diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,33252 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.998419971559487,
+  "eval_steps": 500,
+  "global_step": 47460,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002106704587349239,
+      "grad_norm": 1.0077738761901855,
+      "learning_rate": 0.00019999997809141818,
+      "loss": 2.2638,
+      "step": 10
+    },
+    {
+      "epoch": 0.004213409174698478,
+      "grad_norm": 0.7950726747512817,
+      "learning_rate": 0.00019999991236568227,
+      "loss": 2.0614,
+      "step": 20
+    },
+    {
+      "epoch": 0.006320113762047717,
+      "grad_norm": 0.8017520308494568,
+      "learning_rate": 0.0001999998028228211,
+      "loss": 2.0796,
+      "step": 30
+    },
+    {
+      "epoch": 0.008426818349396957,
+      "grad_norm": 0.7065390944480896,
+      "learning_rate": 0.00019999964946288266,
+      "loss": 2.0147,
+      "step": 40
+    },
+    {
+      "epoch": 0.010533522936746195,
+      "grad_norm": 0.7359225749969482,
+      "learning_rate": 0.00019999945228593418,
+      "loss": 2.0113,
+      "step": 50
+    },
+    {
+      "epoch": 0.012640227524095434,
+      "grad_norm": 0.7210670113563538,
+      "learning_rate": 0.000199999211292062,
+      "loss": 1.9811,
+      "step": 60
+    },
+    {
+      "epoch": 0.014746932111444672,
+      "grad_norm": 0.6569874286651611,
+      "learning_rate": 0.00019999892648137174,
+      "loss": 1.9632,
+      "step": 70
+    },
+    {
+      "epoch": 0.016853636698793913,
+      "grad_norm": 0.6518391966819763,
+      "learning_rate": 0.0001999985978539882,
+      "loss": 1.9901,
+      "step": 80
+    },
+    {
+      "epoch": 0.01896034128614315,
+      "grad_norm": 0.6628450155258179,
+      "learning_rate": 0.00019999822541005537,
+      "loss": 1.9529,
+      "step": 90
+    },
+    {
+      "epoch": 0.02106704587349239,
+      "grad_norm": 0.6857717037200928,
+      "learning_rate": 0.00019999780914973646,
+      "loss": 1.9697,
+      "step": 100
+    },
+    {
+      "epoch": 0.02317375046084163,
+      "grad_norm": 0.7159650325775146,
+      "learning_rate": 0.00019999734907321385,
+      "loss": 1.9493,
+      "step": 110
+    },
+    {
+      "epoch": 0.025280455048190868,
+      "grad_norm": 0.6563044786453247,
+      "learning_rate": 0.00019999684518068916,
+      "loss": 1.9791,
+      "step": 120
+    },
+    {
+      "epoch": 0.027387159635540108,
+      "grad_norm": 0.6270883083343506,
+      "learning_rate": 0.0001999962974723831,
+      "loss": 1.9267,
+      "step": 130
+    },
+    {
+      "epoch": 0.029493864222889344,
+      "grad_norm": 0.6255055069923401,
+      "learning_rate": 0.00019999570594853575,
+      "loss": 1.9642,
+      "step": 140
+    },
+    {
+      "epoch": 0.03160056881023859,
+      "grad_norm": 0.6271708011627197,
+      "learning_rate": 0.00019999507060940625,
+      "loss": 1.9391,
+      "step": 150
+    },
+    {
+      "epoch": 0.03370727339758783,
+      "grad_norm": 0.641325056552887,
+      "learning_rate": 0.00019999439145527303,
+      "loss": 1.9436,
+      "step": 160
+    },
+    {
+      "epoch": 0.03581397798493706,
+      "grad_norm": 0.6543128490447998,
+      "learning_rate": 0.00019999366848643364,
+      "loss": 1.9596,
+      "step": 170
+    },
+    {
+      "epoch": 0.0379206825722863,
+      "grad_norm": 0.634831964969635,
+      "learning_rate": 0.00019999290170320485,
+      "loss": 1.9341,
+      "step": 180
+    },
+    {
+      "epoch": 0.04002738715963554,
+      "grad_norm": 0.6254514455795288,
+      "learning_rate": 0.0001999920911059227,
+      "loss": 1.9737,
+      "step": 190
+    },
+    {
+      "epoch": 0.04213409174698478,
+      "grad_norm": 0.6236160397529602,
+      "learning_rate": 0.0001999912366949423,
+      "loss": 1.9373,
+      "step": 200
+    },
+    {
+      "epoch": 0.04424079633433402,
+      "grad_norm": 0.6445717215538025,
+      "learning_rate": 0.00019999033847063811,
+      "loss": 1.928,
+      "step": 210
+    },
+    {
+      "epoch": 0.04634750092168326,
+      "grad_norm": 0.6292468905448914,
+      "learning_rate": 0.00019998939643340365,
+      "loss": 1.9223,
+      "step": 220
+    },
+    {
+      "epoch": 0.0484542055090325,
+      "grad_norm": 0.6503478288650513,
+      "learning_rate": 0.00019998841058365167,
+      "loss": 1.9438,
+      "step": 230
+    },
+    {
+      "epoch": 0.050560910096381737,
+      "grad_norm": 0.6267730593681335,
+      "learning_rate": 0.00019998738092181421,
+      "loss": 1.9529,
+      "step": 240
+    },
+    {
+      "epoch": 0.052667614683730976,
+      "grad_norm": 0.6830586791038513,
+      "learning_rate": 0.00019998630744834243,
+      "loss": 1.9406,
+      "step": 250
+    },
+    {
+      "epoch": 0.054774319271080216,
+      "grad_norm": 0.6437792181968689,
+      "learning_rate": 0.0001999851901637066,
+      "loss": 1.985,
+      "step": 260
+    },
+    {
+      "epoch": 0.05688102385842945,
+      "grad_norm": 0.6661207675933838,
+      "learning_rate": 0.00019998402906839643,
+      "loss": 1.8874,
+      "step": 270
+    },
+    {
+      "epoch": 0.05898772844577869,
+      "grad_norm": 0.6596685647964478,
+      "learning_rate": 0.00019998282416292055,
+      "loss": 1.8807,
+      "step": 280
+    },
+    {
+      "epoch": 0.06109443303312793,
+      "grad_norm": 0.6325543522834778,
+      "learning_rate": 0.00019998157544780698,
+      "loss": 1.9462,
+      "step": 290
+    },
+    {
+      "epoch": 0.06320113762047717,
+      "grad_norm": 0.7503716349601746,
+      "learning_rate": 0.00019998028292360286,
+      "loss": 1.9079,
+      "step": 300
+    },
+    {
+      "epoch": 0.06530784220782641,
+      "grad_norm": 0.6907737255096436,
+      "learning_rate": 0.00019997894659087457,
+      "loss": 1.8995,
+      "step": 310
+    },
+    {
+      "epoch": 0.06741454679517565,
+      "grad_norm": 0.6293074488639832,
+      "learning_rate": 0.00019997756645020757,
+      "loss": 1.9193,
+      "step": 320
+    },
+    {
+      "epoch": 0.06952125138252488,
+      "grad_norm": 0.6826409101486206,
+      "learning_rate": 0.0001999761425022067,
+      "loss": 1.9169,
+      "step": 330
+    },
+    {
+      "epoch": 0.07162795596987412,
+      "grad_norm": 0.6254687309265137,
+      "learning_rate": 0.0001999746747474958,
+      "loss": 1.8822,
+      "step": 340
+    },
+    {
+      "epoch": 0.07373466055722336,
+      "grad_norm": 0.6506015658378601,
+      "learning_rate": 0.00019997316318671806,
+      "loss": 1.8803,
+      "step": 350
+    },
+    {
+      "epoch": 0.0758413651445726,
+      "grad_norm": 0.639499306678772,
+      "learning_rate": 0.00019997160782053578,
+      "loss": 1.8657,
+      "step": 360
+    },
+    {
+      "epoch": 0.07794806973192184,
+      "grad_norm": 0.603563666343689,
+      "learning_rate": 0.0001999700086496305,
+      "loss": 1.9138,
+      "step": 370
+    },
+    {
+      "epoch": 0.08005477431927108,
+      "grad_norm": 0.6491252183914185,
+      "learning_rate": 0.0001999683656747029,
+      "loss": 1.945,
+      "step": 380
+    },
+    {
+      "epoch": 0.08216147890662032,
+      "grad_norm": 0.6904931664466858,
+      "learning_rate": 0.00019996667889647288,
+      "loss": 1.9429,
+      "step": 390
+    },
+    {
+      "epoch": 0.08426818349396956,
+      "grad_norm": 0.6672365069389343,
+      "learning_rate": 0.00019996494831567958,
+      "loss": 1.8993,
+      "step": 400
+    },
+    {
+      "epoch": 0.0863748880813188,
+      "grad_norm": 0.635400116443634,
+      "learning_rate": 0.00019996317393308126,
+      "loss": 1.8849,
+      "step": 410
+    },
+    {
+      "epoch": 0.08848159266866804,
+      "grad_norm": 0.6338510513305664,
+      "learning_rate": 0.00019996135574945544,
+      "loss": 1.9232,
+      "step": 420
+    },
+    {
+      "epoch": 0.09058829725601728,
+      "grad_norm": 0.718330979347229,
+      "learning_rate": 0.00019995949376559874,
+      "loss": 1.9022,
+      "step": 430
+    },
+    {
+      "epoch": 0.09269500184336651,
+      "grad_norm": 0.6823340058326721,
+      "learning_rate": 0.00019995758798232704,
+      "loss": 1.9344,
+      "step": 440
+    },
+    {
+      "epoch": 0.09480170643071575,
+      "grad_norm": 0.6402202248573303,
+      "learning_rate": 0.00019995563840047542,
+      "loss": 1.9258,
+      "step": 450
+    },
+    {
+      "epoch": 0.096908411018065,
+      "grad_norm": 0.5626136660575867,
+      "learning_rate": 0.00019995364502089813,
+      "loss": 1.8337,
+      "step": 460
+    },
+    {
+      "epoch": 0.09901511560541423,
+      "grad_norm": 0.6894583702087402,
+      "learning_rate": 0.00019995160784446863,
+      "loss": 1.9246,
+      "step": 470
+    },
+    {
+      "epoch": 0.10112182019276347,
+      "grad_norm": 0.6594771146774292,
+      "learning_rate": 0.00019994952687207954,
+      "loss": 1.9241,
+      "step": 480
+    },
+    {
+      "epoch": 0.10322852478011271,
+      "grad_norm": 0.624462902545929,
+      "learning_rate": 0.00019994740210464268,
+      "loss": 1.9486,
+      "step": 490
+    },
+    {
+      "epoch": 0.10533522936746195,
+      "grad_norm": 0.6428165435791016,
+      "learning_rate": 0.00019994523354308904,
+      "loss": 1.8711,
+      "step": 500
+    },
+    {
+      "epoch": 0.10744193395481119,
+      "grad_norm": 0.626460075378418,
+      "learning_rate": 0.00019994302118836883,
+      "loss": 1.8453,
+      "step": 510
+    },
+    {
+      "epoch": 0.10954863854216043,
+      "grad_norm": 0.5991859436035156,
+      "learning_rate": 0.00019994076504145148,
+      "loss": 1.8691,
+      "step": 520
+    },
+    {
+      "epoch": 0.11165534312950967,
+      "grad_norm": 0.6435895562171936,
+      "learning_rate": 0.00019993846510332552,
+      "loss": 1.9216,
+      "step": 530
+    },
+    {
+      "epoch": 0.1137620477168589,
+      "grad_norm": 0.6111456751823425,
+      "learning_rate": 0.00019993612137499876,
+      "loss": 1.846,
+      "step": 540
+    },
+    {
+      "epoch": 0.11586875230420814,
+      "grad_norm": 0.6688899993896484,
+      "learning_rate": 0.0001999337338574981,
+      "loss": 1.8676,
+      "step": 550
+    },
+    {
+      "epoch": 0.11797545689155738,
+      "grad_norm": 0.6680476069450378,
+      "learning_rate": 0.00019993130255186977,
+      "loss": 1.8763,
+      "step": 560
+    },
+    {
+      "epoch": 0.12008216147890662,
+      "grad_norm": 0.6650174856185913,
+      "learning_rate": 0.00019992882745917902,
+      "loss": 1.8757,
+      "step": 570
+    },
+    {
+      "epoch": 0.12218886606625586,
+      "grad_norm": 0.6582191586494446,
+      "learning_rate": 0.0001999263085805104,
+      "loss": 1.921,
+      "step": 580
+    },
+    {
+      "epoch": 0.1242955706536051,
+      "grad_norm": 0.6733909249305725,
+      "learning_rate": 0.00019992374591696761,
+      "loss": 1.9209,
+      "step": 590
+    },
+    {
+      "epoch": 0.12640227524095435,
+      "grad_norm": 0.6196283102035522,
+      "learning_rate": 0.00019992113946967353,
+      "loss": 1.8838,
+      "step": 600
+    },
+    {
+      "epoch": 0.1285089798283036,
+      "grad_norm": 0.6418552398681641,
+      "learning_rate": 0.00019991848923977022,
+      "loss": 1.9044,
+      "step": 610
+    },
+    {
+      "epoch": 0.13061568441565283,
+      "grad_norm": 0.6484676599502563,
+      "learning_rate": 0.00019991579522841892,
+      "loss": 1.8783,
+      "step": 620
+    },
+    {
+      "epoch": 0.13272238900300207,
+      "grad_norm": 0.6386022567749023,
+      "learning_rate": 0.00019991305743680013,
+      "loss": 1.9412,
+      "step": 630
+    },
+    {
+      "epoch": 0.1348290935903513,
+      "grad_norm": 0.7087893486022949,
+      "learning_rate": 0.00019991027586611343,
+      "loss": 1.955,
+      "step": 640
+    },
+    {
+      "epoch": 0.13693579817770052,
+      "grad_norm": 0.6675140857696533,
+      "learning_rate": 0.00019990745051757765,
+      "loss": 1.9606,
+      "step": 650
+    },
+    {
+      "epoch": 0.13904250276504976,
+      "grad_norm": 0.6183488965034485,
+      "learning_rate": 0.00019990458139243077,
+      "loss": 1.8419,
+      "step": 660
+    },
+    {
+      "epoch": 0.141149207352399,
+      "grad_norm": 0.6724327206611633,
+      "learning_rate": 0.00019990166849192994,
+      "loss": 1.9343,
+      "step": 670
+    },
+    {
+      "epoch": 0.14325591193974824,
+      "grad_norm": 0.650532066822052,
+      "learning_rate": 0.00019989871181735148,
+      "loss": 1.8345,
+      "step": 680
+    },
+    {
+      "epoch": 0.14536261652709748,
+      "grad_norm": 0.608699381351471,
+      "learning_rate": 0.000199895711369991,
+      "loss": 1.9123,
+      "step": 690
+    },
+    {
+      "epoch": 0.14746932111444672,
+      "grad_norm": 0.6671069264411926,
+      "learning_rate": 0.00019989266715116316,
+      "loss": 1.8696,
+      "step": 700
+    },
+    {
+      "epoch": 0.14957602570179596,
+      "grad_norm": 0.619253396987915,
+      "learning_rate": 0.0001998895791622019,
+      "loss": 1.9529,
+      "step": 710
+    },
+    {
+      "epoch": 0.1516827302891452,
+      "grad_norm": 0.6708328127861023,
+      "learning_rate": 0.00019988644740446022,
+      "loss": 1.9158,
+      "step": 720
+    },
+    {
+      "epoch": 0.15378943487649444,
+      "grad_norm": 0.6542178392410278,
+      "learning_rate": 0.0001998832718793104,
+      "loss": 1.8742,
+      "step": 730
+    },
+    {
+      "epoch": 0.15589613946384367,
+      "grad_norm": 0.6505009531974792,
+      "learning_rate": 0.00019988005258814387,
+      "loss": 1.8314,
+      "step": 740
+    },
+    {
+      "epoch": 0.15800284405119291,
+      "grad_norm": 0.6684021353721619,
+      "learning_rate": 0.00019987678953237127,
+      "loss": 1.8651,
+      "step": 750
+    },
+    {
+      "epoch": 0.16010954863854215,
+      "grad_norm": 0.6631053686141968,
+      "learning_rate": 0.00019987348271342228,
+      "loss": 1.8857,
+      "step": 760
+    },
+    {
+      "epoch": 0.1622162532258914,
+      "grad_norm": 0.7295246124267578,
+      "learning_rate": 0.00019987013213274593,
+      "loss": 1.9163,
+      "step": 770
+    },
+    {
+      "epoch": 0.16432295781324063,
+      "grad_norm": 0.6868222951889038,
+      "learning_rate": 0.00019986673779181033,
+      "loss": 1.8221,
+      "step": 780
+    },
+    {
+      "epoch": 0.16642966240058987,
+      "grad_norm": 0.6467426419258118,
+      "learning_rate": 0.00019986329969210278,
+      "loss": 1.8954,
+      "step": 790
+    },
+    {
+      "epoch": 0.1685363669879391,
+      "grad_norm": 0.6961290240287781,
+      "learning_rate": 0.00019985981783512976,
+      "loss": 1.9336,
+      "step": 800
+    },
+    {
+      "epoch": 0.17064307157528835,
+      "grad_norm": 0.6626701951026917,
+      "learning_rate": 0.00019985629222241694,
+      "loss": 1.8577,
+      "step": 810
+    },
+    {
+      "epoch": 0.1727497761626376,
+      "grad_norm": 0.6900150775909424,
+      "learning_rate": 0.0001998527228555091,
+      "loss": 1.8339,
+      "step": 820
+    },
+    {
+      "epoch": 0.17485648074998683,
+      "grad_norm": 0.6740383505821228,
+      "learning_rate": 0.00019984910973597027,
+      "loss": 1.8174,
+      "step": 830
+    },
+    {
+      "epoch": 0.17696318533733607,
+      "grad_norm": 0.6543059945106506,
+      "learning_rate": 0.0001998454528653836,
+      "loss": 1.8637,
+      "step": 840
+    },
+    {
+      "epoch": 0.1790698899246853,
+      "grad_norm": 0.652992308139801,
+      "learning_rate": 0.00019984175224535146,
+      "loss": 1.888,
+      "step": 850
+    },
+    {
+      "epoch": 0.18117659451203455,
+      "grad_norm": 0.6753103137016296,
+      "learning_rate": 0.00019983800787749532,
+      "loss": 1.9282,
+      "step": 860
+    },
+    {
+      "epoch": 0.1832832990993838,
+      "grad_norm": 0.6452760696411133,
+      "learning_rate": 0.00019983421976345586,
+      "loss": 1.846,
+      "step": 870
+    },
+    {
+      "epoch": 0.18539000368673303,
+      "grad_norm": 0.7148582339286804,
+      "learning_rate": 0.00019983038790489296,
+      "loss": 1.8172,
+      "step": 880
+    },
+    {
+      "epoch": 0.18749670827408227,
+      "grad_norm": 0.644650399684906,
+      "learning_rate": 0.00019982651230348556,
+      "loss": 1.8717,
+      "step": 890
+    },
+    {
+      "epoch": 0.1896034128614315,
+      "grad_norm": 0.6961220502853394,
+      "learning_rate": 0.0001998225929609319,
+      "loss": 1.85,
+      "step": 900
+    },
+    {
+      "epoch": 0.19171011744878075,
+      "grad_norm": 0.5916398763656616,
+      "learning_rate": 0.00019981862987894934,
+      "loss": 1.8181,
+      "step": 910
+    },
+    {
+      "epoch": 0.19381682203613,
+      "grad_norm": 0.697644829750061,
+      "learning_rate": 0.0001998146230592743,
+      "loss": 1.8774,
+      "step": 920
+    },
+    {
+      "epoch": 0.19592352662347923,
+      "grad_norm": 0.6732763648033142,
+      "learning_rate": 0.00019981057250366253,
+      "loss": 1.8786,
+      "step": 930
+    },
+    {
+      "epoch": 0.19803023121082847,
+      "grad_norm": 0.6978752613067627,
+      "learning_rate": 0.00019980647821388886,
+      "loss": 1.8784,
+      "step": 940
+    },
+    {
+      "epoch": 0.2001369357981777,
+      "grad_norm": 0.7141478657722473,
+      "learning_rate": 0.00019980234019174729,
+      "loss": 1.847,
+      "step": 950
+    },
+    {
+      "epoch": 0.20224364038552695,
+      "grad_norm": 0.6787569522857666,
+      "learning_rate": 0.00019979815843905097,
+      "loss": 1.8903,
+      "step": 960
+    },
+    {
+      "epoch": 0.20435034497287619,
+      "grad_norm": 0.6798018217086792,
+      "learning_rate": 0.0001997939329576322,
+      "loss": 1.9078,
+      "step": 970
+    },
+    {
+      "epoch": 0.20645704956022543,
+      "grad_norm": 0.6333872675895691,
+      "learning_rate": 0.00019978966374934254,
+      "loss": 1.8625,
+      "step": 980
+    },
+    {
+      "epoch": 0.20856375414757466,
+      "grad_norm": 0.706005871295929,
+      "learning_rate": 0.0001997853508160526,
+      "loss": 1.8134,
+      "step": 990
+    },
+    {
+      "epoch": 0.2106704587349239,
+      "grad_norm": 0.6944781541824341,
+      "learning_rate": 0.00019978099415965213,
+      "loss": 1.8938,
+      "step": 1000
+    },
+    {
+      "epoch": 0.21277716332227314,
+      "grad_norm": 0.7021796107292175,
+      "learning_rate": 0.00019977659378205015,
+      "loss": 1.8474,
+      "step": 1010
+    },
+    {
+      "epoch": 0.21488386790962238,
+      "grad_norm": 0.6823903918266296,
+      "learning_rate": 0.0001997721496851748,
+      "loss": 1.883,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21699057249697162,
+      "grad_norm": 0.6894837617874146,
+      "learning_rate": 0.0001997676618709733,
+      "loss": 1.8624,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21909727708432086,
+      "grad_norm": 0.7047960758209229,
+      "learning_rate": 0.00019976313034141213,
+      "loss": 1.8421,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2212039816716701,
+      "grad_norm": 0.6393168568611145,
+      "learning_rate": 0.00019975855509847686,
+      "loss": 1.8069,
+      "step": 1050
+    },
+    {
+      "epoch": 0.22331068625901934,
+      "grad_norm": 0.6958315372467041,
+      "learning_rate": 0.00019975393614417224,
+      "loss": 1.8484,
+      "step": 1060
+    },
+    {
+      "epoch": 0.22541739084636858,
+      "grad_norm": 0.716953694820404,
+      "learning_rate": 0.00019974927348052215,
+      "loss": 1.8619,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2275240954337178,
+      "grad_norm": 0.7667638659477234,
+      "learning_rate": 0.00019974456710956964,
+      "loss": 1.8564,
+      "step": 1080
+    },
+    {
+      "epoch": 0.22963080002106703,
+      "grad_norm": 0.6643611192703247,
+      "learning_rate": 0.0001997398170333769,
+      "loss": 1.8949,
+      "step": 1090
+    },
+    {
+      "epoch": 0.23173750460841627,
+      "grad_norm": 0.7317732572555542,
+      "learning_rate": 0.00019973502325402532,
+      "loss": 1.8381,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2338442091957655,
+      "grad_norm": 0.7321386933326721,
+      "learning_rate": 0.00019973018577361536,
+      "loss": 1.8545,
+      "step": 1110
+    },
+    {
+      "epoch": 0.23595091378311475,
+      "grad_norm": 0.6243638396263123,
+      "learning_rate": 0.00019972530459426663,
+      "loss": 1.8488,
+      "step": 1120
+    },
+    {
+      "epoch": 0.238057618370464,
+      "grad_norm": 0.6985125541687012,
+      "learning_rate": 0.00019972037971811802,
+      "loss": 1.8695,
+      "step": 1130
+    },
+    {
+      "epoch": 0.24016432295781323,
+      "grad_norm": 0.6656931042671204,
+      "learning_rate": 0.00019971541114732741,
+      "loss": 1.8663,
+      "step": 1140
+    },
+    {
+      "epoch": 0.24227102754516247,
+      "grad_norm": 0.6580804586410522,
+      "learning_rate": 0.00019971039888407187,
+      "loss": 1.8632,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2443777321325117,
