diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11313 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9995807518516795,
+  "eval_steps": 10500,
+  "global_step": 16098,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001863325103647459,
+      "grad_norm": 8.962045669555664,
+      "learning_rate": 0.00019987576096409493,
+      "loss": 2.809,
+      "step": 10
+    },
+    {
+      "epoch": 0.003726650207294918,
+      "grad_norm": 3.3185620307922363,
+      "learning_rate": 0.00019975152192818984,
+      "loss": 0.4593,
+      "step": 20
+    },
+    {
+      "epoch": 0.0055899753109423765,
+      "grad_norm": 3.248004674911499,
+      "learning_rate": 0.00019962728289228476,
+      "loss": 0.2631,
+      "step": 30
+    },
+    {
+      "epoch": 0.007453300414589836,
+      "grad_norm": 1.8456782102584839,
+      "learning_rate": 0.00019950304385637967,
+      "loss": 0.186,
+      "step": 40
+    },
+    {
+      "epoch": 0.009316625518237294,
+      "grad_norm": 1.2301431894302368,
+      "learning_rate": 0.0001993788048204746,
+      "loss": 0.1519,
+      "step": 50
+    },
+    {
+      "epoch": 0.011179950621884753,
+      "grad_norm": 0.8274269104003906,
+      "learning_rate": 0.0001992545657845695,
+      "loss": 0.1339,
+      "step": 60
+    },
+    {
+      "epoch": 0.013043275725532212,
+      "grad_norm": 1.3983819484710693,
+      "learning_rate": 0.00019913032674866442,
+      "loss": 0.1403,
+      "step": 70
+    },
+    {
+      "epoch": 0.014906600829179672,
+      "grad_norm": 0.9112318158149719,
+      "learning_rate": 0.00019900608771275934,
+      "loss": 0.119,
+      "step": 80
+    },
+    {
+      "epoch": 0.01676992593282713,
+      "grad_norm": 0.9845889806747437,
+      "learning_rate": 0.00019888184867685428,
+      "loss": 0.1003,
+      "step": 90
+    },
+    {
+      "epoch": 0.018633251036474587,
+      "grad_norm": 0.6126352548599243,
+      "learning_rate": 0.0001987576096409492,
+      "loss": 0.1063,
+      "step": 100
+    },
+    {
+      "epoch": 0.02049657614012205,
+      "grad_norm": 0.6967755556106567,
+      "learning_rate": 0.0001986333706050441,
+      "loss": 0.1097,
+      "step": 110
+    },
+    {
+      "epoch": 0.022359901243769506,
+      "grad_norm": 0.994688093662262,
+      "learning_rate": 0.00019850913156913903,
+      "loss": 0.1087,
+      "step": 120
+    },
+    {
+      "epoch": 0.024223226347416967,
+      "grad_norm": 0.9732158184051514,
+      "learning_rate": 0.00019838489253323397,
+      "loss": 0.0942,
+      "step": 130
+    },
+    {
+      "epoch": 0.026086551451064425,
+      "grad_norm": 0.45771920680999756,
+      "learning_rate": 0.00019826065349732889,
+      "loss": 0.0915,
+      "step": 140
+    },
+    {
+      "epoch": 0.027949876554711883,
+      "grad_norm": 0.6761167645454407,
+      "learning_rate": 0.0001981364144614238,
+      "loss": 0.0882,
+      "step": 150
+    },
+    {
+      "epoch": 0.029813201658359344,
+      "grad_norm": 0.8559990525245667,
+      "learning_rate": 0.00019801217542551872,
+      "loss": 0.0842,
+      "step": 160
+    },
+    {
+      "epoch": 0.0316765267620068,
+      "grad_norm": 0.6518134474754333,
+      "learning_rate": 0.00019788793638961363,
+      "loss": 0.0891,
+      "step": 170
+    },
+    {
+      "epoch": 0.03353985186565426,
+      "grad_norm": 0.6274904608726501,
+      "learning_rate": 0.00019776369735370855,
+      "loss": 0.0851,
+      "step": 180
+    },
+    {
+      "epoch": 0.03540317696930172,
+      "grad_norm": 0.4390016496181488,
+      "learning_rate": 0.00019763945831780346,
+      "loss": 0.0932,
+      "step": 190
+    },
+    {
+      "epoch": 0.037266502072949174,
+      "grad_norm": 0.5347189903259277,
+      "learning_rate": 0.00019751521928189838,
+      "loss": 0.0766,
+      "step": 200
+    },
+    {
+      "epoch": 0.03912982717659664,
+      "grad_norm": 2.5851993560791016,
+      "learning_rate": 0.0001973909802459933,
+      "loss": 0.0769,
+      "step": 210
+    },
+    {
+      "epoch": 0.0409931522802441,
+      "grad_norm": 0.7917025089263916,
+      "learning_rate": 0.0001972667412100882,
+      "loss": 0.0891,
+      "step": 220
+    },
+    {
+      "epoch": 0.042856477383891554,
+      "grad_norm": 1.9254189729690552,
+      "learning_rate": 0.00019714250217418313,
+      "loss": 0.0748,
+      "step": 230
+    },
+    {
+      "epoch": 0.04471980248753901,
+      "grad_norm": 0.4789351224899292,
+      "learning_rate": 0.00019701826313827804,
+      "loss": 0.0762,
+      "step": 240
+    },
+    {
+      "epoch": 0.04658312759118647,
+      "grad_norm": 0.4679170548915863,
+      "learning_rate": 0.00019689402410237296,
+      "loss": 0.0716,
+      "step": 250
+    },
+    {
+      "epoch": 0.048446452694833934,
+      "grad_norm": 0.6142864227294922,
+      "learning_rate": 0.0001967697850664679,
+      "loss": 0.0717,
+      "step": 260
+    },
+    {
+      "epoch": 0.05030977779848139,
+      "grad_norm": 0.6055127382278442,
+      "learning_rate": 0.00019664554603056282,
+      "loss": 0.0784,
+      "step": 270
+    },
+    {
+      "epoch": 0.05217310290212885,
+      "grad_norm": 0.5234378576278687,
+      "learning_rate": 0.00019652130699465773,
+      "loss": 0.0773,
+      "step": 280
+    },
+    {
+      "epoch": 0.05403642800577631,
+      "grad_norm": 0.6473028063774109,
+      "learning_rate": 0.00019639706795875265,
+      "loss": 0.0801,
+      "step": 290
+    },
+    {
+      "epoch": 0.055899753109423765,
+      "grad_norm": 0.5021067261695862,
+      "learning_rate": 0.00019627282892284757,
+      "loss": 0.0821,
+      "step": 300
+    },
+    {
+      "epoch": 0.05776307821307122,
+      "grad_norm": 0.5899055600166321,
+      "learning_rate": 0.00019614858988694248,
+      "loss": 0.0829,
+      "step": 310
+    },
+    {
+      "epoch": 0.05962640331671869,
+      "grad_norm": 0.9202219247817993,
+      "learning_rate": 0.0001960243508510374,
+      "loss": 0.0723,
+      "step": 320
+    },
+    {
+      "epoch": 0.061489728420366145,
+      "grad_norm": 0.45627737045288086,
+      "learning_rate": 0.0001959001118151323,
+      "loss": 0.0815,
+      "step": 330
+    },
+    {
+      "epoch": 0.0633530535240136,
+      "grad_norm": 0.4791790246963501,
+      "learning_rate": 0.00019577587277922726,
+      "loss": 0.0754,
+      "step": 340
+    },
+    {
+      "epoch": 0.06521637862766107,
+      "grad_norm": 0.48401352763175964,
+      "learning_rate": 0.00019565163374332217,
+      "loss": 0.0834,
+      "step": 350
+    },
+    {
+      "epoch": 0.06707970373130852,
+      "grad_norm": 0.45857805013656616,
+      "learning_rate": 0.0001955273947074171,
+      "loss": 0.0794,
+      "step": 360
+    },
+    {
+      "epoch": 0.06894302883495598,
+      "grad_norm": 0.3174554705619812,
+      "learning_rate": 0.000195403155671512,
+      "loss": 0.0567,
+      "step": 370
+    },
+    {
+      "epoch": 0.07080635393860343,
+      "grad_norm": 0.4005182087421417,
+      "learning_rate": 0.00019527891663560692,
+      "loss": 0.0593,
+      "step": 380
+    },
+    {
+      "epoch": 0.0726696790422509,
+      "grad_norm": 0.6048309206962585,
+      "learning_rate": 0.00019515467759970183,
+      "loss": 0.0802,
+      "step": 390
+    },
+    {
+      "epoch": 0.07453300414589835,
+      "grad_norm": 1.142407774925232,
+      "learning_rate": 0.00019503043856379675,
+      "loss": 0.0695,
+      "step": 400
+    },
+    {
+      "epoch": 0.07639632924954581,
+      "grad_norm": 0.4419887065887451,
+      "learning_rate": 0.00019490619952789167,
+      "loss": 0.0601,
+      "step": 410
+    },
+    {
+      "epoch": 0.07825965435319328,
+      "grad_norm": 0.33916333317756653,
+      "learning_rate": 0.00019478196049198658,
+      "loss": 0.052,
+      "step": 420
+    },
+    {
+      "epoch": 0.08012297945684073,
+      "grad_norm": 0.8268319964408875,
+      "learning_rate": 0.00019465772145608152,
+      "loss": 0.0602,
+      "step": 430
+    },
+    {
+      "epoch": 0.0819863045604882,
+      "grad_norm": 0.3288143575191498,
+      "learning_rate": 0.00019453348242017644,
+      "loss": 0.0559,
+      "step": 440
+    },
+    {
+      "epoch": 0.08384962966413564,
+      "grad_norm": 1.1518080234527588,
+      "learning_rate": 0.00019440924338427136,
+      "loss": 0.0586,
+      "step": 450
+    },
+    {
+      "epoch": 0.08571295476778311,
+      "grad_norm": 0.4662655293941498,
+      "learning_rate": 0.00019428500434836627,
+      "loss": 0.0828,
+      "step": 460
+    },
+    {
+      "epoch": 0.08757627987143057,
+      "grad_norm": 0.5718135237693787,
+      "learning_rate": 0.0001941607653124612,
+      "loss": 0.0538,
+      "step": 470
+    },
+    {
+      "epoch": 0.08943960497507802,
+      "grad_norm": 0.40107667446136475,
+      "learning_rate": 0.0001940365262765561,
+      "loss": 0.0492,
+      "step": 480
+    },
+    {
+      "epoch": 0.09130293007872549,
+      "grad_norm": 3.8866326808929443,
+      "learning_rate": 0.00019391228724065102,
+      "loss": 0.0554,
+      "step": 490
+    },
+    {
+      "epoch": 0.09316625518237294,
+      "grad_norm": 0.42105963826179504,
+      "learning_rate": 0.00019378804820474594,
+      "loss": 0.0597,
+      "step": 500
+    },
+    {
+      "epoch": 0.0950295802860204,
+      "grad_norm": 0.49508970975875854,
+      "learning_rate": 0.00019366380916884085,
+      "loss": 0.0541,
+      "step": 510
+    },
+    {
+      "epoch": 0.09689290538966787,
+      "grad_norm": 0.35285356640815735,
+      "learning_rate": 0.00019353957013293577,
+      "loss": 0.0489,
+      "step": 520
+    },
+    {
+      "epoch": 0.09875623049331532,
+      "grad_norm": 0.5713452100753784,
+      "learning_rate": 0.00019341533109703068,
+      "loss": 0.0606,
+      "step": 530
+    },
+    {
+      "epoch": 0.10061955559696278,
+      "grad_norm": 0.5628998279571533,
+      "learning_rate": 0.0001932910920611256,
+      "loss": 0.0604,
+      "step": 540
+    },
+    {
+      "epoch": 0.10248288070061023,
+      "grad_norm": 0.4256053864955902,
+      "learning_rate": 0.00019316685302522054,
+      "loss": 0.054,
+      "step": 550
+    },
+    {
+      "epoch": 0.1043462058042577,
+      "grad_norm": 0.4267830550670624,
+      "learning_rate": 0.00019304261398931546,
+      "loss": 0.0687,
+      "step": 560
+    },
+    {
+      "epoch": 0.10620953090790515,
+      "grad_norm": 0.5250473618507385,
+      "learning_rate": 0.00019291837495341037,
+      "loss": 0.0587,
+      "step": 570
+    },
+    {
+      "epoch": 0.10807285601155261,
+      "grad_norm": 0.45175454020500183,
+      "learning_rate": 0.0001927941359175053,
+      "loss": 0.0532,
+      "step": 580
+    },
+    {
+      "epoch": 0.10993618111520008,
+      "grad_norm": 0.40816521644592285,
+      "learning_rate": 0.0001926698968816002,
+      "loss": 0.0583,
+      "step": 590
+    },
+    {
+      "epoch": 0.11179950621884753,
+      "grad_norm": 0.5700634717941284,
+      "learning_rate": 0.00019254565784569515,
+      "loss": 0.0507,
+      "step": 600
+    },
+    {
+      "epoch": 0.113662831322495,
+      "grad_norm": 0.32277169823646545,
+      "learning_rate": 0.00019242141880979006,
+      "loss": 0.0485,
+      "step": 610
+    },
+    {
+      "epoch": 0.11552615642614245,
+      "grad_norm": 0.5027480721473694,
+      "learning_rate": 0.00019229717977388498,
+      "loss": 0.0493,
+      "step": 620
+    },
+    {
+      "epoch": 0.11738948152978991,
+      "grad_norm": 0.43033790588378906,
+      "learning_rate": 0.0001921729407379799,
+      "loss": 0.0498,
+      "step": 630
+    },
+    {
+      "epoch": 0.11925280663343737,
+      "grad_norm": 0.42897236347198486,
+      "learning_rate": 0.0001920487017020748,
+      "loss": 0.0542,
+      "step": 640
+    },
+    {
+      "epoch": 0.12111613173708483,
+      "grad_norm": 0.46961647272109985,
+      "learning_rate": 0.00019192446266616973,
+      "loss": 0.0532,
+      "step": 650
+    },
+    {
+      "epoch": 0.12297945684073229,
+      "grad_norm": 0.3467908799648285,
+      "learning_rate": 0.00019180022363026464,
+      "loss": 0.0702,
+      "step": 660
+    },
+    {
+      "epoch": 0.12484278194437974,
+      "grad_norm": 0.3707217574119568,
+      "learning_rate": 0.00019167598459435956,
+      "loss": 0.0496,
+      "step": 670
+    },
+    {
+      "epoch": 0.1267061070480272,
+      "grad_norm": 0.28889259696006775,
+      "learning_rate": 0.00019155174555845447,
+      "loss": 0.0507,
+      "step": 680
+    },
+    {
+      "epoch": 0.12856943215167466,
+      "grad_norm": 0.3841610550880432,
+      "learning_rate": 0.0001914275065225494,
+      "loss": 0.0518,
+      "step": 690
+    },
+    {
+      "epoch": 0.13043275725532213,
+      "grad_norm": 1.2005867958068848,
+      "learning_rate": 0.0001913032674866443,
+      "loss": 0.051,
+      "step": 700
+    },
+    {
+      "epoch": 0.13229608235896959,
+      "grad_norm": 0.3136664628982544,
+      "learning_rate": 0.00019117902845073922,
+      "loss": 0.0512,
+      "step": 710
+    },
+    {
+      "epoch": 0.13415940746261704,
+      "grad_norm": 0.2575332224369049,
+      "learning_rate": 0.00019105478941483414,
+      "loss": 0.0471,
+      "step": 720
+    },
+    {
+      "epoch": 0.1360227325662645,
+      "grad_norm": 0.44971975684165955,
+      "learning_rate": 0.00019093055037892905,
+      "loss": 0.0542,
+      "step": 730
+    },
+    {
+      "epoch": 0.13788605766991197,
+      "grad_norm": 0.40713998675346375,
+      "learning_rate": 0.00019080631134302397,
+      "loss": 0.0456,
+      "step": 740
+    },
+    {
+      "epoch": 0.13974938277355942,
+      "grad_norm": 0.4077875018119812,
+      "learning_rate": 0.00019068207230711888,
+      "loss": 0.0532,
+      "step": 750
+    },
+    {
+      "epoch": 0.14161270787720687,
+      "grad_norm": 0.4080505073070526,
+      "learning_rate": 0.00019055783327121383,
+      "loss": 0.0598,
+      "step": 760
+    },
+    {
+      "epoch": 0.14347603298085435,
+      "grad_norm": 0.30817773938179016,
+      "learning_rate": 0.00019043359423530874,
+      "loss": 0.0451,
+      "step": 770
+    },
+    {
+      "epoch": 0.1453393580845018,
+      "grad_norm": 0.27631935477256775,
+      "learning_rate": 0.00019030935519940366,
+      "loss": 0.0469,
+      "step": 780
+    },
+    {
+      "epoch": 0.14720268318814925,
+      "grad_norm": 0.8231372237205505,
+      "learning_rate": 0.0001901851161634986,
+      "loss": 0.0465,
+      "step": 790
+    },
+    {
+      "epoch": 0.1490660082917967,
+      "grad_norm": 0.3672221302986145,
+      "learning_rate": 0.00019006087712759352,
+      "loss": 0.0489,
+      "step": 800
+    },
+    {
+      "epoch": 0.15092933339544418,
+      "grad_norm": 0.32903629541397095,
+      "learning_rate": 0.00018993663809168843,
+      "loss": 0.0382,
+      "step": 810
+    },
+    {
+      "epoch": 0.15279265849909163,
+      "grad_norm": 0.291094034910202,
+      "learning_rate": 0.00018981239905578335,
+      "loss": 0.0478,
+      "step": 820
+    },
+    {
+      "epoch": 0.15465598360273908,
+      "grad_norm": 0.45158231258392334,
+      "learning_rate": 0.00018968816001987827,
+      "loss": 0.0443,
+      "step": 830
+    },
+    {
+      "epoch": 0.15651930870638656,
+      "grad_norm": 0.3328520953655243,
+      "learning_rate": 0.00018956392098397318,
+      "loss": 0.0369,
+      "step": 840
+    },
+    {
+      "epoch": 0.158382633810034,
+      "grad_norm": 0.3631761372089386,
+      "learning_rate": 0.0001894396819480681,
+      "loss": 0.0606,
+      "step": 850
+    },
+    {
+      "epoch": 0.16024595891368146,
+      "grad_norm": 0.37086474895477295,
+      "learning_rate": 0.000189315442912163,
+      "loss": 0.046,
+      "step": 860
+    },
+    {
+      "epoch": 0.16210928401732894,
+      "grad_norm": 0.43115702271461487,
+      "learning_rate": 0.00018919120387625793,
+      "loss": 0.0401,
+      "step": 870
+    },
+    {
+      "epoch": 0.1639726091209764,
+      "grad_norm": 0.44107526540756226,
+      "learning_rate": 0.00018906696484035284,
+      "loss": 0.0441,
+      "step": 880
+    },
+    {
+      "epoch": 0.16583593422462384,
+      "grad_norm": 0.3664778172969818,
+      "learning_rate": 0.00018894272580444776,
+      "loss": 0.0498,
+      "step": 890
+    },
+    {
+      "epoch": 0.1676992593282713,
+      "grad_norm": 0.24749009311199188,
+      "learning_rate": 0.00018881848676854268,
+      "loss": 0.0408,
+      "step": 900
+    },
+    {
+      "epoch": 0.16956258443191877,
+      "grad_norm": 0.41872236132621765,
+      "learning_rate": 0.0001886942477326376,
+      "loss": 0.0568,
+      "step": 910
+    },
+    {
+      "epoch": 0.17142590953556622,
+      "grad_norm": 0.25826430320739746,
+      "learning_rate": 0.0001885700086967325,
+      "loss": 0.0477,
+      "step": 920
+    },
+    {
+      "epoch": 0.17328923463921367,
+      "grad_norm": 0.29239359498023987,
+      "learning_rate": 0.00018844576966082742,
+      "loss": 0.0472,
+      "step": 930
+    },
+    {
+      "epoch": 0.17515255974286115,
+      "grad_norm": 0.8567617535591125,
+      "learning_rate": 0.00018832153062492234,
+      "loss": 0.0396,
+      "step": 940
+    },
+    {
+      "epoch": 0.1770158848465086,
+      "grad_norm": 0.47980305552482605,
+      "learning_rate": 0.00018819729158901728,
+      "loss": 0.0404,
+      "step": 950
+    },
+    {
+      "epoch": 0.17887920995015605,
+      "grad_norm": 0.2695956528186798,
+      "learning_rate": 0.0001880730525531122,
+      "loss": 0.0385,
+      "step": 960
+    },
+    {
+      "epoch": 0.1807425350538035,
+      "grad_norm": 0.3030606508255005,
+      "learning_rate": 0.00018794881351720711,
+      "loss": 0.0433,
+      "step": 970
+    },
+    {
+      "epoch": 0.18260586015745098,
+      "grad_norm": 0.5657457709312439,
+      "learning_rate": 0.00018782457448130203,
+      "loss": 0.0492,
+      "step": 980
+    },
+    {
+      "epoch": 0.18446918526109843,
+      "grad_norm": 0.4033714532852173,
+      "learning_rate": 0.00018770033544539695,
+      "loss": 0.0439,
+      "step": 990
+    },
+    {
+      "epoch": 0.18633251036474588,
+      "grad_norm": 0.21745193004608154,
+      "learning_rate": 0.00018757609640949186,
+      "loss": 0.0353,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18819583546839336,
+      "grad_norm": 0.3541043698787689,
+      "learning_rate": 0.0001874518573735868,
+      "loss": 0.0387,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1900591605720408,
+      "grad_norm": 0.38949277997016907,
+      "learning_rate": 0.00018732761833768172,
+      "loss": 0.0477,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19192248567568826,
+      "grad_norm": 0.2696048617362976,
+      "learning_rate": 0.00018720337930177664,
+      "loss": 0.0467,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19378581077933574,
+      "grad_norm": 0.40931448340415955,
+      "learning_rate": 0.00018707914026587155,
+      "loss": 0.0387,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1956491358829832,
+      "grad_norm": 1.0407683849334717,
+      "learning_rate": 0.00018695490122996647,
+      "loss": 0.0418,
+      "step": 1050
+    },
+    {
+      "epoch": 0.19751246098663064,
+      "grad_norm": 0.8484430313110352,
+      "learning_rate": 0.00018683066219406138,
+      "loss": 0.0379,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1993757860902781,
+      "grad_norm": 0.2715752422809601,
+      "learning_rate": 0.0001867064231581563,
+      "loss": 0.0472,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20123911119392557,
+      "grad_norm": 0.3399510383605957,
+      "learning_rate": 0.00018658218412225122,
+      "loss": 0.0412,
+      "step": 1080
+    },
+    {
+      "epoch": 0.20310243629757302,
+      "grad_norm": 0.36955976486206055,
+      "learning_rate": 0.00018645794508634613,
+      "loss": 0.0414,
+      "step": 1090
+    },
+    {
+      "epoch": 0.20496576140122047,
+      "grad_norm": 0.43797340989112854,
+      "learning_rate": 0.00018633370605044105,
+      "loss": 0.0439,
+      "step": 1100
+    },
+    {
+      "epoch": 0.20682908650486795,
+      "grad_norm": 0.2698400020599365,
+      "learning_rate": 0.00018620946701453596,
+      "loss": 0.0426,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2086924116085154,
+      "grad_norm": 0.39951518177986145,
+      "learning_rate": 0.0001860852279786309,
+      "loss": 0.0436,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21055573671216285,
+      "grad_norm": 0.24244333803653717,
+      "learning_rate": 0.00018596098894272582,
+      "loss": 0.0424,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2124190618158103,
+      "grad_norm": 1.0004985332489014,
+      "learning_rate": 0.00018583674990682074,
+      "loss": 0.0586,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21428238691945778,
+      "grad_norm": 0.40292367339134216,
+      "learning_rate": 0.00018571251087091565,
+      "loss": 0.0416,
+      "step": 1150
+    },
+    {
+      "epoch": 0.21614571202310523,
+      "grad_norm": 2.8957316875457764,
+      "learning_rate": 0.00018558827183501057,
+      "loss": 0.0486,
+      "step": 1160
+    },
+    {
+      "epoch": 0.21800903712675268,
+      "grad_norm": 0.3073020875453949,
+      "learning_rate": 0.00018546403279910548,
+      "loss": 0.048,
+      "step": 1170
+    },
+    {
+      "epoch": 0.21987236223040016,
+      "grad_norm": 0.3981713354587555,
+      "learning_rate": 0.0001853397937632004,
+      "loss": 0.0406,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2217356873340476,
+      "grad_norm": 0.3430505394935608,
+      "learning_rate": 0.00018521555472729532,
+      "loss": 0.0342,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22359901243769506,
+      "grad_norm": 0.2942293584346771,
+      "learning_rate": 0.00018509131569139023,
+      "loss": 0.0373,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22546233754134254,
+      "grad_norm": 0.2513025403022766,
+      "learning_rate": 0.00018496707665548515,
+      "loss": 0.0437,
+      "step": 1210
+    },
+    {
+      "epoch": 0.22732566264499,
+      "grad_norm": 0.5296605825424194,
+      "learning_rate": 0.0001848428376195801,
+      "loss": 0.0431,
+      "step": 1220
+    },
+    {
+      "epoch": 0.22918898774863744,
+      "grad_norm": 0.28046369552612305,
+      "learning_rate": 0.000184718598583675,
+      "loss": 0.0393,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2310523128522849,
+      "grad_norm": 0.4907744526863098,
+      "learning_rate": 0.00018459435954776992,
+      "loss": 0.043,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23291563795593237,
+      "grad_norm": 0.829265296459198,
+      "learning_rate": 0.00018447012051186484,
+      "loss": 0.0333,
+      "step": 1250
+    },
+    {
+      "epoch": 0.23477896305957982,
+      "grad_norm": 0.21454624831676483,
+      "learning_rate": 0.00018434588147595975,
+      "loss": 0.0303,
+      "step": 1260
+    },
+    {
+      "epoch": 0.23664228816322727,
+      "grad_norm": 0.2321559488773346,
+      "learning_rate": 0.00018422164244005467,
+      "loss": 0.0525,
+      "step": 1270
+    },
+    {
+      "epoch": 0.23850561326687475,
+      "grad_norm": 0.30932319164276123,
+      "learning_rate": 0.00018409740340414959,
+      "loss": 0.0295,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2403689383705222,
+      "grad_norm": 0.2902001738548279,
+      "learning_rate": 0.00018397316436824453,
+      "loss": 0.0408,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24223226347416965,
+      "grad_norm": 0.3144378066062927,
+      "learning_rate": 0.00018384892533233944,
+      "loss": 0.0321,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2440955885778171,
+      "grad_norm": 0.20050252974033356,
+      "learning_rate": 0.00018372468629643436,
+      "loss": 0.0317,
+      "step": 1310
+    },
+    {
+      "epoch": 0.24595891368146458,
+      "grad_norm": 0.6081618666648865,
+      "learning_rate": 0.00018360044726052928,
+      "loss": 0.0334,
+      "step": 1320
+    },
+    {
+      "epoch": 0.24782223878511203,
+      "grad_norm": 0.2507849931716919,
+      "learning_rate": 0.0001834762082246242,
+      "loss": 0.0486,
+      "step": 1330
+    },
+    {
+      "epoch": 0.24968556388875948,
+      "grad_norm": 0.25980520248413086,
+      "learning_rate": 0.0001833519691887191,
+      "loss": 0.0454,
+      "step": 1340
+    },
+    {
+      "epoch": 0.25154888899240696,
+      "grad_norm": 0.2981957197189331,
+      "learning_rate": 0.00018322773015281402,
+      "loss": 0.0347,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2534122140960544,
+      "grad_norm": 0.21605946123600006,
+      "learning_rate": 0.00018310349111690894,
+      "loss": 0.0347,
+      "step": 1360
+    },
+    {
+      "epoch": 0.25527553919970186,
+      "grad_norm": 0.16565227508544922,
+      "learning_rate": 0.00018297925208100385,
+      "loss": 0.0348,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2571388643033493,
+      "grad_norm": 0.23056954145431519,
+      "learning_rate": 0.00018285501304509877,
+      "loss": 0.0313,
+      "step": 1380
+    },
+    {
+      "epoch": 0.25900218940699676,
+      "grad_norm": 0.1978144496679306,
+      "learning_rate": 0.0001827307740091937,
+      "loss": 0.0324,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26086551451064427,
+      "grad_norm": 0.4201340675354004,
+      "learning_rate": 0.0001826065349732886,
+      "loss": 0.038,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2627288396142917,
+      "grad_norm": 0.2694622874259949,
+      "learning_rate": 0.00018248229593738352,
+      "loss": 0.0617,
+      "step": 1410
+    },
+    {
+      "epoch": 0.26459216471793917,
+      "grad_norm": 0.2527810037136078,
+      "learning_rate": 0.00018235805690147843,
+      "loss": 0.0278,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2664554898215866,
+      "grad_norm": 0.1482096016407013,
+      "learning_rate": 0.00018223381786557338,
+      "loss": 0.035,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2683188149252341,
+      "grad_norm": 0.36221590638160706,
+      "learning_rate": 0.0001821095788296683,
+      "loss": 0.0306,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2701821400288815,
+      "grad_norm": 1.1332191228866577,
+      "learning_rate": 0.0001819853397937632,
+      "loss": 0.0383,
+      "step": 1450
+    },
+    {
+      "epoch": 0.272045465132529,
+      "grad_norm": 0.16091710329055786,
+      "learning_rate": 0.00018186110075785815,
+      "loss": 0.0553,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2739087902361765,
+      "grad_norm": 0.2549150288105011,
+      "learning_rate": 0.00018173686172195307,
+      "loss": 0.0335,
+      "step": 1470
+    },
+    {
+      "epoch": 0.27577211533982393,
+      "grad_norm": 0.6201621890068054,
+      "learning_rate": 0.00018161262268604798,
+      "loss": 0.0336,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2776354404434714,
+      "grad_norm": 0.24882057309150696,
+      "learning_rate": 0.0001814883836501429,
+      "loss": 0.0407,
+      "step": 1490
+    },
+    {
+      "epoch": 0.27949876554711883,
+      "grad_norm": 0.2643592059612274,
+      "learning_rate": 0.00018136414461423781,
+      "loss": 0.0336,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2813620906507663,
+      "grad_norm": 0.2674497067928314,
+      "learning_rate": 0.00018123990557833273,
+      "loss": 0.036,
+      "step": 1510
+    },
+    {
+      "epoch": 0.28322541575441373,
+      "grad_norm": 0.1852717250585556,
+      "learning_rate": 0.00018111566654242765,
+      "loss": 0.0306,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2850887408580612,
+      "grad_norm": 0.16907139122486115,
+      "learning_rate": 0.00018099142750652256,
+      "loss": 0.0309,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2869520659617087,
+      "grad_norm": 0.2268272340297699,
+      "learning_rate": 0.00018086718847061748,
+      "loss": 0.0338,
+      "step": 1540
+    },
+    {
+      "epoch": 0.28881539106535614,
+      "grad_norm": 0.1555815041065216,
+      "learning_rate": 0.0001807429494347124,
+      "loss": 0.0341,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2906787161690036,
+      "grad_norm": 0.23342694342136383,
+      "learning_rate": 0.0001806187103988073,
+      "loss": 0.0301,
+      "step": 1560
+    },
+    {
+      "epoch": 0.29254204127265104,
+      "grad_norm": 0.2339681088924408,
+      "learning_rate": 0.00018049447136290223,
+      "loss": 0.0278,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2944053663762985,
+      "grad_norm": 0.2596907317638397,
+      "learning_rate": 0.00018037023232699714,
+      "loss": 0.0316,
+      "step": 1580
+    },
+    {
+      "epoch": 0.29626869147994594,
+      "grad_norm": 0.1681392937898636,
+      "learning_rate": 0.00018024599329109206,
+      "loss": 0.0384,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2981320165835934,
+      "grad_norm": 0.3269426226615906,
+      "learning_rate": 0.00018012175425518697,
+      "loss": 0.03,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2999953416872409,
+      "grad_norm": 0.37414050102233887,
+      "learning_rate": 0.0001799975152192819,
+      "loss": 0.0377,
+      "step": 1610
+    },
+    {
+      "epoch": 0.30185866679088835,
+      "grad_norm": 0.237156942486763,
+      "learning_rate": 0.0001798732761833768,
+      "loss": 0.0361,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3037219918945358,
+      "grad_norm": 0.1954265832901001,
+      "learning_rate": 0.00017974903714747175,
+      "loss": 0.0458,
+      "step": 1630
+    },
+    {
+      "epoch": 0.30558531699818325,
+      "grad_norm": 0.4432564079761505,
+      "learning_rate": 0.00017962479811156666,
+      "loss": 0.0329,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3074486421018307,
+      "grad_norm": 0.20280279219150543,
+      "learning_rate": 0.00017950055907566158,
+      "loss": 0.0346,
+      "step": 1650
+    },
+    {
+      "epoch": 0.30931196720547816,
+      "grad_norm": 0.18664799630641937,
+      "learning_rate": 0.0001793763200397565,
+      "loss": 0.0325,
+      "step": 1660
+    },
+    {
+      "epoch": 0.31117529230912566,
+      "grad_norm": 0.4096941351890564,
+      "learning_rate": 0.00017925208100385144,
+      "loss": 0.0326,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3130386174127731,
+      "grad_norm": 0.14809350669384003,
+      "learning_rate": 0.00017912784196794635,
+      "loss": 0.0305,
+      "step": 1680
+    },
+    {
+      "epoch": 0.31490194251642056,
+      "grad_norm": 0.1550980508327484,
+      "learning_rate": 0.00017900360293204127,
+      "loss": 0.0419,
+      "step": 1690
+    },
+    {
+      "epoch": 0.316765267620068,
+      "grad_norm": 0.2249947041273117,
+      "learning_rate": 0.00017887936389613619,
+      "loss": 0.0329,
+      "step": 1700
+    },
+    {
+      "epoch": 0.31862859272371546,
+      "grad_norm": 0.28853726387023926,
+      "learning_rate": 0.0001787551248602311,
+      "loss": 0.0395,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3204919178273629,
+      "grad_norm": 0.2440110743045807,
+      "learning_rate": 0.00017863088582432602,
+      "loss": 0.0374,
+      "step": 1720
+    },
+    {
+      "epoch": 0.32235524293101037,
+      "grad_norm": 0.2219010591506958,
+      "learning_rate": 0.00017850664678842093,
+      "loss": 0.0325,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3242185680346579,
+      "grad_norm": 1.0239918231964111,
+      "learning_rate": 0.00017838240775251585,
+      "loss": 0.0354,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3260818931383053,
+      "grad_norm": 0.27604353427886963,
+      "learning_rate": 0.00017825816871661076,
+      "loss": 0.0291,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3279452182419528,
+      "grad_norm": 0.25242915749549866,
+      "learning_rate": 0.00017813392968070568,
+      "loss": 0.0417,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3298085433456002,
+      "grad_norm": 0.1866777390241623,
+      "learning_rate": 0.0001780096906448006,
+      "loss": 0.0315,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3316718684492477,
+      "grad_norm": 0.26006487011909485,
+      "learning_rate": 0.0001778854516088955,
+      "loss": 0.0281,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3335351935528951,
+      "grad_norm": 0.2924884557723999,
+      "learning_rate": 0.00017776121257299043,
+      "loss": 0.0371,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3353985186565426,
+      "grad_norm": 0.1723160594701767,
+      "learning_rate": 0.00017763697353708537,
+      "loss": 0.0344,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3372618437601901,
+      "grad_norm": 0.3166184425354004,
+      "learning_rate": 0.00017751273450118029,
+      "loss": 0.0309,
+      "step": 1810
+    },
+    {
+      "epoch": 0.33912516886383753,
+      "grad_norm": 0.17607541382312775,
+      "learning_rate": 0.0001773884954652752,
+      "loss": 0.0406,
+      "step": 1820
+    },
+    {
+      "epoch": 0.340988493967485,
+      "grad_norm": 0.36102867126464844,
+      "learning_rate": 0.00017726425642937012,
+      "loss": 0.0399,
+      "step": 1830
+    },
+    {
+      "epoch": 0.34285181907113244,
+      "grad_norm": 0.23942948877811432,
+      "learning_rate": 0.00017714001739346503,
+      "loss": 0.0387,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3447151441747799,
+      "grad_norm": 0.33895015716552734,
+      "learning_rate": 0.00017701577835755995,
+      "loss": 0.0347,
+      "step": 1850
+    },
+    {
+      "epoch": 0.34657846927842734,
+      "grad_norm": 0.21242383122444153,
+      "learning_rate": 0.00017689153932165487,
+      "loss": 0.0288,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3484417943820748,
+      "grad_norm": 0.264288067817688,
+      "learning_rate": 0.00017676730028574978,
+      "loss": 0.0379,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3503051194857223,
+      "grad_norm": 0.17818303406238556,
+      "learning_rate": 0.0001766430612498447,
+      "loss": 0.03,
+      "step": 1880
+    },
+    {
+      "epoch": 0.35216844458936974,
+      "grad_norm": 0.2825514078140259,
+      "learning_rate": 0.00017651882221393964,
+      "loss": 0.0275,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3540317696930172,
+      "grad_norm": 0.36010608077049255,
+      "learning_rate": 0.00017639458317803456,
+      "loss": 0.038,
+      "step": 1900
+    },
+    {
+      "epoch": 0.35589509479666465,
+      "grad_norm": 0.22792565822601318,
+      "learning_rate": 0.00017627034414212947,
+      "loss": 0.0351,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3577584199003121,
+      "grad_norm": 0.2936442196369171,
+      "learning_rate": 0.0001761461051062244,
+      "loss": 0.0421,
+      "step": 1920
+    },
+    {
+      "epoch": 0.35962174500395955,
+      "grad_norm": 0.2216373234987259,
+      "learning_rate": 0.0001760218660703193,
+      "loss": 0.027,
+      "step": 1930
+    },
+    {
+      "epoch": 0.361485070107607,
+      "grad_norm": 2.177262783050537,
+      "learning_rate": 0.00017589762703441422,
+      "loss": 0.031,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3633483952112545,
+      "grad_norm": 0.14397388696670532,
+      "learning_rate": 0.00017577338799850913,
+      "loss": 0.0373,
+      "step": 1950
+    },
+    {
+      "epoch": 0.36521172031490196,
+      "grad_norm": 0.11747460812330246,
+      "learning_rate": 0.00017564914896260405,
+      "loss": 0.0284,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3670750454185494,
+      "grad_norm": 0.23211225867271423,
+      "learning_rate": 0.000175524909926699,
+      "loss": 0.0307,
+      "step": 1970
+    },
+    {
+      "epoch": 0.36893837052219686,
+      "grad_norm": 0.6417592763900757,
+      "learning_rate": 0.0001754006708907939,
+      "loss": 0.0291,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3708016956258443,
+      "grad_norm": 0.22125494480133057,
+      "learning_rate": 0.00017527643185488882,
+      "loss": 0.0313,
+      "step": 1990
+    },
+    {
+      "epoch": 0.37266502072949176,
+      "grad_norm": 0.22006738185882568,
+      "learning_rate": 0.00017515219281898374,
+      "loss": 0.0343,
+      "step": 2000
+    },
+    {
+      "epoch": 0.37452834583313926,
+      "grad_norm": 0.4610983729362488,
+      "learning_rate": 0.00017502795378307866,
