.6629175597890224,
+      "grad_norm": 1.3838763236999512,
+      "learning_rate": 3.548575195213467e-07,
+      "loss": 0.9027,
+      "step": 27525
+    },
+    {
+      "epoch": 0.6630379807808097,
+      "grad_norm": 1.4422403573989868,
+      "learning_rate": 3.547307575296623e-07,
+      "loss": 0.8957,
+      "step": 27530
+    },
+    {
+      "epoch": 0.663158401772597,
+      "grad_norm": 1.3933910131454468,
+      "learning_rate": 3.546039955379779e-07,
+      "loss": 0.9401,
+      "step": 27535
+    },
+    {
+      "epoch": 0.6632788227643843,
+      "grad_norm": 1.6570963859558105,
+      "learning_rate": 3.544772335462934e-07,
+      "loss": 0.8739,
+      "step": 27540
+    },
+    {
+      "epoch": 0.6633992437561715,
+      "grad_norm": 1.5029670000076294,
+      "learning_rate": 3.5435047155460907e-07,
+      "loss": 0.8745,
+      "step": 27545
+    },
+    {
+      "epoch": 0.6635196647479589,
+      "grad_norm": 1.5351386070251465,
+      "learning_rate": 3.542237095629246e-07,
+      "loss": 0.8826,
+      "step": 27550
+    },
+    {
+      "epoch": 0.6636400857397462,
+      "grad_norm": 1.940496563911438,
+      "learning_rate": 3.540969475712402e-07,
+      "loss": 0.9065,
+      "step": 27555
+    },
+    {
+      "epoch": 0.6637605067315334,
+      "grad_norm": 1.5638905763626099,
+      "learning_rate": 3.5397018557955585e-07,
+      "loss": 0.902,
+      "step": 27560
+    },
+    {
+      "epoch": 0.6638809277233207,
+      "grad_norm": 1.4888930320739746,
+      "learning_rate": 3.538434235878714e-07,
+      "loss": 0.9199,
+      "step": 27565
+    },
+    {
+      "epoch": 0.664001348715108,
+      "grad_norm": 1.898266315460205,
+      "learning_rate": 3.53716661596187e-07,
+      "loss": 0.9166,
+      "step": 27570
+    },
+    {
+      "epoch": 0.6641217697068953,
+      "grad_norm": 1.5693520307540894,
+      "learning_rate": 3.5358989960450257e-07,
+      "loss": 0.9261,
+      "step": 27575
+    },
+    {
+      "epoch": 0.6642421906986826,
+      "grad_norm": 1.8452355861663818,
+      "learning_rate": 3.5346313761281816e-07,
+      "loss": 0.8723,
+      "step": 27580
+    },
+    {
+      "epoch": 0.6643626116904698,
+      "grad_norm": 1.486311435699463,
+      "learning_rate": 3.533363756211337e-07,
+      "loss": 0.8415,
+      "step": 27585
+    },
+    {
+      "epoch": 0.6644830326822572,
+      "grad_norm": 1.572955846786499,
+      "learning_rate": 3.5320961362944934e-07,
+      "loss": 0.8646,
+      "step": 27590
+    },
+    {
+      "epoch": 0.6646034536740445,
+      "grad_norm": 1.6016361713409424,
+      "learning_rate": 3.5308285163776494e-07,
+      "loss": 0.855,
+      "step": 27595
+    },
+    {
+      "epoch": 0.6647238746658317,
+      "grad_norm": 1.5524715185165405,
+      "learning_rate": 3.529560896460805e-07,
+      "loss": 0.9067,
+      "step": 27600
+    },
+    {
+      "epoch": 0.664844295657619,
+      "grad_norm": 1.272606372833252,
+      "learning_rate": 3.528293276543961e-07,
+      "loss": 0.902,
+      "step": 27605
+    },
+    {
+      "epoch": 0.6649647166494064,
+      "grad_norm": 1.6640665531158447,
+      "learning_rate": 3.5270256566271166e-07,
+      "loss": 0.9027,
+      "step": 27610
+    },
+    {
+      "epoch": 0.6650851376411936,
+      "grad_norm": 1.495080590248108,
+      "learning_rate": 3.5257580367102725e-07,
+      "loss": 0.835,
+      "step": 27615
+    },
+    {
+      "epoch": 0.6652055586329809,
+      "grad_norm": 1.670669674873352,
+      "learning_rate": 3.5244904167934284e-07,
+      "loss": 0.9034,
+      "step": 27620
+    },
+    {
+      "epoch": 0.6653259796247681,
+      "grad_norm": 1.649122953414917,
+      "learning_rate": 3.5232227968765843e-07,
+      "loss": 0.8889,
+      "step": 27625
+    },
+    {
+      "epoch": 0.6654464006165555,
+      "grad_norm": 1.7025654315948486,
+      "learning_rate": 3.5219551769597397e-07,
+      "loss": 0.8757,
+      "step": 27630
+    },
+    {
+      "epoch": 0.6655668216083428,
+      "grad_norm": 1.565008521080017,
+      "learning_rate": 3.520687557042896e-07,
+      "loss": 0.8086,
+      "step": 27635
+    },
+    {
+      "epoch": 0.66568724260013,
+      "grad_norm": 1.5095890760421753,
+      "learning_rate": 3.519419937126052e-07,
+      "loss": 0.9017,
+      "step": 27640
+    },
+    {
+      "epoch": 0.6658076635919173,
+      "grad_norm": 1.465999722480774,
+      "learning_rate": 3.5181523172092075e-07,
+      "loss": 0.8848,
+      "step": 27645
+    },
+    {
+      "epoch": 0.6659280845837047,
+      "grad_norm": 1.6381109952926636,
+      "learning_rate": 3.516884697292364e-07,
+      "loss": 0.9635,
+      "step": 27650
+    },
+    {
+      "epoch": 0.6660485055754919,
+      "grad_norm": 1.6575664281845093,
+      "learning_rate": 3.5156170773755193e-07,
+      "loss": 0.934,
+      "step": 27655
+    },
+    {
+      "epoch": 0.6661689265672792,
+      "grad_norm": 1.5238529443740845,
+      "learning_rate": 3.514349457458675e-07,
+      "loss": 0.9337,
+      "step": 27660
+    },
+    {
+      "epoch": 0.6662893475590665,
+      "grad_norm": 1.5721980333328247,
+      "learning_rate": 3.513081837541831e-07,
+      "loss": 0.919,
+      "step": 27665
+    },
+    {
+      "epoch": 0.6664097685508538,
+      "grad_norm": 1.770738959312439,
+      "learning_rate": 3.511814217624987e-07,
+      "loss": 0.8885,
+      "step": 27670
+    },
+    {
+      "epoch": 0.6665301895426411,
+      "grad_norm": 1.657386064529419,
+      "learning_rate": 3.510546597708143e-07,
+      "loss": 0.9222,
+      "step": 27675
+    },
+    {
+      "epoch": 0.6666506105344283,
+      "grad_norm": 1.6715120077133179,
+      "learning_rate": 3.509278977791299e-07,
+      "loss": 0.9324,
+      "step": 27680
+    },
+    {
+      "epoch": 0.6667710315262156,
+      "grad_norm": 1.4019960165023804,
+      "learning_rate": 3.508011357874455e-07,
+      "loss": 0.9458,
+      "step": 27685
+    },
+    {
+      "epoch": 0.666891452518003,
+      "grad_norm": 1.474033236503601,
+      "learning_rate": 3.50674373795761e-07,
+      "loss": 0.8673,
+      "step": 27690
+    },
+    {
+      "epoch": 0.6670118735097902,
+      "grad_norm": 1.807705283164978,
+      "learning_rate": 3.5054761180407667e-07,
+      "loss": 0.8754,
+      "step": 27695
+    },
+    {
+      "epoch": 0.6671322945015775,
+      "grad_norm": 1.7381840944290161,
+      "learning_rate": 3.504208498123922e-07,
+      "loss": 0.8682,
+      "step": 27700
+    },
+    {
+      "epoch": 0.6672527154933648,
+      "grad_norm": 1.5708887577056885,
+      "learning_rate": 3.502940878207078e-07,
+      "loss": 0.9015,
+      "step": 27705
+    },
+    {
+      "epoch": 0.6673731364851521,
+      "grad_norm": 1.4795726537704468,
+      "learning_rate": 3.5016732582902344e-07,
+      "loss": 0.8973,
+      "step": 27710
+    },
+    {
+      "epoch": 0.6674935574769394,
+      "grad_norm": 1.4762171506881714,
+      "learning_rate": 3.50040563837339e-07,
+      "loss": 0.932,
+      "step": 27715
+    },
+    {
+      "epoch": 0.6676139784687267,
+      "grad_norm": 1.4258259534835815,
+      "learning_rate": 3.499138018456546e-07,
+      "loss": 0.9464,
+      "step": 27720
+    },
+    {
+      "epoch": 0.6677343994605139,
+      "grad_norm": 1.4672174453735352,
+      "learning_rate": 3.4978703985397016e-07,
+      "loss": 0.9,
+      "step": 27725
+    },
+    {
+      "epoch": 0.6678548204523013,
+      "grad_norm": 1.41704261302948,
+      "learning_rate": 3.4966027786228576e-07,
+      "loss": 0.9241,
+      "step": 27730
+    },
+    {
+      "epoch": 0.6679752414440885,
+      "grad_norm": 1.6913530826568604,
+      "learning_rate": 3.4953351587060135e-07,
+      "loss": 0.897,
+      "step": 27735
+    },
+    {
+      "epoch": 0.6680956624358758,
+      "grad_norm": 1.5633080005645752,
+      "learning_rate": 3.4940675387891694e-07,
+      "loss": 0.8569,
+      "step": 27740
+    },
+    {
+      "epoch": 0.6682160834276631,
+      "grad_norm": 1.7641186714172363,
+      "learning_rate": 3.4927999188723253e-07,
+      "loss": 0.8692,
+      "step": 27745
+    },
+    {
+      "epoch": 0.6683365044194504,
+      "grad_norm": 1.6391301155090332,
+      "learning_rate": 3.491532298955481e-07,
+      "loss": 0.9202,
+      "step": 27750
+    },
+    {
+      "epoch": 0.6684569254112377,
+      "grad_norm": 1.598775863647461,
+      "learning_rate": 3.490264679038637e-07,
+      "loss": 0.9319,
+      "step": 27755
+    },
+    {
+      "epoch": 0.668577346403025,
+      "grad_norm": 1.6690956354141235,
+      "learning_rate": 3.4889970591217925e-07,
+      "loss": 0.8681,
+      "step": 27760
+    },
+    {
+      "epoch": 0.6686977673948122,
+      "grad_norm": 1.440314531326294,
+      "learning_rate": 3.487729439204949e-07,
+      "loss": 0.8995,
+      "step": 27765
+    },
+    {
+      "epoch": 0.6688181883865996,
+      "grad_norm": 1.5245696306228638,
+      "learning_rate": 3.4864618192881044e-07,
+      "loss": 0.8909,
+      "step": 27770
+    },
+    {
+      "epoch": 0.6689386093783869,
+      "grad_norm": 1.6076589822769165,
+      "learning_rate": 3.4851941993712603e-07,
+      "loss": 0.8378,
+      "step": 27775
+    },
+    {
+      "epoch": 0.6690590303701741,
+      "grad_norm": 1.628814935684204,
+      "learning_rate": 3.4839265794544167e-07,
+      "loss": 0.8412,
+      "step": 27780
+    },
+    {
+      "epoch": 0.6691794513619614,
+      "grad_norm": 1.5785472393035889,
+      "learning_rate": 3.482658959537572e-07,
+      "loss": 0.913,
+      "step": 27785
+    },
+    {
+      "epoch": 0.6692998723537487,
+      "grad_norm": 1.8055250644683838,
+      "learning_rate": 3.481391339620728e-07,
+      "loss": 0.9016,
+      "step": 27790
+    },
+    {
+      "epoch": 0.669420293345536,
+      "grad_norm": 1.6282498836517334,
+      "learning_rate": 3.480123719703884e-07,
+      "loss": 0.9203,
+      "step": 27795
+    },
+    {
+      "epoch": 0.6695407143373233,
+      "grad_norm": 1.3604364395141602,
+      "learning_rate": 3.47885609978704e-07,
+      "loss": 0.9183,
+      "step": 27800
+    },
+    {
+      "epoch": 0.6696611353291105,
+      "grad_norm": 1.5119647979736328,
+      "learning_rate": 3.477588479870195e-07,
+      "loss": 0.9655,
+      "step": 27805
+    },
+    {
+      "epoch": 0.6697815563208979,
+      "grad_norm": 1.5084353685379028,
+      "learning_rate": 3.4763208599533517e-07,
+      "loss": 0.9002,
+      "step": 27810
+    },
+    {
+      "epoch": 0.6699019773126852,
+      "grad_norm": 1.4556913375854492,
+      "learning_rate": 3.475053240036507e-07,
+      "loss": 0.896,
+      "step": 27815
+    },
+    {
+      "epoch": 0.6700223983044724,
+      "grad_norm": 1.5710209608078003,
+      "learning_rate": 3.473785620119663e-07,
+      "loss": 0.9056,
+      "step": 27820
+    },
+    {
+      "epoch": 0.6701428192962597,
+      "grad_norm": 2.3126919269561768,
+      "learning_rate": 3.4725180002028195e-07,
+      "loss": 0.8748,
+      "step": 27825
+    },
+    {
+      "epoch": 0.670263240288047,
+      "grad_norm": 1.6674573421478271,
+      "learning_rate": 3.471250380285975e-07,
+      "loss": 0.9199,
+      "step": 27830
+    },
+    {
+      "epoch": 0.6703836612798343,
