diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/last-checkpoint/trainer_state.json"
@@ -0,0 +1,7033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.1414777349414636,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00014147773494146358,
+      "grad_norm": 1.199510931968689,
+      "learning_rate": 0.00019999950652018584,
+      "loss": 1.3974,
+      "step": 1
+    },
+    {
+      "epoch": 0.00028295546988292715,
+      "grad_norm": 2.689831495285034,
+      "learning_rate": 0.0001999980260856137,
+      "loss": 2.1709,
+      "step": 2
+    },
+    {
+      "epoch": 0.0004244332048243908,
+      "grad_norm": 3.1065266132354736,
+      "learning_rate": 0.000199995558710895,
+      "loss": 1.6518,
+      "step": 3
+    },
+    {
+      "epoch": 0.0005659109397658543,
+      "grad_norm": 2.5816352367401123,
+      "learning_rate": 0.00019999210442038162,
+      "loss": 1.1529,
+      "step": 4
+    },
+    {
+      "epoch": 0.0007073886747073179,
+      "grad_norm": 3.101250171661377,
+      "learning_rate": 0.00019998766324816607,
+      "loss": 0.6761,
+      "step": 5
+    },
+    {
+      "epoch": 0.0008488664096487816,
+      "grad_norm": 3.0736050605773926,
+      "learning_rate": 0.0001999822352380809,
+      "loss": 0.3526,
+      "step": 6
+    },
+    {
+      "epoch": 0.0009903441445902451,
+      "grad_norm": 4.289332866668701,
+      "learning_rate": 0.00019997582044369843,
+      "loss": 0.3733,
+      "step": 7
+    },
+    {
+      "epoch": 0.0011318218795317086,
+      "grad_norm": 3.8449699878692627,
+      "learning_rate": 0.00019996841892833,
+      "loss": 0.2397,
+      "step": 8
+    },
+    {
+      "epoch": 0.0012732996144731723,
+      "grad_norm": 7.1344475746154785,
+      "learning_rate": 0.00019996003076502565,
+      "loss": 0.3184,
+      "step": 9
+    },
+    {
+      "epoch": 0.0014147773494146358,
+      "grad_norm": 0.7714758515357971,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 0.0618,
+      "step": 10
+    },
+    {
+      "epoch": 0.0015562550843560995,
+      "grad_norm": 1.3134876489639282,
+      "learning_rate": 0.0001999402948354973,
+      "loss": 0.0192,
+      "step": 11
+    },
+    {
+      "epoch": 0.0016977328192975631,
+      "grad_norm": 1.4721983671188354,
+      "learning_rate": 0.00019992894726405893,
+      "loss": 0.0161,
+      "step": 12
+    },
+    {
+      "epoch": 0.0018392105542390266,
+      "grad_norm": 0.1445562243461609,
+      "learning_rate": 0.000199916613434254,
+      "loss": 0.005,
+      "step": 13
+    },
+    {
+      "epoch": 0.0019806882891804903,
+      "grad_norm": 2.260159730911255,
+      "learning_rate": 0.0001999032934678125,
+      "loss": 0.0221,
+      "step": 14
+    },
+    {
+      "epoch": 0.002122166024121954,
+      "grad_norm": 0.06021130084991455,
+      "learning_rate": 0.00019988898749619702,
+      "loss": 0.0017,
+      "step": 15
+    },
+    {
+      "epoch": 0.0022636437590634172,
+      "grad_norm": 0.09766138345003128,
+      "learning_rate": 0.00019987369566060176,
+      "loss": 0.0013,
+      "step": 16
+    },
+    {
+      "epoch": 0.002405121494004881,
+      "grad_norm": 0.1634778529405594,
+      "learning_rate": 0.00019985741811195097,
+      "loss": 0.0025,
+      "step": 17
+    },
+    {
+      "epoch": 0.0025465992289463446,
+      "grad_norm": 2.076612949371338,
+      "learning_rate": 0.00019984015501089752,
+      "loss": 0.024,
+      "step": 18
+    },
+    {
+      "epoch": 0.0026880769638878083,
+      "grad_norm": 2.0792248249053955,
+      "learning_rate": 0.0001998219065278212,
+      "loss": 0.016,
+      "step": 19
+    },
+    {
+      "epoch": 0.0028295546988292715,
+      "grad_norm": 0.0070343296974897385,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 0.0002,
+      "step": 20
+    },
+    {
+      "epoch": 0.0029710324337707352,
+      "grad_norm": 0.5132514834403992,
+      "learning_rate": 0.00019978245414574417,
+      "loss": 0.0027,
+      "step": 21
+    },
+    {
+      "epoch": 0.003112510168712199,
+      "grad_norm": 0.015104507096111774,
+      "learning_rate": 0.00019976125063612252,
+      "loss": 0.0003,
+      "step": 22
+    },
+    {
+      "epoch": 0.0032539879036536626,
+      "grad_norm": 0.03561875969171524,
+      "learning_rate": 0.00019973906252323238,
+      "loss": 0.0004,
+      "step": 23
+    },
+    {
+      "epoch": 0.0033954656385951263,
+      "grad_norm": 1.239057183265686,
+      "learning_rate": 0.0001997158900260614,
+      "loss": 0.0068,
+      "step": 24
+    },
+    {
+      "epoch": 0.0035369433735365895,
+      "grad_norm": 0.004514714702963829,
+      "learning_rate": 0.0001996917333733128,
+      "loss": 0.0001,
+      "step": 25
+    },
+    {
+      "epoch": 0.0036784211084780532,
+      "grad_norm": 0.0027808404993265867,
+      "learning_rate": 0.00019966659280340297,
+      "loss": 0.0001,
+      "step": 26
+    },
+    {
+      "epoch": 0.003819898843419517,
+      "grad_norm": 1.516070008277893,
+      "learning_rate": 0.00019964046856445924,
+      "loss": 0.0102,
+      "step": 27
+    },
+    {
+      "epoch": 0.003961376578360981,
+      "grad_norm": 0.058614715933799744,
+      "learning_rate": 0.00019961336091431727,
+      "loss": 0.0002,
+      "step": 28
+    },
+    {
+      "epoch": 0.004102854313302444,
+      "grad_norm": 0.12913420796394348,
+      "learning_rate": 0.00019958527012051857,
+      "loss": 0.0023,
+      "step": 29
+    },
+    {
+      "epoch": 0.004244332048243908,
+      "grad_norm": 0.010314574465155602,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 0.0002,
+      "step": 30
+    },
+    {
+      "epoch": 0.004385809783185371,
+      "grad_norm": 0.0008627488277852535,
+      "learning_rate": 0.00019952614022063084,
+      "loss": 0.0,
+      "step": 31
+    },
+    {
+      "epoch": 0.0045272875181268345,
+      "grad_norm": NaN,
+      "learning_rate": 0.00019952614022063084,
+      "loss": 0.2811,
+      "step": 32
+    },
+    {
+      "epoch": 0.004668765253068298,
+      "grad_norm": 1.236807942390442,
+      "learning_rate": 0.00019949510169813003,
+      "loss": 0.0072,
+      "step": 33
+    },
+    {
+      "epoch": 0.004810242988009762,
+      "grad_norm": 0.003006057580932975,
+      "learning_rate": 0.00019946308119914323,
+      "loss": 0.0001,
+      "step": 34
+    },
+    {
+      "epoch": 0.0049517207229512255,
+      "grad_norm": 0.43032628297805786,
+      "learning_rate": 0.0001994300790396999,
+      "loss": 0.0035,
+      "step": 35
+    },
+    {
+      "epoch": 0.005093198457892689,
+      "grad_norm": 0.001233522198162973,
+      "learning_rate": 0.000199396095545518,
+      "loss": 0.0,
+      "step": 36
+    },
+    {
+      "epoch": 0.005234676192834153,
+      "grad_norm": 0.0023691693786531687,
+      "learning_rate": 0.00019936113105200085,
+      "loss": 0.0001,
+      "step": 37
+    },
+    {
+      "epoch": 0.005376153927775617,
+      "grad_norm": 0.025574322789907455,
+      "learning_rate": 0.00019932518590423394,
+      "loss": 0.0002,
+      "step": 38
+    },
+    {
+      "epoch": 0.00551763166271708,
+      "grad_norm": 0.09084252268075943,
+      "learning_rate": 0.00019928826045698136,
+      "loss": 0.0003,
+      "step": 39
+    },
+    {
+      "epoch": 0.005659109397658543,
+      "grad_norm": 1.9598499536514282,
+      "learning_rate": 0.0001992503550746824,
+      "loss": 0.0094,
+      "step": 40
+    },
+    {
+      "epoch": 0.005800587132600007,
+      "grad_norm": 0.001224718289449811,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 0.0,
+      "step": 41
+    },
+    {
+      "epoch": 0.0059420648675414705,
+      "grad_norm": 0.0008977500256150961,
+      "learning_rate": 0.0001991716060110563,
+      "loss": 0.0,
+      "step": 42
+    },
+    {
+      "epoch": 0.006083542602482934,
+      "grad_norm": 0.000888747803401202,
+      "learning_rate": 0.00019913076310695068,
+      "loss": 0.0,
+      "step": 43
+    },
+    {
+      "epoch": 0.006225020337424398,
+      "grad_norm": 0.0025476876180619,
+      "learning_rate": 0.00019908894182223388,
+      "loss": 0.0001,
+      "step": 44
+    },
+    {
+      "epoch": 0.0063664980723658615,
+      "grad_norm": 0.0013534630415961146,
+      "learning_rate": 0.00019904614256966512,
+      "loss": 0.0,
+      "step": 45
+    },
+    {
+      "epoch": 0.006507975807307325,
+      "grad_norm": 0.0009306861320510507,
+      "learning_rate": 0.00019900236577165576,
+      "loss": 0.0001,
+      "step": 46
+    },
+    {
+      "epoch": 0.006649453542248789,
+      "grad_norm": 0.024325449019670486,
+      "learning_rate": 0.0001989576118602651,
+      "loss": 0.0004,
+      "step": 47
+    },
+    {
+      "epoch": 0.006790931277190253,
+      "grad_norm": 0.32905998826026917,
+      "learning_rate": 0.00019891188127719618,
+      "loss": 0.0031,
+      "step": 48
+    },
+    {
+      "epoch": 0.006932409012131715,
+      "grad_norm": 0.002287359442561865,
+      "learning_rate": 0.0001988651744737914,
+      "loss": 0.0001,
+      "step": 49
+    },
+    {
+      "epoch": 0.007073886747073179,
+      "grad_norm": 0.03360757231712341,
+      "learning_rate": 0.00019881749191102808,
+      "loss": 0.0004,
+      "step": 50
+    },
+    {
+      "epoch": 0.007215364482014643,
+      "grad_norm": 0.0011812775628641248,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 0.0,
+      "step": 51
+    },
+    {
+      "epoch": 0.0073568422169561064,
+      "grad_norm": 0.014647467993199825,
+      "learning_rate": 0.00019871920139948192,
+      "loss": 0.0002,
+      "step": 52
+    },
+    {
+      "epoch": 0.00749831995189757,
+      "grad_norm": 0.021761754527688026,
+      "learning_rate": 0.0001986685944207868,
+      "loss": 0.0003,
+      "step": 53
+    },
+    {
+      "epoch": 0.007639797686839034,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001986685944207868,
+      "loss": 0.9778,
+      "step": 54
+    },
+    {
+      "epoch": 0.0077812754217804975,
+      "grad_norm": 0.0026818220503628254,
+      "learning_rate": 0.0001986170136228989,
+      "loss": 0.0,
+      "step": 55
+    },
+    {
+      "epoch": 0.007922753156721961,
+      "grad_norm": 0.562186062335968,
+      "learning_rate": 0.00019856445951489982,
+      "loss": 0.096,
+      "step": 56
+    },
+    {
+      "epoch": 0.008064230891663424,
+      "grad_norm": 0.020677773281931877,
+      "learning_rate": 0.0001985109326154774,
+      "loss": 0.0005,
+      "step": 57
+    },
+    {
+      "epoch": 0.008205708626604889,
+      "grad_norm": 0.01492348127067089,
+      "learning_rate": 0.00019845643345292054,
+      "loss": 0.0003,
+      "step": 58
+    },
+    {
+      "epoch": 0.008347186361546351,
+      "grad_norm": 0.06748542934656143,
+      "learning_rate": 0.00019840096256511398,
+      "loss": 0.0011,
+      "step": 59
+    },
+    {
+      "epoch": 0.008488664096487816,
+      "grad_norm": 0.007372274994850159,
+      "learning_rate": 0.00019834452049953297,
+      "loss": 0.0001,
+      "step": 60
+    },
+    {
+      "epoch": 0.008630141831429279,
+      "grad_norm": 0.002576321130618453,
+      "learning_rate": 0.00019828710781323792,
+      "loss": 0.0001,
+      "step": 61
+    },
+    {
+      "epoch": 0.008771619566370742,
+      "grad_norm": 0.000969462504144758,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 0.0,
+      "step": 62
+    },
+    {
+      "epoch": 0.008913097301312206,
+      "grad_norm": 0.031211236491799355,
+      "learning_rate": 0.0001981693728546399,
+      "loss": 0.0006,
+      "step": 63
+    },
+    {
+      "epoch": 0.009054575036253669,
+      "grad_norm": 0.0024660963099449873,
+      "learning_rate": 0.0001981090517443334,
+      "loss": 0.0001,
+      "step": 64
+    },
+    {
+      "epoch": 0.009196052771195133,
+      "grad_norm": 0.002796717919409275,
+      "learning_rate": 0.00019804776233729444,
+      "loss": 0.0001,
+      "step": 65
+    },
+    {
+      "epoch": 0.009337530506136596,
+      "grad_norm": 0.006275609601289034,
+      "learning_rate": 0.0001979855052384247,
+      "loss": 0.0002,
+      "step": 66
+    },
+    {
+      "epoch": 0.00947900824107806,
+      "grad_norm": 0.0006806784658692777,
+      "learning_rate": 0.00019792228106217658,
+      "loss": 0.0,
+      "step": 67
+    },
+    {
+      "epoch": 0.009620485976019524,
+      "grad_norm": 0.004091021604835987,
+      "learning_rate": 0.00019785809043254722,
+      "loss": 0.0002,
+      "step": 68
+    },
+    {
+      "epoch": 0.009761963710960988,
+      "grad_norm": 0.0047629051841795444,
+      "learning_rate": 0.0001977929339830722,
+      "loss": 0.0002,
+      "step": 69
+    },
+    {
+      "epoch": 0.009903441445902451,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001977929339830722,
+      "loss": 6.0759,
+      "step": 70
+    },
+    {
+      "epoch": 0.010044919180843914,
+      "grad_norm": 0.001333454973064363,
+      "learning_rate": 0.00019772681235681936,
+      "loss": 0.0001,
+      "step": 71
+    },
+    {
+      "epoch": 0.010186396915785378,
+      "grad_norm": 0.02889617159962654,
+      "learning_rate": 0.00019765972620638248,
+      "loss": 0.0006,
+      "step": 72
+    },
+    {
+      "epoch": 0.010327874650726841,
+      "grad_norm": 0.004033127333968878,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 0.0002,
+      "step": 73
+    },
+    {
+      "epoch": 0.010469352385668306,
+      "grad_norm": 0.003473761025816202,
+      "learning_rate": 0.00019752266299092236,
+      "loss": 0.0002,
+      "step": 74
+    },
+    {
+      "epoch": 0.010610830120609769,
+      "grad_norm": 0.0035185706801712513,
+      "learning_rate": 0.00019745268727865774,
+      "loss": 0.0002,
+      "step": 75
+    },
+    {
+      "epoch": 0.010752307855551233,
+      "grad_norm": 0.008891044184565544,
+      "learning_rate": 0.0001973817497477129,
+      "loss": 0.0003,
+      "step": 76
+    },
+    {
+      "epoch": 0.010893785590492696,
+      "grad_norm": 0.001775395474396646,
+      "learning_rate": 0.00019730985109821266,
+      "loss": 0.0001,
+      "step": 77
+    },
+    {
+      "epoch": 0.01103526332543416,
+      "grad_norm": 0.004444313235580921,
+      "learning_rate": 0.00019723699203976766,
+      "loss": 0.0002,
+      "step": 78
+    },
+    {
+      "epoch": 0.011176741060375623,
+      "grad_norm": 7.77717399597168,
+      "learning_rate": 0.0001971631732914674,
+      "loss": 0.1504,
+      "step": 79
+    },
+    {
+      "epoch": 0.011318218795317086,
+      "grad_norm": 0.004705169703811407,
+      "learning_rate": 0.0001970883955818731,
+      "loss": 0.0002,
+      "step": 80
+    },
+    {
+      "epoch": 0.01145969653025855,
+      "grad_norm": 0.03359982743859291,
+      "learning_rate": 0.0001970126596490106,
+      "loss": 0.0009,
+      "step": 81
+    },
+    {
+      "epoch": 0.011601174265200014,
+      "grad_norm": 0.01374754961580038,
+      "learning_rate": 0.00019693596624036292,
+      "loss": 0.0004,
+      "step": 82
+    },
+    {
+      "epoch": 0.011742652000141478,
+      "grad_norm": 0.8847774267196655,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 0.0095,
+      "step": 83
+    },
+    {
+      "epoch": 0.011884129735082941,
+      "grad_norm": 0.23337610065937042,
+      "learning_rate": 0.00019677971003288655,
+      "loss": 0.0033,
+      "step": 84
+    },
+    {
+      "epoch": 0.012025607470024405,
+      "grad_norm": 0.13195404410362244,
+      "learning_rate": 0.00019670014877624353,
+      "loss": 0.0041,
+      "step": 85
+    },
+    {
+      "epoch": 0.012167085204965868,
+      "grad_norm": 0.10853024572134018,
+      "learning_rate": 0.00019661963312817148,
+      "loss": 0.002,
+      "step": 86
+    },
+    {
+      "epoch": 0.012308562939907333,
+      "grad_norm": 0.3537059724330902,
+      "learning_rate": 0.0001965381638833274,
+      "loss": 0.0033,
+      "step": 87
+    },
+    {
+      "epoch": 0.012450040674848796,
+      "grad_norm": 0.009482644498348236,
+      "learning_rate": 0.00019645574184577982,
+      "loss": 0.0004,
+      "step": 88
+    },
+    {
+      "epoch": 0.012591518409790258,
+      "grad_norm": 0.07353401929140091,
+      "learning_rate": 0.000196372367829001,
+      "loss": 0.0012,
+      "step": 89
+    },
+    {
+      "epoch": 0.012732996144731723,
+      "grad_norm": 0.38067692518234253,
+      "learning_rate": 0.00019628804265585877,
+      "loss": 0.0089,
+      "step": 90
+    },
+    {
+      "epoch": 0.012874473879673186,
+      "grad_norm": 0.0006739358650520444,
+      "learning_rate": 0.0001962027671586086,
+      "loss": 0.0,
+      "step": 91
+    },
+    {
+      "epoch": 0.01301595161461465,
+      "grad_norm": 20.609447479248047,
+      "learning_rate": 0.0001961165421788852,
+      "loss": 0.5171,
+      "step": 92
+    },
+    {
+      "epoch": 0.013157429349556113,
+      "grad_norm": 0.0008163169259205461,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 0.0,
+      "step": 93
+    },
+    {
+      "epoch": 0.013298907084497578,
+      "grad_norm": 0.005171321332454681,
+      "learning_rate": 0.0001959412471854043,
+      "loss": 0.0001,
+      "step": 94
+    },
+    {
+      "epoch": 0.01344038481943904,
+      "grad_norm": 0.0035722258035093546,
+      "learning_rate": 0.0001958521789017376,
+      "loss": 0.0,
+      "step": 95
+    },
+    {
+      "epoch": 0.013581862554380505,
+      "grad_norm": 0.004428490065038204,
+      "learning_rate": 0.00019576216459576222,
+      "loss": 0.0,
+      "step": 96
+    },
+    {
+      "epoch": 0.013723340289321968,
+      "grad_norm": 0.0027291348669677973,
+      "learning_rate": 0.00019567120515588308,
+      "loss": 0.0001,
+      "step": 97
+    },
+    {
+      "epoch": 0.01386481802426343,
+      "grad_norm": 0.0007217226084321737,
+      "learning_rate": 0.00019557930147983302,
+      "loss": 0.0,
+      "step": 98
+    },
+    {
+      "epoch": 0.014006295759204895,
+      "grad_norm": 0.0516432449221611,
+      "learning_rate": 0.00019548645447466431,
+      "loss": 0.0004,
+      "step": 99
+    },
+    {
+      "epoch": 0.014147773494146358,
+      "grad_norm": 0.002289633499458432,
+      "learning_rate": 0.00019539266505673938,
+      "loss": 0.0001,
+      "step": 100
+    },
+    {
+      "epoch": 0.014289251229087823,
+      "grad_norm": 0.022498467937111855,
+      "learning_rate": 0.00019529793415172192,
+      "loss": 0.0002,
+      "step": 101
+    },
+    {
+      "epoch": 0.014430728964029286,
+      "grad_norm": 0.0024731154553592205,
+      "learning_rate": 0.00019520226269456768,
+      "loss": 0.0,
+      "step": 102
+    },
+    {
+      "epoch": 0.01457220669897075,
+      "grad_norm": 0.053761184215545654,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 0.0003,
+      "step": 103
+    },
+    {
+      "epoch": 0.014713684433912213,
+      "grad_norm": 52.32304000854492,
+      "learning_rate": 0.00019500810191007718,
+      "loss": 0.5584,
+      "step": 104
+    },
+    {
+      "epoch": 0.014855162168853677,
+      "grad_norm": 0.0023259480949491262,
+      "learning_rate": 0.00019490961449902946,
+      "loss": 0.0,
+      "step": 105
+    },
+    {
+      "epoch": 0.01499663990379514,
+      "grad_norm": 0.00143586122430861,
+      "learning_rate": 0.0001948101903684032,
+      "loss": 0.0001,
+      "step": 106
+    },
+    {
+      "epoch": 0.015138117638736603,
+      "grad_norm": 0.0017758137546479702,
+      "learning_rate": 0.00019470983049947444,
+      "loss": 0.0001,
+      "step": 107
+    },
+    {
+      "epoch": 0.015279595373678068,
+      "grad_norm": 0.008223861455917358,
+      "learning_rate": 0.00019460853588275454,
+      "loss": 0.0001,
+      "step": 108
+    },
+    {
+      "epoch": 0.01542107310861953,
+      "grad_norm": 0.002562491688877344,
+      "learning_rate": 0.00019450630751798048,
+      "loss": 0.0001,
+      "step": 109
+    },
+    {
+      "epoch": 0.015562550843560995,
+      "grad_norm": 0.0011813660385087132,
+      "learning_rate": 0.000194403146414105,
+      "loss": 0.0,
+      "step": 110
+    },
+    {
+      "epoch": 0.01570402857850246,
+      "grad_norm": 0.001858884934335947,
+      "learning_rate": 0.00019429905358928646,
+      "loss": 0.0,
+      "step": 111
+    },
+    {
+      "epoch": 0.015845506313443922,
+      "grad_norm": 30.78669548034668,
+      "learning_rate": 0.00019419403007087907,
+      "loss": 0.0985,
+      "step": 112
+    },
+    {
+      "epoch": 0.015986984048385385,
+      "grad_norm": 0.007476686500012875,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 0.0002,
+      "step": 113
+    },
+    {
+      "epoch": 0.016128461783326848,
+      "grad_norm": 0.0010357332648709416,
+      "learning_rate": 0.00019398119510863197,
+      "loss": 0.0,
+      "step": 114
+    },
+    {
+      "epoch": 0.01626993951826831,
+      "grad_norm": 0.001794468960724771,
+      "learning_rate": 0.00019387338576538744,
+      "loss": 0.0,
+      "step": 115
+    },
+    {
+      "epoch": 0.016411417253209777,
+      "grad_norm": 14.643719673156738,
+      "learning_rate": 0.00019376464992972356,
+      "loss": 0.4845,
+      "step": 116
+    },
+    {
+      "epoch": 0.01655289498815124,
+      "grad_norm": 10.48317813873291,
+      "learning_rate": 0.00019365498867481923,
+      "loss": 0.5411,
+      "step": 117
+    },
+    {
+      "epoch": 0.016694372723092703,
+      "grad_norm": 0.001224672538228333,
+      "learning_rate": 0.00019354440308298675,
+      "loss": 0.0,
+      "step": 118
+    },
+    {
+      "epoch": 0.016835850458034166,
+      "grad_norm": 0.0019753577653318644,
+      "learning_rate": 0.00019343289424566122,
+      "loss": 0.0,
+      "step": 119
+    },
+    {
+      "epoch": 0.016977328192975632,
+      "grad_norm": 0.006218386348336935,
+      "learning_rate": 0.00019332046326338986,
+      "loss": 0.0001,
+      "step": 120
+    },
+    {
+      "epoch": 0.017118805927917095,
+      "grad_norm": 0.004642636049538851,
+      "learning_rate": 0.0001932071112458211,
+      "loss": 0.0001,
+      "step": 121
+    },
+    {
+      "epoch": 0.017260283662858558,
+      "grad_norm": 0.004355429671704769,
+      "learning_rate": 0.00019309283931169356,
+      "loss": 0.0001,
+      "step": 122
+    },
+    {
+      "epoch": 0.01740176139780002,
+      "grad_norm": 0.006458196323364973,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 0.0002,
+      "step": 123
+    },
+    {
+      "epoch": 0.017543239132741483,