+      "grad_norm": 0.6782211661338806,
+      "learning_rate": 0.0001997053429305477,
+      "loss": 1.876,
+      "step": 1160
+    },
+    {
+      "epoch": 0.24648443671986095,
+      "grad_norm": 0.6574906706809998,
+      "learning_rate": 0.00019970024328897022,
+      "loss": 1.8532,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2485911413072102,
+      "grad_norm": 0.7327693104743958,
+      "learning_rate": 0.00019969509996157396,
+      "loss": 1.8929,
+      "step": 1180
+    },
+    {
+      "epoch": 0.25069784589455946,
+      "grad_norm": 0.7190077900886536,
+      "learning_rate": 0.0001996899129506126,
+      "loss": 1.9107,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2528045504819087,
+      "grad_norm": 0.7218247652053833,
+      "learning_rate": 0.0001996846822583589,
+      "loss": 1.9151,
+      "step": 1200
+    },
+    {
+      "epoch": 0.25491125506925794,
+      "grad_norm": 0.7065020799636841,
+      "learning_rate": 0.0001996794078871048,
+      "loss": 1.8707,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2570179596566072,
+      "grad_norm": 0.688412606716156,
+      "learning_rate": 0.00019967408983916145,
+      "loss": 1.8346,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2591246642439564,
+      "grad_norm": 0.6871249079704285,
+      "learning_rate": 0.000199668728116859,
+      "loss": 1.825,
+      "step": 1230
+    },
+    {
+      "epoch": 0.26123136883130565,
+      "grad_norm": 0.6805330514907837,
+      "learning_rate": 0.00019966332272254684,
+      "loss": 1.8608,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2633380734186549,
+      "grad_norm": 0.6569370627403259,
+      "learning_rate": 0.00019965787365859344,
+      "loss": 1.8489,
+      "step": 1250
+    },
+    {
+      "epoch": 0.26544477800600413,
+      "grad_norm": 0.6645459532737732,
+      "learning_rate": 0.00019965238092738643,
+      "loss": 1.8218,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2675514825933534,
+      "grad_norm": 0.6701253056526184,
+      "learning_rate": 0.0001996468445313326,
+      "loss": 1.8862,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2696581871807026,
+      "grad_norm": 0.7218180298805237,
+      "learning_rate": 0.00019964126447285778,
+      "loss": 1.9143,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2717648917680518,
+      "grad_norm": 0.8006751537322998,
+      "learning_rate": 0.00019963564075440703,
+      "loss": 1.8406,
+      "step": 1290
+    },
+    {
+      "epoch": 0.27387159635540104,
+      "grad_norm": 0.7272390127182007,
+      "learning_rate": 0.00019962997337844452,
+      "loss": 1.8889,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2759783009427503,
+      "grad_norm": 0.6612052321434021,
+      "learning_rate": 0.0001996242623474535,
+      "loss": 1.8533,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2780850055300995,
+      "grad_norm": 0.7371575236320496,
+      "learning_rate": 0.0001996185076639364,
+      "loss": 1.8072,
+      "step": 1320
+    },
+    {
+      "epoch": 0.28019171011744876,
+      "grad_norm": 0.6426770687103271,
+      "learning_rate": 0.00019961270933041477,
+      "loss": 1.8731,
+      "step": 1330
+    },
+    {
+      "epoch": 0.282298414704798,
+      "grad_norm": 0.6850872039794922,
+      "learning_rate": 0.00019960686734942922,
+      "loss": 1.9205,
+      "step": 1340
+    },
+    {
+      "epoch": 0.28440511929214723,
+      "grad_norm": 0.7293111681938171,
+      "learning_rate": 0.00019960098172353962,
+      "loss": 1.8028,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2865118238794965,
+      "grad_norm": 0.6498401761054993,
+      "learning_rate": 0.0001995950524553248,
+      "loss": 1.7996,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2886185284668457,
+      "grad_norm": 0.7127597332000732,
+      "learning_rate": 0.00019958907954738288,
+      "loss": 1.8245,
+      "step": 1370
+    },
+    {
+      "epoch": 0.29072523305419495,
+      "grad_norm": 0.6359906196594238,
+      "learning_rate": 0.00019958306300233098,
+      "loss": 1.8264,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2928319376415442,
+      "grad_norm": 0.7264698147773743,
+      "learning_rate": 0.00019957700282280537,
+      "loss": 1.8531,
+      "step": 1390
+    },
+    {
+      "epoch": 0.29493864222889343,
+      "grad_norm": 0.6813586354255676,
+      "learning_rate": 0.00019957089901146148,
+      "loss": 1.8633,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2970453468162427,
+      "grad_norm": 0.6944628357887268,
+      "learning_rate": 0.00019956475157097378,
+      "loss": 1.8383,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2991520514035919,
+      "grad_norm": 0.6405615210533142,
+      "learning_rate": 0.00019955856050403594,
+      "loss": 1.8386,
+      "step": 1420
+    },
+    {
+      "epoch": 0.30125875599094115,
+      "grad_norm": 0.6815128326416016,
+      "learning_rate": 0.0001995523258133607,
+      "loss": 1.92,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3033654605782904,
+      "grad_norm": 0.6566641330718994,
+      "learning_rate": 0.00019954604750167993,
+      "loss": 1.898,
+      "step": 1440
+    },
+    {
+      "epoch": 0.30547216516563963,
+      "grad_norm": 0.7111663222312927,
+      "learning_rate": 0.0001995397255717446,
+      "loss": 1.8138,
+      "step": 1450
+    },
+    {
+      "epoch": 0.30757886975298887,
+      "grad_norm": 0.7194084525108337,
+      "learning_rate": 0.0001995333600263248,
+      "loss": 1.8619,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3096855743403381,
+      "grad_norm": 0.7005071640014648,
+      "learning_rate": 0.00019952695086820975,
+      "loss": 1.8261,
+      "step": 1470
+    },
+    {
+      "epoch": 0.31179227892768735,
+      "grad_norm": 0.7344840168952942,
+      "learning_rate": 0.00019952049810020771,
+      "loss": 1.8191,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3138989835150366,
+      "grad_norm": 0.692154049873352,
+      "learning_rate": 0.00019951400172514618,
+      "loss": 1.8496,
+      "step": 1490
+    },
+    {
+      "epoch": 0.31600568810238583,
+      "grad_norm": 0.7312874794006348,
+      "learning_rate": 0.00019950746174587163,
+      "loss": 1.8621,
+      "step": 1500
+    },
+    {
+      "epoch": 0.31811239268973507,
+      "grad_norm": 0.7678685188293457,
+      "learning_rate": 0.0001995008781652497,
+      "loss": 1.8418,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3202190972770843,
+      "grad_norm": 0.7713690400123596,
+      "learning_rate": 0.00019949425098616513,
+      "loss": 1.853,
+      "step": 1520
+    },
+    {
+      "epoch": 0.32232580186443355,
+      "grad_norm": 0.7388377785682678,
+      "learning_rate": 0.0001994875802115218,
+      "loss": 1.8708,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3244325064517828,
+      "grad_norm": 0.6781003475189209,
+      "learning_rate": 0.00019948086584424256,
+      "loss": 1.8333,
+      "step": 1540
+    },
+    {
+      "epoch": 0.326539211039132,
+      "grad_norm": 0.7277612686157227,
+      "learning_rate": 0.00019947410788726956,
+      "loss": 1.7997,
+      "step": 1550
+    },
+    {
+      "epoch": 0.32864591562648127,
+      "grad_norm": 0.7328895926475525,
+      "learning_rate": 0.0001994673063435639,
+      "loss": 1.8654,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3307526202138305,
+      "grad_norm": 0.779045820236206,
+      "learning_rate": 0.00019946046121610583,
+      "loss": 1.8928,
+      "step": 1570
+    },
+    {
+      "epoch": 0.33285932480117975,
+      "grad_norm": 0.7018063068389893,
+      "learning_rate": 0.00019945357250789468,
+      "loss": 1.8213,
+      "step": 1580
+    },
+    {
+      "epoch": 0.334966029388529,
+      "grad_norm": 0.6804325580596924,
+      "learning_rate": 0.00019944664022194885,
+      "loss": 1.8256,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3370727339758782,
+      "grad_norm": 0.6910491585731506,
+      "learning_rate": 0.00019943966436130597,
+      "loss": 1.8203,
+      "step": 1600
+    },
+    {
+      "epoch": 0.33917943856322746,
+      "grad_norm": 0.6203540563583374,
+      "learning_rate": 0.00019943264492902258,
+      "loss": 1.8504,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3412861431505767,
+      "grad_norm": 0.6889249682426453,
+      "learning_rate": 0.0001994255819281744,
+      "loss": 1.8939,
+      "step": 1620
+    },
+    {
+      "epoch": 0.34339284773792594,
+      "grad_norm": 0.7516829967498779,
+      "learning_rate": 0.00019941847536185633,
+      "loss": 1.8601,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3454995523252752,
+      "grad_norm": 0.6868574619293213,
+      "learning_rate": 0.00019941132523318216,
+      "loss": 1.8381,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3476062569126244,
+      "grad_norm": 0.7086262106895447,
+      "learning_rate": 0.0001994041315452849,
+      "loss": 1.8145,
+      "step": 1650
+    },
+    {
+      "epoch": 0.34971296149997366,
+      "grad_norm": 0.7216698527336121,
+      "learning_rate": 0.00019939689430131666,
+      "loss": 1.8728,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3518196660873229,
+      "grad_norm": 0.6712501645088196,
+      "learning_rate": 0.00019938961350444854,
+      "loss": 1.8133,
+      "step": 1670
+    },
+    {
+      "epoch": 0.35392637067467214,
+      "grad_norm": 0.64311683177948,
+      "learning_rate": 0.0001993822891578708,
+      "loss": 1.8468,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3560330752620214,
+      "grad_norm": 0.7252877354621887,
+      "learning_rate": 0.00019937492126479277,
+      "loss": 1.8592,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3581397798493706,
+      "grad_norm": 0.6842496991157532,
+      "learning_rate": 0.00019936750982844283,
+      "loss": 1.8267,
+      "step": 1700
+    },
+    {
+      "epoch": 0.36024648443671986,
+      "grad_norm": 0.7731478214263916,
+      "learning_rate": 0.00019936005485206851,
+      "loss": 1.8334,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3623531890240691,
+      "grad_norm": 2.383643865585327,
+      "learning_rate": 0.00019935255633893632,
+      "loss": 1.7964,
+      "step": 1720
+    },
+    {
+      "epoch": 0.36445989361141834,
+      "grad_norm": 0.7606905698776245,
+      "learning_rate": 0.0001993450142923319,
+      "loss": 1.7642,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3665665981987676,
+      "grad_norm": 0.748777449131012,
+      "learning_rate": 0.00019933742871556,
+      "loss": 1.8177,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3686733027861168,
+      "grad_norm": 0.6804649829864502,
+      "learning_rate": 0.00019932979961194435,
+      "loss": 1.7606,
+      "step": 1750
+    },
+    {
+      "epoch": 0.37078000737346606,
+      "grad_norm": 0.6883798837661743,
+      "learning_rate": 0.00019932212698482786,
+      "loss": 1.8529,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3728867119608153,
+      "grad_norm": 0.6991207003593445,
+      "learning_rate": 0.00019931441083757245,
+      "loss": 1.842,
+      "step": 1770
+    },
+    {
+      "epoch": 0.37499341654816454,
+      "grad_norm": 0.8275269269943237,
+      "learning_rate": 0.00019930665117355906,
+      "loss": 1.7748,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3771001211355138,
+      "grad_norm": 0.740403950214386,
+      "learning_rate": 0.0001992988479961878,
+      "loss": 1.8512,
+      "step": 1790
+    },
+    {
+      "epoch": 0.379206825722863,
+      "grad_norm": 0.7021561861038208,
+      "learning_rate": 0.00019929100130887782,
+      "loss": 1.7875,
+      "step": 1800
+    },
+    {
+      "epoch": 0.38131353031021226,
+      "grad_norm": 0.6865299344062805,
+      "learning_rate": 0.00019928311111506726,
+      "loss": 1.7992,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3834202348975615,
+      "grad_norm": 0.7252370119094849,
+      "learning_rate": 0.00019927517741821343,
+      "loss": 1.8761,
+      "step": 1820
+    },
+    {
+      "epoch": 0.38552693948491074,
+      "grad_norm": 0.7330300211906433,
+      "learning_rate": 0.0001992672002217926,
+      "loss": 1.8956,
+      "step": 1830
+    },
+    {
+      "epoch": 0.38763364407226,
+      "grad_norm": 0.6942179203033447,
+      "learning_rate": 0.00019925917952930022,
+      "loss": 1.8465,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3897403486596092,
+      "grad_norm": 0.7793520092964172,
+      "learning_rate": 0.00019925111534425068,
+      "loss": 1.7717,
+      "step": 1850
+    },
+    {
+      "epoch": 0.39184705324695845,
+      "grad_norm": 0.7142687439918518,
+      "learning_rate": 0.0001992430076701775,
+      "loss": 1.7862,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3939537578343077,
+      "grad_norm": 0.6944627165794373,
+      "learning_rate": 0.00019923485651063318,
+      "loss": 1.8167,
+      "step": 1870
+    },
+    {
+      "epoch": 0.39606046242165693,
+      "grad_norm": 0.7327067852020264,
+      "learning_rate": 0.0001992266618691894,
+      "loss": 1.8431,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3981671670090062,
+      "grad_norm": 0.7131524682044983,
+      "learning_rate": 0.0001992184237494368,
+      "loss": 1.8566,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4002738715963554,
+      "grad_norm": 0.7777301073074341,
+      "learning_rate": 0.00019921014215498506,
+      "loss": 1.8612,
+      "step": 1900
+    },
+    {
+      "epoch": 0.40238057618370465,
+      "grad_norm": 0.684638500213623,
+      "learning_rate": 0.00019920181708946296,
+      "loss": 1.8838,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4044872807710539,
+      "grad_norm": 0.7215536236763,
+      "learning_rate": 0.00019919344855651833,
+      "loss": 1.8362,
+      "step": 1920
+    },
+    {
+      "epoch": 0.40659398535840313,
+      "grad_norm": 0.7496854066848755,
+      "learning_rate": 0.00019918503655981797,
+      "loss": 1.8092,
+      "step": 1930
+    },
+    {
+      "epoch": 0.40870068994575237,
+      "grad_norm": 0.7422535419464111,
+      "learning_rate": 0.00019917658110304783,
+      "loss": 1.8325,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4108073945331016,
+      "grad_norm": 0.7733063101768494,
+      "learning_rate": 0.0001991680821899128,
+      "loss": 1.7758,
+      "step": 1950
+    },
+    {
+      "epoch": 0.41291409912045085,
+      "grad_norm": 0.662631094455719,
+      "learning_rate": 0.0001991595398241369,
+      "loss": 1.7993,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4150208037078001,
+      "grad_norm": 0.7115535140037537,
+      "learning_rate": 0.00019915095400946319,
+      "loss": 1.8554,
+      "step": 1970
+    },
+    {
+      "epoch": 0.41712750829514933,
+      "grad_norm": 0.6702825427055359,
+      "learning_rate": 0.00019914232474965365,
+      "loss": 1.808,
+      "step": 1980
+    },
+    {
+      "epoch": 0.41923421288249857,
+      "grad_norm": 0.6962928771972656,
+      "learning_rate": 0.00019913365204848939,
+      "loss": 1.7712,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4213409174698478,
+      "grad_norm": 0.7009411454200745,
+      "learning_rate": 0.0001991249359097706,
+      "loss": 1.8069,
+      "step": 2000
+    },
+    {
+      "epoch": 0.42344762205719705,
+      "grad_norm": 0.7244513630867004,
+      "learning_rate": 0.00019911617633731638,
+      "loss": 1.8486,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4255543266445463,
+      "grad_norm": 0.7286465764045715,
+      "learning_rate": 0.00019910737333496498,
+      "loss": 1.7949,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4276610312318955,
+      "grad_norm": 0.686680018901825,
+      "learning_rate": 0.00019909852690657359,
+      "loss": 1.811,
+      "step": 2030
+    },
+    {
+      "epoch": 0.42976773581924477,
+      "grad_norm": 0.7796157002449036,
+      "learning_rate": 0.00019908963705601846,
+      "loss": 1.8168,
+      "step": 2040
+    },
+    {
+      "epoch": 0.431874440406594,
+      "grad_norm": 0.6823425889015198,
+      "learning_rate": 0.00019908070378719492,
+      "loss": 1.7983,
+      "step": 2050
+    },
+    {
+      "epoch": 0.43398114499394325,
+      "grad_norm": 0.7402235269546509,
+      "learning_rate": 0.00019907172710401723,
+      "loss": 1.8272,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4360878495812925,
+      "grad_norm": 0.7159191370010376,
+      "learning_rate": 0.0001990627070104187,
+      "loss": 1.8816,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4381945541686417,
+      "grad_norm": 0.7955963015556335,
+      "learning_rate": 0.00019905364351035173,
+      "loss": 1.8434,
+      "step": 2080
+    },
+    {
+      "epoch": 0.44030125875599097,
+      "grad_norm": 0.6838492751121521,
+      "learning_rate": 0.0001990445366077877,
+      "loss": 1.8187,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4424079633433402,
+      "grad_norm": 0.7640077471733093,
+      "learning_rate": 0.0001990353863067169,
+      "loss": 1.8395,
+      "step": 2100
+    },
+    {
+      "epoch": 0.44451466793068944,
+      "grad_norm": 0.8374277353286743,
+      "learning_rate": 0.0001990261926111488,
+      "loss": 1.813,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4466213725180387,
+      "grad_norm": 0.7484899759292603,
+      "learning_rate": 0.00019901695552511183,
+      "loss": 1.8341,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4487280771053879,
+      "grad_norm": 0.7072715759277344,
+      "learning_rate": 0.0001990076750526534,
+      "loss": 1.8297,
+      "step": 2130
+    },
+    {
+      "epoch": 0.45083478169273716,
+      "grad_norm": 0.6923781633377075,
+      "learning_rate": 0.00019899835119783997,
+      "loss": 1.8126,
+      "step": 2140
+    },
+    {
+      "epoch": 0.45294148628008635,
+      "grad_norm": 0.8046467304229736,
+      "learning_rate": 0.00019898898396475694,
+      "loss": 1.8306,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4550481908674356,
+      "grad_norm": 0.7048581838607788,
+      "learning_rate": 0.00019897957335750878,
+      "loss": 1.8026,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4571548954547848,
+      "grad_norm": 0.6879782676696777,
+      "learning_rate": 0.000198970119380219,
+      "loss": 1.8211,
+      "step": 2170
+    },
+    {
+      "epoch": 0.45926160004213407,
+      "grad_norm": 0.7135127782821655,
+      "learning_rate": 0.00019896062203703002,
+      "loss": 1.8329,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4613683046294833,
+      "grad_norm": 0.7031205892562866,
+      "learning_rate": 0.00019895108133210335,
+      "loss": 1.7945,
+      "step": 2190
+    },
+    {
+      "epoch": 0.46347500921683255,
+      "grad_norm": 0.7131096720695496,
+      "learning_rate": 0.00019894149726961937,
+      "loss": 1.8469,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4655817138041818,
+      "grad_norm": 0.7171880006790161,
+      "learning_rate": 0.00019893186985377761,
+      "loss": 1.8642,
+      "step": 2210
+    },
+    {
+      "epoch": 0.467688418391531,
+      "grad_norm": 0.8296772241592407,
+      "learning_rate": 0.00019892219908879653,
+      "loss": 1.8306,
+      "step": 2220
+    },
+    {
+      "epoch": 0.46979512297888026,
+      "grad_norm": 0.7418431639671326,
+      "learning_rate": 0.0001989124849789136,
+      "loss": 1.8617,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4719018275662295,
+      "grad_norm": 0.6901628971099854,
+      "learning_rate": 0.00019890272752838518,
+      "loss": 1.7787,
+      "step": 2240
+    },
+    {
+      "epoch": 0.47400853215357874,
+      "grad_norm": 0.7321175336837769,
+      "learning_rate": 0.00019889292674148682,
+      "loss": 1.7973,
+      "step": 2250
+    },
+    {
+      "epoch": 0.476115236740928,
+      "grad_norm": 0.7562801837921143,
+      "learning_rate": 0.00019888308262251285,
+      "loss": 1.908,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4782219413282772,
+      "grad_norm": 0.7009602785110474,
+      "learning_rate": 0.0001988731951757768,
+      "loss": 1.8732,
+      "step": 2270
+    },
+    {
+      "epoch": 0.48032864591562646,
+      "grad_norm": 0.7295694947242737,
+      "learning_rate": 0.00019886326440561093,
+      "loss": 1.875,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4824353505029757,
+      "grad_norm": 0.7495467662811279,
+      "learning_rate": 0.0001988532903163667,
+      "loss": 1.8208,
+      "step": 2290
+    },
+    {
+      "epoch": 0.48454205509032494,
+      "grad_norm": 0.7268190979957581,
+      "learning_rate": 0.00019884327291241446,
+      "loss": 1.8246,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4866487596776742,
+      "grad_norm": 0.671644926071167,
+      "learning_rate": 0.0001988332121981436,
+      "loss": 1.8348,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4887554642650234,
+      "grad_norm": 0.7536865472793579,
+      "learning_rate": 0.00019882310817796235,