+      "loss": 0.0375,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3763916709367867,
+      "grad_norm": 0.29008451104164124,
+      "learning_rate": 0.00017490371474717357,
+      "loss": 0.0347,
+      "step": 2020
+    },
+    {
+      "epoch": 0.37825499604043417,
+      "grad_norm": 0.20057004690170288,
+      "learning_rate": 0.0001747794757112685,
+      "loss": 0.0305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3801183211440816,
+      "grad_norm": 0.22764652967453003,
+      "learning_rate": 0.0001746552366753634,
+      "loss": 0.0317,
+      "step": 2040
+    },
+    {
+      "epoch": 0.38198164624772907,
+      "grad_norm": 0.19526733458042145,
+      "learning_rate": 0.00017453099763945832,
+      "loss": 0.0394,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3838449713513765,
+      "grad_norm": 0.2324579507112503,
+      "learning_rate": 0.00017440675860355324,
+      "loss": 0.0248,
+      "step": 2060
+    },
+    {
+      "epoch": 0.38570829645502397,
+      "grad_norm": 0.21112307906150818,
+      "learning_rate": 0.00017428251956764815,
+      "loss": 0.0323,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3875716215586715,
+      "grad_norm": 0.3266056478023529,
+      "learning_rate": 0.00017415828053174307,
+      "loss": 0.0297,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3894349466623189,
+      "grad_norm": 0.13637982308864594,
+      "learning_rate": 0.00017403404149583798,
+      "loss": 0.0262,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3912982717659664,
+      "grad_norm": 0.16478992998600006,
+      "learning_rate": 0.00017390980245993293,
+      "loss": 0.0313,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3931615968696138,
+      "grad_norm": 0.16533003747463226,
+      "learning_rate": 0.00017378556342402784,
+      "loss": 0.0291,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3950249219732613,
+      "grad_norm": 0.2503752112388611,
+      "learning_rate": 0.00017366132438812276,
+      "loss": 0.0293,
+      "step": 2120
+    },
+    {
+      "epoch": 0.39688824707690873,
+      "grad_norm": 0.36518189311027527,
+      "learning_rate": 0.00017353708535221767,
+      "loss": 0.0377,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3987515721805562,
+      "grad_norm": 0.1784435659646988,
+      "learning_rate": 0.0001734128463163126,
+      "loss": 0.0217,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4006148972842037,
+      "grad_norm": 0.2254277467727661,
+      "learning_rate": 0.00017328860728040753,
+      "loss": 0.0334,
+      "step": 2150
+    },
+    {
+      "epoch": 0.40247822238785114,
+      "grad_norm": 0.23453503847122192,
+      "learning_rate": 0.00017316436824450245,
+      "loss": 0.0342,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4043415474914986,
+      "grad_norm": 0.17916861176490784,
+      "learning_rate": 0.00017304012920859736,
+      "loss": 0.0366,
+      "step": 2170
+    },
+    {
+      "epoch": 0.40620487259514604,
+      "grad_norm": 0.20418667793273926,
+      "learning_rate": 0.00017291589017269228,
+      "loss": 0.028,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4080681976987935,
+      "grad_norm": 0.32335731387138367,
+      "learning_rate": 0.0001727916511367872,
+      "loss": 0.0336,
+      "step": 2190
+    },
+    {
+      "epoch": 0.40993152280244094,
+      "grad_norm": 0.19899824261665344,
+      "learning_rate": 0.0001726674121008821,
+      "loss": 0.0398,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4117948479060884,
+      "grad_norm": 0.22357973456382751,
+      "learning_rate": 0.00017254317306497703,
+      "loss": 0.0315,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4136581730097359,
+      "grad_norm": 0.12390164285898209,
+      "learning_rate": 0.00017241893402907194,
+      "loss": 0.0265,
+      "step": 2220
+    },
+    {
+      "epoch": 0.41552149811338335,
+      "grad_norm": 0.2476750761270523,
+      "learning_rate": 0.00017229469499316686,
+      "loss": 0.0405,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4173848232170308,
+      "grad_norm": 0.24649333953857422,
+      "learning_rate": 0.00017217045595726177,
+      "loss": 0.0327,
+      "step": 2240
+    },
+    {
+      "epoch": 0.41924814832067825,
+      "grad_norm": 0.30495890974998474,
+      "learning_rate": 0.0001720462169213567,
+      "loss": 0.0319,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4211114734243257,
+      "grad_norm": 0.2112802118062973,
+      "learning_rate": 0.0001719219778854516,
+      "loss": 0.0294,
+      "step": 2260
+    },
+    {
+      "epoch": 0.42297479852797315,
+      "grad_norm": 0.25762444734573364,
+      "learning_rate": 0.00017179773884954652,
+      "loss": 0.0291,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4248381236316206,
+      "grad_norm": 0.3094237744808197,
+      "learning_rate": 0.00017167349981364144,
+      "loss": 0.0391,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4267014487352681,
+      "grad_norm": 0.17324984073638916,
+      "learning_rate": 0.00017154926077773635,
+      "loss": 0.0256,
+      "step": 2290
+    },
+    {
+      "epoch": 0.42856477383891556,
+      "grad_norm": 0.12840083241462708,
+      "learning_rate": 0.00017142502174183127,
+      "loss": 0.04,
+      "step": 2300
+    },
+    {
+      "epoch": 0.430428098942563,
+      "grad_norm": 0.23374256491661072,
+      "learning_rate": 0.0001713007827059262,
+      "loss": 0.0277,
+      "step": 2310
+    },
+    {
+      "epoch": 0.43229142404621046,
+      "grad_norm": 0.27365607023239136,
+      "learning_rate": 0.00017117654367002113,
+      "loss": 0.0365,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4341547491498579,
+      "grad_norm": 0.12370350956916809,
+      "learning_rate": 0.00017105230463411604,
+      "loss": 0.0449,
+      "step": 2330
+    },
+    {
+      "epoch": 0.43601807425350536,
+      "grad_norm": 0.3122682571411133,
+      "learning_rate": 0.000170928065598211,
+      "loss": 0.0263,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4378813993571528,
+      "grad_norm": 0.20058439671993256,
+      "learning_rate": 0.0001708038265623059,
+      "loss": 0.0296,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4397447244608003,
+      "grad_norm": 0.16738611459732056,
+      "learning_rate": 0.00017067958752640082,
+      "loss": 0.0376,
+      "step": 2360
+    },
+    {
+      "epoch": 0.44160804956444777,
+      "grad_norm": 0.11145570129156113,
+      "learning_rate": 0.00017055534849049573,
+      "loss": 0.0227,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4434713746680952,
+      "grad_norm": 0.2827621102333069,
+      "learning_rate": 0.00017043110945459065,
+      "loss": 0.0281,
+      "step": 2380
+    },
+    {
+      "epoch": 0.44533469977174267,
+      "grad_norm": 0.13350103795528412,
+      "learning_rate": 0.00017030687041868557,
+      "loss": 0.038,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4471980248753901,
+      "grad_norm": 0.10742458701133728,
+      "learning_rate": 0.00017018263138278048,
+      "loss": 0.0295,
+      "step": 2400
+    },
+    {
+      "epoch": 0.44906134997903757,
+      "grad_norm": 0.21985341608524323,
+      "learning_rate": 0.0001700583923468754,
+      "loss": 0.0301,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4509246750826851,
+      "grad_norm": 0.22676043212413788,
+      "learning_rate": 0.0001699341533109703,
+      "loss": 0.0323,
+      "step": 2420
+    },
+    {
+      "epoch": 0.45278800018633253,
+      "grad_norm": 0.39583566784858704,
+      "learning_rate": 0.00016980991427506523,
+      "loss": 0.0283,
+      "step": 2430
+    },
+    {
+      "epoch": 0.45465132528998,
+      "grad_norm": 0.33214402198791504,
+      "learning_rate": 0.00016968567523916014,
+      "loss": 0.0328,
+      "step": 2440
+    },
+    {
+      "epoch": 0.45651465039362743,
+      "grad_norm": 0.19369973242282867,
+      "learning_rate": 0.00016956143620325506,
+      "loss": 0.0251,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4583779754972749,
+      "grad_norm": 0.1785402148962021,
+      "learning_rate": 0.00016943719716734998,
+      "loss": 0.0338,
+      "step": 2460
+    },
+    {
+      "epoch": 0.46024130060092233,
+      "grad_norm": 0.30956077575683594,
+      "learning_rate": 0.0001693129581314449,
+      "loss": 0.0251,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4621046257045698,
+      "grad_norm": 0.17932891845703125,
+      "learning_rate": 0.0001691887190955398,
+      "loss": 0.0291,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4639679508082173,
+      "grad_norm": 0.584335446357727,
+      "learning_rate": 0.00016906448005963475,
+      "loss": 0.0223,
+      "step": 2490
+    },
+    {
+      "epoch": 0.46583127591186474,
+      "grad_norm": 0.2823520004749298,
+      "learning_rate": 0.00016894024102372967,
+      "loss": 0.0289,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4676946010155122,
+      "grad_norm": 0.12712427973747253,
+      "learning_rate": 0.00016881600198782458,
+      "loss": 0.0321,
+      "step": 2510
+    },
+    {
+      "epoch": 0.46955792611915964,
+      "grad_norm": 0.18848009407520294,
+      "learning_rate": 0.0001686917629519195,
+      "loss": 0.0183,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4714212512228071,
+      "grad_norm": 0.5352602601051331,
+      "learning_rate": 0.00016856752391601441,
+      "loss": 0.0417,
+      "step": 2530
+    },
+    {
+      "epoch": 0.47328457632645454,
+      "grad_norm": 0.1589186191558838,
+      "learning_rate": 0.00016844328488010933,
+      "loss": 0.0258,
+      "step": 2540
+    },
+    {
+      "epoch": 0.475147901430102,
+      "grad_norm": 0.4766451120376587,
+      "learning_rate": 0.00016831904584420427,
+      "loss": 0.0224,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4770112265337495,
+      "grad_norm": 0.11691979318857193,
+      "learning_rate": 0.0001681948068082992,
+      "loss": 0.0264,
+      "step": 2560
+    },
+    {
+      "epoch": 0.47887455163739695,
+      "grad_norm": 0.19135166704654694,
+      "learning_rate": 0.0001680705677723941,
+      "loss": 0.0311,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4807378767410444,
+      "grad_norm": 0.1519075334072113,
+      "learning_rate": 0.00016794632873648902,
+      "loss": 0.0316,
+      "step": 2580
+    },
+    {
+      "epoch": 0.48260120184469185,
+      "grad_norm": 0.211242213845253,
+      "learning_rate": 0.00016782208970058394,
+      "loss": 0.0267,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4844645269483393,
+      "grad_norm": 0.1120869442820549,
+      "learning_rate": 0.00016769785066467885,
+      "loss": 0.0275,
+      "step": 2600
+    },
+    {
+      "epoch": 0.48632785205198675,
+      "grad_norm": 0.12016676366329193,
+      "learning_rate": 0.00016757361162877377,
+      "loss": 0.0269,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4881911771556342,
+      "grad_norm": 0.17459681630134583,
+      "learning_rate": 0.00016744937259286868,
+      "loss": 0.0275,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4900545022592817,
+      "grad_norm": 0.4547879993915558,
+      "learning_rate": 0.0001673251335569636,
+      "loss": 0.0445,
+      "step": 2630
+    },
+    {
+      "epoch": 0.49191782736292916,
+      "grad_norm": 0.20323753356933594,
+      "learning_rate": 0.00016720089452105851,
+      "loss": 0.0274,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4937811524665766,
+      "grad_norm": 0.14843535423278809,
+      "learning_rate": 0.00016707665548515343,
+      "loss": 0.0261,
+      "step": 2650
+    },
+    {
+      "epoch": 0.49564447757022406,
+      "grad_norm": 0.3513728678226471,
+      "learning_rate": 0.00016695241644924837,
+      "loss": 0.0292,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4975078026738715,
+      "grad_norm": 0.10806536674499512,
+      "learning_rate": 0.0001668281774133433,
+      "loss": 0.0263,
+      "step": 2670
+    },
+    {
+      "epoch": 0.49937112777751896,
+      "grad_norm": 0.2897528409957886,
+      "learning_rate": 0.0001667039383774382,
+      "loss": 0.0265,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5012344528811664,
+      "grad_norm": 0.26667869091033936,
+      "learning_rate": 0.00016657969934153312,
+      "loss": 0.0319,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5030977779848139,
+      "grad_norm": 0.21708862483501434,
+      "learning_rate": 0.00016645546030562804,
+      "loss": 0.0254,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5049611030884613,
+      "grad_norm": 0.1377273052930832,
+      "learning_rate": 0.00016633122126972295,
+      "loss": 0.021,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5068244281921088,
+      "grad_norm": 0.18345075845718384,
+      "learning_rate": 0.00016620698223381787,
+      "loss": 0.0329,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5086877532957563,
+      "grad_norm": 0.1460709273815155,
+      "learning_rate": 0.00016608274319791278,
+      "loss": 0.0263,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5105510783994037,
+      "grad_norm": 0.32654863595962524,
+      "learning_rate": 0.0001659585041620077,
+      "loss": 0.0226,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5124144035030512,
+      "grad_norm": 0.166800856590271,
+      "learning_rate": 0.00016583426512610262,
+      "loss": 0.0235,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5142777286066986,
+      "grad_norm": 0.15170325338840485,
+      "learning_rate": 0.00016571002609019756,
+      "loss": 0.0267,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5161410537103461,
+      "grad_norm": 0.2023041844367981,
+      "learning_rate": 0.00016558578705429247,
+      "loss": 0.0218,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5180043788139935,
+      "grad_norm": 0.32125502824783325,
+      "learning_rate": 0.0001654615480183874,
+      "loss": 0.0295,
+      "step": 2780
+    },
+    {
+      "epoch": 0.519867703917641,
+      "grad_norm": 0.21208274364471436,
+      "learning_rate": 0.0001653373089824823,
+      "loss": 0.0242,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5217310290212885,
+      "grad_norm": 0.22862495481967926,
+      "learning_rate": 0.00016521306994657722,
+      "loss": 0.038,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5235943541249359,
+      "grad_norm": 0.1913980096578598,
+      "learning_rate": 0.00016508883091067214,
+      "loss": 0.0308,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5254576792285834,
+      "grad_norm": 0.14556388556957245,
+      "learning_rate": 0.00016496459187476705,
+      "loss": 0.0329,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5273210043322308,
+      "grad_norm": 0.19217199087142944,
+      "learning_rate": 0.000164840352838862,
+      "loss": 0.0253,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5291843294358783,
+      "grad_norm": 0.11398417502641678,
+      "learning_rate": 0.0001647161138029569,
+      "loss": 0.0205,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5310476545395257,
+      "grad_norm": 0.16399814188480377,
+      "learning_rate": 0.00016459187476705183,
+      "loss": 0.0222,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5329109796431732,
+      "grad_norm": 0.4391205906867981,
+      "learning_rate": 0.00016446763573114674,
+      "loss": 0.0288,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5347743047468208,
+      "grad_norm": 0.1345457136631012,
+      "learning_rate": 0.00016434339669524166,
+      "loss": 0.0236,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5366376298504681,
+      "grad_norm": 0.18585018813610077,
+      "learning_rate": 0.00016421915765933658,
+      "loss": 0.0262,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5385009549541157,
+      "grad_norm": 0.26273420453071594,
+      "learning_rate": 0.0001640949186234315,
+      "loss": 0.0278,
+      "step": 2890
+    },
+    {
+      "epoch": 0.540364280057763,
+      "grad_norm": 0.21691930294036865,
+      "learning_rate": 0.0001639706795875264,
+      "loss": 0.0246,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5422276051614106,
+      "grad_norm": 0.21845707297325134,
+      "learning_rate": 0.00016384644055162132,
+      "loss": 0.0361,
+      "step": 2910
+    },
+    {
+      "epoch": 0.544090930265058,
+      "grad_norm": 0.16350382566452026,
+      "learning_rate": 0.00016372220151571624,
+      "loss": 0.0252,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5459542553687055,
+      "grad_norm": 0.16099347174167633,
+      "learning_rate": 0.00016359796247981115,
+      "loss": 0.0219,
+      "step": 2930
+    },
+    {
+      "epoch": 0.547817580472353,
+      "grad_norm": 0.16874344646930695,
+      "learning_rate": 0.00016347372344390607,
+      "loss": 0.0204,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5496809055760004,
+      "grad_norm": 0.45683175325393677,
+      "learning_rate": 0.00016334948440800099,
+      "loss": 0.0257,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5515442306796479,
+      "grad_norm": 0.3147335648536682,
+      "learning_rate": 0.0001632252453720959,
+      "loss": 0.0238,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5534075557832953,
+      "grad_norm": 0.14535823464393616,
+      "learning_rate": 0.00016310100633619082,
+      "loss": 0.0317,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5552708808869428,
+      "grad_norm": 0.17752012610435486,
+      "learning_rate": 0.00016297676730028576,
+      "loss": 0.0332,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5571342059905902,
+      "grad_norm": 0.1567634791135788,
+      "learning_rate": 0.00016285252826438068,
+      "loss": 0.026,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5589975310942377,
+      "grad_norm": 0.12023455649614334,
+      "learning_rate": 0.0001627282892284756,
+      "loss": 0.0244,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5608608561978852,
+      "grad_norm": 0.20631061494350433,
+      "learning_rate": 0.00016260405019257054,
+      "loss": 0.0289,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5627241813015326,
+      "grad_norm": 0.15248391032218933,
+      "learning_rate": 0.00016247981115666545,
+      "loss": 0.0245,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5645875064051801,
+      "grad_norm": 0.1587265431880951,
+      "learning_rate": 0.00016235557212076037,
+      "loss": 0.0375,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5664508315088275,
+      "grad_norm": 0.12269464135169983,
+      "learning_rate": 0.00016223133308485528,
+      "loss": 0.0212,
+      "step": 3040
+    },
+    {
+      "epoch": 0.568314156612475,
+      "grad_norm": 0.1672634780406952,
+      "learning_rate": 0.0001621070940489502,
+      "loss": 0.0285,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5701774817161224,
+      "grad_norm": 0.1095886081457138,
+      "learning_rate": 0.00016198285501304511,
+      "loss": 0.0245,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5720408068197699,
+      "grad_norm": 0.19466912746429443,
+      "learning_rate": 0.00016185861597714003,
+      "loss": 0.03,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5739041319234174,
+      "grad_norm": 0.13806433975696564,
+      "learning_rate": 0.00016173437694123495,
+      "loss": 0.0307,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5757674570270648,
+      "grad_norm": 0.08115004748106003,
+      "learning_rate": 0.00016161013790532986,
+      "loss": 0.0239,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5776307821307123,
+      "grad_norm": 0.13298851251602173,
+      "learning_rate": 0.00016148589886942478,
+      "loss": 0.0298,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5794941072343597,
+      "grad_norm": 0.1280878335237503,
+      "learning_rate": 0.0001613616598335197,
+      "loss": 0.0222,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5813574323380072,
+      "grad_norm": 0.18590818345546722,
+      "learning_rate": 0.0001612374207976146,
+      "loss": 0.0247,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5832207574416546,
+      "grad_norm": 0.1540619432926178,
+      "learning_rate": 0.00016111318176170953,
+      "loss": 0.028,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5850840825453021,
+      "grad_norm": 0.18428578972816467,
+      "learning_rate": 0.00016098894272580444,
+      "loss": 0.0297,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5869474076489496,
+      "grad_norm": 0.155143141746521,
+      "learning_rate": 0.00016086470368989936,
+      "loss": 0.0252,
+      "step": 3150
+    },
+    {
+      "epoch": 0.588810732752597,
+      "grad_norm": 0.10080347210168839,
+      "learning_rate": 0.00016074046465399427,
+      "loss": 0.032,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5906740578562445,
+      "grad_norm": 0.12629786133766174,
+      "learning_rate": 0.00016061622561808922,
+      "loss": 0.0233,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5925373829598919,
+      "grad_norm": 0.11422615498304367,
+      "learning_rate": 0.00016049198658218413,
+      "loss": 0.0226,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5944007080635394,
+      "grad_norm": 0.34021082520484924,
+      "learning_rate": 0.00016036774754627905,
+      "loss": 0.0282,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5962640331671868,
+      "grad_norm": 0.1320790946483612,
+      "learning_rate": 0.00016024350851037396,
+      "loss": 0.0285,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5981273582708343,
+      "grad_norm": 0.20607562363147736,
+      "learning_rate": 0.00016011926947446888,
+      "loss": 0.0373,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5999906833744818,
+      "grad_norm": 0.17851243913173676,
+      "learning_rate": 0.00015999503043856382,
+      "loss": 0.0227,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6018540084781292,
+      "grad_norm": 0.118630051612854,
+      "learning_rate": 0.00015987079140265874,
+      "loss": 0.0261,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6037173335817767,
+      "grad_norm": 0.1931089609861374,
+      "learning_rate": 0.00015974655236675365,
+      "loss": 0.0242,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6055806586854241,
+      "grad_norm": 0.1010395735502243,
+      "learning_rate": 0.00015962231333084857,
+      "loss": 0.0472,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6074439837890716,
+      "grad_norm": 0.3368585705757141,
+      "learning_rate": 0.00015949807429494348,
+      "loss": 0.0236,
+      "step": 3260
+    },
+    {
+      "epoch": 0.609307308892719,
+      "grad_norm": 0.23170095682144165,
+      "learning_rate": 0.0001593738352590384,
+      "loss": 0.0188,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6111706339963665,
+      "grad_norm": 0.20712672173976898,
+      "learning_rate": 0.00015924959622313332,
+      "loss": 0.0342,
+      "step": 3280
+    },
+    {
+      "epoch": 0.613033959100014,
+      "grad_norm": 0.2359304428100586,
+      "learning_rate": 0.00015912535718722823,
+      "loss": 0.0226,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6148972842036614,
+      "grad_norm": 0.19130489230155945,
+      "learning_rate": 0.00015900111815132315,
+      "loss": 0.0277,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6167606093073089,
+      "grad_norm": 0.10351112484931946,
+      "learning_rate": 0.00015887687911541806,
+      "loss": 0.0246,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6186239344109563,
+      "grad_norm": 0.15079782903194427,
+      "learning_rate": 0.00015875264007951298,
+      "loss": 0.0247,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6204872595146038,
+      "grad_norm": 0.13707318902015686,
+      "learning_rate": 0.0001586284010436079,
+      "loss": 0.0246,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6223505846182513,
+      "grad_norm": 0.3334473669528961,
+      "learning_rate": 0.00015850416200770284,
+      "loss": 0.026,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6242139097218987,
+      "grad_norm": 0.14594854414463043,
+      "learning_rate": 0.00015837992297179775,
+      "loss": 0.0231,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6260772348255462,
+      "grad_norm": 0.17928946018218994,
+      "learning_rate": 0.00015825568393589267,
+      "loss": 0.0243,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6279405599291936,
+      "grad_norm": 0.18053315579891205,
+      "learning_rate": 0.00015813144489998759,
+      "loss": 0.0184,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6298038850328411,
+      "grad_norm": 0.1368819922208786,
+      "learning_rate": 0.0001580072058640825,
+      "loss": 0.0229,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6316672101364885,
+      "grad_norm": 0.1753019094467163,
+      "learning_rate": 0.00015788296682817742,
+      "loss": 0.0244,
+      "step": 3390
+    },
+    {
+      "epoch": 0.633530535240136,
+      "grad_norm": 0.189280703663826,
+      "learning_rate": 0.00015775872779227233,
+      "loss": 0.023,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6353938603437835,
+      "grad_norm": 0.10799846053123474,
+      "learning_rate": 0.00015763448875636725,
+      "loss": 0.0232,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6372571854474309,
+      "grad_norm": 0.14953063428401947,
+      "learning_rate": 0.00015751024972046216,
+      "loss": 0.0242,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6391205105510784,
+      "grad_norm": 0.46299856901168823,
+      "learning_rate": 0.0001573860106845571,
+      "loss": 0.0346,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6409838356547258,
+      "grad_norm": 0.21256397664546967,
+      "learning_rate": 0.00015726177164865202,
+      "loss": 0.0284,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6428471607583733,
+      "grad_norm": 0.21798713505268097,
+      "learning_rate": 0.00015713753261274694,
+      "loss": 0.0299,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6447104858620207,
+      "grad_norm": 0.16905297338962555,
+      "learning_rate": 0.00015701329357684186,
+      "loss": 0.0206,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6465738109656682,
+      "grad_norm": 0.11575949192047119,
+      "learning_rate": 0.00015688905454093677,
+      "loss": 0.0247,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6484371360693157,
+      "grad_norm": 0.11626104265451431,
+      "learning_rate": 0.0001567648155050317,
+      "loss": 0.0233,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6503004611729631,
+      "grad_norm": 0.214884415268898,
+      "learning_rate": 0.0001566405764691266,
+      "loss": 0.0177,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6521637862766106,
+      "grad_norm": 0.14941707253456116,
+      "learning_rate": 0.00015651633743322152,
+      "loss": 0.0218,
+      "step": 3500
+    },
+    {
+      "epoch": 0.654027111380258,
+      "grad_norm": 0.1662212312221527,
+      "learning_rate": 0.00015639209839731643,
+      "loss": 0.0262,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6558904364839055,
+      "grad_norm": 0.1752852499485016,
+      "learning_rate": 0.00015626785936141138,
+      "loss": 0.024,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6577537615875529,
+      "grad_norm": 0.15072235465049744,
+      "learning_rate": 0.0001561436203255063,
+      "loss": 0.0234,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6596170866912004,
+      "grad_norm": 0.18537260591983795,
+      "learning_rate": 0.0001560193812896012,
+      "loss": 0.0268,
+      "step": 3540
+    },
+    {
+      "epoch": 0.661480411794848,
+      "grad_norm": 0.10309719294309616,
+      "learning_rate": 0.00015589514225369612,
+      "loss": 0.029,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6633437368984954,
+      "grad_norm": 0.14345987141132355,
+      "learning_rate": 0.00015577090321779104,
+      "loss": 0.0192,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6652070620021429,
+      "grad_norm": 0.12496069073677063,
+      "learning_rate": 0.00015564666418188596,
+      "loss": 0.0173,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6670703871057903,
+      "grad_norm": 0.15995663404464722,
+      "learning_rate": 0.00015552242514598087,
+      "loss": 0.02,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6689337122094378,
+      "grad_norm": 0.1138169914484024,
+      "learning_rate": 0.0001553981861100758,
+      "loss": 0.0156,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6707970373130852,
+      "grad_norm": 0.2230171263217926,
+      "learning_rate": 0.0001552739470741707,
+      "loss": 0.0297,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6726603624167327,
+      "grad_norm": 0.08599012345075607,
+      "learning_rate": 0.00015514970803826562,
+      "loss": 0.0292,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6745236875203802,
+      "grad_norm": 0.1836289018392563,
+      "learning_rate": 0.00015502546900236054,
+      "loss": 0.0229,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6763870126240276,
+      "grad_norm": 0.20711427927017212,
+      "learning_rate": 0.00015490122996645545,
+      "loss": 0.0206,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6782503377276751,
+      "grad_norm": 0.23631249368190765,
+      "learning_rate": 0.0001547769909305504,
+      "loss": 0.0188,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6801136628313225,
+      "grad_norm": 0.28623124957084656,
+      "learning_rate": 0.0001546527518946453,
+      "loss": 0.0279,
+      "step": 3650
+    },
+    {
+      "epoch": 0.68197698793497,
+      "grad_norm": 0.16205276548862457,
+      "learning_rate": 0.00015452851285874023,
+      "loss": 0.0205,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6838403130386174,
+      "grad_norm": 0.3658526539802551,
+      "learning_rate": 0.00015440427382283514,
+      "loss": 0.0345,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6857036381422649,
+      "grad_norm": 0.19878637790679932,
+      "learning_rate": 0.00015428003478693006,
+      "loss": 0.0188,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6875669632459124,
+      "grad_norm": 0.148567795753479,
+      "learning_rate": 0.000154155795751025,
+      "loss": 0.018,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6894302883495598,
+      "grad_norm": 0.0717538595199585,
+      "learning_rate": 0.00015403155671511992,
+      "loss": 0.0187,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6912936134532073,
+      "grad_norm": 0.18270643055438995,
+      "learning_rate": 0.00015390731767921483,
+      "loss": 0.0323,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6931569385568547,
+      "grad_norm": 0.10219421982765198,
+      "learning_rate": 0.00015378307864330975,
+      "loss": 0.0207,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6950202636605022,
+      "grad_norm": 0.15821602940559387,
+      "learning_rate": 0.00015365883960740466,
+      "loss": 0.0324,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6968835887641496,
+      "grad_norm": 0.23627109825611115,
+      "learning_rate": 0.00015353460057149958,
+      "loss": 0.0237,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6987469138677971,
+      "grad_norm": 0.1554834395647049,
+      "learning_rate": 0.0001534103615355945,
+      "loss": 0.0254,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7006102389714446,
+      "grad_norm": 0.13305582106113434,
+      "learning_rate": 0.0001532861224996894,
+      "loss": 0.021,
+      "step": 3760
+    },
+    {
+      "epoch": 0.702473564075092,
+      "grad_norm": 0.11506448686122894,
+      "learning_rate": 0.00015316188346378433,
+      "loss": 0.0342,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7043368891787395,
+      "grad_norm": 0.33708521723747253,
+      "learning_rate": 0.00015303764442787924,
+      "loss": 0.0231,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7062002142823869,
+      "grad_norm": 0.15879079699516296,
+      "learning_rate": 0.00015291340539197416,
+      "loss": 0.0318,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7080635393860344,
+      "grad_norm": 0.13867144286632538,
+      "learning_rate": 0.00015278916635606907,
+      "loss": 0.0282,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7099268644896818,
+      "grad_norm": 0.1752207726240158,
+      "learning_rate": 0.000152664927320164,
+      "loss": 0.0259,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7117901895933293,
+      "grad_norm": 0.1792452335357666,
+      "learning_rate": 0.0001525406882842589,
+      "loss": 0.0285,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7136535146969768,
+      "grad_norm": 0.14697203040122986,
+      "learning_rate": 0.00015241644924835382,
+      "loss": 0.0277,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7155168398006242,
+      "grad_norm": 0.1709863245487213,
+      "learning_rate": 0.00015229221021244874,
+      "loss": 0.0226,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7173801649042717,
+      "grad_norm": 0.2217591106891632,
+      "learning_rate": 0.00015216797117654365,
+      "loss": 0.0214,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7192434900079191,
+      "grad_norm": 0.13732624053955078,
+      "learning_rate": 0.0001520437321406386,
+      "loss": 0.0227,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7211068151115666,
+      "grad_norm": 0.1472344845533371,
+      "learning_rate": 0.0001519194931047335,
+      "loss": 0.0196,
+      "step": 3870
+    },
+    {
+      "epoch": 0.722970140215214,
+      "grad_norm": 0.26150020956993103,
+      "learning_rate": 0.00015179525406882843,
+      "loss": 0.0278,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7248334653188615,
+      "grad_norm": 0.12479466199874878,
+      "learning_rate": 0.00015167101503292337,
+      "loss": 0.0277,
+      "step": 3890
+    },
+    {
+      "epoch": 0.726696790422509,
+      "grad_norm": 0.09242439270019531,
+      "learning_rate": 0.00015154677599701829,
+      "loss": 0.0209,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7285601155261564,
+      "grad_norm": 0.1810086965560913,
+      "learning_rate": 0.0001514225369611132,
+      "loss": 0.0228,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7304234406298039,
+      "grad_norm": 0.10720740258693695,
+      "learning_rate": 0.00015129829792520812,
+      "loss": 0.0179,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7322867657334513,
+      "grad_norm": 0.12426480650901794,
+      "learning_rate": 0.00015117405888930303,
+      "loss": 0.0203,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7341500908370988,
+      "grad_norm": 0.11750097572803497,
+      "learning_rate": 0.00015104981985339795,
+      "loss": 0.0211,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7360134159407462,
+      "grad_norm": 0.12620018422603607,
+      "learning_rate": 0.00015092558081749287,
+      "loss": 0.0186,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7378767410443937,
+      "grad_norm": 0.13408900797367096,
+      "learning_rate": 0.00015080134178158778,
+      "loss": 0.0271,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7397400661480412,
+      "grad_norm": 0.12148467451334,
+      "learning_rate": 0.0001506771027456827,
+      "loss": 0.0214,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7416033912516886,
+      "grad_norm": 0.13083823025226593,
+      "learning_rate": 0.0001505528637097776,
+      "loss": 0.0186,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7434667163553361,
+      "grad_norm": 0.10845719277858734,
+      "learning_rate": 0.00015042862467387253,
+      "loss": 0.0233,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7453300414589835,
+      "grad_norm": 0.14539609849452972,
+      "learning_rate": 0.00015030438563796744,
+      "loss": 0.0267,
+      "step": 4000
+    },
+    {
+      "epoch": 0.747193366562631,
+      "grad_norm": 0.1348503679037094,
+      "learning_rate": 0.00015018014660206236,
+      "loss": 0.0264,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7490566916662785,
+      "grad_norm": 0.12877729535102844,
+      "learning_rate": 0.00015005590756615728,
+      "loss": 0.0224,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7509200167699259,
+      "grad_norm": 0.3890345096588135,
+      "learning_rate": 0.00014993166853025222,
+      "loss": 0.0212,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7527833418735734,