+      "grad_norm": 0.014439636841416359,
+      "learning_rate": 0.00019286154021410173,
+      "loss": 0.0003,
+      "step": 124
+    },
+    {
+      "epoch": 0.01768471686768295,
+      "grad_norm": 0.008952593430876732,
+      "learning_rate": 0.00019274451533346615,
+      "loss": 0.0002,
+      "step": 125
+    },
+    {
+      "epoch": 0.017826194602624412,
+      "grad_norm": 0.00849517434835434,
+      "learning_rate": 0.00019262657510190666,
+      "loss": 0.0003,
+      "step": 126
+    },
+    {
+      "epoch": 0.017967672337565875,
+      "grad_norm": 0.21418528258800507,
+      "learning_rate": 0.0001925077206834458,
+      "loss": 0.0017,
+      "step": 127
+    },
+    {
+      "epoch": 0.018109150072507338,
+      "grad_norm": 0.0048170918598771095,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.0001,
+      "step": 128
+    },
+    {
+      "epoch": 0.018250627807448804,
+      "grad_norm": 0.016278916969895363,
+      "learning_rate": 0.0001922672739870115,
+      "loss": 0.0003,
+      "step": 129
+    },
+    {
+      "epoch": 0.018392105542390267,
+      "grad_norm": 0.001837101997807622,
+      "learning_rate": 0.00019214568408214985,
+      "loss": 0.0001,
+      "step": 130
+    },
+    {
+      "epoch": 0.01853358327733173,
+      "grad_norm": 29.20079231262207,
+      "learning_rate": 0.00019202318473658705,
+      "loss": 7.677,
+      "step": 131
+    },
+    {
+      "epoch": 0.018675061012273193,
+      "grad_norm": 0.25970369577407837,
+      "learning_rate": 0.00019189977715934213,
+      "loss": 0.0016,
+      "step": 132
+    },
+    {
+      "epoch": 0.018816538747214655,
+      "grad_norm": 0.04528359696269035,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 0.0004,
+      "step": 133
+    },
+    {
+      "epoch": 0.01895801648215612,
+      "grad_norm": 0.005976614076644182,
+      "learning_rate": 0.0001916502421906898,
+      "loss": 0.0001,
+      "step": 134
+    },
+    {
+      "epoch": 0.019099494217097585,
+      "grad_norm": 0.8626922965049744,
+      "learning_rate": 0.00019152411726209176,
+      "loss": 0.0374,
+      "step": 135
+    },
+    {
+      "epoch": 0.019240971952039047,
+      "grad_norm": 0.3195020854473114,
+      "learning_rate": 0.00019139708902740613,
+      "loss": 0.0096,
+      "step": 136
+    },
+    {
+      "epoch": 0.01938244968698051,
+      "grad_norm": 0.020532745867967606,
+      "learning_rate": 0.0001912691587403503,
+      "loss": 0.0009,
+      "step": 137
+    },
+    {
+      "epoch": 0.019523927421921976,
+      "grad_norm": 0.09249910712242126,
+      "learning_rate": 0.00019114032766354453,
+      "loss": 0.0015,
+      "step": 138
+    },
+    {
+      "epoch": 0.01966540515686344,
+      "grad_norm": 0.018980201333761215,
+      "learning_rate": 0.00019101059706849957,
+      "loss": 0.0009,
+      "step": 139
+    },
+    {
+      "epoch": 0.019806882891804902,
+      "grad_norm": 0.028726842254400253,
+      "learning_rate": 0.00019087996823560402,
+      "loss": 0.001,
+      "step": 140
+    },
+    {
+      "epoch": 0.019948360626746365,
+      "grad_norm": 0.007452077232301235,
+      "learning_rate": 0.0001907484424541117,
+      "loss": 0.0003,
+      "step": 141
+    },
+    {
+      "epoch": 0.020089838361687828,
+      "grad_norm": 0.021318119019269943,
+      "learning_rate": 0.00019061602102212898,
+      "loss": 0.0004,
+      "step": 142
+    },
+    {
+      "epoch": 0.020231316096629294,
+      "grad_norm": 6.968769073486328,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 0.1187,
+      "step": 143
+    },
+    {
+      "epoch": 0.020372793831570757,
+      "grad_norm": 0.025080297142267227,
+      "learning_rate": 0.0001903484964433035,
+      "loss": 0.0005,
+      "step": 144
+    },
+    {
+      "epoch": 0.02051427156651222,
+      "grad_norm": 0.02062026597559452,
+      "learning_rate": 0.00019021339593682028,
+      "loss": 0.0008,
+      "step": 145
+    },
+    {
+      "epoch": 0.020655749301453682,
+      "grad_norm": 0.14519056677818298,
+      "learning_rate": 0.00019007740506053983,
+      "loss": 0.0027,
+      "step": 146
+    },
+    {
+      "epoch": 0.02079722703639515,
+      "grad_norm": 0.02562364749610424,
+      "learning_rate": 0.0001899405251566371,
+      "loss": 0.001,
+      "step": 147
+    },
+    {
+      "epoch": 0.02093870477133661,
+      "grad_norm": 0.018246619030833244,
+      "learning_rate": 0.00018980275757606157,
+      "loss": 0.0007,
+      "step": 148
+    },
+    {
+      "epoch": 0.021080182506278074,
+      "grad_norm": 0.005684189964085817,
+      "learning_rate": 0.00018966410367852362,
+      "loss": 0.0003,
+      "step": 149
+    },
+    {
+      "epoch": 0.021221660241219537,
+      "grad_norm": 0.42502322793006897,
+      "learning_rate": 0.00018952456483248119,
+      "loss": 0.0112,
+      "step": 150
+    },
+    {
+      "epoch": 0.021363137976161,
+      "grad_norm": 2.7156591415405273,
+      "learning_rate": 0.0001893841424151264,
+      "loss": 0.22,
+      "step": 151
+    },
+    {
+      "epoch": 0.021504615711102466,
+      "grad_norm": 0.004702973645180464,
+      "learning_rate": 0.0001892428378123718,
+      "loss": 0.0003,
+      "step": 152
+    },
+    {
+      "epoch": 0.02164609344604393,
+      "grad_norm": 0.003851711517199874,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 0.0002,
+      "step": 153
+    },
+    {
+      "epoch": 0.021787571180985392,
+      "grad_norm": 0.004598681814968586,
+      "learning_rate": 0.00018895758763783383,
+      "loss": 0.0002,
+      "step": 154
+    },
+    {
+      "epoch": 0.021929048915926855,
+      "grad_norm": 3.063443183898926,
+      "learning_rate": 0.00018881364488135448,
+      "loss": 0.0125,
+      "step": 155
+    },
+    {
+      "epoch": 0.02207052665086832,
+      "grad_norm": 0.005910379812121391,
+      "learning_rate": 0.00018866882557005567,
+      "loss": 0.0002,
+      "step": 156
+    },
+    {
+      "epoch": 0.022212004385809784,
+      "grad_norm": 0.005766018759459257,
+      "learning_rate": 0.00018852313113324552,
+      "loss": 0.0003,
+      "step": 157
+    },
+    {
+      "epoch": 0.022353482120751247,
+      "grad_norm": 0.004247742705047131,
+      "learning_rate": 0.00018837656300886937,
+      "loss": 0.0002,
+      "step": 158
+    },
+    {
+      "epoch": 0.02249495985569271,
+      "grad_norm": 0.007876652292907238,
+      "learning_rate": 0.00018822912264349534,
+      "loss": 0.0004,
+      "step": 159
+    },
+    {
+      "epoch": 0.022636437590634172,
+      "grad_norm": 0.007455014158040285,
+      "learning_rate": 0.00018808081149230036,
+      "loss": 0.0004,
+      "step": 160
+    },
+    {
+      "epoch": 0.02277791532557564,
+      "grad_norm": 0.03861529007554054,
+      "learning_rate": 0.00018793163101905563,
+      "loss": 0.0006,
+      "step": 161
+    },
+    {
+      "epoch": 0.0229193930605171,
+      "grad_norm": 0.8474092483520508,
+      "learning_rate": 0.00018778158269611218,
+      "loss": 0.0075,
+      "step": 162
+    },
+    {
+      "epoch": 0.023060870795458564,
+      "grad_norm": 0.050449665635824203,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 0.001,
+      "step": 163
+    },
+    {
+      "epoch": 0.023202348530400027,
+      "grad_norm": 0.007101717870682478,
+      "learning_rate": 0.0001874788884333453,
+      "loss": 0.0002,
+      "step": 164
+    },
+    {
+      "epoch": 0.023343826265341493,
+      "grad_norm": 5.2260355949401855,
+      "learning_rate": 0.00018732624548099204,
+      "loss": 0.0506,
+      "step": 165
+    },
+    {
+      "epoch": 0.023485304000282956,
+      "grad_norm": 0.03760479390621185,
+      "learning_rate": 0.0001871727406538509,
+      "loss": 0.0008,
+      "step": 166
+    },
+    {
+      "epoch": 0.02362678173522442,
+      "grad_norm": 0.02344103716313839,
+      "learning_rate": 0.0001870183754669526,
+      "loss": 0.0005,
+      "step": 167
+    },
+    {
+      "epoch": 0.023768259470165882,
+      "grad_norm": 0.007347187492996454,
+      "learning_rate": 0.00018686315144381913,
+      "loss": 0.0003,
+      "step": 168
+    },
+    {
+      "epoch": 0.023909737205107345,
+      "grad_norm": 4.132694244384766,
+      "learning_rate": 0.000186707070116449,
+      "loss": 0.0651,
+      "step": 169
+    },
+    {
+      "epoch": 0.02405121494004881,
+      "grad_norm": 8.260241508483887,
+      "learning_rate": 0.0001865501330253019,
+      "loss": 0.1644,
+      "step": 170
+    },
+    {
+      "epoch": 0.024192692674990274,
+      "grad_norm": 0.0025873123668134212,
+      "learning_rate": 0.00018639234171928353,
+      "loss": 0.0002,
+      "step": 171
+    },
+    {
+      "epoch": 0.024334170409931737,
+      "grad_norm": 0.022175127640366554,
+      "learning_rate": 0.0001862336977557304,
+      "loss": 0.0005,
+      "step": 172
+    },
+    {
+      "epoch": 0.0244756481448732,
+      "grad_norm": 0.0020805918611586094,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 0.0001,
+      "step": 173
+    },
+    {
+      "epoch": 0.024617125879814666,
+      "grad_norm": 0.00254829996265471,
+      "learning_rate": 0.00018591385812742725,
+      "loss": 0.0001,
+      "step": 174
+    },
+    {
+      "epoch": 0.02475860361475613,
+      "grad_norm": 0.003948370926082134,
+      "learning_rate": 0.00018575266561936523,
+      "loss": 0.0002,
+      "step": 175
+    },
+    {
+      "epoch": 0.02490008134969759,
+      "grad_norm": 62.64409637451172,
+      "learning_rate": 0.00018559062676711332,
+      "loss": 0.2596,
+      "step": 176
+    },
+    {
+      "epoch": 0.025041559084639054,
+      "grad_norm": 0.003057444002479315,
+      "learning_rate": 0.0001854277431699295,
+      "loss": 0.0002,
+      "step": 177
+    },
+    {
+      "epoch": 0.025183036819580517,
+      "grad_norm": 0.002523846924304962,
+      "learning_rate": 0.00018526401643540922,
+      "loss": 0.0001,
+      "step": 178
+    },
+    {
+      "epoch": 0.025324514554521983,
+      "grad_norm": 0.022342350333929062,
+      "learning_rate": 0.00018509944817946922,
+      "loss": 0.0006,
+      "step": 179
+    },
+    {
+      "epoch": 0.025465992289463446,
+      "grad_norm": 0.7193881273269653,
+      "learning_rate": 0.00018493404002633166,
+      "loss": 0.0397,
+      "step": 180
+    },
+    {
+      "epoch": 0.02560747002440491,
+      "grad_norm": 0.0033624654170125723,
+      "learning_rate": 0.00018476779360850832,
+      "loss": 0.0001,
+      "step": 181
+    },
+    {
+      "epoch": 0.02574894775934637,
+      "grad_norm": 4.625955581665039,
+      "learning_rate": 0.00018460071056678422,
+      "loss": 1.1598,
+      "step": 182
+    },
+    {
+      "epoch": 0.025890425494287838,
+      "grad_norm": 0.002986474661156535,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 0.0002,
+      "step": 183
+    },
+    {
+      "epoch": 0.0260319032292293,
+      "grad_norm": 0.0038495738990604877,
+      "learning_rate": 0.00018426404121604323,
+      "loss": 0.0003,
+      "step": 184
+    },
+    {
+      "epoch": 0.026173380964170764,
+      "grad_norm": 0.003486146917566657,
+      "learning_rate": 0.00018409445822981693,
+      "loss": 0.0003,
+      "step": 185
+    },
+    {
+      "epoch": 0.026314858699112226,
+      "grad_norm": 0.014306096360087395,
+      "learning_rate": 0.00018392404526523817,
+      "loss": 0.0011,
+      "step": 186
+    },
+    {
+      "epoch": 0.02645633643405369,
+      "grad_norm": 0.01424553245306015,
+      "learning_rate": 0.0001837528040042142,
+      "loss": 0.001,
+      "step": 187
+    },
+    {
+      "epoch": 0.026597814168995156,
+      "grad_norm": 0.011042344383895397,
+      "learning_rate": 0.00018358073613682706,
+      "loss": 0.0008,
+      "step": 188
+    },
+    {
+      "epoch": 0.02673929190393662,
+      "grad_norm": 0.034611593931913376,
+      "learning_rate": 0.00018340784336131713,
+      "loss": 0.0017,
+      "step": 189
+    },
+    {
+      "epoch": 0.02688076963887808,
+      "grad_norm": 0.00807161070406437,
+      "learning_rate": 0.00018323412738406635,
+      "loss": 0.0006,
+      "step": 190
+    },
+    {
+      "epoch": 0.027022247373819544,
+      "grad_norm": 0.015202701091766357,
+      "learning_rate": 0.00018305958991958127,
+      "loss": 0.0009,
+      "step": 191
+    },
+    {
+      "epoch": 0.02716372510876101,
+      "grad_norm": 0.11597127467393875,
+      "learning_rate": 0.0001828842326904762,
+      "loss": 0.0045,
+      "step": 192
+    },
+    {
+      "epoch": 0.027305202843702473,
+      "grad_norm": 0.13688474893569946,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 0.0021,
+      "step": 193
+    },
+    {
+      "epoch": 0.027446680578643936,
+      "grad_norm": 0.02395058050751686,
+      "learning_rate": 0.00018253106586929997,
+      "loss": 0.0015,
+      "step": 194
+    },
+    {
+      "epoch": 0.0275881583135854,
+      "grad_norm": 0.024011526256799698,
+      "learning_rate": 0.00018235325976284275,
+      "loss": 0.0017,
+      "step": 195
+    },
+    {
+      "epoch": 0.02772963604852686,
+      "grad_norm": 0.015601936727762222,
+      "learning_rate": 0.00018217464086295904,
+      "loss": 0.0008,
+      "step": 196
+    },
+    {
+      "epoch": 0.027871113783468328,
+      "grad_norm": 0.008236625231802464,
+      "learning_rate": 0.00018199521093254523,
+      "loss": 0.0007,
+      "step": 197
+    },
+    {
+      "epoch": 0.02801259151840979,
+      "grad_norm": 0.007339195813983679,
+      "learning_rate": 0.00018181497174250236,
+      "loss": 0.0006,
+      "step": 198
+    },
+    {
+      "epoch": 0.028154069253351253,
+      "grad_norm": 0.03408842161297798,
+      "learning_rate": 0.00018163392507171842,
+      "loss": 0.0013,
+      "step": 199
+    },
+    {
+      "epoch": 0.028295546988292716,
+      "grad_norm": 0.010708988644182682,
+      "learning_rate": 0.00018145207270705096,
+      "loss": 0.0008,
+      "step": 200
+    },
+    {
+      "epoch": 0.028437024723234183,
+      "grad_norm": 0.04180416092276573,
+      "learning_rate": 0.0001812694164433094,
+      "loss": 0.0008,
+      "step": 201
+    },
+    {
+      "epoch": 0.028578502458175645,
+      "grad_norm": 0.015216993167996407,
+      "learning_rate": 0.00018108595808323736,
+      "loss": 0.0004,
+      "step": 202
+    },
+    {
+      "epoch": 0.028719980193117108,
+      "grad_norm": 0.023775571957230568,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.0014,
+      "step": 203
+    },
+    {
+      "epoch": 0.02886145792805857,
+      "grad_norm": 0.007402106188237667,
+      "learning_rate": 0.00018071664232464002,
+      "loss": 0.0005,
+      "step": 204
+    },
+    {
+      "epoch": 0.029002935663000034,
+      "grad_norm": 0.051116202026605606,
+      "learning_rate": 0.0001805307885711122,
+      "loss": 0.0011,
+      "step": 205
+    },
+    {
+      "epoch": 0.0291444133979415,
+      "grad_norm": 0.006247265264391899,
+      "learning_rate": 0.00018034414001121278,
+      "loss": 0.0004,
+      "step": 206
+    },
+    {
+      "epoch": 0.029285891132882963,
+      "grad_norm": 0.006299880333244801,
+      "learning_rate": 0.00018015669848708767,
+      "loss": 0.0004,
+      "step": 207
+    },
+    {
+      "epoch": 0.029427368867824426,
+      "grad_norm": 0.11377950757741928,
+      "learning_rate": 0.00017996846584870908,
+      "loss": 0.0009,
+      "step": 208
+    },
+    {
+      "epoch": 0.02956884660276589,
+      "grad_norm": 0.0020067133009433746,
+      "learning_rate": 0.0001797794439538571,
+      "loss": 0.0002,
+      "step": 209
+    },
+    {
+      "epoch": 0.029710324337707355,
+      "grad_norm": 0.40424010157585144,
+      "learning_rate": 0.0001795896346681016,
+      "loss": 0.0108,
+      "step": 210
+    },
+    {
+      "epoch": 0.029851802072648818,
+      "grad_norm": 0.0018006038153544068,
+      "learning_rate": 0.00017939903986478355,
+      "loss": 0.0001,
+      "step": 211
+    },
+    {
+      "epoch": 0.02999327980759028,
+      "grad_norm": 0.00778124388307333,
+      "learning_rate": 0.00017920766142499672,
+      "loss": 0.0004,
+      "step": 212
+    },
+    {
+      "epoch": 0.030134757542531743,
+      "grad_norm": 0.0031390658114105463,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 0.0001,
+      "step": 213
+    },
+    {
+      "epoch": 0.030276235277473206,
+      "grad_norm": 0.03861818090081215,
+      "learning_rate": 0.00017882256119904403,
+      "loss": 0.0015,
+      "step": 214
+    },
+    {
+      "epoch": 0.030417713012414672,
+      "grad_norm": 0.6883278489112854,
+      "learning_rate": 0.00017862884321366188,
+      "loss": 0.004,
+      "step": 215
+    },
+    {
+      "epoch": 0.030559190747356135,
+      "grad_norm": 0.0016816803254187107,
+      "learning_rate": 0.000178434349193341,
+      "loss": 0.0001,
+      "step": 216
+    },
+    {
+      "epoch": 0.030700668482297598,
+      "grad_norm": 0.022568391636013985,
+      "learning_rate": 0.0001782390810576588,
+      "loss": 0.0007,
+      "step": 217
+    },
+    {
+      "epoch": 0.03084214621723906,
+      "grad_norm": 0.001979145687073469,
+      "learning_rate": 0.000178043040733833,
+      "loss": 0.0001,
+      "step": 218
+    },
+    {
+      "epoch": 0.030983623952180527,
+      "grad_norm": 0.006649916060268879,
+      "learning_rate": 0.00017784623015670238,
+      "loss": 0.0002,
+      "step": 219
+    },
+    {
+      "epoch": 0.03112510168712199,
+      "grad_norm": 0.0012332225451245904,
+      "learning_rate": 0.00017764865126870786,
+      "loss": 0.0001,
+      "step": 220
+    },
+    {
+      "epoch": 0.03126657942206345,
+      "grad_norm": 0.0019407098880037665,
+      "learning_rate": 0.00017745030601987337,
+      "loss": 0.0001,
+      "step": 221
+    },
+    {
+      "epoch": 0.03140805715700492,
+      "grad_norm": 0.0025570588186383247,
+      "learning_rate": 0.00017725119636778644,
+      "loss": 0.0001,
+      "step": 222
+    },
+    {
+      "epoch": 0.03154953489194638,
+      "grad_norm": 10.877803802490234,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 0.2045,
+      "step": 223
+    },
+    {
+      "epoch": 0.031691012626887845,
+      "grad_norm": 0.001555653172545135,
+      "learning_rate": 0.00017685069172190766,
+      "loss": 0.0001,
+      "step": 224
+    },
+    {
+      "epoch": 0.03183249036182931,
+      "grad_norm": 0.0014934063656255603,
+      "learning_rate": 0.00017664930068093498,
+      "loss": 0.0001,
+      "step": 225
+    },
+    {
+      "epoch": 0.03197396809677077,
+      "grad_norm": 0.0034368853084743023,
+      "learning_rate": 0.00017644715314230918,
+      "loss": 0.0001,
+      "step": 226
+    },
+    {
+      "epoch": 0.03211544583171223,
+      "grad_norm": 0.0021537458524107933,
+      "learning_rate": 0.0001762442511011448,
+      "loss": 0.0001,
+      "step": 227
+    },
+    {
+      "epoch": 0.032256923566653696,
+      "grad_norm": 0.002286655129864812,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 0.0001,
+      "step": 228
+    },
+    {
+      "epoch": 0.03239840130159516,
+      "grad_norm": 0.005433680955320597,
+      "learning_rate": 0.0001758361915288722,
+      "loss": 0.0003,
+      "step": 229
+    },
+    {
+      "epoch": 0.03253987903653662,
+      "grad_norm": 0.0013771787052974105,
+      "learning_rate": 0.0001756310380251472,
+      "loss": 0.0001,
+      "step": 230
+    },
+    {
+      "epoch": 0.03268135677147809,
+      "grad_norm": 0.0018148127710446715,
+      "learning_rate": 0.00017542513807361037,
+      "loss": 0.0001,
+      "step": 231
+    },
+    {
+      "epoch": 0.032822834506419554,
+      "grad_norm": 0.00181623047683388,
+      "learning_rate": 0.00017521849370641114,
+      "loss": 0.0001,
+      "step": 232
+    },
+    {
+      "epoch": 0.03296431224136102,
+      "grad_norm": 0.01041108276695013,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 0.0003,
+      "step": 233
+    },
+    {
+      "epoch": 0.03310578997630248,
+      "grad_norm": 0.004963583778589964,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 0.0002,
+      "step": 234
+    },
+    {
+      "epoch": 0.03324726771124394,
+      "grad_norm": 0.007771691773086786,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 0.0003,
+      "step": 235
+    },
+    {
+      "epoch": 0.033388745446185406,
+      "grad_norm": 0.0008118674159049988,
+      "learning_rate": 0.00017438451298070252,
+      "loss": 0.0001,
+      "step": 236
+    },
+    {
+      "epoch": 0.03353022318112687,
+      "grad_norm": 0.004826071672141552,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 0.0001,
+      "step": 237
+    },
+    {
+      "epoch": 0.03367170091606833,
+      "grad_norm": 0.0054870243184268475,
+      "learning_rate": 0.000173963109497861,
+      "loss": 0.0001,
+      "step": 238
+    },
+    {
+      "epoch": 0.033813178651009794,
+      "grad_norm": 0.31345561146736145,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 0.0022,
+      "step": 239
+    },
+    {
+      "epoch": 0.033954656385951264,
+      "grad_norm": 0.0009997932938858867,
+      "learning_rate": 0.0001735387860781016,
+      "loss": 0.0,
+      "step": 240
+    },
+    {
+      "epoch": 0.03409613412089273,
+      "grad_norm": 0.0014913649065420032,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 0.0001,
+      "step": 241
+    },
+    {
+      "epoch": 0.03423761185583419,
+      "grad_norm": 0.001048844656907022,
+      "learning_rate": 0.00017311155947298643,
+      "loss": 0.0001,
+      "step": 242
+    },
+    {
+      "epoch": 0.03437908959077565,
+      "grad_norm": 0.005004017613828182,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.0001,
+      "step": 243
+    },
+    {
+      "epoch": 0.034520567325717115,
+      "grad_norm": 0.0016325002070516348,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 0.0001,
+      "step": 244
+    },
+    {
+      "epoch": 0.03466204506065858,
+      "grad_norm": 2.20041561126709,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 0.1624,
+      "step": 245
+    },
+    {
+      "epoch": 0.03480352279560004,
+      "grad_norm": 0.0004976244526915252,
+      "learning_rate": 0.00017224846428533499,
+      "loss": 0.0,
+      "step": 246
+    },
+    {
+      "epoch": 0.0349450005305415,
+      "grad_norm": 0.0007265186868607998,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 0.0001,
+      "step": 247
+    },
+    {