+      "loss": 1.7981,
+      "step": 2320
+    },
+    {
+      "epoch": 0.49086216885237266,
+      "grad_norm": 0.8593438863754272,
+      "learning_rate": 0.00019881296085629807,
+      "loss": 1.8558,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4929688734397219,
+      "grad_norm": 0.7384650707244873,
+      "learning_rate": 0.00019880277023759702,
+      "loss": 1.8659,
+      "step": 2340
+    },
+    {
+      "epoch": 0.49507557802707114,
+      "grad_norm": 0.7357537150382996,
+      "learning_rate": 0.0001987925363263244,
+      "loss": 1.7887,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4971822826144204,
+      "grad_norm": 0.7231765389442444,
+      "learning_rate": 0.00019878225912696446,
+      "loss": 1.8046,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4992889872017696,
+      "grad_norm": 0.7143809199333191,
+      "learning_rate": 0.00019877193864402038,
+      "loss": 1.8156,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5013956917891189,
+      "grad_norm": 0.7593514323234558,
+      "learning_rate": 0.00019876157488201424,
+      "loss": 1.8299,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5035023963764681,
+      "grad_norm": 0.7489436864852905,
+      "learning_rate": 0.00019875116784548723,
+      "loss": 1.7569,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5056091009638174,
+      "grad_norm": 0.7284204363822937,
+      "learning_rate": 0.0001987407175389994,
+      "loss": 1.8094,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5077158055511666,
+      "grad_norm": 0.7947770357131958,
+      "learning_rate": 0.00019873022396712972,
+      "loss": 1.8225,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5098225101385159,
+      "grad_norm": 0.6892939209938049,
+      "learning_rate": 0.00019871968713447625,
+      "loss": 1.7943,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5119292147258651,
+      "grad_norm": 0.7589520215988159,
+      "learning_rate": 0.00019870910704565588,
+      "loss": 1.774,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5140359193132144,
+      "grad_norm": 0.7198192477226257,
+      "learning_rate": 0.00019869848370530452,
+      "loss": 1.7809,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5161426239005635,
+      "grad_norm": 0.7147354483604431,
+      "learning_rate": 0.00019868781711807705,
+      "loss": 1.8524,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5182493284879128,
+      "grad_norm": 0.8181036710739136,
+      "learning_rate": 0.0001986771072886472,
+      "loss": 1.8625,
+      "step": 2460
+    },
+    {
+      "epoch": 0.520356033075262,
+      "grad_norm": 0.7021554112434387,
+      "learning_rate": 0.00019866635422170775,
+      "loss": 1.8081,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5224627376626113,
+      "grad_norm": 0.7589981555938721,
+      "learning_rate": 0.00019865555792197042,
+      "loss": 1.8463,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5245694422499605,
+      "grad_norm": 0.734856128692627,
+      "learning_rate": 0.00019864471839416576,
+      "loss": 1.852,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5266761468373098,
+      "grad_norm": 0.7465302348136902,
+      "learning_rate": 0.00019863383564304346,
+      "loss": 1.809,
+      "step": 2500
+    },
+    {
+      "epoch": 0.528782851424659,
+      "grad_norm": 0.7393603324890137,
+      "learning_rate": 0.00019862290967337192,
+      "loss": 1.842,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5308895560120083,
+      "grad_norm": 0.7171728610992432,
+      "learning_rate": 0.00019861194048993863,
+      "loss": 1.7747,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5329962605993575,
+      "grad_norm": 0.7123204469680786,
+      "learning_rate": 0.00019860092809755,
+      "loss": 1.8511,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5351029651867067,
+      "grad_norm": 0.8142136931419373,
+      "learning_rate": 0.00019858987250103132,
+      "loss": 1.8167,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5372096697740559,
+      "grad_norm": 0.7203823924064636,
+      "learning_rate": 0.00019857877370522685,
+      "loss": 1.8407,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5393163743614052,
+      "grad_norm": 0.7605892419815063,
+      "learning_rate": 0.00019856763171499978,
+      "loss": 1.8308,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5414230789487544,
+      "grad_norm": 0.735731840133667,
+      "learning_rate": 0.00019855644653523217,
+      "loss": 1.808,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5435297835361036,
+      "grad_norm": 0.7104590535163879,
+      "learning_rate": 0.0001985452181708251,
+      "loss": 1.7968,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5456364881234529,
+      "grad_norm": 0.734384298324585,
+      "learning_rate": 0.00019853394662669847,
+      "loss": 1.8222,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5477431927108021,
+      "grad_norm": 0.7150630950927734,
+      "learning_rate": 0.00019852263190779122,
+      "loss": 1.8305,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5498498972981514,
+      "grad_norm": 0.7069328427314758,
+      "learning_rate": 0.0001985112740190611,
+      "loss": 1.8839,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5519566018855006,
+      "grad_norm": 0.8343731164932251,
+      "learning_rate": 0.00019849987296548477,
+      "loss": 1.8147,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5540633064728498,
+      "grad_norm": 0.74041748046875,
+      "learning_rate": 0.00019848842875205792,
+      "loss": 1.8146,
+      "step": 2630
+    },
+    {
+      "epoch": 0.556170011060199,
+      "grad_norm": 0.7421287298202515,
+      "learning_rate": 0.00019847694138379506,
+      "loss": 1.7814,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5582767156475483,
+      "grad_norm": 0.7727276682853699,
+      "learning_rate": 0.0001984654108657296,
+      "loss": 1.82,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5603834202348975,
+      "grad_norm": 0.7713900208473206,
+      "learning_rate": 0.00019845383720291392,
+      "loss": 1.8117,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5624901248222468,
+      "grad_norm": 0.7313315272331238,
+      "learning_rate": 0.00019844222040041928,
+      "loss": 1.8402,
+      "step": 2670
+    },
+    {
+      "epoch": 0.564596829409596,
+      "grad_norm": 0.7062333226203918,
+      "learning_rate": 0.00019843056046333577,
+      "loss": 1.8251,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5667035339969453,
+      "grad_norm": 0.7643407583236694,
+      "learning_rate": 0.00019841885739677251,
+      "loss": 1.8659,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5688102385842945,
+      "grad_norm": 0.7446190714836121,
+      "learning_rate": 0.0001984071112058574,
+      "loss": 1.8089,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5709169431716438,
+      "grad_norm": 0.7492159605026245,
+      "learning_rate": 0.00019839532189573733,
+      "loss": 1.7645,
+      "step": 2710
+    },
+    {
+      "epoch": 0.573023647758993,
+      "grad_norm": 0.707715630531311,
+      "learning_rate": 0.00019838348947157804,
+      "loss": 1.7683,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5751303523463422,
+      "grad_norm": 0.7064851522445679,
+      "learning_rate": 0.0001983716139385641,
+      "loss": 1.8434,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5772370569336914,
+      "grad_norm": 0.6879447102546692,
+      "learning_rate": 0.00019835969530189912,
+      "loss": 1.8176,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5793437615210407,
+      "grad_norm": 0.7283410429954529,
+      "learning_rate": 0.00019834773356680547,
+      "loss": 1.8101,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5814504661083899,
+      "grad_norm": 0.7824443578720093,
+      "learning_rate": 0.00019833572873852444,
+      "loss": 1.8454,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5835571706957392,
+      "grad_norm": 0.7730831503868103,
+      "learning_rate": 0.0001983236808223162,
+      "loss": 1.8517,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5856638752830884,
+      "grad_norm": 0.7812346816062927,
+      "learning_rate": 0.0001983115898234598,
+      "loss": 1.8139,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5877705798704377,
+      "grad_norm": 0.7695793509483337,
+      "learning_rate": 0.0001982994557472532,
+      "loss": 1.8299,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5898772844577869,
+      "grad_norm": 0.7079188227653503,
+      "learning_rate": 0.00019828727859901317,
+      "loss": 1.7954,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5919839890451362,
+      "grad_norm": 0.9112276434898376,
+      "learning_rate": 0.00019827505838407544,
+      "loss": 1.8438,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5940906936324853,
+      "grad_norm": 0.7448698282241821,
+      "learning_rate": 0.00019826279510779454,
+      "loss": 1.7831,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5961973982198346,
+      "grad_norm": 0.8215844035148621,
+      "learning_rate": 0.00019825048877554385,
+      "loss": 1.8194,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5983041028071838,
+      "grad_norm": 0.7165418863296509,
+      "learning_rate": 0.0001982381393927157,
+      "loss": 1.8716,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6004108073945331,
+      "grad_norm": 0.7175021171569824,
+      "learning_rate": 0.00019822574696472126,
+      "loss": 1.8433,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6025175119818823,
+      "grad_norm": 0.8004611730575562,
+      "learning_rate": 0.00019821331149699048,
+      "loss": 1.8758,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6046242165692316,
+      "grad_norm": 0.731736958026886,
+      "learning_rate": 0.00019820083299497228,
+      "loss": 1.7633,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6067309211565808,
+      "grad_norm": 0.7083932757377625,
+      "learning_rate": 0.00019818831146413434,
+      "loss": 1.7994,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6088376257439301,
+      "grad_norm": 0.740691065788269,
+      "learning_rate": 0.0001981757469099633,
+      "loss": 1.7944,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6109443303312793,
+      "grad_norm": 0.7464807629585266,
+      "learning_rate": 0.0001981631393379645,
+      "loss": 1.8093,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6130510349186286,
+      "grad_norm": 0.7055980563163757,
+      "learning_rate": 0.00019815048875366234,
+      "loss": 1.8142,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6151577395059777,
+      "grad_norm": 0.7619740962982178,
+      "learning_rate": 0.00019813779516259986,
+      "loss": 1.7805,
+      "step": 2920
+    },
+    {
+      "epoch": 0.617264444093327,
+      "grad_norm": 0.8591639995574951,
+      "learning_rate": 0.00019812505857033904,
+      "loss": 1.8294,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6193711486806762,
+      "grad_norm": 0.7257503867149353,
+      "learning_rate": 0.0001981122789824607,
+      "loss": 1.7968,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6214778532680255,
+      "grad_norm": 0.8203638195991516,
+      "learning_rate": 0.00019809945640456453,
+      "loss": 1.8513,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6235845578553747,
+      "grad_norm": 0.7143609523773193,
+      "learning_rate": 0.000198086590842269,
+      "loss": 1.7997,
+      "step": 2960
+    },
+    {
+      "epoch": 0.625691262442724,
+      "grad_norm": 0.7339126467704773,
+      "learning_rate": 0.0001980736823012114,
+      "loss": 1.8356,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6277979670300732,
+      "grad_norm": 0.755310595035553,
+      "learning_rate": 0.0001980607307870479,
+      "loss": 1.771,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6299046716174225,
+      "grad_norm": 0.7657110691070557,
+      "learning_rate": 0.00019804773630545353,
+      "loss": 1.8006,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6320113762047717,
+      "grad_norm": 0.7267753481864929,
+      "learning_rate": 0.0001980346988621221,
+      "loss": 1.895,
+      "step": 3000
+    },
+    {
+      "epoch": 0.634118080792121,
+      "grad_norm": 0.7973232865333557,
+      "learning_rate": 0.00019802161846276615,
+      "loss": 1.8724,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6362247853794701,
+      "grad_norm": 0.7586084008216858,
+      "learning_rate": 0.00019800849511311726,
+      "loss": 1.8659,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6383314899668194,
+      "grad_norm": 0.7559179663658142,
+      "learning_rate": 0.00019799532881892564,
+      "loss": 1.8208,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6404381945541686,
+      "grad_norm": 0.7547093629837036,
+      "learning_rate": 0.0001979821195859604,
+      "loss": 1.7948,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6425448991415179,
+      "grad_norm": 0.7194349765777588,
+      "learning_rate": 0.0001979688674200095,
+      "loss": 1.795,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6446516037288671,
+      "grad_norm": 0.7290440797805786,
+      "learning_rate": 0.00019795557232687956,
+      "loss": 1.8149,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6467583083162164,
+      "grad_norm": 0.7215458750724792,
+      "learning_rate": 0.0001979422343123962,
+      "loss": 1.8181,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6488650129035656,
+      "grad_norm": 0.7262182831764221,
+      "learning_rate": 0.00019792885338240374,
+      "loss": 1.7897,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6509717174909149,
+      "grad_norm": 0.7412195205688477,
+      "learning_rate": 0.0001979154295427653,
+      "loss": 1.8387,
+      "step": 3090
+    },
+    {
+      "epoch": 0.653078422078264,
+      "grad_norm": 0.7855926156044006,
+      "learning_rate": 0.00019790196279936286,
+      "loss": 1.8165,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6551851266656133,
+      "grad_norm": 0.7217236161231995,
+      "learning_rate": 0.00019788845315809713,
+      "loss": 1.7652,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6572918312529625,
+      "grad_norm": 0.7485338449478149,
+      "learning_rate": 0.0001978749006248877,
+      "loss": 1.7949,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6593985358403118,
+      "grad_norm": 0.7767829895019531,
+      "learning_rate": 0.00019786130520567285,
+      "loss": 1.802,
+      "step": 3130
+    },
+    {
+      "epoch": 0.661505240427661,
+      "grad_norm": 0.7325339317321777,
+      "learning_rate": 0.00019784766690640975,
+      "loss": 1.7796,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6636119450150103,
+      "grad_norm": 0.7544266581535339,
+      "learning_rate": 0.00019783398573307428,
+      "loss": 1.8225,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6657186496023595,
+      "grad_norm": 1.14772367477417,
+      "learning_rate": 0.00019782026169166118,
+      "loss": 1.8231,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6678253541897088,
+      "grad_norm": 0.7514392733573914,
+      "learning_rate": 0.0001978064947881839,
+      "loss": 1.8369,
+      "step": 3170
+    },
+    {
+      "epoch": 0.669932058777058,
+      "grad_norm": 0.7889571785926819,
+      "learning_rate": 0.00019779268502867473,
+      "loss": 1.779,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6720387633644073,
+      "grad_norm": 0.7332549095153809,
+      "learning_rate": 0.00019777883241918468,
+      "loss": 1.8519,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6741454679517564,
+      "grad_norm": 0.7040186524391174,
+      "learning_rate": 0.00019776493696578365,
+      "loss": 1.7989,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6762521725391057,
+      "grad_norm": 0.7270476818084717,
+      "learning_rate": 0.00019775099867456013,
+      "loss": 1.8423,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6783588771264549,
+      "grad_norm": 0.7685533165931702,
+      "learning_rate": 0.00019773701755162158,
+      "loss": 1.7827,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6804655817138042,
+      "grad_norm": 0.6836286187171936,
+      "learning_rate": 0.00019772299360309406,
+      "loss": 1.8167,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6825722863011534,
+      "grad_norm": 0.7396418452262878,
+      "learning_rate": 0.0001977089268351225,
+      "loss": 1.8,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6846789908885027,
+      "grad_norm": 0.7958587408065796,
+      "learning_rate": 0.00019769481725387053,
+      "loss": 1.8342,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6867856954758519,
+      "grad_norm": 0.7690047025680542,
+      "learning_rate": 0.00019768066486552065,
+      "loss": 1.8248,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6888924000632012,
+      "grad_norm": 0.7547836303710938,
+      "learning_rate": 0.0001976664696762739,
+      "loss": 1.8619,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6909991046505504,
+      "grad_norm": 0.8090780377388,
+      "learning_rate": 0.00019765223169235035,
+      "loss": 1.8464,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6931058092378997,
+      "grad_norm": 0.7340956926345825,
+      "learning_rate": 0.00019763795091998858,
+      "loss": 1.8414,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6952125138252488,
+      "grad_norm": 0.7637834548950195,
+      "learning_rate": 0.00019762362736544607,
+      "loss": 1.775,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6973192184125981,
+      "grad_norm": 0.7355379462242126,
+      "learning_rate": 0.00019760926103499897,
+      "loss": 1.8039,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6994259229999473,
+      "grad_norm": 0.7598356008529663,
+      "learning_rate": 0.0001975948519349422,
+      "loss": 1.7958,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7015326275872966,
+      "grad_norm": 0.6928766369819641,
+      "learning_rate": 0.00019758040007158948,
+      "loss": 1.7421,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7036393321746458,
+      "grad_norm": 0.7078458666801453,
+      "learning_rate": 0.00019756590545127312,
+      "loss": 1.8463,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7057460367619951,
+      "grad_norm": 0.7719635963439941,
+      "learning_rate": 0.00019755136808034425,
+      "loss": 1.7731,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7078527413493443,
+      "grad_norm": 0.7915549278259277,
+      "learning_rate": 0.00019753678796517282,
+      "loss": 1.8302,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7099594459366936,
+      "grad_norm": 0.7216295599937439,
+      "learning_rate": 0.00019752216511214737,
+      "loss": 1.8054,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7120661505240428,
+      "grad_norm": 0.7865151762962341,
+      "learning_rate": 0.0001975074995276752,
+      "loss": 1.8076,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7141728551113921,
+      "grad_norm": 0.7931540608406067,
+      "learning_rate": 0.00019749279121818235,
+      "loss": 1.8389,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7162795596987412,
+      "grad_norm": 0.7390506267547607,
+      "learning_rate": 0.00019747804019011367,
+      "loss": 1.7799,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7183862642860905,
+      "grad_norm": 0.7065187096595764,
+      "learning_rate": 0.00019746324644993255,
+      "loss": 1.7853,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7204929688734397,
+      "grad_norm": 0.7845907211303711,
+      "learning_rate": 0.00019744841000412123,
+      "loss": 1.8037,
+      "step": 3420
+    },
+    {
+      "epoch": 0.722599673460789,
+      "grad_norm": 0.7040785551071167,
+      "learning_rate": 0.0001974335308591806,
+      "loss": 1.7391,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7247063780481382,
+      "grad_norm": 0.7547016739845276,
+      "learning_rate": 0.00019741860902163029,
+      "loss": 1.8104,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7268130826354874,
+      "grad_norm": 0.7265505790710449,
+      "learning_rate": 0.0001974036444980086,
+      "loss": 1.8016,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7289197872228367,
+      "grad_norm": 0.6882880926132202,
+      "learning_rate": 0.0001973886372948726,
+      "loss": 1.7781,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7310264918101859,
+      "grad_norm": 0.7181904911994934,
+      "learning_rate": 0.00019737358741879802,
+      "loss": 1.8162,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7331331963975352,
+      "grad_norm": 0.7640955448150635,
+      "learning_rate": 0.00019735849487637929,
+      "loss": 1.7987,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7352399009848843,
+      "grad_norm": 0.7534769773483276,