+      "grad_norm": 1.5254837274551392,
+      "learning_rate": 3.469982760369131e-07,
+      "loss": 0.9121,
+      "step": 27835
+    },
+    {
+      "epoch": 0.6705040822716216,
+      "grad_norm": 1.6415302753448486,
+      "learning_rate": 3.4687151404522867e-07,
+      "loss": 0.9034,
+      "step": 27840
+    },
+    {
+      "epoch": 0.6706245032634088,
+      "grad_norm": 1.5428940057754517,
+      "learning_rate": 3.4674475205354426e-07,
+      "loss": 0.906,
+      "step": 27845
+    },
+    {
+      "epoch": 0.6707449242551962,
+      "grad_norm": 1.6544368267059326,
+      "learning_rate": 3.466179900618598e-07,
+      "loss": 0.9137,
+      "step": 27850
+    },
+    {
+      "epoch": 0.6708653452469835,
+      "grad_norm": 1.7646950483322144,
+      "learning_rate": 3.4649122807017544e-07,
+      "loss": 0.8985,
+      "step": 27855
+    },
+    {
+      "epoch": 0.6709857662387707,
+      "grad_norm": 1.6507251262664795,
+      "learning_rate": 3.4636446607849104e-07,
+      "loss": 0.8856,
+      "step": 27860
+    },
+    {
+      "epoch": 0.671106187230558,
+      "grad_norm": 1.5486855506896973,
+      "learning_rate": 3.462377040868066e-07,
+      "loss": 0.8904,
+      "step": 27865
+    },
+    {
+      "epoch": 0.6712266082223454,
+      "grad_norm": 1.3910833597183228,
+      "learning_rate": 3.461109420951222e-07,
+      "loss": 0.9062,
+      "step": 27870
+    },
+    {
+      "epoch": 0.6713470292141326,
+      "grad_norm": 1.4404289722442627,
+      "learning_rate": 3.4598418010343776e-07,
+      "loss": 0.9364,
+      "step": 27875
+    },
+    {
+      "epoch": 0.6714674502059199,
+      "grad_norm": 1.5207289457321167,
+      "learning_rate": 3.4585741811175335e-07,
+      "loss": 0.9202,
+      "step": 27880
+    },
+    {
+      "epoch": 0.6715878711977071,
+      "grad_norm": 1.6543253660202026,
+      "learning_rate": 3.4573065612006894e-07,
+      "loss": 0.9125,
+      "step": 27885
+    },
+    {
+      "epoch": 0.6717082921894945,
+      "grad_norm": 1.3949698209762573,
+      "learning_rate": 3.4560389412838453e-07,
+      "loss": 0.8908,
+      "step": 27890
+    },
+    {
+      "epoch": 0.6718287131812818,
+      "grad_norm": 1.6345555782318115,
+      "learning_rate": 3.454771321367001e-07,
+      "loss": 0.9174,
+      "step": 27895
+    },
+    {
+      "epoch": 0.671949134173069,
+      "grad_norm": 1.3611493110656738,
+      "learning_rate": 3.453503701450157e-07,
+      "loss": 0.9037,
+      "step": 27900
+    },
+    {
+      "epoch": 0.6720695551648563,
+      "grad_norm": 1.530236005783081,
+      "learning_rate": 3.452236081533313e-07,
+      "loss": 0.911,
+      "step": 27905
+    },
+    {
+      "epoch": 0.6721899761566437,
+      "grad_norm": 1.83746337890625,
+      "learning_rate": 3.4509684616164685e-07,
+      "loss": 0.9357,
+      "step": 27910
+    },
+    {
+      "epoch": 0.6723103971484309,
+      "grad_norm": 1.508793830871582,
+      "learning_rate": 3.449700841699625e-07,
+      "loss": 0.8985,
+      "step": 27915
+    },
+    {
+      "epoch": 0.6724308181402182,
+      "grad_norm": 1.5518765449523926,
+      "learning_rate": 3.4484332217827803e-07,
+      "loss": 0.8257,
+      "step": 27920
+    },
+    {
+      "epoch": 0.6725512391320055,
+      "grad_norm": 1.5395053625106812,
+      "learning_rate": 3.447165601865936e-07,
+      "loss": 0.9241,
+      "step": 27925
+    },
+    {
+      "epoch": 0.6726716601237928,
+      "grad_norm": 1.581311583518982,
+      "learning_rate": 3.4458979819490927e-07,
+      "loss": 0.9167,
+      "step": 27930
+    },
+    {
+      "epoch": 0.6727920811155801,
+      "grad_norm": 1.6826280355453491,
+      "learning_rate": 3.444630362032248e-07,
+      "loss": 0.9198,
+      "step": 27935
+    },
+    {
+      "epoch": 0.6729125021073673,
+      "grad_norm": 1.5144160985946655,
+      "learning_rate": 3.443362742115404e-07,
+      "loss": 0.8347,
+      "step": 27940
+    },
+    {
+      "epoch": 0.6730329230991546,
+      "grad_norm": 1.4464256763458252,
+      "learning_rate": 3.44209512219856e-07,
+      "loss": 0.9136,
+      "step": 27945
+    },
+    {
+      "epoch": 0.673153344090942,
+      "grad_norm": 1.5539331436157227,
+      "learning_rate": 3.440827502281716e-07,
+      "loss": 0.8851,
+      "step": 27950
+    },
+    {
+      "epoch": 0.6732737650827292,
+      "grad_norm": 1.602190613746643,
+      "learning_rate": 3.439559882364871e-07,
+      "loss": 0.8836,
+      "step": 27955
+    },
+    {
+      "epoch": 0.6733941860745165,
+      "grad_norm": 1.5354704856872559,
+      "learning_rate": 3.4382922624480277e-07,
+      "loss": 0.886,
+      "step": 27960
+    },
+    {
+      "epoch": 0.6735146070663038,
+      "grad_norm": 1.4831632375717163,
+      "learning_rate": 3.437024642531183e-07,
+      "loss": 0.9186,
+      "step": 27965
+    },
+    {
+      "epoch": 0.6736350280580911,
+      "grad_norm": 1.542081356048584,
+      "learning_rate": 3.435757022614339e-07,
+      "loss": 0.9533,
+      "step": 27970
+    },
+    {
+      "epoch": 0.6737554490498784,
+      "grad_norm": 1.5842245817184448,
+      "learning_rate": 3.4344894026974954e-07,
+      "loss": 0.8567,
+      "step": 27975
+    },
+    {
+      "epoch": 0.6738758700416657,
+      "grad_norm": 1.6727997064590454,
+      "learning_rate": 3.433221782780651e-07,
+      "loss": 0.9581,
+      "step": 27980
+    },
+    {
+      "epoch": 0.6739962910334529,
+      "grad_norm": 1.609487533569336,
+      "learning_rate": 3.4319541628638067e-07,
+      "loss": 0.8881,
+      "step": 27985
+    },
+    {
+      "epoch": 0.6741167120252403,
+      "grad_norm": 1.6405456066131592,
+      "learning_rate": 3.4306865429469626e-07,
+      "loss": 0.9186,
+      "step": 27990
+    },
+    {
+      "epoch": 0.6742371330170275,
+      "grad_norm": 1.876981496810913,
+      "learning_rate": 3.4294189230301186e-07,
+      "loss": 0.8679,
+      "step": 27995
+    },
+    {
+      "epoch": 0.6743575540088148,
+      "grad_norm": 1.5377568006515503,
+      "learning_rate": 3.428151303113274e-07,
+      "loss": 0.9266,
+      "step": 28000
+    },
+    {
+      "epoch": 0.6744779750006021,
+      "grad_norm": 1.7596919536590576,
+      "learning_rate": 3.4268836831964304e-07,
+      "loss": 0.9209,
+      "step": 28005
+    },
+    {
+      "epoch": 0.6745983959923894,
+      "grad_norm": 1.688023328781128,
+      "learning_rate": 3.4256160632795863e-07,
+      "loss": 0.9065,
+      "step": 28010
+    },
+    {
+      "epoch": 0.6747188169841767,
+      "grad_norm": 1.5075517892837524,
+      "learning_rate": 3.4243484433627417e-07,
+      "loss": 0.903,
+      "step": 28015
+    },
+    {
+      "epoch": 0.674839237975964,
+      "grad_norm": 1.4978880882263184,
+      "learning_rate": 3.423080823445898e-07,
+      "loss": 0.8766,
+      "step": 28020
+    },
+    {
+      "epoch": 0.6749596589677512,
+      "grad_norm": 1.665744662284851,
+      "learning_rate": 3.4218132035290535e-07,
+      "loss": 0.8692,
+      "step": 28025
+    },
+    {
+      "epoch": 0.6750800799595386,
+      "grad_norm": 1.611232042312622,
+      "learning_rate": 3.4205455836122095e-07,
+      "loss": 0.8826,
+      "step": 28030
+    },
+    {
+      "epoch": 0.6752005009513259,
+      "grad_norm": 2.0650761127471924,
+      "learning_rate": 3.4192779636953654e-07,
+      "loss": 0.9662,
+      "step": 28035
+    },
+    {
+      "epoch": 0.6753209219431131,
+      "grad_norm": 1.5076416730880737,
+      "learning_rate": 3.4180103437785213e-07,
+      "loss": 0.8655,
+      "step": 28040
+    },
+    {
+      "epoch": 0.6754413429349004,
+      "grad_norm": 1.4803639650344849,
+      "learning_rate": 3.416742723861677e-07,
+      "loss": 0.8003,
+      "step": 28045
+    },
+    {
+      "epoch": 0.6755617639266877,
+      "grad_norm": 1.8180280923843384,
+      "learning_rate": 3.415475103944833e-07,
+      "loss": 0.8784,
+      "step": 28050
+    },
+    {
+      "epoch": 0.675682184918475,
+      "grad_norm": 1.4070686101913452,
+      "learning_rate": 3.414207484027989e-07,
+      "loss": 0.8566,
+      "step": 28055
+    },
+    {
+      "epoch": 0.6758026059102623,
+      "grad_norm": 1.5639938116073608,
+      "learning_rate": 3.4129398641111444e-07,
+      "loss": 0.8849,
+      "step": 28060
+    },
+    {
+      "epoch": 0.6759230269020495,
+      "grad_norm": 1.6492422819137573,
+      "learning_rate": 3.411672244194301e-07,
+      "loss": 0.8641,
+      "step": 28065
+    },
+    {
+      "epoch": 0.6760434478938369,
+      "grad_norm": 1.6205191612243652,
+      "learning_rate": 3.4104046242774563e-07,
+      "loss": 0.8943,
+      "step": 28070
+    },
+    {
+      "epoch": 0.6761638688856242,
+      "grad_norm": 1.5223482847213745,
+      "learning_rate": 3.409137004360612e-07,
+      "loss": 0.8655,
+      "step": 28075
+    },
+    {
+      "epoch": 0.6762842898774114,
+      "grad_norm": 1.5024832487106323,
+      "learning_rate": 3.4078693844437686e-07,
+      "loss": 0.9364,
+      "step": 28080
+    },
+    {
+      "epoch": 0.6764047108691987,
+      "grad_norm": 1.5253880023956299,
+      "learning_rate": 3.406601764526924e-07,
+      "loss": 0.9013,
+      "step": 28085
+    },
+    {
+      "epoch": 0.6765251318609861,
+      "grad_norm": 1.6916687488555908,
+      "learning_rate": 3.40533414461008e-07,
+      "loss": 0.9208,
+      "step": 28090
+    },
+    {
+      "epoch": 0.6766455528527733,
+      "grad_norm": 1.7585076093673706,
+      "learning_rate": 3.404066524693236e-07,
+      "loss": 0.8898,
+      "step": 28095
+    },
+    {
+      "epoch": 0.6767659738445606,
+      "grad_norm": 1.453704833984375,
+      "learning_rate": 3.402798904776392e-07,
+      "loss": 0.9072,
+      "step": 28100
+    },
+    {
+      "epoch": 0.6768863948363478,
+      "grad_norm": 1.501059651374817,
+      "learning_rate": 3.401531284859547e-07,
+      "loss": 0.8604,
+      "step": 28105
+    },
+    {
+      "epoch": 0.6770068158281352,
+      "grad_norm": 1.4455554485321045,
+      "learning_rate": 3.4002636649427036e-07,
+      "loss": 0.9066,
+      "step": 28110
+    },
+    {
+      "epoch": 0.6771272368199225,
+      "grad_norm": 1.4117891788482666,
+      "learning_rate": 3.398996045025859e-07,
+      "loss": 0.8527,
+      "step": 28115
+    },
+    {
+      "epoch": 0.6772476578117097,
+      "grad_norm": 1.7244776487350464,
+      "learning_rate": 3.397728425109015e-07,
+      "loss": 0.8595,
+      "step": 28120
+    },
+    {
+      "epoch": 0.677368078803497,
+      "grad_norm": 1.7050223350524902,
+      "learning_rate": 3.3964608051921714e-07,
+      "loss": 0.941,
+      "step": 28125
+    },
+    {
+      "epoch": 0.6774884997952844,
+      "grad_norm": 1.3838149309158325,
+      "learning_rate": 3.395193185275327e-07,
+      "loss": 0.9236,
+      "step": 28130
+    },
+    {
+      "epoch": 0.6776089207870716,
+      "grad_norm": 1.6386252641677856,
+      "learning_rate": 3.3939255653584827e-07,
+      "loss": 0.8627,
+      "step": 28135
+    },
+    {
+      "epoch": 0.6777293417788589,
+      "grad_norm": 1.5708414316177368,
+      "learning_rate": 3.3926579454416386e-07,
+      "loss": 0.942,
+      "step": 28140
+    },
+    {
+      "epoch": 0.6778497627706461,