+      "epoch": 0.035086478265482966,
+      "grad_norm": 0.03309307619929314,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 0.0,
+      "step": 248
+    },
+    {
+      "epoch": 0.035227956000424436,
+      "grad_norm": 0.0008523318101651967,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 0.0001,
+      "step": 249
+    },
+    {
+      "epoch": 0.0353694337353659,
+      "grad_norm": 0.0026479423977434635,
+      "learning_rate": 0.00017137396022764214,
+      "loss": 0.0002,
+      "step": 250
+    },
+    {
+      "epoch": 0.03551091147030736,
+      "grad_norm": 0.0009089628583751619,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 0.0001,
+      "step": 251
+    },
+    {
+      "epoch": 0.035652389205248824,
+      "grad_norm": 0.001088530640117824,
+      "learning_rate": 0.0001709324729572274,
+      "loss": 0.0001,
+      "step": 252
+    },
+    {
+      "epoch": 0.03579386694019029,
+      "grad_norm": 0.0062820520251989365,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.0001,
+      "step": 253
+    },
+    {
+      "epoch": 0.03593534467513175,
+      "grad_norm": 0.09086458384990692,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 0.0018,
+      "step": 254
+    },
+    {
+      "epoch": 0.03607682241007321,
+      "grad_norm": 26.014320373535156,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 3.5957,
+      "step": 255
+    },
+    {
+      "epoch": 0.036218300145014676,
+      "grad_norm": 0.0020564054138958454,
+      "learning_rate": 0.00017004111507838064,
+      "loss": 0.0002,
+      "step": 256
+    },
+    {
+      "epoch": 0.03635977787995614,
+      "grad_norm": 0.0016233263304457068,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 0.0001,
+      "step": 257
+    },
+    {
+      "epoch": 0.03650125561489761,
+      "grad_norm": 0.0012816754169762135,
+      "learning_rate": 0.00016959127965923142,
+      "loss": 0.0001,
+      "step": 258
+    },
+    {
+      "epoch": 0.03664273334983907,
+      "grad_norm": 0.0034658850636333227,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 0.0002,
+      "step": 259
+    },
+    {
+      "epoch": 0.036784211084780534,
+      "grad_norm": 0.02739444188773632,
+      "learning_rate": 0.00016913869689552064,
+      "loss": 0.0003,
+      "step": 260
+    },
+    {
+      "epoch": 0.036925688819722,
+      "grad_norm": 0.011004602536559105,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 0.0005,
+      "step": 261
+    },
+    {
+      "epoch": 0.03706716655466346,
+      "grad_norm": 0.0057389517314732075,
+      "learning_rate": 0.00016868338465444085,
+      "loss": 0.0002,
+      "step": 262
+    },
+    {
+      "epoch": 0.03720864428960492,
+      "grad_norm": 3.8611650466918945,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 1.4247,
+      "step": 263
+    },
+    {
+      "epoch": 0.037350122024546385,
+      "grad_norm": 0.01210610382258892,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 0.0004,
+      "step": 264
+    },
+    {
+      "epoch": 0.03749159975948785,
+      "grad_norm": 0.013306356966495514,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 0.0008,
+      "step": 265
+    },
+    {
+      "epoch": 0.03763307749442931,
+      "grad_norm": 0.27052658796310425,
+      "learning_rate": 0.00016776464374701025,
+      "loss": 0.0099,
+      "step": 266
+    },
+    {
+      "epoch": 0.03777455522937078,
+      "grad_norm": 0.008403144776821136,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 0.0005,
+      "step": 267
+    },
+    {
+      "epoch": 0.03791603296431224,
+      "grad_norm": 0.13167297840118408,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 0.0027,
+      "step": 268
+    },
+    {
+      "epoch": 0.038057510699253706,
+      "grad_norm": 0.026871370151638985,
+      "learning_rate": 0.000167068557653672,
+      "loss": 0.001,
+      "step": 269
+    },
+    {
+      "epoch": 0.03819898843419517,
+      "grad_norm": 0.03469863906502724,
+      "learning_rate": 0.0001668352020167793,
+      "loss": 0.0015,
+      "step": 270
+    },
+    {
+      "epoch": 0.03834046616913663,
+      "grad_norm": 11.140663146972656,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 0.0507,
+      "step": 271
+    },
+    {
+      "epoch": 0.038481943904078095,
+      "grad_norm": 0.062388550490140915,
+      "learning_rate": 0.00016636651414324587,
+      "loss": 0.0023,
+      "step": 272
+    },
+    {
+      "epoch": 0.03862342163901956,
+      "grad_norm": 0.009884512983262539,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 0.0006,
+      "step": 273
+    },
+    {
+      "epoch": 0.03876489937396102,
+      "grad_norm": 0.011747513897716999,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 0.0007,
+      "step": 274
+    },
+    {
+      "epoch": 0.03890637710890248,
+      "grad_norm": 0.016286730766296387,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 0.0007,
+      "step": 275
+    },
+    {
+      "epoch": 0.03904785484384395,
+      "grad_norm": 0.008195940405130386,
+      "learning_rate": 0.00016542129689358612,
+      "loss": 0.0005,
+      "step": 276
+    },
+    {
+      "epoch": 0.039189332578785416,
+      "grad_norm": 0.14791356027126312,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 0.0012,
+      "step": 277
+    },
+    {
+      "epoch": 0.03933081031372688,
+      "grad_norm": 0.04902546480298042,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 0.0014,
+      "step": 278
+    },
+    {
+      "epoch": 0.03947228804866834,
+      "grad_norm": 0.08834591507911682,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 0.0009,
+      "step": 279
+    },
+    {
+      "epoch": 0.039613765783609804,
+      "grad_norm": 0.02635873667895794,
+      "learning_rate": 0.00016446574886275913,
+      "loss": 0.0006,
+      "step": 280
+    },
+    {
+      "epoch": 0.03975524351855127,
+      "grad_norm": 0.005103811621665955,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 0.0003,
+      "step": 281
+    },
+    {
+      "epoch": 0.03989672125349273,
+      "grad_norm": 0.004541166592389345,
+      "learning_rate": 0.00016398414789511786,
+      "loss": 0.0004,
+      "step": 282
+    },
+    {
+      "epoch": 0.04003819898843419,
+      "grad_norm": 0.0033859983086586,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.0002,
+      "step": 283
+    },
+    {
+      "epoch": 0.040179676723375655,
+      "grad_norm": 0.003963902592658997,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 0.0003,
+      "step": 284
+    },
+    {
+      "epoch": 0.040321154458317125,
+      "grad_norm": 0.005432737059891224,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 0.0003,
+      "step": 285
+    },
+    {
+      "epoch": 0.04046263219325859,
+      "grad_norm": 0.014246117323637009,
+      "learning_rate": 0.00016301338711853693,
+      "loss": 0.0003,
+      "step": 286
+    },
+    {
+      "epoch": 0.04060410992820005,
+      "grad_norm": 0.060390155762434006,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 0.0006,
+      "step": 287
+    },
+    {
+      "epoch": 0.040745587663141514,
+      "grad_norm": 0.003241261001676321,
+      "learning_rate": 0.00016252426563357055,
+      "loss": 0.0002,
+      "step": 288
+    },
+    {
+      "epoch": 0.040887065398082977,
+      "grad_norm": 0.0031794258393347263,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 0.0003,
+      "step": 289
+    },
+    {
+      "epoch": 0.04102854313302444,
+      "grad_norm": 0.008073396049439907,
+      "learning_rate": 0.00016203267579765563,
+      "loss": 0.0002,
+      "step": 290
+    },
+    {
+      "epoch": 0.0411700208679659,
+      "grad_norm": 0.002360474318265915,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 0.0002,
+      "step": 291
+    },
+    {
+      "epoch": 0.041311498602907365,
+      "grad_norm": 0.004607622046023607,
+      "learning_rate": 0.00016153863701791717,
+      "loss": 0.0003,
+      "step": 292
+    },
+    {
+      "epoch": 0.04145297633784883,
+      "grad_norm": 0.002519590314477682,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 0.0002,
+      "step": 293
+    },
+    {
+      "epoch": 0.0415944540727903,
+      "grad_norm": 0.001880524680018425,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 0.0001,
+      "step": 294
+    },
+    {
+      "epoch": 0.04173593180773176,
+      "grad_norm": 0.0013052290305495262,
+      "learning_rate": 0.00016079302976946055,
+      "loss": 0.0001,
+      "step": 295
+    },
+    {
+      "epoch": 0.04187740954267322,
+      "grad_norm": 0.0029020386282354593,
+      "learning_rate": 0.00016054329073810015,
+      "loss": 0.0002,
+      "step": 296
+    },
+    {
+      "epoch": 0.042018887277614686,
+      "grad_norm": 0.0025680111721158028,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 0.0002,
+      "step": 297
+    },
+    {
+      "epoch": 0.04216036501255615,
+      "grad_norm": 0.0024828522000461817,
+      "learning_rate": 0.00016004202253258842,
+      "loss": 0.0001,
+      "step": 298
+    },
+    {
+      "epoch": 0.04230184274749761,
+      "grad_norm": 0.0020577250979840755,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 0.0001,
+      "step": 299
+    },
+    {
+      "epoch": 0.042443320482439074,
+      "grad_norm": 0.0016125338152050972,
+      "learning_rate": 0.00015953838397083552,
+      "loss": 0.0001,
+      "step": 300
+    },
+    {
+      "epoch": 0.04258479821738054,
+      "grad_norm": 0.001345417695119977,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 0.0001,
+      "step": 301
+    },
+    {
+      "epoch": 0.042726275952322,
+      "grad_norm": 0.0021328881848603487,
+      "learning_rate": 0.00015903239493562948,
+      "loss": 0.0001,
+      "step": 302
+    },
+    {
+      "epoch": 0.04286775368726347,
+      "grad_norm": 0.0024942061863839626,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.0001,
+      "step": 303
+    },
+    {
+      "epoch": 0.04300923142220493,
+      "grad_norm": 0.004291863180696964,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 0.0002,
+      "step": 304
+    },
+    {
+      "epoch": 0.043150709157146395,
+      "grad_norm": 0.005543672014027834,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 0.0003,
+      "step": 305
+    },
+    {
+      "epoch": 0.04329218689208786,
+      "grad_norm": 0.002027965849265456,
+      "learning_rate": 0.00015801344543918495,
+      "loss": 0.0001,
+      "step": 306
+    },
+    {
+      "epoch": 0.04343366462702932,
+      "grad_norm": 0.002267736243084073,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 0.0001,
+      "step": 307
+    },
+    {
+      "epoch": 0.043575142361970784,
+      "grad_norm": 0.02030862867832184,
+      "learning_rate": 0.00015750052520432787,
+      "loss": 0.0004,
+      "step": 308
+    },
+    {
+      "epoch": 0.04371662009691225,
+      "grad_norm": 0.005449316464364529,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 0.0002,
+      "step": 309
+    },
+    {
+      "epoch": 0.04385809783185371,
+      "grad_norm": 0.0010752433445304632,
+      "learning_rate": 0.00015698533494719238,
+      "loss": 0.0001,
+      "step": 310
+    },
+    {
+      "epoch": 0.04399957556679517,
+      "grad_norm": 0.0009665637626312673,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 0.0001,
+      "step": 311
+    },
+    {
+      "epoch": 0.04414105330173664,
+      "grad_norm": 0.0021984372287988663,
+      "learning_rate": 0.00015646789500660773,
+      "loss": 0.0001,
+      "step": 312
+    },
+    {
+      "epoch": 0.044282531036678105,
+      "grad_norm": 0.0006756667862646282,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 0.0001,
+      "step": 313
+    },
+    {
+      "epoch": 0.04442400877161957,
+      "grad_norm": 0.0008639508741907775,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 0.0001,
+      "step": 314
+    },
+    {
+      "epoch": 0.04456548650656103,
+      "grad_norm": 0.0005970299243927002,
+      "learning_rate": 0.00015568756164881882,
+      "loss": 0.0,
+      "step": 315
+    },
+    {
+      "epoch": 0.04470696424150249,
+      "grad_norm": 0.0009843648876994848,
+      "learning_rate": 0.00015542634787366942,
+      "loss": 0.0001,
+      "step": 316
+    },
+    {
+      "epoch": 0.044848441976443956,
+      "grad_norm": 0.0009396934765391052,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 0.0001,
+      "step": 317
+    },
+    {
+      "epoch": 0.04498991971138542,
+      "grad_norm": 0.0009946058271452785,
+      "learning_rate": 0.0001549022817998132,
+      "loss": 0.0001,
+      "step": 318
+    },
+    {
+      "epoch": 0.04513139744632688,
+      "grad_norm": 0.0008691767579875886,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 0.0001,
+      "step": 319
+    },
+    {
+      "epoch": 0.045272875181268345,
+      "grad_norm": 0.001260230434127152,
+      "learning_rate": 0.00015437604827787927,
+      "loss": 0.0001,
+      "step": 320
+    },
+    {
+      "epoch": 0.045414352916209814,
+      "grad_norm": 0.0007110429578460753,
+      "learning_rate": 0.00015411212521268758,
+      "loss": 0.0001,
+      "step": 321
+    },
+    {
+      "epoch": 0.04555583065115128,
+      "grad_norm": 0.011615327559411526,
+      "learning_rate": 0.00015384766808266602,
+      "loss": 0.0003,
+      "step": 322
+    },
+    {
+      "epoch": 0.04569730838609274,
+      "grad_norm": 0.000552054145373404,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 0.0,
+      "step": 323
+    },
+    {
+      "epoch": 0.0458387861210342,
+      "grad_norm": 0.001125704264268279,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 0.0001,
+      "step": 324
+    },
+    {
+      "epoch": 0.045980263855975666,
+      "grad_norm": 0.0021504261530935764,
+      "learning_rate": 0.0001530511184306734,
+      "loss": 0.0001,
+      "step": 325
+    },
+    {
+      "epoch": 0.04612174159091713,
+      "grad_norm": 0.002731619169935584,
+      "learning_rate": 0.00015278455119450664,
+      "loss": 0.0001,
+      "step": 326
+    },
+    {
+      "epoch": 0.04626321932585859,
+      "grad_norm": 0.0007831475813873112,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 0.0001,
+      "step": 327
+    },
+    {
+      "epoch": 0.046404697060800054,
+      "grad_norm": 0.20330454409122467,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 0.0041,
+      "step": 328
+    },
+    {
+      "epoch": 0.04654617479574152,
+      "grad_norm": 0.0007988195866346359,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 0.0001,
+      "step": 329
+    },
+    {
+      "epoch": 0.04668765253068299,
+      "grad_norm": 17.48381805419922,
+      "learning_rate": 0.00015171309901381572,
+      "loss": 0.2524,
+      "step": 330
+    },
+    {
+      "epoch": 0.04682913026562445,
+      "grad_norm": 0.0008586746989749372,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 0.0001,
+      "step": 331
+    },
+    {
+      "epoch": 0.04697060800056591,
+      "grad_norm": 0.0005710547557100654,
+      "learning_rate": 0.00015117430001143452,
+      "loss": 0.0,
+      "step": 332
+    },
+    {
+      "epoch": 0.047112085735507375,
+      "grad_norm": 0.0006703097024001181,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 0.0001,
+      "step": 333
+    },
+    {
+      "epoch": 0.04725356347044884,
+      "grad_norm": 0.002217133529484272,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 0.0001,
+      "step": 334
+    },
+    {
+      "epoch": 0.0473950412053903,
+      "grad_norm": 3.155515193939209,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 1.6251,
+      "step": 335
+    },
+    {
+      "epoch": 0.047536518940331764,
+      "grad_norm": 0.004736212082207203,
+      "learning_rate": 0.000150090662536071,
+      "loss": 0.0003,
+      "step": 336
+    },
+    {
+      "epoch": 0.047677996675273226,
+      "grad_norm": 0.002677865093573928,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 0.0001,
+      "step": 337
+    },
+    {
+      "epoch": 0.04781947441021469,
+      "grad_norm": 0.0027205327060073614,
+      "learning_rate": 0.00014954586684324078,
+      "loss": 0.0002,
+      "step": 338
+    },
+    {
+      "epoch": 0.04796095214515616,
+      "grad_norm": 0.009435261599719524,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 0.0006,
+      "step": 339
+    },
+    {
+      "epoch": 0.04810242988009762,
+      "grad_norm": 0.0023822993971407413,
+      "learning_rate": 0.00014899911516442365,
+      "loss": 0.0002,
+      "step": 340
+    },
+    {
+      "epoch": 0.048243907615039085,
+      "grad_norm": 0.0032713001128286123,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 0.0003,
+      "step": 341
+    },
+    {
+      "epoch": 0.04838538534998055,
+      "grad_norm": 0.006985257379710674,
+      "learning_rate": 0.0001484504290844398,
+      "loss": 0.0005,
+      "step": 342
+    },
+    {
+      "epoch": 0.04852686308492201,
+      "grad_norm": 0.005185326095670462,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 0.0004,
+      "step": 343
+    },
+    {
+      "epoch": 0.04866834081986347,
+      "grad_norm": 0.010701478458940983,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 0.0007,
+      "step": 344
+    },
+    {
+      "epoch": 0.048809818554804936,
+      "grad_norm": 0.006417698226869106,
+      "learning_rate": 0.0001476238203667939,
+      "loss": 0.0005,
+      "step": 345
+    },
+    {
+      "epoch": 0.0489512962897464,
+      "grad_norm": 0.022306060418486595,
+      "learning_rate": 0.0001473473404412312,
+      "loss": 0.0016,
+      "step": 346
+    },
+    {
+      "epoch": 0.04909277402468786,
+      "grad_norm": 0.01026198361068964,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 0.0008,
+      "step": 347
+    },
+    {
+      "epoch": 0.04923425175962933,
+      "grad_norm": 0.009428659453988075,
+      "learning_rate": 0.00014679298142605734,
+      "loss": 0.0007,
+      "step": 348
+    },
+    {
+      "epoch": 0.049375729494570794,
+      "grad_norm": 0.018124036490917206,
+      "learning_rate": 0.00014651510780774583,
+      "loss": 0.0012,
+      "step": 349
+    },
+    {
+      "epoch": 0.04951720722951226,
+      "grad_norm": 0.015100178308784962,
+      "learning_rate": 0.00014623677510409918,
+      "loss": 0.0012,
+      "step": 350
+    },
+    {
+      "epoch": 0.04965868496445372,
+      "grad_norm": 0.011049584485590458,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 0.0009,
+      "step": 351
+    },
+    {
+      "epoch": 0.04980016269939518,
+      "grad_norm": 0.018350180238485336,
+      "learning_rate": 0.00014567874343342997,
+      "loss": 0.0012,
+      "step": 352
+    },
+    {
+      "epoch": 0.049941640434336645,
+      "grad_norm": 0.005243629217147827,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 0.0005,
+      "step": 353
+    },
+    {
+      "epoch": 0.05008311816927811,
+      "grad_norm": 0.009776602499186993,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 0.0008,
+      "step": 354
+    },
+    {
+      "epoch": 0.05022459590421957,
+      "grad_norm": 0.07710767537355423,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 0.0023,
+      "step": 355
+    },
+    {
+      "epoch": 0.050366073639161034,
+      "grad_norm": 0.010807436890900135,
+      "learning_rate": 0.00014455729223768966,
+      "loss": 0.0008,
+      "step": 356
+    },
+    {
+      "epoch": 0.050507551374102504,
+      "grad_norm": 0.010125311091542244,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 0.0007,
+      "step": 357
+    },
+    {
+      "epoch": 0.050649029109043966,
+      "grad_norm": 0.004309076350182295,
+      "learning_rate": 0.00014399391698559152,
+      "loss": 0.0004,
+      "step": 358
+    },
+    {
+      "epoch": 0.05079050684398543,
+      "grad_norm": 0.00469523249194026,
+      "learning_rate": 0.0001437115766650933,
+      "loss": 0.0003,
+      "step": 359
+    },
+    {
+      "epoch": 0.05093198457892689,
+      "grad_norm": 0.01330766174942255,
+      "learning_rate": 0.00014342880492898048,
+      "loss": 0.0006,
+      "step": 360
+    },
+    {
+      "epoch": 0.051073462313868355,
+      "grad_norm": 0.007151063531637192,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 0.0004,
+      "step": 361
+    },
+    {
+      "epoch": 0.05121494004880982,
+      "grad_norm": 0.002502172952517867,
+      "learning_rate": 0.00014286197837751286,
+      "loss": 0.0002,
+      "step": 362
+    },
+    {
+      "epoch": 0.05135641778375128,
+      "grad_norm": 0.0047018746845424175,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 0.0003,
+      "step": 363
+    },
+    {
+      "epoch": 0.05149789551869274,
+      "grad_norm": 0.002204613294452429,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 0.0002,
+      "step": 364
+    },
+    {
+      "epoch": 0.051639373253634206,
+      "grad_norm": 0.01051216758787632,
+      "learning_rate": 0.00014200857284118066,
+      "loss": 0.0005,
+      "step": 365
+    },
+    {
+      "epoch": 0.051780850988575676,
+      "grad_norm": 0.003874387824907899,
+      "learning_rate": 0.00014172327136617656,
+      "loss": 0.0003,
+      "step": 366
+    },
+    {
+      "epoch": 0.05192232872351714,
+      "grad_norm": 0.0045910426415503025,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 0.0003,
+      "step": 367
+    },
+    {
+      "epoch": 0.0520638064584586,
+      "grad_norm": 0.593174159526825,
+      "learning_rate": 0.00014115143586051088,
+      "loss": 0.0112,
+      "step": 368
+    },
+    {
+      "epoch": 0.052205284193400064,
+      "grad_norm": 0.010553068481385708,
+      "learning_rate": 0.00014086490747363493,
+      "loss": 0.0003,
+      "step": 369
+    },
+    {
+      "epoch": 0.05234676192834153,
+      "grad_norm": 0.10641700774431229,
+      "learning_rate": 0.00014057797576662,
+      "loss": 0.001,
+      "step": 370
+    },
+    {
+      "epoch": 0.05248823966328299,
+      "grad_norm": 0.0037969814147800207,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 0.0002,
+      "step": 371
+    },
+    {
+      "epoch": 0.05262971739822445,
+      "grad_norm": 0.0030788243748247623,
+      "learning_rate": 0.00014000291372372647,
+      "loss": 0.0002,
+      "step": 372
+    },
+    {
+      "epoch": 0.052771195133165916,
+      "grad_norm": 0.0028270557522773743,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 0.0002,
+      "step": 373
+    },
+    {
+      "epoch": 0.05291267286810738,