+      "grad_norm": 0.21587280929088593,
+      "learning_rate": 0.00014980742949434713,
+      "loss": 0.0295,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7546466669772208,
+      "grad_norm": 0.10658788681030273,
+      "learning_rate": 0.00014968319045844205,
+      "loss": 0.0189,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7565099920808683,
+      "grad_norm": 0.10835613310337067,
+      "learning_rate": 0.00014955895142253697,
+      "loss": 0.0233,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7583733171845157,
+      "grad_norm": 0.22986513376235962,
+      "learning_rate": 0.00014943471238663188,
+      "loss": 0.0272,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7602366422881632,
+      "grad_norm": 0.11246643960475922,
+      "learning_rate": 0.0001493104733507268,
+      "loss": 0.021,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7620999673918107,
+      "grad_norm": 0.16520898044109344,
+      "learning_rate": 0.00014918623431482171,
+      "loss": 0.0192,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7639632924954581,
+      "grad_norm": 0.40314796566963196,
+      "learning_rate": 0.00014906199527891666,
+      "loss": 0.0397,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7658266175991056,
+      "grad_norm": 0.0877520963549614,
+      "learning_rate": 0.00014893775624301157,
+      "loss": 0.0253,
+      "step": 4110
+    },
+    {
+      "epoch": 0.767689942702753,
+      "grad_norm": 0.16271091997623444,
+      "learning_rate": 0.0001488135172071065,
+      "loss": 0.022,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7695532678064005,
+      "grad_norm": 0.21588853001594543,
+      "learning_rate": 0.0001486892781712014,
+      "loss": 0.023,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7714165929100479,
+      "grad_norm": 0.2022632360458374,
+      "learning_rate": 0.00014856503913529632,
+      "loss": 0.0188,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7732799180136954,
+      "grad_norm": 0.11369698494672775,
+      "learning_rate": 0.00014844080009939124,
+      "loss": 0.0278,
+      "step": 4150
+    },
+    {
+      "epoch": 0.775143243117343,
+      "grad_norm": 0.4832044243812561,
+      "learning_rate": 0.00014831656106348615,
+      "loss": 0.0239,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7770065682209903,
+      "grad_norm": 0.1868063062429428,
+      "learning_rate": 0.00014819232202758107,
+      "loss": 0.0256,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7788698933246379,
+      "grad_norm": 0.1666928231716156,
+      "learning_rate": 0.00014806808299167598,
+      "loss": 0.0267,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7807332184282852,
+      "grad_norm": 0.14221766591072083,
+      "learning_rate": 0.0001479438439557709,
+      "loss": 0.0174,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7825965435319328,
+      "grad_norm": 0.11577267944812775,
+      "learning_rate": 0.00014781960491986584,
+      "loss": 0.0243,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7844598686355801,
+      "grad_norm": 0.1612447202205658,
+      "learning_rate": 0.00014769536588396076,
+      "loss": 0.0221,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7863231937392277,
+      "grad_norm": 0.11659658700227737,
+      "learning_rate": 0.00014757112684805567,
+      "loss": 0.0433,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7881865188428752,
+      "grad_norm": 0.31002146005630493,
+      "learning_rate": 0.0001474468878121506,
+      "loss": 0.0232,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7900498439465226,
+      "grad_norm": 0.6392256021499634,
+      "learning_rate": 0.0001473226487762455,
+      "loss": 0.0297,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7919131690501701,
+      "grad_norm": 0.1782752126455307,
+      "learning_rate": 0.00014719840974034042,
+      "loss": 0.0183,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7937764941538175,
+      "grad_norm": 0.1695147305727005,
+      "learning_rate": 0.00014707417070443534,
+      "loss": 0.0201,
+      "step": 4260
+    },
+    {
+      "epoch": 0.795639819257465,
+      "grad_norm": 0.19251610338687897,
+      "learning_rate": 0.00014694993166853025,
+      "loss": 0.0222,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7975031443611124,
+      "grad_norm": 0.12733601033687592,
+      "learning_rate": 0.00014682569263262517,
+      "loss": 0.0164,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7993664694647599,
+      "grad_norm": 0.14594994485378265,
+      "learning_rate": 0.00014670145359672008,
+      "loss": 0.0219,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8012297945684074,
+      "grad_norm": 0.17092078924179077,
+      "learning_rate": 0.000146577214560815,
+      "loss": 0.0237,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8030931196720548,
+      "grad_norm": 0.1365964561700821,
+      "learning_rate": 0.00014645297552490994,
+      "loss": 0.0198,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8049564447757023,
+      "grad_norm": 0.28921860456466675,
+      "learning_rate": 0.00014632873648900486,
+      "loss": 0.0353,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8068197698793497,
+      "grad_norm": 0.17445138096809387,
+      "learning_rate": 0.00014620449745309977,
+      "loss": 0.022,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8086830949829972,
+      "grad_norm": 0.1306416541337967,
+      "learning_rate": 0.0001460802584171947,
+      "loss": 0.0174,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8105464200866446,
+      "grad_norm": 0.19088391959667206,
+      "learning_rate": 0.0001459560193812896,
+      "loss": 0.021,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8124097451902921,
+      "grad_norm": 0.11422597616910934,
+      "learning_rate": 0.00014583178034538452,
+      "loss": 0.0194,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8142730702939396,
+      "grad_norm": 0.18395616114139557,
+      "learning_rate": 0.00014570754130947946,
+      "loss": 0.0262,
+      "step": 4370
+    },
+    {
+      "epoch": 0.816136395397587,
+      "grad_norm": 0.3011746108531952,
+      "learning_rate": 0.00014558330227357438,
+      "loss": 0.0255,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8179997205012345,
+      "grad_norm": 0.1338394582271576,
+      "learning_rate": 0.0001454590632376693,
+      "loss": 0.0228,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8198630456048819,
+      "grad_norm": 0.2081775963306427,
+      "learning_rate": 0.0001453348242017642,
+      "loss": 0.0205,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8217263707085294,
+      "grad_norm": 0.28005295991897583,
+      "learning_rate": 0.00014521058516585913,
+      "loss": 0.0212,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8235896958121768,
+      "grad_norm": 0.12276031076908112,
+      "learning_rate": 0.00014508634612995404,
+      "loss": 0.0213,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8254530209158243,
+      "grad_norm": 0.11224307119846344,
+      "learning_rate": 0.00014496210709404896,
+      "loss": 0.0238,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8273163460194718,
+      "grad_norm": 0.15174053609371185,
+      "learning_rate": 0.00014483786805814388,
+      "loss": 0.0193,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8291796711231192,
+      "grad_norm": 0.13163380324840546,
+      "learning_rate": 0.0001447136290222388,
+      "loss": 0.0181,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8310429962267667,
+      "grad_norm": 0.1232830286026001,
+      "learning_rate": 0.0001445893899863337,
+      "loss": 0.0238,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8329063213304141,
+      "grad_norm": 0.13815902173519135,
+      "learning_rate": 0.00014446515095042862,
+      "loss": 0.0181,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8347696464340616,
+      "grad_norm": 0.17953529953956604,
+      "learning_rate": 0.00014434091191452354,
+      "loss": 0.0184,
+      "step": 4480
+    },
+    {
+      "epoch": 0.836632971537709,
+      "grad_norm": 0.15750688314437866,
+      "learning_rate": 0.00014421667287861845,
+      "loss": 0.0288,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8384962966413565,
+      "grad_norm": 0.14837270975112915,
+      "learning_rate": 0.00014409243384271337,
+      "loss": 0.0166,
+      "step": 4500
+    },
+    {
+      "epoch": 0.840359621745004,
+      "grad_norm": 0.19016404449939728,
+      "learning_rate": 0.00014396819480680829,
+      "loss": 0.0248,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8422229468486514,
+      "grad_norm": 0.10917269438505173,
+      "learning_rate": 0.00014384395577090323,
+      "loss": 0.0216,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8440862719522989,
+      "grad_norm": 0.12285543233156204,
+      "learning_rate": 0.00014371971673499814,
+      "loss": 0.0283,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8459495970559463,
+      "grad_norm": 0.14190028607845306,
+      "learning_rate": 0.00014359547769909306,
+      "loss": 0.0187,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8478129221595938,
+      "grad_norm": 0.1671181321144104,
+      "learning_rate": 0.00014347123866318798,
+      "loss": 0.0201,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8496762472632412,
+      "grad_norm": 0.17794017493724823,
+      "learning_rate": 0.00014334699962728292,
+      "loss": 0.0146,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8515395723668887,
+      "grad_norm": 0.11655906587839127,
+      "learning_rate": 0.00014322276059137784,
+      "loss": 0.0216,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8534028974705362,
+      "grad_norm": 0.10486368089914322,
+      "learning_rate": 0.00014309852155547275,
+      "loss": 0.0246,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8552662225741836,
+      "grad_norm": 0.12611308693885803,
+      "learning_rate": 0.00014297428251956767,
+      "loss": 0.0211,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8571295476778311,
+      "grad_norm": 0.22120191156864166,
+      "learning_rate": 0.00014285004348366258,
+      "loss": 0.0198,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8589928727814785,
+      "grad_norm": 0.21815341711044312,
+      "learning_rate": 0.0001427258044477575,
+      "loss": 0.023,
+      "step": 4610
+    },
+    {
+      "epoch": 0.860856197885126,
+      "grad_norm": 0.11714211106300354,
+      "learning_rate": 0.00014260156541185241,
+      "loss": 0.0286,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8627195229887734,
+      "grad_norm": 0.1074879840016365,
+      "learning_rate": 0.00014247732637594733,
+      "loss": 0.0198,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8645828480924209,
+      "grad_norm": 0.1850721836090088,
+      "learning_rate": 0.00014235308734004225,
+      "loss": 0.0244,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8664461731960684,
+      "grad_norm": 0.18282188475131989,
+      "learning_rate": 0.00014222884830413716,
+      "loss": 0.0183,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8683094982997158,
+      "grad_norm": 0.20297600328922272,
+      "learning_rate": 0.00014210460926823208,
+      "loss": 0.0209,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8701728234033633,
+      "grad_norm": 0.10349154472351074,
+      "learning_rate": 0.000141980370232327,
+      "loss": 0.0359,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8720361485070107,
+      "grad_norm": 0.25054970383644104,
+      "learning_rate": 0.0001418561311964219,
+      "loss": 0.0185,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8738994736106582,
+      "grad_norm": 0.1190393716096878,
+      "learning_rate": 0.00014173189216051683,
+      "loss": 0.0193,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8757627987143056,
+      "grad_norm": 0.24704180657863617,
+      "learning_rate": 0.00014160765312461174,
+      "loss": 0.0209,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8776261238179531,
+      "grad_norm": 0.14505434036254883,
+      "learning_rate": 0.00014148341408870668,
+      "loss": 0.0189,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8794894489216006,
+      "grad_norm": 0.1667511761188507,
+      "learning_rate": 0.0001413591750528016,
+      "loss": 0.0198,
+      "step": 4720
+    },
+    {
+      "epoch": 0.881352774025248,
+      "grad_norm": 0.15274561941623688,
+      "learning_rate": 0.00014123493601689652,
+      "loss": 0.0163,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8832160991288955,
+      "grad_norm": 0.1500704288482666,
+      "learning_rate": 0.00014111069698099143,
+      "loss": 0.0153,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8850794242325429,
+      "grad_norm": 0.4478381276130676,
+      "learning_rate": 0.00014098645794508635,
+      "loss": 0.0395,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8869427493361904,
+      "grad_norm": 0.2906605005264282,
+      "learning_rate": 0.00014086221890918126,
+      "loss": 0.0224,
+      "step": 4760
+    },
+    {
+      "epoch": 0.888806074439838,
+      "grad_norm": 0.1912839412689209,
+      "learning_rate": 0.0001407379798732762,
+      "loss": 0.0279,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8906693995434853,
+      "grad_norm": 0.08297467231750488,
+      "learning_rate": 0.00014061374083737112,
+      "loss": 0.0201,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8925327246471328,
+      "grad_norm": 0.18305924534797668,
+      "learning_rate": 0.00014048950180146604,
+      "loss": 0.0206,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8943960497507802,
+      "grad_norm": 0.1690533459186554,
+      "learning_rate": 0.00014036526276556095,
+      "loss": 0.0176,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8962593748544277,
+      "grad_norm": 0.16443593800067902,
+      "learning_rate": 0.00014024102372965587,
+      "loss": 0.0209,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8981226999580751,
+      "grad_norm": 0.19682125747203827,
+      "learning_rate": 0.00014011678469375078,
+      "loss": 0.0193,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8999860250617226,
+      "grad_norm": 0.15023230016231537,
+      "learning_rate": 0.0001399925456578457,
+      "loss": 0.0198,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9018493501653702,
+      "grad_norm": 0.10205192863941193,
+      "learning_rate": 0.00013986830662194062,
+      "loss": 0.0209,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9037126752690176,
+      "grad_norm": 0.10337179154157639,
+      "learning_rate": 0.00013974406758603553,
+      "loss": 0.0172,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9055760003726651,
+      "grad_norm": 0.18049459159374237,
+      "learning_rate": 0.00013961982855013045,
+      "loss": 0.0234,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9074393254763125,
+      "grad_norm": 0.11622573435306549,
+      "learning_rate": 0.00013949558951422536,
+      "loss": 0.0207,
+      "step": 4870
+    },
+    {
+      "epoch": 0.90930265057996,
+      "grad_norm": 0.22294333577156067,
+      "learning_rate": 0.00013937135047832028,
+      "loss": 0.0185,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9111659756836074,
+      "grad_norm": 0.37016913294792175,
+      "learning_rate": 0.00013924711144241522,
+      "loss": 0.0238,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9130293007872549,
+      "grad_norm": 0.14420120418071747,
+      "learning_rate": 0.00013912287240651014,
+      "loss": 0.0168,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9148926258909024,
+      "grad_norm": 0.13367877900600433,
+      "learning_rate": 0.00013899863337060505,
+      "loss": 0.018,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9167559509945498,
+      "grad_norm": 0.1635744571685791,
+      "learning_rate": 0.00013887439433469997,
+      "loss": 0.0155,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9186192760981973,
+      "grad_norm": 0.14747698605060577,
+      "learning_rate": 0.00013875015529879489,
+      "loss": 0.0221,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9204826012018447,
+      "grad_norm": 0.1708410084247589,
+      "learning_rate": 0.0001386259162628898,
+      "loss": 0.0197,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9223459263054922,
+      "grad_norm": 0.09329384565353394,
+      "learning_rate": 0.00013850167722698472,
+      "loss": 0.0176,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9242092514091396,
+      "grad_norm": 0.15747447311878204,
+      "learning_rate": 0.00013837743819107963,
+      "loss": 0.0204,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9260725765127871,
+      "grad_norm": 0.14883270859718323,
+      "learning_rate": 0.00013825319915517455,
+      "loss": 0.019,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9279359016164346,
+      "grad_norm": 0.13427717983722687,
+      "learning_rate": 0.0001381289601192695,
+      "loss": 0.0275,
+      "step": 4980
+    },
+    {
+      "epoch": 0.929799226720082,
+      "grad_norm": 0.11976408958435059,
+      "learning_rate": 0.0001380047210833644,
+      "loss": 0.0182,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9316625518237295,
+      "grad_norm": 0.12686076760292053,
+      "learning_rate": 0.00013788048204745932,
+      "loss": 0.0257,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9335258769273769,
+      "grad_norm": 0.16047020256519318,
+      "learning_rate": 0.00013775624301155424,
+      "loss": 0.0192,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9353892020310244,
+      "grad_norm": 0.15550382435321808,
+      "learning_rate": 0.00013763200397564916,
+      "loss": 0.0217,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9372525271346718,
+      "grad_norm": 0.11465305089950562,
+      "learning_rate": 0.00013750776493974407,
+      "loss": 0.0222,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9391158522383193,
+      "grad_norm": 0.1337389051914215,
+      "learning_rate": 0.000137383525903839,
+      "loss": 0.0186,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9409791773419668,
+      "grad_norm": 0.19774989783763885,
+      "learning_rate": 0.0001372592868679339,
+      "loss": 0.0192,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9428425024456142,
+      "grad_norm": 0.1269395351409912,
+      "learning_rate": 0.00013713504783202885,
+      "loss": 0.0188,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9447058275492617,
+      "grad_norm": 0.25458306074142456,
+      "learning_rate": 0.00013701080879612376,
+      "loss": 0.0192,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9465691526529091,
+      "grad_norm": 0.17495116591453552,
+      "learning_rate": 0.00013688656976021868,
+      "loss": 0.0159,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9484324777565566,
+      "grad_norm": 0.07973502576351166,
+      "learning_rate": 0.0001367623307243136,
+      "loss": 0.0158,
+      "step": 5090
+    },
+    {
+      "epoch": 0.950295802860204,
+      "grad_norm": 0.13139723241329193,
+      "learning_rate": 0.0001366380916884085,
+      "loss": 0.0202,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9521591279638515,
+      "grad_norm": 0.2671772837638855,
+      "learning_rate": 0.00013651385265250342,
+      "loss": 0.0199,
+      "step": 5110
+    },
+    {
+      "epoch": 0.954022453067499,
+      "grad_norm": 0.13597209751605988,
+      "learning_rate": 0.00013638961361659834,
+      "loss": 0.021,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9558857781711464,
+      "grad_norm": 0.15831440687179565,
+      "learning_rate": 0.00013626537458069326,
+      "loss": 0.0236,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9577491032747939,
+      "grad_norm": 0.08694765716791153,
+      "learning_rate": 0.00013614113554478817,
+      "loss": 0.0191,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9596124283784413,
+      "grad_norm": 0.12419598549604416,
+      "learning_rate": 0.0001360168965088831,
+      "loss": 0.0259,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9614757534820888,
+      "grad_norm": 0.4500528872013092,
+      "learning_rate": 0.000135892657472978,
+      "loss": 0.0233,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9633390785857362,
+      "grad_norm": 0.13462555408477783,
+      "learning_rate": 0.00013576841843707292,
+      "loss": 0.0239,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9652024036893837,
+      "grad_norm": 0.11178270727396011,
+      "learning_rate": 0.00013564417940116784,
+      "loss": 0.0204,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9670657287930312,
+      "grad_norm": 0.09808334708213806,
+      "learning_rate": 0.00013551994036526278,
+      "loss": 0.0176,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9689290538966786,
+      "grad_norm": 0.12407030910253525,
+      "learning_rate": 0.0001353957013293577,
+      "loss": 0.0201,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9707923790003261,
+      "grad_norm": 0.11156706511974335,
+      "learning_rate": 0.0001352714622934526,
+      "loss": 0.0176,
+      "step": 5210
+    },
+    {
+      "epoch": 0.9726557041039735,
+      "grad_norm": 0.16406476497650146,
+      "learning_rate": 0.00013514722325754753,
+      "loss": 0.0208,
+      "step": 5220
+    },
+    {
+      "epoch": 0.974519029207621,
+      "grad_norm": 0.08763796091079712,
+      "learning_rate": 0.00013502298422164247,
+      "loss": 0.0236,
+      "step": 5230
+    },
+    {
+      "epoch": 0.9763823543112684,
+      "grad_norm": 0.10975922644138336,
+      "learning_rate": 0.00013489874518573738,
+      "loss": 0.0172,
+      "step": 5240
+    },
+    {
+      "epoch": 0.9782456794149159,
+      "grad_norm": 0.14900168776512146,
+      "learning_rate": 0.0001347745061498323,
+      "loss": 0.0176,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9801090045185634,
+      "grad_norm": 0.08646497875452042,
+      "learning_rate": 0.00013465026711392722,
+      "loss": 0.0255,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9819723296222108,
+      "grad_norm": 0.12455034255981445,
+      "learning_rate": 0.00013452602807802213,
+      "loss": 0.0148,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9838356547258583,
+      "grad_norm": 0.08824939280748367,
+      "learning_rate": 0.00013440178904211705,
+      "loss": 0.0219,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9856989798295057,
+      "grad_norm": 0.12328789383172989,
+      "learning_rate": 0.00013427755000621196,
+      "loss": 0.0182,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9875623049331532,
+      "grad_norm": 0.07896114885807037,
+      "learning_rate": 0.00013415331097030688,
+      "loss": 0.0203,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9894256300368006,
+      "grad_norm": 1.816659688949585,
+      "learning_rate": 0.0001340290719344018,
+      "loss": 0.0186,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9912889551404481,
+      "grad_norm": 0.1731538623571396,
+      "learning_rate": 0.0001339048328984967,
+      "loss": 0.0132,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9931522802440956,
+      "grad_norm": 0.1737430840730667,
+      "learning_rate": 0.00013378059386259163,
+      "loss": 0.0206,
+      "step": 5330
+    },
+    {
+      "epoch": 0.995015605347743,
+      "grad_norm": 0.10039955377578735,
+      "learning_rate": 0.00013365635482668654,
+      "loss": 0.0182,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9968789304513905,
+      "grad_norm": 0.07095132768154144,
+      "learning_rate": 0.00013353211579078146,
+      "loss": 0.0187,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9987422555550379,
+      "grad_norm": 0.08819052577018738,
+      "learning_rate": 0.00013340787675487637,
+      "loss": 0.0162,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0006055806586853,
+      "grad_norm": 0.10541682690382004,
+      "learning_rate": 0.0001332836377189713,
+      "loss": 0.0218,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0024689057623328,
+      "grad_norm": 0.09925888478755951,
+      "learning_rate": 0.0001331593986830662,
+      "loss": 0.017,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0043322308659803,
+      "grad_norm": 0.130414217710495,
+      "learning_rate": 0.00013303515964716112,
+      "loss": 0.0183,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0061955559696278,
+      "grad_norm": 0.10512223094701767,
+      "learning_rate": 0.00013291092061125606,
+      "loss": 0.0177,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0080588810732753,
+      "grad_norm": 0.09342774003744125,
+      "learning_rate": 0.00013278668157535098,
+      "loss": 0.0128,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0099222061769226,
+      "grad_norm": 0.14016024768352509,
+      "learning_rate": 0.0001326624425394459,
+      "loss": 0.0174,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0117855312805701,
+      "grad_norm": 0.33078333735466003,
+      "learning_rate": 0.00013253820350354084,
+      "loss": 0.0141,
+      "step": 5430
+    },
+    {
+      "epoch": 1.0136488563842176,
+      "grad_norm": 0.14986422657966614,
+      "learning_rate": 0.00013241396446763575,
+      "loss": 0.0138,
+      "step": 5440
+    },
+    {
+      "epoch": 1.0155121814878652,
+      "grad_norm": 0.05468583106994629,
+      "learning_rate": 0.00013228972543173067,
+      "loss": 0.0134,
+      "step": 5450
+    },
+    {
+      "epoch": 1.0173755065915127,
+      "grad_norm": 0.15035457909107208,
+      "learning_rate": 0.00013216548639582559,
+      "loss": 0.0138,
+      "step": 5460
+    },
+    {
+      "epoch": 1.01923883169516,
+      "grad_norm": 0.12102866917848587,
+      "learning_rate": 0.0001320412473599205,
+      "loss": 0.0121,
+      "step": 5470
+    },
+    {
+      "epoch": 1.0211021567988074,
+      "grad_norm": 0.1015734151005745,
+      "learning_rate": 0.00013191700832401542,
+      "loss": 0.0156,
+      "step": 5480
+    },
+    {
+      "epoch": 1.022965481902455,
+      "grad_norm": 0.08906779438257217,
+      "learning_rate": 0.00013179276928811033,
+      "loss": 0.0137,
+      "step": 5490
+    },
+    {
+      "epoch": 1.0248288070061025,
+      "grad_norm": 0.04403156042098999,
+      "learning_rate": 0.00013166853025220525,
+      "loss": 0.0115,
+      "step": 5500
+    },
+    {
+      "epoch": 1.0266921321097497,
+      "grad_norm": 0.1241898313164711,
+      "learning_rate": 0.00013154429121630017,
+      "loss": 0.0176,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0285554572133973,
+      "grad_norm": 0.1671934872865677,
+      "learning_rate": 0.00013142005218039508,
+      "loss": 0.0136,
+      "step": 5520
+    },
+    {
+      "epoch": 1.0304187823170448,
+      "grad_norm": 0.6383609771728516,
+      "learning_rate": 0.00013129581314449,
+      "loss": 0.022,
+      "step": 5530
+    },
+    {
+      "epoch": 1.0322821074206923,
+      "grad_norm": 0.08012861758470535,
+      "learning_rate": 0.0001311715741085849,
+      "loss": 0.0141,
+      "step": 5540
+    },
+    {
+      "epoch": 1.0341454325243398,
+      "grad_norm": 0.11658889800310135,
+      "learning_rate": 0.00013104733507267983,
+      "loss": 0.0152,
+      "step": 5550
+    },
+    {
+      "epoch": 1.036008757627987,
+      "grad_norm": 0.13740360736846924,
+      "learning_rate": 0.00013092309603677474,
+      "loss": 0.0198,
+      "step": 5560
+    },
+    {
+      "epoch": 1.0378720827316346,
+      "grad_norm": 0.8119271993637085,
+      "learning_rate": 0.0001307988570008697,
+      "loss": 0.0151,
+      "step": 5570
+    },
+    {
+      "epoch": 1.039735407835282,
+      "grad_norm": 0.2707085907459259,
+      "learning_rate": 0.0001306746179649646,
+      "loss": 0.0163,
+      "step": 5580
+    },
+    {
+      "epoch": 1.0415987329389296,
+      "grad_norm": 0.10634835809469223,
+      "learning_rate": 0.00013055037892905952,
+      "loss": 0.0121,
+      "step": 5590
+    },
+    {
+      "epoch": 1.043462058042577,
+      "grad_norm": 0.12598338723182678,
+      "learning_rate": 0.00013042613989315443,
+      "loss": 0.0166,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0453253831462244,
+      "grad_norm": 0.13499614596366882,
+      "learning_rate": 0.00013030190085724935,
+      "loss": 0.0119,
+      "step": 5610
+    },
+    {
+      "epoch": 1.0471887082498719,
+      "grad_norm": 0.1680738776922226,
+      "learning_rate": 0.00013017766182134427,
+      "loss": 0.024,
+      "step": 5620
+    },
+    {
+      "epoch": 1.0490520333535194,
+      "grad_norm": 0.1279182881116867,
+      "learning_rate": 0.00013005342278543918,
+      "loss": 0.0166,
+      "step": 5630
+    },
+    {
+      "epoch": 1.0509153584571669,
+      "grad_norm": 0.07233988493680954,
+      "learning_rate": 0.0001299291837495341,
+      "loss": 0.0157,
+      "step": 5640
+    },
+    {
+      "epoch": 1.0527786835608142,
+      "grad_norm": 0.1104668378829956,
+      "learning_rate": 0.00012980494471362904,
+      "loss": 0.0156,
+      "step": 5650
+    },
+    {
+      "epoch": 1.0546420086644617,
+      "grad_norm": 0.1559237837791443,
+      "learning_rate": 0.00012968070567772396,
+      "loss": 0.0147,
+      "step": 5660
+    },
+    {
+      "epoch": 1.0565053337681092,
+      "grad_norm": 0.04395401477813721,
+      "learning_rate": 0.00012955646664181887,
+      "loss": 0.0134,
+      "step": 5670
+    },
+    {
+      "epoch": 1.0583686588717567,
+      "grad_norm": 0.07985333353281021,
+      "learning_rate": 0.0001294322276059138,
+      "loss": 0.0146,
+      "step": 5680
+    },
+    {
+      "epoch": 1.0602319839754042,
+      "grad_norm": 0.06879911571741104,
+      "learning_rate": 0.0001293079885700087,
+      "loss": 0.0202,
+      "step": 5690
+    },
+    {
+      "epoch": 1.0620953090790515,
+      "grad_norm": 0.10023193806409836,
+      "learning_rate": 0.00012918374953410362,
+      "loss": 0.0202,
+      "step": 5700
+    },
+    {
+      "epoch": 1.063958634182699,
+      "grad_norm": 0.07399614155292511,
+      "learning_rate": 0.00012905951049819854,
+      "loss": 0.0119,
+      "step": 5710
+    },
+    {
+      "epoch": 1.0658219592863465,
+      "grad_norm": 0.10685840249061584,
+      "learning_rate": 0.00012893527146229345,
+      "loss": 0.0153,
+      "step": 5720
+    },
+    {
+      "epoch": 1.067685284389994,
+      "grad_norm": 0.12774063646793365,
+      "learning_rate": 0.00012881103242638837,
+      "loss": 0.0153,
+      "step": 5730
+    },
+    {
+      "epoch": 1.0695486094936415,
+      "grad_norm": 0.09937337785959244,
+      "learning_rate": 0.0001286867933904833,
+      "loss": 0.0383,
+      "step": 5740
+    },
+    {
+      "epoch": 1.0714119345972888,
+      "grad_norm": 0.09235163778066635,
+      "learning_rate": 0.00012856255435457823,
+      "loss": 0.0173,
+      "step": 5750
+    },
+    {
+      "epoch": 1.0732752597009363,
+      "grad_norm": 0.14451561868190765,
+      "learning_rate": 0.00012843831531867314,
+      "loss": 0.0125,
+      "step": 5760
+    },
+    {
+      "epoch": 1.0751385848045838,
+      "grad_norm": 0.07140475511550903,
+      "learning_rate": 0.00012831407628276806,
+      "loss": 0.0152,
+      "step": 5770
+    },
+    {
+      "epoch": 1.0770019099082313,
+      "grad_norm": 0.10524556040763855,
+      "learning_rate": 0.00012818983724686297,
+      "loss": 0.0167,
+      "step": 5780
+    },
+    {
+      "epoch": 1.0788652350118788,
+      "grad_norm": 0.1626797616481781,
+      "learning_rate": 0.0001280655982109579,
+      "loss": 0.0164,
+      "step": 5790
+    },
+    {
+      "epoch": 1.080728560115526,
+      "grad_norm": 0.20539982616901398,
+      "learning_rate": 0.0001279413591750528,
+      "loss": 0.0145,
+      "step": 5800
+    },
+    {
+      "epoch": 1.0825918852191736,
+      "grad_norm": 0.11382734030485153,
+      "learning_rate": 0.00012781712013914772,
+      "loss": 0.0149,
+      "step": 5810
+    },
+    {
+      "epoch": 1.084455210322821,
+      "grad_norm": 0.07617650181055069,
+      "learning_rate": 0.00012769288110324264,
+      "loss": 0.0141,
+      "step": 5820
+    },
+    {
+      "epoch": 1.0863185354264686,
+      "grad_norm": 0.14019963145256042,
+      "learning_rate": 0.00012756864206733755,
+      "loss": 0.017,
+      "step": 5830
+    },
+    {
+      "epoch": 1.088181860530116,
+      "grad_norm": 0.097674660384655,
+      "learning_rate": 0.00012744440303143247,
+      "loss": 0.019,
+      "step": 5840
+    },
+    {
+      "epoch": 1.0900451856337634,
+      "grad_norm": 0.13726158440113068,
+      "learning_rate": 0.00012732016399552738,
+      "loss": 0.0166,
+      "step": 5850
+    },
+    {
+      "epoch": 1.091908510737411,
+      "grad_norm": 0.2003697007894516,
+      "learning_rate": 0.00012719592495962233,
+      "loss": 0.0154,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0937718358410584,
+      "grad_norm": 0.08333192020654678,
+      "learning_rate": 0.00012707168592371724,
+      "loss": 0.0143,
+      "step": 5870
+    },
+    {
+      "epoch": 1.095635160944706,
+      "grad_norm": 0.13497716188430786,
+      "learning_rate": 0.00012694744688781216,
+      "loss": 0.018,
+      "step": 5880
+    },
+    {
+      "epoch": 1.0974984860483532,
+      "grad_norm": 0.09423158317804337,
+      "learning_rate": 0.00012682320785190707,
+      "loss": 0.0203,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0993618111520007,
+      "grad_norm": 0.1061846911907196,
+      "learning_rate": 0.000126698968816002,
+      "loss": 0.0106,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1012251362556482,
+      "grad_norm": 0.10075803101062775,
+      "learning_rate": 0.00012657472978009693,
+      "loss": 0.0104,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1030884613592957,
+      "grad_norm": 0.19989649951457977,
+      "learning_rate": 0.00012645049074419185,
+      "loss": 0.0205,
+      "step": 5920
+    },
+    {
+      "epoch": 1.104951786462943,
+      "grad_norm": 0.10290393233299255,
+      "learning_rate": 0.00012632625170828676,
+      "loss": 0.0101,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1068151115665905,
+      "grad_norm": 0.1811441034078598,
+      "learning_rate": 0.00012620201267238168,
+      "loss": 0.0169,
+      "step": 5940
+    },
+    {
+      "epoch": 1.108678436670238,
+      "grad_norm": 0.07941067963838577,
+      "learning_rate": 0.0001260777736364766,
+      "loss": 0.0156,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1105417617738855,
+      "grad_norm": 0.15073469281196594,
+      "learning_rate": 0.0001259535346005715,
+      "loss": 0.0219,
+      "step": 5960
+    },
+    {
+      "epoch": 1.112405086877533,
+      "grad_norm": 0.09446485340595245,
+      "learning_rate": 0.00012582929556466643,
+      "loss": 0.0122,
+      "step": 5970
+    },
+    {
+      "epoch": 1.1142684119811803,
+      "grad_norm": 0.08050578832626343,
+      "learning_rate": 0.00012570505652876134,
+      "loss": 0.012,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1161317370848278,
+      "grad_norm": 0.08295507729053497,
+      "learning_rate": 0.00012558081749285626,
+      "loss": 0.0153,
+      "step": 5990
+    },
+    {
+      "epoch": 1.1179950621884753,
+      "grad_norm": 0.09732520580291748,
+      "learning_rate": 0.00012545657845695118,
+      "loss": 0.0203,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1198583872921228,
+      "grad_norm": 0.12038660049438477,
+      "learning_rate": 0.0001253323394210461,
+      "loss": 0.0182,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1217217123957703,
+      "grad_norm": 0.10337748378515244,
+      "learning_rate": 0.000125208100385141,
+      "loss": 0.0118,
+      "step": 6020
+    },
+    {
+      "epoch": 1.1235850374994176,
+      "grad_norm": 0.15960311889648438,
+      "learning_rate": 0.00012508386134923592,
+      "loss": 0.0167,
+      "step": 6030
+    },
+    {
+      "epoch": 1.1254483626030651,
+      "grad_norm": 0.12002138048410416,
+      "learning_rate": 0.00012495962231333084,
+      "loss": 0.0141,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1273116877067126,
+      "grad_norm": 0.1650262326002121,
+      "learning_rate": 0.00012483538327742575,
+      "loss": 0.018,
+      "step": 6050
+    },
+    {
+      "epoch": 1.1291750128103601,
+      "grad_norm": 0.13276247680187225,
+      "learning_rate": 0.00012471114424152067,