+      "grad_norm": 1.4829083681106567,
+      "learning_rate": 3.3913903255247945e-07,
+      "loss": 0.8537,
+      "step": 28145
+    },
+    {
+      "epoch": 0.6779701837624335,
+      "grad_norm": 1.8113583326339722,
+      "learning_rate": 3.39012270560795e-07,
+      "loss": 0.9425,
+      "step": 28150
+    },
+    {
+      "epoch": 0.6780906047542208,
+      "grad_norm": 1.42169988155365,
+      "learning_rate": 3.3888550856911063e-07,
+      "loss": 0.928,
+      "step": 28155
+    },
+    {
+      "epoch": 0.678211025746008,
+      "grad_norm": 1.514338493347168,
+      "learning_rate": 3.387587465774262e-07,
+      "loss": 0.8995,
+      "step": 28160
+    },
+    {
+      "epoch": 0.6783314467377953,
+      "grad_norm": 1.3799099922180176,
+      "learning_rate": 3.3863198458574176e-07,
+      "loss": 0.8719,
+      "step": 28165
+    },
+    {
+      "epoch": 0.6784518677295827,
+      "grad_norm": 1.5651031732559204,
+      "learning_rate": 3.385052225940574e-07,
+      "loss": 0.9245,
+      "step": 28170
+    },
+    {
+      "epoch": 0.6785722887213699,
+      "grad_norm": 1.80035400390625,
+      "learning_rate": 3.3837846060237295e-07,
+      "loss": 0.9217,
+      "step": 28175
+    },
+    {
+      "epoch": 0.6786927097131572,
+      "grad_norm": 1.7736549377441406,
+      "learning_rate": 3.3825169861068854e-07,
+      "loss": 0.918,
+      "step": 28180
+    },
+    {
+      "epoch": 0.6788131307049445,
+      "grad_norm": 1.5086334943771362,
+      "learning_rate": 3.3812493661900413e-07,
+      "loss": 0.9244,
+      "step": 28185
+    },
+    {
+      "epoch": 0.6789335516967318,
+      "grad_norm": 1.5210703611373901,
+      "learning_rate": 3.379981746273197e-07,
+      "loss": 0.8856,
+      "step": 28190
+    },
+    {
+      "epoch": 0.6790539726885191,
+      "grad_norm": 1.441664218902588,
+      "learning_rate": 3.3787141263563526e-07,
+      "loss": 0.8924,
+      "step": 28195
+    },
+    {
+      "epoch": 0.6791743936803063,
+      "grad_norm": 1.375878095626831,
+      "learning_rate": 3.377446506439509e-07,
+      "loss": 0.9211,
+      "step": 28200
+    },
+    {
+      "epoch": 0.6792948146720936,
+      "grad_norm": 1.7333893775939941,
+      "learning_rate": 3.376178886522665e-07,
+      "loss": 0.9314,
+      "step": 28205
+    },
+    {
+      "epoch": 0.679415235663881,
+      "grad_norm": 1.5676325559616089,
+      "learning_rate": 3.3749112666058204e-07,
+      "loss": 0.9135,
+      "step": 28210
+    },
+    {
+      "epoch": 0.6795356566556682,
+      "grad_norm": 1.579702377319336,
+      "learning_rate": 3.373643646688977e-07,
+      "loss": 0.8987,
+      "step": 28215
+    },
+    {
+      "epoch": 0.6796560776474555,
+      "grad_norm": 2.8164761066436768,
+      "learning_rate": 3.372376026772132e-07,
+      "loss": 0.8973,
+      "step": 28220
+    },
+    {
+      "epoch": 0.6797764986392428,
+      "grad_norm": 1.6582890748977661,
+      "learning_rate": 3.3711084068552887e-07,
+      "loss": 0.9338,
+      "step": 28225
+    },
+    {
+      "epoch": 0.67989691963103,
+      "grad_norm": 1.664105772972107,
+      "learning_rate": 3.369840786938444e-07,
+      "loss": 0.9562,
+      "step": 28230
+    },
+    {
+      "epoch": 0.6800173406228174,
+      "grad_norm": 1.8200933933258057,
+      "learning_rate": 3.3685731670216e-07,
+      "loss": 0.8899,
+      "step": 28235
+    },
+    {
+      "epoch": 0.6801377616146047,
+      "grad_norm": 1.5525609254837036,
+      "learning_rate": 3.3673055471047564e-07,
+      "loss": 0.8831,
+      "step": 28240
+    },
+    {
+      "epoch": 0.6802581826063919,
+      "grad_norm": 1.448174238204956,
+      "learning_rate": 3.366037927187912e-07,
+      "loss": 0.901,
+      "step": 28245
+    },
+    {
+      "epoch": 0.6803786035981793,
+      "grad_norm": 1.6512342691421509,
+      "learning_rate": 3.3647703072710677e-07,
+      "loss": 0.9541,
+      "step": 28250
+    },
+    {
+      "epoch": 0.6804990245899665,
+      "grad_norm": 1.595133662223816,
+      "learning_rate": 3.3635026873542236e-07,
+      "loss": 0.9343,
+      "step": 28255
+    },
+    {
+      "epoch": 0.6806194455817538,
+      "grad_norm": 1.513055443763733,
+      "learning_rate": 3.3622350674373796e-07,
+      "loss": 0.9306,
+      "step": 28260
+    },
+    {
+      "epoch": 0.6807398665735411,
+      "grad_norm": 1.5822969675064087,
+      "learning_rate": 3.360967447520535e-07,
+      "loss": 0.8789,
+      "step": 28265
+    },
+    {
+      "epoch": 0.6808602875653283,
+      "grad_norm": 1.7590924501419067,
+      "learning_rate": 3.3596998276036914e-07,
+      "loss": 0.8886,
+      "step": 28270
+    },
+    {
+      "epoch": 0.6809807085571157,
+      "grad_norm": 1.41705322265625,
+      "learning_rate": 3.3584322076868473e-07,
+      "loss": 0.9205,
+      "step": 28275
+    },
+    {
+      "epoch": 0.681101129548903,
+      "grad_norm": 1.8668830394744873,
+      "learning_rate": 3.3571645877700027e-07,
+      "loss": 0.9275,
+      "step": 28280
+    },
+    {
+      "epoch": 0.6812215505406902,
+      "grad_norm": 1.4133549928665161,
+      "learning_rate": 3.355896967853159e-07,
+      "loss": 0.9073,
+      "step": 28285
+    },
+    {
+      "epoch": 0.6813419715324776,
+      "grad_norm": 1.3410143852233887,
+      "learning_rate": 3.3546293479363145e-07,
+      "loss": 0.8635,
+      "step": 28290
+    },
+    {
+      "epoch": 0.6814623925242649,
+      "grad_norm": 1.4225311279296875,
+      "learning_rate": 3.3533617280194705e-07,
+      "loss": 0.9126,
+      "step": 28295
+    },
+    {
+      "epoch": 0.6815828135160521,
+      "grad_norm": 1.5505460500717163,
+      "learning_rate": 3.3520941081026264e-07,
+      "loss": 0.8903,
+      "step": 28300
+    },
+    {
+      "epoch": 0.6817032345078394,
+      "grad_norm": 1.7633575201034546,
+      "learning_rate": 3.3508264881857823e-07,
+      "loss": 0.8563,
+      "step": 28305
+    },
+    {
+      "epoch": 0.6818236554996266,
+      "grad_norm": 1.2661945819854736,
+      "learning_rate": 3.349558868268938e-07,
+      "loss": 0.8984,
+      "step": 28310
+    },
+    {
+      "epoch": 0.681944076491414,
+      "grad_norm": 1.5636026859283447,
+      "learning_rate": 3.348291248352094e-07,
+      "loss": 0.8653,
+      "step": 28315
+    },
+    {
+      "epoch": 0.6820644974832013,
+      "grad_norm": 1.41295325756073,
+      "learning_rate": 3.34702362843525e-07,
+      "loss": 0.8966,
+      "step": 28320
+    },
+    {
+      "epoch": 0.6821849184749885,
+      "grad_norm": 1.695952296257019,
+      "learning_rate": 3.3457560085184054e-07,
+      "loss": 0.9029,
+      "step": 28325
+    },
+    {
+      "epoch": 0.6823053394667759,
+      "grad_norm": 1.8036835193634033,
+      "learning_rate": 3.344488388601562e-07,
+      "loss": 0.933,
+      "step": 28330
+    },
+    {
+      "epoch": 0.6824257604585632,
+      "grad_norm": 1.3599246740341187,
+      "learning_rate": 3.3432207686847173e-07,
+      "loss": 0.8424,
+      "step": 28335
+    },
+    {
+      "epoch": 0.6825461814503504,
+      "grad_norm": 1.4054001569747925,
+      "learning_rate": 3.341953148767873e-07,
+      "loss": 0.8867,
+      "step": 28340
+    },
+    {
+      "epoch": 0.6826666024421377,
+      "grad_norm": 1.5154298543930054,
+      "learning_rate": 3.3406855288510296e-07,
+      "loss": 0.8846,
+      "step": 28345
+    },
+    {
+      "epoch": 0.682787023433925,
+      "grad_norm": 1.7169283628463745,
+      "learning_rate": 3.339417908934185e-07,
+      "loss": 0.9178,
+      "step": 28350
+    },
+    {
+      "epoch": 0.6829074444257123,
+      "grad_norm": 1.7609354257583618,
+      "learning_rate": 3.338150289017341e-07,
+      "loss": 0.8968,
+      "step": 28355
+    },
+    {
+      "epoch": 0.6830278654174996,
+      "grad_norm": 1.993127703666687,
+      "learning_rate": 3.336882669100497e-07,
+      "loss": 0.9229,
+      "step": 28360
+    },
+    {
+      "epoch": 0.6831482864092868,
+      "grad_norm": 1.6624497175216675,
+      "learning_rate": 3.335615049183653e-07,
+      "loss": 0.8973,
+      "step": 28365
+    },
+    {
+      "epoch": 0.6832687074010741,
+      "grad_norm": 1.5066438913345337,
+      "learning_rate": 3.334347429266808e-07,
+      "loss": 0.9035,
+      "step": 28370
+    },
+    {
+      "epoch": 0.6833891283928615,
+      "grad_norm": 1.6899200677871704,
+      "learning_rate": 3.3330798093499646e-07,
+      "loss": 0.9272,
+      "step": 28375
+    },
+    {
+      "epoch": 0.6835095493846487,
+      "grad_norm": 1.4545860290527344,
+      "learning_rate": 3.33181218943312e-07,
+      "loss": 0.8128,
+      "step": 28380
+    },
+    {
+      "epoch": 0.683629970376436,
+      "grad_norm": 1.4383794069290161,
+      "learning_rate": 3.330544569516276e-07,
+      "loss": 0.862,
+      "step": 28385
+    },
+    {
+      "epoch": 0.6837503913682234,
+      "grad_norm": 1.5542722940444946,
+      "learning_rate": 3.3292769495994324e-07,
+      "loss": 0.9214,
+      "step": 28390
+    },
+    {
+      "epoch": 0.6838708123600106,
+      "grad_norm": 1.58388090133667,
+      "learning_rate": 3.328009329682588e-07,
+      "loss": 0.9244,
+      "step": 28395
+    },
+    {
+      "epoch": 0.6839912333517979,
+      "grad_norm": 1.826556921005249,
+      "learning_rate": 3.3267417097657437e-07,
+      "loss": 0.9191,
+      "step": 28400
+    },
+    {
+      "epoch": 0.6841116543435851,
+      "grad_norm": 1.6162747144699097,
+      "learning_rate": 3.3254740898488996e-07,
+      "loss": 0.9699,
+      "step": 28405
+    },
+    {
+      "epoch": 0.6842320753353724,
+      "grad_norm": 1.5720504522323608,
+      "learning_rate": 3.3242064699320555e-07,
+      "loss": 0.8742,
+      "step": 28410
+    },
+    {
+      "epoch": 0.6843524963271598,
+      "grad_norm": 1.6315302848815918,
+      "learning_rate": 3.322938850015211e-07,
+      "loss": 0.9017,
+      "step": 28415
+    },
+    {
+      "epoch": 0.684472917318947,
+      "grad_norm": 1.5458885431289673,
+      "learning_rate": 3.3216712300983673e-07,
+      "loss": 0.9638,
+      "step": 28420
+    },
+    {
+      "epoch": 0.6845933383107343,
+      "grad_norm": 1.3931212425231934,
+      "learning_rate": 3.320403610181523e-07,
+      "loss": 0.8568,
+      "step": 28425
+    },
+    {
+      "epoch": 0.6847137593025217,
+      "grad_norm": 1.4961457252502441,
+      "learning_rate": 3.3191359902646787e-07,
+      "loss": 0.868,
+      "step": 28430
+    },
+    {
+      "epoch": 0.6848341802943089,
+      "grad_norm": 1.6866486072540283,
+      "learning_rate": 3.317868370347835e-07,
+      "loss": 0.8903,
+      "step": 28435
+    },
+    {
+      "epoch": 0.6849546012860962,
+      "grad_norm": 1.539838433265686,
+      "learning_rate": 3.3166007504309905e-07,
+      "loss": 0.8707,
+      "step": 28440
+    },
+    {
+      "epoch": 0.6850750222778835,
+      "grad_norm": 1.6249197721481323,
+      "learning_rate": 3.3153331305141464e-07,
+      "loss": 0.9353,
+      "step": 28445
+    },
+    {
+      "epoch": 0.6851954432696707,
+      "grad_norm": 1.4114655256271362,
+      "learning_rate": 3.3140655105973023e-07,
+      "loss": 0.8629,
+      "step": 28450
+    },
+    {
+      "epoch": 0.6853158642614581,
+      "grad_norm": 1.4835420846939087,