+      "grad_norm": 0.022989150136709213,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 0.0006,
+      "step": 374
+    },
+    {
+      "epoch": 0.05305415060304885,
+      "grad_norm": 0.008212580345571041,
+      "learning_rate": 0.00013913736668372026,
+      "loss": 0.0004,
+      "step": 375
+    },
+    {
+      "epoch": 0.05319562833799031,
+      "grad_norm": 0.0014105787267908454,
+      "learning_rate": 0.00013884807466313663,
+      "loss": 0.0001,
+      "step": 376
+    },
+    {
+      "epoch": 0.053337106072931774,
+      "grad_norm": 0.0034487394150346518,
+      "learning_rate": 0.00013855839922773968,
+      "loss": 0.0002,
+      "step": 377
+    },
+    {
+      "epoch": 0.05347858380787324,
+      "grad_norm": 0.0014626358170062304,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.0001,
+      "step": 378
+    },
+    {
+      "epoch": 0.0536200615428147,
+      "grad_norm": 0.001634497893974185,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 0.0001,
+      "step": 379
+    },
+    {
+      "epoch": 0.05376153927775616,
+      "grad_norm": 0.3064191937446594,
+      "learning_rate": 0.00013768710104121627,
+      "loss": 0.0022,
+      "step": 380
+    },
+    {
+      "epoch": 0.053903017012697625,
+      "grad_norm": 0.0029049685690551996,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 0.0002,
+      "step": 381
+    },
+    {
+      "epoch": 0.05404449474763909,
+      "grad_norm": 0.005087097175419331,
+      "learning_rate": 0.0001371043710237051,
+      "loss": 0.0003,
+      "step": 382
+    },
+    {
+      "epoch": 0.05418597248258055,
+      "grad_norm": 0.008555252104997635,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 0.0004,
+      "step": 383
+    },
+    {
+      "epoch": 0.05432745021752202,
+      "grad_norm": 0.0014736767625436187,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 0.0001,
+      "step": 384
+    },
+    {
+      "epoch": 0.05446892795246348,
+      "grad_norm": 0.0015276784542948008,
+      "learning_rate": 0.00013622753667045457,
+      "loss": 0.0001,
+      "step": 385
+    },
+    {
+      "epoch": 0.054610405687404946,
+      "grad_norm": 0.001212657312862575,
+      "learning_rate": 0.00013593453960058908,
+      "loss": 0.0001,
+      "step": 386
+    },
+    {
+      "epoch": 0.05475188342234641,
+      "grad_norm": 0.0023078841622918844,
+      "learning_rate": 0.00013564118787132506,
+      "loss": 0.0001,
+      "step": 387
+    },
+    {
+      "epoch": 0.05489336115728787,
+      "grad_norm": 0.002347096335142851,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 0.0002,
+      "step": 388
+    },
+    {
+      "epoch": 0.055034838892229335,
+      "grad_norm": 0.004173379857093096,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 0.0001,
+      "step": 389
+    },
+    {
+      "epoch": 0.0551763166271708,
+      "grad_norm": 0.000805948453489691,
+      "learning_rate": 0.0001347590336971037,
+      "loss": 0.0001,
+      "step": 390
+    },
+    {
+      "epoch": 0.05531779436211226,
+      "grad_norm": 0.00531614338979125,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 0.0001,
+      "step": 391
+    },
+    {
+      "epoch": 0.05545927209705372,
+      "grad_norm": 0.009082508273422718,
+      "learning_rate": 0.00013416921078914835,
+      "loss": 0.0003,
+      "step": 392
+    },
+    {
+      "epoch": 0.05560074983199519,
+      "grad_norm": 0.022566402330994606,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 0.0003,
+      "step": 393
+    },
+    {
+      "epoch": 0.055742227566936656,
+      "grad_norm": 0.004918436054140329,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 0.0001,
+      "step": 394
+    },
+    {
+      "epoch": 0.05588370530187812,
+      "grad_norm": 0.0007803619955666363,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 0.0001,
+      "step": 395
+    },
+    {
+      "epoch": 0.05602518303681958,
+      "grad_norm": 0.0024644150398671627,
+      "learning_rate": 0.00013298554148588528,
+      "loss": 0.0001,
+      "step": 396
+    },
+    {
+      "epoch": 0.056166660771761044,
+      "grad_norm": 0.008953128010034561,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 0.0006,
+      "step": 397
+    },
+    {
+      "epoch": 0.05630813850670251,
+      "grad_norm": 0.0007975678890943527,
+      "learning_rate": 0.00013239174181981495,
+      "loss": 0.0001,
+      "step": 398
+    },
+    {
+      "epoch": 0.05644961624164397,
+      "grad_norm": 0.0007346520433202386,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 0.0001,
+      "step": 399
+    },
+    {
+      "epoch": 0.05659109397658543,
+      "grad_norm": 0.025135928764939308,
+      "learning_rate": 0.00013179666338324108,
+      "loss": 0.0002,
+      "step": 400
+    },
+    {
+      "epoch": 0.056732571711526895,
+      "grad_norm": 0.0011148619232699275,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 0.0001,
+      "step": 401
+    },
+    {
+      "epoch": 0.056874049446468365,
+      "grad_norm": 0.0004933550371788442,
+      "learning_rate": 0.0001312003296688415,
+      "loss": 0.0,
+      "step": 402
+    },
+    {
+      "epoch": 0.05701552718140983,
+      "grad_norm": 0.000759758404456079,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.0001,
+      "step": 403
+    },
+    {
+      "epoch": 0.05715700491635129,
+      "grad_norm": 0.0008426797576248646,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 0.0001,
+      "step": 404
+    },
+    {
+      "epoch": 0.057298482651292754,
+      "grad_norm": 0.0015815444057807326,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 0.0001,
+      "step": 405
+    },
+    {
+      "epoch": 0.057439960386234216,
+      "grad_norm": 0.00043021078454330564,
+      "learning_rate": 0.00013000399062412763,
+      "loss": 0.0,
+      "step": 406
+    },
+    {
+      "epoch": 0.05758143812117568,
+      "grad_norm": 0.0008661262108944356,
+      "learning_rate": 0.0001297041581577035,
+      "loss": 0.0001,
+      "step": 407
+    },
+    {
+      "epoch": 0.05772291585611714,
+      "grad_norm": 0.004798825830221176,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 0.0001,
+      "step": 408
+    },
+    {
+      "epoch": 0.057864393591058605,
+      "grad_norm": 0.0007471769349649549,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 0.0001,
+      "step": 409
+    },
+    {
+      "epoch": 0.05800587132600007,
+      "grad_norm": 0.012662540189921856,
+      "learning_rate": 0.00012880291360147693,
+      "loss": 0.0001,
+      "step": 410
+    },
+    {
+      "epoch": 0.05814734906094154,
+      "grad_norm": 0.0005973036750219762,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 0.0,
+      "step": 411
+    },
+    {
+      "epoch": 0.058288826795883,
+      "grad_norm": 0.0005400076624937356,
+      "learning_rate": 0.00012820065759001293,
+      "loss": 0.0,
+      "step": 412
+    },
+    {
+      "epoch": 0.05843030453082446,
+      "grad_norm": 0.0010891173733398318,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 0.0001,
+      "step": 413
+    },
+    {
+      "epoch": 0.058571782265765926,
+      "grad_norm": 0.0004290946526452899,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 0.0,
+      "step": 414
+    },
+    {
+      "epoch": 0.05871326000070739,
+      "grad_norm": 0.0004823437484446913,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 0.0,
+      "step": 415
+    },
+    {
+      "epoch": 0.05885473773564885,
+      "grad_norm": 0.005344636272639036,
+      "learning_rate": 0.00012699282944604967,
+      "loss": 0.0001,
+      "step": 416
+    },
+    {
+      "epoch": 0.058996215470590314,
+      "grad_norm": 3.188809394836426,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 1.2313,
+      "step": 417
+    },
+    {
+      "epoch": 0.05913769320553178,
+      "grad_norm": 0.0019267729949206114,
+      "learning_rate": 0.0001263873049965373,
+      "loss": 0.0001,
+      "step": 418
+    },
+    {
+      "epoch": 0.05927917094047324,
+      "grad_norm": 0.0005391717422753572,
+      "learning_rate": 0.00012608415062898972,
+      "loss": 0.0,
+      "step": 419
+    },
+    {
+      "epoch": 0.05942064867541471,
+      "grad_norm": 0.005409176927059889,
+      "learning_rate": 0.000125780738821406,
+      "loss": 0.0003,
+      "step": 420
+    },
+    {
+      "epoch": 0.05956212641035617,
+      "grad_norm": 0.013438250869512558,
+      "learning_rate": 0.00012547707256833823,
+      "loss": 0.0006,
+      "step": 421
+    },
+    {
+      "epoch": 0.059703604145297635,
+      "grad_norm": 0.060834258794784546,
+      "learning_rate": 0.00012517315486684972,
+      "loss": 0.0019,
+      "step": 422
+    },
+    {
+      "epoch": 0.0598450818802391,
+      "grad_norm": 0.30073702335357666,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 0.0043,
+      "step": 423
+    },
+    {
+      "epoch": 0.05998655961518056,
+      "grad_norm": 0.5793721675872803,
+      "learning_rate": 0.00012456457711924266,
+      "loss": 0.023,
+      "step": 424
+    },
+    {
+      "epoch": 0.060128037350122024,
+      "grad_norm": 0.01727374456822872,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 0.0012,
+      "step": 425
+    },
+    {
+      "epoch": 0.06026951508506349,
+      "grad_norm": 0.039440348744392395,
+      "learning_rate": 0.0001239550296041922,
+      "loss": 0.002,
+      "step": 426
+    },
+    {
+      "epoch": 0.06041099282000495,
+      "grad_norm": 0.01554853469133377,
+      "learning_rate": 0.00012364989970237248,
+      "loss": 0.0007,
+      "step": 427
+    },
+    {
+      "epoch": 0.06055247055494641,
+      "grad_norm": 0.020410068333148956,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 0.0013,
+      "step": 428
+    },
+    {
+      "epoch": 0.06069394828988788,
+      "grad_norm": 3.5123672485351562,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 0.159,
+      "step": 429
+    },
+    {
+      "epoch": 0.060835426024829345,
+      "grad_norm": 0.010013503022491932,
+      "learning_rate": 0.00012273312156466464,
+      "loss": 0.0008,
+      "step": 430
+    },
+    {
+      "epoch": 0.06097690375977081,
+      "grad_norm": 0.060755547136068344,
+      "learning_rate": 0.00012242707609493814,
+      "loss": 0.0035,
+      "step": 431
+    },
+    {
+      "epoch": 0.06111838149471227,
+      "grad_norm": 0.005364975892007351,
+      "learning_rate": 0.00012212080927902474,
+      "loss": 0.0004,
+      "step": 432
+    },
+    {
+      "epoch": 0.06125985922965373,
+      "grad_norm": 0.014719590544700623,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 0.0007,
+      "step": 433
+    },
+    {
+      "epoch": 0.061401336964595196,
+      "grad_norm": 0.016079336404800415,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 0.0008,
+      "step": 434
+    },
+    {
+      "epoch": 0.06154281469953666,
+      "grad_norm": 0.03487243875861168,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 0.0018,
+      "step": 435
+    },
+    {
+      "epoch": 0.06168429243447812,
+      "grad_norm": 0.03025330975651741,
+      "learning_rate": 0.00012089358904024117,
+      "loss": 0.0015,
+      "step": 436
+    },
+    {
+      "epoch": 0.061825770169419585,
+      "grad_norm": 0.02542041428387165,
+      "learning_rate": 0.00012058626087698814,
+      "loss": 0.0008,
+      "step": 437
+    },
+    {
+      "epoch": 0.061967247904361054,
+      "grad_norm": 0.011038844473659992,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 0.0005,
+      "step": 438
+    },
+    {
+      "epoch": 0.06210872563930252,
+      "grad_norm": 0.020355192944407463,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 0.0009,
+      "step": 439
+    },
+    {
+      "epoch": 0.06225020337424398,
+      "grad_norm": 0.008449536748230457,
+      "learning_rate": 0.000119663069461542,
+      "loss": 0.0005,
+      "step": 440
+    },
+    {
+      "epoch": 0.06239168110918544,
+      "grad_norm": 0.008639265783131123,
+      "learning_rate": 0.00011935494680508606,
+      "loss": 0.0004,
+      "step": 441
+    },
+    {
+      "epoch": 0.0625331588441269,
+      "grad_norm": 0.020484648644924164,
+      "learning_rate": 0.00011904663312311901,
+      "loss": 0.0012,
+      "step": 442
+    },
+    {
+      "epoch": 0.06267463657906837,
+      "grad_norm": 0.0050698923878371716,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 0.0003,
+      "step": 443
+    },
+    {
+      "epoch": 0.06281611431400984,
+      "grad_norm": 0.004197041038423777,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 0.0003,
+      "step": 444
+    },
+    {
+      "epoch": 0.0629575920489513,
+      "grad_norm": 0.029855944216251373,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 0.0012,
+      "step": 445
+    },
+    {
+      "epoch": 0.06309906978389276,
+      "grad_norm": 0.002124859020113945,
+      "learning_rate": 0.000117811529026421,
+      "loss": 0.0001,
+      "step": 446
+    },
+    {
+      "epoch": 0.06324054751883422,
+      "grad_norm": 0.0030106690246611834,
+      "learning_rate": 0.00011750230589752762,
+      "loss": 0.0001,
+      "step": 447
+    },
+    {
+      "epoch": 0.06338202525377569,
+      "grad_norm": 0.003053938504308462,
+      "learning_rate": 0.00011719291002794096,
+      "loss": 0.0002,
+      "step": 448
+    },
+    {
+      "epoch": 0.06352350298871715,
+      "grad_norm": 0.0032461783848702908,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 0.0002,
+      "step": 449
+    },
+    {
+      "epoch": 0.06366498072365862,
+      "grad_norm": 0.02754434570670128,
+      "learning_rate": 0.00011657361228281199,
+      "loss": 0.0013,
+      "step": 450
+    },
+    {
+      "epoch": 0.06380645845860007,
+      "grad_norm": 0.0038834463339298964,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 0.0002,
+      "step": 451
+    },
+    {
+      "epoch": 0.06394793619354154,
+      "grad_norm": 0.0011302996426820755,
+      "learning_rate": 0.00011595366023984864,
+      "loss": 0.0001,
+      "step": 452
+    },
+    {
+      "epoch": 0.06408941392848301,
+      "grad_norm": 0.004102921113371849,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 0.0001,
+      "step": 453
+    },
+    {
+      "epoch": 0.06423089166342447,
+      "grad_norm": 0.0013566426932811737,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 0.0001,
+      "step": 454
+    },
+    {
+      "epoch": 0.06437236939836594,
+      "grad_norm": 0.0034373244270682335,
+      "learning_rate": 0.00011502255891207572,
+      "loss": 0.0001,
+      "step": 455
+    },
+    {
+      "epoch": 0.06451384713330739,
+      "grad_norm": 0.0010898098116740584,
+      "learning_rate": 0.00011471189118386375,
+      "loss": 0.0001,
+      "step": 456
+    },
+    {
+      "epoch": 0.06465532486824886,
+      "grad_norm": 0.0038209185004234314,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 0.0002,
+      "step": 457
+    },
+    {
+      "epoch": 0.06479680260319032,
+      "grad_norm": 0.0053474209271371365,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 0.0003,
+      "step": 458
+    },
+    {
+      "epoch": 0.06493828033813179,
+      "grad_norm": 0.0008347167167812586,
+      "learning_rate": 0.0001137790290684638,
+      "loss": 0.0001,
+      "step": 459
+    },
+    {
+      "epoch": 0.06507975807307324,
+      "grad_norm": 0.0008003226830624044,
+      "learning_rate": 0.00011346779894971527,
+      "loss": 0.0001,
+      "step": 460
+    },
+    {
+      "epoch": 0.06522123580801471,
+      "grad_norm": 0.0006106279324740171,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 0.0,
+      "step": 461
+    },
+    {
+      "epoch": 0.06536271354295618,
+      "grad_norm": 0.0030305362306535244,
+      "learning_rate": 0.0001128449430200303,
+      "loss": 0.0002,
+      "step": 462
+    },
+    {
+      "epoch": 0.06550419127789764,
+      "grad_norm": 0.003015139140188694,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 0.0002,
+      "step": 463
+    },
+    {
+      "epoch": 0.06564566901283911,
+      "grad_norm": 0.002168464008718729,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 0.0001,
+      "step": 464
+    },
+    {
+      "epoch": 0.06578714674778056,
+      "grad_norm": 0.0038204342126846313,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 0.0003,
+      "step": 465
+    },
+    {
+      "epoch": 0.06592862448272203,
+      "grad_norm": 0.0014754064613953233,
+      "learning_rate": 0.00011159773448089614,
+      "loss": 0.0001,
+      "step": 466
+    },
+    {
+      "epoch": 0.06607010221766349,
+      "grad_norm": 0.0011957096867263317,
+      "learning_rate": 0.00011128563848734816,
+      "loss": 0.0001,
+      "step": 467
+    },
+    {
+      "epoch": 0.06621157995260496,
+      "grad_norm": 0.002384569263085723,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 0.0,
+      "step": 468
+    },
+    {
+      "epoch": 0.06635305768754642,
+      "grad_norm": 0.0006383927538990974,
+      "learning_rate": 0.000110661115427526,
+      "loss": 0.0,
+      "step": 469
+    },
+    {
+      "epoch": 0.06649453542248789,
+      "grad_norm": 0.0007738766144029796,
+      "learning_rate": 0.00011034869452504226,
+      "loss": 0.0001,
+      "step": 470
+    },
+    {
+      "epoch": 0.06663601315742936,
+      "grad_norm": 0.11662693321704865,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 0.0009,
+      "step": 471
+    },
+    {
+      "epoch": 0.06677749089237081,
+      "grad_norm": 0.000526952208019793,
+      "learning_rate": 0.00010972354939223996,
+      "loss": 0.0,
+      "step": 472
+    },
+    {
+      "epoch": 0.06691896862731228,
+      "grad_norm": 0.001668064622208476,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 0.0001,
+      "step": 473
+    },
+    {
+      "epoch": 0.06706044636225374,
+      "grad_norm": 0.003478636732324958,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 0.0002,
+      "step": 474
+    },
+    {
+      "epoch": 0.0672019240971952,
+      "grad_norm": 0.0026004633400589228,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 0.0001,
+      "step": 475
+    },
+    {
+      "epoch": 0.06734340183213666,
+      "grad_norm": 0.0019350722432136536,
+      "learning_rate": 0.00010847213221420736,
+      "loss": 0.0001,
+      "step": 476
+    },
+    {
+      "epoch": 0.06748487956707813,
+      "grad_norm": 0.0009413132793270051,
+      "learning_rate": 0.00010815906115681578,
+      "loss": 0.0001,
+      "step": 477
+    },
+    {
+      "epoch": 0.06762635730201959,
+      "grad_norm": 0.00028766566538251936,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 0.0,
+      "step": 478
+    },
+    {
+      "epoch": 0.06776783503696106,
+      "grad_norm": 0.00023354949371423572,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 0.0,
+      "step": 479
+    },
+    {
+      "epoch": 0.06790931277190253,
+      "grad_norm": 0.002164566656574607,
+      "learning_rate": 0.0001072193771882861,
+      "loss": 0.0001,
+      "step": 480
+    },
+    {
+      "epoch": 0.06805079050684398,
+      "grad_norm": 0.0007285153842531145,
+      "learning_rate": 0.00010690600257144061,
+      "loss": 0.0001,
+      "step": 481
+    },
+    {
+      "epoch": 0.06819226824178545,
+      "grad_norm": 0.0002895043871831149,
+      "learning_rate": 0.0001065925597951378,
+      "loss": 0.0,
+      "step": 482
+    },
+    {
+      "epoch": 0.06833374597672691,
+      "grad_norm": 0.001868347986601293,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 0.0001,
+      "step": 483
+    },
+    {
+      "epoch": 0.06847522371166838,
+      "grad_norm": 0.0008055320358835161,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 0.0001,
+      "step": 484
+    },
+    {
+      "epoch": 0.06861670144660983,
+      "grad_norm": 0.0038549823220819235,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 0.0001,
+      "step": 485
+    },
+    {
+      "epoch": 0.0687581791815513,
+      "grad_norm": 0.003560754470527172,
+      "learning_rate": 0.00010533816897587606,
+      "loss": 0.0003,
+      "step": 486
+    },
+    {
+      "epoch": 0.06889965691649276,
+      "grad_norm": 0.0006623827503062785,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 0.0001,
+      "step": 487
+    },
+    {
+      "epoch": 0.06904113465143423,
+      "grad_norm": 0.030813848599791527,
+      "learning_rate": 0.00010471064507096426,
+      "loss": 0.0003,
+      "step": 488
+    },
+    {
+      "epoch": 0.0691826123863757,
+      "grad_norm": 0.003694392740726471,
+      "learning_rate": 0.0001043968118317865,
+      "loss": 0.0002,
+      "step": 489
+    },
+    {
+      "epoch": 0.06932409012131716,
+      "grad_norm": 0.0010670871706679463,
+      "learning_rate": 0.00010408293519785101,
+      "loss": 0.0001,
+      "step": 490
+    },
+    {
+      "epoch": 0.06946556785625863,
+      "grad_norm": 0.0027516393456608057,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 0.0001,
+      "step": 491
+    },
+    {
+      "epoch": 0.06960704559120008,
+      "grad_norm": 0.0029371054843068123,
+      "learning_rate": 0.00010345506413744726,
+      "loss": 0.0001,
+      "step": 492
+    },
+    {
+      "epoch": 0.06974852332614155,
+      "grad_norm": 0.0013108784332871437,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 0.0001,
+      "step": 493
+    },
+    {
+      "epoch": 0.069890001061083,
+      "grad_norm": 0.0004981230013072491,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 0.0,
+      "step": 494
+    },
+    {
+      "epoch": 0.07003147879602448,
+      "grad_norm": 0.00029754231218248606,
+      "learning_rate": 0.00010251300954433376,
+      "loss": 0.0,
+      "step": 495
+    },
+    {
+      "epoch": 0.07017295653096593,
+      "grad_norm": 0.0013778514694422483,
+      "learning_rate": 0.00010219893760925052,
+      "loss": 0.0,
+      "step": 496
+    },
+    {
+      "epoch": 0.0703144342659074,
+      "grad_norm": 0.0023718923330307007,
+      "learning_rate": 0.00010188484397154084,
+      "loss": 0.0001,
+      "step": 497
+    },
+    {
+      "epoch": 0.07045591200084887,
+      "grad_norm": 0.0007462603389285505,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 0.0001,
+      "step": 498
+    },
+    {
+      "epoch": 0.07059738973579033,
+      "grad_norm": 0.00037132593570277095,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 0.0,
+      "step": 499
+    },
+    {
+      "epoch": 0.0707388674707318,