+      "loss": 0.0209,
+      "step": 6060
+    },
+    {
+      "epoch": 1.1310383379140077,
+      "grad_norm": 0.0923129990696907,
+      "learning_rate": 0.0001245869052056156,
+      "loss": 0.0173,
+      "step": 6070
+    },
+    {
+      "epoch": 1.132901663017655,
+      "grad_norm": 0.1202143207192421,
+      "learning_rate": 0.00012446266616971053,
+      "loss": 0.0158,
+      "step": 6080
+    },
+    {
+      "epoch": 1.1347649881213024,
+      "grad_norm": 0.12622275948524475,
+      "learning_rate": 0.00012433842713380544,
+      "loss": 0.0145,
+      "step": 6090
+    },
+    {
+      "epoch": 1.13662831322495,
+      "grad_norm": 0.21728819608688354,
+      "learning_rate": 0.0001242141880979004,
+      "loss": 0.0137,
+      "step": 6100
+    },
+    {
+      "epoch": 1.1384916383285975,
+      "grad_norm": 0.14400696754455566,
+      "learning_rate": 0.0001240899490619953,
+      "loss": 0.0147,
+      "step": 6110
+    },
+    {
+      "epoch": 1.140354963432245,
+      "grad_norm": 0.04675738885998726,
+      "learning_rate": 0.00012396571002609022,
+      "loss": 0.0141,
+      "step": 6120
+    },
+    {
+      "epoch": 1.1422182885358922,
+      "grad_norm": 0.11618246138095856,
+      "learning_rate": 0.00012384147099018514,
+      "loss": 0.0124,
+      "step": 6130
+    },
+    {
+      "epoch": 1.1440816136395398,
+      "grad_norm": 0.08097544312477112,
+      "learning_rate": 0.00012371723195428005,
+      "loss": 0.0112,
+      "step": 6140
+    },
+    {
+      "epoch": 1.1459449387431873,
+      "grad_norm": 0.1423315703868866,
+      "learning_rate": 0.00012359299291837497,
+      "loss": 0.0138,
+      "step": 6150
+    },
+    {
+      "epoch": 1.1478082638468348,
+      "grad_norm": 0.4167300760746002,
+      "learning_rate": 0.00012346875388246988,
+      "loss": 0.0171,
+      "step": 6160
+    },
+    {
+      "epoch": 1.149671588950482,
+      "grad_norm": 0.12244168668985367,
+      "learning_rate": 0.0001233445148465648,
+      "loss": 0.0135,
+      "step": 6170
+    },
+    {
+      "epoch": 1.1515349140541296,
+      "grad_norm": 0.10434919595718384,
+      "learning_rate": 0.00012322027581065971,
+      "loss": 0.0128,
+      "step": 6180
+    },
+    {
+      "epoch": 1.153398239157777,
+      "grad_norm": 0.11119112372398376,
+      "learning_rate": 0.00012309603677475463,
+      "loss": 0.0122,
+      "step": 6190
+    },
+    {
+      "epoch": 1.1552615642614246,
+      "grad_norm": 0.14864900708198547,
+      "learning_rate": 0.00012297179773884955,
+      "loss": 0.0119,
+      "step": 6200
+    },
+    {
+      "epoch": 1.1571248893650719,
+      "grad_norm": 0.9202861189842224,
+      "learning_rate": 0.00012284755870294446,
+      "loss": 0.0149,
+      "step": 6210
+    },
+    {
+      "epoch": 1.1589882144687194,
+      "grad_norm": 0.10615832358598709,
+      "learning_rate": 0.00012272331966703938,
+      "loss": 0.0161,
+      "step": 6220
+    },
+    {
+      "epoch": 1.1608515395723669,
+      "grad_norm": 0.05958676338195801,
+      "learning_rate": 0.0001225990806311343,
+      "loss": 0.0137,
+      "step": 6230
+    },
+    {
+      "epoch": 1.1627148646760144,
+      "grad_norm": 0.10617449879646301,
+      "learning_rate": 0.0001224748415952292,
+      "loss": 0.012,
+      "step": 6240
+    },
+    {
+      "epoch": 1.1645781897796619,
+      "grad_norm": 0.141194686293602,
+      "learning_rate": 0.00012235060255932412,
+      "loss": 0.0129,
+      "step": 6250
+    },
+    {
+      "epoch": 1.1664415148833092,
+      "grad_norm": 0.09088177978992462,
+      "learning_rate": 0.00012222636352341907,
+      "loss": 0.0125,
+      "step": 6260
+    },
+    {
+      "epoch": 1.1683048399869567,
+      "grad_norm": 0.12682999670505524,
+      "learning_rate": 0.00012210212448751398,
+      "loss": 0.012,
+      "step": 6270
+    },
+    {
+      "epoch": 1.1701681650906042,
+      "grad_norm": 0.13394199311733246,
+      "learning_rate": 0.00012197788545160891,
+      "loss": 0.0159,
+      "step": 6280
+    },
+    {
+      "epoch": 1.1720314901942517,
+      "grad_norm": 0.12738922238349915,
+      "learning_rate": 0.00012185364641570383,
+      "loss": 0.0155,
+      "step": 6290
+    },
+    {
+      "epoch": 1.1738948152978992,
+      "grad_norm": 0.11358192563056946,
+      "learning_rate": 0.00012172940737979874,
+      "loss": 0.0114,
+      "step": 6300
+    },
+    {
+      "epoch": 1.1757581404015465,
+      "grad_norm": 0.09697134792804718,
+      "learning_rate": 0.00012160516834389366,
+      "loss": 0.0126,
+      "step": 6310
+    },
+    {
+      "epoch": 1.177621465505194,
+      "grad_norm": 0.12397624552249908,
+      "learning_rate": 0.00012148092930798858,
+      "loss": 0.0133,
+      "step": 6320
+    },
+    {
+      "epoch": 1.1794847906088415,
+      "grad_norm": 0.07363367825746536,
+      "learning_rate": 0.00012135669027208349,
+      "loss": 0.0108,
+      "step": 6330
+    },
+    {
+      "epoch": 1.181348115712489,
+      "grad_norm": 0.10165391117334366,
+      "learning_rate": 0.00012123245123617841,
+      "loss": 0.0162,
+      "step": 6340
+    },
+    {
+      "epoch": 1.1832114408161365,
+      "grad_norm": 0.07939434796571732,
+      "learning_rate": 0.00012110821220027332,
+      "loss": 0.011,
+      "step": 6350
+    },
+    {
+      "epoch": 1.1850747659197838,
+      "grad_norm": 0.11631353199481964,
+      "learning_rate": 0.00012098397316436825,
+      "loss": 0.0135,
+      "step": 6360
+    },
+    {
+      "epoch": 1.1869380910234313,
+      "grad_norm": 0.12175942212343216,
+      "learning_rate": 0.00012085973412846317,
+      "loss": 0.012,
+      "step": 6370
+    },
+    {
+      "epoch": 1.1888014161270788,
+      "grad_norm": 0.06976991146802902,
+      "learning_rate": 0.00012073549509255808,
+      "loss": 0.0232,
+      "step": 6380
+    },
+    {
+      "epoch": 1.1906647412307263,
+      "grad_norm": 0.07860373705625534,
+      "learning_rate": 0.000120611256056653,
+      "loss": 0.0183,
+      "step": 6390
+    },
+    {
+      "epoch": 1.1925280663343738,
+      "grad_norm": 0.10476142168045044,
+      "learning_rate": 0.00012048701702074792,
+      "loss": 0.015,
+      "step": 6400
+    },
+    {
+      "epoch": 1.194391391438021,
+      "grad_norm": 0.15761305391788483,
+      "learning_rate": 0.00012036277798484283,
+      "loss": 0.0163,
+      "step": 6410
+    },
+    {
+      "epoch": 1.1962547165416686,
+      "grad_norm": 0.11684102565050125,
+      "learning_rate": 0.00012023853894893775,
+      "loss": 0.0114,
+      "step": 6420
+    },
+    {
+      "epoch": 1.198118041645316,
+      "grad_norm": 0.07757461071014404,
+      "learning_rate": 0.00012011429991303269,
+      "loss": 0.0106,
+      "step": 6430
+    },
+    {
+      "epoch": 1.1999813667489636,
+      "grad_norm": 0.08250425010919571,
+      "learning_rate": 0.0001199900608771276,
+      "loss": 0.0162,
+      "step": 6440
+    },
+    {
+      "epoch": 1.201844691852611,
+      "grad_norm": 0.07098899036645889,
+      "learning_rate": 0.00011986582184122252,
+      "loss": 0.0094,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2037080169562584,
+      "grad_norm": 0.12341819703578949,
+      "learning_rate": 0.00011974158280531744,
+      "loss": 0.014,
+      "step": 6460
+    },
+    {
+      "epoch": 1.205571342059906,
+      "grad_norm": 0.12068308889865875,
+      "learning_rate": 0.00011961734376941235,
+      "loss": 0.0144,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2074346671635534,
+      "grad_norm": 0.12204370647668839,
+      "learning_rate": 0.00011949310473350728,
+      "loss": 0.0139,
+      "step": 6480
+    },
+    {
+      "epoch": 1.209297992267201,
+      "grad_norm": 0.07349750399589539,
+      "learning_rate": 0.0001193688656976022,
+      "loss": 0.0134,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2111613173708482,
+      "grad_norm": 0.09023798257112503,
+      "learning_rate": 0.00011924462666169712,
+      "loss": 0.0133,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2130246424744957,
+      "grad_norm": 0.11921370029449463,
+      "learning_rate": 0.00011912038762579203,
+      "loss": 0.0202,
+      "step": 6510
+    },
+    {
+      "epoch": 1.2148879675781432,
+      "grad_norm": 0.16381864249706268,
+      "learning_rate": 0.00011899614858988695,
+      "loss": 0.013,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2167512926817907,
+      "grad_norm": 0.18260565400123596,
+      "learning_rate": 0.00011887190955398186,
+      "loss": 0.0152,
+      "step": 6530
+    },
+    {
+      "epoch": 1.218614617785438,
+      "grad_norm": 0.17924544215202332,
+      "learning_rate": 0.00011874767051807678,
+      "loss": 0.0156,
+      "step": 6540
+    },
+    {
+      "epoch": 1.2204779428890855,
+      "grad_norm": 0.0538809597492218,
+      "learning_rate": 0.0001186234314821717,
+      "loss": 0.0096,
+      "step": 6550
+    },
+    {
+      "epoch": 1.222341267992733,
+      "grad_norm": 0.15469981729984283,
+      "learning_rate": 0.00011849919244626661,
+      "loss": 0.0122,
+      "step": 6560
+    },
+    {
+      "epoch": 1.2242045930963805,
+      "grad_norm": 0.10429839044809341,
+      "learning_rate": 0.00011837495341036154,
+      "loss": 0.0153,
+      "step": 6570
+    },
+    {
+      "epoch": 1.226067918200028,
+      "grad_norm": 0.09104898571968079,
+      "learning_rate": 0.00011825071437445646,
+      "loss": 0.0136,
+      "step": 6580
+    },
+    {
+      "epoch": 1.2279312433036753,
+      "grad_norm": 0.11124694347381592,
+      "learning_rate": 0.00011812647533855137,
+      "loss": 0.012,
+      "step": 6590
+    },
+    {
+      "epoch": 1.2297945684073228,
+      "grad_norm": 0.18156588077545166,
+      "learning_rate": 0.00011800223630264631,
+      "loss": 0.013,
+      "step": 6600
+    },
+    {
+      "epoch": 1.2316578935109703,
+      "grad_norm": 0.10969128459692001,
+      "learning_rate": 0.00011787799726674123,
+      "loss": 0.0124,
+      "step": 6610
+    },
+    {
+      "epoch": 1.2335212186146178,
+      "grad_norm": 0.1498788446187973,
+      "learning_rate": 0.00011775375823083615,
+      "loss": 0.0131,
+      "step": 6620
+    },
+    {
+      "epoch": 1.2353845437182653,
+      "grad_norm": 0.08221983164548874,
+      "learning_rate": 0.00011762951919493106,
+      "loss": 0.0151,
+      "step": 6630
+    },
+    {
+      "epoch": 1.2372478688219126,
+      "grad_norm": 0.10865967720746994,
+      "learning_rate": 0.00011750528015902598,
+      "loss": 0.0115,
+      "step": 6640
+    },
+    {
+      "epoch": 1.2391111939255601,
+      "grad_norm": 0.09154286235570908,
+      "learning_rate": 0.00011738104112312089,
+      "loss": 0.0118,
+      "step": 6650
+    },
+    {
+      "epoch": 1.2409745190292076,
+      "grad_norm": 0.2195647954940796,
+      "learning_rate": 0.00011725680208721581,
+      "loss": 0.024,
+      "step": 6660
+    },
+    {
+      "epoch": 1.2428378441328551,
+      "grad_norm": 0.09780694544315338,
+      "learning_rate": 0.00011713256305131072,
+      "loss": 0.0121,
+      "step": 6670
+    },
+    {
+      "epoch": 1.2447011692365026,
+      "grad_norm": 0.12990465760231018,
+      "learning_rate": 0.00011700832401540564,
+      "loss": 0.0118,
+      "step": 6680
+    },
+    {
+      "epoch": 1.24656449434015,
+      "grad_norm": 0.09181374311447144,
+      "learning_rate": 0.00011688408497950057,
+      "loss": 0.015,
+      "step": 6690
+    },
+    {
+      "epoch": 1.2484278194437974,
+      "grad_norm": 0.0914270430803299,
+      "learning_rate": 0.00011675984594359549,
+      "loss": 0.0085,
+      "step": 6700
+    },
+    {
+      "epoch": 1.250291144547445,
+      "grad_norm": 0.09390436112880707,
+      "learning_rate": 0.0001166356069076904,
+      "loss": 0.0143,
+      "step": 6710
+    },
+    {
+      "epoch": 1.2521544696510924,
+      "grad_norm": 0.0553867444396019,
+      "learning_rate": 0.00011651136787178532,
+      "loss": 0.0096,
+      "step": 6720
+    },
+    {
+      "epoch": 1.25401779475474,
+      "grad_norm": 0.08669324219226837,
+      "learning_rate": 0.00011638712883588023,
+      "loss": 0.012,
+      "step": 6730
+    },
+    {
+      "epoch": 1.2558811198583872,
+      "grad_norm": 0.10883240401744843,
+      "learning_rate": 0.00011626288979997515,
+      "loss": 0.0156,
+      "step": 6740
+    },
+    {
+      "epoch": 1.2577444449620347,
+      "grad_norm": 0.39218905568122864,
+      "learning_rate": 0.00011613865076407006,
+      "loss": 0.0404,
+      "step": 6750
+    },
+    {
+      "epoch": 1.2596077700656823,
+      "grad_norm": 0.10930469632148743,
+      "learning_rate": 0.00011601441172816498,
+      "loss": 0.014,
+      "step": 6760
+    },
+    {
+      "epoch": 1.2614710951693295,
+      "grad_norm": 0.19345052540302277,
+      "learning_rate": 0.00011589017269225992,
+      "loss": 0.0147,
+      "step": 6770
+    },
+    {
+      "epoch": 1.2633344202729773,
+      "grad_norm": 0.0582694374024868,
+      "learning_rate": 0.00011576593365635484,
+      "loss": 0.0116,
+      "step": 6780
+    },
+    {
+      "epoch": 1.2651977453766245,
+      "grad_norm": 0.0844726487994194,
+      "learning_rate": 0.00011564169462044975,
+      "loss": 0.0152,
+      "step": 6790
+    },
+    {
+      "epoch": 1.267061070480272,
+      "grad_norm": 0.11363263428211212,
+      "learning_rate": 0.00011551745558454467,
+      "loss": 0.0128,
+      "step": 6800
+    },
+    {
+      "epoch": 1.2689243955839196,
+      "grad_norm": 0.13044427335262299,
+      "learning_rate": 0.0001153932165486396,
+      "loss": 0.014,
+      "step": 6810
+    },
+    {
+      "epoch": 1.2707877206875668,
+      "grad_norm": 0.1002730280160904,
+      "learning_rate": 0.00011526897751273452,
+      "loss": 0.0139,
+      "step": 6820
+    },
+    {
+      "epoch": 1.2726510457912144,
+      "grad_norm": 0.10550817102193832,
+      "learning_rate": 0.00011514473847682943,
+      "loss": 0.0125,
+      "step": 6830
+    },
+    {
+      "epoch": 1.2745143708948619,
+      "grad_norm": 0.11775978654623032,
+      "learning_rate": 0.00011502049944092435,
+      "loss": 0.0103,
+      "step": 6840
+    },
+    {
+      "epoch": 1.2763776959985094,
+      "grad_norm": 0.19106002151966095,
+      "learning_rate": 0.00011489626040501926,
+      "loss": 0.0153,
+      "step": 6850
+    },
+    {
+      "epoch": 1.2782410211021569,
+      "grad_norm": 0.07046420127153397,
+      "learning_rate": 0.00011477202136911418,
+      "loss": 0.0204,
+      "step": 6860
+    },
+    {
+      "epoch": 1.2801043462058042,
+      "grad_norm": 0.10768242925405502,
+      "learning_rate": 0.0001146477823332091,
+      "loss": 0.0111,
+      "step": 6870
+    },
+    {
+      "epoch": 1.2819676713094517,
+      "grad_norm": 0.08488526940345764,
+      "learning_rate": 0.00011452354329730401,
+      "loss": 0.0166,
+      "step": 6880
+    },
+    {
+      "epoch": 1.2838309964130992,
+      "grad_norm": 0.11364570260047913,
+      "learning_rate": 0.00011439930426139893,
+      "loss": 0.0127,
+      "step": 6890
+    },
+    {
+      "epoch": 1.2856943215167467,
+      "grad_norm": 0.09781802445650101,
+      "learning_rate": 0.00011427506522549386,
+      "loss": 0.0122,
+      "step": 6900
+    },
+    {
+      "epoch": 1.2875576466203942,
+      "grad_norm": 0.11134841293096542,
+      "learning_rate": 0.00011415082618958877,
+      "loss": 0.0222,
+      "step": 6910
+    },
+    {
+      "epoch": 1.2894209717240415,
+      "grad_norm": 0.09454433619976044,
+      "learning_rate": 0.00011402658715368369,
+      "loss": 0.014,
+      "step": 6920
+    },
+    {
+      "epoch": 1.291284296827689,
+      "grad_norm": 0.13544605672359467,
+      "learning_rate": 0.0001139023481177786,
+      "loss": 0.0172,
+      "step": 6930
+    },
+    {
+      "epoch": 1.2931476219313365,
+      "grad_norm": 0.06879796087741852,
+      "learning_rate": 0.00011377810908187355,
+      "loss": 0.0136,
+      "step": 6940
+    },
+    {
+      "epoch": 1.295010947034984,
+      "grad_norm": 0.10580771416425705,
+      "learning_rate": 0.00011365387004596846,
+      "loss": 0.0208,
+      "step": 6950
+    },
+    {
+      "epoch": 1.2968742721386315,
+      "grad_norm": 0.1340237557888031,
+      "learning_rate": 0.00011352963101006338,
+      "loss": 0.0117,
+      "step": 6960
+    },
+    {
+      "epoch": 1.2987375972422788,
+      "grad_norm": 0.14873532950878143,
+      "learning_rate": 0.0001134053919741583,
+      "loss": 0.0134,
+      "step": 6970
+    },
+    {
+      "epoch": 1.3006009223459263,
+      "grad_norm": 0.11179360747337341,
+      "learning_rate": 0.00011328115293825321,
+      "loss": 0.0142,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3024642474495738,
+      "grad_norm": 0.10709336400032043,
+      "learning_rate": 0.00011315691390234813,
+      "loss": 0.014,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3043275725532213,
+      "grad_norm": 0.20911122858524323,
+      "learning_rate": 0.00011303267486644304,
+      "loss": 0.018,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3061908976568688,
+      "grad_norm": 0.0970635861158371,
+      "learning_rate": 0.00011290843583053796,
+      "loss": 0.0157,
+      "step": 7010
+    },
+    {
+      "epoch": 1.308054222760516,
+      "grad_norm": 0.13426648080348969,
+      "learning_rate": 0.00011278419679463287,
+      "loss": 0.0128,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3099175478641636,
+      "grad_norm": 0.11847269535064697,
+      "learning_rate": 0.0001126599577587278,
+      "loss": 0.0085,
+      "step": 7030
+    },
+    {
+      "epoch": 1.311780872967811,
+      "grad_norm": 0.07585333287715912,
+      "learning_rate": 0.00011253571872282272,
+      "loss": 0.0127,
+      "step": 7040
+    },
+    {
+      "epoch": 1.3136441980714584,
+      "grad_norm": 0.11813944578170776,
+      "learning_rate": 0.00011241147968691763,
+      "loss": 0.0142,
+      "step": 7050
+    },
+    {
+      "epoch": 1.315507523175106,
+      "grad_norm": 0.11280883848667145,
+      "learning_rate": 0.00011228724065101255,
+      "loss": 0.0093,
+      "step": 7060
+    },
+    {
+      "epoch": 1.3173708482787534,
+      "grad_norm": 0.07703553140163422,
+      "learning_rate": 0.00011216300161510747,
+      "loss": 0.014,
+      "step": 7070
+    },
+    {
+      "epoch": 1.319234173382401,
+      "grad_norm": 0.1067732498049736,
+      "learning_rate": 0.00011203876257920238,
+      "loss": 0.012,
+      "step": 7080
+    },
+    {
+      "epoch": 1.3210974984860484,
+      "grad_norm": 0.12903323769569397,
+      "learning_rate": 0.0001119145235432973,
+      "loss": 0.0122,
+      "step": 7090
+    },
+    {
+      "epoch": 1.3229608235896957,
+      "grad_norm": 0.07112511992454529,
+      "learning_rate": 0.00011179028450739221,
+      "loss": 0.0149,
+      "step": 7100
+    },
+    {
+      "epoch": 1.3248241486933432,
+      "grad_norm": 0.08477038890123367,
+      "learning_rate": 0.00011166604547148716,
+      "loss": 0.0133,
+      "step": 7110
+    },
+    {
+      "epoch": 1.3266874737969907,
+      "grad_norm": 0.07981168478727341,
+      "learning_rate": 0.00011154180643558207,
+      "loss": 0.0127,
+      "step": 7120
+    },
+    {
+      "epoch": 1.3285507989006382,
+      "grad_norm": 0.15720146894454956,
+      "learning_rate": 0.00011141756739967699,
+      "loss": 0.0132,
+      "step": 7130
+    },
+    {
+      "epoch": 1.3304141240042857,
+      "grad_norm": 0.19825328886508942,
+      "learning_rate": 0.0001112933283637719,
+      "loss": 0.015,
+      "step": 7140
+    },
+    {
+      "epoch": 1.332277449107933,
+      "grad_norm": 0.11979183554649353,
+      "learning_rate": 0.00011116908932786683,
+      "loss": 0.0146,
+      "step": 7150
+    },
+    {
+      "epoch": 1.3341407742115805,
+      "grad_norm": 0.1986120492219925,
+      "learning_rate": 0.00011104485029196175,
+      "loss": 0.0146,
+      "step": 7160
+    },
+    {
+      "epoch": 1.336004099315228,
+      "grad_norm": 0.20982642471790314,
+      "learning_rate": 0.00011092061125605666,
+      "loss": 0.0122,
+      "step": 7170
+    },
+    {
+      "epoch": 1.3378674244188755,
+      "grad_norm": 0.12676341831684113,
+      "learning_rate": 0.00011079637222015158,
+      "loss": 0.0182,
+      "step": 7180
+    },
+    {
+      "epoch": 1.339730749522523,
+      "grad_norm": 0.072712741792202,
+      "learning_rate": 0.0001106721331842465,
+      "loss": 0.0169,
+      "step": 7190
+    },
+    {
+      "epoch": 1.3415940746261703,
+      "grad_norm": 0.07570649683475494,
+      "learning_rate": 0.00011054789414834141,
+      "loss": 0.0148,
+      "step": 7200
+    },
+    {
+      "epoch": 1.3434573997298178,
+      "grad_norm": 0.0949859768152237,
+      "learning_rate": 0.00011042365511243633,
+      "loss": 0.0126,
+      "step": 7210
+    },
+    {
+      "epoch": 1.3453207248334653,
+      "grad_norm": 0.0748329609632492,
+      "learning_rate": 0.00011029941607653124,
+      "loss": 0.016,
+      "step": 7220
+    },
+    {
+      "epoch": 1.3471840499371128,
+      "grad_norm": 0.10189643502235413,
+      "learning_rate": 0.00011017517704062616,
+      "loss": 0.0129,
+      "step": 7230
+    },
+    {
+      "epoch": 1.3490473750407603,
+      "grad_norm": 0.11995694786310196,
+      "learning_rate": 0.00011005093800472109,
+      "loss": 0.02,
+      "step": 7240
+    },
+    {
+      "epoch": 1.3509107001444076,
+      "grad_norm": 0.04982425644993782,
+      "learning_rate": 0.000109926698968816,
+      "loss": 0.0187,
+      "step": 7250
+    },
+    {
+      "epoch": 1.3527740252480551,
+      "grad_norm": 0.12379368394613266,
+      "learning_rate": 0.00010980245993291092,
+      "loss": 0.013,
+      "step": 7260
+    },
+    {
+      "epoch": 1.3546373503517026,
+      "grad_norm": 0.15470531582832336,
+      "learning_rate": 0.00010967822089700584,
+      "loss": 0.0111,
+      "step": 7270
+    },
+    {
+      "epoch": 1.3565006754553501,
+      "grad_norm": 0.14728385210037231,
+      "learning_rate": 0.00010955398186110078,
+      "loss": 0.0142,
+      "step": 7280
+    },
+    {
+      "epoch": 1.3583640005589976,
+      "grad_norm": 0.04794132709503174,
+      "learning_rate": 0.0001094297428251957,
+      "loss": 0.0132,
+      "step": 7290
+    },
+    {
+      "epoch": 1.360227325662645,
+      "grad_norm": 0.12964306771755219,
+      "learning_rate": 0.00010930550378929061,
+      "loss": 0.0133,
+      "step": 7300
+    },
+    {
+      "epoch": 1.3620906507662924,
+      "grad_norm": 0.2551344931125641,
+      "learning_rate": 0.00010918126475338553,
+      "loss": 0.0222,
+      "step": 7310
+    },
+    {
+      "epoch": 1.36395397586994,
+      "grad_norm": 0.06458201259374619,
+      "learning_rate": 0.00010905702571748044,
+      "loss": 0.0134,
+      "step": 7320
+    },
+    {
+      "epoch": 1.3658173009735874,
+      "grad_norm": 0.09689020365476608,
+      "learning_rate": 0.00010893278668157536,
+      "loss": 0.0097,
+      "step": 7330
+    },
+    {
+      "epoch": 1.367680626077235,
+      "grad_norm": 0.12692950665950775,
+      "learning_rate": 0.00010880854764567027,
+      "loss": 0.017,
+      "step": 7340
+    },
+    {
+      "epoch": 1.3695439511808822,
+      "grad_norm": 0.18245796859264374,
+      "learning_rate": 0.00010868430860976519,
+      "loss": 0.0139,
+      "step": 7350
+    },
+    {
+      "epoch": 1.3714072762845297,
+      "grad_norm": 0.1438094675540924,
+      "learning_rate": 0.00010856006957386012,
+      "loss": 0.0125,
+      "step": 7360
+    },
+    {
+      "epoch": 1.3732706013881772,
+      "grad_norm": 0.05395403876900673,
+      "learning_rate": 0.00010843583053795503,
+      "loss": 0.0151,
+      "step": 7370
+    },
+    {
+      "epoch": 1.3751339264918245,
+      "grad_norm": 0.06818302720785141,
+      "learning_rate": 0.00010831159150204995,
+      "loss": 0.0112,
+      "step": 7380
+    },
+    {
+      "epoch": 1.376997251595472,
+      "grad_norm": 0.09590564668178558,
+      "learning_rate": 0.00010818735246614487,
+      "loss": 0.0126,
+      "step": 7390
+    },
+    {
+      "epoch": 1.3788605766991195,
+      "grad_norm": 0.09465160220861435,
+      "learning_rate": 0.00010806311343023978,
+      "loss": 0.0118,
+      "step": 7400
+    },
+    {
+      "epoch": 1.380723901802767,
+      "grad_norm": 0.07145224511623383,
+      "learning_rate": 0.0001079388743943347,
+      "loss": 0.0105,
+      "step": 7410
+    },
+    {
+      "epoch": 1.3825872269064146,
+      "grad_norm": 0.12949031591415405,
+      "learning_rate": 0.00010781463535842961,
+      "loss": 0.0154,
+      "step": 7420
+    },
+    {
+      "epoch": 1.3844505520100618,
+      "grad_norm": 0.10481762140989304,
+      "learning_rate": 0.00010769039632252453,
+      "loss": 0.0193,
+      "step": 7430
+    },
+    {
+      "epoch": 1.3863138771137093,
+      "grad_norm": 0.08996088802814484,
+      "learning_rate": 0.00010756615728661945,
+      "loss": 0.013,
+      "step": 7440
+    },
+    {
+      "epoch": 1.3881772022173569,
+      "grad_norm": 0.07961221039295197,
+      "learning_rate": 0.00010744191825071439,
+      "loss": 0.0161,
+      "step": 7450
+    },
+    {
+      "epoch": 1.3900405273210044,
+      "grad_norm": 0.11494274437427521,
+      "learning_rate": 0.0001073176792148093,
+      "loss": 0.0117,
+      "step": 7460
+    },
+    {
+      "epoch": 1.3919038524246519,
+      "grad_norm": 0.16665033996105194,
+      "learning_rate": 0.00010719344017890422,
+      "loss": 0.013,
+      "step": 7470
+    },
+    {
+      "epoch": 1.3937671775282992,
+      "grad_norm": 0.09289297461509705,
+      "learning_rate": 0.00010706920114299915,
+      "loss": 0.015,
+      "step": 7480
+    },
+    {
+      "epoch": 1.3956305026319467,
+      "grad_norm": 0.06463921070098877,
+      "learning_rate": 0.00010694496210709406,
+      "loss": 0.0099,
+      "step": 7490
+    },
+    {
+      "epoch": 1.3974938277355942,
+      "grad_norm": 0.15225523710250854,
+      "learning_rate": 0.00010682072307118898,
+      "loss": 0.0124,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3993571528392417,
+      "grad_norm": 0.07819876074790955,
+      "learning_rate": 0.0001066964840352839,
+      "loss": 0.0106,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4012204779428892,
+      "grad_norm": 0.2872188985347748,
+      "learning_rate": 0.00010657224499937881,
+      "loss": 0.0191,
+      "step": 7520
+    },
+    {
+      "epoch": 1.4030838030465365,
+      "grad_norm": 0.13932381570339203,
+      "learning_rate": 0.00010644800596347373,
+      "loss": 0.015,
+      "step": 7530
+    },
+    {
+      "epoch": 1.404947128150184,
+      "grad_norm": 0.09094579517841339,
+      "learning_rate": 0.00010632376692756864,
+      "loss": 0.0141,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4068104532538315,
+      "grad_norm": 0.07863321900367737,
+      "learning_rate": 0.00010619952789166356,
+      "loss": 0.0123,
+      "step": 7550
+    },
+    {
+      "epoch": 1.408673778357479,
+      "grad_norm": 0.08263008296489716,
+      "learning_rate": 0.00010607528885575848,
+      "loss": 0.0128,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4105371034611265,
+      "grad_norm": 0.055199749767780304,
+      "learning_rate": 0.0001059510498198534,
+      "loss": 0.0158,
+      "step": 7570
+    },
+    {
+      "epoch": 1.4124004285647738,
+      "grad_norm": 0.09670916199684143,
+      "learning_rate": 0.00010582681078394832,
+      "loss": 0.0107,
+      "step": 7580
+    },
+    {
+      "epoch": 1.4142637536684213,
+      "grad_norm": 0.1617942899465561,
+      "learning_rate": 0.00010570257174804324,
+      "loss": 0.0173,
+      "step": 7590
+    },
+    {
+      "epoch": 1.4161270787720688,
+      "grad_norm": 0.099526546895504,
+      "learning_rate": 0.00010557833271213815,
+      "loss": 0.0127,
+      "step": 7600
+    },
+    {
+      "epoch": 1.4179904038757163,
+      "grad_norm": 0.18963083624839783,
+      "learning_rate": 0.00010545409367623307,
+      "loss": 0.011,
+      "step": 7610
+    },
+    {
+      "epoch": 1.4198537289793638,
+      "grad_norm": 0.11049555242061615,
+      "learning_rate": 0.00010532985464032798,
+      "loss": 0.0115,
+      "step": 7620
+    },
+    {
+      "epoch": 1.421717054083011,
+      "grad_norm": 0.0848945677280426,
+      "learning_rate": 0.00010520561560442293,
+      "loss": 0.0128,
+      "step": 7630
+    },
+    {
+      "epoch": 1.4235803791866586,
+      "grad_norm": 0.09841426461935043,
+      "learning_rate": 0.00010508137656851784,
+      "loss": 0.0116,
+      "step": 7640
+    },
+    {
+      "epoch": 1.425443704290306,
+      "grad_norm": 0.11386366933584213,
+      "learning_rate": 0.00010495713753261276,
+      "loss": 0.0141,
+      "step": 7650
+    },
+    {
+      "epoch": 1.4273070293939534,
+      "grad_norm": 0.10864854604005814,
+      "learning_rate": 0.00010483289849670767,
+      "loss": 0.0144,
+      "step": 7660
+    },
+    {
+      "epoch": 1.429170354497601,
+      "grad_norm": 0.09958741068840027,
+      "learning_rate": 0.00010470865946080259,
+      "loss": 0.0097,
+      "step": 7670
+    },
+    {
+      "epoch": 1.4310336796012484,
+      "grad_norm": 0.12464049458503723,
+      "learning_rate": 0.0001045844204248975,
+      "loss": 0.0146,
+      "step": 7680
+    },
+    {
+      "epoch": 1.432897004704896,
+      "grad_norm": 0.10106810927391052,
+      "learning_rate": 0.00010446018138899244,
+      "loss": 0.0082,
+      "step": 7690
+    },
+    {
+      "epoch": 1.4347603298085434,
+      "grad_norm": 0.10265982151031494,
+      "learning_rate": 0.00010433594235308735,
+      "loss": 0.0137,
+      "step": 7700
+    },
+    {
+      "epoch": 1.4366236549121907,
+      "grad_norm": 0.25061050057411194,
+      "learning_rate": 0.00010421170331718227,
+      "loss": 0.0153,
+      "step": 7710
+    },
+    {
+      "epoch": 1.4384869800158382,
+      "grad_norm": 0.15667365491390228,
+      "learning_rate": 0.00010408746428127718,
+      "loss": 0.0119,
+      "step": 7720
+    },
+    {
+      "epoch": 1.4403503051194857,
+      "grad_norm": 0.08727956563234329,
+      "learning_rate": 0.0001039632252453721,
+      "loss": 0.0156,
+      "step": 7730
+    },
+    {
+      "epoch": 1.4422136302231332,
+      "grad_norm": 0.05918029323220253,
+      "learning_rate": 0.00010383898620946701,
+      "loss": 0.0137,
+      "step": 7740
+    },
+    {
+      "epoch": 1.4440769553267807,
+      "grad_norm": 0.07321096211671829,
+      "learning_rate": 0.00010371474717356193,
+      "loss": 0.0114,
+      "step": 7750
+    },
+    {
+      "epoch": 1.445940280430428,
+      "grad_norm": 0.11011398583650589,
+      "learning_rate": 0.00010359050813765685,
+      "loss": 0.0125,
+      "step": 7760
+    },
+    {
+      "epoch": 1.4478036055340755,
+      "grad_norm": 0.10525079816579819,
+      "learning_rate": 0.00010346626910175176,
+      "loss": 0.0119,
+      "step": 7770
+    },
+    {
+      "epoch": 1.449666930637723,
+      "grad_norm": 0.09164225310087204,
+      "learning_rate": 0.00010334203006584669,
+      "loss": 0.0114,
+      "step": 7780
+    },
+    {
+      "epoch": 1.4515302557413705,
+      "grad_norm": 0.33539170026779175,
+      "learning_rate": 0.0001032177910299416,
+      "loss": 0.0199,
+      "step": 7790
+    },
+    {
+      "epoch": 1.453393580845018,
+      "grad_norm": 0.11161132156848907,
+      "learning_rate": 0.00010309355199403654,
+      "loss": 0.0142,
+      "step": 7800
+    },
+    {
+      "epoch": 1.4552569059486653,
+      "grad_norm": 0.06660373508930206,
+      "learning_rate": 0.00010296931295813145,
+      "loss": 0.0106,
+      "step": 7810
+    },
+    {
+      "epoch": 1.4571202310523128,
+      "grad_norm": 0.08710351586341858,
+      "learning_rate": 0.00010284507392222638,
+      "loss": 0.0137,
+      "step": 7820
+    },
+    {
+      "epoch": 1.4589835561559603,
+      "grad_norm": 0.0928850769996643,
+      "learning_rate": 0.0001027208348863213,
+      "loss": 0.0123,
+      "step": 7830
+    },
+    {
+      "epoch": 1.4608468812596078,
+      "grad_norm": 0.05278251692652702,
+      "learning_rate": 0.00010259659585041621,
+      "loss": 0.0141,
+      "step": 7840
+    },
+    {
+      "epoch": 1.4627102063632553,
+      "grad_norm": 0.15453903377056122,
+      "learning_rate": 0.00010247235681451113,
+      "loss": 0.0112,
+      "step": 7850
+    },
+    {
+      "epoch": 1.4645735314669026,
+      "grad_norm": 0.14208067953586578,
+      "learning_rate": 0.00010234811777860604,
+      "loss": 0.0151,
+      "step": 7860
+    },
+    {
+      "epoch": 1.4664368565705501,
+      "grad_norm": 0.12577812373638153,
+      "learning_rate": 0.00010222387874270096,
+      "loss": 0.0102,
+      "step": 7870
+    },
+    {
+      "epoch": 1.4683001816741976,
+      "grad_norm": 0.06806960701942444,
+      "learning_rate": 0.00010209963970679588,
+      "loss": 0.0146,
+      "step": 7880
+    },
+    {
+      "epoch": 1.4701635067778451,
+      "grad_norm": 0.23436611890792847,
+      "learning_rate": 0.00010197540067089079,
+      "loss": 0.0141,
+      "step": 7890
+    },
+    {
+      "epoch": 1.4720268318814926,
+      "grad_norm": 0.12878048419952393,
+      "learning_rate": 0.00010185116163498571,
+      "loss": 0.0093,
+      "step": 7900
+    },
+    {
+      "epoch": 1.47389015698514,
+      "grad_norm": 0.054260533303022385,
+      "learning_rate": 0.00010172692259908064,
+      "loss": 0.0124,
+      "step": 7910
+    },
+    {
+      "epoch": 1.4757534820887874,
+      "grad_norm": 0.09768477827310562,
+      "learning_rate": 0.00010160268356317555,
+      "loss": 0.0118,
+      "step": 7920
+    },
+    {
+      "epoch": 1.477616807192435,
+      "grad_norm": 0.10654323548078537,
+      "learning_rate": 0.00010147844452727047,
+      "loss": 0.0117,
+      "step": 7930
+    },
+    {
+      "epoch": 1.4794801322960824,
+      "grad_norm": 0.10613780468702316,
+      "learning_rate": 0.00010135420549136538,
+      "loss": 0.0136,
+      "step": 7940
+    },
+    {
+      "epoch": 1.48134345739973,
+      "grad_norm": 0.11709728837013245,
+      "learning_rate": 0.0001012299664554603,
+      "loss": 0.0113,
+      "step": 7950
+    },
+    {
+      "epoch": 1.4832067825033772,
+      "grad_norm": 0.05041942372918129,
+      "learning_rate": 0.00010110572741955522,
+      "loss": 0.0101,
+      "step": 7960
+    },
+    {
+      "epoch": 1.4850701076070247,
+      "grad_norm": 0.06611751019954681,
+      "learning_rate": 0.00010098148838365016,
+      "loss": 0.0119,
+      "step": 7970
+    },
+    {
+      "epoch": 1.4869334327106722,
+      "grad_norm": 0.11693181842565536,
+      "learning_rate": 0.00010085724934774507,
+      "loss": 0.0139,
+      "step": 7980
+    },
+    {
+      "epoch": 1.4887967578143195,
+      "grad_norm": 0.1092601865530014,
+      "learning_rate": 0.00010073301031183999,
+      "loss": 0.0104,
+      "step": 7990
+    },
+    {
+      "epoch": 1.490660082917967,
+      "grad_norm": 0.09393884241580963,
+      "learning_rate": 0.0001006087712759349,
+      "loss": 0.0109,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4925234080216145,
+      "grad_norm": 0.12773072719573975,
+      "learning_rate": 0.00010048453224002982,
+      "loss": 0.0175,
+      "step": 8010
+    },
+    {
+      "epoch": 1.494386733125262,
+      "grad_norm": 0.10976378619670868,
+      "learning_rate": 0.00010036029320412474,
+      "loss": 0.0125,
+      "step": 8020
+    },
+    {
+      "epoch": 1.4962500582289096,
+      "grad_norm": 0.09273622930049896,
+      "learning_rate": 0.00010023605416821967,
+      "loss": 0.0101,
+      "step": 8030
+    },
+    {
+      "epoch": 1.4981133833325568,
+      "grad_norm": 0.10113687813282013,
+      "learning_rate": 0.00010011181513231458,
+      "loss": 0.01,
+      "step": 8040
+    },
+    {
+      "epoch": 1.4999767084362043,
+      "grad_norm": 0.09726346284151077,
+      "learning_rate": 9.99875760964095e-05,
+      "loss": 0.0131,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5018400335398518,
+      "grad_norm": 0.09806101024150848,
+      "learning_rate": 9.986333706050441e-05,
+      "loss": 0.0123,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5037033586434994,
+      "grad_norm": 0.0637512356042862,
+      "learning_rate": 9.973909802459933e-05,
+      "loss": 0.0102,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5055666837471469,