+      "learning_rate": 3.312797890680458e-07,
+      "loss": 0.8831,
+      "step": 28455
+    },
+    {
+      "epoch": 0.6854362852532453,
+      "grad_norm": 1.542344570159912,
+      "learning_rate": 3.311530270763614e-07,
+      "loss": 0.8706,
+      "step": 28460
+    },
+    {
+      "epoch": 0.6855567062450326,
+      "grad_norm": 1.5637487173080444,
+      "learning_rate": 3.31026265084677e-07,
+      "loss": 0.8511,
+      "step": 28465
+    },
+    {
+      "epoch": 0.68567712723682,
+      "grad_norm": 1.612994909286499,
+      "learning_rate": 3.308995030929926e-07,
+      "loss": 0.8995,
+      "step": 28470
+    },
+    {
+      "epoch": 0.6857975482286072,
+      "grad_norm": 1.5317628383636475,
+      "learning_rate": 3.3077274110130814e-07,
+      "loss": 0.9169,
+      "step": 28475
+    },
+    {
+      "epoch": 0.6859179692203945,
+      "grad_norm": 1.5903844833374023,
+      "learning_rate": 3.306459791096238e-07,
+      "loss": 0.9222,
+      "step": 28480
+    },
+    {
+      "epoch": 0.6860383902121818,
+      "grad_norm": 1.6235851049423218,
+      "learning_rate": 3.305192171179393e-07,
+      "loss": 0.9109,
+      "step": 28485
+    },
+    {
+      "epoch": 0.686158811203969,
+      "grad_norm": 1.5382553339004517,
+      "learning_rate": 3.303924551262549e-07,
+      "loss": 0.9307,
+      "step": 28490
+    },
+    {
+      "epoch": 0.6862792321957564,
+      "grad_norm": 1.549065113067627,
+      "learning_rate": 3.3026569313457056e-07,
+      "loss": 0.9015,
+      "step": 28495
+    },
+    {
+      "epoch": 0.6863996531875437,
+      "grad_norm": 1.5097498893737793,
+      "learning_rate": 3.301389311428861e-07,
+      "loss": 0.8801,
+      "step": 28500
+    },
+    {
+      "epoch": 0.6865200741793309,
+      "grad_norm": 1.4290951490402222,
+      "learning_rate": 3.300121691512017e-07,
+      "loss": 0.9129,
+      "step": 28505
+    },
+    {
+      "epoch": 0.6866404951711182,
+      "grad_norm": 1.4810277223587036,
+      "learning_rate": 3.298854071595173e-07,
+      "loss": 0.8924,
+      "step": 28510
+    },
+    {
+      "epoch": 0.6867609161629055,
+      "grad_norm": 1.5234373807907104,
+      "learning_rate": 3.2975864516783287e-07,
+      "loss": 0.8969,
+      "step": 28515
+    },
+    {
+      "epoch": 0.6868813371546928,
+      "grad_norm": 1.4685182571411133,
+      "learning_rate": 3.296318831761484e-07,
+      "loss": 0.8968,
+      "step": 28520
+    },
+    {
+      "epoch": 0.6870017581464801,
+      "grad_norm": 1.7230477333068848,
+      "learning_rate": 3.2950512118446406e-07,
+      "loss": 0.9227,
+      "step": 28525
+    },
+    {
+      "epoch": 0.6871221791382673,
+      "grad_norm": 1.5892548561096191,
+      "learning_rate": 3.293783591927796e-07,
+      "loss": 0.915,
+      "step": 28530
+    },
+    {
+      "epoch": 0.6872426001300547,
+      "grad_norm": 1.4972862005233765,
+      "learning_rate": 3.292515972010952e-07,
+      "loss": 0.9013,
+      "step": 28535
+    },
+    {
+      "epoch": 0.687363021121842,
+      "grad_norm": 1.53257417678833,
+      "learning_rate": 3.2912483520941083e-07,
+      "loss": 0.8379,
+      "step": 28540
+    },
+    {
+      "epoch": 0.6874834421136292,
+      "grad_norm": 1.5534003973007202,
+      "learning_rate": 3.2899807321772637e-07,
+      "loss": 0.8975,
+      "step": 28545
+    },
+    {
+      "epoch": 0.6876038631054165,
+      "grad_norm": 1.609209418296814,
+      "learning_rate": 3.2887131122604196e-07,
+      "loss": 0.8671,
+      "step": 28550
+    },
+    {
+      "epoch": 0.6877242840972039,
+      "grad_norm": 1.4327870607376099,
+      "learning_rate": 3.2874454923435755e-07,
+      "loss": 0.86,
+      "step": 28555
+    },
+    {
+      "epoch": 0.6878447050889911,
+      "grad_norm": 1.5099050998687744,
+      "learning_rate": 3.2861778724267315e-07,
+      "loss": 0.9027,
+      "step": 28560
+    },
+    {
+      "epoch": 0.6879651260807784,
+      "grad_norm": 1.4186595678329468,
+      "learning_rate": 3.284910252509887e-07,
+      "loss": 0.8845,
+      "step": 28565
+    },
+    {
+      "epoch": 0.6880855470725656,
+      "grad_norm": 1.5847558975219727,
+      "learning_rate": 3.2836426325930433e-07,
+      "loss": 0.9053,
+      "step": 28570
+    },
+    {
+      "epoch": 0.688205968064353,
+      "grad_norm": 1.4546008110046387,
+      "learning_rate": 3.282375012676199e-07,
+      "loss": 0.9295,
+      "step": 28575
+    },
+    {
+      "epoch": 0.6883263890561403,
+      "grad_norm": 1.7883195877075195,
+      "learning_rate": 3.2811073927593546e-07,
+      "loss": 0.9353,
+      "step": 28580
+    },
+    {
+      "epoch": 0.6884468100479275,
+      "grad_norm": 1.451277256011963,
+      "learning_rate": 3.279839772842511e-07,
+      "loss": 0.876,
+      "step": 28585
+    },
+    {
+      "epoch": 0.6885672310397148,
+      "grad_norm": 1.5470596551895142,
+      "learning_rate": 3.2785721529256664e-07,
+      "loss": 0.9338,
+      "step": 28590
+    },
+    {
+      "epoch": 0.6886876520315022,
+      "grad_norm": 1.6704994440078735,
+      "learning_rate": 3.2773045330088224e-07,
+      "loss": 0.8935,
+      "step": 28595
+    },
+    {
+      "epoch": 0.6888080730232894,
+      "grad_norm": 1.406724452972412,
+      "learning_rate": 3.2760369130919783e-07,
+      "loss": 0.8856,
+      "step": 28600
+    },
+    {
+      "epoch": 0.6889284940150767,
+      "grad_norm": 1.718907117843628,
+      "learning_rate": 3.274769293175134e-07,
+      "loss": 0.9068,
+      "step": 28605
+    },
+    {
+      "epoch": 0.689048915006864,
+      "grad_norm": 1.6455481052398682,
+      "learning_rate": 3.27350167325829e-07,
+      "loss": 0.8716,
+      "step": 28610
+    },
+    {
+      "epoch": 0.6891693359986513,
+      "grad_norm": 1.4469377994537354,
+      "learning_rate": 3.272234053341446e-07,
+      "loss": 0.887,
+      "step": 28615
+    },
+    {
+      "epoch": 0.6892897569904386,
+      "grad_norm": 1.5895036458969116,
+      "learning_rate": 3.270966433424602e-07,
+      "loss": 0.9508,
+      "step": 28620
+    },
+    {
+      "epoch": 0.6894101779822258,
+      "grad_norm": 1.68479585647583,
+      "learning_rate": 3.2696988135077573e-07,
+      "loss": 0.8584,
+      "step": 28625
+    },
+    {
+      "epoch": 0.6895305989740131,
+      "grad_norm": 1.654051661491394,
+      "learning_rate": 3.268431193590914e-07,
+      "loss": 0.86,
+      "step": 28630
+    },
+    {
+      "epoch": 0.6896510199658005,
+      "grad_norm": 1.661611795425415,
+      "learning_rate": 3.267163573674069e-07,
+      "loss": 0.8986,
+      "step": 28635
+    },
+    {
+      "epoch": 0.6897714409575877,
+      "grad_norm": 1.7625168561935425,
+      "learning_rate": 3.265895953757225e-07,
+      "loss": 0.8978,
+      "step": 28640
+    },
+    {
+      "epoch": 0.689891861949375,
+      "grad_norm": 1.7721279859542847,
+      "learning_rate": 3.2646283338403815e-07,
+      "loss": 0.9409,
+      "step": 28645
+    },
+    {
+      "epoch": 0.6900122829411623,
+      "grad_norm": 1.6151334047317505,
+      "learning_rate": 3.263360713923537e-07,
+      "loss": 0.9559,
+      "step": 28650
+    },
+    {
+      "epoch": 0.6901327039329496,
+      "grad_norm": 1.4370307922363281,
+      "learning_rate": 3.262093094006693e-07,
+      "loss": 0.8596,
+      "step": 28655
+    },
+    {
+      "epoch": 0.6902531249247369,
+      "grad_norm": 1.4641938209533691,
+      "learning_rate": 3.260825474089849e-07,
+      "loss": 0.9553,
+      "step": 28660
+    },
+    {
+      "epoch": 0.6903735459165242,
+      "grad_norm": 1.4579864740371704,
+      "learning_rate": 3.2595578541730047e-07,
+      "loss": 0.8337,
+      "step": 28665
+    },
+    {
+      "epoch": 0.6904939669083114,
+      "grad_norm": 1.6612746715545654,
+      "learning_rate": 3.25829023425616e-07,
+      "loss": 0.8416,
+      "step": 28670
+    },
+    {
+      "epoch": 0.6906143879000988,
+      "grad_norm": 1.6550942659378052,
+      "learning_rate": 3.2570226143393165e-07,
+      "loss": 0.9109,
+      "step": 28675
+    },
+    {
+      "epoch": 0.690734808891886,
+      "grad_norm": 1.4688982963562012,
+      "learning_rate": 3.255754994422472e-07,
+      "loss": 0.9009,
+      "step": 28680
+    },
+    {
+      "epoch": 0.6908552298836733,
+      "grad_norm": 1.709791898727417,
+      "learning_rate": 3.254487374505628e-07,
+      "loss": 0.8808,
+      "step": 28685
+    },
+    {
+      "epoch": 0.6909756508754606,
+      "grad_norm": 1.4528766870498657,
+      "learning_rate": 3.2532197545887843e-07,
+      "loss": 0.9381,
+      "step": 28690
+    },
+    {
+      "epoch": 0.6910960718672479,
+      "grad_norm": 1.5410945415496826,
+      "learning_rate": 3.2519521346719397e-07,
+      "loss": 0.853,
+      "step": 28695
+    },
+    {
+      "epoch": 0.6912164928590352,
+      "grad_norm": 1.4717358350753784,
+      "learning_rate": 3.2506845147550956e-07,
+      "loss": 0.8683,
+      "step": 28700
+    },
+    {
+      "epoch": 0.6913369138508225,
+      "grad_norm": 1.528515100479126,
+      "learning_rate": 3.2494168948382515e-07,
+      "loss": 0.93,
+      "step": 28705
+    },
+    {
+      "epoch": 0.6914573348426097,
+      "grad_norm": 1.6877847909927368,
+      "learning_rate": 3.2481492749214074e-07,
+      "loss": 0.9553,
+      "step": 28710
+    },
+    {
+      "epoch": 0.6915777558343971,
+      "grad_norm": 1.7174590826034546,
+      "learning_rate": 3.246881655004563e-07,
+      "loss": 0.8947,
+      "step": 28715
+    },
+    {
+      "epoch": 0.6916981768261843,
+      "grad_norm": 1.3972924947738647,
+      "learning_rate": 3.245614035087719e-07,
+      "loss": 0.9128,
+      "step": 28720
+    },
+    {
+      "epoch": 0.6918185978179716,
+      "grad_norm": 1.4659849405288696,
+      "learning_rate": 3.244346415170875e-07,
+      "loss": 0.9376,
+      "step": 28725
+    },
+    {
+      "epoch": 0.6919390188097589,
+      "grad_norm": 1.6079254150390625,
+      "learning_rate": 3.2430787952540306e-07,
+      "loss": 0.8994,
+      "step": 28730
+    },
+    {
+      "epoch": 0.6920594398015462,
+      "grad_norm": 1.5413719415664673,
+      "learning_rate": 3.241811175337187e-07,
+      "loss": 0.9299,
+      "step": 28735
+    },
+    {
+      "epoch": 0.6921798607933335,
+      "grad_norm": 1.6940799951553345,
+      "learning_rate": 3.2405435554203424e-07,
+      "loss": 0.85,
+      "step": 28740
+    },
+    {
+      "epoch": 0.6923002817851208,
+      "grad_norm": 1.489127278327942,
+      "learning_rate": 3.239275935503499e-07,
+      "loss": 0.931,
+      "step": 28745
+    },
+    {
+      "epoch": 0.692420702776908,
+      "grad_norm": 1.5167713165283203,
+      "learning_rate": 3.238008315586654e-07,
+      "loss": 0.9065,
+      "step": 28750
+    },
+    {
+      "epoch": 0.6925411237686954,
+      "grad_norm": 1.5457457304000854,
+      "learning_rate": 3.23674069566981e-07,
+      "loss": 0.9039,
+      "step": 28755
+    },
+    {
+      "epoch": 0.6926615447604827,
+      "grad_norm": 1.664921760559082,
+      "learning_rate": 3.2354730757529666e-07,
+      "loss": 0.867,
+      "step": 28760
+    },
+    {
+      "epoch": 0.6927819657522699,
+      "grad_norm": 1.631435751914978,