+      "learning_rate": 0.00019734335967422947,
+      "loss": 1.7754,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7373466055722336,
+      "grad_norm": 0.7754014730453491,
+      "learning_rate": 0.00019732818181898045,
+      "loss": 1.827,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7394533101595828,
+      "grad_norm": 0.6991292834281921,
+      "learning_rate": 0.0001973129613172827,
+      "loss": 1.8061,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7415600147469321,
+      "grad_norm": 0.7496746182441711,
+      "learning_rate": 0.00019729769817580542,
+      "loss": 1.8001,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7436667193342813,
+      "grad_norm": 0.7338647246360779,
+      "learning_rate": 0.0001972823924012365,
+      "loss": 1.8216,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7457734239216306,
+      "grad_norm": 0.742842435836792,
+      "learning_rate": 0.0001972670440002825,
+      "loss": 1.7468,
+      "step": 3540
+    },
+    {
+      "epoch": 0.7478801285089798,
+      "grad_norm": 0.7462572455406189,
+      "learning_rate": 0.00019725165297966859,
+      "loss": 1.7863,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7499868330963291,
+      "grad_norm": 0.7699198126792908,
+      "learning_rate": 0.00019723621934613874,
+      "loss": 1.7659,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7520935376836783,
+      "grad_norm": 0.7163071632385254,
+      "learning_rate": 0.00019722074310645553,
+      "loss": 1.8067,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7542002422710276,
+      "grad_norm": 0.7273010015487671,
+      "learning_rate": 0.0001972052242674002,
+      "loss": 1.8497,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7563069468583767,
+      "grad_norm": 0.7326186895370483,
+      "learning_rate": 0.00019718966283577264,
+      "loss": 1.7821,
+      "step": 3590
+    },
+    {
+      "epoch": 0.758413651445726,
+      "grad_norm": 0.7297202944755554,
+      "learning_rate": 0.00019717405881839145,
+      "loss": 1.8288,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7605203560330752,
+      "grad_norm": 0.7100248336791992,
+      "learning_rate": 0.00019715841222209387,
+      "loss": 1.7949,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7626270606204245,
+      "grad_norm": 0.7175480723381042,
+      "learning_rate": 0.0001971427230537358,
+      "loss": 1.807,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7647337652077737,
+      "grad_norm": 0.7053601741790771,
+      "learning_rate": 0.0001971269913201918,
+      "loss": 1.8223,
+      "step": 3630
+    },
+    {
+      "epoch": 0.766840469795123,
+      "grad_norm": 0.7398414611816406,
+      "learning_rate": 0.00019711121702835504,
+      "loss": 1.7607,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7689471743824722,
+      "grad_norm": 0.7213431596755981,
+      "learning_rate": 0.00019709540018513736,
+      "loss": 1.7927,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7710538789698215,
+      "grad_norm": 0.7234872579574585,
+      "learning_rate": 0.00019707954079746927,
+      "loss": 1.8027,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7731605835571707,
+      "grad_norm": 0.8808572292327881,
+      "learning_rate": 0.0001970636388722999,
+      "loss": 1.7833,
+      "step": 3670
+    },
+    {
+      "epoch": 0.77526728814452,
+      "grad_norm": 0.7086265683174133,
+      "learning_rate": 0.00019704769441659703,
+      "loss": 1.7386,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7773739927318691,
+      "grad_norm": 0.7601612210273743,
+      "learning_rate": 0.00019703170743734706,
+      "loss": 1.7816,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7794806973192184,
+      "grad_norm": 0.7857274413108826,
+      "learning_rate": 0.000197015677941555,
+      "loss": 1.7497,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7815874019065676,
+      "grad_norm": 0.7167443633079529,
+      "learning_rate": 0.00019699960593624462,
+      "loss": 1.7963,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7836941064939169,
+      "grad_norm": 0.6912304759025574,
+      "learning_rate": 0.00019698349142845814,
+      "loss": 1.738,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7858008110812661,
+      "grad_norm": 0.7966861724853516,
+      "learning_rate": 0.00019696733442525646,
+      "loss": 1.8344,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7879075156686154,
+      "grad_norm": 0.8376454710960388,
+      "learning_rate": 0.00019695113493371918,
+      "loss": 1.7743,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7900142202559646,
+      "grad_norm": 0.7219945192337036,
+      "learning_rate": 0.00019693489296094443,
+      "loss": 1.8116,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7921209248433139,
+      "grad_norm": 0.7250286936759949,
+      "learning_rate": 0.00019691860851404897,
+      "loss": 1.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.794227629430663,
+      "grad_norm": 0.8389397859573364,
+      "learning_rate": 0.00019690228160016817,
+      "loss": 1.7662,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7963343340180123,
+      "grad_norm": 0.8005920052528381,
+      "learning_rate": 0.00019688591222645607,
+      "loss": 1.8361,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7984410386053615,
+      "grad_norm": 0.7611145377159119,
+      "learning_rate": 0.00019686950040008526,
+      "loss": 1.8219,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8005477431927108,
+      "grad_norm": 0.7690491080284119,
+      "learning_rate": 0.0001968530461282469,
+      "loss": 1.78,
+      "step": 3800
+    },
+    {
+      "epoch": 0.80265444778006,
+      "grad_norm": 0.7202954292297363,
+      "learning_rate": 0.00019683654941815077,
+      "loss": 1.8325,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8047611523674093,
+      "grad_norm": 0.7846165299415588,
+      "learning_rate": 0.00019682001027702533,
+      "loss": 1.7509,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8068678569547585,
+      "grad_norm": 0.7994365096092224,
+      "learning_rate": 0.00019680342871211752,
+      "loss": 1.8076,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8089745615421078,
+      "grad_norm": 0.854131281375885,
+      "learning_rate": 0.00019678680473069293,
+      "loss": 1.8089,
+      "step": 3840
+    },
+    {
+      "epoch": 0.811081266129457,
+      "grad_norm": 0.7317761778831482,
+      "learning_rate": 0.0001967701383400357,
+      "loss": 1.8047,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8131879707168063,
+      "grad_norm": 0.7426530718803406,
+      "learning_rate": 0.00019675342954744853,
+      "loss": 1.7675,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8152946753041554,
+      "grad_norm": 0.7657327055931091,
+      "learning_rate": 0.00019673667836025283,
+      "loss": 1.8319,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8174013798915047,
+      "grad_norm": 0.773003339767456,
+      "learning_rate": 0.00019671988478578843,
+      "loss": 1.8963,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8195080844788539,
+      "grad_norm": 0.7750910520553589,
+      "learning_rate": 0.00019670304883141382,
+      "loss": 1.8312,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8216147890662032,
+      "grad_norm": 0.7256291508674622,
+      "learning_rate": 0.00019668617050450603,
+      "loss": 1.7888,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8237214936535524,
+      "grad_norm": 0.7944165468215942,
+      "learning_rate": 0.00019666924981246066,
+      "loss": 1.8584,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8258281982409017,
+      "grad_norm": 0.7136659622192383,
+      "learning_rate": 0.0001966522867626919,
+      "loss": 1.8315,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8279349028282509,
+      "grad_norm": 0.8189762830734253,
+      "learning_rate": 0.00019663528136263246,
+      "loss": 1.8193,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8300416074156002,
+      "grad_norm": 0.7890966534614563,
+      "learning_rate": 0.0001966182336197336,
+      "loss": 1.8204,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8321483120029494,
+      "grad_norm": 0.8017906546592712,
+      "learning_rate": 0.00019660114354146525,
+      "loss": 1.7891,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8342550165902987,
+      "grad_norm": 0.8445520997047424,
+      "learning_rate": 0.00019658401113531565,
+      "loss": 1.8108,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8363617211776478,
+      "grad_norm": 0.7506213188171387,
+      "learning_rate": 0.00019656683640879185,
+      "loss": 1.7811,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8384684257649971,
+      "grad_norm": 0.7670649290084839,
+      "learning_rate": 0.00019654961936941932,
+      "loss": 1.8,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8405751303523463,
+      "grad_norm": 0.86253422498703,
+      "learning_rate": 0.000196532360024742,
+      "loss": 1.7841,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8426818349396956,
+      "grad_norm": 0.8307934999465942,
+      "learning_rate": 0.00019651505838232255,
+      "loss": 1.7813,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8447885395270448,
+      "grad_norm": 0.7400214076042175,
+      "learning_rate": 0.00019649771444974197,
+      "loss": 1.7305,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8468952441143941,
+      "grad_norm": 0.7180853486061096,
+      "learning_rate": 0.00019648032823459994,
+      "loss": 1.7569,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8490019487017433,
+      "grad_norm": 0.7013576626777649,
+      "learning_rate": 0.00019646289974451455,
+      "loss": 1.796,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8511086532890926,
+      "grad_norm": 0.7643212676048279,
+      "learning_rate": 0.00019644542898712252,
+      "loss": 1.7791,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8532153578764418,
+      "grad_norm": 0.7359760999679565,
+      "learning_rate": 0.00019642791597007902,
+      "loss": 1.7836,
+      "step": 4050
+    },
+    {
+      "epoch": 0.855322062463791,
+      "grad_norm": 0.7171246409416199,
+      "learning_rate": 0.00019641036070105778,
+      "loss": 1.7218,
+      "step": 4060
+    },
+    {
+      "epoch": 0.8574287670511402,
+      "grad_norm": 0.7775237560272217,
+      "learning_rate": 0.000196392763187751,
+      "loss": 1.7838,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8595354716384895,
+      "grad_norm": 0.687788188457489,
+      "learning_rate": 0.00019637512343786937,
+      "loss": 1.8259,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8616421762258387,
+      "grad_norm": 0.7279845476150513,
+      "learning_rate": 0.00019635744145914222,
+      "loss": 1.755,
+      "step": 4090
+    },
+    {
+      "epoch": 0.863748880813188,
+      "grad_norm": 0.8731548190116882,
+      "learning_rate": 0.0001963397172593172,
+      "loss": 1.8539,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8658555854005372,
+      "grad_norm": 0.7026403546333313,
+      "learning_rate": 0.00019632195084616063,
+      "loss": 1.7384,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8679622899878865,
+      "grad_norm": 0.7707716226577759,
+      "learning_rate": 0.0001963041422274572,
+      "loss": 1.821,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8700689945752357,
+      "grad_norm": 0.7273789644241333,
+      "learning_rate": 0.00019628629141101012,
+      "loss": 1.8124,
+      "step": 4130
+    },
+    {
+      "epoch": 0.872175699162585,
+      "grad_norm": 0.7353452444076538,
+      "learning_rate": 0.00019626839840464119,
+      "loss": 1.7925,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8742824037499342,
+      "grad_norm": 0.7055097818374634,
+      "learning_rate": 0.00019625046321619053,
+      "loss": 1.8403,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8763891083372835,
+      "grad_norm": 0.8632143139839172,
+      "learning_rate": 0.0001962324858535169,
+      "loss": 1.78,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8784958129246326,
+      "grad_norm": 0.7952572107315063,
+      "learning_rate": 0.00019621446632449744,
+      "loss": 1.7858,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8806025175119819,
+      "grad_norm": 0.7177860736846924,
+      "learning_rate": 0.00019619640463702779,
+      "loss": 1.7926,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8827092220993311,
+      "grad_norm": 0.7487980723381042,
+      "learning_rate": 0.0001961783007990221,
+      "loss": 1.7805,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8848159266866804,
+      "grad_norm": 0.8172057271003723,
+      "learning_rate": 0.0001961601548184129,
+      "loss": 1.7751,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8869226312740296,
+      "grad_norm": 0.7768619656562805,
+      "learning_rate": 0.00019614196670315133,
+      "loss": 1.7962,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8890293358613789,
+      "grad_norm": 0.7874531149864197,
+      "learning_rate": 0.00019612373646120683,
+      "loss": 1.7485,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8911360404487281,
+      "grad_norm": 0.7843292355537415,
+      "learning_rate": 0.0001961054641005674,
+      "loss": 1.8358,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8932427450360774,
+      "grad_norm": 0.7430649995803833,
+      "learning_rate": 0.00019608714962923948,
+      "loss": 1.8252,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8953494496234266,
+      "grad_norm": 0.7348790764808655,
+      "learning_rate": 0.00019606879305524794,
+      "loss": 1.8759,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8974561542107758,
+      "grad_norm": 0.7689461708068848,
+      "learning_rate": 0.00019605039438663614,
+      "loss": 1.8029,
+      "step": 4260
+    },
+    {
+      "epoch": 0.899562858798125,
+      "grad_norm": 0.7884606719017029,
+      "learning_rate": 0.0001960319536314658,
+      "loss": 1.7901,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9016695633854743,
+      "grad_norm": 0.7943284511566162,
+      "learning_rate": 0.0001960134707978172,
+      "loss": 1.7719,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9037762679728235,
+      "grad_norm": 0.9755631685256958,
+      "learning_rate": 0.0001959949458937889,
+      "loss": 1.8401,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9058829725601727,
+      "grad_norm": 0.7475715279579163,
+      "learning_rate": 0.0001959763789274981,
+      "loss": 1.7304,
+      "step": 4300
+    },
+    {
+      "epoch": 0.907989677147522,
+      "grad_norm": 0.731900155544281,
+      "learning_rate": 0.00019595776990708022,
+      "loss": 1.779,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9100963817348712,
+      "grad_norm": 0.7448343634605408,
+      "learning_rate": 0.0001959391188406893,
+      "loss": 1.7347,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9122030863222205,
+      "grad_norm": 0.7453986406326294,
+      "learning_rate": 0.00019592042573649763,
+      "loss": 1.7833,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9143097909095697,
+      "grad_norm": 0.7423632144927979,
+      "learning_rate": 0.00019590169060269602,
+      "loss": 1.7099,
+      "step": 4340
+    },
+    {
+      "epoch": 0.916416495496919,
+      "grad_norm": 0.7491478323936462,
+      "learning_rate": 0.0001958829134474937,
+      "loss": 1.7697,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9185232000842681,
+      "grad_norm": 0.7573466300964355,
+      "learning_rate": 0.00019586409427911825,
+      "loss": 1.8192,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9206299046716174,
+      "grad_norm": 0.7873802185058594,
+      "learning_rate": 0.00019584523310581573,
+      "loss": 1.7532,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9227366092589666,
+      "grad_norm": 0.782875120639801,
+      "learning_rate": 0.00019582632993585052,
+      "loss": 1.7863,
+      "step": 4380
+    },
+    {
+      "epoch": 0.9248433138463159,
+      "grad_norm": 0.7400102615356445,
+      "learning_rate": 0.00019580738477750553,
+      "loss": 1.8116,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9269500184336651,
+      "grad_norm": 0.7774225473403931,
+      "learning_rate": 0.00019578839763908192,
+      "loss": 1.7869,
+      "step": 4400
+    },
+    {
+      "epoch": 0.9290567230210144,
+      "grad_norm": 0.7710514068603516,
+      "learning_rate": 0.00019576936852889936,
+      "loss": 1.7993,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9311634276083636,
+      "grad_norm": 0.6935902833938599,
+      "learning_rate": 0.00019575029745529582,
+      "loss": 1.7855,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9332701321957129,
+      "grad_norm": 0.7460892796516418,
+      "learning_rate": 0.00019573118442662776,
+      "loss": 1.8266,
+      "step": 4430
+    },
+    {
+      "epoch": 0.935376836783062,
+      "grad_norm": 0.8303509950637817,
+      "learning_rate": 0.00019571202945126994,
+      "loss": 1.7451,
+      "step": 4440
+    },
+    {
+      "epoch": 0.9374835413704113,
+      "grad_norm": 0.7582190632820129,
+      "learning_rate": 0.00019569283253761553,
+      "loss": 1.7904,
+      "step": 4450
+    },
+    {
+      "epoch": 0.9395902459577605,
+      "grad_norm": 0.7921029329299927,
+      "learning_rate": 0.00019567359369407605,
+      "loss": 1.7872,
+      "step": 4460
+    },
+    {
+      "epoch": 0.9416969505451098,
+      "grad_norm": 0.7588982582092285,
+      "learning_rate": 0.00019565431292908146,
+      "loss": 1.8489,
+      "step": 4470
+    },
+    {
+      "epoch": 0.943803655132459,
+      "grad_norm": 0.7762647867202759,
+      "learning_rate": 0.00019563499025107998,
+      "loss": 1.8482,
+      "step": 4480
+    },
+    {
+      "epoch": 0.9459103597198083,
+      "grad_norm": 0.7484056949615479,
+      "learning_rate": 0.00019561562566853836,
+      "loss": 1.7979,
+      "step": 4490
+    },
+    {
+      "epoch": 0.9480170643071575,
+      "grad_norm": 0.7211093902587891,
+      "learning_rate": 0.0001955962191899415,
+      "loss": 1.7849,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9501237688945068,
+      "grad_norm": 0.7542027235031128,
+      "learning_rate": 0.00019557677082379286,
+      "loss": 1.7533,
+      "step": 4510
+    },
+    {
+      "epoch": 0.952230473481856,
+      "grad_norm": 0.7238128185272217,
+      "learning_rate": 0.0001955572805786141,
+      "loss": 1.8403,
+      "step": 4520
+    },
+    {
+      "epoch": 0.9543371780692053,
+      "grad_norm": 0.7738656997680664,
+      "learning_rate": 0.0001955377484629453,
+      "loss": 1.8056,
+      "step": 4530
+    },
+    {
+      "epoch": 0.9564438826565544,
+      "grad_norm": 0.7569396495819092,
+      "learning_rate": 0.0001955181744853449,
+      "loss": 1.8334,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9585505872439037,
+      "grad_norm": 0.7401841878890991,
+      "learning_rate": 0.00019549855865438965,
+      "loss": 1.8209,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9606572918312529,
+      "grad_norm": 0.8001863956451416,
+      "learning_rate": 0.00019547890097867468,
+      "loss": 1.7749,
+      "step": 4560
+    },
+    {
+      "epoch": 0.9627639964186022,
+      "grad_norm": 0.8146688938140869,
+      "learning_rate": 0.00019545920146681338,
+      "loss": 1.7778,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9648707010059514,
+      "grad_norm": 0.7350276708602905,
+      "learning_rate": 0.00019543946012743756,
+      "loss": 1.7766,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9669774055933007,
+      "grad_norm": 0.7190563082695007,
+      "learning_rate": 0.0001954196769691973,
+      "loss": 1.7803,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9690841101806499,
+      "grad_norm": 0.7367439866065979,
+      "learning_rate": 0.00019539985200076098,
+      "loss": 1.7911,
+      "step": 4600
+    },
+    {
+      "epoch": 0.9711908147679992,
+      "grad_norm": 0.760611355304718,
+      "learning_rate": 0.0001953799852308154,
+      "loss": 1.8031,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9732975193553484,
+      "grad_norm": 0.7771376967430115,
+      "learning_rate": 0.00019536007666806556,
+      "loss": 1.764,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9754042239426977,
+      "grad_norm": 0.7223167419433594,
+      "learning_rate": 0.00019534012632123484,
+      "loss": 1.7894,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9775109285300468,
+      "grad_norm": 0.72389155626297,
+      "learning_rate": 0.00019532013419906497,
+      "loss": 1.8032,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9796176331173961,
+      "grad_norm": 0.7463617324829102,
+      "learning_rate": 0.00019530010031031586,
+      "loss": 1.7631,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9817243377047453,
+      "grad_norm": 0.7694106698036194,