+      "grad_norm": 0.13974900543689728,
+      "learning_rate": 9.961485898869425e-05,
+      "loss": 0.0107,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5074300088507941,
+      "grad_norm": 0.10583356767892838,
+      "learning_rate": 9.949061995278916e-05,
+      "loss": 0.0093,
+      "step": 8090
+    },
+    {
+      "epoch": 1.5092933339544417,
+      "grad_norm": 0.08961586654186249,
+      "learning_rate": 9.936638091688409e-05,
+      "loss": 0.0103,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5111566590580892,
+      "grad_norm": 0.12322380393743515,
+      "learning_rate": 9.924214188097901e-05,
+      "loss": 0.0126,
+      "step": 8110
+    },
+    {
+      "epoch": 1.5130199841617367,
+      "grad_norm": 0.09430979937314987,
+      "learning_rate": 9.911790284507392e-05,
+      "loss": 0.0165,
+      "step": 8120
+    },
+    {
+      "epoch": 1.5148833092653842,
+      "grad_norm": 0.08856257051229477,
+      "learning_rate": 9.899366380916885e-05,
+      "loss": 0.0106,
+      "step": 8130
+    },
+    {
+      "epoch": 1.5167466343690315,
+      "grad_norm": 0.26079264283180237,
+      "learning_rate": 9.886942477326377e-05,
+      "loss": 0.0123,
+      "step": 8140
+    },
+    {
+      "epoch": 1.518609959472679,
+      "grad_norm": 0.06865710020065308,
+      "learning_rate": 9.874518573735868e-05,
+      "loss": 0.0154,
+      "step": 8150
+    },
+    {
+      "epoch": 1.5204732845763265,
+      "grad_norm": 0.04991608113050461,
+      "learning_rate": 9.86209467014536e-05,
+      "loss": 0.0097,
+      "step": 8160
+    },
+    {
+      "epoch": 1.5223366096799738,
+      "grad_norm": 0.1879420429468155,
+      "learning_rate": 9.849670766554852e-05,
+      "loss": 0.015,
+      "step": 8170
+    },
+    {
+      "epoch": 1.5241999347836215,
+      "grad_norm": 0.08976439386606216,
+      "learning_rate": 9.837246862964345e-05,
+      "loss": 0.0089,
+      "step": 8180
+    },
+    {
+      "epoch": 1.5260632598872688,
+      "grad_norm": 0.08291597664356232,
+      "learning_rate": 9.824822959373836e-05,
+      "loss": 0.0122,
+      "step": 8190
+    },
+    {
+      "epoch": 1.5279265849909163,
+      "grad_norm": 0.15329314768314362,
+      "learning_rate": 9.812399055783328e-05,
+      "loss": 0.0159,
+      "step": 8200
+    },
+    {
+      "epoch": 1.5297899100945638,
+      "grad_norm": 0.05117841437458992,
+      "learning_rate": 9.799975152192819e-05,
+      "loss": 0.008,
+      "step": 8210
+    },
+    {
+      "epoch": 1.531653235198211,
+      "grad_norm": 0.15377692878246307,
+      "learning_rate": 9.787551248602311e-05,
+      "loss": 0.0302,
+      "step": 8220
+    },
+    {
+      "epoch": 1.5335165603018588,
+      "grad_norm": 0.1379365473985672,
+      "learning_rate": 9.775127345011802e-05,
+      "loss": 0.016,
+      "step": 8230
+    },
+    {
+      "epoch": 1.535379885405506,
+      "grad_norm": 0.1009238213300705,
+      "learning_rate": 9.762703441421295e-05,
+      "loss": 0.0113,
+      "step": 8240
+    },
+    {
+      "epoch": 1.5372432105091536,
+      "grad_norm": 0.05966337025165558,
+      "learning_rate": 9.750279537830787e-05,
+      "loss": 0.0108,
+      "step": 8250
+    },
+    {
+      "epoch": 1.539106535612801,
+      "grad_norm": 0.129754900932312,
+      "learning_rate": 9.737855634240279e-05,
+      "loss": 0.0137,
+      "step": 8260
+    },
+    {
+      "epoch": 1.5409698607164484,
+      "grad_norm": 0.09081212431192398,
+      "learning_rate": 9.725431730649771e-05,
+      "loss": 0.0116,
+      "step": 8270
+    },
+    {
+      "epoch": 1.542833185820096,
+      "grad_norm": 0.10083411633968353,
+      "learning_rate": 9.713007827059263e-05,
+      "loss": 0.0122,
+      "step": 8280
+    },
+    {
+      "epoch": 1.5446965109237434,
+      "grad_norm": 0.10409017652273178,
+      "learning_rate": 9.700583923468755e-05,
+      "loss": 0.01,
+      "step": 8290
+    },
+    {
+      "epoch": 1.5465598360273909,
+      "grad_norm": 0.07942856103181839,
+      "learning_rate": 9.688160019878246e-05,
+      "loss": 0.0103,
+      "step": 8300
+    },
+    {
+      "epoch": 1.5484231611310384,
+      "grad_norm": 0.09349773079156876,
+      "learning_rate": 9.675736116287738e-05,
+      "loss": 0.0122,
+      "step": 8310
+    },
+    {
+      "epoch": 1.5502864862346857,
+      "grad_norm": 0.05473365634679794,
+      "learning_rate": 9.66331221269723e-05,
+      "loss": 0.0101,
+      "step": 8320
+    },
+    {
+      "epoch": 1.5521498113383334,
+      "grad_norm": 0.12536931037902832,
+      "learning_rate": 9.650888309106721e-05,
+      "loss": 0.0154,
+      "step": 8330
+    },
+    {
+      "epoch": 1.5540131364419807,
+      "grad_norm": 0.08710866421461105,
+      "learning_rate": 9.638464405516214e-05,
+      "loss": 0.0121,
+      "step": 8340
+    },
+    {
+      "epoch": 1.5558764615456282,
+      "grad_norm": 0.2000790685415268,
+      "learning_rate": 9.626040501925705e-05,
+      "loss": 0.012,
+      "step": 8350
+    },
+    {
+      "epoch": 1.5577397866492757,
+      "grad_norm": 0.1262202113866806,
+      "learning_rate": 9.613616598335198e-05,
+      "loss": 0.0186,
+      "step": 8360
+    },
+    {
+      "epoch": 1.559603111752923,
+      "grad_norm": 0.102536141872406,
+      "learning_rate": 9.60119269474469e-05,
+      "loss": 0.0123,
+      "step": 8370
+    },
+    {
+      "epoch": 1.5614664368565707,
+      "grad_norm": 0.1080302819609642,
+      "learning_rate": 9.588768791154182e-05,
+      "loss": 0.0107,
+      "step": 8380
+    },
+    {
+      "epoch": 1.563329761960218,
+      "grad_norm": 0.04270603507757187,
+      "learning_rate": 9.576344887563673e-05,
+      "loss": 0.0123,
+      "step": 8390
+    },
+    {
+      "epoch": 1.5651930870638655,
+      "grad_norm": 0.07982005923986435,
+      "learning_rate": 9.563920983973165e-05,
+      "loss": 0.0225,
+      "step": 8400
+    },
+    {
+      "epoch": 1.567056412167513,
+      "grad_norm": 0.1117529571056366,
+      "learning_rate": 9.551497080382656e-05,
+      "loss": 0.0171,
+      "step": 8410
+    },
+    {
+      "epoch": 1.5689197372711603,
+      "grad_norm": 0.06128223240375519,
+      "learning_rate": 9.539073176792148e-05,
+      "loss": 0.0122,
+      "step": 8420
+    },
+    {
+      "epoch": 1.5707830623748078,
+      "grad_norm": 0.10701828449964523,
+      "learning_rate": 9.52664927320164e-05,
+      "loss": 0.0106,
+      "step": 8430
+    },
+    {
+      "epoch": 1.5726463874784553,
+      "grad_norm": 0.09437397867441177,
+      "learning_rate": 9.514225369611132e-05,
+      "loss": 0.0132,
+      "step": 8440
+    },
+    {
+      "epoch": 1.5745097125821028,
+      "grad_norm": 0.07649078965187073,
+      "learning_rate": 9.501801466020624e-05,
+      "loss": 0.0149,
+      "step": 8450
+    },
+    {
+      "epoch": 1.5763730376857503,
+      "grad_norm": 0.12064909189939499,
+      "learning_rate": 9.489377562430116e-05,
+      "loss": 0.0155,
+      "step": 8460
+    },
+    {
+      "epoch": 1.5782363627893976,
+      "grad_norm": 0.10540273040533066,
+      "learning_rate": 9.476953658839609e-05,
+      "loss": 0.0152,
+      "step": 8470
+    },
+    {
+      "epoch": 1.5800996878930451,
+      "grad_norm": 0.09898658096790314,
+      "learning_rate": 9.4645297552491e-05,
+      "loss": 0.0161,
+      "step": 8480
+    },
+    {
+      "epoch": 1.5819630129966926,
+      "grad_norm": 0.08065871894359589,
+      "learning_rate": 9.452105851658592e-05,
+      "loss": 0.0091,
+      "step": 8490
+    },
+    {
+      "epoch": 1.58382633810034,
+      "grad_norm": 0.04499693587422371,
+      "learning_rate": 9.439681948068083e-05,
+      "loss": 0.0105,
+      "step": 8500
+    },
+    {
+      "epoch": 1.5856896632039876,
+      "grad_norm": 0.04815511777997017,
+      "learning_rate": 9.427258044477575e-05,
+      "loss": 0.011,
+      "step": 8510
+    },
+    {
+      "epoch": 1.587552988307635,
+      "grad_norm": 0.11696461588144302,
+      "learning_rate": 9.414834140887068e-05,
+      "loss": 0.0113,
+      "step": 8520
+    },
+    {
+      "epoch": 1.5894163134112824,
+      "grad_norm": 0.06907574832439423,
+      "learning_rate": 9.40241023729656e-05,
+      "loss": 0.0149,
+      "step": 8530
+    },
+    {
+      "epoch": 1.59127963851493,
+      "grad_norm": 0.11363095790147781,
+      "learning_rate": 9.389986333706051e-05,
+      "loss": 0.0147,
+      "step": 8540
+    },
+    {
+      "epoch": 1.5931429636185772,
+      "grad_norm": 0.17413190007209778,
+      "learning_rate": 9.377562430115543e-05,
+      "loss": 0.0134,
+      "step": 8550
+    },
+    {
+      "epoch": 1.595006288722225,
+      "grad_norm": 0.08960302174091339,
+      "learning_rate": 9.365138526525034e-05,
+      "loss": 0.0102,
+      "step": 8560
+    },
+    {
+      "epoch": 1.5968696138258722,
+      "grad_norm": 0.023280244320631027,
+      "learning_rate": 9.352714622934527e-05,
+      "loss": 0.011,
+      "step": 8570
+    },
+    {
+      "epoch": 1.5987329389295197,
+      "grad_norm": 0.07135169953107834,
+      "learning_rate": 9.340290719344019e-05,
+      "loss": 0.0095,
+      "step": 8580
+    },
+    {
+      "epoch": 1.6005962640331672,
+      "grad_norm": 0.05813858285546303,
+      "learning_rate": 9.32786681575351e-05,
+      "loss": 0.0098,
+      "step": 8590
+    },
+    {
+      "epoch": 1.6024595891368145,
+      "grad_norm": 0.08756371587514877,
+      "learning_rate": 9.315442912163002e-05,
+      "loss": 0.0173,
+      "step": 8600
+    },
+    {
+      "epoch": 1.6043229142404623,
+      "grad_norm": 0.12145531177520752,
+      "learning_rate": 9.303019008572495e-05,
+      "loss": 0.0127,
+      "step": 8610
+    },
+    {
+      "epoch": 1.6061862393441095,
+      "grad_norm": 0.11375415325164795,
+      "learning_rate": 9.290595104981986e-05,
+      "loss": 0.0113,
+      "step": 8620
+    },
+    {
+      "epoch": 1.608049564447757,
+      "grad_norm": 0.056099046021699905,
+      "learning_rate": 9.278171201391478e-05,
+      "loss": 0.0118,
+      "step": 8630
+    },
+    {
+      "epoch": 1.6099128895514045,
+      "grad_norm": 0.11111286282539368,
+      "learning_rate": 9.26574729780097e-05,
+      "loss": 0.0131,
+      "step": 8640
+    },
+    {
+      "epoch": 1.6117762146550518,
+      "grad_norm": 0.08085739612579346,
+      "learning_rate": 9.253323394210461e-05,
+      "loss": 0.016,
+      "step": 8650
+    },
+    {
+      "epoch": 1.6136395397586996,
+      "grad_norm": 0.11837384849786758,
+      "learning_rate": 9.240899490619953e-05,
+      "loss": 0.0162,
+      "step": 8660
+    },
+    {
+      "epoch": 1.6155028648623468,
+      "grad_norm": 0.06581897288560867,
+      "learning_rate": 9.228475587029444e-05,
+      "loss": 0.0087,
+      "step": 8670
+    },
+    {
+      "epoch": 1.6173661899659943,
+      "grad_norm": 0.1268416792154312,
+      "learning_rate": 9.216051683438937e-05,
+      "loss": 0.0094,
+      "step": 8680
+    },
+    {
+      "epoch": 1.6192295150696419,
+      "grad_norm": 0.08039700239896774,
+      "learning_rate": 9.203627779848429e-05,
+      "loss": 0.0132,
+      "step": 8690
+    },
+    {
+      "epoch": 1.6210928401732891,
+      "grad_norm": 0.09226547926664352,
+      "learning_rate": 9.191203876257922e-05,
+      "loss": 0.0114,
+      "step": 8700
+    },
+    {
+      "epoch": 1.6229561652769366,
+      "grad_norm": 0.15534359216690063,
+      "learning_rate": 9.178779972667413e-05,
+      "loss": 0.0154,
+      "step": 8710
+    },
+    {
+      "epoch": 1.6248194903805842,
+      "grad_norm": 0.11844916641712189,
+      "learning_rate": 9.166356069076905e-05,
+      "loss": 0.0136,
+      "step": 8720
+    },
+    {
+      "epoch": 1.6266828154842317,
+      "grad_norm": 0.08913256973028183,
+      "learning_rate": 9.153932165486396e-05,
+      "loss": 0.0162,
+      "step": 8730
+    },
+    {
+      "epoch": 1.6285461405878792,
+      "grad_norm": 0.08861401677131653,
+      "learning_rate": 9.141508261895888e-05,
+      "loss": 0.0162,
+      "step": 8740
+    },
+    {
+      "epoch": 1.6304094656915264,
+      "grad_norm": 0.0848320946097374,
+      "learning_rate": 9.12908435830538e-05,
+      "loss": 0.0143,
+      "step": 8750
+    },
+    {
+      "epoch": 1.632272790795174,
+      "grad_norm": 0.12847183644771576,
+      "learning_rate": 9.116660454714871e-05,
+      "loss": 0.0271,
+      "step": 8760
+    },
+    {
+      "epoch": 1.6341361158988215,
+      "grad_norm": 0.09224189072847366,
+      "learning_rate": 9.104236551124363e-05,
+      "loss": 0.0099,
+      "step": 8770
+    },
+    {
+      "epoch": 1.6359994410024687,
+      "grad_norm": 0.09641575813293457,
+      "learning_rate": 9.091812647533856e-05,
+      "loss": 0.0098,
+      "step": 8780
+    },
+    {
+      "epoch": 1.6378627661061165,
+      "grad_norm": 0.07251239567995071,
+      "learning_rate": 9.079388743943347e-05,
+      "loss": 0.012,
+      "step": 8790
+    },
+    {
+      "epoch": 1.6397260912097638,
+      "grad_norm": 0.06500142812728882,
+      "learning_rate": 9.06696484035284e-05,
+      "loss": 0.013,
+      "step": 8800
+    },
+    {
+      "epoch": 1.6415894163134113,
+      "grad_norm": 0.106007419526577,
+      "learning_rate": 9.054540936762332e-05,
+      "loss": 0.0072,
+      "step": 8810
+    },
+    {
+      "epoch": 1.6434527414170588,
+      "grad_norm": 0.09809418767690659,
+      "learning_rate": 9.042117033171823e-05,
+      "loss": 0.0111,
+      "step": 8820
+    },
+    {
+      "epoch": 1.645316066520706,
+      "grad_norm": 0.07123219221830368,
+      "learning_rate": 9.029693129581315e-05,
+      "loss": 0.0111,
+      "step": 8830
+    },
+    {
+      "epoch": 1.6471793916243538,
+      "grad_norm": 0.1504216492176056,
+      "learning_rate": 9.017269225990806e-05,
+      "loss": 0.0138,
+      "step": 8840
+    },
+    {
+      "epoch": 1.649042716728001,
+      "grad_norm": 0.11574763059616089,
+      "learning_rate": 9.004845322400298e-05,
+      "loss": 0.0101,
+      "step": 8850
+    },
+    {
+      "epoch": 1.6509060418316486,
+      "grad_norm": 0.09216564893722534,
+      "learning_rate": 8.99242141880979e-05,
+      "loss": 0.0118,
+      "step": 8860
+    },
+    {
+      "epoch": 1.652769366935296,
+      "grad_norm": 0.050066202878952026,
+      "learning_rate": 8.979997515219283e-05,
+      "loss": 0.0115,
+      "step": 8870
+    },
+    {
+      "epoch": 1.6546326920389434,
+      "grad_norm": 0.07031989097595215,
+      "learning_rate": 8.967573611628774e-05,
+      "loss": 0.0126,
+      "step": 8880
+    },
+    {
+      "epoch": 1.656496017142591,
+      "grad_norm": 0.05028266832232475,
+      "learning_rate": 8.955149708038266e-05,
+      "loss": 0.0138,
+      "step": 8890
+    },
+    {
+      "epoch": 1.6583593422462384,
+      "grad_norm": 0.08418164402246475,
+      "learning_rate": 8.942725804447757e-05,
+      "loss": 0.0102,
+      "step": 8900
+    },
+    {
+      "epoch": 1.6602226673498859,
+      "grad_norm": 0.07395261526107788,
+      "learning_rate": 8.93030190085725e-05,
+      "loss": 0.0123,
+      "step": 8910
+    },
+    {
+      "epoch": 1.6620859924535334,
+      "grad_norm": 0.051781509071588516,
+      "learning_rate": 8.917877997266742e-05,
+      "loss": 0.0102,
+      "step": 8920
+    },
+    {
+      "epoch": 1.6639493175571807,
+      "grad_norm": 0.12200374156236649,
+      "learning_rate": 8.905454093676233e-05,
+      "loss": 0.0091,
+      "step": 8930
+    },
+    {
+      "epoch": 1.6658126426608284,
+      "grad_norm": 0.09385799616575241,
+      "learning_rate": 8.893030190085725e-05,
+      "loss": 0.0133,
+      "step": 8940
+    },
+    {
+      "epoch": 1.6676759677644757,
+      "grad_norm": 0.10892565548419952,
+      "learning_rate": 8.880606286495218e-05,
+      "loss": 0.0091,
+      "step": 8950
+    },
+    {
+      "epoch": 1.6695392928681232,
+      "grad_norm": 0.06951303780078888,
+      "learning_rate": 8.86818238290471e-05,
+      "loss": 0.0083,
+      "step": 8960
+    },
+    {
+      "epoch": 1.6714026179717707,
+      "grad_norm": 0.09501393139362335,
+      "learning_rate": 8.855758479314201e-05,
+      "loss": 0.0145,
+      "step": 8970
+    },
+    {
+      "epoch": 1.673265943075418,
+      "grad_norm": 0.10199866443872452,
+      "learning_rate": 8.843334575723693e-05,
+      "loss": 0.0107,
+      "step": 8980
+    },
+    {
+      "epoch": 1.6751292681790655,
+      "grad_norm": 0.07333105802536011,
+      "learning_rate": 8.830910672133184e-05,
+      "loss": 0.0111,
+      "step": 8990
+    },
+    {
+      "epoch": 1.676992593282713,
+      "grad_norm": 0.07807095348834991,
+      "learning_rate": 8.818486768542676e-05,
+      "loss": 0.0119,
+      "step": 9000
+    },
+    {
+      "epoch": 1.6788559183863605,
+      "grad_norm": 0.06234030798077583,
+      "learning_rate": 8.806062864952169e-05,
+      "loss": 0.0127,
+      "step": 9010
+    },
+    {
+      "epoch": 1.680719243490008,
+      "grad_norm": 0.07110045850276947,
+      "learning_rate": 8.79363896136166e-05,
+      "loss": 0.0099,
+      "step": 9020
+    },
+    {
+      "epoch": 1.6825825685936553,
+      "grad_norm": 0.0808567926287651,
+      "learning_rate": 8.781215057771152e-05,
+      "loss": 0.0115,
+      "step": 9030
+    },
+    {
+      "epoch": 1.6844458936973028,
+      "grad_norm": 0.07250918447971344,
+      "learning_rate": 8.768791154180645e-05,
+      "loss": 0.0106,
+      "step": 9040
+    },
+    {
+      "epoch": 1.6863092188009503,
+      "grad_norm": 0.05763540416955948,
+      "learning_rate": 8.756367250590136e-05,
+      "loss": 0.0132,
+      "step": 9050
+    },
+    {
+      "epoch": 1.6881725439045976,
+      "grad_norm": 0.07025442272424698,
+      "learning_rate": 8.743943346999628e-05,
+      "loss": 0.0122,
+      "step": 9060
+    },
+    {
+      "epoch": 1.6900358690082453,
+      "grad_norm": 0.12990747392177582,
+      "learning_rate": 8.73151944340912e-05,
+      "loss": 0.0137,
+      "step": 9070
+    },
+    {
+      "epoch": 1.6918991941118926,
+      "grad_norm": 0.07888498157262802,
+      "learning_rate": 8.719095539818611e-05,
+      "loss": 0.0071,
+      "step": 9080
+    },
+    {
+      "epoch": 1.69376251921554,
+      "grad_norm": 0.05839758738875389,
+      "learning_rate": 8.706671636228103e-05,
+      "loss": 0.0122,
+      "step": 9090
+    },
+    {
+      "epoch": 1.6956258443191876,
+      "grad_norm": 0.05912087857723236,
+      "learning_rate": 8.694247732637594e-05,
+      "loss": 0.0125,
+      "step": 9100
+    },
+    {
+      "epoch": 1.697489169422835,
+      "grad_norm": 0.13204564154148102,
+      "learning_rate": 8.681823829047086e-05,
+      "loss": 0.0104,
+      "step": 9110
+    },
+    {
+      "epoch": 1.6993524945264826,
+      "grad_norm": 0.12108772248029709,
+      "learning_rate": 8.669399925456579e-05,
+      "loss": 0.0143,
+      "step": 9120
+    },
+    {
+      "epoch": 1.70121581963013,
+      "grad_norm": 0.10718534886837006,
+      "learning_rate": 8.656976021866072e-05,
+      "loss": 0.0108,
+      "step": 9130
+    },
+    {
+      "epoch": 1.7030791447337774,
+      "grad_norm": 0.06931287050247192,
+      "learning_rate": 8.644552118275563e-05,
+      "loss": 0.0108,
+      "step": 9140
+    },
+    {
+      "epoch": 1.704942469837425,
+      "grad_norm": 0.12279055267572403,
+      "learning_rate": 8.632128214685055e-05,
+      "loss": 0.0131,
+      "step": 9150
+    },
+    {
+      "epoch": 1.7068057949410722,
+      "grad_norm": 0.06814446300268173,
+      "learning_rate": 8.619704311094547e-05,
+      "loss": 0.0101,
+      "step": 9160
+    },
+    {
+      "epoch": 1.70866912004472,
+      "grad_norm": 0.043399911373853683,
+      "learning_rate": 8.607280407504038e-05,
+      "loss": 0.0077,
+      "step": 9170
+    },
+    {
+      "epoch": 1.7105324451483672,
+      "grad_norm": 0.10185623914003372,
+      "learning_rate": 8.59485650391353e-05,
+      "loss": 0.0154,
+      "step": 9180
+    },
+    {
+      "epoch": 1.7123957702520147,
+      "grad_norm": 0.0697004571557045,
+      "learning_rate": 8.582432600323021e-05,
+      "loss": 0.0089,
+      "step": 9190
+    },
+    {
+      "epoch": 1.7142590953556622,
+      "grad_norm": 0.08959820121526718,
+      "learning_rate": 8.570008696732513e-05,
+      "loss": 0.0125,
+      "step": 9200
+    },
+    {
+      "epoch": 1.7161224204593095,
+      "grad_norm": 0.07031527906656265,
+      "learning_rate": 8.557584793142006e-05,
+      "loss": 0.0068,
+      "step": 9210
+    },
+    {
+      "epoch": 1.7179857455629572,
+      "grad_norm": 0.14567942917346954,
+      "learning_rate": 8.545160889551497e-05,
+      "loss": 0.0149,
+      "step": 9220
+    },
+    {
+      "epoch": 1.7198490706666045,
+      "grad_norm": 0.05701775848865509,
+      "learning_rate": 8.532736985960989e-05,
+      "loss": 0.0128,
+      "step": 9230
+    },
+    {
+      "epoch": 1.721712395770252,
+      "grad_norm": 0.1378256380558014,
+      "learning_rate": 8.520313082370482e-05,
+      "loss": 0.0117,
+      "step": 9240
+    },
+    {
+      "epoch": 1.7235757208738995,
+      "grad_norm": 0.1644400954246521,
+      "learning_rate": 8.507889178779973e-05,
+      "loss": 0.0126,
+      "step": 9250
+    },
+    {
+      "epoch": 1.7254390459775468,
+      "grad_norm": 0.1157638356089592,
+      "learning_rate": 8.495465275189465e-05,
+      "loss": 0.0091,
+      "step": 9260
+    },
+    {
+      "epoch": 1.7273023710811946,
+      "grad_norm": 0.06827396899461746,
+      "learning_rate": 8.483041371598957e-05,
+      "loss": 0.0115,
+      "step": 9270
+    },
+    {
+      "epoch": 1.7291656961848418,
+      "grad_norm": 0.07707670331001282,
+      "learning_rate": 8.470617468008448e-05,
+      "loss": 0.0104,
+      "step": 9280
+    },
+    {
+      "epoch": 1.7310290212884893,
+      "grad_norm": 0.09398248046636581,
+      "learning_rate": 8.45819356441794e-05,
+      "loss": 0.0119,
+      "step": 9290
+    },
+    {
+      "epoch": 1.7328923463921369,
+      "grad_norm": 0.051631245762109756,
+      "learning_rate": 8.445769660827433e-05,
+      "loss": 0.0094,
+      "step": 9300
+    },
+    {
+      "epoch": 1.7347556714957841,
+      "grad_norm": 0.06436185538768768,
+      "learning_rate": 8.433345757236924e-05,
+      "loss": 0.0131,
+      "step": 9310
+    },
+    {
+      "epoch": 1.7366189965994316,
+      "grad_norm": 0.07613561302423477,
+      "learning_rate": 8.420921853646416e-05,
+      "loss": 0.0107,
+      "step": 9320
+    },
+    {
+      "epoch": 1.7384823217030791,
+      "grad_norm": 0.07821378856897354,
+      "learning_rate": 8.408497950055907e-05,
+      "loss": 0.0156,
+      "step": 9330
+    },
+    {
+      "epoch": 1.7403456468067267,
+      "grad_norm": 0.06976544111967087,
+      "learning_rate": 8.396074046465399e-05,
+      "loss": 0.0108,
+      "step": 9340
+    },
+    {
+      "epoch": 1.7422089719103742,
+      "grad_norm": 0.09211380779743195,
+      "learning_rate": 8.383650142874892e-05,
+      "loss": 0.0106,
+      "step": 9350
+    },
+    {
+      "epoch": 1.7440722970140214,
+      "grad_norm": 0.09161022305488586,
+      "learning_rate": 8.371226239284384e-05,
+      "loss": 0.0144,
+      "step": 9360
+    },
+    {
+      "epoch": 1.745935622117669,
+      "grad_norm": 0.10251525789499283,
+      "learning_rate": 8.358802335693875e-05,
+      "loss": 0.0096,
+      "step": 9370
+    },
+    {
+      "epoch": 1.7477989472213165,
+      "grad_norm": 0.09510977566242218,
+      "learning_rate": 8.346378432103368e-05,
+      "loss": 0.013,
+      "step": 9380
+    },
+    {
+      "epoch": 1.7496622723249637,
+      "grad_norm": 0.13077139854431152,
+      "learning_rate": 8.33395452851286e-05,
+      "loss": 0.0106,
+      "step": 9390
+    },
+    {
+      "epoch": 1.7515255974286115,
+      "grad_norm": 0.09187795966863632,
+      "learning_rate": 8.321530624922351e-05,
+      "loss": 0.0089,
+      "step": 9400
+    },
+    {
+      "epoch": 1.7533889225322588,
+      "grad_norm": 0.13412202894687653,
+      "learning_rate": 8.309106721331843e-05,
+      "loss": 0.0107,
+      "step": 9410
+    },
+    {
+      "epoch": 1.7552522476359063,
+      "grad_norm": 0.07389464974403381,
+      "learning_rate": 8.296682817741334e-05,
+      "loss": 0.0109,
+      "step": 9420
+    },
+    {
+      "epoch": 1.7571155727395538,
+      "grad_norm": 0.10995054244995117,
+      "learning_rate": 8.284258914150826e-05,
+      "loss": 0.0124,
+      "step": 9430
+    },
+    {
+      "epoch": 1.758978897843201,
+      "grad_norm": 0.055603329092264175,
+      "learning_rate": 8.271835010560318e-05,
+      "loss": 0.012,
+      "step": 9440
+    },
+    {
+      "epoch": 1.7608422229468488,
+      "grad_norm": 0.1301773190498352,
+      "learning_rate": 8.25941110696981e-05,
+      "loss": 0.0166,
+      "step": 9450
+    },
+    {
+      "epoch": 1.762705548050496,
+      "grad_norm": 0.1451479196548462,
+      "learning_rate": 8.246987203379302e-05,
+      "loss": 0.0164,
+      "step": 9460
+    },
+    {
+      "epoch": 1.7645688731541436,
+      "grad_norm": 0.05127796530723572,
+      "learning_rate": 8.234563299788795e-05,
+      "loss": 0.0094,
+      "step": 9470
+    },
+    {
+      "epoch": 1.766432198257791,
+      "grad_norm": 0.11545635014772415,
+      "learning_rate": 8.222139396198287e-05,
+      "loss": 0.0161,
+      "step": 9480
+    },
+    {
+      "epoch": 1.7682955233614384,
+      "grad_norm": 0.10210834443569183,
+      "learning_rate": 8.209715492607778e-05,
+      "loss": 0.0093,
+      "step": 9490
+    },
+    {
+      "epoch": 1.770158848465086,
+      "grad_norm": 0.11474502831697464,
+      "learning_rate": 8.19729158901727e-05,
+      "loss": 0.0093,
+      "step": 9500
+    },
+    {
+      "epoch": 1.7720221735687334,
+      "grad_norm": 0.07989396154880524,
+      "learning_rate": 8.184867685426761e-05,
+      "loss": 0.011,
+      "step": 9510
+    },
+    {
+      "epoch": 1.7738854986723809,
+      "grad_norm": 0.0556521899998188,
+      "learning_rate": 8.172443781836253e-05,
+      "loss": 0.0126,
+      "step": 9520
+    },
+    {
+      "epoch": 1.7757488237760284,
+      "grad_norm": 0.1595865786075592,
+      "learning_rate": 8.160019878245745e-05,
+      "loss": 0.0107,
+      "step": 9530
+    },
+    {
+      "epoch": 1.7776121488796757,
+      "grad_norm": 0.11151939630508423,
+      "learning_rate": 8.147595974655236e-05,
+      "loss": 0.0114,
+      "step": 9540
+    },
+    {
+      "epoch": 1.7794754739833234,
+      "grad_norm": 0.1147758886218071,
+      "learning_rate": 8.135172071064729e-05,
+      "loss": 0.0097,
+      "step": 9550
+    },
+    {
+      "epoch": 1.7813387990869707,
+      "grad_norm": 0.09774407744407654,
+      "learning_rate": 8.12274816747422e-05,
+      "loss": 0.0114,
+      "step": 9560
+    },
+    {
+      "epoch": 1.7832021241906182,
+      "grad_norm": 0.08929464966058731,
+      "learning_rate": 8.110324263883714e-05,
+      "loss": 0.0103,
+      "step": 9570
+    },
+    {
+      "epoch": 1.7850654492942657,
+      "grad_norm": 0.05338837951421738,
+      "learning_rate": 8.097900360293205e-05,
+      "loss": 0.0073,
+      "step": 9580
+    },
+    {
+      "epoch": 1.786928774397913,
+      "grad_norm": 0.0910460576415062,
+      "learning_rate": 8.085476456702697e-05,
+      "loss": 0.0086,
+      "step": 9590
+    },
+    {
+      "epoch": 1.7887920995015605,
+      "grad_norm": 0.12020589411258698,
+      "learning_rate": 8.073052553112188e-05,
+      "loss": 0.0104,
+      "step": 9600
+    },
+    {
+      "epoch": 1.790655424605208,
+      "grad_norm": 0.0601915568113327,
+      "learning_rate": 8.06062864952168e-05,
+      "loss": 0.0085,
+      "step": 9610
+    },
+    {
+      "epoch": 1.7925187497088555,
+      "grad_norm": 0.0929914340376854,
+      "learning_rate": 8.048204745931171e-05,
+      "loss": 0.011,
+      "step": 9620
+    },
+    {
+      "epoch": 1.794382074812503,
+      "grad_norm": 0.1210533082485199,
+      "learning_rate": 8.035780842340663e-05,
+      "loss": 0.0095,
+      "step": 9630
+    },
+    {
+      "epoch": 1.7962453999161503,
+      "grad_norm": 0.08702056854963303,
+      "learning_rate": 8.023356938750156e-05,
+      "loss": 0.0083,
+      "step": 9640
+    },
+    {
+      "epoch": 1.7981087250197978,
+      "grad_norm": 0.09298733621835709,
+      "learning_rate": 8.010933035159648e-05,
+      "loss": 0.0116,
+      "step": 9650
+    },
+    {
+      "epoch": 1.7999720501234453,
+      "grad_norm": 0.06856974214315414,
+      "learning_rate": 7.998509131569139e-05,
+      "loss": 0.0124,
+      "step": 9660
+    },
+    {
+      "epoch": 1.8018353752270926,
+      "grad_norm": 0.10700483620166779,
+      "learning_rate": 7.986085227978631e-05,
+      "loss": 0.0099,
+      "step": 9670
+    },
+    {
+      "epoch": 1.8036987003307403,
+      "grad_norm": 0.23971673846244812,
+      "learning_rate": 7.973661324388124e-05,
+      "loss": 0.013,
+      "step": 9680
+    },
+    {
+      "epoch": 1.8055620254343876,
+      "grad_norm": 0.05989585071802139,
+      "learning_rate": 7.961237420797615e-05,
+      "loss": 0.0081,
+      "step": 9690
+    },
+    {
+      "epoch": 1.807425350538035,
+      "grad_norm": 0.06265757232904434,
+      "learning_rate": 7.948813517207107e-05,
+      "loss": 0.0086,
+      "step": 9700
+    },
+    {
+      "epoch": 1.8092886756416826,
+      "grad_norm": 0.09412927180528641,
+      "learning_rate": 7.936389613616598e-05,
+      "loss": 0.0123,
+      "step": 9710
+    },
+    {
+      "epoch": 1.81115200074533,
+      "grad_norm": 0.09182179719209671,
+      "learning_rate": 7.923965710026091e-05,
+      "loss": 0.0109,
+      "step": 9720
+    },
+    {
+      "epoch": 1.8130153258489776,
+      "grad_norm": 0.06258631497621536,
+      "learning_rate": 7.911541806435583e-05,
+      "loss": 0.0112,
+      "step": 9730
+    },
+    {
+      "epoch": 1.814878650952625,
+      "grad_norm": 0.06989934295415878,
+      "learning_rate": 7.899117902845075e-05,
+      "loss": 0.0087,
+      "step": 9740
+    },
+    {
+      "epoch": 1.8167419760562724,
+      "grad_norm": 0.09968297928571701,
+      "learning_rate": 7.886693999254566e-05,
+      "loss": 0.007,
+      "step": 9750
+    },
+    {
+      "epoch": 1.81860530115992,
+      "grad_norm": 0.07616010308265686,
+      "learning_rate": 7.874270095664058e-05,
+      "loss": 0.009,
+      "step": 9760
+    },
+    {
+      "epoch": 1.8204686262635672,
+      "grad_norm": 0.055172596126794815,
+      "learning_rate": 7.861846192073549e-05,
+      "loss": 0.0123,
+      "step": 9770
+    },
+    {
+      "epoch": 1.822331951367215,
+      "grad_norm": 0.09393038600683212,
+      "learning_rate": 7.849422288483041e-05,
+      "loss": 0.0096,
+      "step": 9780
+    },
+    {
+      "epoch": 1.8241952764708622,
+      "grad_norm": 0.04935317859053612,
+      "learning_rate": 7.836998384892534e-05,
+      "loss": 0.0097,
+      "step": 9790
+    },
+    {
+      "epoch": 1.8260586015745097,
+      "grad_norm": 0.08303016424179077,
+      "learning_rate": 7.824574481302025e-05,
+      "loss": 0.0092,
+      "step": 9800
+    },
+    {
+      "epoch": 1.8279219266781572,
+      "grad_norm": 0.07915756106376648,
+      "learning_rate": 7.812150577711518e-05,
+      "loss": 0.011,
+      "step": 9810
+    },
+    {
+      "epoch": 1.8297852517818045,
+      "grad_norm": 0.08060972392559052,
+      "learning_rate": 7.79972667412101e-05,
+      "loss": 0.0127,
+      "step": 9820
+    },
+    {
+      "epoch": 1.8316485768854522,
+      "grad_norm": 0.06542019546031952,
+      "learning_rate": 7.787302770530501e-05,
+      "loss": 0.0078,
+      "step": 9830
+    },
+    {
+      "epoch": 1.8335119019890995,
+      "grad_norm": 0.1199030801653862,
+      "learning_rate": 7.774878866939993e-05,
+      "loss": 0.0263,
+      "step": 9840
+    },
+    {
+      "epoch": 1.835375227092747,
+      "grad_norm": 0.05325644463300705,
+      "learning_rate": 7.762454963349485e-05,
+      "loss": 0.0144,
+      "step": 9850
+    },
+    {
+      "epoch": 1.8372385521963945,
+      "grad_norm": 0.10904191434383392,
+      "learning_rate": 7.750031059758976e-05,
+      "loss": 0.0097,
+      "step": 9860
+    },
+    {
+      "epoch": 1.8391018773000418,
+      "grad_norm": 0.06219085678458214,
+      "learning_rate": 7.737607156168468e-05,
+      "loss": 0.0106,
+      "step": 9870
+    },
+    {
+      "epoch": 1.8409652024036895,
+      "grad_norm": 0.13541138172149658,
+      "learning_rate": 7.72518325257796e-05,
+      "loss": 0.0106,
+      "step": 9880
+    },
+    {
+      "epoch": 1.8428285275073368,
+      "grad_norm": 0.02818934991955757,
+      "learning_rate": 7.712759348987452e-05,
+      "loss": 0.0078,
+      "step": 9890
+    },
+    {
+      "epoch": 1.8446918526109843,
+      "grad_norm": 0.07589732110500336,
+      "learning_rate": 7.700335445396944e-05,
+      "loss": 0.0104,
+      "step": 9900
+    },
+    {
+      "epoch": 1.8465551777146318,
+      "grad_norm": 0.07196231186389923,
+      "learning_rate": 7.687911541806437e-05,
+      "loss": 0.01,
+      "step": 9910
+    },
+    {
+      "epoch": 1.8484185028182791,
+      "grad_norm": 0.09193772822618484,
+      "learning_rate": 7.675487638215928e-05,
+      "loss": 0.0093,
+      "step": 9920
+    },
+    {
+      "epoch": 1.8502818279219266,
+      "grad_norm": 0.1183035671710968,
+      "learning_rate": 7.66306373462542e-05,
+      "loss": 0.0129,
+      "step": 9930
+    },
+    {
+      "epoch": 1.8521451530255741,
+      "grad_norm": 0.19073550403118134,
+      "learning_rate": 7.650639831034912e-05,
+      "loss": 0.0133,
+      "step": 9940
+    },
+    {
+      "epoch": 1.8540084781292216,
+      "grad_norm": 0.09900416433811188,
+      "learning_rate": 7.638215927444403e-05,
+      "loss": 0.0202,
+      "step": 9950
+    },
+    {
+      "epoch": 1.8558718032328692,
+      "grad_norm": 0.1016170009970665,
+      "learning_rate": 7.625792023853895e-05,
+      "loss": 0.0104,
+      "step": 9960
+    },
+    {
+      "epoch": 1.8577351283365164,
+      "grad_norm": 0.06790608912706375,
+      "learning_rate": 7.613368120263386e-05,
+      "loss": 0.012,
+      "step": 9970
+    },
+    {
+      "epoch": 1.859598453440164,
+      "grad_norm": 0.058812689036130905,
+      "learning_rate": 7.600944216672879e-05,
+      "loss": 0.0078,
+      "step": 9980
+    },
+    {
+      "epoch": 1.8614617785438115,
+      "grad_norm": 0.1043725460767746,
+      "learning_rate": 7.588520313082371e-05,
+      "loss": 0.0076,
+      "step": 9990
+    },
+    {
+      "epoch": 1.8633251036474587,
+      "grad_norm": 0.08266306668519974,
+      "learning_rate": 7.576096409491862e-05,
+      "loss": 0.0095,
+      "step": 10000
+    },
+    {
+      "epoch": 1.8651884287511065,
+      "grad_norm": 0.27030545473098755,
+      "learning_rate": 7.563672505901355e-05,
+      "loss": 0.0129,
+      "step": 10010
+    },
+    {
+      "epoch": 1.8670517538547537,
+      "grad_norm": 0.06581460684537888,
+      "learning_rate": 7.551248602310847e-05,
+      "loss": 0.0107,
+      "step": 10020
+    },
+    {
+      "epoch": 1.8689150789584013,
+      "grad_norm": 0.046561527997255325,
+      "learning_rate": 7.538824698720338e-05,
+      "loss": 0.0115,
+      "step": 10030
+    },
+    {
+      "epoch": 1.8707784040620488,
+      "grad_norm": 0.0690365731716156,
+      "learning_rate": 7.52640079512983e-05,
+      "loss": 0.0081,
+      "step": 10040
+    },
+    {
+      "epoch": 1.872641729165696,
+      "grad_norm": 0.08930855244398117,
+      "learning_rate": 7.513976891539322e-05,
+      "loss": 0.0094,
+      "step": 10050
+    },
+    {
+      "epoch": 1.8745050542693438,
+      "grad_norm": 0.056025706231594086,
+      "learning_rate": 7.501552987948813e-05,
+      "loss": 0.0095,
+      "step": 10060
+    },
+    {
+      "epoch": 1.876368379372991,
+      "grad_norm": 0.0407416895031929,
+      "learning_rate": 7.489129084358306e-05,
+      "loss": 0.0115,
+      "step": 10070
+    },
+    {
+      "epoch": 1.8782317044766386,
+      "grad_norm": 0.09971708804368973,
+      "learning_rate": 7.476705180767798e-05,
+      "loss": 0.0104,
+      "step": 10080
+    },
+    {
+      "epoch": 1.880095029580286,
+      "grad_norm": 0.0750962570309639,
+      "learning_rate": 7.46428127717729e-05,
+      "loss": 0.0086,
+      "step": 10090
+    },
+    {
+      "epoch": 1.8819583546839334,
+      "grad_norm": 0.06815943866968155,
+      "learning_rate": 7.451857373586781e-05,