+      "grad_norm": 0.000680119323078543,
+      "learning_rate": 0.00010094246384331442,
+      "loss": 0.0001,
+      "step": 500
+    },
+    {
+      "epoch": 0.07088034520567325,
+      "grad_norm": 0.0008801145595498383,
+      "learning_rate": 0.00010062831439655591,
+      "loss": 0.0001,
+      "step": 501
+    },
+    {
+      "epoch": 0.07102182294061472,
+      "grad_norm": 0.0019729931373149157,
+      "learning_rate": 0.00010031415874858797,
+      "loss": 0.0002,
+      "step": 502
+    },
+    {
+      "epoch": 0.07116330067555618,
+      "grad_norm": 0.0009173870203085244,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 503
+    },
+    {
+      "epoch": 0.07130477841049765,
+      "grad_norm": 0.0003767950984183699,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 0.0,
+      "step": 504
+    },
+    {
+      "epoch": 0.0714462561454391,
+      "grad_norm": 0.0010468692053109407,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 0.0001,
+      "step": 505
+    },
+    {
+      "epoch": 0.07158773388038057,
+      "grad_norm": 0.0019238515524193645,
+      "learning_rate": 9.90575361566856e-05,
+      "loss": 0.0001,
+      "step": 506
+    },
+    {
+      "epoch": 0.07172921161532204,
+      "grad_norm": 0.00045130791841074824,
+      "learning_rate": 9.874339601166473e-05,
+      "loss": 0.0,
+      "step": 507
+    },
+    {
+      "epoch": 0.0718706893502635,
+      "grad_norm": 0.0017480598762631416,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 0.0001,
+      "step": 508
+    },
+    {
+      "epoch": 0.07201216708520497,
+      "grad_norm": 0.0005805511609651148,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 0.0,
+      "step": 509
+    },
+    {
+      "epoch": 0.07215364482014643,
+      "grad_norm": 0.00020824585226364434,
+      "learning_rate": 9.78010623907495e-05,
+      "loss": 0.0,
+      "step": 510
+    },
+    {
+      "epoch": 0.0722951225550879,
+      "grad_norm": 0.000254087004577741,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 0.0,
+      "step": 511
+    },
+    {
+      "epoch": 0.07243660029002935,
+      "grad_norm": 0.0006141222547739744,
+      "learning_rate": 9.717294332297268e-05,
+      "loss": 0.0,
+      "step": 512
+    },
+    {
+      "epoch": 0.07257807802497082,
+      "grad_norm": 0.0007048989646136761,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 0.0001,
+      "step": 513
+    },
+    {
+      "epoch": 0.07271955575991228,
+      "grad_norm": 0.00048093812074512243,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 0.0,
+      "step": 514
+    },
+    {
+      "epoch": 0.07286103349485375,
+      "grad_norm": 0.0009035237017087638,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 0.0001,
+      "step": 515
+    },
+    {
+      "epoch": 0.07300251122979522,
+      "grad_norm": 0.0005546315223909914,
+      "learning_rate": 9.591706480214901e-05,
+      "loss": 0.0,
+      "step": 516
+    },
+    {
+      "epoch": 0.07314398896473667,
+      "grad_norm": 0.017050910741090775,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 0.0005,
+      "step": 517
+    },
+    {
+      "epoch": 0.07328546669967814,
+      "grad_norm": 0.015889961272478104,
+      "learning_rate": 9.528935492903575e-05,
+      "loss": 0.0002,
+      "step": 518
+    },
+    {
+      "epoch": 0.0734269444346196,
+      "grad_norm": 0.0003646703262347728,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 0.0,
+      "step": 519
+    },
+    {
+      "epoch": 0.07356842216956107,
+      "grad_norm": 0.0006750635802745819,
+      "learning_rate": 9.466183102412395e-05,
+      "loss": 0.0,
+      "step": 520
+    },
+    {
+      "epoch": 0.07370989990450252,
+      "grad_norm": 0.0002909269533120096,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 0.0,
+      "step": 521
+    },
+    {
+      "epoch": 0.073851377639444,
+      "grad_norm": 0.0032838047482073307,
+      "learning_rate": 9.403451786098294e-05,
+      "loss": 0.0001,
+      "step": 522
+    },
+    {
+      "epoch": 0.07399285537438545,
+      "grad_norm": 0.0004030920681543648,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 0.0,
+      "step": 523
+    },
+    {
+      "epoch": 0.07413433310932692,
+      "grad_norm": 0.003876444650813937,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 0.0001,
+      "step": 524
+    },
+    {
+      "epoch": 0.07427581084426839,
+      "grad_norm": 0.0003544877690728754,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 0.0,
+      "step": 525
+    },
+    {
+      "epoch": 0.07441728857920984,
+      "grad_norm": 0.0003599386545829475,
+      "learning_rate": 9.278062281171393e-05,
+      "loss": 0.0,
+      "step": 526
+    },
+    {
+      "epoch": 0.07455876631415131,
+      "grad_norm": 0.0009949669474735856,
+      "learning_rate": 9.246731944720675e-05,
+      "loss": 0.0001,
+      "step": 527
+    },
+    {
+      "epoch": 0.07470024404909277,
+      "grad_norm": 0.0006730380118824542,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 0.0,
+      "step": 528
+    },
+    {
+      "epoch": 0.07484172178403424,
+      "grad_norm": 0.0006648433045484126,
+      "learning_rate": 9.184093884318425e-05,
+      "loss": 0.0,
+      "step": 529
+    },
+    {
+      "epoch": 0.0749831995189757,
+      "grad_norm": 0.0010465396335348487,
+      "learning_rate": 9.152786778579267e-05,
+      "loss": 0.0001,
+      "step": 530
+    },
+    {
+      "epoch": 0.07512467725391717,
+      "grad_norm": 0.0002523198490962386,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 0.0,
+      "step": 531
+    },
+    {
+      "epoch": 0.07526615498885862,
+      "grad_norm": 0.0005815126351080835,
+      "learning_rate": 9.090197960964301e-05,
+      "loss": 0.0,
+      "step": 532
+    },
+    {
+      "epoch": 0.07540763272380009,
+      "grad_norm": 0.0027641430497169495,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 0.0001,
+      "step": 533
+    },
+    {
+      "epoch": 0.07554911045874156,
+      "grad_norm": 0.0037838981952518225,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 0.0002,
+      "step": 534
+    },
+    {
+      "epoch": 0.07569058819368302,
+      "grad_norm": 0.00022687749878969043,
+      "learning_rate": 8.99638285148785e-05,
+      "loss": 0.0,
+      "step": 535
+    },
+    {
+      "epoch": 0.07583206592862449,
+      "grad_norm": 0.0002952757349703461,
+      "learning_rate": 8.965130547495776e-05,
+      "loss": 0.0,
+      "step": 536
+    },
+    {
+      "epoch": 0.07597354366356594,
+      "grad_norm": 0.00020262067846488208,
+      "learning_rate": 8.933888457247402e-05,
+      "loss": 0.0,
+      "step": 537
+    },
+    {
+      "epoch": 0.07611502139850741,
+      "grad_norm": 0.00022801620070822537,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 0.0,
+      "step": 538
+    },
+    {
+      "epoch": 0.07625649913344887,
+      "grad_norm": 0.0029370381962507963,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 0.0001,
+      "step": 539
+    },
+    {
+      "epoch": 0.07639797686839034,
+      "grad_norm": 0.00019757526752073318,
+      "learning_rate": 8.840226551910387e-05,
+      "loss": 0.0,
+      "step": 540
+    },
+    {
+      "epoch": 0.0765394546033318,
+      "grad_norm": 0.00035157534875907004,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 0.0,
+      "step": 541
+    },
+    {
+      "epoch": 0.07668093233827326,
+      "grad_norm": 0.00033226568484678864,
+      "learning_rate": 8.777842000601105e-05,
+      "loss": 0.0,
+      "step": 542
+    },
+    {
+      "epoch": 0.07682241007321473,
+      "grad_norm": 0.0004599531239364296,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 0.0,
+      "step": 543
+    },
+    {
+      "epoch": 0.07696388780815619,
+      "grad_norm": 0.000987498089671135,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 0.0001,
+      "step": 544
+    },
+    {
+      "epoch": 0.07710536554309766,
+      "grad_norm": 0.0003310377069283277,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 0.0,
+      "step": 545
+    },
+    {
+      "epoch": 0.07724684327803912,
+      "grad_norm": 0.0006641054060310125,
+      "learning_rate": 8.653220105028474e-05,
+      "loss": 0.0001,
+      "step": 546
+    },
+    {
+      "epoch": 0.07738832101298058,
+      "grad_norm": 0.0002603815810289234,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 0.0,
+      "step": 547
+    },
+    {
+      "epoch": 0.07752979874792204,
+      "grad_norm": 0.004001593217253685,
+      "learning_rate": 8.590987680624174e-05,
+      "loss": 0.0001,
+      "step": 548
+    },
+    {
+      "epoch": 0.07767127648286351,
+      "grad_norm": 0.00022212194744497538,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 0.0,
+      "step": 549
+    },
+    {
+      "epoch": 0.07781275421780497,
+      "grad_norm": 0.07186271995306015,
+      "learning_rate": 8.528810881613626e-05,
+      "loss": 0.0022,
+      "step": 550
+    },
+    {
+      "epoch": 0.07795423195274644,
+      "grad_norm": 0.0003144246293231845,
+      "learning_rate": 8.497744108792429e-05,
+      "loss": 0.0,
+      "step": 551
+    },
+    {
+      "epoch": 0.0780957096876879,
+      "grad_norm": 0.005859481170773506,
+      "learning_rate": 8.466692162630392e-05,
+      "loss": 0.0003,
+      "step": 552
+    },
+    {
+      "epoch": 0.07823718742262936,
+      "grad_norm": 0.000740467687137425,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 0.0,
+      "step": 553
+    },
+    {
+      "epoch": 0.07837866515757083,
+      "grad_norm": 0.0005886468570679426,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 0.0,
+      "step": 554
+    },
+    {
+      "epoch": 0.07852014289251229,
+      "grad_norm": 0.0009009242057800293,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 0.0001,
+      "step": 555
+    },
+    {
+      "epoch": 0.07866162062745376,
+      "grad_norm": 0.0015645289095118642,
+      "learning_rate": 8.342638771718802e-05,
+      "loss": 0.0001,
+      "step": 556
+    },
+    {
+      "epoch": 0.07880309836239521,
+      "grad_norm": 0.0002594326506368816,
+      "learning_rate": 8.311665552872662e-05,
+      "loss": 0.0,
+      "step": 557
+    },
+    {
+      "epoch": 0.07894457609733668,
+      "grad_norm": 0.00021320579980965704,
+      "learning_rate": 8.280708997205904e-05,
+      "loss": 0.0,
+      "step": 558
+    },
+    {
+      "epoch": 0.07908605383227814,
+      "grad_norm": 0.00017182575538754463,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 0.0,
+      "step": 559
+    },
+    {
+      "epoch": 0.07922753156721961,
+      "grad_norm": 0.00021368375746533275,
+      "learning_rate": 8.218847097357898e-05,
+      "loss": 0.0,
+      "step": 560
+    },
+    {
+      "epoch": 0.07936900930216108,
+      "grad_norm": 0.0002579468709882349,
+      "learning_rate": 8.187942363728625e-05,
+      "loss": 0.0,
+      "step": 561
+    },
+    {
+      "epoch": 0.07951048703710253,
+      "grad_norm": 0.00014378316700458527,
+      "learning_rate": 8.157055514376666e-05,
+      "loss": 0.0,
+      "step": 562
+    },
+    {
+      "epoch": 0.079651964772044,
+      "grad_norm": 0.0006984220235608518,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 0.0,
+      "step": 563
+    },
+    {
+      "epoch": 0.07979344250698546,
+      "grad_norm": 0.00012296480417717248,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 0.0,
+      "step": 564
+    },
+    {
+      "epoch": 0.07993492024192693,
+      "grad_norm": 0.0001581730757607147,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 0.0,
+      "step": 565
+    },
+    {
+      "epoch": 0.08007639797686839,
+      "grad_norm": 0.0001595177163835615,
+      "learning_rate": 8.033693053845801e-05,
+      "loss": 0.0,
+      "step": 566
+    },
+    {
+      "epoch": 0.08021787571180986,
+      "grad_norm": 0.8796876668930054,
+      "learning_rate": 8.002900194855932e-05,
+      "loss": 0.4228,
+      "step": 567
+    },
+    {
+      "epoch": 0.08035935344675131,
+      "grad_norm": 0.0006363423308357596,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 0.0,
+      "step": 568
+    },
+    {
+      "epoch": 0.08050083118169278,
+      "grad_norm": 0.0015658728079870343,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 0.0001,
+      "step": 569
+    },
+    {
+      "epoch": 0.08064230891663425,
+      "grad_norm": 0.0007438261527568102,
+      "learning_rate": 7.910641095975886e-05,
+      "loss": 0.0001,
+      "step": 570
+    },
+    {
+      "epoch": 0.0807837866515757,
+      "grad_norm": 0.0014448019210249186,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 0.0,
+      "step": 571
+    },
+    {
+      "epoch": 0.08092526438651718,
+      "grad_norm": 0.00021074670075904578,
+      "learning_rate": 7.849237629828869e-05,
+      "loss": 0.0,
+      "step": 572
+    },
+    {
+      "epoch": 0.08106674212145863,
+      "grad_norm": 0.00042521391878835857,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 0.0,
+      "step": 573
+    },
+    {
+      "epoch": 0.0812082198564001,
+      "grad_norm": 0.0009682802483439445,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 0.0001,
+      "step": 574
+    },
+    {
+      "epoch": 0.08134969759134156,
+      "grad_norm": 0.0012320525711402297,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 0.0001,
+      "step": 575
+    },
+    {
+      "epoch": 0.08149117532628303,
+      "grad_norm": 0.00030932322260923684,
+      "learning_rate": 7.726687843533538e-05,
+      "loss": 0.0,
+      "step": 576
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 0.0066994656808674335,
+      "learning_rate": 7.696105733234098e-05,
+      "loss": 0.0004,
+      "step": 577
+    },
+    {
+      "epoch": 0.08177413079616595,
+      "grad_norm": 0.0007564704865217209,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 0.0001,
+      "step": 578
+    },
+    {
+      "epoch": 0.08191560853110742,
+      "grad_norm": 0.0006317924126051366,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 0.0001,
+      "step": 579
+    },
+    {
+      "epoch": 0.08205708626604888,
+      "grad_norm": 0.00094227195950225,
+      "learning_rate": 7.604497039580785e-05,
+      "loss": 0.0001,
+      "step": 580
+    },
+    {
+      "epoch": 0.08219856400099035,
+      "grad_norm": 0.0027162383776158094,
+      "learning_rate": 7.574007692045928e-05,
+      "loss": 0.0002,
+      "step": 581
+    },
+    {
+      "epoch": 0.0823400417359318,
+      "grad_norm": 0.003974114544689655,
+      "learning_rate": 7.543542288075739e-05,
+      "loss": 0.0001,
+      "step": 582
+    },
+    {
+      "epoch": 0.08248151947087327,
+      "grad_norm": 0.004181701224297285,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 0.0003,
+      "step": 583
+    },
+    {
+      "epoch": 0.08262299720581473,
+      "grad_norm": 0.0016781368758529425,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 0.0001,
+      "step": 584
+    },
+    {
+      "epoch": 0.0827644749407562,
+      "grad_norm": 0.0009342627017758787,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 0.0001,
+      "step": 585
+    },
+    {
+      "epoch": 0.08290595267569766,
+      "grad_norm": 0.05183006078004837,
+      "learning_rate": 7.421926117859403e-05,
+      "loss": 0.0026,
+      "step": 586
+    },
+    {
+      "epoch": 0.08304743041063913,
+      "grad_norm": 0.0014916006475687027,
+      "learning_rate": 7.391584937101033e-05,
+      "loss": 0.0001,
+      "step": 587
+    },
+    {
+      "epoch": 0.0831889081455806,
+      "grad_norm": 0.001441052882000804,
+      "learning_rate": 7.361269500346274e-05,
+      "loss": 0.0001,
+      "step": 588
+    },
+    {
+      "epoch": 0.08333038588052205,
+      "grad_norm": 0.003165285335853696,
+      "learning_rate": 7.330980106796246e-05,
+      "loss": 0.0002,
+      "step": 589
+    },
+    {
+      "epoch": 0.08347186361546352,
+      "grad_norm": 0.00124356709420681,
+      "learning_rate": 7.300717055395039e-05,
+      "loss": 0.0001,
+      "step": 590
+    },
+    {
+      "epoch": 0.08361334135040498,
+      "grad_norm": 0.0007717587286606431,
+      "learning_rate": 7.270480644826749e-05,
+      "loss": 0.0001,
+      "step": 591
+    },
+    {
+      "epoch": 0.08375481908534645,
+      "grad_norm": 0.0012677700724452734,
+      "learning_rate": 7.240271173512546e-05,
+      "loss": 0.0001,
+      "step": 592
+    },
+    {
+      "epoch": 0.0838962968202879,
+      "grad_norm": 0.008508299477398396,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 0.0005,
+      "step": 593
+    },
+    {
+      "epoch": 0.08403777455522937,
+      "grad_norm": 0.0027715216856449842,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 0.0002,
+      "step": 594
+    },
+    {
+      "epoch": 0.08417925229017083,
+      "grad_norm": 0.0006361504783853889,
+      "learning_rate": 7.149807375300239e-05,
+      "loss": 0.0001,
+      "step": 595
+    },
+    {
+      "epoch": 0.0843207300251123,
+      "grad_norm": 0.0007495923200622201,
+      "learning_rate": 7.119708639852312e-05,
+      "loss": 0.0001,
+      "step": 596
+    },
+    {
+      "epoch": 0.08446220776005377,
+      "grad_norm": 0.0008743112557567656,
+      "learning_rate": 7.089638331717284e-05,
+      "loss": 0.0001,
+      "step": 597
+    },
+    {
+      "epoch": 0.08460368549499522,
+      "grad_norm": 0.006306476425379515,
+      "learning_rate": 7.059596747676962e-05,
+      "loss": 0.0004,
+      "step": 598
+    },
+    {
+      "epoch": 0.08474516322993669,
+      "grad_norm": 0.2450258582830429,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 0.002,
+      "step": 599
+    },
+    {
+      "epoch": 0.08488664096487815,
+      "grad_norm": 0.0009690061560831964,
+      "learning_rate": 6.999600937587239e-05,
+      "loss": 0.0001,
+      "step": 600
+    },
+    {
+      "epoch": 0.08502811869981962,
+      "grad_norm": 0.004833643790334463,
+      "learning_rate": 6.969647303672262e-05,
+      "loss": 0.0003,
+      "step": 601
+    },
+    {
+      "epoch": 0.08516959643476107,
+      "grad_norm": 0.00044146020081825554,
+      "learning_rate": 6.939723578114993e-05,
+      "loss": 0.0,
+      "step": 602
+    },
+    {
+      "epoch": 0.08531107416970254,
+      "grad_norm": 0.000619858386926353,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.0001,
+      "step": 603
+    },
+    {
+      "epoch": 0.085452551904644,
+      "grad_norm": 0.0006960814935155213,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 0.0001,
+      "step": 604
+    },
+    {
+      "epoch": 0.08559402963958547,
+      "grad_norm": 0.019819116219878197,
+      "learning_rate": 6.850134803446954e-05,
+      "loss": 0.0012,
+      "step": 605
+    },
+    {
+      "epoch": 0.08573550737452694,
+      "grad_norm": 0.0005196069832891226,
+      "learning_rate": 6.820333661675893e-05,
+      "loss": 0.0,
+      "step": 606
+    },
+    {
+      "epoch": 0.0858769851094684,
+      "grad_norm": 0.0008297892636619508,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 0.0001,
+      "step": 607
+    },
+    {
+      "epoch": 0.08601846284440987,
+      "grad_norm": 23.886945724487305,
+      "learning_rate": 6.760825818018508e-05,
+      "loss": 0.3313,
+      "step": 608
+    },
+    {
+      "epoch": 0.08615994057935132,
+      "grad_norm": 0.001881247735582292,
+      "learning_rate": 6.731119703450577e-05,
+      "loss": 0.0001,
+      "step": 609
+    },
+    {
+      "epoch": 0.08630141831429279,
+      "grad_norm": 0.0037710980977863073,
+      "learning_rate": 6.701445851411472e-05,
+      "loss": 0.0002,
+      "step": 610
+    },
+    {
+      "epoch": 0.08644289604923425,
+      "grad_norm": 0.001079973648302257,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 0.0001,
+      "step": 611
+    },
+    {
+      "epoch": 0.08658437378417572,
+      "grad_norm": 0.002796902321279049,
+      "learning_rate": 6.642196106074194e-05,
+      "loss": 0.0001,
+      "step": 612
+    },
+    {
+      "epoch": 0.08672585151911717,
+      "grad_norm": 0.023120107129216194,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 0.0005,
+      "step": 613
+    },
+    {
+      "epoch": 0.08686732925405864,
+      "grad_norm": 0.04147057607769966,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 0.0011,
+      "step": 614
+    },
+    {
+      "epoch": 0.08700880698900011,
+      "grad_norm": 0.033874575048685074,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 0.001,
+      "step": 615
+    },
+    {
+      "epoch": 0.08715028472394157,
+      "grad_norm": 0.07600700855255127,
+      "learning_rate": 6.52409663028963e-05,
+      "loss": 0.0032,
+      "step": 616
+    },
+    {
+      "epoch": 0.08729176245888304,
+      "grad_norm": 0.019974149763584137,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 0.001,
+      "step": 617
+    },
+    {
+      "epoch": 0.0874332401938245,
+      "grad_norm": 0.015204976312816143,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 0.0006,
+      "step": 618
+    },
+    {
+      "epoch": 0.08757471792876596,
+      "grad_norm": 0.01299876905977726,
+      "learning_rate": 6.435881212867493e-05,
+      "loss": 0.0005,
+      "step": 619
+    },
+    {
+      "epoch": 0.08771619566370742,
+      "grad_norm": 0.02097712829709053,
+      "learning_rate": 6.406546039941094e-05,
+      "loss": 0.0007,
+      "step": 620
+    },
+    {
+      "epoch": 0.08785767339864889,
+      "grad_norm": 0.020041853189468384,
+      "learning_rate": 6.377246332954544e-05,
+      "loss": 0.0007,
+      "step": 621
+    },
+    {
+      "epoch": 0.08799915113359034,
+      "grad_norm": 0.021356869488954544,
+      "learning_rate": 6.347982381084123e-05,
+      "loss": 0.0007,
+      "step": 622
+    },
+    {
+      "epoch": 0.08814062886853181,
+      "grad_norm": 0.006782687269151211,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 0.0002,
+      "step": 623
+    },
+    {
+      "epoch": 0.08828210660347328,
+      "grad_norm": 0.008033067919313908,
+      "learning_rate": 6.289562897629492e-05,
+      "loss": 0.0002,
+      "step": 624
+    },
+    {
+      "epoch": 0.08842358433841474,
+      "grad_norm": 0.0033709348645061255,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 0.0001,
+      "step": 625
+    },
+    {
+      "epoch": 0.08856506207335621,
+      "grad_norm": 0.005288586486130953,
+      "learning_rate": 6.231289895878375e-05,
+      "loss": 0.0002,
+      "step": 626
+    },
+    {