+      "learning_rate": 0.00019528002466376586,
+      "loss": 1.7619,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9838310422920946,
+      "grad_norm": 0.7413203120231628,
+      "learning_rate": 0.0001952599072682115,
+      "loss": 1.7901,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9859377468794438,
+      "grad_norm": 0.7745965123176575,
+      "learning_rate": 0.00019523974813246767,
+      "loss": 1.7842,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9880444514667931,
+      "grad_norm": 0.7581157088279724,
+      "learning_rate": 0.0001952195472653675,
+      "loss": 1.8425,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9901511560541423,
+      "grad_norm": 0.7829221487045288,
+      "learning_rate": 0.00019519930467576247,
+      "loss": 1.8011,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9922578606414916,
+      "grad_norm": 0.7388656735420227,
+      "learning_rate": 0.0001951790203725223,
+      "loss": 1.7878,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9943645652288408,
+      "grad_norm": 1.5744229555130005,
+      "learning_rate": 0.00019515869436453502,
+      "loss": 1.8004,
+      "step": 4720
+    },
+    {
+      "epoch": 0.99647126981619,
+      "grad_norm": 0.787341296672821,
+      "learning_rate": 0.00019513832666070687,
+      "loss": 1.7781,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9985779744035392,
+      "grad_norm": 0.8090167045593262,
+      "learning_rate": 0.00019511791726996243,
+      "loss": 1.8597,
+      "step": 4740
+    },
+    {
+      "epoch": 1.0006846789908885,
+      "grad_norm": 0.7438858151435852,
+      "learning_rate": 0.00019509746620124447,
+      "loss": 1.7137,
+      "step": 4750
+    },
+    {
+      "epoch": 1.0027913835782378,
+      "grad_norm": 0.7524532079696655,
+      "learning_rate": 0.00019507697346351414,
+      "loss": 1.7519,
+      "step": 4760
+    },
+    {
+      "epoch": 1.004898088165587,
+      "grad_norm": 0.7667739391326904,
+      "learning_rate": 0.00019505643906575073,
+      "loss": 1.7693,
+      "step": 4770
+    },
+    {
+      "epoch": 1.0070047927529362,
+      "grad_norm": 0.8733193874359131,
+      "learning_rate": 0.00019503586301695183,
+      "loss": 1.7616,
+      "step": 4780
+    },
+    {
+      "epoch": 1.0091114973402855,
+      "grad_norm": 0.7520270943641663,
+      "learning_rate": 0.00019501524532613328,
+      "loss": 1.7154,
+      "step": 4790
+    },
+    {
+      "epoch": 1.0112182019276348,
+      "grad_norm": 0.8560175895690918,
+      "learning_rate": 0.0001949945860023292,
+      "loss": 1.7157,
+      "step": 4800
+    },
+    {
+      "epoch": 1.0133249065149839,
+      "grad_norm": 0.7532521486282349,
+      "learning_rate": 0.00019497388505459188,
+      "loss": 1.7118,
+      "step": 4810
+    },
+    {
+      "epoch": 1.0154316111023332,
+      "grad_norm": 0.7637246251106262,
+      "learning_rate": 0.00019495314249199194,
+      "loss": 1.7412,
+      "step": 4820
+    },
+    {
+      "epoch": 1.0175383156896824,
+      "grad_norm": 0.789354681968689,
+      "learning_rate": 0.0001949323583236181,
+      "loss": 1.739,
+      "step": 4830
+    },
+    {
+      "epoch": 1.0196450202770317,
+      "grad_norm": 0.8701115846633911,
+      "learning_rate": 0.00019491153255857748,
+      "loss": 1.7189,
+      "step": 4840
+    },
+    {
+      "epoch": 1.0217517248643808,
+      "grad_norm": 0.7776268720626831,
+      "learning_rate": 0.0001948906652059953,
+      "loss": 1.7785,
+      "step": 4850
+    },
+    {
+      "epoch": 1.0238584294517301,
+      "grad_norm": 0.8140941858291626,
+      "learning_rate": 0.00019486975627501502,
+      "loss": 1.7281,
+      "step": 4860
+    },
+    {
+      "epoch": 1.0259651340390794,
+      "grad_norm": 0.8021364808082581,
+      "learning_rate": 0.00019484880577479835,
+      "loss": 1.7601,
+      "step": 4870
+    },
+    {
+      "epoch": 1.0280718386264287,
+      "grad_norm": 0.8303508758544922,
+      "learning_rate": 0.00019482781371452524,
+      "loss": 1.7261,
+      "step": 4880
+    },
+    {
+      "epoch": 1.0301785432137778,
+      "grad_norm": 0.7315002083778381,
+      "learning_rate": 0.0001948067801033938,
+      "loss": 1.7464,
+      "step": 4890
+    },
+    {
+      "epoch": 1.032285247801127,
+      "grad_norm": 0.8632476329803467,
+      "learning_rate": 0.00019478570495062037,
+      "loss": 1.772,
+      "step": 4900
+    },
+    {
+      "epoch": 1.0343919523884764,
+      "grad_norm": 0.7853039503097534,
+      "learning_rate": 0.00019476458826543945,
+      "loss": 1.7331,
+      "step": 4910
+    },
+    {
+      "epoch": 1.0364986569758257,
+      "grad_norm": 0.7495045065879822,
+      "learning_rate": 0.0001947434300571038,
+      "loss": 1.7609,
+      "step": 4920
+    },
+    {
+      "epoch": 1.0386053615631747,
+      "grad_norm": 0.8120619654655457,
+      "learning_rate": 0.00019472223033488431,
+      "loss": 1.7377,
+      "step": 4930
+    },
+    {
+      "epoch": 1.040712066150524,
+      "grad_norm": 0.7205514311790466,
+      "learning_rate": 0.00019470098910807015,
+      "loss": 1.7342,
+      "step": 4940
+    },
+    {
+      "epoch": 1.0428187707378733,
+      "grad_norm": 0.8088648319244385,
+      "learning_rate": 0.0001946797063859686,
+      "loss": 1.7257,
+      "step": 4950
+    },
+    {
+      "epoch": 1.0449254753252226,
+      "grad_norm": 0.7667573094367981,
+      "learning_rate": 0.00019465838217790516,
+      "loss": 1.7089,
+      "step": 4960
+    },
+    {
+      "epoch": 1.0470321799125717,
+      "grad_norm": 0.8235226273536682,
+      "learning_rate": 0.00019463701649322343,
+      "loss": 1.7422,
+      "step": 4970
+    },
+    {
+      "epoch": 1.049138884499921,
+      "grad_norm": 0.8100693225860596,
+      "learning_rate": 0.00019461560934128533,
+      "loss": 1.7587,
+      "step": 4980
+    },
+    {
+      "epoch": 1.0512455890872703,
+      "grad_norm": 0.7639106512069702,
+      "learning_rate": 0.00019459416073147083,
+      "loss": 1.775,
+      "step": 4990
+    },
+    {
+      "epoch": 1.0533522936746196,
+      "grad_norm": 0.7470769286155701,
+      "learning_rate": 0.00019457267067317808,
+      "loss": 1.7528,
+      "step": 5000
+    },
+    {
+      "epoch": 1.0554589982619687,
+      "grad_norm": 0.7330575585365295,
+      "learning_rate": 0.00019455113917582346,
+      "loss": 1.725,
+      "step": 5010
+    },
+    {
+      "epoch": 1.057565702849318,
+      "grad_norm": 0.8469204902648926,
+      "learning_rate": 0.00019452956624884144,
+      "loss": 1.7515,
+      "step": 5020
+    },
+    {
+      "epoch": 1.0596724074366672,
+      "grad_norm": 0.8233253359794617,
+      "learning_rate": 0.00019450795190168466,
+      "loss": 1.7137,
+      "step": 5030
+    },
+    {
+      "epoch": 1.0617791120240165,
+      "grad_norm": 0.7793976664543152,
+      "learning_rate": 0.0001944862961438239,
+      "loss": 1.7661,
+      "step": 5040
+    },
+    {
+      "epoch": 1.0638858166113656,
+      "grad_norm": 0.748579204082489,
+      "learning_rate": 0.00019446459898474813,
+      "loss": 1.7186,
+      "step": 5050
+    },
+    {
+      "epoch": 1.065992521198715,
+      "grad_norm": 0.8824768662452698,
+      "learning_rate": 0.00019444286043396444,
+      "loss": 1.7756,
+      "step": 5060
+    },
+    {
+      "epoch": 1.0680992257860642,
+      "grad_norm": 0.8221313953399658,
+      "learning_rate": 0.000194421080500998,
+      "loss": 1.7871,
+      "step": 5070
+    },
+    {
+      "epoch": 1.0702059303734135,
+      "grad_norm": 0.7716003060340881,
+      "learning_rate": 0.0001943992591953922,
+      "loss": 1.7361,
+      "step": 5080
+    },
+    {
+      "epoch": 1.0723126349607626,
+      "grad_norm": 0.7575770616531372,
+      "learning_rate": 0.00019437739652670847,
+      "loss": 1.7479,
+      "step": 5090
+    },
+    {
+      "epoch": 1.0744193395481119,
+      "grad_norm": 0.8454105854034424,
+      "learning_rate": 0.00019435549250452645,
+      "loss": 1.7889,
+      "step": 5100
+    },
+    {
+      "epoch": 1.0765260441354612,
+      "grad_norm": 0.809290885925293,
+      "learning_rate": 0.00019433354713844386,
+      "loss": 1.8065,
+      "step": 5110
+    },
+    {
+      "epoch": 1.0786327487228102,
+      "grad_norm": 0.8279988765716553,
+      "learning_rate": 0.00019431156043807652,
+      "loss": 1.7569,
+      "step": 5120
+    },
+    {
+      "epoch": 1.0807394533101595,
+      "grad_norm": 0.8030558824539185,
+      "learning_rate": 0.00019428953241305838,
+      "loss": 1.7719,
+      "step": 5130
+    },
+    {
+      "epoch": 1.0828461578975088,
+      "grad_norm": 0.7470248341560364,
+      "learning_rate": 0.00019426746307304153,
+      "loss": 1.7697,
+      "step": 5140
+    },
+    {
+      "epoch": 1.0849528624848581,
+      "grad_norm": 0.8372868299484253,
+      "learning_rate": 0.00019424535242769606,
+      "loss": 1.7387,
+      "step": 5150
+    },
+    {
+      "epoch": 1.0870595670722074,
+      "grad_norm": 0.74366694688797,
+      "learning_rate": 0.0001942232004867103,
+      "loss": 1.7477,
+      "step": 5160
+    },
+    {
+      "epoch": 1.0891662716595565,
+      "grad_norm": 0.9802844524383545,
+      "learning_rate": 0.00019420100725979058,
+      "loss": 1.6741,
+      "step": 5170
+    },
+    {
+      "epoch": 1.0912729762469058,
+      "grad_norm": 0.9001051187515259,
+      "learning_rate": 0.0001941787727566613,
+      "loss": 1.7635,
+      "step": 5180
+    },
+    {
+      "epoch": 1.093379680834255,
+      "grad_norm": 0.8312892317771912,
+      "learning_rate": 0.00019415649698706507,
+      "loss": 1.7846,
+      "step": 5190
+    },
+    {
+      "epoch": 1.0954863854216041,
+      "grad_norm": 0.7780260443687439,
+      "learning_rate": 0.0001941341799607624,
+      "loss": 1.7114,
+      "step": 5200
+    },
+    {
+      "epoch": 1.0975930900089534,
+      "grad_norm": 0.7878116965293884,
+      "learning_rate": 0.00019411182168753205,
+      "loss": 1.7928,
+      "step": 5210
+    },
+    {
+      "epoch": 1.0996997945963027,
+      "grad_norm": 0.7883005738258362,
+      "learning_rate": 0.0001940894221771708,
+      "loss": 1.7648,
+      "step": 5220
+    },
+    {
+      "epoch": 1.101806499183652,
+      "grad_norm": 0.7752858400344849,
+      "learning_rate": 0.00019406698143949338,
+      "loss": 1.73,
+      "step": 5230
+    },
+    {
+      "epoch": 1.1039132037710013,
+      "grad_norm": 0.9342918992042542,
+      "learning_rate": 0.00019404449948433276,
+      "loss": 1.783,
+      "step": 5240
+    },
+    {
+      "epoch": 1.1060199083583504,
+      "grad_norm": 0.7494181394577026,
+      "learning_rate": 0.00019402197632153992,
+      "loss": 1.7905,
+      "step": 5250
+    },
+    {
+      "epoch": 1.1081266129456997,
+      "grad_norm": 0.7930409908294678,
+      "learning_rate": 0.0001939994119609838,
+      "loss": 1.7753,
+      "step": 5260
+    },
+    {
+      "epoch": 1.110233317533049,
+      "grad_norm": 0.7505367994308472,
+      "learning_rate": 0.0001939768064125515,
+      "loss": 1.7385,
+      "step": 5270
+    },
+    {
+      "epoch": 1.112340022120398,
+      "grad_norm": 0.912246823310852,
+      "learning_rate": 0.00019395415968614813,
+      "loss": 1.7465,
+      "step": 5280
+    },
+    {
+      "epoch": 1.1144467267077474,
+      "grad_norm": 0.8005505204200745,
+      "learning_rate": 0.00019393147179169685,
+      "loss": 1.7409,
+      "step": 5290
+    },
+    {
+      "epoch": 1.1165534312950967,
+      "grad_norm": 0.811210036277771,
+      "learning_rate": 0.00019390874273913884,
+      "loss": 1.7744,
+      "step": 5300
+    },
+    {
+      "epoch": 1.118660135882446,
+      "grad_norm": 0.7788394093513489,
+      "learning_rate": 0.00019388597253843334,
+      "loss": 1.78,
+      "step": 5310
+    },
+    {
+      "epoch": 1.120766840469795,
+      "grad_norm": 0.7960549592971802,
+      "learning_rate": 0.00019386316119955756,
+      "loss": 1.7695,
+      "step": 5320
+    },
+    {
+      "epoch": 1.1228735450571443,
+      "grad_norm": 0.7453601956367493,
+      "learning_rate": 0.00019384030873250682,
+      "loss": 1.7466,
+      "step": 5330
+    },
+    {
+      "epoch": 1.1249802496444936,
+      "grad_norm": 0.7400311231613159,
+      "learning_rate": 0.00019381741514729443,
+      "loss": 1.7426,
+      "step": 5340
+    },
+    {
+      "epoch": 1.127086954231843,
+      "grad_norm": 0.7591160535812378,
+      "learning_rate": 0.00019379448045395167,
+      "loss": 1.7722,
+      "step": 5350
+    },
+    {
+      "epoch": 1.129193658819192,
+      "grad_norm": 0.7780054211616516,
+      "learning_rate": 0.00019377150466252797,
+      "loss": 1.756,
+      "step": 5360
+    },
+    {
+      "epoch": 1.1313003634065413,
+      "grad_norm": 0.8573557734489441,
+      "learning_rate": 0.00019374848778309055,
+      "loss": 1.751,
+      "step": 5370
+    },
+    {
+      "epoch": 1.1334070679938906,
+      "grad_norm": 0.8030732870101929,
+      "learning_rate": 0.0001937254298257248,
+      "loss": 1.7632,
+      "step": 5380
+    },
+    {
+      "epoch": 1.1355137725812399,
+      "grad_norm": 0.8074278831481934,
+      "learning_rate": 0.00019370233080053407,
+      "loss": 1.7779,
+      "step": 5390
+    },
+    {
+      "epoch": 1.137620477168589,
+      "grad_norm": 0.7617718577384949,
+      "learning_rate": 0.0001936791907176397,
+      "loss": 1.7212,
+      "step": 5400
+    },
+    {
+      "epoch": 1.1397271817559382,
+      "grad_norm": 0.9997255802154541,
+      "learning_rate": 0.000193656009587181,
+      "loss": 1.684,
+      "step": 5410
+    },
+    {
+      "epoch": 1.1418338863432875,
+      "grad_norm": 0.9460029006004333,
+      "learning_rate": 0.0001936327874193153,
+      "loss": 1.8257,
+      "step": 5420
+    },
+    {
+      "epoch": 1.1439405909306368,
+      "grad_norm": 0.8431273102760315,
+      "learning_rate": 0.00019360952422421793,
+      "loss": 1.7567,
+      "step": 5430
+    },
+    {
+      "epoch": 1.146047295517986,
+      "grad_norm": 0.7903361320495605,
+      "learning_rate": 0.00019358622001208205,
+      "loss": 1.7924,
+      "step": 5440
+    },
+    {
+      "epoch": 1.1481540001053352,
+      "grad_norm": 0.8683080673217773,
+      "learning_rate": 0.00019356287479311903,
+      "loss": 1.8095,
+      "step": 5450
+    },
+    {
+      "epoch": 1.1502607046926845,
+      "grad_norm": 0.8753440976142883,
+      "learning_rate": 0.00019353948857755803,
+      "loss": 1.7625,
+      "step": 5460
+    },
+    {
+      "epoch": 1.1523674092800338,
+      "grad_norm": 0.9305017590522766,
+      "learning_rate": 0.0001935160613756462,
+      "loss": 1.7848,
+      "step": 5470
+    },
+    {
+      "epoch": 1.1544741138673829,
+      "grad_norm": 0.8456583023071289,
+      "learning_rate": 0.00019349259319764874,
+      "loss": 1.7629,
+      "step": 5480
+    },
+    {
+      "epoch": 1.1565808184547322,
+      "grad_norm": 0.8240710496902466,
+      "learning_rate": 0.00019346908405384867,
+      "loss": 1.7405,
+      "step": 5490
+    },
+    {
+      "epoch": 1.1586875230420814,
+      "grad_norm": 0.8952575922012329,
+      "learning_rate": 0.00019344553395454707,
+      "loss": 1.7578,
+      "step": 5500
+    },
+    {
+      "epoch": 1.1607942276294307,
+      "grad_norm": 0.8688398599624634,
+      "learning_rate": 0.00019342194291006295,
+      "loss": 1.7185,
+      "step": 5510
+    },
+    {
+      "epoch": 1.1629009322167798,
+      "grad_norm": 0.8205084800720215,
+      "learning_rate": 0.00019339831093073318,
+      "loss": 1.7478,
+      "step": 5520
+    },
+    {
+      "epoch": 1.165007636804129,
+      "grad_norm": 0.8122640252113342,
+      "learning_rate": 0.00019337463802691264,
+      "loss": 1.76,
+      "step": 5530
+    },
+    {
+      "epoch": 1.1671143413914784,
+      "grad_norm": 0.8044484257698059,
+      "learning_rate": 0.00019335092420897417,
+      "loss": 1.7624,
+      "step": 5540
+    },
+    {
+      "epoch": 1.1692210459788277,
+      "grad_norm": 0.8380444049835205,
+      "learning_rate": 0.0001933271694873084,
+      "loss": 1.7228,
+      "step": 5550
+    },
+    {
+      "epoch": 1.1713277505661768,
+      "grad_norm": 0.8047568202018738,
+      "learning_rate": 0.00019330337387232408,
+      "loss": 1.7534,
+      "step": 5560
+    },
+    {
+      "epoch": 1.173434455153526,
+      "grad_norm": 0.8417034149169922,
+      "learning_rate": 0.00019327953737444768,
+      "loss": 1.7706,
+      "step": 5570
+    },
+    {
+      "epoch": 1.1755411597408754,
+      "grad_norm": 0.8022201657295227,
+      "learning_rate": 0.00019325566000412376,
+      "loss": 1.7586,
+      "step": 5580
+    },
+    {
+      "epoch": 1.1776478643282247,
+      "grad_norm": 0.7806428074836731,
+      "learning_rate": 0.00019323174177181463,
+      "loss": 1.7009,
+      "step": 5590
+    },
+    {
+      "epoch": 1.1797545689155737,
+      "grad_norm": 0.7925698757171631,
+      "learning_rate": 0.00019320778268800066,
+      "loss": 1.724,
+      "step": 5600
+    },
+    {
+      "epoch": 1.181861273502923,
+      "grad_norm": 0.7943428158760071,
+      "learning_rate": 0.00019318378276318,
+      "loss": 1.7897,
+      "step": 5610
+    },
+    {
+      "epoch": 1.1839679780902723,
+      "grad_norm": 0.8270592093467712,
+      "learning_rate": 0.0001931597420078687,
+      "loss": 1.7325,
+      "step": 5620
+    },
+    {
+      "epoch": 1.1860746826776216,
+      "grad_norm": 0.7894768118858337,
+      "learning_rate": 0.00019313566043260082,
+      "loss": 1.7382,
+      "step": 5630
+    },
+    {
+      "epoch": 1.1881813872649707,
+      "grad_norm": 0.9318056702613831,
+      "learning_rate": 0.0001931115380479281,
+      "loss": 1.7403,
+      "step": 5640
+    },
+    {
+      "epoch": 1.19028809185232,
+      "grad_norm": 0.8299248218536377,
+      "learning_rate": 0.00019308737486442045,
+      "loss": 1.7449,
+      "step": 5650
+    },
+    {
+      "epoch": 1.1923947964396693,
+      "grad_norm": 0.8641102910041809,
+      "learning_rate": 0.00019306317089266535,
+      "loss": 1.7894,
+      "step": 5660
+    },
+    {
+      "epoch": 1.1945015010270186,
+      "grad_norm": 0.8760716319084167,
+      "learning_rate": 0.00019303892614326836,
+      "loss": 1.8106,
+      "step": 5670
+    },
+    {
+      "epoch": 1.1966082056143676,
+      "grad_norm": 0.8323030471801758,
+      "learning_rate": 0.00019301464062685284,
+      "loss": 1.6792,
+      "step": 5680
+    },
+    {
+      "epoch": 1.198714910201717,
+      "grad_norm": 0.8567124009132385,
+      "learning_rate": 0.00019299031435406,
+      "loss": 1.7333,
+      "step": 5690
+    },
+    {
+      "epoch": 1.2008216147890662,
+      "grad_norm": 0.8274356722831726,
+      "learning_rate": 0.00019296594733554892,
+      "loss": 1.7383,
+      "step": 5700
+    },
+    {
+      "epoch": 1.2029283193764155,
+      "grad_norm": 0.8282963037490845,
+      "learning_rate": 0.0001929415395819965,
+      "loss": 1.7545,
+      "step": 5710
+    },
+    {
+      "epoch": 1.2050350239637646,
+      "grad_norm": 0.9670590162277222,
+      "learning_rate": 0.00019291709110409762,
+      "loss": 1.7156,
+      "step": 5720
+    },
+    {
+      "epoch": 1.207141728551114,
+      "grad_norm": 0.892984926700592,
+      "learning_rate": 0.00019289260191256483,
+      "loss": 1.7138,
+      "step": 5730
+    },
+    {
+      "epoch": 1.2092484331384632,
+      "grad_norm": 0.8116103410720825,
+      "learning_rate": 0.00019286807201812867,
+      "loss": 1.7758,
+      "step": 5740
+    },
+    {
+      "epoch": 1.2113551377258125,
+      "grad_norm": 0.9112514853477478,
+      "learning_rate": 0.00019284350143153737,
+      "loss": 1.7703,
+      "step": 5750
+    },
+    {
+      "epoch": 1.2134618423131616,
+      "grad_norm": 0.9463765025138855,
+      "learning_rate": 0.0001928188901635571,
+      "loss": 1.7667,
+      "step": 5760
+    },
+    {
+      "epoch": 1.2155685469005109,
+      "grad_norm": 0.9577860832214355,
+      "learning_rate": 0.0001927942382249718,
+      "loss": 1.6904,
+      "step": 5770
+    },
+    {
+      "epoch": 1.2176752514878602,
+      "grad_norm": 0.7787550687789917,
+      "learning_rate": 0.00019276954562658327,
+      "loss": 1.7321,
+      "step": 5780
+    },
+    {
+      "epoch": 1.2197819560752095,
+      "grad_norm": 0.8212473392486572,
+      "learning_rate": 0.00019274481237921114,
+      "loss": 1.6719,
+      "step": 5790
+    },
+    {
+      "epoch": 1.2218886606625585,
+      "grad_norm": 0.8447032570838928,
+      "learning_rate": 0.00019272003849369273,
+      "loss": 1.7175,
+      "step": 5800
+    },
+    {
+      "epoch": 1.2239953652499078,
+      "grad_norm": 0.8335479497909546,
+      "learning_rate": 0.00019269522398088332,
+      "loss": 1.8072,
+      "step": 5810
+    },
+    {
+      "epoch": 1.2261020698372571,
+      "grad_norm": 0.8188406825065613,
+      "learning_rate": 0.00019267036885165588,
+      "loss": 1.8029,