+      "learning_rate": 3.234205455836122e-07,
+      "loss": 0.8717,
+      "step": 28765
+    },
+    {
+      "epoch": 0.6929023867440572,
+      "grad_norm": 1.630450963973999,
+      "learning_rate": 3.232937835919278e-07,
+      "loss": 0.9026,
+      "step": 28770
+    },
+    {
+      "epoch": 0.6930228077358445,
+      "grad_norm": 1.4385497570037842,
+      "learning_rate": 3.231670216002434e-07,
+      "loss": 0.9005,
+      "step": 28775
+    },
+    {
+      "epoch": 0.6931432287276318,
+      "grad_norm": 1.5163872241973877,
+      "learning_rate": 3.2304025960855897e-07,
+      "loss": 0.9388,
+      "step": 28780
+    },
+    {
+      "epoch": 0.6932636497194191,
+      "grad_norm": 1.693488597869873,
+      "learning_rate": 3.229134976168745e-07,
+      "loss": 0.8862,
+      "step": 28785
+    },
+    {
+      "epoch": 0.6933840707112063,
+      "grad_norm": 1.6429200172424316,
+      "learning_rate": 3.2278673562519016e-07,
+      "loss": 0.8752,
+      "step": 28790
+    },
+    {
+      "epoch": 0.6935044917029937,
+      "grad_norm": 1.5503240823745728,
+      "learning_rate": 3.2265997363350575e-07,
+      "loss": 0.9339,
+      "step": 28795
+    },
+    {
+      "epoch": 0.693624912694781,
+      "grad_norm": 1.4487390518188477,
+      "learning_rate": 3.225332116418213e-07,
+      "loss": 0.911,
+      "step": 28800
+    },
+    {
+      "epoch": 0.6937453336865682,
+      "grad_norm": 1.561008095741272,
+      "learning_rate": 3.2240644965013693e-07,
+      "loss": 0.8992,
+      "step": 28805
+    },
+    {
+      "epoch": 0.6938657546783555,
+      "grad_norm": 1.6827517747879028,
+      "learning_rate": 3.2227968765845247e-07,
+      "loss": 0.9336,
+      "step": 28810
+    },
+    {
+      "epoch": 0.6939861756701429,
+      "grad_norm": 1.4995176792144775,
+      "learning_rate": 3.2215292566676806e-07,
+      "loss": 0.9022,
+      "step": 28815
+    },
+    {
+      "epoch": 0.6941065966619301,
+      "grad_norm": 1.3722862005233765,
+      "learning_rate": 3.2202616367508365e-07,
+      "loss": 0.866,
+      "step": 28820
+    },
+    {
+      "epoch": 0.6942270176537174,
+      "grad_norm": 1.6332498788833618,
+      "learning_rate": 3.2189940168339925e-07,
+      "loss": 0.9745,
+      "step": 28825
+    },
+    {
+      "epoch": 0.6943474386455046,
+      "grad_norm": 1.4901752471923828,
+      "learning_rate": 3.217726396917148e-07,
+      "loss": 0.9024,
+      "step": 28830
+    },
+    {
+      "epoch": 0.694467859637292,
+      "grad_norm": 1.6924899816513062,
+      "learning_rate": 3.2164587770003043e-07,
+      "loss": 0.8887,
+      "step": 28835
+    },
+    {
+      "epoch": 0.6945882806290793,
+      "grad_norm": 1.6854546070098877,
+      "learning_rate": 3.21519115708346e-07,
+      "loss": 0.919,
+      "step": 28840
+    },
+    {
+      "epoch": 0.6947087016208665,
+      "grad_norm": 1.4148783683776855,
+      "learning_rate": 3.2139235371666156e-07,
+      "loss": 0.9169,
+      "step": 28845
+    },
+    {
+      "epoch": 0.6948291226126538,
+      "grad_norm": 1.6648613214492798,
+      "learning_rate": 3.212655917249772e-07,
+      "loss": 0.8992,
+      "step": 28850
+    },
+    {
+      "epoch": 0.6949495436044412,
+      "grad_norm": 1.4850813150405884,
+      "learning_rate": 3.2113882973329274e-07,
+      "loss": 0.908,
+      "step": 28855
+    },
+    {
+      "epoch": 0.6950699645962284,
+      "grad_norm": 1.6685442924499512,
+      "learning_rate": 3.2101206774160834e-07,
+      "loss": 0.8785,
+      "step": 28860
+    },
+    {
+      "epoch": 0.6951903855880157,
+      "grad_norm": 1.6255873441696167,
+      "learning_rate": 3.2088530574992393e-07,
+      "loss": 0.9498,
+      "step": 28865
+    },
+    {
+      "epoch": 0.695310806579803,
+      "grad_norm": 1.4400304555892944,
+      "learning_rate": 3.207585437582395e-07,
+      "loss": 0.9274,
+      "step": 28870
+    },
+    {
+      "epoch": 0.6954312275715903,
+      "grad_norm": 1.3725250959396362,
+      "learning_rate": 3.206317817665551e-07,
+      "loss": 0.9407,
+      "step": 28875
+    },
+    {
+      "epoch": 0.6955516485633776,
+      "grad_norm": 1.577254295349121,
+      "learning_rate": 3.205050197748707e-07,
+      "loss": 0.882,
+      "step": 28880
+    },
+    {
+      "epoch": 0.6956720695551648,
+      "grad_norm": 1.5213263034820557,
+      "learning_rate": 3.203782577831863e-07,
+      "loss": 0.8686,
+      "step": 28885
+    },
+    {
+      "epoch": 0.6957924905469521,
+      "grad_norm": 1.3886874914169312,
+      "learning_rate": 3.2025149579150183e-07,
+      "loss": 0.8521,
+      "step": 28890
+    },
+    {
+      "epoch": 0.6959129115387395,
+      "grad_norm": 1.5748800039291382,
+      "learning_rate": 3.201247337998175e-07,
+      "loss": 0.8903,
+      "step": 28895
+    },
+    {
+      "epoch": 0.6960333325305267,
+      "grad_norm": 1.5380878448486328,
+      "learning_rate": 3.19997971808133e-07,
+      "loss": 0.938,
+      "step": 28900
+    },
+    {
+      "epoch": 0.696153753522314,
+      "grad_norm": 1.555604100227356,
+      "learning_rate": 3.198712098164486e-07,
+      "loss": 0.8907,
+      "step": 28905
+    },
+    {
+      "epoch": 0.6962741745141013,
+      "grad_norm": 1.5657362937927246,
+      "learning_rate": 3.1974444782476425e-07,
+      "loss": 0.8877,
+      "step": 28910
+    },
+    {
+      "epoch": 0.6963945955058886,
+      "grad_norm": 1.5190544128417969,
+      "learning_rate": 3.196176858330798e-07,
+      "loss": 0.887,
+      "step": 28915
+    },
+    {
+      "epoch": 0.6965150164976759,
+      "grad_norm": 1.5380806922912598,
+      "learning_rate": 3.194909238413954e-07,
+      "loss": 0.8979,
+      "step": 28920
+    },
+    {
+      "epoch": 0.6966354374894632,
+      "grad_norm": 1.5136362314224243,
+      "learning_rate": 3.19364161849711e-07,
+      "loss": 0.9221,
+      "step": 28925
+    },
+    {
+      "epoch": 0.6967558584812504,
+      "grad_norm": 1.7348721027374268,
+      "learning_rate": 3.1923739985802657e-07,
+      "loss": 0.862,
+      "step": 28930
+    },
+    {
+      "epoch": 0.6968762794730378,
+      "grad_norm": 1.4176759719848633,
+      "learning_rate": 3.191106378663421e-07,
+      "loss": 0.8833,
+      "step": 28935
+    },
+    {
+      "epoch": 0.696996700464825,
+      "grad_norm": 1.5287283658981323,
+      "learning_rate": 3.1898387587465775e-07,
+      "loss": 0.9482,
+      "step": 28940
+    },
+    {
+      "epoch": 0.6971171214566123,
+      "grad_norm": 1.7669847011566162,
+      "learning_rate": 3.188571138829733e-07,
+      "loss": 0.8706,
+      "step": 28945
+    },
+    {
+      "epoch": 0.6972375424483996,
+      "grad_norm": 1.6608628034591675,
+      "learning_rate": 3.187303518912889e-07,
+      "loss": 0.8909,
+      "step": 28950
+    },
+    {
+      "epoch": 0.6973579634401869,
+      "grad_norm": 1.3679029941558838,
+      "learning_rate": 3.1860358989960453e-07,
+      "loss": 0.8428,
+      "step": 28955
+    },
+    {
+      "epoch": 0.6974783844319742,
+      "grad_norm": 1.4520623683929443,
+      "learning_rate": 3.1847682790792007e-07,
+      "loss": 0.913,
+      "step": 28960
+    },
+    {
+      "epoch": 0.6975988054237615,
+      "grad_norm": 1.913512945175171,
+      "learning_rate": 3.1835006591623566e-07,
+      "loss": 0.8596,
+      "step": 28965
+    },
+    {
+      "epoch": 0.6977192264155487,
+      "grad_norm": 1.517116665840149,
+      "learning_rate": 3.1822330392455125e-07,
+      "loss": 0.9363,
+      "step": 28970
+    },
+    {
+      "epoch": 0.6978396474073361,
+      "grad_norm": 2.0575098991394043,
+      "learning_rate": 3.1809654193286684e-07,
+      "loss": 0.8777,
+      "step": 28975
+    },
+    {
+      "epoch": 0.6979600683991234,
+      "grad_norm": 1.5659877061843872,
+      "learning_rate": 3.179697799411824e-07,
+      "loss": 0.8989,
+      "step": 28980
+    },
+    {
+      "epoch": 0.6980804893909106,
+      "grad_norm": 1.4551013708114624,
+      "learning_rate": 3.17843017949498e-07,
+      "loss": 0.8814,
+      "step": 28985
+    },
+    {
+      "epoch": 0.6982009103826979,
+      "grad_norm": 1.452511191368103,
+      "learning_rate": 3.177162559578136e-07,
+      "loss": 0.9256,
+      "step": 28990
+    },
+    {
+      "epoch": 0.6983213313744852,
+      "grad_norm": 1.4275633096694946,
+      "learning_rate": 3.1758949396612916e-07,
+      "loss": 0.9033,
+      "step": 28995
+    },
+    {
+      "epoch": 0.6984417523662725,
+      "grad_norm": 1.5840181112289429,
+      "learning_rate": 3.174627319744448e-07,
+      "loss": 0.8965,
+      "step": 29000
+    },
+    {
+      "epoch": 0.6985621733580598,
+      "grad_norm": 1.48139226436615,
+      "learning_rate": 3.1733596998276034e-07,
+      "loss": 0.9428,
+      "step": 29005
+    },
+    {
+      "epoch": 0.698682594349847,
+      "grad_norm": 1.5555086135864258,
+      "learning_rate": 3.1720920799107593e-07,
+      "loss": 0.8771,
+      "step": 29010
+    },
+    {
+      "epoch": 0.6988030153416344,
+      "grad_norm": 1.6042619943618774,
+      "learning_rate": 3.170824459993915e-07,
+      "loss": 0.8988,
+      "step": 29015
+    },
+    {
+      "epoch": 0.6989234363334217,
+      "grad_norm": 1.6112546920776367,
+      "learning_rate": 3.169556840077071e-07,
+      "loss": 0.9054,
+      "step": 29020
+    },
+    {
+      "epoch": 0.6990438573252089,
+      "grad_norm": 1.7235147953033447,
+      "learning_rate": 3.168289220160227e-07,
+      "loss": 0.9407,
+      "step": 29025
+    },
+    {
+      "epoch": 0.6991642783169962,
+      "grad_norm": 1.510565161705017,
+      "learning_rate": 3.167021600243383e-07,
+      "loss": 0.8459,
+      "step": 29030
+    },
+    {
+      "epoch": 0.6992846993087835,
+      "grad_norm": 1.5521827936172485,
+      "learning_rate": 3.165753980326539e-07,
+      "loss": 0.9397,
+      "step": 29035
+    },
+    {
+      "epoch": 0.6994051203005708,
+      "grad_norm": 1.5342634916305542,
+      "learning_rate": 3.1644863604096943e-07,
+      "loss": 0.9075,
+      "step": 29040
+    },
+    {
+      "epoch": 0.6995255412923581,
+      "grad_norm": 1.5240405797958374,
+      "learning_rate": 3.1632187404928507e-07,
+      "loss": 0.9667,
+      "step": 29045
+    },
+    {
+      "epoch": 0.6996459622841453,
+      "grad_norm": 1.4187840223312378,
+      "learning_rate": 3.161951120576006e-07,
+      "loss": 0.8983,
+      "step": 29050
+    },
+    {
+      "epoch": 0.6997663832759327,
+      "grad_norm": 1.578110933303833,
+      "learning_rate": 3.160683500659162e-07,
+      "loss": 0.8728,
+      "step": 29055
+    },
+    {
+      "epoch": 0.69988680426772,
+      "grad_norm": 1.782132625579834,
+      "learning_rate": 3.1594158807423185e-07,
+      "loss": 0.9311,
+      "step": 29060
+    },
+    {
+      "epoch": 0.7000072252595072,
+      "grad_norm": 1.8534215688705444,
+      "learning_rate": 3.158148260825474e-07,
+      "loss": 0.8793,
+      "step": 29065
+    },
+    {
+      "epoch": 0.7001276462512945,
+      "grad_norm": 1.4021070003509521,
+      "learning_rate": 3.15688064090863e-07,
+      "loss": 0.889,
+      "step": 29070
+    },
+    {
+      "epoch": 0.7002480672430819,
+      "grad_norm": 1.5727077722549438,
+      "learning_rate": 3.1556130209917857e-07,