+      "loss": 0.009,
+      "step": 10100
+    },
+    {
+      "epoch": 1.883821679787581,
+      "grad_norm": 0.08812274783849716,
+      "learning_rate": 7.439433469996272e-05,
+      "loss": 0.0105,
+      "step": 10110
+    },
+    {
+      "epoch": 1.8856850048912284,
+      "grad_norm": 0.06751228123903275,
+      "learning_rate": 7.427009566405765e-05,
+      "loss": 0.007,
+      "step": 10120
+    },
+    {
+      "epoch": 1.8875483299948759,
+      "grad_norm": 0.1457456648349762,
+      "learning_rate": 7.414585662815257e-05,
+      "loss": 0.0121,
+      "step": 10130
+    },
+    {
+      "epoch": 1.8894116550985234,
+      "grad_norm": 0.05187823250889778,
+      "learning_rate": 7.402161759224749e-05,
+      "loss": 0.0112,
+      "step": 10140
+    },
+    {
+      "epoch": 1.8912749802021707,
+      "grad_norm": 0.05900781601667404,
+      "learning_rate": 7.389737855634242e-05,
+      "loss": 0.0089,
+      "step": 10150
+    },
+    {
+      "epoch": 1.8931383053058184,
+      "grad_norm": 0.09933795034885406,
+      "learning_rate": 7.377313952043733e-05,
+      "loss": 0.0102,
+      "step": 10160
+    },
+    {
+      "epoch": 1.8950016304094657,
+      "grad_norm": 0.09261105209589005,
+      "learning_rate": 7.364890048453225e-05,
+      "loss": 0.0111,
+      "step": 10170
+    },
+    {
+      "epoch": 1.8968649555131132,
+      "grad_norm": 0.0641617551445961,
+      "learning_rate": 7.352466144862716e-05,
+      "loss": 0.0082,
+      "step": 10180
+    },
+    {
+      "epoch": 1.8987282806167607,
+      "grad_norm": 0.09815941751003265,
+      "learning_rate": 7.340042241272208e-05,
+      "loss": 0.0115,
+      "step": 10190
+    },
+    {
+      "epoch": 1.900591605720408,
+      "grad_norm": 0.12402988225221634,
+      "learning_rate": 7.3276183376817e-05,
+      "loss": 0.0092,
+      "step": 10200
+    },
+    {
+      "epoch": 1.9024549308240555,
+      "grad_norm": 0.07279510796070099,
+      "learning_rate": 7.315194434091191e-05,
+      "loss": 0.0119,
+      "step": 10210
+    },
+    {
+      "epoch": 1.904318255927703,
+      "grad_norm": 0.12178485840559006,
+      "learning_rate": 7.302770530500683e-05,
+      "loss": 0.0094,
+      "step": 10220
+    },
+    {
+      "epoch": 1.9061815810313505,
+      "grad_norm": 0.07118740677833557,
+      "learning_rate": 7.290346626910176e-05,
+      "loss": 0.0117,
+      "step": 10230
+    },
+    {
+      "epoch": 1.908044906134998,
+      "grad_norm": 0.15892274677753448,
+      "learning_rate": 7.277922723319668e-05,
+      "loss": 0.0082,
+      "step": 10240
+    },
+    {
+      "epoch": 1.9099082312386453,
+      "grad_norm": 0.058475472033023834,
+      "learning_rate": 7.26549881972916e-05,
+      "loss": 0.0105,
+      "step": 10250
+    },
+    {
+      "epoch": 1.9117715563422928,
+      "grad_norm": 0.07305944710969925,
+      "learning_rate": 7.253074916138652e-05,
+      "loss": 0.0103,
+      "step": 10260
+    },
+    {
+      "epoch": 1.9136348814459403,
+      "grad_norm": 0.05083195120096207,
+      "learning_rate": 7.240651012548143e-05,
+      "loss": 0.0097,
+      "step": 10270
+    },
+    {
+      "epoch": 1.9154982065495876,
+      "grad_norm": 0.10659267753362656,
+      "learning_rate": 7.228227108957635e-05,
+      "loss": 0.0127,
+      "step": 10280
+    },
+    {
+      "epoch": 1.9173615316532353,
+      "grad_norm": 0.05836571007966995,
+      "learning_rate": 7.215803205367126e-05,
+      "loss": 0.0118,
+      "step": 10290
+    },
+    {
+      "epoch": 1.9192248567568826,
+      "grad_norm": 0.07619835436344147,
+      "learning_rate": 7.203379301776618e-05,
+      "loss": 0.0128,
+      "step": 10300
+    },
+    {
+      "epoch": 1.92108818186053,
+      "grad_norm": 0.0604756698012352,
+      "learning_rate": 7.19095539818611e-05,
+      "loss": 0.0094,
+      "step": 10310
+    },
+    {
+      "epoch": 1.9229515069641776,
+      "grad_norm": 0.08139246702194214,
+      "learning_rate": 7.178531494595602e-05,
+      "loss": 0.008,
+      "step": 10320
+    },
+    {
+      "epoch": 1.9248148320678249,
+      "grad_norm": 0.04249119386076927,
+      "learning_rate": 7.166107591005094e-05,
+      "loss": 0.0117,
+      "step": 10330
+    },
+    {
+      "epoch": 1.9266781571714726,
+      "grad_norm": 0.06616081297397614,
+      "learning_rate": 7.153683687414586e-05,
+      "loss": 0.0099,
+      "step": 10340
+    },
+    {
+      "epoch": 1.92854148227512,
+      "grad_norm": 0.1366560459136963,
+      "learning_rate": 7.141259783824079e-05,
+      "loss": 0.0097,
+      "step": 10350
+    },
+    {
+      "epoch": 1.9304048073787674,
+      "grad_norm": 0.06568538397550583,
+      "learning_rate": 7.12883588023357e-05,
+      "loss": 0.0084,
+      "step": 10360
+    },
+    {
+      "epoch": 1.932268132482415,
+      "grad_norm": 0.06210561469197273,
+      "learning_rate": 7.116411976643062e-05,
+      "loss": 0.0112,
+      "step": 10370
+    },
+    {
+      "epoch": 1.9341314575860622,
+      "grad_norm": 0.0465511716902256,
+      "learning_rate": 7.103988073052553e-05,
+      "loss": 0.0127,
+      "step": 10380
+    },
+    {
+      "epoch": 1.93599478268971,
+      "grad_norm": 0.06106647104024887,
+      "learning_rate": 7.091564169462045e-05,
+      "loss": 0.0098,
+      "step": 10390
+    },
+    {
+      "epoch": 1.9378581077933572,
+      "grad_norm": 0.05775139108300209,
+      "learning_rate": 7.079140265871536e-05,
+      "loss": 0.0098,
+      "step": 10400
+    },
+    {
+      "epoch": 1.9397214328970047,
+      "grad_norm": 0.038371358066797256,
+      "learning_rate": 7.06671636228103e-05,
+      "loss": 0.0109,
+      "step": 10410
+    },
+    {
+      "epoch": 1.9415847580006522,
+      "grad_norm": 0.0670790821313858,
+      "learning_rate": 7.054292458690521e-05,
+      "loss": 0.0111,
+      "step": 10420
+    },
+    {
+      "epoch": 1.9434480831042995,
+      "grad_norm": 0.13203051686286926,
+      "learning_rate": 7.041868555100013e-05,
+      "loss": 0.0102,
+      "step": 10430
+    },
+    {
+      "epoch": 1.9453114082079472,
+      "grad_norm": 0.06752142310142517,
+      "learning_rate": 7.029444651509504e-05,
+      "loss": 0.0147,
+      "step": 10440
+    },
+    {
+      "epoch": 1.9471747333115945,
+      "grad_norm": 0.1533273160457611,
+      "learning_rate": 7.017020747918997e-05,
+      "loss": 0.0132,
+      "step": 10450
+    },
+    {
+      "epoch": 1.949038058415242,
+      "grad_norm": 0.06826525926589966,
+      "learning_rate": 7.004596844328489e-05,
+      "loss": 0.0088,
+      "step": 10460
+    },
+    {
+      "epoch": 1.9509013835188895,
+      "grad_norm": 0.10859847068786621,
+      "learning_rate": 6.99217294073798e-05,
+      "loss": 0.0135,
+      "step": 10470
+    },
+    {
+      "epoch": 1.9527647086225368,
+      "grad_norm": 0.08629245311021805,
+      "learning_rate": 6.979749037147472e-05,
+      "loss": 0.01,
+      "step": 10480
+    },
+    {
+      "epoch": 1.9546280337261843,
+      "grad_norm": 0.08627601712942123,
+      "learning_rate": 6.967325133556965e-05,
+      "loss": 0.0083,
+      "step": 10490
+    },
+    {
+      "epoch": 1.9564913588298318,
+      "grad_norm": 0.09831984341144562,
+      "learning_rate": 6.954901229966456e-05,
+      "loss": 0.012,
+      "step": 10500
+    },
+    {
+      "epoch": 1.9564913588298318,
+      "eval_loss": 0.009778189472854137,
+      "eval_runtime": 3.848,
+      "eval_samples_per_second": 51.975,
+      "eval_steps_per_second": 12.994,
+      "step": 10500
+    },
+    {
+      "epoch": 1.9583546839334793,
+      "grad_norm": 0.06444848328828812,
+      "learning_rate": 6.942477326375948e-05,
+      "loss": 0.0122,
+      "step": 10510
+    },
+    {
+      "epoch": 1.9602180090371268,
+      "grad_norm": 0.06674704700708389,
+      "learning_rate": 6.93005342278544e-05,
+      "loss": 0.0075,
+      "step": 10520
+    },
+    {
+      "epoch": 1.9620813341407741,
+      "grad_norm": 0.09635625034570694,
+      "learning_rate": 6.917629519194931e-05,
+      "loss": 0.0103,
+      "step": 10530
+    },
+    {
+      "epoch": 1.9639446592444216,
+      "grad_norm": 0.08825049549341202,
+      "learning_rate": 6.905205615604423e-05,
+      "loss": 0.0132,
+      "step": 10540
+    },
+    {
+      "epoch": 1.9658079843480691,
+      "grad_norm": 0.11532425135374069,
+      "learning_rate": 6.892781712013914e-05,
+      "loss": 0.0097,
+      "step": 10550
+    },
+    {
+      "epoch": 1.9676713094517166,
+      "grad_norm": 0.06080237403512001,
+      "learning_rate": 6.880357808423407e-05,
+      "loss": 0.0073,
+      "step": 10560
+    },
+    {
+      "epoch": 1.9695346345553641,
+      "grad_norm": 0.11655959486961365,
+      "learning_rate": 6.867933904832899e-05,
+      "loss": 0.0095,
+      "step": 10570
+    },
+    {
+      "epoch": 1.9713979596590114,
+      "grad_norm": 0.12663961946964264,
+      "learning_rate": 6.855510001242392e-05,
+      "loss": 0.0091,
+      "step": 10580
+    },
+    {
+      "epoch": 1.973261284762659,
+      "grad_norm": 0.07252644747495651,
+      "learning_rate": 6.843086097651883e-05,
+      "loss": 0.0121,
+      "step": 10590
+    },
+    {
+      "epoch": 1.9751246098663064,
+      "grad_norm": 0.06399507820606232,
+      "learning_rate": 6.830662194061375e-05,
+      "loss": 0.0079,
+      "step": 10600
+    },
+    {
+      "epoch": 1.9769879349699537,
+      "grad_norm": 0.13027061522006989,
+      "learning_rate": 6.818238290470866e-05,
+      "loss": 0.0116,
+      "step": 10610
+    },
+    {
+      "epoch": 1.9788512600736015,
+      "grad_norm": 0.09459857642650604,
+      "learning_rate": 6.805814386880358e-05,
+      "loss": 0.01,
+      "step": 10620
+    },
+    {
+      "epoch": 1.9807145851772487,
+      "grad_norm": 0.06579575687646866,
+      "learning_rate": 6.79339048328985e-05,
+      "loss": 0.0105,
+      "step": 10630
+    },
+    {
+      "epoch": 1.9825779102808962,
+      "grad_norm": 0.08901679515838623,
+      "learning_rate": 6.780966579699341e-05,
+      "loss": 0.008,
+      "step": 10640
+    },
+    {
+      "epoch": 1.9844412353845438,
+      "grad_norm": 0.1649041771888733,
+      "learning_rate": 6.768542676108833e-05,
+      "loss": 0.0141,
+      "step": 10650
+    },
+    {
+      "epoch": 1.986304560488191,
+      "grad_norm": 0.052169207483530045,
+      "learning_rate": 6.756118772518326e-05,
+      "loss": 0.0104,
+      "step": 10660
+    },
+    {
+      "epoch": 1.9881678855918388,
+      "grad_norm": 0.06536156684160233,
+      "learning_rate": 6.743694868927817e-05,
+      "loss": 0.0121,
+      "step": 10670
+    },
+    {
+      "epoch": 1.990031210695486,
+      "grad_norm": 0.1085677444934845,
+      "learning_rate": 6.73127096533731e-05,
+      "loss": 0.0109,
+      "step": 10680
+    },
+    {
+      "epoch": 1.9918945357991336,
+      "grad_norm": 0.04441511258482933,
+      "learning_rate": 6.718847061746802e-05,
+      "loss": 0.007,
+      "step": 10690
+    },
+    {
+      "epoch": 1.993757860902781,
+      "grad_norm": 0.06156736612319946,
+      "learning_rate": 6.706423158156293e-05,
+      "loss": 0.008,
+      "step": 10700
+    },
+    {
+      "epoch": 1.9956211860064283,
+      "grad_norm": 0.05205359682440758,
+      "learning_rate": 6.693999254565785e-05,
+      "loss": 0.0076,
+      "step": 10710
+    },
+    {
+      "epoch": 1.997484511110076,
+      "grad_norm": 0.11469718813896179,
+      "learning_rate": 6.681575350975277e-05,
+      "loss": 0.0138,
+      "step": 10720
+    },
+    {
+      "epoch": 1.9993478362137234,
+      "grad_norm": 0.10553640872240067,
+      "learning_rate": 6.669151447384768e-05,
+      "loss": 0.0099,
+      "step": 10730
+    },
+    {
+      "epoch": 2.0012111613173706,
+      "grad_norm": 0.06447623670101166,
+      "learning_rate": 6.65672754379426e-05,
+      "loss": 0.007,
+      "step": 10740
+    },
+    {
+      "epoch": 2.0030744864210184,
+      "grad_norm": 0.102451391518116,
+      "learning_rate": 6.644303640203753e-05,
+      "loss": 0.006,
+      "step": 10750
+    },
+    {
+      "epoch": 2.0049378115246657,
+      "grad_norm": 0.030060315504670143,
+      "learning_rate": 6.631879736613244e-05,
+      "loss": 0.0116,
+      "step": 10760
+    },
+    {
+      "epoch": 2.0068011366283134,
+      "grad_norm": 0.05747315660119057,
+      "learning_rate": 6.619455833022736e-05,
+      "loss": 0.0067,
+      "step": 10770
+    },
+    {
+      "epoch": 2.0086644617319607,
+      "grad_norm": 0.09092244505882263,
+      "learning_rate": 6.607031929432227e-05,
+      "loss": 0.0068,
+      "step": 10780
+    },
+    {
+      "epoch": 2.010527786835608,
+      "grad_norm": 0.05836044251918793,
+      "learning_rate": 6.59460802584172e-05,
+      "loss": 0.0073,
+      "step": 10790
+    },
+    {
+      "epoch": 2.0123911119392557,
+      "grad_norm": 0.05856243893504143,
+      "learning_rate": 6.582184122251212e-05,
+      "loss": 0.0052,
+      "step": 10800
+    },
+    {
+      "epoch": 2.014254437042903,
+      "grad_norm": 0.08944112807512283,
+      "learning_rate": 6.569760218660703e-05,
+      "loss": 0.0072,
+      "step": 10810
+    },
+    {
+      "epoch": 2.0161177621465507,
+      "grad_norm": 0.06631997972726822,
+      "learning_rate": 6.557336315070195e-05,
+      "loss": 0.0063,
+      "step": 10820
+    },
+    {
+      "epoch": 2.017981087250198,
+      "grad_norm": 0.09783359616994858,
+      "learning_rate": 6.544912411479687e-05,
+      "loss": 0.006,
+      "step": 10830
+    },
+    {
+      "epoch": 2.0198444123538453,
+      "grad_norm": 0.1494947224855423,
+      "learning_rate": 6.53248850788918e-05,
+      "loss": 0.0071,
+      "step": 10840
+    },
+    {
+      "epoch": 2.021707737457493,
+      "grad_norm": 0.07631028443574905,
+      "learning_rate": 6.520064604298671e-05,
+      "loss": 0.0072,
+      "step": 10850
+    },
+    {
+      "epoch": 2.0235710625611403,
+      "grad_norm": 0.08964214473962784,
+      "learning_rate": 6.507640700708163e-05,
+      "loss": 0.0062,
+      "step": 10860
+    },
+    {
+      "epoch": 2.025434387664788,
+      "grad_norm": 0.0452706478536129,
+      "learning_rate": 6.495216797117654e-05,
+      "loss": 0.0061,
+      "step": 10870
+    },
+    {
+      "epoch": 2.0272977127684353,
+      "grad_norm": 0.050183720886707306,
+      "learning_rate": 6.482792893527146e-05,
+      "loss": 0.0065,
+      "step": 10880
+    },
+    {
+      "epoch": 2.0291610378720826,
+      "grad_norm": 0.06710134446620941,
+      "learning_rate": 6.470368989936639e-05,
+      "loss": 0.0041,
+      "step": 10890
+    },
+    {
+      "epoch": 2.0310243629757303,
+      "grad_norm": 0.0552317276597023,
+      "learning_rate": 6.45794508634613e-05,
+      "loss": 0.0122,
+      "step": 10900
+    },
+    {
+      "epoch": 2.0328876880793776,
+      "grad_norm": 0.07701552659273148,
+      "learning_rate": 6.445521182755622e-05,
+      "loss": 0.0061,
+      "step": 10910
+    },
+    {
+      "epoch": 2.0347510131830253,
+      "grad_norm": 0.021767565980553627,
+      "learning_rate": 6.433097279165115e-05,
+      "loss": 0.0058,
+      "step": 10920
+    },
+    {
+      "epoch": 2.0366143382866726,
+      "grad_norm": 0.07532000541687012,
+      "learning_rate": 6.420673375574607e-05,
+      "loss": 0.0065,
+      "step": 10930
+    },
+    {
+      "epoch": 2.03847766339032,
+      "grad_norm": 0.06637753546237946,
+      "learning_rate": 6.408249471984098e-05,
+      "loss": 0.011,
+      "step": 10940
+    },
+    {
+      "epoch": 2.0403409884939676,
+      "grad_norm": 0.08492667973041534,
+      "learning_rate": 6.39582556839359e-05,
+      "loss": 0.0137,
+      "step": 10950
+    },
+    {
+      "epoch": 2.042204313597615,
+      "grad_norm": 0.06736136972904205,
+      "learning_rate": 6.383401664803081e-05,
+      "loss": 0.007,
+      "step": 10960
+    },
+    {
+      "epoch": 2.0440676387012626,
+      "grad_norm": 0.14675526320934296,
+      "learning_rate": 6.370977761212573e-05,
+      "loss": 0.0077,
+      "step": 10970
+    },
+    {
+      "epoch": 2.04593096380491,
+      "grad_norm": 0.05233339965343475,
+      "learning_rate": 6.358553857622064e-05,
+      "loss": 0.007,
+      "step": 10980
+    },
+    {
+      "epoch": 2.047794288908557,
+      "grad_norm": 0.05443247780203819,
+      "learning_rate": 6.346129954031556e-05,
+      "loss": 0.0073,
+      "step": 10990
+    },
+    {
+      "epoch": 2.049657614012205,
+      "grad_norm": 0.07367470115423203,
+      "learning_rate": 6.333706050441049e-05,
+      "loss": 0.0071,
+      "step": 11000
+    },
+    {
+      "epoch": 2.051520939115852,
+      "grad_norm": 0.04883608594536781,
+      "learning_rate": 6.321282146850542e-05,
+      "loss": 0.0063,
+      "step": 11010
+    },
+    {
+      "epoch": 2.0533842642194995,
+      "grad_norm": 0.06024446710944176,
+      "learning_rate": 6.308858243260033e-05,
+      "loss": 0.0078,
+      "step": 11020
+    },
+    {
+      "epoch": 2.055247589323147,
+      "grad_norm": 0.11617977172136307,
+      "learning_rate": 6.296434339669525e-05,
+      "loss": 0.0065,
+      "step": 11030
+    },
+    {
+      "epoch": 2.0571109144267945,
+      "grad_norm": 0.07751832902431488,
+      "learning_rate": 6.284010436079017e-05,
+      "loss": 0.0064,
+      "step": 11040
+    },
+    {
+      "epoch": 2.0589742395304422,
+      "grad_norm": 0.07423734664916992,
+      "learning_rate": 6.271586532488508e-05,
+      "loss": 0.0053,
+      "step": 11050
+    },
+    {
+      "epoch": 2.0608375646340895,
+      "grad_norm": 0.11013174802064896,
+      "learning_rate": 6.259162628898e-05,
+      "loss": 0.0081,
+      "step": 11060
+    },
+    {
+      "epoch": 2.062700889737737,
+      "grad_norm": 0.03833114728331566,
+      "learning_rate": 6.246738725307491e-05,
+      "loss": 0.0064,
+      "step": 11070
+    },
+    {
+      "epoch": 2.0645642148413845,
+      "grad_norm": 0.13632163405418396,
+      "learning_rate": 6.234314821716983e-05,
+      "loss": 0.0067,
+      "step": 11080
+    },
+    {
+      "epoch": 2.066427539945032,
+      "grad_norm": 0.05485612899065018,
+      "learning_rate": 6.221890918126476e-05,
+      "loss": 0.0069,
+      "step": 11090
+    },
+    {
+      "epoch": 2.0682908650486795,
+      "grad_norm": 0.1530752331018448,
+      "learning_rate": 6.209467014535967e-05,
+      "loss": 0.0052,
+      "step": 11100
+    },
+    {
+      "epoch": 2.070154190152327,
+      "grad_norm": 0.06478920578956604,
+      "learning_rate": 6.197043110945459e-05,
+      "loss": 0.0069,
+      "step": 11110
+    },
+    {
+      "epoch": 2.072017515255974,
+      "grad_norm": 0.050055887550115585,
+      "learning_rate": 6.184619207354952e-05,
+      "loss": 0.0072,
+      "step": 11120
+    },
+    {
+      "epoch": 2.073880840359622,
+      "grad_norm": 0.06400181353092194,
+      "learning_rate": 6.172195303764444e-05,
+      "loss": 0.0111,
+      "step": 11130
+    },
+    {
+      "epoch": 2.075744165463269,
+      "grad_norm": 0.07093937695026398,
+      "learning_rate": 6.159771400173935e-05,
+      "loss": 0.0073,
+      "step": 11140
+    },
+    {
+      "epoch": 2.077607490566917,
+      "grad_norm": 0.07676272094249725,
+      "learning_rate": 6.147347496583427e-05,
+      "loss": 0.0062,
+      "step": 11150
+    },
+    {
+      "epoch": 2.079470815670564,
+      "grad_norm": 0.06639111787080765,
+      "learning_rate": 6.134923592992918e-05,
+      "loss": 0.0098,
+      "step": 11160
+    },
+    {
+      "epoch": 2.0813341407742114,
+      "grad_norm": 0.14330090582370758,
+      "learning_rate": 6.12249968940241e-05,
+      "loss": 0.0083,
+      "step": 11170
+    },
+    {
+      "epoch": 2.083197465877859,
+      "grad_norm": 0.05557774752378464,
+      "learning_rate": 6.110075785811903e-05,
+      "loss": 0.006,
+      "step": 11180
+    },
+    {
+      "epoch": 2.0850607909815064,
+      "grad_norm": 0.07704410701990128,
+      "learning_rate": 6.0976518822213944e-05,
+      "loss": 0.0081,
+      "step": 11190
+    },
+    {
+      "epoch": 2.086924116085154,
+      "grad_norm": 0.1603652834892273,
+      "learning_rate": 6.0852279786308867e-05,
+      "loss": 0.0079,
+      "step": 11200
+    },
+    {
+      "epoch": 2.0887874411888014,
+      "grad_norm": 0.04601491987705231,
+      "learning_rate": 6.072804075040378e-05,
+      "loss": 0.0075,
+      "step": 11210
+    },
+    {
+      "epoch": 2.0906507662924487,
+      "grad_norm": 0.13114094734191895,
+      "learning_rate": 6.06038017144987e-05,
+      "loss": 0.0095,
+      "step": 11220
+    },
+    {
+      "epoch": 2.0925140913960965,
+      "grad_norm": 0.04922521114349365,
+      "learning_rate": 6.0479562678593614e-05,
+      "loss": 0.009,
+      "step": 11230
+    },
+    {
+      "epoch": 2.0943774164997437,
+      "grad_norm": 0.06616795063018799,
+      "learning_rate": 6.035532364268853e-05,
+      "loss": 0.0077,
+      "step": 11240
+    },
+    {
+      "epoch": 2.0962407416033915,
+      "grad_norm": 0.06306971609592438,
+      "learning_rate": 6.023108460678345e-05,
+      "loss": 0.0079,
+      "step": 11250
+    },
+    {
+      "epoch": 2.0981040667070388,
+      "grad_norm": 0.04845070093870163,
+      "learning_rate": 6.010684557087837e-05,
+      "loss": 0.0078,
+      "step": 11260
+    },
+    {
+      "epoch": 2.099967391810686,
+      "grad_norm": 0.047751642763614655,
+      "learning_rate": 5.99826065349733e-05,
+      "loss": 0.0083,
+      "step": 11270
+    },
+    {
+      "epoch": 2.1018307169143338,
+      "grad_norm": 0.07718124240636826,
+      "learning_rate": 5.9858367499068213e-05,
+      "loss": 0.0072,
+      "step": 11280
+    },
+    {
+      "epoch": 2.103694042017981,
+      "grad_norm": 0.13909615576267242,
+      "learning_rate": 5.973412846316313e-05,
+      "loss": 0.0064,
+      "step": 11290
+    },
+    {
+      "epoch": 2.1055573671216283,
+      "grad_norm": 0.10713133960962296,
+      "learning_rate": 5.9609889427258045e-05,
+      "loss": 0.0074,
+      "step": 11300
+    },
+    {
+      "epoch": 2.107420692225276,
+      "grad_norm": 0.08075924217700958,
+      "learning_rate": 5.948565039135297e-05,
+      "loss": 0.0075,
+      "step": 11310
+    },
+    {
+      "epoch": 2.1092840173289233,
+      "grad_norm": 0.0693148747086525,
+      "learning_rate": 5.9361411355447883e-05,
+      "loss": 0.0053,
+      "step": 11320
+    },
+    {
+      "epoch": 2.111147342432571,
+      "grad_norm": 0.08559976518154144,
+      "learning_rate": 5.92371723195428e-05,
+      "loss": 0.0088,
+      "step": 11330
+    },
+    {
+      "epoch": 2.1130106675362184,
+      "grad_norm": 0.05133243277668953,
+      "learning_rate": 5.9112933283637715e-05,
+      "loss": 0.0067,
+      "step": 11340
+    },
+    {
+      "epoch": 2.1148739926398656,
+      "grad_norm": 0.08110009133815765,
+      "learning_rate": 5.8988694247732644e-05,
+      "loss": 0.0061,
+      "step": 11350
+    },
+    {
+      "epoch": 2.1167373177435134,
+      "grad_norm": 0.07096893340349197,
+      "learning_rate": 5.886445521182756e-05,
+      "loss": 0.0063,
+      "step": 11360
+    },
+    {
+      "epoch": 2.1186006428471607,
+      "grad_norm": 0.06264513731002808,
+      "learning_rate": 5.874021617592248e-05,
+      "loss": 0.0057,
+      "step": 11370
+    },
+    {
+      "epoch": 2.1204639679508084,
+      "grad_norm": 0.11413908004760742,
+      "learning_rate": 5.86159771400174e-05,
+      "loss": 0.007,
+      "step": 11380
+    },
+    {
+      "epoch": 2.1223272930544557,
+      "grad_norm": 0.07597790658473969,
+      "learning_rate": 5.8491738104112314e-05,
+      "loss": 0.0068,
+      "step": 11390
+    },
+    {
+      "epoch": 2.124190618158103,
+      "grad_norm": 0.03327861428260803,
+      "learning_rate": 5.836749906820723e-05,
+      "loss": 0.0057,
+      "step": 11400
+    },
+    {
+      "epoch": 2.1260539432617507,
+      "grad_norm": 0.10454593598842621,
+      "learning_rate": 5.8243260032302146e-05,
+      "loss": 0.0105,
+      "step": 11410
+    },
+    {
+      "epoch": 2.127917268365398,
+      "grad_norm": 0.0536537803709507,
+      "learning_rate": 5.811902099639707e-05,
+      "loss": 0.007,
+      "step": 11420
+    },
+    {
+      "epoch": 2.1297805934690457,
+      "grad_norm": 0.07390698790550232,
+      "learning_rate": 5.7994781960491984e-05,
+      "loss": 0.0063,
+      "step": 11430
+    },
+    {
+      "epoch": 2.131643918572693,
+      "grad_norm": 0.060124851763248444,
+      "learning_rate": 5.7870542924586914e-05,
+      "loss": 0.0054,
+      "step": 11440
+    },
+    {
+      "epoch": 2.1335072436763403,
+      "grad_norm": 0.06745817512273788,
+      "learning_rate": 5.774630388868183e-05,
+      "loss": 0.0075,
+      "step": 11450
+    },
+    {
+      "epoch": 2.135370568779988,
+      "grad_norm": 0.04149980843067169,
+      "learning_rate": 5.7622064852776745e-05,
+      "loss": 0.0082,
+      "step": 11460
+    },
+    {
+      "epoch": 2.1372338938836353,
+      "grad_norm": 0.08865126222372055,
+      "learning_rate": 5.749782581687166e-05,
+      "loss": 0.005,
+      "step": 11470
+    },
+    {
+      "epoch": 2.139097218987283,
+      "grad_norm": 0.0498250387609005,
+      "learning_rate": 5.7373586780966584e-05,
+      "loss": 0.0061,
+      "step": 11480
+    },
+    {
+      "epoch": 2.1409605440909303,
+      "grad_norm": 0.051534079015254974,
+      "learning_rate": 5.72493477450615e-05,
+      "loss": 0.0083,
+      "step": 11490
+    },
+    {
+      "epoch": 2.1428238691945776,
+      "grad_norm": 0.07920075207948685,
+      "learning_rate": 5.7125108709156415e-05,
+      "loss": 0.0125,
+      "step": 11500
+    },
+    {
+      "epoch": 2.1446871942982253,
+      "grad_norm": 0.0594758465886116,
+      "learning_rate": 5.700086967325133e-05,
+      "loss": 0.007,
+      "step": 11510
+    },
+    {
+      "epoch": 2.1465505194018726,
+      "grad_norm": 0.0814061388373375,
+      "learning_rate": 5.687663063734626e-05,
+      "loss": 0.0073,
+      "step": 11520
+    },
+    {
+      "epoch": 2.1484138445055203,
+      "grad_norm": 0.061178576201200485,
+      "learning_rate": 5.6752391601441176e-05,
+      "loss": 0.0059,
+      "step": 11530
+    },
+    {
+      "epoch": 2.1502771696091676,
+      "grad_norm": 0.046416446566581726,
+      "learning_rate": 5.66281525655361e-05,
+      "loss": 0.0072,
+      "step": 11540
+    },
+    {
+      "epoch": 2.152140494712815,
+      "grad_norm": 0.0688127651810646,
+      "learning_rate": 5.6503913529631015e-05,
+      "loss": 0.0069,
+      "step": 11550
+    },
+    {
+      "epoch": 2.1540038198164626,
+      "grad_norm": 0.15987776219844818,
+      "learning_rate": 5.637967449372593e-05,
+      "loss": 0.0076,
+      "step": 11560
+    },
+    {
+      "epoch": 2.15586714492011,
+      "grad_norm": 0.06576576083898544,
+      "learning_rate": 5.6255435457820846e-05,
+      "loss": 0.0082,
+      "step": 11570
+    },
+    {
+      "epoch": 2.1577304700237576,
+      "grad_norm": 0.09326786547899246,
+      "learning_rate": 5.613119642191577e-05,
+      "loss": 0.0047,
+      "step": 11580
+    },
+    {
+      "epoch": 2.159593795127405,
+      "grad_norm": 0.05868074297904968,
+      "learning_rate": 5.6006957386010685e-05,
+      "loss": 0.0077,
+      "step": 11590
+    },
+    {
+      "epoch": 2.161457120231052,
+      "grad_norm": 0.03512386605143547,
+      "learning_rate": 5.58827183501056e-05,
+      "loss": 0.0074,
+      "step": 11600
+    },
+    {
+      "epoch": 2.1633204453347,
+      "grad_norm": 0.09879045188426971,
+      "learning_rate": 5.575847931420053e-05,
+      "loss": 0.0074,
+      "step": 11610
+    },
+    {
+      "epoch": 2.165183770438347,
+      "grad_norm": 0.08072607964277267,
+      "learning_rate": 5.5634240278295446e-05,
+      "loss": 0.0068,
+      "step": 11620
+    },
+    {
+      "epoch": 2.1670470955419945,
+      "grad_norm": 0.06294963508844376,
+      "learning_rate": 5.551000124239036e-05,
+      "loss": 0.0054,
+      "step": 11630
+    },
+    {
+      "epoch": 2.168910420645642,
+      "grad_norm": 0.13824103772640228,
+      "learning_rate": 5.5385762206485284e-05,
+      "loss": 0.0074,
+      "step": 11640
+    },
+    {
+      "epoch": 2.1707737457492895,
+      "grad_norm": 0.09912306070327759,
+      "learning_rate": 5.52615231705802e-05,
+      "loss": 0.0094,
+      "step": 11650
+    },
+    {
+      "epoch": 2.1726370708529372,
+      "grad_norm": 0.1159447655081749,
+      "learning_rate": 5.5137284134675116e-05,
+      "loss": 0.0061,
+      "step": 11660
+    },
+    {
+      "epoch": 2.1745003959565845,
+      "grad_norm": 0.11456414312124252,
+      "learning_rate": 5.501304509877003e-05,
+      "loss": 0.0071,
+      "step": 11670
+    },
+    {
+      "epoch": 2.176363721060232,
+      "grad_norm": 0.06627790629863739,
+      "learning_rate": 5.488880606286495e-05,
+      "loss": 0.0057,
+      "step": 11680
+    },
+    {
+      "epoch": 2.1782270461638795,
+      "grad_norm": 0.09946975111961365,
+      "learning_rate": 5.476456702695988e-05,
+      "loss": 0.0094,
+      "step": 11690
+    },
+    {
+      "epoch": 2.180090371267527,
+      "grad_norm": 0.02680755965411663,
+      "learning_rate": 5.46403279910548e-05,
+      "loss": 0.0068,
+      "step": 11700
+    },
+    {
+      "epoch": 2.1819536963711745,
+      "grad_norm": 0.10289502143859863,
+      "learning_rate": 5.4516088955149715e-05,
+      "loss": 0.006,
+      "step": 11710
+    },
+    {
+      "epoch": 2.183817021474822,
+      "grad_norm": 0.0671938881278038,
+      "learning_rate": 5.439184991924463e-05,
+      "loss": 0.0066,
+      "step": 11720
+    },
+    {
+      "epoch": 2.185680346578469,
+      "grad_norm": 0.059153467416763306,
+      "learning_rate": 5.426761088333955e-05,
+      "loss": 0.0062,
+      "step": 11730
+    },
+    {
+      "epoch": 2.187543671682117,
+      "grad_norm": 0.06136234104633331,
+      "learning_rate": 5.414337184743446e-05,
+      "loss": 0.009,
+      "step": 11740
+    },
+    {
+      "epoch": 2.189406996785764,
+      "grad_norm": 0.08328571915626526,
+      "learning_rate": 5.4019132811529385e-05,
+      "loss": 0.0072,
+      "step": 11750
+    },
+    {
+      "epoch": 2.191270321889412,
+      "grad_norm": 0.08631386607885361,
+      "learning_rate": 5.38948937756243e-05,
+      "loss": 0.0064,
+      "step": 11760
+    },
+    {
+      "epoch": 2.193133646993059,
+      "grad_norm": 0.11763385683298111,
+      "learning_rate": 5.377065473971922e-05,
+      "loss": 0.0066,
+      "step": 11770
+    },
+    {
+      "epoch": 2.1949969720967064,
+      "grad_norm": 0.0526675283908844,
+      "learning_rate": 5.3646415703814146e-05,
+      "loss": 0.0042,
+      "step": 11780
+    },
+    {
+      "epoch": 2.196860297200354,
+      "grad_norm": 0.06366570293903351,
+      "learning_rate": 5.352217666790906e-05,
+      "loss": 0.0083,
+      "step": 11790
+    },
+    {
+      "epoch": 2.1987236223040014,
+      "grad_norm": 0.05300299450755119,
+      "learning_rate": 5.339793763200398e-05,
+      "loss": 0.0058,
+      "step": 11800
+    },
+    {
+      "epoch": 2.200586947407649,
+      "grad_norm": 0.04540044814348221,
+      "learning_rate": 5.32736985960989e-05,
+      "loss": 0.0057,
+      "step": 11810
+    },
+    {
+      "epoch": 2.2024502725112964,
+      "grad_norm": 0.03275251388549805,
+      "learning_rate": 5.3149459560193816e-05,
+      "loss": 0.0075,
+      "step": 11820
+    },
+    {
+      "epoch": 2.2043135976149437,
+      "grad_norm": 0.03645419701933861,
+      "learning_rate": 5.302522052428873e-05,
+      "loss": 0.0049,
+      "step": 11830
+    },
+    {
+      "epoch": 2.2061769227185914,
+      "grad_norm": 0.09000103920698166,
+      "learning_rate": 5.290098148838365e-05,
+      "loss": 0.0075,
+      "step": 11840
+    },
+    {
+      "epoch": 2.2080402478222387,
+      "grad_norm": 0.09125344455242157,
+      "learning_rate": 5.2776742452478564e-05,
+      "loss": 0.0068,
+      "step": 11850
+    },
+    {
+      "epoch": 2.209903572925886,
+      "grad_norm": 0.06729969382286072,
+      "learning_rate": 5.265250341657349e-05,
+      "loss": 0.0049,
+      "step": 11860
+    },
+    {
+      "epoch": 2.2117668980295337,
+      "grad_norm": 0.0643710121512413,
+      "learning_rate": 5.2528264380668416e-05,
+      "loss": 0.0069,
+      "step": 11870
+    },
+    {
+      "epoch": 2.213630223133181,
+      "grad_norm": 0.0587821826338768,
+      "learning_rate": 5.240402534476333e-05,
+      "loss": 0.0077,
+      "step": 11880
+    },
+    {
+      "epoch": 2.2154935482368288,
+      "grad_norm": 0.04202594608068466,
+      "learning_rate": 5.227978630885825e-05,
+      "loss": 0.0057,
+      "step": 11890
+    },
+    {
+      "epoch": 2.217356873340476,
+      "grad_norm": 0.05304161086678505,
+      "learning_rate": 5.215554727295316e-05,
+      "loss": 0.0054,
+      "step": 11900
+    },
+    {
+      "epoch": 2.2192201984441233,
+      "grad_norm": 0.05053735896945,
+      "learning_rate": 5.203130823704808e-05,
+      "loss": 0.0067,
+      "step": 11910
+    },
+    {
+      "epoch": 2.221083523547771,
+      "grad_norm": 0.05211859941482544,
+      "learning_rate": 5.1907069201143e-05,
+      "loss": 0.0084,
+      "step": 11920
+    },
+    {
+      "epoch": 2.2229468486514183,
+      "grad_norm": 0.09272083640098572,
+      "learning_rate": 5.178283016523792e-05,
+      "loss": 0.0062,
+      "step": 11930
+    },
+    {
+      "epoch": 2.224810173755066,
+      "grad_norm": 0.05025614798069,
+      "learning_rate": 5.165859112933283e-05,
+      "loss": 0.0075,
+      "step": 11940
+    },
+    {
+      "epoch": 2.2266734988587134,
+      "grad_norm": 0.11683744937181473,
+      "learning_rate": 5.153435209342776e-05,
+      "loss": 0.0072,
+      "step": 11950
+    },
+    {
+      "epoch": 2.2285368239623606,
+      "grad_norm": 0.051841359585523605,
+      "learning_rate": 5.141011305752268e-05,
+      "loss": 0.0101,
+      "step": 11960
+    },
+    {
+      "epoch": 2.2304001490660084,
+      "grad_norm": 0.04127165675163269,
+      "learning_rate": 5.1285874021617594e-05,
+      "loss": 0.006,
+      "step": 11970
+    },
+    {
+      "epoch": 2.2322634741696556,
+      "grad_norm": 0.07321721315383911,
+      "learning_rate": 5.1161634985712517e-05,
+      "loss": 0.0062,
+      "step": 11980
+    },
+    {
+      "epoch": 2.2341267992733034,
+      "grad_norm": 0.10085262358188629,
+      "learning_rate": 5.103739594980743e-05,
+      "loss": 0.0068,
+      "step": 11990
+    },
+    {
+      "epoch": 2.2359901243769507,
+      "grad_norm": 0.07528786361217499,
+      "learning_rate": 5.091315691390235e-05,
+      "loss": 0.0082,
+      "step": 12000
+    },
+    {
+      "epoch": 2.237853449480598,
+      "grad_norm": 0.08790519088506699,
+      "learning_rate": 5.0788917877997264e-05,
+      "loss": 0.0066,
+      "step": 12010
+    },
+    {
+      "epoch": 2.2397167745842457,
+      "grad_norm": 0.0749954879283905,
+      "learning_rate": 5.0664678842092187e-05,
+      "loss": 0.0067,
+      "step": 12020
+    },
+    {
+      "epoch": 2.241580099687893,
+      "grad_norm": 0.04362763091921806,
+      "learning_rate": 5.05404398061871e-05,
+      "loss": 0.0052,
+      "step": 12030
+    },
+    {
+      "epoch": 2.2434434247915407,
+      "grad_norm": 0.09141118079423904,
+      "learning_rate": 5.041620077028203e-05,
+      "loss": 0.0061,
+      "step": 12040
+    },
+    {
+      "epoch": 2.245306749895188,
+      "grad_norm": 0.05899785831570625,
+      "learning_rate": 5.029196173437695e-05,
+      "loss": 0.0077,
+      "step": 12050
+    },
+    {
+      "epoch": 2.2471700749988353,
+      "grad_norm": 0.06144483759999275,
+      "learning_rate": 5.016772269847186e-05,
+      "loss": 0.0058,
+      "step": 12060
+    },
+    {
+      "epoch": 2.249033400102483,
+      "grad_norm": 0.06769667565822601,
+      "learning_rate": 5.004348366256678e-05,
+      "loss": 0.0062,
+      "step": 12070
+    },
+    {
+      "epoch": 2.2508967252061303,
+      "grad_norm": 0.05238569527864456,
+      "learning_rate": 4.99192446266617e-05,
+      "loss": 0.0081,
+      "step": 12080
+    },
+    {
+      "epoch": 2.252760050309778,
+      "grad_norm": 0.09564723819494247,
+      "learning_rate": 4.979500559075662e-05,
+      "loss": 0.0063,
+      "step": 12090
+    },
+    {
+      "epoch": 2.2546233754134253,
+      "grad_norm": 0.07752995193004608,
+      "learning_rate": 4.967076655485154e-05,
+      "loss": 0.0056,
+      "step": 12100