+      "step": 5820
+    },
+    {
+      "epoch": 1.2282087744246062,
+      "grad_norm": 0.9411261677742004,
+      "learning_rate": 0.00019264547311690128,
+      "loss": 1.6936,
+      "step": 5830
+    },
+    {
+      "epoch": 1.2303154790119555,
+      "grad_norm": 0.8654270768165588,
+      "learning_rate": 0.0001926205367875281,
+      "loss": 1.74,
+      "step": 5840
+    },
+    {
+      "epoch": 1.2324221835993048,
+      "grad_norm": 0.8441559672355652,
+      "learning_rate": 0.0001925955598744627,
+      "loss": 1.7549,
+      "step": 5850
+    },
+    {
+      "epoch": 1.234528888186654,
+      "grad_norm": 0.820061206817627,
+      "learning_rate": 0.0001925705423886493,
+      "loss": 1.6773,
+      "step": 5860
+    },
+    {
+      "epoch": 1.2366355927740034,
+      "grad_norm": 0.7828805446624756,
+      "learning_rate": 0.00019254548434104985,
+      "loss": 1.7523,
+      "step": 5870
+    },
+    {
+      "epoch": 1.2387422973613524,
+      "grad_norm": 0.8304197192192078,
+      "learning_rate": 0.00019252038574264405,
+      "loss": 1.7367,
+      "step": 5880
+    },
+    {
+      "epoch": 1.2408490019487017,
+      "grad_norm": 0.8142679333686829,
+      "learning_rate": 0.0001924952466044294,
+      "loss": 1.7818,
+      "step": 5890
+    },
+    {
+      "epoch": 1.242955706536051,
+      "grad_norm": 0.8129520416259766,
+      "learning_rate": 0.00019247006693742113,
+      "loss": 1.783,
+      "step": 5900
+    },
+    {
+      "epoch": 1.2450624111234,
+      "grad_norm": 0.9476183652877808,
+      "learning_rate": 0.00019244484675265232,
+      "loss": 1.7761,
+      "step": 5910
+    },
+    {
+      "epoch": 1.2471691157107494,
+      "grad_norm": 0.8139955997467041,
+      "learning_rate": 0.00019241958606117373,
+      "loss": 1.7095,
+      "step": 5920
+    },
+    {
+      "epoch": 1.2492758202980987,
+      "grad_norm": 0.774530291557312,
+      "learning_rate": 0.00019239428487405382,
+      "loss": 1.7884,
+      "step": 5930
+    },
+    {
+      "epoch": 1.251382524885448,
+      "grad_norm": 0.7705311179161072,
+      "learning_rate": 0.00019236894320237894,
+      "loss": 1.7389,
+      "step": 5940
+    },
+    {
+      "epoch": 1.2534892294727973,
+      "grad_norm": 0.8137152194976807,
+      "learning_rate": 0.00019234356105725297,
+      "loss": 1.7329,
+      "step": 5950
+    },
+    {
+      "epoch": 1.2555959340601464,
+      "grad_norm": 0.8168460726737976,
+      "learning_rate": 0.00019231813844979777,
+      "loss": 1.7827,
+      "step": 5960
+    },
+    {
+      "epoch": 1.2577026386474957,
+      "grad_norm": 0.7712127566337585,
+      "learning_rate": 0.0001922926753911527,
+      "loss": 1.7593,
+      "step": 5970
+    },
+    {
+      "epoch": 1.259809343234845,
+      "grad_norm": 0.8201771974563599,
+      "learning_rate": 0.00019226717189247503,
+      "loss": 1.7471,
+      "step": 5980
+    },
+    {
+      "epoch": 1.261916047822194,
+      "grad_norm": 0.8837641477584839,
+      "learning_rate": 0.00019224162796493968,
+      "loss": 1.7401,
+      "step": 5990
+    },
+    {
+      "epoch": 1.2640227524095433,
+      "grad_norm": 0.8871500492095947,
+      "learning_rate": 0.00019221604361973919,
+      "loss": 1.7056,
+      "step": 6000
+    },
+    {
+      "epoch": 1.2661294569968926,
+      "grad_norm": 0.8061213493347168,
+      "learning_rate": 0.00019219041886808392,
+      "loss": 1.7369,
+      "step": 6010
+    },
+    {
+      "epoch": 1.268236161584242,
+      "grad_norm": 0.8275548219680786,
+      "learning_rate": 0.00019216475372120197,
+      "loss": 1.7914,
+      "step": 6020
+    },
+    {
+      "epoch": 1.2703428661715912,
+      "grad_norm": 0.8748604655265808,
+      "learning_rate": 0.00019213904819033903,
+      "loss": 1.696,
+      "step": 6030
+    },
+    {
+      "epoch": 1.2724495707589403,
+      "grad_norm": 0.9023202061653137,
+      "learning_rate": 0.00019211330228675855,
+      "loss": 1.7967,
+      "step": 6040
+    },
+    {
+      "epoch": 1.2745562753462896,
+      "grad_norm": 0.8412946462631226,
+      "learning_rate": 0.00019208751602174163,
+      "loss": 1.7548,
+      "step": 6050
+    },
+    {
+      "epoch": 1.2766629799336389,
+      "grad_norm": 0.8570536971092224,
+      "learning_rate": 0.00019206168940658712,
+      "loss": 1.7311,
+      "step": 6060
+    },
+    {
+      "epoch": 1.278769684520988,
+      "grad_norm": 0.8899838924407959,
+      "learning_rate": 0.00019203582245261148,
+      "loss": 1.7736,
+      "step": 6070
+    },
+    {
+      "epoch": 1.2808763891083372,
+      "grad_norm": 0.828941285610199,
+      "learning_rate": 0.0001920099151711489,
+      "loss": 1.7372,
+      "step": 6080
+    },
+    {
+      "epoch": 1.2829830936956865,
+      "grad_norm": 0.8117388486862183,
+      "learning_rate": 0.00019198396757355118,
+      "loss": 1.8239,
+      "step": 6090
+    },
+    {
+      "epoch": 1.2850897982830358,
+      "grad_norm": 0.9626308679580688,
+      "learning_rate": 0.00019195797967118785,
+      "loss": 1.7581,
+      "step": 6100
+    },
+    {
+      "epoch": 1.2871965028703851,
+      "grad_norm": 0.7775184512138367,
+      "learning_rate": 0.00019193195147544607,
+      "loss": 1.7391,
+      "step": 6110
+    },
+    {
+      "epoch": 1.2893032074577342,
+      "grad_norm": 0.7835120558738708,
+      "learning_rate": 0.00019190588299773062,
+      "loss": 1.7508,
+      "step": 6120
+    },
+    {
+      "epoch": 1.2914099120450835,
+      "grad_norm": 0.8474375605583191,
+      "learning_rate": 0.00019187977424946405,
+      "loss": 1.8031,
+      "step": 6130
+    },
+    {
+      "epoch": 1.2935166166324328,
+      "grad_norm": 0.8936663269996643,
+      "learning_rate": 0.0001918536252420864,
+      "loss": 1.6969,
+      "step": 6140
+    },
+    {
+      "epoch": 1.2956233212197819,
+      "grad_norm": 0.7543231844902039,
+      "learning_rate": 0.00019182743598705542,
+      "loss": 1.7573,
+      "step": 6150
+    },
+    {
+      "epoch": 1.2977300258071311,
+      "grad_norm": 0.7518043518066406,
+      "learning_rate": 0.00019180120649584653,
+      "loss": 1.7773,
+      "step": 6160
+    },
+    {
+      "epoch": 1.2998367303944804,
+      "grad_norm": 0.7887030243873596,
+      "learning_rate": 0.00019177493677995276,
+      "loss": 1.7501,
+      "step": 6170
+    },
+    {
+      "epoch": 1.3019434349818297,
+      "grad_norm": 0.7868638038635254,
+      "learning_rate": 0.00019174862685088472,
+      "loss": 1.7665,
+      "step": 6180
+    },
+    {
+      "epoch": 1.304050139569179,
+      "grad_norm": 0.8161253929138184,
+      "learning_rate": 0.0001917222767201707,
+      "loss": 1.76,
+      "step": 6190
+    },
+    {
+      "epoch": 1.306156844156528,
+      "grad_norm": 0.8441837430000305,
+      "learning_rate": 0.00019169588639935658,
+      "loss": 1.7396,
+      "step": 6200
+    },
+    {
+      "epoch": 1.3082635487438774,
+      "grad_norm": 0.8492400050163269,
+      "learning_rate": 0.00019166945590000584,
+      "loss": 1.7393,
+      "step": 6210
+    },
+    {
+      "epoch": 1.3103702533312267,
+      "grad_norm": 0.8949893712997437,
+      "learning_rate": 0.00019164298523369956,
+      "loss": 1.7993,
+      "step": 6220
+    },
+    {
+      "epoch": 1.3124769579185758,
+      "grad_norm": 0.8137794733047485,
+      "learning_rate": 0.00019161647441203646,
+      "loss": 1.7052,
+      "step": 6230
+    },
+    {
+      "epoch": 1.314583662505925,
+      "grad_norm": 0.8320265412330627,
+      "learning_rate": 0.0001915899234466328,
+      "loss": 1.7183,
+      "step": 6240
+    },
+    {
+      "epoch": 1.3166903670932744,
+      "grad_norm": 0.759221613407135,
+      "learning_rate": 0.00019156333234912252,
+      "loss": 1.7243,
+      "step": 6250
+    },
+    {
+      "epoch": 1.3187970716806237,
+      "grad_norm": 0.8010006546974182,
+      "learning_rate": 0.00019153670113115703,
+      "loss": 1.7496,
+      "step": 6260
+    },
+    {
+      "epoch": 1.320903776267973,
+      "grad_norm": 0.8241715431213379,
+      "learning_rate": 0.0001915100298044054,
+      "loss": 1.7967,
+      "step": 6270
+    },
+    {
+      "epoch": 1.323010480855322,
+      "grad_norm": 0.9702495336532593,
+      "learning_rate": 0.00019148331838055423,
+      "loss": 1.7791,
+      "step": 6280
+    },
+    {
+      "epoch": 1.3251171854426713,
+      "grad_norm": 0.814001739025116,
+      "learning_rate": 0.0001914565668713077,
+      "loss": 1.7507,
+      "step": 6290
+    },
+    {
+      "epoch": 1.3272238900300206,
+      "grad_norm": 0.8167542815208435,
+      "learning_rate": 0.00019142977528838762,
+      "loss": 1.7606,
+      "step": 6300
+    },
+    {
+      "epoch": 1.3293305946173697,
+      "grad_norm": 0.8223824501037598,
+      "learning_rate": 0.00019140294364353324,
+      "loss": 1.6936,
+      "step": 6310
+    },
+    {
+      "epoch": 1.331437299204719,
+      "grad_norm": 0.8717813491821289,
+      "learning_rate": 0.00019137607194850146,
+      "loss": 1.7051,
+      "step": 6320
+    },
+    {
+      "epoch": 1.3335440037920683,
+      "grad_norm": 0.8642485737800598,
+      "learning_rate": 0.00019134916021506666,
+      "loss": 1.7115,
+      "step": 6330
+    },
+    {
+      "epoch": 1.3356507083794176,
+      "grad_norm": 0.851425290107727,
+      "learning_rate": 0.00019132220845502086,
+      "loss": 1.7759,
+      "step": 6340
+    },
+    {
+      "epoch": 1.3377574129667669,
+      "grad_norm": 0.8192858099937439,
+      "learning_rate": 0.00019129521668017347,
+      "loss": 1.7782,
+      "step": 6350
+    },
+    {
+      "epoch": 1.339864117554116,
+      "grad_norm": 0.8982135057449341,
+      "learning_rate": 0.0001912681849023516,
+      "loss": 1.7843,
+      "step": 6360
+    },
+    {
+      "epoch": 1.3419708221414652,
+      "grad_norm": 0.8544314503669739,
+      "learning_rate": 0.00019124111313339976,
+      "loss": 1.7804,
+      "step": 6370
+    },
+    {
+      "epoch": 1.3440775267288145,
+      "grad_norm": 0.8088465929031372,
+      "learning_rate": 0.00019121400138518007,
+      "loss": 1.7797,
+      "step": 6380
+    },
+    {
+      "epoch": 1.3461842313161636,
+      "grad_norm": 0.8711520433425903,
+      "learning_rate": 0.00019118684966957207,
+      "loss": 1.6855,
+      "step": 6390
+    },
+    {
+      "epoch": 1.348290935903513,
+      "grad_norm": 0.8697302937507629,
+      "learning_rate": 0.00019115965799847292,
+      "loss": 1.7711,
+      "step": 6400
+    },
+    {
+      "epoch": 1.3503976404908622,
+      "grad_norm": 0.7881453037261963,
+      "learning_rate": 0.00019113242638379725,
+      "loss": 1.7652,
+      "step": 6410
+    },
+    {
+      "epoch": 1.3525043450782115,
+      "grad_norm": 0.8739869594573975,
+      "learning_rate": 0.00019110515483747716,
+      "loss": 1.7741,
+      "step": 6420
+    },
+    {
+      "epoch": 1.3546110496655608,
+      "grad_norm": 0.9500364661216736,
+      "learning_rate": 0.0001910778433714622,
+      "loss": 1.7113,
+      "step": 6430
+    },
+    {
+      "epoch": 1.3567177542529099,
+      "grad_norm": 0.8580523729324341,
+      "learning_rate": 0.00019105049199771962,
+      "loss": 1.7662,
+      "step": 6440
+    },
+    {
+      "epoch": 1.3588244588402592,
+      "grad_norm": 0.7940788865089417,
+      "learning_rate": 0.00019102310072823393,
+      "loss": 1.7336,
+      "step": 6450
+    },
+    {
+      "epoch": 1.3609311634276082,
+      "grad_norm": 0.80447918176651,
+      "learning_rate": 0.0001909956695750072,
+      "loss": 1.6681,
+      "step": 6460
+    },
+    {
+      "epoch": 1.3630378680149575,
+      "grad_norm": 0.8043936491012573,
+      "learning_rate": 0.000190968198550059,
+      "loss": 1.7039,
+      "step": 6470
+    },
+    {
+      "epoch": 1.3651445726023068,
+      "grad_norm": 0.8359491229057312,
+      "learning_rate": 0.0001909406876654264,
+      "loss": 1.711,
+      "step": 6480
+    },
+    {
+      "epoch": 1.3672512771896561,
+      "grad_norm": 0.8909677267074585,
+      "learning_rate": 0.00019091313693316383,
+      "loss": 1.7488,
+      "step": 6490
+    },
+    {
+      "epoch": 1.3693579817770054,
+      "grad_norm": 0.9276780486106873,
+      "learning_rate": 0.00019088554636534323,
+      "loss": 1.7168,
+      "step": 6500
+    },
+    {
+      "epoch": 1.3714646863643545,
+      "grad_norm": 0.8368661403656006,
+      "learning_rate": 0.00019085791597405404,
+      "loss": 1.8085,
+      "step": 6510
+    },
+    {
+      "epoch": 1.3735713909517038,
+      "grad_norm": 1.0867797136306763,
+      "learning_rate": 0.0001908302457714031,
+      "loss": 1.7569,
+      "step": 6520
+    },
+    {
+      "epoch": 1.375678095539053,
+      "grad_norm": 0.7681411504745483,
+      "learning_rate": 0.00019080253576951473,
+      "loss": 1.7583,
+      "step": 6530
+    },
+    {
+      "epoch": 1.3777848001264021,
+      "grad_norm": 0.8406849503517151,
+      "learning_rate": 0.00019077478598053063,
+      "loss": 1.7632,
+      "step": 6540
+    },
+    {
+      "epoch": 1.3798915047137514,
+      "grad_norm": 0.771959125995636,
+      "learning_rate": 0.00019074699641661,
+      "loss": 1.6977,
+      "step": 6550
+    },
+    {
+      "epoch": 1.3819982093011007,
+      "grad_norm": 0.8810305595397949,
+      "learning_rate": 0.00019071916708992943,
+      "loss": 1.7648,
+      "step": 6560
+    },
+    {
+      "epoch": 1.38410491388845,
+      "grad_norm": 0.7824996709823608,
+      "learning_rate": 0.00019069129801268294,
+      "loss": 1.8061,
+      "step": 6570
+    },
+    {
+      "epoch": 1.3862116184757993,
+      "grad_norm": 0.822420060634613,
+      "learning_rate": 0.00019066338919708197,
+      "loss": 1.7184,
+      "step": 6580
+    },
+    {
+      "epoch": 1.3883183230631484,
+      "grad_norm": 0.8344022035598755,
+      "learning_rate": 0.00019063544065535534,
+      "loss": 1.7274,
+      "step": 6590
+    },
+    {
+      "epoch": 1.3904250276504977,
+      "grad_norm": 0.8080721497535706,
+      "learning_rate": 0.00019060745239974936,
+      "loss": 1.7137,
+      "step": 6600
+    },
+    {
+      "epoch": 1.392531732237847,
+      "grad_norm": 0.8242797255516052,
+      "learning_rate": 0.0001905794244425277,
+      "loss": 1.7193,
+      "step": 6610
+    },
+    {
+      "epoch": 1.394638436825196,
+      "grad_norm": 0.8020373582839966,
+      "learning_rate": 0.00019055135679597136,
+      "loss": 1.7942,
+      "step": 6620
+    },
+    {
+      "epoch": 1.3967451414125454,
+      "grad_norm": 0.8053699731826782,
+      "learning_rate": 0.0001905232494723788,
+      "loss": 1.7781,
+      "step": 6630
+    },
+    {
+      "epoch": 1.3988518459998946,
+      "grad_norm": 0.8109655380249023,
+      "learning_rate": 0.00019049510248406586,
+      "loss": 1.7311,
+      "step": 6640
+    },
+    {
+      "epoch": 1.400958550587244,
+      "grad_norm": 0.8344601392745972,
+      "learning_rate": 0.00019046691584336577,
+      "loss": 1.7843,
+      "step": 6650
+    },
+    {
+      "epoch": 1.4030652551745932,
+      "grad_norm": 0.7590084075927734,
+      "learning_rate": 0.0001904386895626291,
+      "loss": 1.7729,
+      "step": 6660
+    },
+    {
+      "epoch": 1.4051719597619423,
+      "grad_norm": 0.7660609483718872,
+      "learning_rate": 0.0001904104236542238,
+      "loss": 1.7351,
+      "step": 6670
+    },
+    {
+      "epoch": 1.4072786643492916,
+      "grad_norm": 0.8946515917778015,
+      "learning_rate": 0.0001903821181305352,
+      "loss": 1.7527,
+      "step": 6680
+    },
+    {
+      "epoch": 1.409385368936641,
+      "grad_norm": 0.8283473253250122,
+      "learning_rate": 0.00019035377300396597,
+      "loss": 1.7567,
+      "step": 6690
+    },
+    {
+      "epoch": 1.41149207352399,
+      "grad_norm": 0.7549485564231873,
+      "learning_rate": 0.00019032538828693616,
+      "loss": 1.7559,
+      "step": 6700
+    },
+    {
+      "epoch": 1.4135987781113393,
+      "grad_norm": 0.8404794931411743,
+      "learning_rate": 0.0001902969639918831,
+      "loss": 1.7181,
+      "step": 6710
+    },
+    {
+      "epoch": 1.4157054826986886,
+      "grad_norm": 0.826153576374054,
+      "learning_rate": 0.00019026850013126157,
+      "loss": 1.7724,
+      "step": 6720
+    },
+    {
+      "epoch": 1.4178121872860379,
+      "grad_norm": 0.8652869462966919,
+      "learning_rate": 0.0001902399967175436,
+      "loss": 1.6797,
+      "step": 6730
+    },
+    {
+      "epoch": 1.4199188918733872,
+      "grad_norm": 0.8194015622138977,
+      "learning_rate": 0.00019021145376321857,
+      "loss": 1.7175,
+      "step": 6740
+    },
+    {
+      "epoch": 1.4220255964607362,
+      "grad_norm": 0.8753089904785156,
+      "learning_rate": 0.0001901828712807932,
+      "loss": 1.756,
+      "step": 6750
+    },
+    {
+      "epoch": 1.4241323010480855,
+      "grad_norm": 0.7799382209777832,
+      "learning_rate": 0.0001901542492827915,
+      "loss": 1.7679,
+      "step": 6760
+    },
+    {
+      "epoch": 1.4262390056354348,
+      "grad_norm": 0.805819571018219,
+      "learning_rate": 0.00019012558778175485,
+      "loss": 1.7729,
+      "step": 6770
+    },
+    {
+      "epoch": 1.428345710222784,
+      "grad_norm": 0.7864932417869568,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 1.7852,
+      "step": 6780
+    },
+    {
+      "epoch": 1.4304524148101332,
+      "grad_norm": 0.8187804222106934,
+      "learning_rate": 0.00019006814632082863,
+      "loss": 1.6702,
+      "step": 6790
+    },
+    {
+      "epoch": 1.4325591193974825,
+      "grad_norm": 0.7939268946647644,
+      "learning_rate": 0.00019003936638610828,
+      "loss": 1.6653,
+      "step": 6800
+    },
+    {
+      "epoch": 1.4346658239848318,
+      "grad_norm": 0.8125925660133362,
+      "learning_rate": 0.00019001054699869133,
+      "loss": 1.7386,
+      "step": 6810
+    },
+    {
+      "epoch": 1.436772528572181,
+      "grad_norm": 0.8372388482093811,
+      "learning_rate": 0.00018998168817120577,
+      "loss": 1.7444,
+      "step": 6820
+    },
+    {
+      "epoch": 1.4388792331595301,
+      "grad_norm": 0.83929842710495,
+      "learning_rate": 0.00018995278991629658,
+      "loss": 1.751,
+      "step": 6830
+    },
+    {
+      "epoch": 1.4409859377468794,
+      "grad_norm": 0.7986190319061279,
+      "learning_rate": 0.00018992385224662623,
+      "loss": 1.746,
+      "step": 6840
+    },
+    {
+      "epoch": 1.4430926423342287,
+      "grad_norm": 0.8739672303199768,
+      "learning_rate": 0.00018989487517487435,
+      "loss": 1.7579,
+      "step": 6850
+    },
+    {
+      "epoch": 1.4451993469215778,
+      "grad_norm": 0.9071808457374573,
+      "learning_rate": 0.0001898658587137379,
+      "loss": 1.7034,
+      "step": 6860
+    },
+    {
+      "epoch": 1.447306051508927,
+      "grad_norm": 0.8445550799369812,
+      "learning_rate": 0.00018983680287593105,
+      "loss": 1.7639,
+      "step": 6870
+    },
+    {
+      "epoch": 1.4494127560962764,
+      "grad_norm": 0.8497149348258972,
+      "learning_rate": 0.00018980770767418526,
+      "loss": 1.7695,
+      "step": 6880
+    },
+    {
+      "epoch": 1.4515194606836257,
+      "grad_norm": 0.7921140789985657,
+      "learning_rate": 0.00018977857312124923,
+      "loss": 1.7391,
+      "step": 6890
+    },
+    {
+      "epoch": 1.453626165270975,
+      "grad_norm": 0.7450403571128845,
+      "learning_rate": 0.00018974939922988883,
+      "loss": 1.7423,
+      "step": 6900
+    },
+    {
+      "epoch": 1.455732869858324,
+      "grad_norm": 0.8325957655906677,
+      "learning_rate": 0.0001897201860128873,
+      "loss": 1.79,
+      "step": 6910
+    },
+    {
+      "epoch": 1.4578395744456734,
+      "grad_norm": 0.9251476526260376,
+      "learning_rate": 0.00018969093348304505,
+      "loss": 1.7301,
+      "step": 6920
+    },
+    {
+      "epoch": 1.4599462790330227,
+      "grad_norm": 0.8292146325111389,
+      "learning_rate": 0.00018966164165317966,
+      "loss": 1.7738,
+      "step": 6930
+    },
+    {
+      "epoch": 1.4620529836203717,
+      "grad_norm": 0.8368434309959412,
+      "learning_rate": 0.000189632310536126,
+      "loss": 1.7129,
+      "step": 6940
+    },
+    {
+      "epoch": 1.464159688207721,
+      "grad_norm": 0.8482025861740112,
+      "learning_rate": 0.00018960294014473615,
+      "loss": 1.778,
+      "step": 6950
+    },
+    {
+      "epoch": 1.4662663927950703,
+      "grad_norm": 0.823621392250061,
+      "learning_rate": 0.00018957353049187936,
+      "loss": 1.7401,
+      "step": 6960
+    },
+    {
+      "epoch": 1.4683730973824196,
+      "grad_norm": 0.7859732508659363,
+      "learning_rate": 0.0001895440815904421,
+      "loss": 1.7523,
+      "step": 6970
+    },
+    {
+      "epoch": 1.470479801969769,
+      "grad_norm": 0.9153429865837097,
+      "learning_rate": 0.00018951459345332807,
+      "loss": 1.7763,
+      "step": 6980
+    },
+    {
+      "epoch": 1.472586506557118,
+      "grad_norm": 0.8432143926620483,