+      "epoch": 0.08870653980829767,
+      "grad_norm": 0.0014732020208612084,
+      "learning_rate": 6.20220904478199e-05,
+      "loss": 0.0,
+      "step": 627
+    },
+    {
+      "epoch": 0.08884801754323914,
+      "grad_norm": 0.002344023436307907,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.0001,
+      "step": 628
+    },
+    {
+      "epoch": 0.08898949527818059,
+      "grad_norm": 0.0037802611477673054,
+      "learning_rate": 6.144160077226036e-05,
+      "loss": 0.0002,
+      "step": 629
+    },
+    {
+      "epoch": 0.08913097301312206,
+      "grad_norm": 0.0017438894137740135,
+      "learning_rate": 6.11519253368634e-05,
+      "loss": 0.0001,
+      "step": 630
+    },
+    {
+      "epoch": 0.08927245074806352,
+      "grad_norm": 0.006431651767343283,
+      "learning_rate": 6.086263331627976e-05,
+      "loss": 0.0001,
+      "step": 631
+    },
+    {
+      "epoch": 0.08941392848300499,
+      "grad_norm": 0.0024249274283647537,
+      "learning_rate": 6.05737275657049e-05,
+      "loss": 0.0001,
+      "step": 632
+    },
+    {
+      "epoch": 0.08955540621794646,
+      "grad_norm": 0.009574075229465961,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 0.0006,
+      "step": 633
+    },
+    {
+      "epoch": 0.08969688395288791,
+      "grad_norm": 0.0018211673013865948,
+      "learning_rate": 5.999708627627354e-05,
+      "loss": 0.0001,
+      "step": 634
+    },
+    {
+      "epoch": 0.08983836168782938,
+      "grad_norm": 0.0010810673702508211,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 0.0,
+      "step": 635
+    },
+    {
+      "epoch": 0.08997983942277084,
+      "grad_norm": 0.0009683024836704135,
+      "learning_rate": 5.9422024233380013e-05,
+      "loss": 0.0001,
+      "step": 636
+    },
+    {
+      "epoch": 0.09012131715771231,
+      "grad_norm": 0.0022739556152373552,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 0.0002,
+      "step": 637
+    },
+    {
+      "epoch": 0.09026279489265376,
+      "grad_norm": 0.0006144489743746817,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 0.0,
+      "step": 638
+    },
+    {
+      "epoch": 0.09040427262759523,
+      "grad_norm": 0.001888429163955152,
+      "learning_rate": 5.856244190067159e-05,
+      "loss": 0.0001,
+      "step": 639
+    },
+    {
+      "epoch": 0.09054575036253669,
+      "grad_norm": 0.0006371226045303047,
+      "learning_rate": 5.82767286338235e-05,
+      "loss": 0.0,
+      "step": 640
+    },
+    {
+      "epoch": 0.09068722809747816,
+      "grad_norm": 0.0011725660879164934,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 0.0001,
+      "step": 641
+    },
+    {
+      "epoch": 0.09082870583241963,
+      "grad_norm": 0.0020375438034534454,
+      "learning_rate": 5.770654029146969e-05,
+      "loss": 0.0001,
+      "step": 642
+    },
+    {
+      "epoch": 0.09097018356736108,
+      "grad_norm": 0.00023711427638772875,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 0.0,
+      "step": 643
+    },
+    {
+      "epoch": 0.09111166130230255,
+      "grad_norm": 0.0010858512250706553,
+      "learning_rate": 5.713802162248718e-05,
+      "loss": 0.0001,
+      "step": 644
+    },
+    {
+      "epoch": 0.09125313903724401,
+      "grad_norm": 0.008106240071356297,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 0.0003,
+      "step": 645
+    },
+    {
+      "epoch": 0.09139461677218548,
+      "grad_norm": 0.0011279614409431815,
+      "learning_rate": 5.657119507101954e-05,
+      "loss": 0.0001,
+      "step": 646
+    },
+    {
+      "epoch": 0.09153609450712694,
+      "grad_norm": 0.001151368604041636,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 0.0,
+      "step": 647
+    },
+    {
+      "epoch": 0.0916775722420684,
+      "grad_norm": 0.002258772263303399,
+      "learning_rate": 5.6006083014408484e-05,
+      "loss": 0.0001,
+      "step": 648
+    },
+    {
+      "epoch": 0.09181904997700986,
+      "grad_norm": 0.0006376850651577115,
+      "learning_rate": 5.572417689610987e-05,
+      "loss": 0.0,
+      "step": 649
+    },
+    {
+      "epoch": 0.09196052771195133,
+      "grad_norm": 0.0002894207718782127,
+      "learning_rate": 5.544270776231038e-05,
+      "loss": 0.0,
+      "step": 650
+    },
+    {
+      "epoch": 0.0921020054468928,
+      "grad_norm": 0.0004874817677773535,
+      "learning_rate": 5.5161678390996796e-05,
+      "loss": 0.0,
+      "step": 651
+    },
+    {
+      "epoch": 0.09224348318183426,
+      "grad_norm": 0.0003829976776614785,
+      "learning_rate": 5.488109155581549e-05,
+      "loss": 0.0,
+      "step": 652
+    },
+    {
+      "epoch": 0.09238496091677573,
+      "grad_norm": 0.0019727759063243866,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 0.0001,
+      "step": 653
+    },
+    {
+      "epoch": 0.09252643865171718,
+      "grad_norm": 0.00044044721289537847,
+      "learning_rate": 5.4321256566570036e-05,
+      "loss": 0.0,
+      "step": 654
+    },
+    {
+      "epoch": 0.09266791638665865,
+      "grad_norm": 0.00258098216727376,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 0.0002,
+      "step": 655
+    },
+    {
+      "epoch": 0.09280939412160011,
+      "grad_norm": 0.0012749796733260155,
+      "learning_rate": 5.3763224895900846e-05,
+      "loss": 0.0001,
+      "step": 656
+    },
+    {
+      "epoch": 0.09295087185654158,
+      "grad_norm": 0.0007071017753332853,
+      "learning_rate": 5.348489219225416e-05,
+      "loss": 0.0,
+      "step": 657
+    },
+    {
+      "epoch": 0.09309234959148303,
+      "grad_norm": 0.0017588926712051034,
+      "learning_rate": 5.320701857394268e-05,
+      "loss": 0.0001,
+      "step": 658
+    },
+    {
+      "epoch": 0.0932338273264245,
+      "grad_norm": 0.002789848716929555,
+      "learning_rate": 5.292960678346675e-05,
+      "loss": 0.0002,
+      "step": 659
+    },
+    {
+      "epoch": 0.09337530506136597,
+      "grad_norm": 0.00038435394526459277,
+      "learning_rate": 5.265265955876879e-05,
+      "loss": 0.0,
+      "step": 660
+    },
+    {
+      "epoch": 0.09351678279630743,
+      "grad_norm": 0.0007107415585778654,
+      "learning_rate": 5.237617963320608e-05,
+      "loss": 0.0,
+      "step": 661
+    },
+    {
+      "epoch": 0.0936582605312489,
+      "grad_norm": 0.0032040157821029425,
+      "learning_rate": 5.210016973552391e-05,
+      "loss": 0.0002,
+      "step": 662
+    },
+    {
+      "epoch": 0.09379973826619036,
+      "grad_norm": 0.0017821533838286996,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 0.0001,
+      "step": 663
+    },
+    {
+      "epoch": 0.09394121600113182,
+      "grad_norm": 0.0003448922943789512,
+      "learning_rate": 5.1549570915560206e-05,
+      "loss": 0.0,
+      "step": 664
+    },
+    {
+      "epoch": 0.09408269373607328,
+      "grad_norm": 0.0010452198330312967,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 0.0,
+      "step": 665
+    },
+    {
+      "epoch": 0.09422417147101475,
+      "grad_norm": 0.0009789979085326195,
+      "learning_rate": 5.100088483557634e-05,
+      "loss": 0.0,
+      "step": 666
+    },
+    {
+      "epoch": 0.0943656492059562,
+      "grad_norm": 0.0005419703666120768,
+      "learning_rate": 5.072726584517086e-05,
+      "loss": 0.0,
+      "step": 667
+    },
+    {
+      "epoch": 0.09450712694089768,
+      "grad_norm": 0.00028989804559387267,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 0.0,
+      "step": 668
+    },
+    {
+      "epoch": 0.09464860467583915,
+      "grad_norm": 0.0018694131867960095,
+      "learning_rate": 5.018148946605092e-05,
+      "loss": 0.0001,
+      "step": 669
+    },
+    {
+      "epoch": 0.0947900824107806,
+      "grad_norm": 0.0001970644952962175,
+      "learning_rate": 4.990933746392899e-05,
+      "loss": 0.0,
+      "step": 670
+    },
+    {
+      "epoch": 0.09493156014572207,
+      "grad_norm": 0.00024093178217299283,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 0.0,
+      "step": 671
+    },
+    {
+      "epoch": 0.09507303788066353,
+      "grad_norm": 0.001606027246452868,
+      "learning_rate": 4.9366519264686725e-05,
+      "loss": 0.0,
+      "step": 672
+    },
+    {
+      "epoch": 0.095214515615605,
+      "grad_norm": 0.0034317609388381243,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 0.0002,
+      "step": 673
+    },
+    {
+      "epoch": 0.09535599335054645,
+      "grad_norm": 0.0005060399416834116,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 0.0,
+      "step": 674
+    },
+    {
+      "epoch": 0.09549747108548792,
+      "grad_norm": 0.00040333837387152016,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 0.0,
+      "step": 675
+    },
+    {
+      "epoch": 0.09563894882042938,
+      "grad_norm": 0.032013844698667526,
+      "learning_rate": 4.828690098618429e-05,
+      "loss": 0.0004,
+      "step": 676
+    },
+    {
+      "epoch": 0.09578042655537085,
+      "grad_norm": 0.0005927220336161554,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 0.0,
+      "step": 677
+    },
+    {
+      "epoch": 0.09592190429031232,
+      "grad_norm": 0.0006017786217853427,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 0.0,
+      "step": 678
+    },
+    {
+      "epoch": 0.09606338202525377,
+      "grad_norm": 0.002094410127028823,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 0.0001,
+      "step": 679
+    },
+    {
+      "epoch": 0.09620485976019524,
+      "grad_norm": 0.0008351746946573257,
+      "learning_rate": 4.721544880549337e-05,
+      "loss": 0.0,
+      "step": 680
+    },
+    {
+      "epoch": 0.0963463374951367,
+      "grad_norm": 0.07186631858348846,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 0.0023,
+      "step": 681
+    },
+    {
+      "epoch": 0.09648781523007817,
+      "grad_norm": 0.0034203226678073406,
+      "learning_rate": 4.668283792628114e-05,
+      "loss": 0.0001,
+      "step": 682
+    },
+    {
+      "epoch": 0.09662929296501963,
+      "grad_norm": 0.0005536339012905955,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 0.0,
+      "step": 683
+    },
+    {
+      "epoch": 0.0967707706999611,
+      "grad_norm": 0.006527572870254517,
+      "learning_rate": 4.615233191733398e-05,
+      "loss": 0.0004,
+      "step": 684
+    },
+    {
+      "epoch": 0.09691224843490255,
+      "grad_norm": 0.00046978742466308177,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 0.0,
+      "step": 685
+    },
+    {
+      "epoch": 0.09705372616984402,
+      "grad_norm": 0.0001826570078264922,
+      "learning_rate": 4.5623951722120736e-05,
+      "loss": 0.0,
+      "step": 686
+    },
+    {
+      "epoch": 0.09719520390478549,
+      "grad_norm": 0.0008781368378549814,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 0.0,
+      "step": 687
+    },
+    {
+      "epoch": 0.09733668163972695,
+      "grad_norm": 0.0002281236374983564,
+      "learning_rate": 4.5097718200186814e-05,
+      "loss": 0.0,
+      "step": 688
+    },
+    {
+      "epoch": 0.09747815937466842,
+      "grad_norm": 0.0003516919387038797,
+      "learning_rate": 4.483541293715698e-05,
+      "loss": 0.0,
+      "step": 689
+    },
+    {
+      "epoch": 0.09761963710960987,
+      "grad_norm": 0.002035178244113922,
+      "learning_rate": 4.457365212633058e-05,
+      "loss": 0.0001,
+      "step": 690
+    },
+    {
+      "epoch": 0.09776111484455134,
+      "grad_norm": 0.0006699900841340423,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 0.0,
+      "step": 691
+    },
+    {
+      "epoch": 0.0979025925794928,
+      "grad_norm": 0.0003746839938685298,
+      "learning_rate": 4.4051774189783315e-05,
+      "loss": 0.0,
+      "step": 692
+    },
+    {
+      "epoch": 0.09804407031443427,
+      "grad_norm": 0.0003929166414309293,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 0.0,
+      "step": 693
+    },
+    {
+      "epoch": 0.09818554804937572,
+      "grad_norm": 0.010091761127114296,
+      "learning_rate": 4.3532104993392306e-05,
+      "loss": 0.0003,
+      "step": 694
+    },
+    {
+      "epoch": 0.09832702578431719,
+      "grad_norm": 0.0012642780784517527,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 0.0001,
+      "step": 695
+    },
+    {
+      "epoch": 0.09846850351925866,
+      "grad_norm": 0.00036933113005943596,
+      "learning_rate": 4.301466505280762e-05,
+      "loss": 0.0,
+      "step": 696
+    },
+    {
+      "epoch": 0.09860998125420012,
+      "grad_norm": 0.0009099773596972227,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 0.0001,
+      "step": 697
+    },
+    {
+      "epoch": 0.09875145898914159,
+      "grad_norm": 0.0003582508652471006,
+      "learning_rate": 4.249947479567218e-05,
+      "loss": 0.0,
+      "step": 698
+    },
+    {
+      "epoch": 0.09889293672408304,
+      "grad_norm": 0.005193689372390509,
+      "learning_rate": 4.224272965777326e-05,
+      "loss": 0.0002,
+      "step": 699
+    },
+    {
+      "epoch": 0.09903441445902451,
+      "grad_norm": 0.0015010442584753036,
+      "learning_rate": 4.1986554560815096e-05,
+      "loss": 0.0001,
+      "step": 700
+    },
+    {
+      "epoch": 0.09917589219396597,
+      "grad_norm": 2.6988134384155273,
+      "learning_rate": 4.173095203314241e-05,
+      "loss": 0.8706,
+      "step": 701
+    },
+    {
+      "epoch": 0.09931736992890744,
+      "grad_norm": 0.0003478722064755857,
+      "learning_rate": 4.1475924597449024e-05,
+      "loss": 0.0,
+      "step": 702
+    },
+    {
+      "epoch": 0.0994588476638489,
+      "grad_norm": 0.00035000426578335464,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 0.0,
+      "step": 703
+    },
+    {
+      "epoch": 0.09960032539879037,
+      "grad_norm": 0.0004221213166601956,
+      "learning_rate": 4.096760506437057e-05,
+      "loss": 0.0,
+      "step": 704
+    },
+    {
+      "epoch": 0.09974180313373184,
+      "grad_norm": 0.0002811127051245421,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 0.0,
+      "step": 705
+    },
+    {
+      "epoch": 0.09988328086867329,
+      "grad_norm": 0.0014768473338335752,
+      "learning_rate": 4.0461616029164526e-05,
+      "loss": 0.0001,
+      "step": 706
+    },
+    {
+      "epoch": 0.10002475860361476,
+      "grad_norm": 0.0011548775946721435,
+      "learning_rate": 4.020950169424815e-05,
+      "loss": 0.0001,
+      "step": 707
+    },
+    {
+      "epoch": 0.10016623633855622,
+      "grad_norm": 0.002059435937553644,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 0.0002,
+      "step": 708
+    },
+    {
+      "epoch": 0.10030771407349769,
+      "grad_norm": 0.0005607010680250823,
+      "learning_rate": 3.9707045831097555e-05,
+      "loss": 0.0,
+      "step": 709
+    },
+    {
+      "epoch": 0.10044919180843914,
+      "grad_norm": 0.0003256005293224007,
+      "learning_rate": 3.945670926189987e-05,
+      "loss": 0.0,
+      "step": 710
+    },
+    {
+      "epoch": 0.10059066954338061,
+      "grad_norm": 0.009765557013452053,
+      "learning_rate": 3.920697023053949e-05,
+      "loss": 0.0002,
+      "step": 711
+    },
+    {
+      "epoch": 0.10073214727832207,
+      "grad_norm": 0.030024884268641472,
+      "learning_rate": 3.895783120183976e-05,
+      "loss": 0.0004,
+      "step": 712
+    },
+    {
+      "epoch": 0.10087362501326354,
+      "grad_norm": 0.0017325967783108354,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 0.0001,
+      "step": 713
+    },
+    {
+      "epoch": 0.10101510274820501,
+      "grad_norm": 0.0008862835238687694,
+      "learning_rate": 3.846136298208285e-05,
+      "loss": 0.0001,
+      "step": 714
+    },
+    {
+      "epoch": 0.10115658048314646,
+      "grad_norm": 0.0033906784374266863,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 0.0003,
+      "step": 715
+    },
+    {
+      "epoch": 0.10129805821808793,
+      "grad_norm": 0.0006109073874540627,
+      "learning_rate": 3.796732420234443e-05,
+      "loss": 0.0,
+      "step": 716
+    },
+    {
+      "epoch": 0.10143953595302939,
+      "grad_norm": 0.008680574595928192,
+      "learning_rate": 3.7721221951188765e-05,
+      "loss": 0.0002,
+      "step": 717
+    },
+    {
+      "epoch": 0.10158101368797086,
+      "grad_norm": 0.0005801686784252524,
+      "learning_rate": 3.747573436642951e-05,
+      "loss": 0.0001,
+      "step": 718
+    },
+    {
+      "epoch": 0.10172249142291231,
+      "grad_norm": 0.00133225554600358,
+      "learning_rate": 3.7230863870929964e-05,
+      "loss": 0.0001,
+      "step": 719
+    },
+    {
+      "epoch": 0.10186396915785378,
+      "grad_norm": 0.0013388440711423755,
+      "learning_rate": 3.698661288146311e-05,
+      "loss": 0.0001,
+      "step": 720
+    },
+    {
+      "epoch": 0.10200544689279524,
+      "grad_norm": 0.002170740393921733,
+      "learning_rate": 3.674298380868756e-05,
+      "loss": 0.0002,
+      "step": 721
+    },
+    {
+      "epoch": 0.10214692462773671,
+      "grad_norm": 0.0010131692979484797,
+      "learning_rate": 3.649997905712396e-05,
+      "loss": 0.0001,
+      "step": 722
+    },
+    {
+      "epoch": 0.10228840236267818,
+      "grad_norm": 0.0006320245447568595,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 0.0001,
+      "step": 723
+    },
+    {
+      "epoch": 0.10242988009761964,
+      "grad_norm": 0.0005717655294574797,
+      "learning_rate": 3.601585210488218e-05,
+      "loss": 0.0001,
+      "step": 724
+    },
+    {
+      "epoch": 0.1025713578325611,
+      "grad_norm": 0.0009512366959825158,
+      "learning_rate": 3.577473468234156e-05,
+      "loss": 0.0001,
+      "step": 725
+    },
+    {
+      "epoch": 0.10271283556750256,
+      "grad_norm": 0.0010945550166070461,
+      "learning_rate": 3.553425113724088e-05,
+      "loss": 0.0001,
+      "step": 726
+    },
+    {
+      "epoch": 0.10285431330244403,
+      "grad_norm": 0.0025489837862551212,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 0.0001,
+      "step": 727
+    },
+    {
+      "epoch": 0.10299579103738549,
+      "grad_norm": 0.016141386702656746,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 0.0013,
+      "step": 728
+    },
+    {
+      "epoch": 0.10313726877232696,
+      "grad_norm": 0.0008998136036098003,
+      "learning_rate": 3.481662746991214e-05,
+      "loss": 0.0001,
+      "step": 729
+    },
+    {
+      "epoch": 0.10327874650726841,
+      "grad_norm": 0.007485687267035246,
+      "learning_rate": 3.4578703106413904e-05,
+      "loss": 0.0006,
+      "step": 730
+    },
+    {
+      "epoch": 0.10342022424220988,
+      "grad_norm": 0.0015185687225311995,
+      "learning_rate": 3.4341424424704375e-05,
+      "loss": 0.0001,
+      "step": 731
+    },
+    {
+      "epoch": 0.10356170197715135,
+      "grad_norm": 0.0037518057506531477,
+      "learning_rate": 3.4104793766628304e-05,
+      "loss": 0.0002,
+      "step": 732
+    },
+    {
+      "epoch": 0.10370317971209281,
+      "grad_norm": 0.002752428874373436,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 0.0001,
+      "step": 733
+    },
+    {
+      "epoch": 0.10384465744703428,
+      "grad_norm": 0.007001630496233702,
+      "learning_rate": 3.363348585675414e-05,
+      "loss": 0.0001,
+      "step": 734
+    },
+    {
+      "epoch": 0.10398613518197573,
+      "grad_norm": 0.005060289520770311,
+      "learning_rate": 3.339881325657484e-05,
+      "loss": 0.0003,
+      "step": 735
+    },
+    {
+      "epoch": 0.1041276129169172,
+      "grad_norm": 0.003094696905463934,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 0.0001,
+      "step": 736
+    },
+    {
+      "epoch": 0.10426909065185866,
+      "grad_norm": 0.0031883700285106897,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 0.0002,
+      "step": 737
+    },
+    {
+      "epoch": 0.10441056838680013,
+      "grad_norm": 0.0017164596356451511,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 0.0001,
+      "step": 738
+    },
+    {
+      "epoch": 0.10455204612174158,
+      "grad_norm": 0.0011868190485984087,
+      "learning_rate": 3.246671918789755e-05,
+      "loss": 0.0001,
+      "step": 739
+    },
+    {
+      "epoch": 0.10469352385668305,
+      "grad_norm": 0.00042919107363559306,
+      "learning_rate": 3.223535625298979e-05,
+      "loss": 0.0,
+      "step": 740
+    },
+    {
+      "epoch": 0.10483500159162452,
+      "grad_norm": 0.0008362437947653234,
+      "learning_rate": 3.200466212775808e-05,
+      "loss": 0.0001,
+      "step": 741
+    },
+    {
+      "epoch": 0.10497647932656598,
+      "grad_norm": 0.0012623334769159555,
+      "learning_rate": 3.1774639089060363e-05,
+      "loss": 0.0001,
+      "step": 742
+    },
+    {
+      "epoch": 0.10511795706150745,
+      "grad_norm": 0.0010986161651089787,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 0.0001,
+      "step": 743
+    },
+    {
+      "epoch": 0.1052594347964489,
+      "grad_norm": 0.0006085491622798145,
+      "learning_rate": 3.1316615345559185e-05,
+      "loss": 0.0001,
+      "step": 744
+    },
+    {
+      "epoch": 0.10540091253139038,
+      "grad_norm": 0.0008765387465246022,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 0.0001,
+      "step": 745
+    },
+    {
+      "epoch": 0.10554239026633183,
+      "grad_norm": 0.0006421508733183146,
+      "learning_rate": 3.086130310447937e-05,
+      "loss": 0.0001,
+      "step": 746
+    },
+    {
+      "epoch": 0.1056838680012733,
+      "grad_norm": 0.0016310947248712182,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 0.0001,
+      "step": 747
+    },
+    {
+      "epoch": 0.10582534573621476,
+      "grad_norm": 0.0014764780644327402,
+      "learning_rate": 3.0408720340768572e-05,
+      "loss": 0.0001,
+      "step": 748
+    },
+    {
+      "epoch": 0.10596682347115623,
+      "grad_norm": 0.0006910095689818263,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 0.0001,
+      "step": 749
+    },
+    {
+      "epoch": 0.1061083012060977,
+      "grad_norm": 0.0015598449390381575,
+      "learning_rate": 2.9958884921619367e-05,
+      "loss": 0.0001,
+      "step": 750
+    },
+    {
+      "epoch": 0.10624977894103915,
+      "grad_norm": 0.0007486201939173043,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 0.0001,
+      "step": 751
+    },
+    {
+      "epoch": 0.10639125667598062,
+      "grad_norm": 0.005332762375473976,
+      "learning_rate": 2.9511814605763855e-05,