+    },
+    {
+      "epoch": 2.2564867005170726,
+      "grad_norm": 0.12467604875564575,
+      "learning_rate": 4.9546527518946456e-05,
+      "loss": 0.0074,
+      "step": 12110
+    },
+    {
+      "epoch": 2.2583500256207203,
+      "grad_norm": 0.03957875445485115,
+      "learning_rate": 4.942228848304137e-05,
+      "loss": 0.0067,
+      "step": 12120
+    },
+    {
+      "epoch": 2.2602133507243676,
+      "grad_norm": 0.09976095706224442,
+      "learning_rate": 4.929804944713629e-05,
+      "loss": 0.0076,
+      "step": 12130
+    },
+    {
+      "epoch": 2.2620766758280153,
+      "grad_norm": 0.09455903619527817,
+      "learning_rate": 4.917381041123122e-05,
+      "loss": 0.0071,
+      "step": 12140
+    },
+    {
+      "epoch": 2.2639400009316626,
+      "grad_norm": 0.15039753913879395,
+      "learning_rate": 4.904957137532613e-05,
+      "loss": 0.0078,
+      "step": 12150
+    },
+    {
+      "epoch": 2.26580332603531,
+      "grad_norm": 0.08308923989534378,
+      "learning_rate": 4.892533233942105e-05,
+      "loss": 0.0073,
+      "step": 12160
+    },
+    {
+      "epoch": 2.2676666511389576,
+      "grad_norm": 0.13981153070926666,
+      "learning_rate": 4.8801093303515964e-05,
+      "loss": 0.0102,
+      "step": 12170
+    },
+    {
+      "epoch": 2.269529976242605,
+      "grad_norm": 0.07559193670749664,
+      "learning_rate": 4.867685426761088e-05,
+      "loss": 0.0068,
+      "step": 12180
+    },
+    {
+      "epoch": 2.2713933013462526,
+      "grad_norm": 0.06467331200838089,
+      "learning_rate": 4.85526152317058e-05,
+      "loss": 0.0053,
+      "step": 12190
+    },
+    {
+      "epoch": 2.2732566264499,
+      "grad_norm": 0.04058116301894188,
+      "learning_rate": 4.8428376195800725e-05,
+      "loss": 0.007,
+      "step": 12200
+    },
+    {
+      "epoch": 2.275119951553547,
+      "grad_norm": 0.04690202698111534,
+      "learning_rate": 4.830413715989564e-05,
+      "loss": 0.0062,
+      "step": 12210
+    },
+    {
+      "epoch": 2.276983276657195,
+      "grad_norm": 0.03539729863405228,
+      "learning_rate": 4.817989812399056e-05,
+      "loss": 0.0078,
+      "step": 12220
+    },
+    {
+      "epoch": 2.278846601760842,
+      "grad_norm": 0.0975772812962532,
+      "learning_rate": 4.805565908808548e-05,
+      "loss": 0.007,
+      "step": 12230
+    },
+    {
+      "epoch": 2.28070992686449,
+      "grad_norm": 0.057454656809568405,
+      "learning_rate": 4.7931420052180395e-05,
+      "loss": 0.008,
+      "step": 12240
+    },
+    {
+      "epoch": 2.282573251968137,
+      "grad_norm": 0.0594792366027832,
+      "learning_rate": 4.780718101627532e-05,
+      "loss": 0.0066,
+      "step": 12250
+    },
+    {
+      "epoch": 2.2844365770717845,
+      "grad_norm": 0.07810505479574203,
+      "learning_rate": 4.7682941980370234e-05,
+      "loss": 0.0052,
+      "step": 12260
+    },
+    {
+      "epoch": 2.286299902175432,
+      "grad_norm": 0.1288454383611679,
+      "learning_rate": 4.7558702944465156e-05,
+      "loss": 0.0078,
+      "step": 12270
+    },
+    {
+      "epoch": 2.2881632272790795,
+      "grad_norm": 0.05404103919863701,
+      "learning_rate": 4.743446390856007e-05,
+      "loss": 0.0056,
+      "step": 12280
+    },
+    {
+      "epoch": 2.290026552382727,
+      "grad_norm": 0.09221246093511581,
+      "learning_rate": 4.731022487265499e-05,
+      "loss": 0.0034,
+      "step": 12290
+    },
+    {
+      "epoch": 2.2918898774863745,
+      "grad_norm": 0.06454599648714066,
+      "learning_rate": 4.718598583674991e-05,
+      "loss": 0.0066,
+      "step": 12300
+    },
+    {
+      "epoch": 2.293753202590022,
+      "grad_norm": 0.23072722554206848,
+      "learning_rate": 4.706174680084483e-05,
+      "loss": 0.0081,
+      "step": 12310
+    },
+    {
+      "epoch": 2.2956165276936695,
+      "grad_norm": 0.03680235520005226,
+      "learning_rate": 4.693750776493975e-05,
+      "loss": 0.0052,
+      "step": 12320
+    },
+    {
+      "epoch": 2.297479852797317,
+      "grad_norm": 0.05998710170388222,
+      "learning_rate": 4.6813268729034665e-05,
+      "loss": 0.0075,
+      "step": 12330
+    },
+    {
+      "epoch": 2.299343177900964,
+      "grad_norm": 0.15091590583324432,
+      "learning_rate": 4.668902969312958e-05,
+      "loss": 0.0079,
+      "step": 12340
+    },
+    {
+      "epoch": 2.301206503004612,
+      "grad_norm": 0.10594821721315384,
+      "learning_rate": 4.6564790657224496e-05,
+      "loss": 0.007,
+      "step": 12350
+    },
+    {
+      "epoch": 2.303069828108259,
+      "grad_norm": 0.0922156348824501,
+      "learning_rate": 4.6440551621319426e-05,
+      "loss": 0.0054,
+      "step": 12360
+    },
+    {
+      "epoch": 2.304933153211907,
+      "grad_norm": 0.08536434918642044,
+      "learning_rate": 4.631631258541434e-05,
+      "loss": 0.008,
+      "step": 12370
+    },
+    {
+      "epoch": 2.306796478315554,
+      "grad_norm": 0.05940347537398338,
+      "learning_rate": 4.619207354950926e-05,
+      "loss": 0.0062,
+      "step": 12380
+    },
+    {
+      "epoch": 2.3086598034192014,
+      "grad_norm": 0.0685829296708107,
+      "learning_rate": 4.606783451360417e-05,
+      "loss": 0.0046,
+      "step": 12390
+    },
+    {
+      "epoch": 2.310523128522849,
+      "grad_norm": 0.0558197982609272,
+      "learning_rate": 4.5943595477699096e-05,
+      "loss": 0.0073,
+      "step": 12400
+    },
+    {
+      "epoch": 2.3123864536264964,
+      "grad_norm": 0.069609135389328,
+      "learning_rate": 4.581935644179401e-05,
+      "loss": 0.0067,
+      "step": 12410
+    },
+    {
+      "epoch": 2.3142497787301437,
+      "grad_norm": 0.037157054990530014,
+      "learning_rate": 4.5695117405888934e-05,
+      "loss": 0.0111,
+      "step": 12420
+    },
+    {
+      "epoch": 2.3161131038337914,
+      "grad_norm": 0.055848293006420135,
+      "learning_rate": 4.557087836998385e-05,
+      "loss": 0.0059,
+      "step": 12430
+    },
+    {
+      "epoch": 2.3179764289374387,
+      "grad_norm": 0.09797196090221405,
+      "learning_rate": 4.544663933407877e-05,
+      "loss": 0.007,
+      "step": 12440
+    },
+    {
+      "epoch": 2.3198397540410864,
+      "grad_norm": 0.045269809663295746,
+      "learning_rate": 4.532240029817369e-05,
+      "loss": 0.0052,
+      "step": 12450
+    },
+    {
+      "epoch": 2.3217030791447337,
+      "grad_norm": 0.04363720491528511,
+      "learning_rate": 4.5198161262268604e-05,
+      "loss": 0.0077,
+      "step": 12460
+    },
+    {
+      "epoch": 2.323566404248381,
+      "grad_norm": 0.06502070277929306,
+      "learning_rate": 4.507392222636353e-05,
+      "loss": 0.0071,
+      "step": 12470
+    },
+    {
+      "epoch": 2.3254297293520287,
+      "grad_norm": 0.043661072850227356,
+      "learning_rate": 4.494968319045844e-05,
+      "loss": 0.0068,
+      "step": 12480
+    },
+    {
+      "epoch": 2.327293054455676,
+      "grad_norm": 0.06712755560874939,
+      "learning_rate": 4.4825444154553365e-05,
+      "loss": 0.0049,
+      "step": 12490
+    },
+    {
+      "epoch": 2.3291563795593238,
+      "grad_norm": 0.06757821142673492,
+      "learning_rate": 4.470120511864828e-05,
+      "loss": 0.0056,
+      "step": 12500
+    },
+    {
+      "epoch": 2.331019704662971,
+      "grad_norm": 0.09899108856916428,
+      "learning_rate": 4.45769660827432e-05,
+      "loss": 0.0059,
+      "step": 12510
+    },
+    {
+      "epoch": 2.3328830297666183,
+      "grad_norm": 0.0666062980890274,
+      "learning_rate": 4.445272704683812e-05,
+      "loss": 0.0048,
+      "step": 12520
+    },
+    {
+      "epoch": 2.334746354870266,
+      "grad_norm": 0.0807105004787445,
+      "learning_rate": 4.432848801093304e-05,
+      "loss": 0.0085,
+      "step": 12530
+    },
+    {
+      "epoch": 2.3366096799739133,
+      "grad_norm": 0.0663641020655632,
+      "learning_rate": 4.420424897502796e-05,
+      "loss": 0.0053,
+      "step": 12540
+    },
+    {
+      "epoch": 2.338473005077561,
+      "grad_norm": 0.07503866404294968,
+      "learning_rate": 4.4080009939122874e-05,
+      "loss": 0.0049,
+      "step": 12550
+    },
+    {
+      "epoch": 2.3403363301812083,
+      "grad_norm": 0.03897761553525925,
+      "learning_rate": 4.395577090321779e-05,
+      "loss": 0.0055,
+      "step": 12560
+    },
+    {
+      "epoch": 2.3421996552848556,
+      "grad_norm": 0.11116714030504227,
+      "learning_rate": 4.383153186731271e-05,
+      "loss": 0.008,
+      "step": 12570
+    },
+    {
+      "epoch": 2.3440629803885034,
+      "grad_norm": 0.052072007209062576,
+      "learning_rate": 4.3707292831407634e-05,
+      "loss": 0.0079,
+      "step": 12580
+    },
+    {
+      "epoch": 2.3459263054921506,
+      "grad_norm": 0.06437745690345764,
+      "learning_rate": 4.358305379550255e-05,
+      "loss": 0.0065,
+      "step": 12590
+    },
+    {
+      "epoch": 2.3477896305957984,
+      "grad_norm": 0.028404507786035538,
+      "learning_rate": 4.3458814759597466e-05,
+      "loss": 0.0052,
+      "step": 12600
+    },
+    {
+      "epoch": 2.3496529556994457,
+      "grad_norm": 0.10816732794046402,
+      "learning_rate": 4.333457572369239e-05,
+      "loss": 0.0111,
+      "step": 12610
+    },
+    {
+      "epoch": 2.351516280803093,
+      "grad_norm": 0.11128159612417221,
+      "learning_rate": 4.3210336687787305e-05,
+      "loss": 0.0091,
+      "step": 12620
+    },
+    {
+      "epoch": 2.3533796059067407,
+      "grad_norm": 0.046126995235681534,
+      "learning_rate": 4.308609765188222e-05,
+      "loss": 0.007,
+      "step": 12630
+    },
+    {
+      "epoch": 2.355242931010388,
+      "grad_norm": 0.03195841982960701,
+      "learning_rate": 4.296185861597714e-05,
+      "loss": 0.0081,
+      "step": 12640
+    },
+    {
+      "epoch": 2.3571062561140357,
+      "grad_norm": 0.05619229003787041,
+      "learning_rate": 4.283761958007206e-05,
+      "loss": 0.0052,
+      "step": 12650
+    },
+    {
+      "epoch": 2.358969581217683,
+      "grad_norm": 0.12804652750492096,
+      "learning_rate": 4.271338054416698e-05,
+      "loss": 0.0077,
+      "step": 12660
+    },
+    {
+      "epoch": 2.3608329063213302,
+      "grad_norm": 0.08166605234146118,
+      "learning_rate": 4.25891415082619e-05,
+      "loss": 0.0075,
+      "step": 12670
+    },
+    {
+      "epoch": 2.362696231424978,
+      "grad_norm": 0.06898462772369385,
+      "learning_rate": 4.246490247235681e-05,
+      "loss": 0.0055,
+      "step": 12680
+    },
+    {
+      "epoch": 2.3645595565286253,
+      "grad_norm": 0.12175576388835907,
+      "learning_rate": 4.2340663436451735e-05,
+      "loss": 0.0053,
+      "step": 12690
+    },
+    {
+      "epoch": 2.366422881632273,
+      "grad_norm": 0.06706222146749496,
+      "learning_rate": 4.221642440054666e-05,
+      "loss": 0.007,
+      "step": 12700
+    },
+    {
+      "epoch": 2.3682862067359203,
+      "grad_norm": 0.05738810822367668,
+      "learning_rate": 4.2092185364641574e-05,
+      "loss": 0.0054,
+      "step": 12710
+    },
+    {
+      "epoch": 2.3701495318395676,
+      "grad_norm": 0.0824676901102066,
+      "learning_rate": 4.196794632873649e-05,
+      "loss": 0.005,
+      "step": 12720
+    },
+    {
+      "epoch": 2.3720128569432153,
+      "grad_norm": 0.10526188462972641,
+      "learning_rate": 4.1843707292831406e-05,
+      "loss": 0.0071,
+      "step": 12730
+    },
+    {
+      "epoch": 2.3738761820468626,
+      "grad_norm": 0.0358128622174263,
+      "learning_rate": 4.171946825692633e-05,
+      "loss": 0.0118,
+      "step": 12740
+    },
+    {
+      "epoch": 2.3757395071505103,
+      "grad_norm": 0.07586889714002609,
+      "learning_rate": 4.159522922102125e-05,
+      "loss": 0.0062,
+      "step": 12750
+    },
+    {
+      "epoch": 2.3776028322541576,
+      "grad_norm": 0.04203260689973831,
+      "learning_rate": 4.1470990185116166e-05,
+      "loss": 0.0061,
+      "step": 12760
+    },
+    {
+      "epoch": 2.379466157357805,
+      "grad_norm": 0.061605483293533325,
+      "learning_rate": 4.134675114921108e-05,
+      "loss": 0.005,
+      "step": 12770
+    },
+    {
+      "epoch": 2.3813294824614526,
+      "grad_norm": 0.0388299860060215,
+      "learning_rate": 4.1222512113306005e-05,
+      "loss": 0.0057,
+      "step": 12780
+    },
+    {
+      "epoch": 2.3831928075651,
+      "grad_norm": 0.07655610889196396,
+      "learning_rate": 4.109827307740092e-05,
+      "loss": 0.0063,
+      "step": 12790
+    },
+    {
+      "epoch": 2.3850561326687476,
+      "grad_norm": 0.05981520935893059,
+      "learning_rate": 4.097403404149584e-05,
+      "loss": 0.0056,
+      "step": 12800
+    },
+    {
+      "epoch": 2.386919457772395,
+      "grad_norm": 0.07374788820743561,
+      "learning_rate": 4.084979500559076e-05,
+      "loss": 0.0055,
+      "step": 12810
+    },
+    {
+      "epoch": 2.388782782876042,
+      "grad_norm": 0.07417966425418854,
+      "learning_rate": 4.0725555969685675e-05,
+      "loss": 0.0048,
+      "step": 12820
+    },
+    {
+      "epoch": 2.39064610797969,
+      "grad_norm": 0.13554275035858154,
+      "learning_rate": 4.06013169337806e-05,
+      "loss": 0.0063,
+      "step": 12830
+    },
+    {
+      "epoch": 2.392509433083337,
+      "grad_norm": 0.2652145028114319,
+      "learning_rate": 4.047707789787551e-05,
+      "loss": 0.0091,
+      "step": 12840
+    },
+    {
+      "epoch": 2.394372758186985,
+      "grad_norm": 0.055324655026197433,
+      "learning_rate": 4.035283886197043e-05,
+      "loss": 0.0052,
+      "step": 12850
+    },
+    {
+      "epoch": 2.396236083290632,
+      "grad_norm": 0.07505489140748978,
+      "learning_rate": 4.022859982606535e-05,
+      "loss": 0.0063,
+      "step": 12860
+    },
+    {
+      "epoch": 2.3980994083942795,
+      "grad_norm": 0.08703186362981796,
+      "learning_rate": 4.0104360790160274e-05,
+      "loss": 0.004,
+      "step": 12870
+    },
+    {
+      "epoch": 2.399962733497927,
+      "grad_norm": 0.09088319540023804,
+      "learning_rate": 3.998012175425519e-05,
+      "loss": 0.0062,
+      "step": 12880
+    },
+    {
+      "epoch": 2.4018260586015745,
+      "grad_norm": 0.059139229357242584,
+      "learning_rate": 3.9855882718350106e-05,
+      "loss": 0.0077,
+      "step": 12890
+    },
+    {
+      "epoch": 2.403689383705222,
+      "grad_norm": 0.04634464904665947,
+      "learning_rate": 3.973164368244502e-05,
+      "loss": 0.0051,
+      "step": 12900
+    },
+    {
+      "epoch": 2.4055527088088695,
+      "grad_norm": 0.09583749622106552,
+      "learning_rate": 3.9607404646539944e-05,
+      "loss": 0.0213,
+      "step": 12910
+    },
+    {
+      "epoch": 2.407416033912517,
+      "grad_norm": 0.09496133774518967,
+      "learning_rate": 3.948316561063487e-05,
+      "loss": 0.0065,
+      "step": 12920
+    },
+    {
+      "epoch": 2.4092793590161645,
+      "grad_norm": 0.04073603078722954,
+      "learning_rate": 3.935892657472978e-05,
+      "loss": 0.006,
+      "step": 12930
+    },
+    {
+      "epoch": 2.411142684119812,
+      "grad_norm": 0.047502193599939346,
+      "learning_rate": 3.92346875388247e-05,
+      "loss": 0.0066,
+      "step": 12940
+    },
+    {
+      "epoch": 2.413006009223459,
+      "grad_norm": 0.09126313030719757,
+      "learning_rate": 3.9110448502919614e-05,
+      "loss": 0.007,
+      "step": 12950
+    },
+    {
+      "epoch": 2.414869334327107,
+      "grad_norm": 0.05916162207722664,
+      "learning_rate": 3.898620946701454e-05,
+      "loss": 0.0057,
+      "step": 12960
+    },
+    {
+      "epoch": 2.416732659430754,
+      "grad_norm": 0.285349577665329,
+      "learning_rate": 3.886197043110946e-05,
+      "loss": 0.0065,
+      "step": 12970
+    },
+    {
+      "epoch": 2.418595984534402,
+      "grad_norm": 0.03480862081050873,
+      "learning_rate": 3.8737731395204375e-05,
+      "loss": 0.0045,
+      "step": 12980
+    },
+    {
+      "epoch": 2.420459309638049,
+      "grad_norm": 0.07367054373025894,
+      "learning_rate": 3.861349235929929e-05,
+      "loss": 0.006,
+      "step": 12990
+    },
+    {
+      "epoch": 2.4223226347416964,
+      "grad_norm": 0.17037363350391388,
+      "learning_rate": 3.8489253323394214e-05,
+      "loss": 0.0052,
+      "step": 13000
+    },
+    {
+      "epoch": 2.424185959845344,
+      "grad_norm": 0.10525022447109222,
+      "learning_rate": 3.836501428748913e-05,
+      "loss": 0.0058,
+      "step": 13010
+    },
+    {
+      "epoch": 2.4260492849489914,
+      "grad_norm": 0.08784069120883942,
+      "learning_rate": 3.824077525158405e-05,
+      "loss": 0.0066,
+      "step": 13020
+    },
+    {
+      "epoch": 2.4279126100526387,
+      "grad_norm": 0.05001769959926605,
+      "learning_rate": 3.811653621567897e-05,
+      "loss": 0.0067,
+      "step": 13030
+    },
+    {
+      "epoch": 2.4297759351562864,
+      "grad_norm": 0.043823111802339554,
+      "learning_rate": 3.799229717977389e-05,
+      "loss": 0.0083,
+      "step": 13040
+    },
+    {
+      "epoch": 2.4316392602599337,
+      "grad_norm": 0.05830995738506317,
+      "learning_rate": 3.7868058143868806e-05,
+      "loss": 0.0066,
+      "step": 13050
+    },
+    {
+      "epoch": 2.4335025853635814,
+      "grad_norm": 0.0677856057882309,
+      "learning_rate": 3.774381910796372e-05,
+      "loss": 0.0062,
+      "step": 13060
+    },
+    {
+      "epoch": 2.4353659104672287,
+      "grad_norm": 0.057640671730041504,
+      "learning_rate": 3.761958007205864e-05,
+      "loss": 0.0057,
+      "step": 13070
+    },
+    {
+      "epoch": 2.437229235570876,
+      "grad_norm": 0.06286245584487915,
+      "learning_rate": 3.749534103615357e-05,
+      "loss": 0.0077,
+      "step": 13080
+    },
+    {
+      "epoch": 2.4390925606745237,
+      "grad_norm": 0.06981087476015091,
+      "learning_rate": 3.737110200024848e-05,
+      "loss": 0.0076,
+      "step": 13090
+    },
+    {
+      "epoch": 2.440955885778171,
+      "grad_norm": 0.0729713961482048,
+      "learning_rate": 3.72468629643434e-05,
+      "loss": 0.005,
+      "step": 13100
+    },
+    {
+      "epoch": 2.4428192108818187,
+      "grad_norm": 0.028910622000694275,
+      "learning_rate": 3.7122623928438315e-05,
+      "loss": 0.0037,
+      "step": 13110
+    },
+    {
+      "epoch": 2.444682535985466,
+      "grad_norm": 0.07171040773391724,
+      "learning_rate": 3.699838489253323e-05,
+      "loss": 0.0055,
+      "step": 13120
+    },
+    {
+      "epoch": 2.4465458610891133,
+      "grad_norm": 0.057173121720552444,
+      "learning_rate": 3.687414585662815e-05,
+      "loss": 0.0057,
+      "step": 13130
+    },
+    {
+      "epoch": 2.448409186192761,
+      "grad_norm": 0.07052276283502579,
+      "learning_rate": 3.6749906820723076e-05,
+      "loss": 0.0048,
+      "step": 13140
+    },
+    {
+      "epoch": 2.4502725112964083,
+      "grad_norm": 0.09433647245168686,
+      "learning_rate": 3.662566778481799e-05,
+      "loss": 0.005,
+      "step": 13150
+    },
+    {
+      "epoch": 2.452135836400056,
+      "grad_norm": 0.05447697639465332,
+      "learning_rate": 3.650142874891291e-05,
+      "loss": 0.0059,
+      "step": 13160
+    },
+    {
+      "epoch": 2.4539991615037033,
+      "grad_norm": 0.06254272907972336,
+      "learning_rate": 3.637718971300783e-05,
+      "loss": 0.0054,
+      "step": 13170
+    },
+    {
+      "epoch": 2.4558624866073506,
+      "grad_norm": 0.030393915250897408,
+      "learning_rate": 3.6252950677102746e-05,
+      "loss": 0.0189,
+      "step": 13180
+    },
+    {
+      "epoch": 2.4577258117109984,
+      "grad_norm": 0.15565629303455353,
+      "learning_rate": 3.612871164119767e-05,
+      "loss": 0.0055,
+      "step": 13190
+    },
+    {
+      "epoch": 2.4595891368146456,
+      "grad_norm": 0.06977463513612747,
+      "learning_rate": 3.6004472605292584e-05,
+      "loss": 0.0062,
+      "step": 13200
+    },
+    {
+      "epoch": 2.4614524619182934,
+      "grad_norm": 0.025816500186920166,
+      "learning_rate": 3.588023356938751e-05,
+      "loss": 0.0049,
+      "step": 13210
+    },
+    {
+      "epoch": 2.4633157870219407,
+      "grad_norm": 0.05781162530183792,
+      "learning_rate": 3.575599453348242e-05,
+      "loss": 0.0071,
+      "step": 13220
+    },
+    {
+      "epoch": 2.465179112125588,
+      "grad_norm": 0.048669327050447464,
+      "learning_rate": 3.563175549757734e-05,
+      "loss": 0.0041,
+      "step": 13230
+    },
+    {
+      "epoch": 2.4670424372292357,
+      "grad_norm": 0.03702136129140854,
+      "learning_rate": 3.550751646167226e-05,
+      "loss": 0.005,
+      "step": 13240
+    },
+    {
+      "epoch": 2.468905762332883,
+      "grad_norm": 0.08766678720712662,
+      "learning_rate": 3.538327742576718e-05,
+      "loss": 0.0067,
+      "step": 13250
+    },
+    {
+      "epoch": 2.4707690874365307,
+      "grad_norm": 0.09898655116558075,
+      "learning_rate": 3.52590383898621e-05,
+      "loss": 0.0066,
+      "step": 13260
+    },
+    {
+      "epoch": 2.472632412540178,
+      "grad_norm": 0.11770444363355637,
+      "learning_rate": 3.5134799353957015e-05,
+      "loss": 0.0052,
+      "step": 13270
+    },
+    {
+      "epoch": 2.4744957376438252,
+      "grad_norm": 0.05603978410363197,
+      "learning_rate": 3.501056031805193e-05,
+      "loss": 0.0085,
+      "step": 13280
+    },
+    {
+      "epoch": 2.476359062747473,
+      "grad_norm": 0.12456575036048889,
+      "learning_rate": 3.488632128214685e-05,
+      "loss": 0.0056,
+      "step": 13290
+    },
+    {
+      "epoch": 2.4782223878511203,
+      "grad_norm": 0.047835420817136765,
+      "learning_rate": 3.4762082246241776e-05,
+      "loss": 0.0058,
+      "step": 13300
+    },
+    {
+      "epoch": 2.480085712954768,
+      "grad_norm": 0.06229006126523018,
+      "learning_rate": 3.463784321033669e-05,
+      "loss": 0.0062,
+      "step": 13310
+    },
+    {
+      "epoch": 2.4819490380584153,
+      "grad_norm": 0.030443726107478142,
+      "learning_rate": 3.451360417443161e-05,
+      "loss": 0.0052,
+      "step": 13320
+    },
+    {
+      "epoch": 2.4838123631620626,
+      "grad_norm": 0.09114624559879303,
+      "learning_rate": 3.4389365138526523e-05,
+      "loss": 0.0066,
+      "step": 13330
+    },
+    {
+      "epoch": 2.4856756882657103,
+      "grad_norm": 0.07564748078584671,
+      "learning_rate": 3.4265126102621446e-05,
+      "loss": 0.0068,
+      "step": 13340
+    },
+    {
+      "epoch": 2.4875390133693576,
+      "grad_norm": 0.07381059229373932,
+      "learning_rate": 3.414088706671636e-05,
+      "loss": 0.0079,
+      "step": 13350
+    },
+    {
+      "epoch": 2.4894023384730053,
+      "grad_norm": 0.03531397879123688,
+      "learning_rate": 3.4016648030811284e-05,
+      "loss": 0.0056,
+      "step": 13360
+    },
+    {
+      "epoch": 2.4912656635766526,
+      "grad_norm": 0.055100537836551666,
+      "learning_rate": 3.38924089949062e-05,
+      "loss": 0.0082,
+      "step": 13370
+    },
+    {
+      "epoch": 2.4931289886803,
+      "grad_norm": 0.07642512023448944,
+      "learning_rate": 3.376816995900112e-05,
+      "loss": 0.0048,
+      "step": 13380
+    },
+    {
+      "epoch": 2.4949923137839476,
+      "grad_norm": 0.09925362467765808,
+      "learning_rate": 3.364393092309604e-05,
+      "loss": 0.007,
+      "step": 13390
+    },
+    {
+      "epoch": 2.496855638887595,
+      "grad_norm": 0.059747230261564255,
+      "learning_rate": 3.3519691887190954e-05,
+      "loss": 0.0051,
+      "step": 13400
+    },
+    {
+      "epoch": 2.4987189639912426,
+      "grad_norm": 0.10106447339057922,
+      "learning_rate": 3.339545285128588e-05,
+      "loss": 0.0071,
+      "step": 13410
+    },
+    {
+      "epoch": 2.50058228909489,
+      "grad_norm": 0.07489795237779617,
+      "learning_rate": 3.327121381538079e-05,
+      "loss": 0.0074,
+      "step": 13420
+    },
+    {
+      "epoch": 2.502445614198537,
+      "grad_norm": 0.054214879870414734,
+      "learning_rate": 3.3146974779475715e-05,
+      "loss": 0.006,
+      "step": 13430
+    },
+    {
+      "epoch": 2.504308939302185,
+      "grad_norm": 0.05765419453382492,
+      "learning_rate": 3.302273574357063e-05,
+      "loss": 0.0053,
+      "step": 13440
+    },
+    {
+      "epoch": 2.506172264405832,
+      "grad_norm": 0.038531653583049774,
+      "learning_rate": 3.289849670766555e-05,
+      "loss": 0.0067,
+      "step": 13450
+    },
+    {
+      "epoch": 2.50803558950948,
+      "grad_norm": 0.08210862427949905,
+      "learning_rate": 3.277425767176047e-05,
+      "loss": 0.0055,
+      "step": 13460
+    },
+    {
+      "epoch": 2.509898914613127,
+      "grad_norm": 0.03968588635325432,
+      "learning_rate": 3.265001863585539e-05,
+      "loss": 0.0044,
+      "step": 13470
+    },
+    {
+      "epoch": 2.5117622397167745,
+      "grad_norm": 0.0714801549911499,
+      "learning_rate": 3.252577959995031e-05,
+      "loss": 0.0073,
+      "step": 13480
+    },
+    {
+      "epoch": 2.513625564820422,
+      "grad_norm": 0.031204110011458397,
+      "learning_rate": 3.2401540564045224e-05,
+      "loss": 0.0045,
+      "step": 13490
+    },
+    {
+      "epoch": 2.5154888899240695,
+      "grad_norm": 0.06460673362016678,
+      "learning_rate": 3.227730152814014e-05,
+      "loss": 0.008,
+      "step": 13500
+    },
+    {
+      "epoch": 2.517352215027717,
+      "grad_norm": 0.05626816302537918,
+      "learning_rate": 3.215306249223506e-05,
+      "loss": 0.0068,
+      "step": 13510
+    },
+    {
+      "epoch": 2.5192155401313645,
+      "grad_norm": 0.0728488489985466,
+      "learning_rate": 3.2028823456329985e-05,
+      "loss": 0.005,
+      "step": 13520
+    },
+    {
+      "epoch": 2.521078865235012,
+      "grad_norm": 0.075608991086483,
+      "learning_rate": 3.19045844204249e-05,
+      "loss": 0.0048,
+      "step": 13530
+    },
+    {
+      "epoch": 2.522942190338659,
+      "grad_norm": 0.0798393115401268,
+      "learning_rate": 3.1780345384519816e-05,
+      "loss": 0.0056,
+      "step": 13540
+    },
+    {
+      "epoch": 2.524805515442307,
+      "grad_norm": 0.07284027338027954,
+      "learning_rate": 3.165610634861473e-05,
+      "loss": 0.0054,
+      "step": 13550
+    },
+    {
+      "epoch": 2.5266688405459545,
+      "grad_norm": 0.07089147716760635,
+      "learning_rate": 3.1531867312709655e-05,
+      "loss": 0.0067,
+      "step": 13560
+    },
+    {
+      "epoch": 2.528532165649602,
+      "grad_norm": 0.11620229482650757,
+      "learning_rate": 3.140762827680457e-05,
+      "loss": 0.0062,
+      "step": 13570
+    },
+    {
+      "epoch": 2.530395490753249,
+      "grad_norm": 0.029851248487830162,
+      "learning_rate": 3.128338924089949e-05,
+      "loss": 0.007,
+      "step": 13580
+    },
+    {
+      "epoch": 2.5322588158568964,
+      "grad_norm": 0.0742005780339241,
+      "learning_rate": 3.115915020499441e-05,
+      "loss": 0.006,
+      "step": 13590
+    },
+    {
+      "epoch": 2.534122140960544,
+      "grad_norm": 0.06323488801717758,
+      "learning_rate": 3.103491116908933e-05,
+      "loss": 0.007,
+      "step": 13600
+    },
+    {
+      "epoch": 2.5359854660641914,
+      "grad_norm": 0.03448924049735069,
+      "learning_rate": 3.091067213318425e-05,
+      "loss": 0.0072,
+      "step": 13610
+    },
+    {
+      "epoch": 2.537848791167839,
+      "grad_norm": 0.04961373656988144,
+      "learning_rate": 3.078643309727916e-05,
+      "loss": 0.0062,
+      "step": 13620
+    },
+    {
+      "epoch": 2.5397121162714864,
+      "grad_norm": 0.049638889729976654,
+      "learning_rate": 3.0662194061374086e-05,
+      "loss": 0.0047,
+      "step": 13630
+    },
+    {
+      "epoch": 2.5415754413751337,
+      "grad_norm": 0.0409548357129097,
+      "learning_rate": 3.053795502546901e-05,
+      "loss": 0.0061,
+      "step": 13640
+    },
+    {
+      "epoch": 2.5434387664787814,
+      "grad_norm": 0.07288318872451782,
+      "learning_rate": 3.0413715989563924e-05,
+      "loss": 0.006,
+      "step": 13650
+    },
+    {
+      "epoch": 2.5453020915824287,
+      "grad_norm": 0.08695163577795029,
+      "learning_rate": 3.028947695365884e-05,
+      "loss": 0.0059,
+      "step": 13660
+    },
+    {
+      "epoch": 2.5471654166860764,
+      "grad_norm": 0.06203850731253624,
+      "learning_rate": 3.016523791775376e-05,
+      "loss": 0.0066,
+      "step": 13670
+    },
+    {
+      "epoch": 2.5490287417897237,
+      "grad_norm": 0.0684581995010376,
+      "learning_rate": 3.0040998881848682e-05,
+      "loss": 0.0061,
+      "step": 13680
+    },
+    {
+      "epoch": 2.550892066893371,
+      "grad_norm": 0.0407898835837841,
+      "learning_rate": 2.9916759845943598e-05,
+      "loss": 0.0042,
+      "step": 13690
+    },
+    {
+      "epoch": 2.5527553919970187,
+      "grad_norm": 0.07625031471252441,
+      "learning_rate": 2.9792520810038517e-05,
+      "loss": 0.0059,
+      "step": 13700
+    },
+    {
+      "epoch": 2.554618717100666,
+      "grad_norm": 0.07639207690954208,
+      "learning_rate": 2.9668281774133433e-05,
+      "loss": 0.0065,
+      "step": 13710
+    },
+    {
+      "epoch": 2.5564820422043137,
+      "grad_norm": 0.09934278577566147,
+      "learning_rate": 2.9544042738228352e-05,
+      "loss": 0.0041,
+      "step": 13720
+    },
+    {
+      "epoch": 2.558345367307961,
+      "grad_norm": 0.03425678610801697,
+      "learning_rate": 2.9419803702323274e-05,
+      "loss": 0.0069,
+      "step": 13730
+    },
+    {
+      "epoch": 2.5602086924116083,
+      "grad_norm": 0.14032401144504547,
+      "learning_rate": 2.929556466641819e-05,
+      "loss": 0.0088,
+      "step": 13740
+    },
+    {
+      "epoch": 2.562072017515256,
+      "grad_norm": 0.04956042021512985,
+      "learning_rate": 2.917132563051311e-05,
+      "loss": 0.0054,
+      "step": 13750
+    },
+    {
+      "epoch": 2.5639353426189033,
+      "grad_norm": 0.013684896752238274,
+      "learning_rate": 2.9047086594608025e-05,
+      "loss": 0.0092,
+      "step": 13760
+    },
+    {
+      "epoch": 2.565798667722551,
+      "grad_norm": 0.06913986057043076,
+      "learning_rate": 2.8922847558702948e-05,
+      "loss": 0.005,
+      "step": 13770
+    },
+    {
+      "epoch": 2.5676619928261983,
+      "grad_norm": 0.06730835139751434,
+      "learning_rate": 2.8798608522797864e-05,
+      "loss": 0.0094,
+      "step": 13780
+    },
+    {
+      "epoch": 2.5695253179298456,
+      "grad_norm": 0.045726943761110306,
+      "learning_rate": 2.8674369486892783e-05,
+      "loss": 0.0043,
+      "step": 13790
+    },
+    {
+      "epoch": 2.5713886430334933,
+      "grad_norm": 0.07426123321056366,
+      "learning_rate": 2.85501304509877e-05,
+      "loss": 0.01,
+      "step": 13800
+    },
+    {
+      "epoch": 2.5732519681371406,
+      "grad_norm": 0.044499099254608154,
+      "learning_rate": 2.842589141508262e-05,
+      "loss": 0.0054,
+      "step": 13810
+    },
+    {
+      "epoch": 2.5751152932407884,
+      "grad_norm": 0.09332112967967987,
+      "learning_rate": 2.830165237917754e-05,
+      "loss": 0.0042,
+      "step": 13820
+    },
+    {
+      "epoch": 2.5769786183444356,
+      "grad_norm": 0.017959794029593468,
+      "learning_rate": 2.8177413343272456e-05,
+      "loss": 0.0052,
+      "step": 13830
+    },
+    {
+      "epoch": 2.578841943448083,
+      "grad_norm": 0.043127838522195816,
+      "learning_rate": 2.8053174307367375e-05,
+      "loss": 0.0056,
+      "step": 13840
+    },
+    {
+      "epoch": 2.5807052685517307,
+      "grad_norm": 0.07607263326644897,
+      "learning_rate": 2.792893527146229e-05,
+      "loss": 0.0041,
+      "step": 13850
+    },
+    {
+      "epoch": 2.582568593655378,
+      "grad_norm": 0.0485481321811676,
+      "learning_rate": 2.7804696235557214e-05,
+      "loss": 0.0077,
+      "step": 13860
+    },
+    {
+      "epoch": 2.5844319187590257,
+      "grad_norm": 0.06554541736841202,
+      "learning_rate": 2.7680457199652133e-05,
+      "loss": 0.0057,
+      "step": 13870
+    },
+    {
+      "epoch": 2.586295243862673,
+      "grad_norm": 0.04134681820869446,
+      "learning_rate": 2.755621816374705e-05,
+      "loss": 0.0045,
+      "step": 13880
+    },
+    {
+      "epoch": 2.5881585689663202,
+      "grad_norm": 0.052181366831064224,
+      "learning_rate": 2.7431979127841968e-05,
+      "loss": 0.0042,
+      "step": 13890
+    },
+    {
+      "epoch": 2.590021894069968,
+      "grad_norm": 0.06621751934289932,
+      "learning_rate": 2.730774009193689e-05,
+      "loss": 0.0067,
+      "step": 13900
+    },
+    {
+      "epoch": 2.5918852191736153,
+      "grad_norm": 0.056223683059215546,
+      "learning_rate": 2.7183501056031806e-05,
+      "loss": 0.0061,
+      "step": 13910
+    },
+    {
+      "epoch": 2.593748544277263,
+      "grad_norm": 0.0572345070540905,
+      "learning_rate": 2.7059262020126726e-05,
+      "loss": 0.0064,
+      "step": 13920
+    },
+    {
+      "epoch": 2.5956118693809103,
+      "grad_norm": 0.04334909841418266,
+      "learning_rate": 2.693502298422164e-05,
+      "loss": 0.0039,
+      "step": 13930
+    },
+    {
+      "epoch": 2.5974751944845575,
+      "grad_norm": 0.08207881450653076,
+      "learning_rate": 2.6810783948316564e-05,
+      "loss": 0.0056,
+      "step": 13940
+    },
+    {
+      "epoch": 2.5993385195882053,
+      "grad_norm": 0.5891373157501221,
+      "learning_rate": 2.6686544912411483e-05,
+      "loss": 0.007,
+      "step": 13950
+    },
+    {
+      "epoch": 2.6012018446918526,
+      "grad_norm": 0.041573237627744675,
+      "learning_rate": 2.65623058765064e-05,
+      "loss": 0.004,
+      "step": 13960
+    },
+    {
+      "epoch": 2.6030651697955003,
+      "grad_norm": 0.071868397295475,
+      "learning_rate": 2.6438066840601318e-05,
+      "loss": 0.0051,
+      "step": 13970
+    },
+    {
+      "epoch": 2.6049284948991476,
+      "grad_norm": 0.07241298258304596,
+      "learning_rate": 2.631382780469624e-05,
+      "loss": 0.0039,
+      "step": 13980
+    },
+    {
+      "epoch": 2.606791820002795,
+      "grad_norm": 0.0609959252178669,
+      "learning_rate": 2.6189588768791157e-05,
+      "loss": 0.0066,
+      "step": 13990
+    },
+    {
+      "epoch": 2.6086551451064426,
+      "grad_norm": 0.051202163100242615,
+      "learning_rate": 2.6065349732886072e-05,
+      "loss": 0.0064,
+      "step": 14000
+    },
+    {
+      "epoch": 2.61051847021009,
+      "grad_norm": 0.06758036464452744,
+      "learning_rate": 2.594111069698099e-05,
+      "loss": 0.0052,
+      "step": 14010
+    },
+    {
+      "epoch": 2.6123817953137376,
+      "grad_norm": 0.01979978010058403,
+      "learning_rate": 2.5816871661075907e-05,
+      "loss": 0.0047,
+      "step": 14020
+    },
+    {
+      "epoch": 2.614245120417385,
+      "grad_norm": 0.05341746285557747,
+      "learning_rate": 2.569263262517083e-05,
+      "loss": 0.0056,
+      "step": 14030
+    },
+    {
+      "epoch": 2.616108445521032,
+      "grad_norm": 0.05054103955626488,
+      "learning_rate": 2.556839358926575e-05,
+      "loss": 0.0057,
+      "step": 14040
+    },
+    {
+      "epoch": 2.61797177062468,
+      "grad_norm": 0.03094291128218174,
+      "learning_rate": 2.5444154553360665e-05,
+      "loss": 0.0041,
+      "step": 14050
+    },
+    {
+      "epoch": 2.619835095728327,
+      "grad_norm": 0.060007113963365555,
+      "learning_rate": 2.5319915517455584e-05,
+      "loss": 0.0062,
+      "step": 14060
+    },
+    {
+      "epoch": 2.621698420831975,
+      "grad_norm": 0.058343224227428436,
+      "learning_rate": 2.5195676481550507e-05,
+      "loss": 0.0069,
+      "step": 14070
+    },
+    {
+      "epoch": 2.623561745935622,
+      "grad_norm": 0.03955060616135597,
+      "learning_rate": 2.5071437445645423e-05,
+      "loss": 0.0053,
+      "step": 14080
+    },
+    {
+      "epoch": 2.6254250710392695,
+      "grad_norm": 0.08612386882305145,
+      "learning_rate": 2.4947198409740342e-05,
+      "loss": 0.0075,
+      "step": 14090
+    },
+    {
+      "epoch": 2.6272883961429168,
+      "grad_norm": 0.07171555608510971,
+      "learning_rate": 2.482295937383526e-05,
+      "loss": 0.0051,
+      "step": 14100
+    },
+    {
+      "epoch": 2.6291517212465645,
+      "grad_norm": 0.06530676782131195,
+      "learning_rate": 2.469872033793018e-05,
+      "loss": 0.0052,
+      "step": 14110
+    },
+    {