+      "loss": 0.9066,
+      "step": 29075
+    },
+    {
+      "epoch": 0.7003684882348691,
+      "grad_norm": 1.6189826726913452,
+      "learning_rate": 3.1543454010749416e-07,
+      "loss": 0.9,
+      "step": 29080
+    },
+    {
+      "epoch": 0.7004889092266564,
+      "grad_norm": 1.587286353111267,
+      "learning_rate": 3.153077781158097e-07,
+      "loss": 0.8862,
+      "step": 29085
+    },
+    {
+      "epoch": 0.7006093302184436,
+      "grad_norm": 1.7006717920303345,
+      "learning_rate": 3.1518101612412535e-07,
+      "loss": 0.9625,
+      "step": 29090
+    },
+    {
+      "epoch": 0.700729751210231,
+      "grad_norm": 1.6427197456359863,
+      "learning_rate": 3.150542541324409e-07,
+      "loss": 0.8765,
+      "step": 29095
+    },
+    {
+      "epoch": 0.7008501722020183,
+      "grad_norm": 1.4622749090194702,
+      "learning_rate": 3.149274921407565e-07,
+      "loss": 0.8812,
+      "step": 29100
+    },
+    {
+      "epoch": 0.7009705931938055,
+      "grad_norm": 1.509885549545288,
+      "learning_rate": 3.148007301490721e-07,
+      "loss": 0.8931,
+      "step": 29105
+    },
+    {
+      "epoch": 0.7010910141855928,
+      "grad_norm": 1.533454418182373,
+      "learning_rate": 3.1467396815738766e-07,
+      "loss": 0.9289,
+      "step": 29110
+    },
+    {
+      "epoch": 0.7012114351773802,
+      "grad_norm": 1.5319466590881348,
+      "learning_rate": 3.1454720616570325e-07,
+      "loss": 0.929,
+      "step": 29115
+    },
+    {
+      "epoch": 0.7013318561691674,
+      "grad_norm": 1.4898655414581299,
+      "learning_rate": 3.1442044417401884e-07,
+      "loss": 0.8779,
+      "step": 29120
+    },
+    {
+      "epoch": 0.7014522771609547,
+      "grad_norm": 1.5144740343093872,
+      "learning_rate": 3.1429368218233444e-07,
+      "loss": 0.8956,
+      "step": 29125
+    },
+    {
+      "epoch": 0.701572698152742,
+      "grad_norm": 1.7286611795425415,
+      "learning_rate": 3.1416692019065e-07,
+      "loss": 0.9521,
+      "step": 29130
+    },
+    {
+      "epoch": 0.7016931191445293,
+      "grad_norm": 2.0020668506622314,
+      "learning_rate": 3.140401581989656e-07,
+      "loss": 0.9005,
+      "step": 29135
+    },
+    {
+      "epoch": 0.7018135401363166,
+      "grad_norm": 1.621082067489624,
+      "learning_rate": 3.139133962072812e-07,
+      "loss": 0.9036,
+      "step": 29140
+    },
+    {
+      "epoch": 0.7019339611281038,
+      "grad_norm": 1.5306613445281982,
+      "learning_rate": 3.1378663421559675e-07,
+      "loss": 0.9365,
+      "step": 29145
+    },
+    {
+      "epoch": 0.7020543821198911,
+      "grad_norm": 1.7362675666809082,
+      "learning_rate": 3.136598722239124e-07,
+      "loss": 0.8838,
+      "step": 29150
+    },
+    {
+      "epoch": 0.7021748031116785,
+      "grad_norm": 1.5057249069213867,
+      "learning_rate": 3.1353311023222793e-07,
+      "loss": 0.8765,
+      "step": 29155
+    },
+    {
+      "epoch": 0.7022952241034657,
+      "grad_norm": 1.4992295503616333,
+      "learning_rate": 3.134063482405435e-07,
+      "loss": 0.9159,
+      "step": 29160
+    },
+    {
+      "epoch": 0.702415645095253,
+      "grad_norm": 1.5106385946273804,
+      "learning_rate": 3.132795862488591e-07,
+      "loss": 0.8865,
+      "step": 29165
+    },
+    {
+      "epoch": 0.7025360660870403,
+      "grad_norm": 1.6429600715637207,
+      "learning_rate": 3.131528242571747e-07,
+      "loss": 0.982,
+      "step": 29170
+    },
+    {
+      "epoch": 0.7026564870788276,
+      "grad_norm": 1.50474214553833,
+      "learning_rate": 3.130260622654903e-07,
+      "loss": 0.823,
+      "step": 29175
+    },
+    {
+      "epoch": 0.7027769080706149,
+      "grad_norm": 1.6841436624526978,
+      "learning_rate": 3.128993002738059e-07,
+      "loss": 0.8769,
+      "step": 29180
+    },
+    {
+      "epoch": 0.7028973290624022,
+      "grad_norm": 1.8708577156066895,
+      "learning_rate": 3.127725382821215e-07,
+      "loss": 0.943,
+      "step": 29185
+    },
+    {
+      "epoch": 0.7030177500541894,
+      "grad_norm": 1.5782015323638916,
+      "learning_rate": 3.12645776290437e-07,
+      "loss": 0.91,
+      "step": 29190
+    },
+    {
+      "epoch": 0.7031381710459768,
+      "grad_norm": 1.6359310150146484,
+      "learning_rate": 3.1251901429875267e-07,
+      "loss": 0.9172,
+      "step": 29195
+    },
+    {
+      "epoch": 0.703258592037764,
+      "grad_norm": 1.3926293849945068,
+      "learning_rate": 3.123922523070682e-07,
+      "loss": 0.9453,
+      "step": 29200
+    },
+    {
+      "epoch": 0.7033790130295513,
+      "grad_norm": 1.5979011058807373,
+      "learning_rate": 3.122654903153838e-07,
+      "loss": 0.9411,
+      "step": 29205
+    },
+    {
+      "epoch": 0.7034994340213386,
+      "grad_norm": 1.612737774848938,
+      "learning_rate": 3.1213872832369944e-07,
+      "loss": 0.8778,
+      "step": 29210
+    },
+    {
+      "epoch": 0.7036198550131259,
+      "grad_norm": 1.5918177366256714,
+      "learning_rate": 3.12011966332015e-07,
+      "loss": 0.9218,
+      "step": 29215
+    },
+    {
+      "epoch": 0.7037402760049132,
+      "grad_norm": 1.55571448802948,
+      "learning_rate": 3.118852043403306e-07,
+      "loss": 0.9203,
+      "step": 29220
+    },
+    {
+      "epoch": 0.7038606969967005,
+      "grad_norm": 1.6012756824493408,
+      "learning_rate": 3.1175844234864617e-07,
+      "loss": 0.9508,
+      "step": 29225
+    },
+    {
+      "epoch": 0.7039811179884877,
+      "grad_norm": 1.6371347904205322,
+      "learning_rate": 3.1163168035696176e-07,
+      "loss": 0.8785,
+      "step": 29230
+    },
+    {
+      "epoch": 0.704101538980275,
+      "grad_norm": 1.51962411403656,
+      "learning_rate": 3.115049183652773e-07,
+      "loss": 0.8845,
+      "step": 29235
+    },
+    {
+      "epoch": 0.7042219599720624,
+      "grad_norm": 1.8233025074005127,
+      "learning_rate": 3.1137815637359294e-07,
+      "loss": 0.9195,
+      "step": 29240
+    },
+    {
+      "epoch": 0.7043423809638496,
+      "grad_norm": 1.4151511192321777,
+      "learning_rate": 3.112513943819085e-07,
+      "loss": 0.8823,
+      "step": 29245
+    },
+    {
+      "epoch": 0.7044628019556369,
+      "grad_norm": 1.5707738399505615,
+      "learning_rate": 3.111246323902241e-07,
+      "loss": 0.918,
+      "step": 29250
+    },
+    {
+      "epoch": 0.7045832229474241,
+      "grad_norm": 1.680002212524414,
+      "learning_rate": 3.109978703985397e-07,
+      "loss": 0.8727,
+      "step": 29255
+    },
+    {
+      "epoch": 0.7047036439392115,
+      "grad_norm": 1.4788126945495605,
+      "learning_rate": 3.1087110840685526e-07,
+      "loss": 0.8589,
+      "step": 29260
+    },
+    {
+      "epoch": 0.7048240649309988,
+      "grad_norm": 1.565387487411499,
+      "learning_rate": 3.107443464151709e-07,
+      "loss": 0.835,
+      "step": 29265
+    },
+    {
+      "epoch": 0.704944485922786,
+      "grad_norm": 1.8114027976989746,
+      "learning_rate": 3.1061758442348644e-07,
+      "loss": 0.8903,
+      "step": 29270
+    },
+    {
+      "epoch": 0.7050649069145734,
+      "grad_norm": 1.5533870458602905,
+      "learning_rate": 3.1049082243180203e-07,
+      "loss": 0.9147,
+      "step": 29275
+    },
+    {
+      "epoch": 0.7051853279063607,
+      "grad_norm": 1.6295989751815796,
+      "learning_rate": 3.103640604401176e-07,
+      "loss": 1.0114,
+      "step": 29280
+    },
+    {
+      "epoch": 0.7053057488981479,
+      "grad_norm": 1.5241661071777344,
+      "learning_rate": 3.102372984484332e-07,
+      "loss": 0.8899,
+      "step": 29285
+    },
+    {
+      "epoch": 0.7054261698899352,
+      "grad_norm": 1.4702953100204468,
+      "learning_rate": 3.101105364567488e-07,
+      "loss": 0.9182,
+      "step": 29290
+    },
+    {
+      "epoch": 0.7055465908817226,
+      "grad_norm": 1.4871397018432617,
+      "learning_rate": 3.099837744650644e-07,
+      "loss": 0.8649,
+      "step": 29295
+    },
+    {
+      "epoch": 0.7056670118735098,
+      "grad_norm": 1.5919462442398071,
+      "learning_rate": 3.0985701247338e-07,
+      "loss": 0.9245,
+      "step": 29300
+    },
+    {
+      "epoch": 0.7057874328652971,
+      "grad_norm": 1.4001860618591309,
+      "learning_rate": 3.0973025048169553e-07,
+      "loss": 0.9326,
+      "step": 29305
+    },
+    {
+      "epoch": 0.7059078538570843,
+      "grad_norm": 1.7274092435836792,
+      "learning_rate": 3.096034884900112e-07,
+      "loss": 0.8802,
+      "step": 29310
+    },
+    {
+      "epoch": 0.7060282748488717,
+      "grad_norm": 1.5036511421203613,
+      "learning_rate": 3.094767264983267e-07,
+      "loss": 0.929,
+      "step": 29315
+    },
+    {
+      "epoch": 0.706148695840659,
+      "grad_norm": 1.533713459968567,
+      "learning_rate": 3.093499645066423e-07,
+      "loss": 0.8605,
+      "step": 29320
+    },
+    {
+      "epoch": 0.7062691168324462,
+      "grad_norm": 1.4081368446350098,
+      "learning_rate": 3.0922320251495795e-07,
+      "loss": 0.9085,
+      "step": 29325
+    },
+    {
+      "epoch": 0.7063895378242335,
+      "grad_norm": 1.609192967414856,
+      "learning_rate": 3.090964405232735e-07,
+      "loss": 0.9238,
+      "step": 29330
+    },
+    {
+      "epoch": 0.7065099588160209,
+      "grad_norm": 1.4841941595077515,
+      "learning_rate": 3.089696785315891e-07,
+      "loss": 0.9246,
+      "step": 29335
+    },
+    {
+      "epoch": 0.7066303798078081,
+      "grad_norm": 1.4160584211349487,
+      "learning_rate": 3.0884291653990467e-07,
+      "loss": 0.8796,
+      "step": 29340
+    },
+    {
+      "epoch": 0.7067508007995954,
+      "grad_norm": 1.6532182693481445,
+      "learning_rate": 3.0871615454822026e-07,
+      "loss": 0.9149,
+      "step": 29345
+    },
+    {
+      "epoch": 0.7068712217913826,
+      "grad_norm": 1.6987322568893433,
+      "learning_rate": 3.085893925565358e-07,
+      "loss": 0.9261,
+      "step": 29350
+    },
+    {
+      "epoch": 0.70699164278317,
+      "grad_norm": 1.4547004699707031,
+      "learning_rate": 3.0846263056485145e-07,
+      "loss": 0.9175,
+      "step": 29355
+    },
+    {
+      "epoch": 0.7071120637749573,
+      "grad_norm": 1.4887008666992188,
+      "learning_rate": 3.0833586857316704e-07,
+      "loss": 0.877,
+      "step": 29360
+    },
+    {
+      "epoch": 0.7072324847667445,
+      "grad_norm": 1.42011296749115,
+      "learning_rate": 3.082091065814826e-07,
+      "loss": 0.946,
+      "step": 29365
+    },
+    {
+      "epoch": 0.7073529057585318,
+      "grad_norm": 1.5739734172821045,
+      "learning_rate": 3.080823445897982e-07,
+      "loss": 0.911,
+      "step": 29370
+    },
+    {
+      "epoch": 0.7074733267503192,
+      "grad_norm": 1.4277094602584839,
+      "learning_rate": 3.0795558259811376e-07,
+      "loss": 0.8827,
+      "step": 29375
+    },
+    {
+      "epoch": 0.7075937477421064,
+      "grad_norm": 1.556277871131897,
+      "learning_rate": 3.0782882060642935e-07,
+      "loss": 0.8724,
+      "step": 29380
+    },
+    {
+      "epoch": 0.7077141687338937,
+      "grad_norm": 1.618062138557434,
+      "learning_rate": 3.0770205861474494e-07,
+      "loss": 0.8724,