+      "learning_rate": 0.00018948506609345813,
+      "loss": 1.733,
+      "step": 6990
+    },
+    {
+      "epoch": 1.4746932111444673,
+      "grad_norm": 0.8629891276359558,
+      "learning_rate": 0.0001894554995237703,
+      "loss": 1.7545,
+      "step": 7000
+    },
+    {
+      "epoch": 1.4767999157318166,
+      "grad_norm": 0.8362427353858948,
+      "learning_rate": 0.00018942589375721985,
+      "loss": 1.7524,
+      "step": 7010
+    },
+    {
+      "epoch": 1.4789066203191656,
+      "grad_norm": 0.7721174359321594,
+      "learning_rate": 0.00018939624880677918,
+      "loss": 1.7503,
+      "step": 7020
+    },
+    {
+      "epoch": 1.481013324906515,
+      "grad_norm": 0.8513792753219604,
+      "learning_rate": 0.00018936656468543784,
+      "loss": 1.7612,
+      "step": 7030
+    },
+    {
+      "epoch": 1.4831200294938642,
+      "grad_norm": 0.804923415184021,
+      "learning_rate": 0.00018933684140620257,
+      "loss": 1.74,
+      "step": 7040
+    },
+    {
+      "epoch": 1.4852267340812135,
+      "grad_norm": 0.88444584608078,
+      "learning_rate": 0.00018930707898209733,
+      "loss": 1.7494,
+      "step": 7050
+    },
+    {
+      "epoch": 1.4873334386685628,
+      "grad_norm": 0.8222391605377197,
+      "learning_rate": 0.00018927727742616313,
+      "loss": 1.7207,
+      "step": 7060
+    },
+    {
+      "epoch": 1.489440143255912,
+      "grad_norm": 0.8316041231155396,
+      "learning_rate": 0.00018924743675145813,
+      "loss": 1.7651,
+      "step": 7070
+    },
+    {
+      "epoch": 1.4915468478432612,
+      "grad_norm": 0.847623348236084,
+      "learning_rate": 0.0001892175569710577,
+      "loss": 1.8181,
+      "step": 7080
+    },
+    {
+      "epoch": 1.4936535524306105,
+      "grad_norm": 0.8623965382575989,
+      "learning_rate": 0.00018918763809805435,
+      "loss": 1.6863,
+      "step": 7090
+    },
+    {
+      "epoch": 1.4957602570179596,
+      "grad_norm": 0.882828950881958,
+      "learning_rate": 0.00018915768014555762,
+      "loss": 1.7619,
+      "step": 7100
+    },
+    {
+      "epoch": 1.4978669616053089,
+      "grad_norm": 0.9039724469184875,
+      "learning_rate": 0.00018912768312669424,
+      "loss": 1.7102,
+      "step": 7110
+    },
+    {
+      "epoch": 1.4999736661926582,
+      "grad_norm": 0.7769089341163635,
+      "learning_rate": 0.0001890976470546081,
+      "loss": 1.7189,
+      "step": 7120
+    },
+    {
+      "epoch": 1.5020803707800074,
+      "grad_norm": 0.7899691462516785,
+      "learning_rate": 0.00018906757194246012,
+      "loss": 1.7211,
+      "step": 7130
+    },
+    {
+      "epoch": 1.5041870753673567,
+      "grad_norm": 0.7718929052352905,
+      "learning_rate": 0.00018903745780342839,
+      "loss": 1.6544,
+      "step": 7140
+    },
+    {
+      "epoch": 1.5062937799547058,
+      "grad_norm": 0.8393471837043762,
+      "learning_rate": 0.00018900730465070802,
+      "loss": 1.7581,
+      "step": 7150
+    },
+    {
+      "epoch": 1.508400484542055,
+      "grad_norm": 0.845622181892395,
+      "learning_rate": 0.0001889771124975113,
+      "loss": 1.6994,
+      "step": 7160
+    },
+    {
+      "epoch": 1.5105071891294042,
+      "grad_norm": 0.7832208275794983,
+      "learning_rate": 0.0001889468813570676,
+      "loss": 1.7505,
+      "step": 7170
+    },
+    {
+      "epoch": 1.5126138937167535,
+      "grad_norm": 0.8510347604751587,
+      "learning_rate": 0.00018891661124262327,
+      "loss": 1.7472,
+      "step": 7180
+    },
+    {
+      "epoch": 1.5147205983041028,
+      "grad_norm": 0.8204689621925354,
+      "learning_rate": 0.00018888630216744193,
+      "loss": 1.7411,
+      "step": 7190
+    },
+    {
+      "epoch": 1.516827302891452,
+      "grad_norm": 0.852558434009552,
+      "learning_rate": 0.00018885595414480405,
+      "loss": 1.7592,
+      "step": 7200
+    },
+    {
+      "epoch": 1.5189340074788014,
+      "grad_norm": 0.8407373428344727,
+      "learning_rate": 0.0001888255671880073,
+      "loss": 1.7416,
+      "step": 7210
+    },
+    {
+      "epoch": 1.5210407120661507,
+      "grad_norm": 0.7926289439201355,
+      "learning_rate": 0.00018879514131036644,
+      "loss": 1.7593,
+      "step": 7220
+    },
+    {
+      "epoch": 1.5231474166534997,
+      "grad_norm": 0.9050236940383911,
+      "learning_rate": 0.00018876467652521317,
+      "loss": 1.7416,
+      "step": 7230
+    },
+    {
+      "epoch": 1.525254121240849,
+      "grad_norm": 0.9141523241996765,
+      "learning_rate": 0.00018873417284589629,
+      "loss": 1.7551,
+      "step": 7240
+    },
+    {
+      "epoch": 1.527360825828198,
+      "grad_norm": 0.8591190576553345,
+      "learning_rate": 0.00018870363028578168,
+      "loss": 1.7524,
+      "step": 7250
+    },
+    {
+      "epoch": 1.5294675304155474,
+      "grad_norm": 0.8687646389007568,
+      "learning_rate": 0.0001886730488582522,
+      "loss": 1.6681,
+      "step": 7260
+    },
+    {
+      "epoch": 1.5315742350028967,
+      "grad_norm": 0.8512703776359558,
+      "learning_rate": 0.0001886424285767078,
+      "loss": 1.8276,
+      "step": 7270
+    },
+    {
+      "epoch": 1.533680939590246,
+      "grad_norm": 0.8147086501121521,
+      "learning_rate": 0.0001886117694545654,
+      "loss": 1.7959,
+      "step": 7280
+    },
+    {
+      "epoch": 1.5357876441775953,
+      "grad_norm": 0.8092646598815918,
+      "learning_rate": 0.0001885810715052589,
+      "loss": 1.7359,
+      "step": 7290
+    },
+    {
+      "epoch": 1.5378943487649446,
+      "grad_norm": 0.8645735383033752,
+      "learning_rate": 0.00018855033474223936,
+      "loss": 1.7274,
+      "step": 7300
+    },
+    {
+      "epoch": 1.5400010533522936,
+      "grad_norm": 0.7911597490310669,
+      "learning_rate": 0.0001885195591789747,
+      "loss": 1.7509,
+      "step": 7310
+    },
+    {
+      "epoch": 1.542107757939643,
+      "grad_norm": 0.8070617914199829,
+      "learning_rate": 0.00018848874482894993,
+      "loss": 1.8309,
+      "step": 7320
+    },
+    {
+      "epoch": 1.544214462526992,
+      "grad_norm": 0.8313998579978943,
+      "learning_rate": 0.00018845789170566702,
+      "loss": 1.776,
+      "step": 7330
+    },
+    {
+      "epoch": 1.5463211671143413,
+      "grad_norm": 0.7963941097259521,
+      "learning_rate": 0.00018842699982264492,
+      "loss": 1.7448,
+      "step": 7340
+    },
+    {
+      "epoch": 1.5484278717016906,
+      "grad_norm": 0.8401038646697998,
+      "learning_rate": 0.0001883960691934196,
+      "loss": 1.7501,
+      "step": 7350
+    },
+    {
+      "epoch": 1.55053457628904,
+      "grad_norm": 0.8527158498764038,
+      "learning_rate": 0.0001883650998315439,
+      "loss": 1.734,
+      "step": 7360
+    },
+    {
+      "epoch": 1.5526412808763892,
+      "grad_norm": 0.8697671294212341,
+      "learning_rate": 0.00018833409175058786,
+      "loss": 1.7513,
+      "step": 7370
+    },
+    {
+      "epoch": 1.5547479854637385,
+      "grad_norm": 0.7973361015319824,
+      "learning_rate": 0.00018830304496413822,
+      "loss": 1.7337,
+      "step": 7380
+    },
+    {
+      "epoch": 1.5568546900510876,
+      "grad_norm": 0.9423846006393433,
+      "learning_rate": 0.00018827195948579886,
+      "loss": 1.6702,
+      "step": 7390
+    },
+    {
+      "epoch": 1.5589613946384369,
+      "grad_norm": 0.8513593673706055,
+      "learning_rate": 0.00018824083532919056,
+      "loss": 1.7395,
+      "step": 7400
+    },
+    {
+      "epoch": 1.561068099225786,
+      "grad_norm": 0.9370805025100708,
+      "learning_rate": 0.000188209672507951,
+      "loss": 1.7967,
+      "step": 7410
+    },
+    {
+      "epoch": 1.5631748038131352,
+      "grad_norm": 0.8711183667182922,
+      "learning_rate": 0.00018817847103573486,
+      "loss": 1.7464,
+      "step": 7420
+    },
+    {
+      "epoch": 1.5652815084004845,
+      "grad_norm": 0.8438371419906616,
+      "learning_rate": 0.00018814723092621375,
+      "loss": 1.7417,
+      "step": 7430
+    },
+    {
+      "epoch": 1.5673882129878338,
+      "grad_norm": 0.7928584218025208,
+      "learning_rate": 0.00018811595219307622,
+      "loss": 1.729,
+      "step": 7440
+    },
+    {
+      "epoch": 1.5694949175751831,
+      "grad_norm": 1.1181422472000122,
+      "learning_rate": 0.00018808463485002767,
+      "loss": 1.784,
+      "step": 7450
+    },
+    {
+      "epoch": 1.5716016221625324,
+      "grad_norm": 0.7837269902229309,
+      "learning_rate": 0.00018805327891079055,
+      "loss": 1.7821,
+      "step": 7460
+    },
+    {
+      "epoch": 1.5737083267498815,
+      "grad_norm": 0.8478876948356628,
+      "learning_rate": 0.00018802188438910405,
+      "loss": 1.7542,
+      "step": 7470
+    },
+    {
+      "epoch": 1.5758150313372308,
+      "grad_norm": 0.8075322508811951,
+      "learning_rate": 0.0001879904512987244,
+      "loss": 1.7268,
+      "step": 7480
+    },
+    {
+      "epoch": 1.5779217359245798,
+      "grad_norm": 1.1477035284042358,
+      "learning_rate": 0.00018795897965342474,
+      "loss": 1.759,
+      "step": 7490
+    },
+    {
+      "epoch": 1.5800284405119291,
+      "grad_norm": 0.9190003275871277,
+      "learning_rate": 0.000187927469466995,
+      "loss": 1.8034,
+      "step": 7500
+    },
+    {
+      "epoch": 1.5821351450992784,
+      "grad_norm": 0.8050284385681152,
+      "learning_rate": 0.00018789592075324203,
+      "loss": 1.7384,
+      "step": 7510
+    },
+    {
+      "epoch": 1.5842418496866277,
+      "grad_norm": 0.9296058416366577,
+      "learning_rate": 0.0001878643335259896,
+      "loss": 1.747,
+      "step": 7520
+    },
+    {
+      "epoch": 1.586348554273977,
+      "grad_norm": 0.8164471983909607,
+      "learning_rate": 0.00018783270779907838,
+      "loss": 1.7186,
+      "step": 7530
+    },
+    {
+      "epoch": 1.5884552588613263,
+      "grad_norm": 0.8429021239280701,
+      "learning_rate": 0.0001878010435863658,
+      "loss": 1.7842,
+      "step": 7540
+    },
+    {
+      "epoch": 1.5905619634486754,
+      "grad_norm": 0.8415732979774475,
+      "learning_rate": 0.00018776934090172627,
+      "loss": 1.7409,
+      "step": 7550
+    },
+    {
+      "epoch": 1.5926686680360247,
+      "grad_norm": 0.8081998229026794,
+      "learning_rate": 0.00018773759975905098,
+      "loss": 1.7561,
+      "step": 7560
+    },
+    {
+      "epoch": 1.5947753726233738,
+      "grad_norm": 0.8396217226982117,
+      "learning_rate": 0.00018770582017224802,
+      "loss": 1.7832,
+      "step": 7570
+    },
+    {
+      "epoch": 1.596882077210723,
+      "grad_norm": 0.9396729469299316,
+      "learning_rate": 0.0001876740021552423,
+      "loss": 1.7721,
+      "step": 7580
+    },
+    {
+      "epoch": 1.5989887817980724,
+      "grad_norm": 0.8377107977867126,
+      "learning_rate": 0.00018764214572197552,
+      "loss": 1.7757,
+      "step": 7590
+    },
+    {
+      "epoch": 1.6010954863854217,
+      "grad_norm": 0.8287121653556824,
+      "learning_rate": 0.00018761025088640632,
+      "loss": 1.7319,
+      "step": 7600
+    },
+    {
+      "epoch": 1.603202190972771,
+      "grad_norm": 0.9031893610954285,
+      "learning_rate": 0.00018757831766251016,
+      "loss": 1.7783,
+      "step": 7610
+    },
+    {
+      "epoch": 1.6053088955601202,
+      "grad_norm": 0.8598982095718384,
+      "learning_rate": 0.00018754634606427914,
+      "loss": 1.8231,
+      "step": 7620
+    },
+    {
+      "epoch": 1.6074156001474693,
+      "grad_norm": 0.8003443479537964,
+      "learning_rate": 0.00018751433610572242,
+      "loss": 1.7699,
+      "step": 7630
+    },
+    {
+      "epoch": 1.6095223047348186,
+      "grad_norm": 0.8151317834854126,
+      "learning_rate": 0.00018748228780086579,
+      "loss": 1.765,
+      "step": 7640
+    },
+    {
+      "epoch": 1.6116290093221677,
+      "grad_norm": 0.825056791305542,
+      "learning_rate": 0.00018745020116375197,
+      "loss": 1.7585,
+      "step": 7650
+    },
+    {
+      "epoch": 1.613735713909517,
+      "grad_norm": 0.8297179937362671,
+      "learning_rate": 0.00018741807620844037,
+      "loss": 1.7263,
+      "step": 7660
+    },
+    {
+      "epoch": 1.6158424184968663,
+      "grad_norm": 0.8049089312553406,
+      "learning_rate": 0.0001873859129490072,
+      "loss": 1.7333,
+      "step": 7670
+    },
+    {
+      "epoch": 1.6179491230842156,
+      "grad_norm": 0.817482590675354,
+      "learning_rate": 0.00018735371139954558,
+      "loss": 1.6743,
+      "step": 7680
+    },
+    {
+      "epoch": 1.6200558276715649,
+      "grad_norm": 0.8606418371200562,
+      "learning_rate": 0.00018732147157416527,
+      "loss": 1.7371,
+      "step": 7690
+    },
+    {
+      "epoch": 1.6221625322589142,
+      "grad_norm": 0.8428048491477966,
+      "learning_rate": 0.00018728919348699283,
+      "loss": 1.7423,
+      "step": 7700
+    },
+    {
+      "epoch": 1.6242692368462632,
+      "grad_norm": 0.9453518986701965,
+      "learning_rate": 0.00018725687715217163,
+      "loss": 1.7329,
+      "step": 7710
+    },
+    {
+      "epoch": 1.6263759414336125,
+      "grad_norm": 0.8569479584693909,
+      "learning_rate": 0.0001872245225838617,
+      "loss": 1.7658,
+      "step": 7720
+    },
+    {
+      "epoch": 1.6284826460209616,
+      "grad_norm": 0.8008100390434265,
+      "learning_rate": 0.00018719212979624,
+      "loss": 1.7586,
+      "step": 7730
+    },
+    {
+      "epoch": 1.630589350608311,
+      "grad_norm": 0.8791933655738831,
+      "learning_rate": 0.0001871596988035001,
+      "loss": 1.8345,
+      "step": 7740
+    },
+    {
+      "epoch": 1.6326960551956602,
+      "grad_norm": 0.848496675491333,
+      "learning_rate": 0.0001871272296198523,
+      "loss": 1.7011,
+      "step": 7750
+    },
+    {
+      "epoch": 1.6348027597830095,
+      "grad_norm": 0.8242083191871643,
+      "learning_rate": 0.0001870947222595237,
+      "loss": 1.7555,
+      "step": 7760
+    },
+    {
+      "epoch": 1.6369094643703588,
+      "grad_norm": 0.8481186032295227,
+      "learning_rate": 0.00018706217673675811,
+      "loss": 1.7048,
+      "step": 7770
+    },
+    {
+      "epoch": 1.6390161689577079,
+      "grad_norm": 0.8180390000343323,
+      "learning_rate": 0.00018702959306581604,
+      "loss": 1.6859,
+      "step": 7780
+    },
+    {
+      "epoch": 1.6411228735450571,
+      "grad_norm": 0.8697353601455688,
+      "learning_rate": 0.00018699697126097476,
+      "loss": 1.7215,
+      "step": 7790
+    },
+    {
+      "epoch": 1.6432295781324062,
+      "grad_norm": 0.8706187605857849,
+      "learning_rate": 0.00018696431133652817,
+      "loss": 1.6866,
+      "step": 7800
+    },
+    {
+      "epoch": 1.6453362827197555,
+      "grad_norm": 0.8431823253631592,
+      "learning_rate": 0.00018693161330678696,
+      "loss": 1.675,
+      "step": 7810
+    },
+    {
+      "epoch": 1.6474429873071048,
+      "grad_norm": 0.8259839415550232,
+      "learning_rate": 0.0001868988771860785,
+      "loss": 1.7079,
+      "step": 7820
+    },
+    {
+      "epoch": 1.649549691894454,
+      "grad_norm": 0.805834949016571,
+      "learning_rate": 0.00018686610298874676,
+      "loss": 1.7089,
+      "step": 7830
+    },
+    {
+      "epoch": 1.6516563964818034,
+      "grad_norm": 0.7511940598487854,
+      "learning_rate": 0.00018683329072915252,
+      "loss": 1.719,
+      "step": 7840
+    },
+    {
+      "epoch": 1.6537631010691527,
+      "grad_norm": 0.8287107348442078,
+      "learning_rate": 0.00018680044042167318,
+      "loss": 1.7804,
+      "step": 7850
+    },
+    {
+      "epoch": 1.6558698056565018,
+      "grad_norm": 0.8526592254638672,
+      "learning_rate": 0.00018676755208070275,
+      "loss": 1.7506,
+      "step": 7860
+    },
+    {
+      "epoch": 1.657976510243851,
+      "grad_norm": 0.8048674464225769,
+      "learning_rate": 0.00018673462572065205,
+      "loss": 1.7624,
+      "step": 7870
+    },
+    {
+      "epoch": 1.6600832148312001,
+      "grad_norm": 0.8441417813301086,
+      "learning_rate": 0.0001867016613559484,
+      "loss": 1.7617,
+      "step": 7880
+    },
+    {
+      "epoch": 1.6621899194185494,
+      "grad_norm": 0.8121191263198853,
+      "learning_rate": 0.00018666865900103597,
+      "loss": 1.7087,
+      "step": 7890
+    },
+    {
+      "epoch": 1.6642966240058987,
+      "grad_norm": 0.8484655618667603,
+      "learning_rate": 0.00018663561867037534,
+      "loss": 1.7656,
+      "step": 7900
+    },
+    {
+      "epoch": 1.666403328593248,
+      "grad_norm": 0.9858533143997192,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 1.6809,
+      "step": 7910
+    },
+    {
+      "epoch": 1.6685100331805973,
+      "grad_norm": 0.7125934362411499,
+      "learning_rate": 0.00018656942413973555,
+      "loss": 1.7411,
+      "step": 7920
+    },
+    {
+      "epoch": 1.6706167377679466,
+      "grad_norm": 0.9220843315124512,
+      "learning_rate": 0.000186536269968761,
+      "loss": 1.8265,
+      "step": 7930
+    },
+    {
+      "epoch": 1.6727234423552957,
+      "grad_norm": 0.934085488319397,
+      "learning_rate": 0.00018650307788004735,
+      "loss": 1.7728,
+      "step": 7940
+    },
+    {
+      "epoch": 1.674830146942645,
+      "grad_norm": 0.8358553647994995,
+      "learning_rate": 0.00018646984788813856,
+      "loss": 1.7261,
+      "step": 7950
+    },
+    {
+      "epoch": 1.676936851529994,
+      "grad_norm": 0.8404299020767212,
+      "learning_rate": 0.00018643658000759493,
+      "loss": 1.7528,
+      "step": 7960
+    },
+    {
+      "epoch": 1.6790435561173433,
+      "grad_norm": 0.8132160305976868,
+      "learning_rate": 0.00018640327425299363,
+      "loss": 1.7522,
+      "step": 7970
+    },
+    {
+      "epoch": 1.6811502607046926,
+      "grad_norm": 0.826701819896698,
+      "learning_rate": 0.0001863699306389282,
+      "loss": 1.7083,
+      "step": 7980
+    },
+    {
+      "epoch": 1.683256965292042,
+      "grad_norm": 0.899210512638092,
+      "learning_rate": 0.00018633654918000892,
+      "loss": 1.7517,
+      "step": 7990
+    },
+    {
+      "epoch": 1.6853636698793912,
+      "grad_norm": 0.7902103662490845,
+      "learning_rate": 0.00018630312989086257,
+      "loss": 1.7133,
+      "step": 8000
+    },
+    {
+      "epoch": 1.6874703744667405,
+      "grad_norm": 0.8423092365264893,
+      "learning_rate": 0.00018626967278613253,
+      "loss": 1.7686,
+      "step": 8010
+    },
+    {
+      "epoch": 1.6895770790540896,
+      "grad_norm": 0.8402092456817627,
+      "learning_rate": 0.0001862361778804788,
+      "loss": 1.737,
+      "step": 8020
+    },
+    {
+      "epoch": 1.691683783641439,
+      "grad_norm": 0.8797973394393921,
+      "learning_rate": 0.00018620264518857786,
+      "loss": 1.7052,
+      "step": 8030
+    },
+    {
+      "epoch": 1.693790488228788,
+      "grad_norm": 0.8236359357833862,
+      "learning_rate": 0.0001861690747251228,
+      "loss": 1.7043,
+      "step": 8040
+    },
+    {
+      "epoch": 1.6958971928161373,
+      "grad_norm": 0.8558099269866943,
+      "learning_rate": 0.00018613546650482322,
+      "loss": 1.7806,
+      "step": 8050
+    },
+    {
+      "epoch": 1.6980038974034866,
+      "grad_norm": 0.8771076798439026,
+      "learning_rate": 0.00018610182054240533,
+      "loss": 1.7177,
+      "step": 8060
+    },
+    {
+      "epoch": 1.7001106019908359,
+      "grad_norm": 0.8418282270431519,
+      "learning_rate": 0.0001860681368526118,
+      "loss": 1.7573,
+      "step": 8070
+    },
+    {
+      "epoch": 1.7022173065781852,
+      "grad_norm": 0.8841100931167603,
+      "learning_rate": 0.00018603441545020187,
+      "loss": 1.7451,
+      "step": 8080
+    },
+    {
+      "epoch": 1.7043240111655344,
+      "grad_norm": 0.8179870247840881,
+      "learning_rate": 0.00018600065634995135,
+      "loss": 1.7527,
+      "step": 8090
+    },
+    {
+      "epoch": 1.7064307157528835,
+      "grad_norm": 0.7647489905357361,
+      "learning_rate": 0.00018596685956665245,
+      "loss": 1.7789,
+      "step": 8100
+    },
+    {
+      "epoch": 1.7085374203402328,
+      "grad_norm": 0.8438899517059326,
+      "learning_rate": 0.000185933025115114,
+      "loss": 1.7211,
+      "step": 8110
+    },
+    {
+      "epoch": 1.7106441249275819,
+      "grad_norm": 0.8097826242446899,
+      "learning_rate": 0.0001858991530101613,
+      "loss": 1.7124,
+      "step": 8120
+    },
+    {
+      "epoch": 1.7127508295149312,
+      "grad_norm": 0.902351438999176,
+      "learning_rate": 0.00018586524326663615,
+      "loss": 1.7037,
+      "step": 8130
+    },
+    {
+      "epoch": 1.7148575341022805,
+      "grad_norm": 0.806341290473938,
+      "learning_rate": 0.0001858312958993968,
+      "loss": 1.7221,
+      "step": 8140
+    },
+    {
+      "epoch": 1.7169642386896298,
+      "grad_norm": 0.8351038098335266,
+      "learning_rate": 0.00018579731092331807,
+      "loss": 1.6798,
+      "step": 8150