+      "loss": 0.0005,
+      "step": 752
+    },
+    {
+      "epoch": 0.10653273441092208,
+      "grad_norm": 0.04735836386680603,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.0004,
+      "step": 753
+    },
+    {
+      "epoch": 0.10667421214586355,
+      "grad_norm": 0.004297490697354078,
+      "learning_rate": 2.9067527042772636e-05,
+      "loss": 0.0004,
+      "step": 754
+    },
+    {
+      "epoch": 0.106815689880805,
+      "grad_norm": 0.0011461108224466443,
+      "learning_rate": 2.8846432279071467e-05,
+      "loss": 0.0001,
+      "step": 755
+    },
+    {
+      "epoch": 0.10695716761574647,
+      "grad_norm": 0.0008430193993262947,
+      "learning_rate": 2.8626039772357882e-05,
+      "loss": 0.0001,
+      "step": 756
+    },
+    {
+      "epoch": 0.10709864535068793,
+      "grad_norm": 0.0005710861296392977,
+      "learning_rate": 2.840635169781688e-05,
+      "loss": 0.0,
+      "step": 757
+    },
+    {
+      "epoch": 0.1072401230856294,
+      "grad_norm": 0.0006396773969754577,
+      "learning_rate": 2.8187370223681132e-05,
+      "loss": 0.0001,
+      "step": 758
+    },
+    {
+      "epoch": 0.10738160082057087,
+      "grad_norm": 0.0046224696561694145,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 0.0001,
+      "step": 759
+    },
+    {
+      "epoch": 0.10752307855551232,
+      "grad_norm": 0.003434223821386695,
+      "learning_rate": 2.775153571466502e-05,
+      "loss": 0.0002,
+      "step": 760
+    },
+    {
+      "epoch": 0.1076645562904538,
+      "grad_norm": 0.005056839436292648,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 0.0002,
+      "step": 761
+    },
+    {
+      "epoch": 0.10780603402539525,
+      "grad_norm": 0.0009986249497160316,
+      "learning_rate": 2.7318553451309726e-05,
+      "loss": 0.0001,
+      "step": 762
+    },
+    {
+      "epoch": 0.10794751176033672,
+      "grad_norm": 0.008566567674279213,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 0.0001,
+      "step": 763
+    },
+    {
+      "epoch": 0.10808898949527818,
+      "grad_norm": 0.001387615455314517,
+      "learning_rate": 2.688844052701359e-05,
+      "loss": 0.0001,
+      "step": 764
+    },
+    {
+      "epoch": 0.10823046723021965,
+      "grad_norm": 0.0008968693437054753,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 0.0001,
+      "step": 765
+    },
+    {
+      "epoch": 0.1083719449651611,
+      "grad_norm": 0.0020030024461448193,
+      "learning_rate": 2.646121392189841e-05,
+      "loss": 0.0001,
+      "step": 766
+    },
+    {
+      "epoch": 0.10851342270010257,
+      "grad_norm": 0.0008027153671719134,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 0.0,
+      "step": 767
+    },
+    {
+      "epoch": 0.10865490043504404,
+      "grad_norm": 0.000605833251029253,
+      "learning_rate": 2.603689050213902e-05,
+      "loss": 0.0001,
+      "step": 768
+    },
+    {
+      "epoch": 0.1087963781699855,
+      "grad_norm": 0.002874531550332904,
+      "learning_rate": 2.582582272612609e-05,
+      "loss": 0.0002,
+      "step": 769
+    },
+    {
+      "epoch": 0.10893785590492697,
+      "grad_norm": 0.0006501192692667246,
+      "learning_rate": 2.561548701929749e-05,
+      "loss": 0.0001,
+      "step": 770
+    },
+    {
+      "epoch": 0.10907933363986842,
+      "grad_norm": 0.000950626446865499,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 0.0001,
+      "step": 771
+    },
+    {
+      "epoch": 0.10922081137480989,
+      "grad_norm": 0.0013473192229866982,
+      "learning_rate": 2.5197020109661772e-05,
+      "loss": 0.0001,
+      "step": 772
+    },
+    {
+      "epoch": 0.10936228910975135,
+      "grad_norm": 0.0013363496400415897,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 0.0001,
+      "step": 773
+    },
+    {
+      "epoch": 0.10950376684469282,
+      "grad_norm": 0.0068521020002663136,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 0.0004,
+      "step": 774
+    },
+    {
+      "epoch": 0.10964524457963427,
+      "grad_norm": 0.0016451867995783687,
+      "learning_rate": 2.4574861926389615e-05,
+      "loss": 0.0001,
+      "step": 775
+    },
+    {
+      "epoch": 0.10978672231457574,
+      "grad_norm": 0.00135321996640414,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 0.0001,
+      "step": 776
+    },
+    {
+      "epoch": 0.10992820004951721,
+      "grad_norm": 0.0008520256378687918,
+      "learning_rate": 2.4163808471127812e-05,
+      "loss": 0.0001,
+      "step": 777
+    },
+    {
+      "epoch": 0.11006967778445867,
+      "grad_norm": 0.0019374149851500988,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 0.0002,
+      "step": 778
+    },
+    {
+      "epoch": 0.11021115551940014,
+      "grad_norm": 0.0019093346782028675,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 0.0002,
+      "step": 779
+    },
+    {
+      "epoch": 0.1103526332543416,
+      "grad_norm": 0.0007253644871525466,
+      "learning_rate": 2.3552846857690846e-05,
+      "loss": 0.0001,
+      "step": 780
+    },
+    {
+      "epoch": 0.11049411098928306,
+      "grad_norm": 0.0013740455033257604,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 0.0001,
+      "step": 781
+    },
+    {
+      "epoch": 0.11063558872422452,
+      "grad_norm": 0.0012033848324790597,
+      "learning_rate": 2.3149308278092342e-05,
+      "loss": 0.0001,
+      "step": 782
+    },
+    {
+      "epoch": 0.11077706645916599,
+      "grad_norm": 0.001106973155401647,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 0.0001,
+      "step": 783
+    },
+    {
+      "epoch": 0.11091854419410745,
+      "grad_norm": 0.00048042790149338543,
+      "learning_rate": 2.2748803632213557e-05,
+      "loss": 0.0,
+      "step": 784
+    },
+    {
+      "epoch": 0.11106002192904892,
+      "grad_norm": 0.001044884673319757,
+      "learning_rate": 2.254969398012663e-05,
+      "loss": 0.0001,
+      "step": 785
+    },
+    {
+      "epoch": 0.11120149966399039,
+      "grad_norm": 0.0007944092503748834,
+      "learning_rate": 2.235134873129213e-05,
+      "loss": 0.0001,
+      "step": 786
+    },
+    {
+      "epoch": 0.11134297739893184,
+      "grad_norm": 0.0007932492881081998,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 0.0001,
+      "step": 787
+    },
+    {
+      "epoch": 0.11148445513387331,
+      "grad_norm": 0.001373193459585309,
+      "learning_rate": 2.195695926616702e-05,
+      "loss": 0.0001,
+      "step": 788
+    },
+    {
+      "epoch": 0.11162593286881477,
+      "grad_norm": 0.006605231203138828,
+      "learning_rate": 2.1760918942341192e-05,
+      "loss": 0.0006,
+      "step": 789
+    },
+    {
+      "epoch": 0.11176741060375624,
+      "grad_norm": 0.0010321062291041017,
+      "learning_rate": 2.1565650806658975e-05,
+      "loss": 0.0001,
+      "step": 790
+    },
+    {
+      "epoch": 0.11190888833869769,
+      "grad_norm": 0.20049892365932465,
+      "learning_rate": 2.137115678633811e-05,
+      "loss": 0.0012,
+      "step": 791
+    },
+    {
+      "epoch": 0.11205036607363916,
+      "grad_norm": 0.003096005879342556,
+      "learning_rate": 2.1177438800956007e-05,
+      "loss": 0.0002,
+      "step": 792
+    },
+    {
+      "epoch": 0.11219184380858062,
+      "grad_norm": 0.0008285019430331886,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 0.0001,
+      "step": 793
+    },
+    {
+      "epoch": 0.11233332154352209,
+      "grad_norm": 0.003056522458791733,
+      "learning_rate": 2.07923385750033e-05,
+      "loss": 0.0002,
+      "step": 794
+    },
+    {
+      "epoch": 0.11247479927846356,
+      "grad_norm": 0.0010699566919356585,
+      "learning_rate": 2.0600960135216462e-05,
+      "loss": 0.0001,
+      "step": 795
+    },
+    {
+      "epoch": 0.11261627701340501,
+      "grad_norm": 0.003565857419744134,
+      "learning_rate": 2.0410365331898416e-05,
+      "loss": 0.0002,
+      "step": 796
+    },
+    {
+      "epoch": 0.11275775474834648,
+      "grad_norm": 0.000508810393512249,
+      "learning_rate": 2.0220556046142893e-05,
+      "loss": 0.0,
+      "step": 797
+    },
+    {
+      "epoch": 0.11289923248328794,
+      "grad_norm": 0.00033257991890423,
+      "learning_rate": 2.0031534151290943e-05,
+      "loss": 0.0,
+      "step": 798
+    },
+    {
+      "epoch": 0.11304071021822941,
+      "grad_norm": 0.0007687319885008037,
+      "learning_rate": 1.9843301512912327e-05,
+      "loss": 0.0001,
+      "step": 799
+    },
+    {
+      "epoch": 0.11318218795317087,
+      "grad_norm": 0.004830488003790379,
+      "learning_rate": 1.965585998878724e-05,
+      "loss": 0.0004,
+      "step": 800
+    },
+    {
+      "epoch": 0.11332366568811233,
+      "grad_norm": 0.0005077541572973132,
+      "learning_rate": 1.946921142888781e-05,
+      "loss": 0.0,
+      "step": 801
+    },
+    {
+      "epoch": 0.11346514342305379,
+      "grad_norm": 0.004109581001102924,
+      "learning_rate": 1.928335767535997e-05,
+      "loss": 0.0003,
+      "step": 802
+    },
+    {
+      "epoch": 0.11360662115799526,
+      "grad_norm": 0.0029293650295585394,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.0002,
+      "step": 803
+    },
+    {
+      "epoch": 0.11374809889293673,
+      "grad_norm": 0.0005266869557090104,
+      "learning_rate": 1.891404191676265e-05,
+      "loss": 0.0,
+      "step": 804
+    },
+    {
+      "epoch": 0.11388957662787819,
+      "grad_norm": 0.0005665577482432127,
+      "learning_rate": 1.8730583556690605e-05,
+      "loss": 0.0001,
+      "step": 805
+    },
+    {
+      "epoch": 0.11403105436281966,
+      "grad_norm": 0.0034227666910737753,
+      "learning_rate": 1.854792729294905e-05,
+      "loss": 0.0001,
+      "step": 806
+    },
+    {
+      "epoch": 0.11417253209776111,
+      "grad_norm": 0.000532831996679306,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 0.0,
+      "step": 807
+    },
+    {
+      "epoch": 0.11431400983270258,
+      "grad_norm": 0.0009800855768844485,
+      "learning_rate": 1.818502825749764e-05,
+      "loss": 0.0001,
+      "step": 808
+    },
+    {
+      "epoch": 0.11445548756764404,
+      "grad_norm": 0.0037957008462399244,
+      "learning_rate": 1.8004789067454764e-05,
+      "loss": 0.0003,
+      "step": 809
+    },
+    {
+      "epoch": 0.11459696530258551,
+      "grad_norm": 0.0018789039459079504,
+      "learning_rate": 1.7825359137040988e-05,
+      "loss": 0.0002,
+      "step": 810
+    },
+    {
+      "epoch": 0.11473844303752696,
+      "grad_norm": 0.0015523998299613595,
+      "learning_rate": 1.7646740237157256e-05,
+      "loss": 0.0001,
+      "step": 811
+    },
+    {
+      "epoch": 0.11487992077246843,
+      "grad_norm": 0.000724561046808958,
+      "learning_rate": 1.7468934130700044e-05,
+      "loss": 0.0001,
+      "step": 812
+    },
+    {
+      "epoch": 0.1150213985074099,
+      "grad_norm": 0.003468384500592947,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 0.0002,
+      "step": 813
+    },
+    {
+      "epoch": 0.11516287624235136,
+      "grad_norm": 0.0004411648551467806,
+      "learning_rate": 1.7115767309523812e-05,
+      "loss": 0.0,
+      "step": 814
+    },
+    {
+      "epoch": 0.11530435397729283,
+      "grad_norm": 0.003343729767948389,
+      "learning_rate": 1.6940410080418723e-05,
+      "loss": 0.0001,
+      "step": 815
+    },
+    {
+      "epoch": 0.11544583171223428,
+      "grad_norm": 1.5019727945327759,
+      "learning_rate": 1.6765872615933677e-05,
+      "loss": 0.4939,
+      "step": 816
+    },
+    {
+      "epoch": 0.11558730944717575,
+      "grad_norm": 0.0021441394928842783,
+      "learning_rate": 1.6592156638682886e-05,
+      "loss": 0.0002,
+      "step": 817
+    },
+    {
+      "epoch": 0.11572878718211721,
+      "grad_norm": 0.0018722140230238438,
+      "learning_rate": 1.6419263863172997e-05,
+      "loss": 0.0001,
+      "step": 818
+    },
+    {
+      "epoch": 0.11587026491705868,
+      "grad_norm": 0.000854533922392875,
+      "learning_rate": 1.6247195995785837e-05,
+      "loss": 0.0001,
+      "step": 819
+    },
+    {
+      "epoch": 0.11601174265200014,
+      "grad_norm": 0.0007223361171782017,
+      "learning_rate": 1.6075954734761845e-05,
+      "loss": 0.0001,
+      "step": 820
+    },
+    {
+      "epoch": 0.1161532203869416,
+      "grad_norm": 0.0026152019854635,
+      "learning_rate": 1.5905541770183096e-05,
+      "loss": 0.0002,
+      "step": 821
+    },
+    {
+      "epoch": 0.11629469812188307,
+      "grad_norm": 0.001397658372297883,
+      "learning_rate": 1.5735958783956794e-05,
+      "loss": 0.0001,
+      "step": 822
+    },
+    {
+      "epoch": 0.11643617585682453,
+      "grad_norm": 0.0011791229480877519,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 0.0001,
+      "step": 823
+    },
+    {
+      "epoch": 0.116577653591766,
+      "grad_norm": 0.0016374592669308186,
+      "learning_rate": 1.539928943321579e-05,
+      "loss": 0.0001,
+      "step": 824
+    },
+    {
+      "epoch": 0.11671913132670746,
+      "grad_norm": 0.002598709659650922,
+      "learning_rate": 1.5232206391491699e-05,
+      "loss": 0.0002,
+      "step": 825
+    },
+    {
+      "epoch": 0.11686060906164893,
+      "grad_norm": 0.0026453689206391573,
+      "learning_rate": 1.5065959973668353e-05,
+      "loss": 0.0002,
+      "step": 826
+    },
+    {
+      "epoch": 0.11700208679659038,
+      "grad_norm": 0.0007108437712304294,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 0.0001,
+      "step": 827
+    },
+    {
+      "epoch": 0.11714356453153185,
+      "grad_norm": 0.0016297996044158936,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 0.0001,
+      "step": 828
+    },
+    {
+      "epoch": 0.11728504226647331,
+      "grad_norm": 0.0010631303302943707,
+      "learning_rate": 1.4572256830070497e-05,
+      "loss": 0.0001,
+      "step": 829
+    },
+    {
+      "epoch": 0.11742652000141478,
+      "grad_norm": 0.041488487273454666,
+      "learning_rate": 1.4409373232886702e-05,
+      "loss": 0.0006,
+      "step": 830
+    },
+    {
+      "epoch": 0.11756799773635625,
+      "grad_norm": 0.002247254131361842,
+      "learning_rate": 1.4247334380634792e-05,
+      "loss": 0.0001,
+      "step": 831
+    },
+    {
+      "epoch": 0.1177094754712977,
+      "grad_norm": 0.0015282740350812674,
+      "learning_rate": 1.4086141872572789e-05,
+      "loss": 0.0001,
+      "step": 832
+    },
+    {
+      "epoch": 0.11785095320623917,
+      "grad_norm": 0.0033260490745306015,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 0.0002,
+      "step": 833
+    },
+    {
+      "epoch": 0.11799243094118063,
+      "grad_norm": 0.00047575324424542487,
+      "learning_rate": 1.3766302244269624e-05,
+      "loss": 0.0,
+      "step": 834
+    },
+    {
+      "epoch": 0.1181339086761221,
+      "grad_norm": 1.9940252304077148,
+      "learning_rate": 1.3607658280716473e-05,
+      "loss": 0.1164,
+      "step": 835
+    },
+    {
+      "epoch": 0.11827538641106355,
+      "grad_norm": 0.013329320587217808,
+      "learning_rate": 1.3449866974698122e-05,
+      "loss": 0.0007,
+      "step": 836
+    },
+    {
+      "epoch": 0.11841686414600502,
+      "grad_norm": 0.009485892951488495,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 0.0007,
+      "step": 837
+    },
+    {
+      "epoch": 0.11855834188094648,
+      "grad_norm": 0.004114389885216951,
+      "learning_rate": 1.3136848556180892e-05,
+      "loss": 0.0004,
+      "step": 838
+    },
+    {
+      "epoch": 0.11869981961588795,
+      "grad_norm": 0.008051803335547447,
+      "learning_rate": 1.2981624533047432e-05,
+      "loss": 0.0002,
+      "step": 839
+    },
+    {
+      "epoch": 0.11884129735082942,
+      "grad_norm": 0.00846549030393362,
+      "learning_rate": 1.2827259346149122e-05,
+      "loss": 0.0005,
+      "step": 840
+    },
+    {
+      "epoch": 0.11898277508577088,
+      "grad_norm": 0.004868239630013704,
+      "learning_rate": 1.2673754519008008e-05,
+      "loss": 0.0004,
+      "step": 841
+    },
+    {
+      "epoch": 0.11912425282071235,
+      "grad_norm": 0.00196613441221416,
+      "learning_rate": 1.2521111566654731e-05,
+      "loss": 0.0001,
+      "step": 842
+    },
+    {
+      "epoch": 0.1192657305556538,
+      "grad_norm": 0.0007714144885540009,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 0.0001,
+      "step": 843
+    },
+    {
+      "epoch": 0.11940720829059527,
+      "grad_norm": 0.0020139142870903015,
+      "learning_rate": 1.2218417303887842e-05,
+      "loss": 0.0002,
+      "step": 844
+    },
+    {
+      "epoch": 0.11954868602553673,
+      "grad_norm": 0.001310071093030274,
+      "learning_rate": 1.206836898094439e-05,
+      "loss": 0.0001,
+      "step": 845
+    },
+    {
+      "epoch": 0.1196901637604782,
+      "grad_norm": 0.00624494906514883,
+      "learning_rate": 1.191918850769964e-05,
+      "loss": 0.0002,
+      "step": 846
+    },
+    {
+      "epoch": 0.11983164149541965,
+      "grad_norm": 0.0040423693135380745,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 0.0002,
+      "step": 847
+    },
+    {
+      "epoch": 0.11997311923036112,
+      "grad_norm": 0.01039808988571167,
+      "learning_rate": 1.1623436991130654e-05,
+      "loss": 0.0007,
+      "step": 848
+    },
+    {
+      "epoch": 0.12011459696530259,
+      "grad_norm": 0.005628874525427818,
+      "learning_rate": 1.1476868866754486e-05,
+      "loss": 0.0004,
+      "step": 849
+    },
+    {
+      "epoch": 0.12025607470024405,
+      "grad_norm": 0.011199036613106728,
+      "learning_rate": 1.1331174429944347e-05,
+      "loss": 0.0007,
+      "step": 850
+    },
+    {
+      "epoch": 0.12039755243518552,
+      "grad_norm": 0.00257562636397779,
+      "learning_rate": 1.1186355118645554e-05,
+      "loss": 0.0001,
+      "step": 851
+    },
+    {
+      "epoch": 0.12053903017012697,
+      "grad_norm": 0.001868070219643414,
+      "learning_rate": 1.1042412362166222e-05,
+      "loss": 0.0001,
+      "step": 852
+    },
+    {
+      "epoch": 0.12068050790506844,
+      "grad_norm": 0.003299275878816843,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 0.0003,
+      "step": 853
+    },
+    {
+      "epoch": 0.1208219856400099,
+      "grad_norm": 0.0014683003537356853,
+      "learning_rate": 1.0757162187628222e-05,
+      "loss": 0.0001,
+      "step": 854
+    },
+    {
+      "epoch": 0.12096346337495137,
+      "grad_norm": 0.004328795243054628,
+      "learning_rate": 1.0615857584873623e-05,
+      "loss": 0.0004,
+      "step": 855
+    },
+    {
+      "epoch": 0.12110494110989282,
+      "grad_norm": 0.002547298092395067,
+      "learning_rate": 1.0475435167518843e-05,
+      "loss": 0.0002,
+      "step": 856
+    },
+    {
+      "epoch": 0.1212464188448343,
+      "grad_norm": 0.0016848307568579912,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 0.0001,
+      "step": 857
+    },
+    {
+      "epoch": 0.12138789657977576,
+      "grad_norm": 0.004747582133859396,
+      "learning_rate": 1.0197242423938446e-05,
+      "loss": 0.0005,
+      "step": 858
+    },
+    {
+      "epoch": 0.12152937431471722,
+      "grad_norm": 0.001700869295746088,
+      "learning_rate": 1.0059474843362892e-05,
+      "loss": 0.0001,
+      "step": 859
+    },
+    {
+      "epoch": 0.12167085204965869,
+      "grad_norm": 0.0012192331487312913,
+      "learning_rate": 9.922594939460194e-06,
+      "loss": 0.0001,
+      "step": 860
+    },
+    {
+      "epoch": 0.12181232978460015,
+      "grad_norm": 0.003007133025676012,
+      "learning_rate": 9.786604063179728e-06,
+      "loss": 0.0002,
+      "step": 861
+    },
+    {
+      "epoch": 0.12195380751954162,
+      "grad_norm": 0.001699776970781386,
+      "learning_rate": 9.651503556696516e-06,
+      "loss": 0.0001,
+      "step": 862
+    },
+    {
+      "epoch": 0.12209528525448307,
+      "grad_norm": 0.0022404356859624386,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 0.0002,
+      "step": 863
+    },
+    {
+      "epoch": 0.12223676298942454,
+      "grad_norm": 0.003892921144142747,
+      "learning_rate": 9.383978977871021e-06,
+      "loss": 0.0002,
+      "step": 864
+    },
+    {
+      "epoch": 0.122378240724366,
+      "grad_norm": 0.0008568639750592411,
+      "learning_rate": 9.251557545888312e-06,
+      "loss": 0.0001,
+      "step": 865
+    },
+    {
+      "epoch": 0.12251971845930747,
+      "grad_norm": 0.001413036254234612,
+      "learning_rate": 9.120031764395987e-06,
+      "loss": 0.0001,
+      "step": 866
+    },
+    {
+      "epoch": 0.12266119619424894,
+      "grad_norm": 0.0006800137925893068,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 0.0001,
+      "step": 867
+    },
+    {
+      "epoch": 0.12280267392919039,
+      "grad_norm": 0.009284158237278461,
+      "learning_rate": 8.85967233645547e-06,
+      "loss": 0.0004,
+      "step": 868
+    },
+    {
+      "epoch": 0.12294415166413186,
+      "grad_norm": 0.002639732789248228,
+      "learning_rate": 8.730841259649725e-06,
+      "loss": 0.0002,
+      "step": 869
+    },
+    {
+      "epoch": 0.12308562939907332,
+      "grad_norm": 0.009310910478234291,
+      "learning_rate": 8.602910972593892e-06,
+      "loss": 0.0007,
+      "step": 870
+    },
+    {
+      "epoch": 0.12322710713401479,
+      "grad_norm": 0.001654073246754706,
+      "learning_rate": 8.475882737908248e-06,
+      "loss": 0.0001,
+      "step": 871
+    },
+    {
+      "epoch": 0.12336858486895624,
+      "grad_norm": 0.007454808335751295,
+      "learning_rate": 8.34975780931021e-06,
+      "loss": 0.0001,
+      "step": 872
+    },
+    {
+      "epoch": 0.12351006260389771,
+      "grad_norm": 0.002037209691479802,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 0.0002,
+      "step": 873
+    },
+    {
+      "epoch": 0.12365154033883917,
+      "grad_norm": 0.003405377734452486,
+      "learning_rate": 8.100222840657878e-06,
+      "loss": 0.0002,
+      "step": 874
+    },
+    {
+      "epoch": 0.12379301807378064,
+      "grad_norm": 0.002389519941061735,
+      "learning_rate": 7.976815263412963e-06,
+      "loss": 0.0002,
+      "step": 875
+    },
+    {
+      "epoch": 0.12393449580872211,
+      "grad_norm": 0.000885749002918601,
+      "learning_rate": 7.854315917850163e-06,
+      "loss": 0.0001,
+      "step": 876
+    },
+    {