+      "step": 29385
+    },
+    {
+      "epoch": 0.707834589725681,
+      "grad_norm": 1.4980077743530273,
+      "learning_rate": 3.0757529662306054e-07,
+      "loss": 0.9093,
+      "step": 29390
+    },
+    {
+      "epoch": 0.7079550107174682,
+      "grad_norm": 1.7104482650756836,
+      "learning_rate": 3.074485346313761e-07,
+      "loss": 0.9211,
+      "step": 29395
+    },
+    {
+      "epoch": 0.7080754317092556,
+      "grad_norm": 1.6199488639831543,
+      "learning_rate": 3.073217726396917e-07,
+      "loss": 0.9073,
+      "step": 29400
+    },
+    {
+      "epoch": 0.7081958527010428,
+      "grad_norm": 1.8200767040252686,
+      "learning_rate": 3.071950106480073e-07,
+      "loss": 0.9386,
+      "step": 29405
+    },
+    {
+      "epoch": 0.7083162736928301,
+      "grad_norm": 1.5339254140853882,
+      "learning_rate": 3.0706824865632285e-07,
+      "loss": 0.8746,
+      "step": 29410
+    },
+    {
+      "epoch": 0.7084366946846175,
+      "grad_norm": 1.588697910308838,
+      "learning_rate": 3.069414866646385e-07,
+      "loss": 0.9552,
+      "step": 29415
+    },
+    {
+      "epoch": 0.7085571156764047,
+      "grad_norm": 1.454702377319336,
+      "learning_rate": 3.0681472467295403e-07,
+      "loss": 0.8876,
+      "step": 29420
+    },
+    {
+      "epoch": 0.708677536668192,
+      "grad_norm": 1.5141271352767944,
+      "learning_rate": 3.066879626812696e-07,
+      "loss": 0.8862,
+      "step": 29425
+    },
+    {
+      "epoch": 0.7087979576599793,
+      "grad_norm": 1.958306074142456,
+      "learning_rate": 3.065612006895852e-07,
+      "loss": 0.9148,
+      "step": 29430
+    },
+    {
+      "epoch": 0.7089183786517665,
+      "grad_norm": 1.6023107767105103,
+      "learning_rate": 3.064344386979008e-07,
+      "loss": 0.8941,
+      "step": 29435
+    },
+    {
+      "epoch": 0.7090387996435539,
+      "grad_norm": 1.651862382888794,
+      "learning_rate": 3.063076767062164e-07,
+      "loss": 0.929,
+      "step": 29440
+    },
+    {
+      "epoch": 0.7091592206353412,
+      "grad_norm": 1.775821328163147,
+      "learning_rate": 3.06180914714532e-07,
+      "loss": 0.9159,
+      "step": 29445
+    },
+    {
+      "epoch": 0.7092796416271284,
+      "grad_norm": 1.7366076707839966,
+      "learning_rate": 3.060541527228476e-07,
+      "loss": 0.9161,
+      "step": 29450
+    },
+    {
+      "epoch": 0.7094000626189157,
+      "grad_norm": 1.4495103359222412,
+      "learning_rate": 3.059273907311631e-07,
+      "loss": 0.9226,
+      "step": 29455
+    },
+    {
+      "epoch": 0.709520483610703,
+      "grad_norm": 1.4665801525115967,
+      "learning_rate": 3.0580062873947877e-07,
+      "loss": 0.8856,
+      "step": 29460
+    },
+    {
+      "epoch": 0.7096409046024903,
+      "grad_norm": 1.6649484634399414,
+      "learning_rate": 3.056738667477943e-07,
+      "loss": 0.8967,
+      "step": 29465
+    },
+    {
+      "epoch": 0.7097613255942776,
+      "grad_norm": 1.6469159126281738,
+      "learning_rate": 3.055471047561099e-07,
+      "loss": 0.8974,
+      "step": 29470
+    },
+    {
+      "epoch": 0.7098817465860648,
+      "grad_norm": 1.42695152759552,
+      "learning_rate": 3.0542034276442554e-07,
+      "loss": 0.9147,
+      "step": 29475
+    },
+    {
+      "epoch": 0.7100021675778522,
+      "grad_norm": 1.5906567573547363,
+      "learning_rate": 3.052935807727411e-07,
+      "loss": 0.8881,
+      "step": 29480
+    },
+    {
+      "epoch": 0.7101225885696395,
+      "grad_norm": 1.5483026504516602,
+      "learning_rate": 3.051668187810567e-07,
+      "loss": 0.934,
+      "step": 29485
+    },
+    {
+      "epoch": 0.7102430095614267,
+      "grad_norm": 1.3973199129104614,
+      "learning_rate": 3.0504005678937227e-07,
+      "loss": 0.8659,
+      "step": 29490
+    },
+    {
+      "epoch": 0.710363430553214,
+      "grad_norm": 1.7336456775665283,
+      "learning_rate": 3.0491329479768786e-07,
+      "loss": 0.9344,
+      "step": 29495
+    },
+    {
+      "epoch": 0.7104838515450014,
+      "grad_norm": 1.6791404485702515,
+      "learning_rate": 3.047865328060034e-07,
+      "loss": 0.905,
+      "step": 29500
+    },
+    {
+      "epoch": 0.7106042725367886,
+      "grad_norm": 1.6338578462600708,
+      "learning_rate": 3.0465977081431904e-07,
+      "loss": 0.8957,
+      "step": 29505
+    },
+    {
+      "epoch": 0.7107246935285759,
+      "grad_norm": 1.517227292060852,
+      "learning_rate": 3.0453300882263463e-07,
+      "loss": 0.8981,
+      "step": 29510
+    },
+    {
+      "epoch": 0.7108451145203631,
+      "grad_norm": 1.650689959526062,
+      "learning_rate": 3.0440624683095017e-07,
+      "loss": 0.9155,
+      "step": 29515
+    },
+    {
+      "epoch": 0.7109655355121505,
+      "grad_norm": 1.6722382307052612,
+      "learning_rate": 3.042794848392658e-07,
+      "loss": 0.8428,
+      "step": 29520
+    },
+    {
+      "epoch": 0.7110859565039378,
+      "grad_norm": 1.6126636266708374,
+      "learning_rate": 3.0415272284758136e-07,
+      "loss": 0.9016,
+      "step": 29525
+    },
+    {
+      "epoch": 0.711206377495725,
+      "grad_norm": 1.598414659500122,
+      "learning_rate": 3.0402596085589695e-07,
+      "loss": 0.9092,
+      "step": 29530
+    },
+    {
+      "epoch": 0.7113267984875123,
+      "grad_norm": 1.6130987405776978,
+      "learning_rate": 3.0389919886421254e-07,
+      "loss": 0.9736,
+      "step": 29535
+    },
+    {
+      "epoch": 0.7114472194792997,
+      "grad_norm": 1.5357635021209717,
+      "learning_rate": 3.0377243687252813e-07,
+      "loss": 0.9343,
+      "step": 29540
+    },
+    {
+      "epoch": 0.7115676404710869,
+      "grad_norm": 1.6397613286972046,
+      "learning_rate": 3.0364567488084367e-07,
+      "loss": 0.9212,
+      "step": 29545
+    },
+    {
+      "epoch": 0.7116880614628742,
+      "grad_norm": 1.4497694969177246,
+      "learning_rate": 3.035189128891593e-07,
+      "loss": 0.9303,
+      "step": 29550
+    },
+    {
+      "epoch": 0.7118084824546616,
+      "grad_norm": 1.4531131982803345,
+      "learning_rate": 3.033921508974749e-07,
+      "loss": 0.9327,
+      "step": 29555
+    },
+    {
+      "epoch": 0.7119289034464488,
+      "grad_norm": 1.509235143661499,
+      "learning_rate": 3.0326538890579045e-07,
+      "loss": 0.9276,
+      "step": 29560
+    },
+    {
+      "epoch": 0.7120493244382361,
+      "grad_norm": 1.6412030458450317,
+      "learning_rate": 3.031386269141061e-07,
+      "loss": 0.9299,
+      "step": 29565
+    },
+    {
+      "epoch": 0.7121697454300233,
+      "grad_norm": 1.4233083724975586,
+      "learning_rate": 3.0301186492242163e-07,
+      "loss": 0.8982,
+      "step": 29570
+    },
+    {
+      "epoch": 0.7122901664218106,
+      "grad_norm": 1.3822218179702759,
+      "learning_rate": 3.028851029307372e-07,
+      "loss": 0.8684,
+      "step": 29575
+    },
+    {
+      "epoch": 0.712410587413598,
+      "grad_norm": 1.3487155437469482,
+      "learning_rate": 3.027583409390528e-07,
+      "loss": 0.8833,
+      "step": 29580
+    },
+    {
+      "epoch": 0.7125310084053852,
+      "grad_norm": 1.3766354322433472,
+      "learning_rate": 3.026315789473684e-07,
+      "loss": 0.7947,
+      "step": 29585
+    },
+    {
+      "epoch": 0.7126514293971725,
+      "grad_norm": 1.5623881816864014,
+      "learning_rate": 3.02504816955684e-07,
+      "loss": 0.8732,
+      "step": 29590
+    },
+    {
+      "epoch": 0.7127718503889598,
+      "grad_norm": 1.7563700675964355,
+      "learning_rate": 3.023780549639996e-07,
+      "loss": 0.9263,
+      "step": 29595
+    },
+    {
+      "epoch": 0.7128922713807471,
+      "grad_norm": 1.507169485092163,
+      "learning_rate": 3.022512929723152e-07,
+      "loss": 0.9059,
+      "step": 29600
+    },
+    {
+      "epoch": 0.7130126923725344,
+      "grad_norm": 1.4992848634719849,
+      "learning_rate": 3.021245309806307e-07,
+      "loss": 0.8569,
+      "step": 29605
+    },
+    {
+      "epoch": 0.7131331133643216,
+      "grad_norm": 1.4725024700164795,
+      "learning_rate": 3.0199776898894636e-07,
+      "loss": 0.9172,
+      "step": 29610
+    },
+    {
+      "epoch": 0.7132535343561089,
+      "grad_norm": 1.655243158340454,
+      "learning_rate": 3.018710069972619e-07,
+      "loss": 0.9222,
+      "step": 29615
+    },
+    {
+      "epoch": 0.7133739553478963,
+      "grad_norm": 1.535508394241333,
+      "learning_rate": 3.017442450055775e-07,
+      "loss": 0.9346,
+      "step": 29620
+    },
+    {
+      "epoch": 0.7134943763396835,
+      "grad_norm": 1.552764892578125,
+      "learning_rate": 3.0161748301389314e-07,
+      "loss": 0.9204,
+      "step": 29625
+    },
+    {
+      "epoch": 0.7136147973314708,
+      "grad_norm": 1.6510361433029175,
+      "learning_rate": 3.014907210222087e-07,
+      "loss": 0.9179,
+      "step": 29630
+    },
+    {
+      "epoch": 0.7137352183232581,
+      "grad_norm": 1.5170223712921143,
+      "learning_rate": 3.0136395903052427e-07,
+      "loss": 0.8754,
+      "step": 29635
+    },
+    {
+      "epoch": 0.7138556393150454,
+      "grad_norm": 1.6754573583602905,
+      "learning_rate": 3.0123719703883986e-07,
+      "loss": 0.9019,
+      "step": 29640
+    },
+    {
+      "epoch": 0.7139760603068327,
+      "grad_norm": 1.57989501953125,
+      "learning_rate": 3.0111043504715545e-07,
+      "loss": 0.9705,
+      "step": 29645
+    },
+    {
+      "epoch": 0.71409648129862,
+      "grad_norm": 1.874380111694336,
+      "learning_rate": 3.00983673055471e-07,
+      "loss": 0.826,
+      "step": 29650
+    },
+    {
+      "epoch": 0.7142169022904072,
+      "grad_norm": 1.5729446411132812,
+      "learning_rate": 3.0085691106378664e-07,
+      "loss": 0.8823,
+      "step": 29655
+    },
+    {
+      "epoch": 0.7143373232821946,
+      "grad_norm": 1.636989712715149,
+      "learning_rate": 3.007301490721022e-07,
+      "loss": 0.8943,
+      "step": 29660
+    },
+    {
+      "epoch": 0.7144577442739818,
+      "grad_norm": 1.538765549659729,
+      "learning_rate": 3.0060338708041777e-07,
+      "loss": 0.9151,
+      "step": 29665
+    },
+    {
+      "epoch": 0.7145781652657691,
+      "grad_norm": 1.378821611404419,
+      "learning_rate": 3.004766250887334e-07,
+      "loss": 0.8793,
+      "step": 29670
+    },
+    {
+      "epoch": 0.7146985862575564,
+      "grad_norm": 1.720047950744629,
+      "learning_rate": 3.0034986309704895e-07,
+      "loss": 0.8541,
+      "step": 29675
+    },
+    {
+      "epoch": 0.7148190072493437,
+      "grad_norm": 1.70628821849823,
+      "learning_rate": 3.0022310110536454e-07,
+      "loss": 0.8768,
+      "step": 29680
+    },
+    {
+      "epoch": 0.714939428241131,
+      "grad_norm": 1.598002314567566,
+      "learning_rate": 3.0009633911368013e-07,
+      "loss": 0.8626,
+      "step": 29685
+    },
+    {
+      "epoch": 0.7150598492329183,
+      "grad_norm": 1.4448331594467163,
+      "learning_rate": 2.999695771219957e-07,
+      "loss": 0.8836,
+      "step": 29690
+    },
+    {
+      "epoch": 0.7151802702247055,
+      "grad_norm": 1.5112354755401611,
+      "learning_rate": 2.9984281513031127e-07,
+      "loss": 0.9152,