+      "epoch": 2.631015046350212,
+      "grad_norm": 0.08867678046226501,
+      "learning_rate": 2.45744813020251e-05,
+      "loss": 0.0088,
+      "step": 14120
+    },
+    {
+      "epoch": 2.6328783714538595,
+      "grad_norm": 0.05115736648440361,
+      "learning_rate": 2.4450242266120015e-05,
+      "loss": 0.0036,
+      "step": 14130
+    },
+    {
+      "epoch": 2.634741696557507,
+      "grad_norm": 0.046469271183013916,
+      "learning_rate": 2.4326003230214934e-05,
+      "loss": 0.007,
+      "step": 14140
+    },
+    {
+      "epoch": 2.636605021661154,
+      "grad_norm": 0.045745860785245895,
+      "learning_rate": 2.4201764194309854e-05,
+      "loss": 0.0055,
+      "step": 14150
+    },
+    {
+      "epoch": 2.638468346764802,
+      "grad_norm": 0.08737187087535858,
+      "learning_rate": 2.407752515840477e-05,
+      "loss": 0.0049,
+      "step": 14160
+    },
+    {
+      "epoch": 2.640331671868449,
+      "grad_norm": 0.0638500526547432,
+      "learning_rate": 2.3953286122499692e-05,
+      "loss": 0.0034,
+      "step": 14170
+    },
+    {
+      "epoch": 2.642194996972097,
+      "grad_norm": 0.04362380877137184,
+      "learning_rate": 2.3829047086594608e-05,
+      "loss": 0.004,
+      "step": 14180
+    },
+    {
+      "epoch": 2.644058322075744,
+      "grad_norm": 0.056783124804496765,
+      "learning_rate": 2.3704808050689527e-05,
+      "loss": 0.0071,
+      "step": 14190
+    },
+    {
+      "epoch": 2.6459216471793914,
+      "grad_norm": 0.050139475613832474,
+      "learning_rate": 2.3580569014784446e-05,
+      "loss": 0.0053,
+      "step": 14200
+    },
+    {
+      "epoch": 2.647784972283039,
+      "grad_norm": 0.04129103943705559,
+      "learning_rate": 2.3456329978879365e-05,
+      "loss": 0.0079,
+      "step": 14210
+    },
+    {
+      "epoch": 2.6496482973866864,
+      "grad_norm": 0.07041865587234497,
+      "learning_rate": 2.3332090942974285e-05,
+      "loss": 0.0054,
+      "step": 14220
+    },
+    {
+      "epoch": 2.651511622490334,
+      "grad_norm": 0.08971253037452698,
+      "learning_rate": 2.3207851907069204e-05,
+      "loss": 0.0053,
+      "step": 14230
+    },
+    {
+      "epoch": 2.6533749475939814,
+      "grad_norm": 0.060063548386096954,
+      "learning_rate": 2.308361287116412e-05,
+      "loss": 0.0042,
+      "step": 14240
+    },
+    {
+      "epoch": 2.6552382726976287,
+      "grad_norm": 0.02573762647807598,
+      "learning_rate": 2.295937383525904e-05,
+      "loss": 0.0041,
+      "step": 14250
+    },
+    {
+      "epoch": 2.6571015978012764,
+      "grad_norm": 0.06436144560575485,
+      "learning_rate": 2.2835134799353958e-05,
+      "loss": 0.0051,
+      "step": 14260
+    },
+    {
+      "epoch": 2.6589649229049237,
+      "grad_norm": 0.04029161483049393,
+      "learning_rate": 2.2710895763448877e-05,
+      "loss": 0.0054,
+      "step": 14270
+    },
+    {
+      "epoch": 2.6608282480085714,
+      "grad_norm": 0.12254491448402405,
+      "learning_rate": 2.2586656727543796e-05,
+      "loss": 0.0078,
+      "step": 14280
+    },
+    {
+      "epoch": 2.6626915731122187,
+      "grad_norm": 0.0719374418258667,
+      "learning_rate": 2.2462417691638716e-05,
+      "loss": 0.0096,
+      "step": 14290
+    },
+    {
+      "epoch": 2.664554898215866,
+      "grad_norm": 0.055200230330228806,
+      "learning_rate": 2.233817865573363e-05,
+      "loss": 0.0062,
+      "step": 14300
+    },
+    {
+      "epoch": 2.6664182233195137,
+      "grad_norm": 0.07559479773044586,
+      "learning_rate": 2.221393961982855e-05,
+      "loss": 0.0064,
+      "step": 14310
+    },
+    {
+      "epoch": 2.668281548423161,
+      "grad_norm": 0.08524436503648758,
+      "learning_rate": 2.208970058392347e-05,
+      "loss": 0.0043,
+      "step": 14320
+    },
+    {
+      "epoch": 2.6701448735268087,
+      "grad_norm": 0.045584194362163544,
+      "learning_rate": 2.196546154801839e-05,
+      "loss": 0.0077,
+      "step": 14330
+    },
+    {
+      "epoch": 2.672008198630456,
+      "grad_norm": 0.15084679424762726,
+      "learning_rate": 2.1841222512113308e-05,
+      "loss": 0.0062,
+      "step": 14340
+    },
+    {
+      "epoch": 2.6738715237341033,
+      "grad_norm": 0.08916497230529785,
+      "learning_rate": 2.1716983476208224e-05,
+      "loss": 0.0053,
+      "step": 14350
+    },
+    {
+      "epoch": 2.675734848837751,
+      "grad_norm": 0.047935601323843,
+      "learning_rate": 2.1592744440303143e-05,
+      "loss": 0.0049,
+      "step": 14360
+    },
+    {
+      "epoch": 2.6775981739413983,
+      "grad_norm": 0.05849865823984146,
+      "learning_rate": 2.1468505404398062e-05,
+      "loss": 0.0061,
+      "step": 14370
+    },
+    {
+      "epoch": 2.679461499045046,
+      "grad_norm": 0.03962693363428116,
+      "learning_rate": 2.134426636849298e-05,
+      "loss": 0.0039,
+      "step": 14380
+    },
+    {
+      "epoch": 2.6813248241486933,
+      "grad_norm": 0.05341466888785362,
+      "learning_rate": 2.12200273325879e-05,
+      "loss": 0.0053,
+      "step": 14390
+    },
+    {
+      "epoch": 2.6831881492523406,
+      "grad_norm": 0.04300933703780174,
+      "learning_rate": 2.109578829668282e-05,
+      "loss": 0.0043,
+      "step": 14400
+    },
+    {
+      "epoch": 2.6850514743559883,
+      "grad_norm": 0.0594099722802639,
+      "learning_rate": 2.0971549260777736e-05,
+      "loss": 0.0087,
+      "step": 14410
+    },
+    {
+      "epoch": 2.6869147994596356,
+      "grad_norm": 0.07009299099445343,
+      "learning_rate": 2.084731022487266e-05,
+      "loss": 0.0054,
+      "step": 14420
+    },
+    {
+      "epoch": 2.6887781245632834,
+      "grad_norm": 0.0624958761036396,
+      "learning_rate": 2.0723071188967574e-05,
+      "loss": 0.0048,
+      "step": 14430
+    },
+    {
+      "epoch": 2.6906414496669306,
+      "grad_norm": 0.0850406363606453,
+      "learning_rate": 2.0598832153062493e-05,
+      "loss": 0.0066,
+      "step": 14440
+    },
+    {
+      "epoch": 2.692504774770578,
+      "grad_norm": 0.07980000972747803,
+      "learning_rate": 2.0474593117157413e-05,
+      "loss": 0.0055,
+      "step": 14450
+    },
+    {
+      "epoch": 2.6943680998742257,
+      "grad_norm": 0.04403868690133095,
+      "learning_rate": 2.035035408125233e-05,
+      "loss": 0.0037,
+      "step": 14460
+    },
+    {
+      "epoch": 2.696231424977873,
+      "grad_norm": 0.06792305409908295,
+      "learning_rate": 2.0226115045347248e-05,
+      "loss": 0.0048,
+      "step": 14470
+    },
+    {
+      "epoch": 2.6980947500815207,
+      "grad_norm": 0.07273188978433609,
+      "learning_rate": 2.0101876009442167e-05,
+      "loss": 0.0044,
+      "step": 14480
+    },
+    {
+      "epoch": 2.699958075185168,
+      "grad_norm": 0.06579861044883728,
+      "learning_rate": 1.9977636973537086e-05,
+      "loss": 0.0061,
+      "step": 14490
+    },
+    {
+      "epoch": 2.7018214002888152,
+      "grad_norm": 0.036138229072093964,
+      "learning_rate": 1.9853397937632005e-05,
+      "loss": 0.0052,
+      "step": 14500
+    },
+    {
+      "epoch": 2.703684725392463,
+      "grad_norm": 0.07591810822486877,
+      "learning_rate": 1.9729158901726924e-05,
+      "loss": 0.0068,
+      "step": 14510
+    },
+    {
+      "epoch": 2.7055480504961102,
+      "grad_norm": 0.11958233267068863,
+      "learning_rate": 1.960491986582184e-05,
+      "loss": 0.0049,
+      "step": 14520
+    },
+    {
+      "epoch": 2.707411375599758,
+      "grad_norm": 0.05322478339076042,
+      "learning_rate": 1.9480680829916763e-05,
+      "loss": 0.0042,
+      "step": 14530
+    },
+    {
+      "epoch": 2.7092747007034053,
+      "grad_norm": 0.060096897184848785,
+      "learning_rate": 1.935644179401168e-05,
+      "loss": 0.0043,
+      "step": 14540
+    },
+    {
+      "epoch": 2.7111380258070525,
+      "grad_norm": 0.0774926245212555,
+      "learning_rate": 1.9232202758106598e-05,
+      "loss": 0.0058,
+      "step": 14550
+    },
+    {
+      "epoch": 2.7130013509107003,
+      "grad_norm": 0.08724630624055862,
+      "learning_rate": 1.9107963722201517e-05,
+      "loss": 0.0047,
+      "step": 14560
+    },
+    {
+      "epoch": 2.7148646760143476,
+      "grad_norm": 0.02089923806488514,
+      "learning_rate": 1.8983724686296436e-05,
+      "loss": 0.0055,
+      "step": 14570
+    },
+    {
+      "epoch": 2.7167280011179953,
+      "grad_norm": 0.04341026023030281,
+      "learning_rate": 1.8859485650391352e-05,
+      "loss": 0.0068,
+      "step": 14580
+    },
+    {
+      "epoch": 2.7185913262216426,
+      "grad_norm": 0.05886482074856758,
+      "learning_rate": 1.8735246614486275e-05,
+      "loss": 0.0041,
+      "step": 14590
+    },
+    {
+      "epoch": 2.72045465132529,
+      "grad_norm": 0.030764520168304443,
+      "learning_rate": 1.861100757858119e-05,
+      "loss": 0.0045,
+      "step": 14600
+    },
+    {
+      "epoch": 2.7223179764289376,
+      "grad_norm": 0.08881295472383499,
+      "learning_rate": 1.848676854267611e-05,
+      "loss": 0.004,
+      "step": 14610
+    },
+    {
+      "epoch": 2.724181301532585,
+      "grad_norm": 0.03445763513445854,
+      "learning_rate": 1.836252950677103e-05,
+      "loss": 0.0044,
+      "step": 14620
+    },
+    {
+      "epoch": 2.7260446266362326,
+      "grad_norm": 0.09047674387693405,
+      "learning_rate": 1.8238290470865945e-05,
+      "loss": 0.0049,
+      "step": 14630
+    },
+    {
+      "epoch": 2.72790795173988,
+      "grad_norm": 0.04898625984787941,
+      "learning_rate": 1.8114051434960867e-05,
+      "loss": 0.0056,
+      "step": 14640
+    },
+    {
+      "epoch": 2.729771276843527,
+      "grad_norm": 0.13687947392463684,
+      "learning_rate": 1.7989812399055783e-05,
+      "loss": 0.0048,
+      "step": 14650
+    },
+    {
+      "epoch": 2.731634601947175,
+      "grad_norm": 0.10672298073768616,
+      "learning_rate": 1.7865573363150702e-05,
+      "loss": 0.0096,
+      "step": 14660
+    },
+    {
+      "epoch": 2.733497927050822,
+      "grad_norm": 0.07122719287872314,
+      "learning_rate": 1.774133432724562e-05,
+      "loss": 0.005,
+      "step": 14670
+    },
+    {
+      "epoch": 2.73536125215447,
+      "grad_norm": 0.11941694468259811,
+      "learning_rate": 1.761709529134054e-05,
+      "loss": 0.0066,
+      "step": 14680
+    },
+    {
+      "epoch": 2.737224577258117,
+      "grad_norm": 0.14367525279521942,
+      "learning_rate": 1.749285625543546e-05,
+      "loss": 0.0057,
+      "step": 14690
+    },
+    {
+      "epoch": 2.7390879023617645,
+      "grad_norm": 0.09323915839195251,
+      "learning_rate": 1.736861721953038e-05,
+      "loss": 0.0053,
+      "step": 14700
+    },
+    {
+      "epoch": 2.7409512274654118,
+      "grad_norm": 0.08907133340835571,
+      "learning_rate": 1.7244378183625295e-05,
+      "loss": 0.0043,
+      "step": 14710
+    },
+    {
+      "epoch": 2.7428145525690595,
+      "grad_norm": 0.05132559314370155,
+      "learning_rate": 1.7120139147720214e-05,
+      "loss": 0.0042,
+      "step": 14720
+    },
+    {
+      "epoch": 2.744677877672707,
+      "grad_norm": 0.08204059302806854,
+      "learning_rate": 1.6995900111815133e-05,
+      "loss": 0.0074,
+      "step": 14730
+    },
+    {
+      "epoch": 2.7465412027763545,
+      "grad_norm": 0.08662567287683487,
+      "learning_rate": 1.6871661075910052e-05,
+      "loss": 0.0056,
+      "step": 14740
+    },
+    {
+      "epoch": 2.748404527880002,
+      "grad_norm": 0.08066625148057938,
+      "learning_rate": 1.674742204000497e-05,
+      "loss": 0.0063,
+      "step": 14750
+    },
+    {
+      "epoch": 2.750267852983649,
+      "grad_norm": 0.08454219996929169,
+      "learning_rate": 1.6623183004099887e-05,
+      "loss": 0.0057,
+      "step": 14760
+    },
+    {
+      "epoch": 2.752131178087297,
+      "grad_norm": 0.07747121155261993,
+      "learning_rate": 1.6498943968194807e-05,
+      "loss": 0.0039,
+      "step": 14770
+    },
+    {
+      "epoch": 2.753994503190944,
+      "grad_norm": 0.0393957644701004,
+      "learning_rate": 1.6374704932289726e-05,
+      "loss": 0.0052,
+      "step": 14780
+    },
+    {
+      "epoch": 2.755857828294592,
+      "grad_norm": 0.03469419479370117,
+      "learning_rate": 1.6250465896384645e-05,
+      "loss": 0.0052,
+      "step": 14790
+    },
+    {
+      "epoch": 2.757721153398239,
+      "grad_norm": 0.07738560438156128,
+      "learning_rate": 1.6126226860479564e-05,
+      "loss": 0.0046,
+      "step": 14800
+    },
+    {
+      "epoch": 2.7595844785018864,
+      "grad_norm": 0.06221096217632294,
+      "learning_rate": 1.6001987824574483e-05,
+      "loss": 0.0038,
+      "step": 14810
+    },
+    {
+      "epoch": 2.761447803605534,
+      "grad_norm": 0.027473099529743195,
+      "learning_rate": 1.58777487886694e-05,
+      "loss": 0.0047,
+      "step": 14820
+    },
+    {
+      "epoch": 2.7633111287091814,
+      "grad_norm": 0.05012982711195946,
+      "learning_rate": 1.575350975276432e-05,
+      "loss": 0.0046,
+      "step": 14830
+    },
+    {
+      "epoch": 2.765174453812829,
+      "grad_norm": 0.07022465020418167,
+      "learning_rate": 1.5629270716859238e-05,
+      "loss": 0.006,
+      "step": 14840
+    },
+    {
+      "epoch": 2.7670377789164764,
+      "grad_norm": 0.03442366048693657,
+      "learning_rate": 1.5505031680954157e-05,
+      "loss": 0.0034,
+      "step": 14850
+    },
+    {
+      "epoch": 2.7689011040201237,
+      "grad_norm": 0.04125374183058739,
+      "learning_rate": 1.5380792645049076e-05,
+      "loss": 0.0058,
+      "step": 14860
+    },
+    {
+      "epoch": 2.7707644291237714,
+      "grad_norm": 0.043936800211668015,
+      "learning_rate": 1.5256553609143995e-05,
+      "loss": 0.0035,
+      "step": 14870
+    },
+    {
+      "epoch": 2.7726277542274187,
+      "grad_norm": 0.03655455261468887,
+      "learning_rate": 1.5132314573238913e-05,
+      "loss": 0.0043,
+      "step": 14880
+    },
+    {
+      "epoch": 2.7744910793310664,
+      "grad_norm": 0.05578184127807617,
+      "learning_rate": 1.5008075537333832e-05,
+      "loss": 0.0038,
+      "step": 14890
+    },
+    {
+      "epoch": 2.7763544044347137,
+      "grad_norm": 0.03317585960030556,
+      "learning_rate": 1.488383650142875e-05,
+      "loss": 0.0051,
+      "step": 14900
+    },
+    {
+      "epoch": 2.778217729538361,
+      "grad_norm": 0.09801258891820908,
+      "learning_rate": 1.475959746552367e-05,
+      "loss": 0.0089,
+      "step": 14910
+    },
+    {
+      "epoch": 2.7800810546420087,
+      "grad_norm": 0.034611500799655914,
+      "learning_rate": 1.4635358429618588e-05,
+      "loss": 0.0046,
+      "step": 14920
+    },
+    {
+      "epoch": 2.781944379745656,
+      "grad_norm": 0.05122116208076477,
+      "learning_rate": 1.4511119393713504e-05,
+      "loss": 0.0064,
+      "step": 14930
+    },
+    {
+      "epoch": 2.7838077048493037,
+      "grad_norm": 0.08390213549137115,
+      "learning_rate": 1.4386880357808424e-05,
+      "loss": 0.0055,
+      "step": 14940
+    },
+    {
+      "epoch": 2.785671029952951,
+      "grad_norm": 0.071256123483181,
+      "learning_rate": 1.4262641321903342e-05,
+      "loss": 0.0057,
+      "step": 14950
+    },
+    {
+      "epoch": 2.7875343550565983,
+      "grad_norm": 0.061553046107292175,
+      "learning_rate": 1.4138402285998261e-05,
+      "loss": 0.0075,
+      "step": 14960
+    },
+    {
+      "epoch": 2.789397680160246,
+      "grad_norm": 0.06200006976723671,
+      "learning_rate": 1.4014163250093179e-05,
+      "loss": 0.0047,
+      "step": 14970
+    },
+    {
+      "epoch": 2.7912610052638933,
+      "grad_norm": 0.04005870595574379,
+      "learning_rate": 1.38899242141881e-05,
+      "loss": 0.0052,
+      "step": 14980
+    },
+    {
+      "epoch": 2.793124330367541,
+      "grad_norm": 0.04067656025290489,
+      "learning_rate": 1.3765685178283017e-05,
+      "loss": 0.0049,
+      "step": 14990
+    },
+    {
+      "epoch": 2.7949876554711883,
+      "grad_norm": 0.10033728182315826,
+      "learning_rate": 1.3641446142377936e-05,
+      "loss": 0.0068,
+      "step": 15000
+    },
+    {
+      "epoch": 2.7968509805748356,
+      "grad_norm": 0.049095358699560165,
+      "learning_rate": 1.3517207106472854e-05,
+      "loss": 0.0044,
+      "step": 15010
+    },
+    {
+      "epoch": 2.7987143056784833,
+      "grad_norm": 0.03246445581316948,
+      "learning_rate": 1.3392968070567775e-05,
+      "loss": 0.004,
+      "step": 15020
+    },
+    {
+      "epoch": 2.8005776307821306,
+      "grad_norm": 0.0769830197095871,
+      "learning_rate": 1.3268729034662692e-05,
+      "loss": 0.006,
+      "step": 15030
+    },
+    {
+      "epoch": 2.8024409558857784,
+      "grad_norm": 0.08352731913328171,
+      "learning_rate": 1.3144489998757611e-05,
+      "loss": 0.007,
+      "step": 15040
+    },
+    {
+      "epoch": 2.8043042809894256,
+      "grad_norm": 0.04997175931930542,
+      "learning_rate": 1.3020250962852529e-05,
+      "loss": 0.0038,
+      "step": 15050
+    },
+    {
+      "epoch": 2.806167606093073,
+      "grad_norm": 0.08544106036424637,
+      "learning_rate": 1.289601192694745e-05,
+      "loss": 0.0067,
+      "step": 15060
+    },
+    {
+      "epoch": 2.8080309311967206,
+      "grad_norm": 0.05177277326583862,
+      "learning_rate": 1.2771772891042365e-05,
+      "loss": 0.0038,
+      "step": 15070
+    },
+    {
+      "epoch": 2.809894256300368,
+      "grad_norm": 0.06597385555505753,
+      "learning_rate": 1.2647533855137283e-05,
+      "loss": 0.0073,
+      "step": 15080
+    },
+    {
+      "epoch": 2.8117575814040157,
+      "grad_norm": 0.047354232519865036,
+      "learning_rate": 1.2523294819232204e-05,
+      "loss": 0.0035,
+      "step": 15090
+    },
+    {
+      "epoch": 2.813620906507663,
+      "grad_norm": 0.04057525470852852,
+      "learning_rate": 1.2399055783327123e-05,
+      "loss": 0.0047,
+      "step": 15100
+    },
+    {
+      "epoch": 2.8154842316113102,
+      "grad_norm": 0.037249766290187836,
+      "learning_rate": 1.227481674742204e-05,
+      "loss": 0.0033,
+      "step": 15110
+    },
+    {
+      "epoch": 2.817347556714958,
+      "grad_norm": 0.11450658738613129,
+      "learning_rate": 1.215057771151696e-05,
+      "loss": 0.0098,
+      "step": 15120
+    },
+    {
+      "epoch": 2.8192108818186052,
+      "grad_norm": 0.03987234830856323,
+      "learning_rate": 1.2026338675611879e-05,
+      "loss": 0.0051,
+      "step": 15130
+    },
+    {
+      "epoch": 2.821074206922253,
+      "grad_norm": 0.027388526126742363,
+      "learning_rate": 1.1902099639706796e-05,
+      "loss": 0.0055,
+      "step": 15140
+    },
+    {
+      "epoch": 2.8229375320259003,
+      "grad_norm": 0.04660959541797638,
+      "learning_rate": 1.1777860603801714e-05,
+      "loss": 0.0042,
+      "step": 15150
+    },
+    {
+      "epoch": 2.8248008571295475,
+      "grad_norm": 0.0806809738278389,
+      "learning_rate": 1.1653621567896633e-05,
+      "loss": 0.0061,
+      "step": 15160
+    },
+    {
+      "epoch": 2.8266641822331953,
+      "grad_norm": 0.04064582288265228,
+      "learning_rate": 1.1529382531991552e-05,
+      "loss": 0.0063,
+      "step": 15170
+    },
+    {
+      "epoch": 2.8285275073368426,
+      "grad_norm": 0.09423641860485077,
+      "learning_rate": 1.140514349608647e-05,
+      "loss": 0.0043,
+      "step": 15180
+    },
+    {
+      "epoch": 2.8303908324404903,
+      "grad_norm": 0.03595752641558647,
+      "learning_rate": 1.1280904460181389e-05,
+      "loss": 0.0046,
+      "step": 15190
+    },
+    {
+      "epoch": 2.8322541575441376,
+      "grad_norm": 0.05157044902443886,
+      "learning_rate": 1.1156665424276308e-05,
+      "loss": 0.0049,
+      "step": 15200
+    },
+    {
+      "epoch": 2.834117482647785,
+      "grad_norm": 0.025802219286561012,
+      "learning_rate": 1.1032426388371227e-05,
+      "loss": 0.004,
+      "step": 15210
+    },
+    {
+      "epoch": 2.8359808077514326,
+      "grad_norm": 0.04284001141786575,
+      "learning_rate": 1.0908187352466145e-05,
+      "loss": 0.0049,
+      "step": 15220
+    },
+    {
+      "epoch": 2.83784413285508,
+      "grad_norm": 0.06535849720239639,
+      "learning_rate": 1.0783948316561064e-05,
+      "loss": 0.0053,
+      "step": 15230
+    },
+    {
+      "epoch": 2.8397074579587276,
+      "grad_norm": 0.034743234515190125,
+      "learning_rate": 1.0659709280655983e-05,
+      "loss": 0.0037,
+      "step": 15240
+    },
+    {
+      "epoch": 2.841570783062375,
+      "grad_norm": 0.04481494799256325,
+      "learning_rate": 1.0535470244750901e-05,
+      "loss": 0.0091,
+      "step": 15250
+    },
+    {
+      "epoch": 2.843434108166022,
+      "grad_norm": 0.027818024158477783,
+      "learning_rate": 1.041123120884582e-05,
+      "loss": 0.0056,
+      "step": 15260
+    },
+    {
+      "epoch": 2.84529743326967,
+      "grad_norm": 0.07148770242929459,
+      "learning_rate": 1.028699217294074e-05,
+      "loss": 0.0042,
+      "step": 15270
+    },
+    {
+      "epoch": 2.847160758373317,
+      "grad_norm": 0.07908165454864502,
+      "learning_rate": 1.0162753137035658e-05,
+      "loss": 0.0052,
+      "step": 15280
+    },
+    {
+      "epoch": 2.849024083476965,
+      "grad_norm": 0.0394151397049427,
+      "learning_rate": 1.0038514101130576e-05,
+      "loss": 0.005,
+      "step": 15290
+    },
+    {
+      "epoch": 2.850887408580612,
+      "grad_norm": 0.07566548883914948,
+      "learning_rate": 9.914275065225493e-06,
+      "loss": 0.0066,
+      "step": 15300
+    },
+    {
+      "epoch": 2.8527507336842595,
+      "grad_norm": 0.043212130665779114,
+      "learning_rate": 9.790036029320413e-06,
+      "loss": 0.006,
+      "step": 15310
+    },
+    {
+      "epoch": 2.8546140587879067,
+      "grad_norm": 0.042301010340452194,
+      "learning_rate": 9.665796993415332e-06,
+      "loss": 0.0042,
+      "step": 15320
+    },
+    {
+      "epoch": 2.8564773838915545,
+      "grad_norm": 0.0396822988986969,
+      "learning_rate": 9.54155795751025e-06,
+      "loss": 0.0041,
+      "step": 15330
+    },
+    {
+      "epoch": 2.858340708995202,
+      "grad_norm": 0.1604224443435669,
+      "learning_rate": 9.417318921605169e-06,
+      "loss": 0.0049,
+      "step": 15340
+    },
+    {
+      "epoch": 2.8602040340988495,
+      "grad_norm": 0.27743810415267944,
+      "learning_rate": 9.293079885700088e-06,
+      "loss": 0.0034,
+      "step": 15350
+    },
+    {
+      "epoch": 2.8620673592024968,
+      "grad_norm": 0.03571480140089989,
+      "learning_rate": 9.168840849795005e-06,
+      "loss": 0.0041,
+      "step": 15360
+    },
+    {
+      "epoch": 2.863930684306144,
+      "grad_norm": 0.057948265224695206,
+      "learning_rate": 9.044601813889924e-06,
+      "loss": 0.0063,
+      "step": 15370
+    },
+    {
+      "epoch": 2.865794009409792,
+      "grad_norm": 0.07889413088560104,
+      "learning_rate": 8.920362777984844e-06,
+      "loss": 0.0053,
+      "step": 15380
+    },
+    {
+      "epoch": 2.867657334513439,
+      "grad_norm": 0.06906258314847946,
+      "learning_rate": 8.796123742079763e-06,
+      "loss": 0.0066,
+      "step": 15390
+    },
+    {
+      "epoch": 2.869520659617087,
+      "grad_norm": 0.11181332916021347,
+      "learning_rate": 8.67188470617468e-06,
+      "loss": 0.0039,
+      "step": 15400
+    },
+    {
+      "epoch": 2.871383984720734,
+      "grad_norm": 0.049036044627428055,
+      "learning_rate": 8.5476456702696e-06,
+      "loss": 0.0044,
+      "step": 15410
+    },
+    {
+      "epoch": 2.8732473098243814,
+      "grad_norm": 0.07270175218582153,
+      "learning_rate": 8.423406634364519e-06,
+      "loss": 0.0043,
+      "step": 15420
+    },
+    {
+      "epoch": 2.875110634928029,
+      "grad_norm": 0.10137112438678741,
+      "learning_rate": 8.299167598459436e-06,
+      "loss": 0.0052,
+      "step": 15430
+    },
+    {
+      "epoch": 2.8769739600316764,
+      "grad_norm": 0.08990975469350815,
+      "learning_rate": 8.174928562554355e-06,
+      "loss": 0.0056,
+      "step": 15440
+    },
+    {
+      "epoch": 2.878837285135324,
+      "grad_norm": 0.022517943754792213,
+      "learning_rate": 8.050689526649273e-06,
+      "loss": 0.004,
+      "step": 15450
+    },
+    {
+      "epoch": 2.8807006102389714,
+      "grad_norm": 0.020834220573306084,
+      "learning_rate": 7.926450490744192e-06,
+      "loss": 0.0055,
+      "step": 15460
+    },
+    {
+      "epoch": 2.8825639353426187,
+      "grad_norm": 0.02067444659769535,
+      "learning_rate": 7.80221145483911e-06,
+      "loss": 0.005,
+      "step": 15470
+    },
+    {
+      "epoch": 2.8844272604462664,
+      "grad_norm": 0.04585389420390129,
+      "learning_rate": 7.677972418934029e-06,
+      "loss": 0.0044,
+      "step": 15480
+    },
+    {
+      "epoch": 2.8862905855499137,
+      "grad_norm": 0.09295323491096497,
+      "learning_rate": 7.553733383028948e-06,
+      "loss": 0.0066,
+      "step": 15490
+    },
+    {
+      "epoch": 2.8881539106535614,
+      "grad_norm": 0.04389965534210205,
+      "learning_rate": 7.429494347123866e-06,
+      "loss": 0.0047,
+      "step": 15500
+    },
+    {
+      "epoch": 2.8900172357572087,
+      "grad_norm": 0.08827071636915207,
+      "learning_rate": 7.305255311218786e-06,
+      "loss": 0.0091,
+      "step": 15510
+    },
+    {
+      "epoch": 2.891880560860856,
+      "grad_norm": 0.06935074180364609,
+      "learning_rate": 7.181016275313704e-06,
+      "loss": 0.0058,
+      "step": 15520
+    },
+    {
+      "epoch": 2.8937438859645037,
+      "grad_norm": 0.07258989661931992,
+      "learning_rate": 7.056777239408622e-06,
+      "loss": 0.0082,
+      "step": 15530
+    },
+    {
+      "epoch": 2.895607211068151,
+      "grad_norm": 0.07443951815366745,
+      "learning_rate": 6.9325382035035415e-06,
+      "loss": 0.0064,
+      "step": 15540
+    },
+    {
+      "epoch": 2.8974705361717987,
+      "grad_norm": 0.045825716108083725,
+      "learning_rate": 6.80829916759846e-06,
+      "loss": 0.0065,
+      "step": 15550
+    },
+    {
+      "epoch": 2.899333861275446,
+      "grad_norm": 0.062278542667627335,
+      "learning_rate": 6.684060131693379e-06,
+      "loss": 0.0052,
+      "step": 15560
+    },
+    {
+      "epoch": 2.9011971863790933,
+      "grad_norm": 0.07567639648914337,
+      "learning_rate": 6.559821095788297e-06,
+      "loss": 0.0045,
+      "step": 15570
+    },
+    {
+      "epoch": 2.903060511482741,
+      "grad_norm": 0.04492925852537155,
+      "learning_rate": 6.435582059883216e-06,
+      "loss": 0.0043,
+      "step": 15580
+    },
+    {
+      "epoch": 2.9049238365863883,
+      "grad_norm": 0.10229244828224182,
+      "learning_rate": 6.311343023978135e-06,
+      "loss": 0.0069,
+      "step": 15590
+    },
+    {
+      "epoch": 2.906787161690036,
+      "grad_norm": 0.01675078086555004,
+      "learning_rate": 6.187103988073053e-06,
+      "loss": 0.0045,
+      "step": 15600
+    },
+    {
+      "epoch": 2.9086504867936833,
+      "grad_norm": 0.03617624193429947,
+      "learning_rate": 6.062864952167972e-06,
+      "loss": 0.0041,
+      "step": 15610
+    },
+    {
+      "epoch": 2.9105138118973306,
+      "grad_norm": 0.06336949020624161,
+      "learning_rate": 5.93862591626289e-06,
+      "loss": 0.0053,
+      "step": 15620
+    },
+    {
+      "epoch": 2.9123771370009783,
+      "grad_norm": 0.09426115453243256,
+      "learning_rate": 5.814386880357809e-06,
+      "loss": 0.0061,
+      "step": 15630
+    },
+    {
+      "epoch": 2.9142404621046256,
+      "grad_norm": 0.06645146012306213,
+      "learning_rate": 5.690147844452727e-06,
+      "loss": 0.0049,
+      "step": 15640
+    },
+    {
+      "epoch": 2.9161037872082733,
+      "grad_norm": 0.06695625931024551,
+      "learning_rate": 5.565908808547646e-06,
+      "loss": 0.0036,
+      "step": 15650
+    },
+    {
+      "epoch": 2.9179671123119206,
+      "grad_norm": 0.03353876248002052,
+      "learning_rate": 5.441669772642564e-06,
+      "loss": 0.0044,
+      "step": 15660
+    },
+    {
+      "epoch": 2.919830437415568,
+      "grad_norm": 0.05638827010989189,
+      "learning_rate": 5.3174307367374834e-06,
+      "loss": 0.005,
+      "step": 15670
+    },
+    {
+      "epoch": 2.9216937625192156,
+      "grad_norm": 0.03586374223232269,
+      "learning_rate": 5.193191700832402e-06,
+      "loss": 0.0043,
+      "step": 15680
+    },
+    {
+      "epoch": 2.923557087622863,
+      "grad_norm": 0.034597247838974,
+      "learning_rate": 5.068952664927321e-06,
+      "loss": 0.0061,
+      "step": 15690
+    },
+    {
+      "epoch": 2.9254204127265107,
+      "grad_norm": 0.03682232275605202,
+      "learning_rate": 4.944713629022239e-06,
+      "loss": 0.0032,
+      "step": 15700
+    },
+    {
+      "epoch": 2.927283737830158,
+      "grad_norm": 0.05273206904530525,
+      "learning_rate": 4.820474593117158e-06,
+      "loss": 0.0044,
+      "step": 15710
+    },
+    {
+      "epoch": 2.9291470629338052,
+      "grad_norm": 0.043077465146780014,
+      "learning_rate": 4.696235557212076e-06,
+      "loss": 0.0044,
+      "step": 15720
+    },
+    {
+      "epoch": 2.931010388037453,
+      "grad_norm": 0.05343864858150482,
+      "learning_rate": 4.571996521306994e-06,
+      "loss": 0.0091,
+      "step": 15730
+    },
+    {
+      "epoch": 2.9328737131411002,
+      "grad_norm": 0.06297837197780609,
+      "learning_rate": 4.447757485401914e-06,
+      "loss": 0.0048,
+      "step": 15740
+    },
+    {
+      "epoch": 2.934737038244748,
+      "grad_norm": 0.055205777287483215,
+      "learning_rate": 4.323518449496832e-06,
+      "loss": 0.0049,
+      "step": 15750
+    },
+    {
+      "epoch": 2.9366003633483952,
+      "grad_norm": 0.05605635046958923,
+      "learning_rate": 4.199279413591751e-06,
+      "loss": 0.0044,
+      "step": 15760
+    },
+    {
+      "epoch": 2.9384636884520425,
+      "grad_norm": 0.06049589440226555,
+      "learning_rate": 4.0750403776866695e-06,
+      "loss": 0.0036,
+      "step": 15770
+    },
+    {
+      "epoch": 2.9403270135556903,
+      "grad_norm": 0.04346240684390068,
+      "learning_rate": 3.950801341781589e-06,
+      "loss": 0.0037,
+      "step": 15780
+    },
+    {
+      "epoch": 2.9421903386593375,
+      "grad_norm": 0.036925263702869415,
+      "learning_rate": 3.826562305876506e-06,
+      "loss": 0.0067,
+      "step": 15790
+    },
+    {
+      "epoch": 2.9440536637629853,
+      "grad_norm": 0.08391120284795761,
+      "learning_rate": 3.702323269971425e-06,
+      "loss": 0.006,
+      "step": 15800
+    },
+    {
+      "epoch": 2.9459169888666326,
+      "grad_norm": 0.06644539535045624,
+      "learning_rate": 3.5780842340663437e-06,
+      "loss": 0.0049,
+      "step": 15810
+    },
+    {
+      "epoch": 2.94778031397028,
+      "grad_norm": 0.14044621586799622,
+      "learning_rate": 3.4538451981612625e-06,
+      "loss": 0.0078,
+      "step": 15820
+    },
+    {
+      "epoch": 2.9496436390739276,
+      "grad_norm": 0.04912354797124863,
+      "learning_rate": 3.3296061622561813e-06,
+      "loss": 0.005,
+      "step": 15830
+    },
+    {
+      "epoch": 2.951506964177575,
+      "grad_norm": 0.030744953081011772,
+      "learning_rate": 3.2053671263510996e-06,
+      "loss": 0.0047,
+      "step": 15840
+    },
+    {
+      "epoch": 2.9533702892812226,
+      "grad_norm": 0.05932426080107689,
+      "learning_rate": 3.081128090446018e-06,
+      "loss": 0.007,
+      "step": 15850
+    },
+    {
+      "epoch": 2.95523361438487,
+      "grad_norm": 0.09052354097366333,
+      "learning_rate": 2.9568890545409367e-06,
+      "loss": 0.0049,
+      "step": 15860
+    },
+    {
+      "epoch": 2.957096939488517,
+      "grad_norm": 0.03159482032060623,
+      "learning_rate": 2.8326500186358555e-06,
+      "loss": 0.0042,
+      "step": 15870
+    },
+    {
+      "epoch": 2.958960264592165,
+      "grad_norm": 0.06023601070046425,
+      "learning_rate": 2.7084109827307743e-06,
+      "loss": 0.0044,
+      "step": 15880
+    },
+    {
+      "epoch": 2.960823589695812,
+      "grad_norm": 0.02894243411719799,
+      "learning_rate": 2.5841719468256926e-06,
+      "loss": 0.0048,
+      "step": 15890
+    },
+    {
+      "epoch": 2.96268691479946,
+      "grad_norm": 0.06545311957597733,
+      "learning_rate": 2.4599329109206114e-06,
+      "loss": 0.0041,
+      "step": 15900
+    },
+    {
+      "epoch": 2.964550239903107,
+      "grad_norm": 0.07171537727117538,
+      "learning_rate": 2.33569387501553e-06,
+      "loss": 0.0051,
+      "step": 15910
+    },
+    {
+      "epoch": 2.9664135650067545,
+      "grad_norm": 0.04423753172159195,
+      "learning_rate": 2.2114548391104485e-06,
+      "loss": 0.0047,
+      "step": 15920
+    },
+    {
+      "epoch": 2.9682768901104017,
+      "grad_norm": 0.022418806329369545,
+      "learning_rate": 2.0872158032053673e-06,
+      "loss": 0.0064,
+      "step": 15930
+    },
+    {
+      "epoch": 2.9701402152140495,
+      "grad_norm": 0.06295423209667206,
+      "learning_rate": 1.9629767673002857e-06,
+      "loss": 0.0046,
+      "step": 15940
+    },
+    {
+      "epoch": 2.972003540317697,
+      "grad_norm": 0.03230300545692444,
+      "learning_rate": 1.8387377313952046e-06,
+      "loss": 0.0038,
+      "step": 15950
+    },
+    {
+      "epoch": 2.9738668654213445,
+      "grad_norm": 0.07564916461706161,
+      "learning_rate": 1.714498695490123e-06,
+      "loss": 0.0065,
+      "step": 15960
+    },
+    {
+      "epoch": 2.9757301905249918,
+      "grad_norm": 0.037815775722265244,
+      "learning_rate": 1.5902596595850416e-06,
+      "loss": 0.005,
+      "step": 15970
+    },
+    {
+      "epoch": 2.977593515628639,
+      "grad_norm": 0.0978199765086174,
+      "learning_rate": 1.4660206236799603e-06,
+      "loss": 0.0036,
+      "step": 15980
+    },
+    {
+      "epoch": 2.979456840732287,
+      "grad_norm": 0.0537540465593338,
+      "learning_rate": 1.341781587774879e-06,
+      "loss": 0.0056,
+      "step": 15990
+    },
+    {
+      "epoch": 2.981320165835934,
+      "grad_norm": 0.05983177572488785,
+      "learning_rate": 1.2175425518697977e-06,
+      "loss": 0.0048,
+      "step": 16000
+    },
+    {
+      "epoch": 2.983183490939582,
+      "grad_norm": 0.05021652951836586,
+      "learning_rate": 1.0933035159647162e-06,
+      "loss": 0.0054,
+      "step": 16010
+    },
+    {
+      "epoch": 2.985046816043229,
+      "grad_norm": 0.05442306399345398,
+      "learning_rate": 9.690644800596348e-07,
+      "loss": 0.0054,
+      "step": 16020
+    },
+    {
+      "epoch": 2.9869101411468764,
+      "grad_norm": 0.023789361119270325,
+      "learning_rate": 8.448254441545534e-07,
+      "loss": 0.0045,
+      "step": 16030
+    },
+    {
+      "epoch": 2.988773466250524,
+      "grad_norm": 0.02481863461434841,
+      "learning_rate": 7.20586408249472e-07,
+      "loss": 0.0044,
+      "step": 16040
+    },
+    {
+      "epoch": 2.9906367913541714,
+      "grad_norm": 0.03807642310857773,
+      "learning_rate": 5.963473723443906e-07,
+      "loss": 0.0046,
+      "step": 16050
+    },
+    {
+      "epoch": 2.992500116457819,
+      "grad_norm": 0.17297309637069702,
+      "learning_rate": 4.7210833643930925e-07,
+      "loss": 0.0045,
+      "step": 16060
+    },
+    {
+      "epoch": 2.9943634415614664,
+      "grad_norm": 0.08204018324613571,
+      "learning_rate": 3.4786930053422787e-07,
+      "loss": 0.0055,
+      "step": 16070
+    },
+    {
+      "epoch": 2.9962267666651137,
+      "grad_norm": 0.07848156243562698,
+      "learning_rate": 2.2363026462914648e-07,
+      "loss": 0.004,
+      "step": 16080
+    },
+    {
+      "epoch": 2.9980900917687614,
+      "grad_norm": 0.1832299679517746,
+      "learning_rate": 9.93912287240651e-08,
+      "loss": 0.0069,
+      "step": 16090
+    },
+    {
+      "epoch": 2.9995807518516795,
+      "step": 16098,
+      "total_flos": 1.2695660559534653e+18,
+      "train_loss": 0.020097828387903093,
+      "train_runtime": 24374.3923,
+      "train_samples_per_second": 10.569,
+      "train_steps_per_second": 0.66
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 16098,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 10500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2695660559534653e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}