+      "epoch": 0.12407597354366356,
+      "grad_norm": 0.0019240875262767076,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 0.0002,
+      "step": 877
+    },
+    {
+      "epoch": 0.12421745127860503,
+      "grad_norm": 0.001213736366480589,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.0001,
+      "step": 878
+    },
+    {
+      "epoch": 0.12435892901354649,
+      "grad_norm": 0.0018200764898210764,
+      "learning_rate": 7.492279316554207e-06,
+      "loss": 0.0002,
+      "step": 879
+    },
+    {
+      "epoch": 0.12450040674848796,
+      "grad_norm": 0.048911500722169876,
+      "learning_rate": 7.3734248980933395e-06,
+      "loss": 0.0009,
+      "step": 880
+    },
+    {
+      "epoch": 0.12464188448342942,
+      "grad_norm": 0.003094620769843459,
+      "learning_rate": 7.255484666533874e-06,
+      "loss": 0.0001,
+      "step": 881
+    },
+    {
+      "epoch": 0.12478336221837089,
+      "grad_norm": 0.02175629325211048,
+      "learning_rate": 7.138459785898266e-06,
+      "loss": 0.0013,
+      "step": 882
+    },
+    {
+      "epoch": 0.12492483995331234,
+      "grad_norm": 0.007283073849976063,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 0.0002,
+      "step": 883
+    },
+    {
+      "epoch": 0.1250663176882538,
+      "grad_norm": 0.0008444397244602442,
+      "learning_rate": 6.907160688306425e-06,
+      "loss": 0.0001,
+      "step": 884
+    },
+    {
+      "epoch": 0.12520779542319527,
+      "grad_norm": 0.0045507000759243965,
+      "learning_rate": 6.7928887541789055e-06,
+      "loss": 0.0003,
+      "step": 885
+    },
+    {
+      "epoch": 0.12534927315813674,
+      "grad_norm": 0.006236658431589603,
+      "learning_rate": 6.679536736610137e-06,
+      "loss": 0.0005,
+      "step": 886
+    },
+    {
+      "epoch": 0.1254907508930782,
+      "grad_norm": 0.0028792882803827524,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 0.0001,
+      "step": 887
+    },
+    {
+      "epoch": 0.12563222862801968,
+      "grad_norm": 0.004536019172519445,
+      "learning_rate": 6.455596917013273e-06,
+      "loss": 0.0002,
+      "step": 888
+    },
+    {
+      "epoch": 0.12577370636296112,
+      "grad_norm": 0.0009695796761661768,
+      "learning_rate": 6.345011325180772e-06,
+      "loss": 0.0001,
+      "step": 889
+    },
+    {
+      "epoch": 0.1259151840979026,
+      "grad_norm": 0.00836182851344347,
+      "learning_rate": 6.235350070276447e-06,
+      "loss": 0.0004,
+      "step": 890
+    },
+    {
+      "epoch": 0.12605666183284406,
+      "grad_norm": 0.0012931537348777056,
+      "learning_rate": 6.126614234612593e-06,
+      "loss": 0.0001,
+      "step": 891
+    },
+    {
+      "epoch": 0.12619813956778553,
+      "grad_norm": 0.004833157639950514,
+      "learning_rate": 6.018804891368035e-06,
+      "loss": 0.0004,
+      "step": 892
+    },
+    {
+      "epoch": 0.12633961730272697,
+      "grad_norm": 0.0011749600525945425,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 0.0001,
+      "step": 893
+    },
+    {
+      "epoch": 0.12648109503766844,
+      "grad_norm": 0.0016763556050136685,
+      "learning_rate": 5.805969929120947e-06,
+      "loss": 0.0001,
+      "step": 894
+    },
+    {
+      "epoch": 0.1266225727726099,
+      "grad_norm": 0.003065883880481124,
+      "learning_rate": 5.700946410713548e-06,
+      "loss": 0.0002,
+      "step": 895
+    },
+    {
+      "epoch": 0.12676405050755138,
+      "grad_norm": 0.015194329433143139,
+      "learning_rate": 5.5968535858950345e-06,
+      "loss": 0.0004,
+      "step": 896
+    },
+    {
+      "epoch": 0.12690552824249285,
+      "grad_norm": 0.007229126524180174,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 0.0006,
+      "step": 897
+    },
+    {
+      "epoch": 0.1270470059774343,
+      "grad_norm": 0.026731327176094055,
+      "learning_rate": 5.39146411724547e-06,
+      "loss": 0.0005,
+      "step": 898
+    },
+    {
+      "epoch": 0.12718848371237576,
+      "grad_norm": 0.001329147256910801,
+      "learning_rate": 5.290169500525577e-06,
+      "loss": 0.0001,
+      "step": 899
+    },
+    {
+      "epoch": 0.12732996144731723,
+      "grad_norm": 0.016030043363571167,
+      "learning_rate": 5.189809631596798e-06,
+      "loss": 0.0008,
+      "step": 900
+    },
+    {
+      "epoch": 0.1274714391822587,
+      "grad_norm": 0.001103358343243599,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 0.0001,
+      "step": 901
+    },
+    {
+      "epoch": 0.12761291691720014,
+      "grad_norm": 0.0012724302941933274,
+      "learning_rate": 4.991898089922819e-06,
+      "loss": 0.0001,
+      "step": 902
+    },
+    {
+      "epoch": 0.1277543946521416,
+      "grad_norm": 0.0033973059616982937,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 0.0003,
+      "step": 903
+    },
+    {
+      "epoch": 0.12789587238708308,
+      "grad_norm": 0.01135913748294115,
+      "learning_rate": 4.797737305432337e-06,
+      "loss": 0.0006,
+      "step": 904
+    },
+    {
+      "epoch": 0.12803735012202455,
+      "grad_norm": 0.0018166914815083146,
+      "learning_rate": 4.702065848278126e-06,
+      "loss": 0.0002,
+      "step": 905
+    },
+    {
+      "epoch": 0.12817882785696602,
+      "grad_norm": 0.0008935577352531254,
+      "learning_rate": 4.607334943260655e-06,
+      "loss": 0.0001,
+      "step": 906
+    },
+    {
+      "epoch": 0.12832030559190746,
+      "grad_norm": 0.0008974755764938891,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 0.0001,
+      "step": 907
+    },
+    {
+      "epoch": 0.12846178332684893,
+      "grad_norm": 0.0007228873437270522,
+      "learning_rate": 4.420698520166988e-06,
+      "loss": 0.0001,
+      "step": 908
+    },
+    {
+      "epoch": 0.1286032610617904,
+      "grad_norm": 0.004292274825274944,
+      "learning_rate": 4.328794844116946e-06,
+      "loss": 0.0002,
+      "step": 909
+    },
+    {
+      "epoch": 0.12874473879673187,
+      "grad_norm": 0.003769136033952236,
+      "learning_rate": 4.237835404237778e-06,
+      "loss": 0.0002,
+      "step": 910
+    },
+    {
+      "epoch": 0.12888621653167331,
+      "grad_norm": 0.044125765562057495,
+      "learning_rate": 4.147821098262405e-06,
+      "loss": 0.0006,
+      "step": 911
+    },
+    {
+      "epoch": 0.12902769426661478,
+      "grad_norm": 0.001430506817996502,
+      "learning_rate": 4.0587528145957235e-06,
+      "loss": 0.0001,
+      "step": 912
+    },
+    {
+      "epoch": 0.12916917200155625,
+      "grad_norm": 0.007022987585514784,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 0.0004,
+      "step": 913
+    },
+    {
+      "epoch": 0.12931064973649772,
+      "grad_norm": 0.0009093685657717288,
+      "learning_rate": 3.883457821114811e-06,
+      "loss": 0.0001,
+      "step": 914
+    },
+    {
+      "epoch": 0.1294521274714392,
+      "grad_norm": 0.0009062972967512906,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 0.0001,
+      "step": 915
+    },
+    {
+      "epoch": 0.12959360520638064,
+      "grad_norm": 0.001498628407716751,
+      "learning_rate": 3.711957344141237e-06,
+      "loss": 0.0001,
+      "step": 916
+    },
+    {
+      "epoch": 0.1297350829413221,
+      "grad_norm": 0.0011825069086626172,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 0.0001,
+      "step": 917
+    },
+    {
+      "epoch": 0.12987656067626357,
+      "grad_norm": 0.005202493164688349,
+      "learning_rate": 3.5442581542201923e-06,
+      "loss": 0.0002,
+      "step": 918
+    },
+    {
+      "epoch": 0.13001803841120504,
+      "grad_norm": 0.0010673865908756852,
+      "learning_rate": 3.461836116672612e-06,
+      "loss": 0.0001,
+      "step": 919
+    },
+    {
+      "epoch": 0.1301595161461465,
+      "grad_norm": 0.003380801063030958,
+      "learning_rate": 3.380366871828522e-06,
+      "loss": 0.0003,
+      "step": 920
+    },
+    {
+      "epoch": 0.13030099388108796,
+      "grad_norm": 0.001819453900679946,
+      "learning_rate": 3.2998512237565005e-06,
+      "loss": 0.0002,
+      "step": 921
+    },
+    {
+      "epoch": 0.13044247161602943,
+      "grad_norm": 0.004628917668014765,
+      "learning_rate": 3.2202899671134546e-06,
+      "loss": 0.0003,
+      "step": 922
+    },
+    {
+      "epoch": 0.1305839493509709,
+      "grad_norm": 0.0009073893306776881,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 0.0001,
+      "step": 923
+    },
+    {
+      "epoch": 0.13072542708591237,
+      "grad_norm": 0.0032710530795156956,
+      "learning_rate": 3.064033759637064e-06,
+      "loss": 0.0003,
+      "step": 924
+    },
+    {
+      "epoch": 0.1308669048208538,
+      "grad_norm": 0.000904114858713001,
+      "learning_rate": 2.9873403509894203e-06,
+      "loss": 0.0001,
+      "step": 925
+    },
+    {
+      "epoch": 0.13100838255579528,
+      "grad_norm": 0.0010823109187185764,
+      "learning_rate": 2.9116044181269007e-06,
+      "loss": 0.0001,
+      "step": 926
+    },
+    {
+      "epoch": 0.13114986029073675,
+      "grad_norm": 0.010690744034945965,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 0.0006,
+      "step": 927
+    },
+    {
+      "epoch": 0.13129133802567822,
+      "grad_norm": 0.007006384897977114,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 0.0003,
+      "step": 928
+    },
+    {
+      "epoch": 0.13143281576061966,
+      "grad_norm": 0.0007068116683512926,
+      "learning_rate": 2.690148901787337e-06,
+      "loss": 0.0001,
+      "step": 929
+    },
+    {
+      "epoch": 0.13157429349556113,
+      "grad_norm": 0.003064696444198489,
+      "learning_rate": 2.618250252287113e-06,
+      "loss": 0.0002,
+      "step": 930
+    },
+    {
+      "epoch": 0.1317157712305026,
+      "grad_norm": 0.0013085821410641074,
+      "learning_rate": 2.5473127213422763e-06,
+      "loss": 0.0001,
+      "step": 931
+    },
+    {
+      "epoch": 0.13185724896544407,
+      "grad_norm": 0.0008828960126265883,
+      "learning_rate": 2.4773370090776626e-06,
+      "loss": 0.0001,
+      "step": 932
+    },
+    {
+      "epoch": 0.13199872670038554,
+      "grad_norm": 2.9277994632720947,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 1.0387,
+      "step": 933
+    },
+    {
+      "epoch": 0.13214020443532698,
+      "grad_norm": 0.0015159097965806723,
+      "learning_rate": 2.3402737936175425e-06,
+      "loss": 0.0001,
+      "step": 934
+    },
+    {
+      "epoch": 0.13228168217026845,
+      "grad_norm": 0.0016068522818386555,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 0.0001,
+      "step": 935
+    },
+    {
+      "epoch": 0.13242315990520992,
+      "grad_norm": 0.001012073247693479,
+      "learning_rate": 2.2070660169278166e-06,
+      "loss": 0.0001,
+      "step": 936
+    },
+    {
+      "epoch": 0.1325646376401514,
+      "grad_norm": 0.004135122988373041,
+      "learning_rate": 2.141909567452793e-06,
+      "loss": 0.0002,
+      "step": 937
+    },
+    {
+      "epoch": 0.13270611537509283,
+      "grad_norm": 0.0009160821209661663,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 0.0001,
+      "step": 938
+    },
+    {
+      "epoch": 0.1328475931100343,
+      "grad_norm": 0.0008047740557231009,
+      "learning_rate": 2.014494761575314e-06,
+      "loss": 0.0001,
+      "step": 939
+    },
+    {
+      "epoch": 0.13298907084497577,
+      "grad_norm": 0.0031038711313158274,
+      "learning_rate": 1.9522376627055583e-06,
+      "loss": 0.0003,
+      "step": 940
+    },
+    {
+      "epoch": 0.13313054857991724,
+      "grad_norm": 0.006230383645743132,
+      "learning_rate": 1.8909482556666024e-06,
+      "loss": 0.0005,
+      "step": 941
+    },
+    {
+      "epoch": 0.1332720263148587,
+      "grad_norm": 0.0016900923801586032,
+      "learning_rate": 1.8306271453601199e-06,
+      "loss": 0.0001,
+      "step": 942
+    },
+    {
+      "epoch": 0.13341350404980015,
+      "grad_norm": 0.0012291180901229382,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 0.0001,
+      "step": 943
+    },
+    {
+      "epoch": 0.13355498178474162,
+      "grad_norm": 0.002149342093616724,
+      "learning_rate": 1.712892186762083e-06,
+      "loss": 0.0002,
+      "step": 944
+    },
+    {
+      "epoch": 0.1336964595196831,
+      "grad_norm": 0.0009066860657185316,
+      "learning_rate": 1.6554795004670388e-06,
+      "loss": 0.0001,
+      "step": 945
+    },
+    {
+      "epoch": 0.13383793725462456,
+      "grad_norm": 0.0023614042438566685,
+      "learning_rate": 1.5990374348860305e-06,
+      "loss": 0.0001,
+      "step": 946
+    },
+    {
+      "epoch": 0.133979414989566,
+      "grad_norm": 0.0028560732025653124,
+      "learning_rate": 1.543566547079467e-06,
+      "loss": 0.0002,
+      "step": 947
+    },
+    {
+      "epoch": 0.13412089272450747,
+      "grad_norm": 0.004639330320060253,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 0.0002,
+      "step": 948
+    },
+    {
+      "epoch": 0.13426237045944894,
+      "grad_norm": 0.0011574679519981146,
+      "learning_rate": 1.4355404851001952e-06,
+      "loss": 0.0001,
+      "step": 949
+    },
+    {
+      "epoch": 0.1344038481943904,
+      "grad_norm": 0.01731015555560589,
+      "learning_rate": 1.3829863771011253e-06,
+      "loss": 0.0002,
+      "step": 950
+    },
+    {
+      "epoch": 0.13454532592933188,
+      "grad_norm": 0.0013583000982180238,
+      "learning_rate": 1.3314055792131964e-06,
+      "loss": 0.0001,
+      "step": 951
+    },
+    {
+      "epoch": 0.13468680366427332,
+      "grad_norm": 0.01872401311993599,
+      "learning_rate": 1.280798600518085e-06,
+      "loss": 0.0004,
+      "step": 952
+    },
+    {
+      "epoch": 0.1348282813992148,
+      "grad_norm": 0.005545449908822775,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 0.0005,
+      "step": 953
+    },
+    {
+      "epoch": 0.13496975913415626,
+      "grad_norm": 0.0026185193564742804,
+      "learning_rate": 1.1825080889719563e-06,
+      "loss": 0.0001,
+      "step": 954
+    },
+    {
+      "epoch": 0.13511123686909773,
+      "grad_norm": 0.0041889348067343235,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 0.0003,
+      "step": 955
+    },
+    {
+      "epoch": 0.13525271460403918,
+      "grad_norm": 0.0009472736855968833,
+      "learning_rate": 1.0881187228038215e-06,
+      "loss": 0.0001,
+      "step": 956
+    },
+    {
+      "epoch": 0.13539419233898065,
+      "grad_norm": 0.0025584392715245485,
+      "learning_rate": 1.0423881397349068e-06,
+      "loss": 0.0001,
+      "step": 957
+    },
+    {
+      "epoch": 0.13553567007392212,
+      "grad_norm": 0.0036006199661642313,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 0.0003,
+      "step": 958
+    },
+    {
+      "epoch": 0.13567714780886359,
+      "grad_norm": 0.004876970779150724,
+      "learning_rate": 9.538574303348813e-07,
+      "loss": 0.0003,
+      "step": 959
+    },
+    {
+      "epoch": 0.13581862554380505,
+      "grad_norm": 0.0027515313122421503,
+      "learning_rate": 9.110581777661331e-07,
+      "loss": 0.0002,
+      "step": 960
+    },
+    {
+      "epoch": 0.1359601032787465,
+      "grad_norm": 0.001473053591325879,
+      "learning_rate": 8.692368930493521e-07,
+      "loss": 0.0001,
+      "step": 961
+    },
+    {
+      "epoch": 0.13610158101368797,
+      "grad_norm": 0.004612274467945099,
+      "learning_rate": 8.283939889437209e-07,
+      "loss": 0.0002,
+      "step": 962
+    },
+    {
+      "epoch": 0.13624305874862944,
+      "grad_norm": 0.0012349090538918972,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 0.0001,
+      "step": 963
+    },
+    {
+      "epoch": 0.1363845364835709,
+      "grad_norm": 0.0017732393462210894,
+      "learning_rate": 7.496449253176274e-07,
+      "loss": 0.0002,
+      "step": 964
+    },
+    {
+      "epoch": 0.13652601421851235,
+      "grad_norm": 0.0008514790097251534,
+      "learning_rate": 7.117395430186414e-07,
+      "loss": 0.0001,
+      "step": 965
+    },
+    {
+      "epoch": 0.13666749195345382,
+      "grad_norm": 0.000827544427011162,
+      "learning_rate": 6.748140957660631e-07,
+      "loss": 0.0001,
+      "step": 966
+    },
+    {
+      "epoch": 0.1368089696883953,
+      "grad_norm": 0.0010190614266321063,
+      "learning_rate": 6.388689479991605e-07,
+      "loss": 0.0001,
+      "step": 967
+    },
+    {
+      "epoch": 0.13695044742333676,
+      "grad_norm": 0.0039400323294103146,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 0.0002,
+      "step": 968
+    },
+    {
+      "epoch": 0.13709192515827823,
+      "grad_norm": 0.002428358420729637,
+      "learning_rate": 5.699209603001076e-07,
+      "loss": 0.0002,
+      "step": 969
+    },
+    {
+      "epoch": 0.13723340289321967,
+      "grad_norm": 0.0017006148118525743,
+      "learning_rate": 5.369188008567672e-07,
+      "loss": 0.0001,
+      "step": 970
+    },
+    {
+      "epoch": 0.13737488062816114,
+      "grad_norm": 0.0008442560210824013,
+      "learning_rate": 5.048983018699827e-07,
+      "loss": 0.0001,
+      "step": 971
+    },
+    {
+      "epoch": 0.1375163583631026,
+      "grad_norm": 0.002315743826329708,
+      "learning_rate": 4.738597793691679e-07,
+      "loss": 0.0002,
+      "step": 972
+    },
+    {
+      "epoch": 0.13765783609804408,
+      "grad_norm": 0.016204334795475006,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 0.0003,
+      "step": 973
+    },
+    {
+      "epoch": 0.13779931383298552,
+      "grad_norm": 0.007021232508122921,
+      "learning_rate": 4.1472987948143473e-07,
+      "loss": 0.0005,
+      "step": 974
+    },
+    {
+      "epoch": 0.137940791567927,
+      "grad_norm": 0.0011094443034380674,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 0.0001,
+      "step": 975
+    },
+    {
+      "epoch": 0.13808226930286846,
+      "grad_norm": 0.0013754486571997404,
+      "learning_rate": 3.595314355407609e-07,
+      "loss": 0.0001,
+      "step": 976
+    },
+    {
+      "epoch": 0.13822374703780993,
+      "grad_norm": 0.0012389803305268288,
+      "learning_rate": 3.3340719659701313e-07,
+      "loss": 0.0001,
+      "step": 977
+    },
+    {
+      "epoch": 0.1383652247727514,
+      "grad_norm": 0.0010980741353705525,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 0.0001,
+      "step": 978
+    },
+    {
+      "epoch": 0.13850670250769284,
+      "grad_norm": 0.006173287518322468,
+      "learning_rate": 2.841099739386066e-07,
+      "loss": 0.0004,
+      "step": 979
+    },
+    {
+      "epoch": 0.1386481802426343,
+      "grad_norm": 1.0002788305282593,
+      "learning_rate": 2.609374767676309e-07,
+      "loss": 0.3774,
+      "step": 980
+    },
+    {
+      "epoch": 0.13878965797757578,
+      "grad_norm": 0.0016743881860747933,
+      "learning_rate": 2.387493638774774e-07,
+      "loss": 0.0001,
+      "step": 981
+    },
+    {
+      "epoch": 0.13893113571251725,
+      "grad_norm": 0.0023470830637961626,
+      "learning_rate": 2.175458542558517e-07,
+      "loss": 0.0001,
+      "step": 982
+    },
+    {
+      "epoch": 0.1390726134474587,
+      "grad_norm": 0.015273729339241982,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 0.0009,
+      "step": 983
+    },
+    {
+      "epoch": 0.13921409118240016,
+      "grad_norm": 0.008181693963706493,
+      "learning_rate": 1.7809347217881966e-07,
+      "loss": 0.0004,
+      "step": 984
+    },
+    {
+      "epoch": 0.13935556891734163,
+      "grad_norm": 0.0018706975970417261,
+      "learning_rate": 1.598449891024978e-07,
+      "loss": 0.0001,
+      "step": 985
+    },
+    {
+      "epoch": 0.1394970466522831,
+      "grad_norm": 0.002527001081034541,
+      "learning_rate": 1.425818880490315e-07,
+      "loss": 0.0002,
+      "step": 986
+    },
+    {
+      "epoch": 0.13963852438722457,
+      "grad_norm": 0.0017283704364672303,
+      "learning_rate": 1.2630433939825327e-07,
+      "loss": 0.0001,
+      "step": 987
+    },
+    {
+      "epoch": 0.139780002122166,
+      "grad_norm": 0.0018094810657203197,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 0.0001,
+      "step": 988
+    },
+    {
+      "epoch": 0.13992147985710748,
+      "grad_norm": 0.001633922685869038,
+      "learning_rate": 9.670653218752934e-08,
+      "loss": 0.0001,
+      "step": 989
+    },
+    {
+      "epoch": 0.14006295759204895,
+      "grad_norm": 0.0005336939357221127,
+      "learning_rate": 8.33865657459909e-08,
+      "loss": 0.0001,
+      "step": 990
+    },
+    {
+      "epoch": 0.14020443532699042,
+      "grad_norm": 0.004103470593690872,
+      "learning_rate": 7.105273594107953e-08,
+      "loss": 0.0002,
+      "step": 991
+    },
+    {
+      "epoch": 0.14034591306193187,
+      "grad_norm": 0.003065005410462618,
+      "learning_rate": 5.970516450271025e-08,
+      "loss": 0.0003,
+      "step": 992
+    },
+    {
+      "epoch": 0.14048739079687333,
+      "grad_norm": 0.0010252476204186678,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 0.0001,
+      "step": 993
+    },
+    {
+      "epoch": 0.1406288685318148,
+      "grad_norm": 0.0008998365374282002,
+      "learning_rate": 3.996923497434635e-08,
+      "loss": 0.0001,
+      "step": 994
+    },
+    {
+      "epoch": 0.14077034626675627,
+      "grad_norm": 0.007965818047523499,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 0.0003,
+      "step": 995
+    },
+    {
+      "epoch": 0.14091182400169774,
+      "grad_norm": 0.0009977156296372414,
+      "learning_rate": 2.417955630159563e-08,
+      "loss": 0.0001,
+      "step": 996
+    },
+    {
+      "epoch": 0.1410533017366392,
+      "grad_norm": 0.0016674831276759505,
+      "learning_rate": 1.7764761919103477e-08,
+      "loss": 0.0001,
+      "step": 997
+    },
+    {
+      "epoch": 0.14119477947158066,
+      "grad_norm": 0.002003878355026245,
+      "learning_rate": 1.2336751833941229e-08,
+      "loss": 0.0002,
+      "step": 998
+    },
+    {
+      "epoch": 0.14133625720652213,
+      "grad_norm": 0.0028543765656650066,
+      "learning_rate": 7.895579618388827e-09,
+      "loss": 0.0001,
+      "step": 999
+    },
+    {
+      "epoch": 0.1414777349414636,
+      "grad_norm": 0.0011507285526022315,
+      "learning_rate": 4.4412891050171765e-09,
+      "loss": 0.0001,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.1662243053568e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}