+      "step": 29695
+    },
+    {
+      "epoch": 0.7153006912164929,
+      "grad_norm": 1.4921990633010864,
+      "learning_rate": 2.997160531386269e-07,
+      "loss": 0.8707,
+      "step": 29700
+    },
+    {
+      "epoch": 0.7154211122082802,
+      "grad_norm": 1.6580222845077515,
+      "learning_rate": 2.995892911469425e-07,
+      "loss": 0.9313,
+      "step": 29705
+    },
+    {
+      "epoch": 0.7155415332000674,
+      "grad_norm": 1.7308915853500366,
+      "learning_rate": 2.9946252915525804e-07,
+      "loss": 0.925,
+      "step": 29710
+    },
+    {
+      "epoch": 0.7156619541918547,
+      "grad_norm": 1.6118513345718384,
+      "learning_rate": 2.993357671635737e-07,
+      "loss": 0.8931,
+      "step": 29715
+    },
+    {
+      "epoch": 0.715782375183642,
+      "grad_norm": 1.7087616920471191,
+      "learning_rate": 2.992090051718892e-07,
+      "loss": 0.8905,
+      "step": 29720
+    },
+    {
+      "epoch": 0.7159027961754293,
+      "grad_norm": 1.477902889251709,
+      "learning_rate": 2.990822431802048e-07,
+      "loss": 0.8966,
+      "step": 29725
+    },
+    {
+      "epoch": 0.7160232171672166,
+      "grad_norm": 1.3396321535110474,
+      "learning_rate": 2.989554811885204e-07,
+      "loss": 0.9194,
+      "step": 29730
+    },
+    {
+      "epoch": 0.7161436381590038,
+      "grad_norm": 1.9310853481292725,
+      "learning_rate": 2.98828719196836e-07,
+      "loss": 0.877,
+      "step": 29735
+    },
+    {
+      "epoch": 0.7162640591507912,
+      "grad_norm": 1.5338441133499146,
+      "learning_rate": 2.987019572051516e-07,
+      "loss": 0.9028,
+      "step": 29740
+    },
+    {
+      "epoch": 0.7163844801425785,
+      "grad_norm": 1.3704942464828491,
+      "learning_rate": 2.985751952134672e-07,
+      "loss": 0.9116,
+      "step": 29745
+    },
+    {
+      "epoch": 0.7165049011343657,
+      "grad_norm": 1.5461375713348389,
+      "learning_rate": 2.984484332217828e-07,
+      "loss": 0.9145,
+      "step": 29750
+    },
+    {
+      "epoch": 0.716625322126153,
+      "grad_norm": 1.5435659885406494,
+      "learning_rate": 2.983216712300983e-07,
+      "loss": 0.9219,
+      "step": 29755
+    },
+    {
+      "epoch": 0.7167457431179404,
+      "grad_norm": 1.4328842163085938,
+      "learning_rate": 2.9819490923841396e-07,
+      "loss": 0.8917,
+      "step": 29760
+    },
+    {
+      "epoch": 0.7168661641097276,
+      "grad_norm": 1.4749674797058105,
+      "learning_rate": 2.980681472467295e-07,
+      "loss": 0.8919,
+      "step": 29765
+    },
+    {
+      "epoch": 0.7169865851015149,
+      "grad_norm": 2.0215656757354736,
+      "learning_rate": 2.9794138525504514e-07,
+      "loss": 0.8829,
+      "step": 29770
+    },
+    {
+      "epoch": 0.7171070060933021,
+      "grad_norm": 1.995144248008728,
+      "learning_rate": 2.9781462326336073e-07,
+      "loss": 0.9261,
+      "step": 29775
+    },
+    {
+      "epoch": 0.7172274270850895,
+      "grad_norm": 1.4340596199035645,
+      "learning_rate": 2.9768786127167627e-07,
+      "loss": 0.892,
+      "step": 29780
+    },
+    {
+      "epoch": 0.7173478480768768,
+      "grad_norm": 1.4539889097213745,
+      "learning_rate": 2.975610992799919e-07,
+      "loss": 0.8956,
+      "step": 29785
+    },
+    {
+      "epoch": 0.717468269068664,
+      "grad_norm": 1.598737120628357,
+      "learning_rate": 2.9743433728830746e-07,
+      "loss": 0.895,
+      "step": 29790
+    },
+    {
+      "epoch": 0.7175886900604513,
+      "grad_norm": 1.61039137840271,
+      "learning_rate": 2.9730757529662305e-07,
+      "loss": 0.971,
+      "step": 29795
+    },
+    {
+      "epoch": 0.7177091110522387,
+      "grad_norm": 1.8494083881378174,
+      "learning_rate": 2.9718081330493864e-07,
+      "loss": 0.8593,
+      "step": 29800
+    },
+    {
+      "epoch": 0.7178295320440259,
+      "grad_norm": 1.4721626043319702,
+      "learning_rate": 2.9705405131325423e-07,
+      "loss": 0.8918,
+      "step": 29805
+    },
+    {
+      "epoch": 0.7179499530358132,
+      "grad_norm": 2.004418134689331,
+      "learning_rate": 2.9692728932156977e-07,
+      "loss": 0.9148,
+      "step": 29810
+    },
+    {
+      "epoch": 0.7180703740276005,
+      "grad_norm": 1.5343745946884155,
+      "learning_rate": 2.968005273298854e-07,
+      "loss": 0.8677,
+      "step": 29815
+    },
+    {
+      "epoch": 0.7181907950193878,
+      "grad_norm": 1.5843580961227417,
+      "learning_rate": 2.96673765338201e-07,
+      "loss": 0.8925,
+      "step": 29820
+    },
+    {
+      "epoch": 0.7183112160111751,
+      "grad_norm": 1.5251814126968384,
+      "learning_rate": 2.9654700334651655e-07,
+      "loss": 0.8998,
+      "step": 29825
+    },
+    {
+      "epoch": 0.7184316370029623,
+      "grad_norm": 1.525067687034607,
+      "learning_rate": 2.964202413548322e-07,
+      "loss": 0.8747,
+      "step": 29830
+    },
+    {
+      "epoch": 0.7185520579947496,
+      "grad_norm": 1.5406259298324585,
+      "learning_rate": 2.9629347936314773e-07,
+      "loss": 0.9172,
+      "step": 29835
+    },
+    {
+      "epoch": 0.718672478986537,
+      "grad_norm": 1.8802706003189087,
+      "learning_rate": 2.961667173714633e-07,
+      "loss": 0.9207,
+      "step": 29840
+    },
+    {
+      "epoch": 0.7187928999783242,
+      "grad_norm": 1.6404411792755127,
+      "learning_rate": 2.960399553797789e-07,
+      "loss": 0.9423,
+      "step": 29845
+    },
+    {
+      "epoch": 0.7189133209701115,
+      "grad_norm": 1.5202003717422485,
+      "learning_rate": 2.959131933880945e-07,
+      "loss": 0.8866,
+      "step": 29850
+    },
+    {
+      "epoch": 0.7190337419618988,
+      "grad_norm": 1.4623454809188843,
+      "learning_rate": 2.957864313964101e-07,
+      "loss": 0.8896,
+      "step": 29855
+    },
+    {
+      "epoch": 0.7191541629536861,
+      "grad_norm": 1.5731388330459595,
+      "learning_rate": 2.956596694047257e-07,
+      "loss": 0.8521,
+      "step": 29860
+    },
+    {
+      "epoch": 0.7192745839454734,
+      "grad_norm": 1.6413127183914185,
+      "learning_rate": 2.955329074130413e-07,
+      "loss": 0.9062,
+      "step": 29865
+    },
+    {
+      "epoch": 0.7193950049372607,
+      "grad_norm": 1.5317072868347168,
+      "learning_rate": 2.954061454213568e-07,
+      "loss": 0.858,
+      "step": 29870
+    },
+    {
+      "epoch": 0.7195154259290479,
+      "grad_norm": 1.4505599737167358,
+      "learning_rate": 2.9527938342967246e-07,
+      "loss": 0.8647,
+      "step": 29875
+    },
+    {
+      "epoch": 0.7196358469208353,
+      "grad_norm": 1.4833729267120361,
+      "learning_rate": 2.95152621437988e-07,
+      "loss": 0.8414,
+      "step": 29880
+    },
+    {
+      "epoch": 0.7197562679126225,
+      "grad_norm": 1.6453678607940674,
+      "learning_rate": 2.950258594463036e-07,
+      "loss": 0.8929,
+      "step": 29885
+    },
+    {
+      "epoch": 0.7198766889044098,
+      "grad_norm": 1.612945556640625,
+      "learning_rate": 2.9489909745461924e-07,
+      "loss": 0.8469,
+      "step": 29890
+    },
+    {
+      "epoch": 0.7199971098961971,
+      "grad_norm": 1.5266822576522827,
+      "learning_rate": 2.947723354629348e-07,
+      "loss": 0.8734,
+      "step": 29895
+    },
+    {
+      "epoch": 0.7201175308879844,
+      "grad_norm": 1.4711270332336426,
+      "learning_rate": 2.9464557347125037e-07,
+      "loss": 0.9144,
+      "step": 29900
+    },
+    {
+      "epoch": 0.7202379518797717,
+      "grad_norm": 1.4607304334640503,
+      "learning_rate": 2.9451881147956596e-07,
+      "loss": 0.9439,
+      "step": 29905
+    },
+    {
+      "epoch": 0.720358372871559,
+      "grad_norm": 1.421431541442871,
+      "learning_rate": 2.9439204948788155e-07,
+      "loss": 0.8784,
+      "step": 29910
+    },
+    {
+      "epoch": 0.7204787938633462,
+      "grad_norm": 1.4766961336135864,
+      "learning_rate": 2.942652874961971e-07,
+      "loss": 0.8832,
+      "step": 29915
+    },
+    {
+      "epoch": 0.7205992148551336,
+      "grad_norm": 1.451119303703308,
+      "learning_rate": 2.9413852550451274e-07,
+      "loss": 0.9138,
+      "step": 29920
+    },
+    {
+      "epoch": 0.7207196358469208,
+      "grad_norm": 1.7488954067230225,
+      "learning_rate": 2.9401176351282833e-07,
+      "loss": 0.8863,
+      "step": 29925
+    },
+    {
+      "epoch": 0.7208400568387081,
+      "grad_norm": 1.5258170366287231,
+      "learning_rate": 2.9388500152114387e-07,
+      "loss": 0.8871,
+      "step": 29930
+    },
+    {
+      "epoch": 0.7209604778304954,
+      "grad_norm": 2.0056021213531494,
+      "learning_rate": 2.937582395294595e-07,
+      "loss": 0.8868,
+      "step": 29935
+    },
+    {
+      "epoch": 0.7210808988222827,
+      "grad_norm": 1.5685434341430664,
+      "learning_rate": 2.9363147753777505e-07,
+      "loss": 0.8433,
+      "step": 29940
+    },
+    {
+      "epoch": 0.72120131981407,
+      "grad_norm": 1.5814381837844849,
+      "learning_rate": 2.9350471554609064e-07,
+      "loss": 0.906,
+      "step": 29945
+    },
+    {
+      "epoch": 0.7213217408058573,
+      "grad_norm": 1.5315696001052856,
+      "learning_rate": 2.9337795355440623e-07,
+      "loss": 0.8956,
+      "step": 29950
+    },
+    {
+      "epoch": 0.7214421617976445,
+      "grad_norm": 1.589921236038208,
+      "learning_rate": 2.9325119156272183e-07,
+      "loss": 0.9191,
+      "step": 29955
+    },
+    {
+      "epoch": 0.7215625827894319,
+      "grad_norm": 1.6248891353607178,
+      "learning_rate": 2.9312442957103737e-07,
+      "loss": 0.9341,
+      "step": 29960
+    },
+    {
+      "epoch": 0.7216830037812192,
+      "grad_norm": 1.5531809329986572,
+      "learning_rate": 2.92997667579353e-07,
+      "loss": 0.8565,
+      "step": 29965
+    },
+    {
+      "epoch": 0.7218034247730064,
+      "grad_norm": 1.4952003955841064,
+      "learning_rate": 2.928709055876686e-07,
+      "loss": 0.9033,
+      "step": 29970
+    },
+    {
+      "epoch": 0.7219238457647937,
+      "grad_norm": 1.6778074502944946,
+      "learning_rate": 2.9274414359598414e-07,
+      "loss": 0.8708,
+      "step": 29975
+    },
+    {
+      "epoch": 0.722044266756581,
+      "grad_norm": 1.4443367719650269,
+      "learning_rate": 2.926173816042998e-07,
+      "loss": 0.9148,
+      "step": 29980
+    },
+    {
+      "epoch": 0.7221646877483683,
+      "grad_norm": 1.5499740839004517,
+      "learning_rate": 2.924906196126153e-07,
+      "loss": 0.9231,
+      "step": 29985
+    },
+    {
+      "epoch": 0.7222851087401556,
+      "grad_norm": 1.434739589691162,
+      "learning_rate": 2.923638576209309e-07,
+      "loss": 0.8409,
+      "step": 29990
+    },
+    {
+      "epoch": 0.7224055297319428,
+      "grad_norm": 1.4947270154953003,
+      "learning_rate": 2.922370956292465e-07,
+      "loss": 0.8939,
+      "step": 29995
+    },
+    {
+      "epoch": 0.7225259507237302,
+      "grad_norm": 1.7011665105819702,
+      "learning_rate": 2.921103336375621e-07,
+      "loss": 0.8744,
+      "step": 30000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 41521,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.180075135360408e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}