+    },
+    {
+      "epoch": 1.719070943276979,
+      "grad_norm": 0.8565513491630554,
+      "learning_rate": 0.00018576328835329117,
+      "loss": 1.7356,
+      "step": 8160
+    },
+    {
+      "epoch": 1.7211776478643284,
+      "grad_norm": 0.8509708046913147,
+      "learning_rate": 0.00018572922820422387,
+      "loss": 1.7857,
+      "step": 8170
+    },
+    {
+      "epoch": 1.7232843524516774,
+      "grad_norm": 0.8077223300933838,
+      "learning_rate": 0.00018569513049104033,
+      "loss": 1.7137,
+      "step": 8180
+    },
+    {
+      "epoch": 1.7253910570390267,
+      "grad_norm": 0.9218313097953796,
+      "learning_rate": 0.00018566099522868119,
+      "loss": 1.7443,
+      "step": 8190
+    },
+    {
+      "epoch": 1.7274977616263758,
+      "grad_norm": 0.828586995601654,
+      "learning_rate": 0.00018562682243210358,
+      "loss": 1.7148,
+      "step": 8200
+    },
+    {
+      "epoch": 1.729604466213725,
+      "grad_norm": 0.966937243938446,
+      "learning_rate": 0.00018559261211628108,
+      "loss": 1.7608,
+      "step": 8210
+    },
+    {
+      "epoch": 1.7317111708010744,
+      "grad_norm": 0.8322204351425171,
+      "learning_rate": 0.00018555836429620358,
+      "loss": 1.7076,
+      "step": 8220
+    },
+    {
+      "epoch": 1.7338178753884237,
+      "grad_norm": 0.8872451782226562,
+      "learning_rate": 0.00018552407898687762,
+      "loss": 1.7599,
+      "step": 8230
+    },
+    {
+      "epoch": 1.735924579975773,
+      "grad_norm": 0.8485720753669739,
+      "learning_rate": 0.00018548975620332598,
+      "loss": 1.6958,
+      "step": 8240
+    },
+    {
+      "epoch": 1.7380312845631223,
+      "grad_norm": 0.839801549911499,
+      "learning_rate": 0.00018545539596058795,
+      "loss": 1.7315,
+      "step": 8250
+    },
+    {
+      "epoch": 1.7401379891504714,
+      "grad_norm": 0.8265219926834106,
+      "learning_rate": 0.0001854209982737192,
+      "loss": 1.7349,
+      "step": 8260
+    },
+    {
+      "epoch": 1.7422446937378206,
+      "grad_norm": 0.8101875185966492,
+      "learning_rate": 0.00018538656315779183,
+      "loss": 1.7198,
+      "step": 8270
+    },
+    {
+      "epoch": 1.7443513983251697,
+      "grad_norm": 0.8307563662528992,
+      "learning_rate": 0.00018535209062789433,
+      "loss": 1.7425,
+      "step": 8280
+    },
+    {
+      "epoch": 1.746458102912519,
+      "grad_norm": 0.8362199068069458,
+      "learning_rate": 0.00018531758069913158,
+      "loss": 1.7387,
+      "step": 8290
+    },
+    {
+      "epoch": 1.7485648074998683,
+      "grad_norm": 0.8315666913986206,
+      "learning_rate": 0.00018528303338662488,
+      "loss": 1.7549,
+      "step": 8300
+    },
+    {
+      "epoch": 1.7506715120872176,
+      "grad_norm": 0.9023050665855408,
+      "learning_rate": 0.00018524844870551185,
+      "loss": 1.7211,
+      "step": 8310
+    },
+    {
+      "epoch": 1.752778216674567,
+      "grad_norm": 0.8308394551277161,
+      "learning_rate": 0.00018521382667094656,
+      "loss": 1.7989,
+      "step": 8320
+    },
+    {
+      "epoch": 1.7548849212619162,
+      "grad_norm": 0.8728174567222595,
+      "learning_rate": 0.0001851791672980993,
+      "loss": 1.7407,
+      "step": 8330
+    },
+    {
+      "epoch": 1.7569916258492653,
+      "grad_norm": 0.8924462795257568,
+      "learning_rate": 0.00018514447060215698,
+      "loss": 1.7964,
+      "step": 8340
+    },
+    {
+      "epoch": 1.7590983304366146,
+      "grad_norm": 0.8875755071640015,
+      "learning_rate": 0.00018510973659832257,
+      "loss": 1.7254,
+      "step": 8350
+    },
+    {
+      "epoch": 1.7612050350239636,
+      "grad_norm": 0.9017901420593262,
+      "learning_rate": 0.0001850749653018156,
+      "loss": 1.7919,
+      "step": 8360
+    },
+    {
+      "epoch": 1.763311739611313,
+      "grad_norm": 0.8241222500801086,
+      "learning_rate": 0.00018504015672787184,
+      "loss": 1.7515,
+      "step": 8370
+    },
+    {
+      "epoch": 1.7654184441986622,
+      "grad_norm": 0.8747454881668091,
+      "learning_rate": 0.00018500531089174341,
+      "loss": 1.6982,
+      "step": 8380
+    },
+    {
+      "epoch": 1.7675251487860115,
+      "grad_norm": 0.8906333446502686,
+      "learning_rate": 0.00018497042780869883,
+      "loss": 1.7056,
+      "step": 8390
+    },
+    {
+      "epoch": 1.7696318533733608,
+      "grad_norm": 0.797450840473175,
+      "learning_rate": 0.00018493550749402278,
+      "loss": 1.7383,
+      "step": 8400
+    },
+    {
+      "epoch": 1.7717385579607101,
+      "grad_norm": 0.813721239566803,
+      "learning_rate": 0.0001849005499630164,
+      "loss": 1.8067,
+      "step": 8410
+    },
+    {
+      "epoch": 1.7738452625480592,
+      "grad_norm": 0.8121708631515503,
+      "learning_rate": 0.00018486555523099712,
+      "loss": 1.7114,
+      "step": 8420
+    },
+    {
+      "epoch": 1.7759519671354085,
+      "grad_norm": 0.8621687293052673,
+      "learning_rate": 0.00018483052331329857,
+      "loss": 1.6977,
+      "step": 8430
+    },
+    {
+      "epoch": 1.7780586717227576,
+      "grad_norm": 0.8001519441604614,
+      "learning_rate": 0.00018479545422527084,
+      "loss": 1.7607,
+      "step": 8440
+    },
+    {
+      "epoch": 1.7801653763101069,
+      "grad_norm": 0.8188821077346802,
+      "learning_rate": 0.0001847603479822801,
+      "loss": 1.7479,
+      "step": 8450
+    },
+    {
+      "epoch": 1.7822720808974561,
+      "grad_norm": 0.8462778329849243,
+      "learning_rate": 0.00018472520459970898,
+      "loss": 1.719,
+      "step": 8460
+    },
+    {
+      "epoch": 1.7843787854848054,
+      "grad_norm": 0.8654899597167969,
+      "learning_rate": 0.00018469002409295628,
+      "loss": 1.7461,
+      "step": 8470
+    },
+    {
+      "epoch": 1.7864854900721547,
+      "grad_norm": 0.8919845819473267,
+      "learning_rate": 0.0001846548064774371,
+      "loss": 1.7215,
+      "step": 8480
+    },
+    {
+      "epoch": 1.788592194659504,
+      "grad_norm": 0.8800879120826721,
+      "learning_rate": 0.00018461955176858285,
+      "loss": 1.7195,
+      "step": 8490
+    },
+    {
+      "epoch": 1.790698899246853,
+      "grad_norm": 0.7641480565071106,
+      "learning_rate": 0.00018458425998184113,
+      "loss": 1.6894,
+      "step": 8500
+    },
+    {
+      "epoch": 1.7928056038342024,
+      "grad_norm": 0.7839869856834412,
+      "learning_rate": 0.00018454893113267572,
+      "loss": 1.6647,
+      "step": 8510
+    },
+    {
+      "epoch": 1.7949123084215515,
+      "grad_norm": 0.8127973079681396,
+      "learning_rate": 0.0001845135652365668,
+      "loss": 1.8154,
+      "step": 8520
+    },
+    {
+      "epoch": 1.7970190130089008,
+      "grad_norm": 0.8238587379455566,
+      "learning_rate": 0.00018447816230901068,
+      "loss": 1.6919,
+      "step": 8530
+    },
+    {
+      "epoch": 1.79912571759625,
+      "grad_norm": 0.9203738570213318,
+      "learning_rate": 0.0001844427223655199,
+      "loss": 1.761,
+      "step": 8540
+    },
+    {
+      "epoch": 1.8012324221835994,
+      "grad_norm": 0.8604192137718201,
+      "learning_rate": 0.00018440724542162328,
+      "loss": 1.8097,
+      "step": 8550
+    },
+    {
+      "epoch": 1.8033391267709487,
+      "grad_norm": 0.8560807704925537,
+      "learning_rate": 0.0001843717314928658,
+      "loss": 1.7408,
+      "step": 8560
+    },
+    {
+      "epoch": 1.805445831358298,
+      "grad_norm": 0.8069592118263245,
+      "learning_rate": 0.00018433618059480864,
+      "loss": 1.7691,
+      "step": 8570
+    },
+    {
+      "epoch": 1.807552535945647,
+      "grad_norm": 0.8375011086463928,
+      "learning_rate": 0.00018430059274302917,
+      "loss": 1.7214,
+      "step": 8580
+    },
+    {
+      "epoch": 1.8096592405329963,
+      "grad_norm": 0.8408818244934082,
+      "learning_rate": 0.000184264967953121,
+      "loss": 1.7175,
+      "step": 8590
+    },
+    {
+      "epoch": 1.8117659451203454,
+      "grad_norm": 0.8032023310661316,
+      "learning_rate": 0.00018422930624069396,
+      "loss": 1.7363,
+      "step": 8600
+    },
+    {
+      "epoch": 1.8138726497076947,
+      "grad_norm": 0.8840620517730713,
+      "learning_rate": 0.00018419360762137395,
+      "loss": 1.7359,
+      "step": 8610
+    },
+    {
+      "epoch": 1.815979354295044,
+      "grad_norm": 0.869219958782196,
+      "learning_rate": 0.00018415787211080304,
+      "loss": 1.7537,
+      "step": 8620
+    },
+    {
+      "epoch": 1.8180860588823933,
+      "grad_norm": 0.8339371085166931,
+      "learning_rate": 0.0001841220997246396,
+      "loss": 1.7803,
+      "step": 8630
+    },
+    {
+      "epoch": 1.8201927634697426,
+      "grad_norm": 0.8503332734107971,
+      "learning_rate": 0.00018408629047855804,
+      "loss": 1.764,
+      "step": 8640
+    },
+    {
+      "epoch": 1.8222994680570916,
+      "grad_norm": 0.8225341439247131,
+      "learning_rate": 0.00018405044438824897,
+      "loss": 1.7486,
+      "step": 8650
+    },
+    {
+      "epoch": 1.824406172644441,
+      "grad_norm": 0.8652873635292053,
+      "learning_rate": 0.00018401456146941908,
+      "loss": 1.6888,
+      "step": 8660
+    },
+    {
+      "epoch": 1.82651287723179,
+      "grad_norm": 0.8248763084411621,
+      "learning_rate": 0.00018397864173779133,
+      "loss": 1.7245,
+      "step": 8670
+    },
+    {
+      "epoch": 1.8286195818191393,
+      "grad_norm": 0.8583594560623169,
+      "learning_rate": 0.00018394268520910466,
+      "loss": 1.7382,
+      "step": 8680
+    },
+    {
+      "epoch": 1.8307262864064886,
+      "grad_norm": 0.8216450214385986,
+      "learning_rate": 0.00018390669189911427,
+      "loss": 1.7708,
+      "step": 8690
+    },
+    {
+      "epoch": 1.832832990993838,
+      "grad_norm": 0.8396652340888977,
+      "learning_rate": 0.00018387066182359133,
+      "loss": 1.7057,
+      "step": 8700
+    },
+    {
+      "epoch": 1.8349396955811872,
+      "grad_norm": 0.8246045708656311,
+      "learning_rate": 0.00018383459499832322,
+      "loss": 1.7709,
+      "step": 8710
+    },
+    {
+      "epoch": 1.8370464001685365,
+      "grad_norm": 0.8746206164360046,
+      "learning_rate": 0.00018379849143911343,
+      "loss": 1.7462,
+      "step": 8720
+    },
+    {
+      "epoch": 1.8391531047558856,
+      "grad_norm": 1.081039547920227,
+      "learning_rate": 0.00018376235116178148,
+      "loss": 1.7569,
+      "step": 8730
+    },
+    {
+      "epoch": 1.8412598093432349,
+      "grad_norm": 0.9471474289894104,
+      "learning_rate": 0.00018372617418216307,
+      "loss": 1.7173,
+      "step": 8740
+    },
+    {
+      "epoch": 1.843366513930584,
+      "grad_norm": 0.848264217376709,
+      "learning_rate": 0.00018368996051610986,
+      "loss": 1.7653,
+      "step": 8750
+    },
+    {
+      "epoch": 1.8454732185179332,
+      "grad_norm": 0.8978562951087952,
+      "learning_rate": 0.00018365371017948964,
+      "loss": 1.7484,
+      "step": 8760
+    },
+    {
+      "epoch": 1.8475799231052825,
+      "grad_norm": 0.8349031209945679,
+      "learning_rate": 0.00018361742318818638,
+      "loss": 1.7595,
+      "step": 8770
+    },
+    {
+      "epoch": 1.8496866276926318,
+      "grad_norm": 0.8227035403251648,
+      "learning_rate": 0.00018358109955809993,
+      "loss": 1.7236,
+      "step": 8780
+    },
+    {
+      "epoch": 1.851793332279981,
+      "grad_norm": 0.7719805240631104,
+      "learning_rate": 0.0001835447393051463,
+      "loss": 1.7278,
+      "step": 8790
+    },
+    {
+      "epoch": 1.8539000368673304,
+      "grad_norm": 1.0519185066223145,
+      "learning_rate": 0.00018350834244525749,
+      "loss": 1.7136,
+      "step": 8800
+    },
+    {
+      "epoch": 1.8560067414546795,
+      "grad_norm": 0.9757561683654785,
+      "learning_rate": 0.00018347190899438158,
+      "loss": 1.7359,
+      "step": 8810
+    },
+    {
+      "epoch": 1.8581134460420288,
+      "grad_norm": 0.7864118814468384,
+      "learning_rate": 0.00018343543896848273,
+      "loss": 1.6817,
+      "step": 8820
+    },
+    {
+      "epoch": 1.8602201506293778,
+      "grad_norm": 0.8366915583610535,
+      "learning_rate": 0.000183398932383541,
+      "loss": 1.7644,
+      "step": 8830
+    },
+    {
+      "epoch": 1.8623268552167271,
+      "grad_norm": 0.8808780312538147,
+      "learning_rate": 0.00018336238925555263,
+      "loss": 1.7644,
+      "step": 8840
+    },
+    {
+      "epoch": 1.8644335598040764,
+      "grad_norm": 0.8752515316009521,
+      "learning_rate": 0.00018332580960052965,
+      "loss": 1.7058,
+      "step": 8850
+    },
+    {
+      "epoch": 1.8665402643914257,
+      "grad_norm": 0.8687223792076111,
+      "learning_rate": 0.00018328919343450035,
+      "loss": 1.7352,
+      "step": 8860
+    },
+    {
+      "epoch": 1.868646968978775,
+      "grad_norm": 0.9593966603279114,
+      "learning_rate": 0.00018325254077350883,
+      "loss": 1.7985,
+      "step": 8870
+    },
+    {
+      "epoch": 1.8707536735661243,
+      "grad_norm": 0.8275451064109802,
+      "learning_rate": 0.00018321585163361527,
+      "loss": 1.7322,
+      "step": 8880
+    },
+    {
+      "epoch": 1.8728603781534734,
+      "grad_norm": 0.8880124092102051,
+      "learning_rate": 0.0001831791260308958,
+      "loss": 1.7701,
+      "step": 8890
+    },
+    {
+      "epoch": 1.8749670827408227,
+      "grad_norm": 0.7356016635894775,
+      "learning_rate": 0.00018314236398144255,
+      "loss": 1.6926,
+      "step": 8900
+    },
+    {
+      "epoch": 1.8770737873281718,
+      "grad_norm": 0.869474470615387,
+      "learning_rate": 0.00018310556550136357,
+      "loss": 1.7184,
+      "step": 8910
+    },
+    {
+      "epoch": 1.879180491915521,
+      "grad_norm": 0.8896689414978027,
+      "learning_rate": 0.00018306873060678296,
+      "loss": 1.6887,
+      "step": 8920
+    },
+    {
+      "epoch": 1.8812871965028704,
+      "grad_norm": 0.9539408683776855,
+      "learning_rate": 0.0001830318593138407,
+      "loss": 1.7284,
+      "step": 8930
+    },
+    {
+      "epoch": 1.8833939010902196,
+      "grad_norm": 0.8492235541343689,
+      "learning_rate": 0.00018299495163869275,
+      "loss": 1.7262,
+      "step": 8940
+    },
+    {
+      "epoch": 1.885500605677569,
+      "grad_norm": 0.8203418850898743,
+      "learning_rate": 0.00018295800759751102,
+      "loss": 1.7769,
+      "step": 8950
+    },
+    {
+      "epoch": 1.8876073102649182,
+      "grad_norm": 0.7786878943443298,
+      "learning_rate": 0.00018292102720648333,
+      "loss": 1.7299,
+      "step": 8960
+    },
+    {
+      "epoch": 1.8897140148522673,
+      "grad_norm": 0.8030612468719482,
+      "learning_rate": 0.0001828840104818134,
+      "loss": 1.7691,
+      "step": 8970
+    },
+    {
+      "epoch": 1.8918207194396166,
+      "grad_norm": 0.9526339769363403,
+      "learning_rate": 0.00018284695743972095,
+      "loss": 1.739,
+      "step": 8980
+    },
+    {
+      "epoch": 1.8939274240269657,
+      "grad_norm": 0.8131811022758484,
+      "learning_rate": 0.00018280986809644158,
+      "loss": 1.7498,
+      "step": 8990
+    },
+    {
+      "epoch": 1.896034128614315,
+      "grad_norm": 0.8120954632759094,
+      "learning_rate": 0.0001827727424682268,
+      "loss": 1.7462,
+      "step": 9000
+    },
+    {
+      "epoch": 1.8981408332016643,
+      "grad_norm": 0.7906435132026672,
+      "learning_rate": 0.00018273558057134393,
+      "loss": 1.7311,
+      "step": 9010
+    },
+    {
+      "epoch": 1.9002475377890136,
+      "grad_norm": 0.9683308601379395,
+      "learning_rate": 0.00018269838242207636,
+      "loss": 1.7781,
+      "step": 9020
+    },
+    {
+      "epoch": 1.9023542423763629,
+      "grad_norm": 0.8596441745758057,
+      "learning_rate": 0.00018266114803672318,
+      "loss": 1.7434,
+      "step": 9030
+    },
+    {
+      "epoch": 1.9044609469637122,
+      "grad_norm": 0.7930140495300293,
+      "learning_rate": 0.0001826238774315995,
+      "loss": 1.7232,
+      "step": 9040
+    },
+    {
+      "epoch": 1.9065676515510612,
+      "grad_norm": 0.8552681803703308,
+      "learning_rate": 0.00018258657062303623,
+      "loss": 1.8003,
+      "step": 9050
+    },
+    {
+      "epoch": 1.9086743561384105,
+      "grad_norm": 0.954117476940155,
+      "learning_rate": 0.00018254922762738008,
+      "loss": 1.7098,
+      "step": 9060
+    },
+    {
+      "epoch": 1.9107810607257596,
+      "grad_norm": 0.9021719098091125,
+      "learning_rate": 0.00018251184846099382,
+      "loss": 1.7276,
+      "step": 9070
+    },
+    {
+      "epoch": 1.912887765313109,
+      "grad_norm": 0.8462482690811157,
+      "learning_rate": 0.00018247443314025583,
+      "loss": 1.6948,
+      "step": 9080
+    },
+    {
+      "epoch": 1.9149944699004582,
+      "grad_norm": 0.8283345699310303,
+      "learning_rate": 0.00018243698168156054,
+      "loss": 1.6674,
+      "step": 9090
+    },
+    {
+      "epoch": 1.9171011744878075,
+      "grad_norm": 0.8391633629798889,
+      "learning_rate": 0.00018239949410131802,
+      "loss": 1.7516,
+      "step": 9100
+    },
+    {
+      "epoch": 1.9192078790751568,
+      "grad_norm": 0.7927471995353699,
+      "learning_rate": 0.00018236197041595432,
+      "loss": 1.7364,
+      "step": 9110
+    },
+    {
+      "epoch": 1.921314583662506,
+      "grad_norm": 0.8896411061286926,
+      "learning_rate": 0.00018232441064191125,
+      "loss": 1.7549,
+      "step": 9120
+    },
+    {
+      "epoch": 1.9234212882498551,
+      "grad_norm": 0.8045653700828552,
+      "learning_rate": 0.00018228681479564644,
+      "loss": 1.7284,
+      "step": 9130
+    },
+    {
+      "epoch": 1.9255279928372044,
+      "grad_norm": 0.8324005603790283,
+      "learning_rate": 0.0001822491828936333,
+      "loss": 1.6741,
+      "step": 9140
+    },
+    {
+      "epoch": 1.9276346974245535,
+      "grad_norm": 0.7971271276473999,
+      "learning_rate": 0.0001822115149523611,
+      "loss": 1.765,
+      "step": 9150
+    },
+    {
+      "epoch": 1.9297414020119028,
+      "grad_norm": 0.892197847366333,
+      "learning_rate": 0.0001821738109883348,
+      "loss": 1.7358,
+      "step": 9160
+    },
+    {
+      "epoch": 1.931848106599252,
+      "grad_norm": 0.8694189786911011,
+      "learning_rate": 0.00018213607101807527,
+      "loss": 1.7133,
+      "step": 9170
+    },
+    {
+      "epoch": 1.9339548111866014,
+      "grad_norm": 0.8469682931900024,
+      "learning_rate": 0.0001820982950581191,
+      "loss": 1.7176,
+      "step": 9180
+    },
+    {
+      "epoch": 1.9360615157739507,
+      "grad_norm": 0.8897395730018616,
+      "learning_rate": 0.0001820604831250186,
+      "loss": 1.6831,
+      "step": 9190
+    },
+    {
+      "epoch": 1.9381682203613,
+      "grad_norm": 0.8338301777839661,
+      "learning_rate": 0.0001820226352353419,
+      "loss": 1.7822,
+      "step": 9200
+    },
+    {
+      "epoch": 1.940274924948649,
+      "grad_norm": 0.8942866921424866,
+      "learning_rate": 0.00018198475140567287,
+      "loss": 1.7657,
+      "step": 9210
+    },
+    {
+      "epoch": 1.9423816295359984,
+      "grad_norm": 0.8633554577827454,
+      "learning_rate": 0.00018194683165261114,
+      "loss": 1.7695,
+      "step": 9220
+    },
+    {
+      "epoch": 1.9444883341233474,
+      "grad_norm": 0.8235793709754944,
+      "learning_rate": 0.00018190887599277207,
+      "loss": 1.8101,
+      "step": 9230
+    },
+    {
+      "epoch": 1.9465950387106967,
+      "grad_norm": 0.8241968154907227,
+      "learning_rate": 0.00018187088444278674,
+      "loss": 1.7691,
+      "step": 9240
+    },
+    {
+      "epoch": 1.948701743298046,
+      "grad_norm": 0.8452653288841248,
+      "learning_rate": 0.000181832857019302,
+      "loss": 1.8259,
+      "step": 9250
+    },
+    {
+      "epoch": 1.9508084478853953,
+      "grad_norm": 0.862368643283844,
+      "learning_rate": 0.00018179479373898035,
+      "loss": 1.7402,
+      "step": 9260
+    },
+    {
+      "epoch": 1.9529151524727446,
+      "grad_norm": 0.7917757034301758,
+      "learning_rate": 0.00018175669461850005,
+      "loss": 1.7411,
+      "step": 9270
+    },
+    {
+      "epoch": 1.955021857060094,
+      "grad_norm": 0.9397031664848328,
+      "learning_rate": 0.00018171855967455506,
+      "loss": 1.6667,
+      "step": 9280
+    },
+    {
+      "epoch": 1.957128561647443,
+      "grad_norm": 0.8887811303138733,
+      "learning_rate": 0.00018168038892385507,
+      "loss": 1.6761,
+      "step": 9290
+    },
+    {
+      "epoch": 1.9592352662347923,
+      "grad_norm": 0.8086221814155579,
+      "learning_rate": 0.00018164218238312535,
+      "loss": 1.7965,
+      "step": 9300
+    },
+    {
+      "epoch": 1.9613419708221413,
+      "grad_norm": 0.8723207712173462,
+      "learning_rate": 0.00018160394006910694,
+      "loss": 1.801,
+      "step": 9310
+    },
+    {
+      "epoch": 1.9634486754094906,
+      "grad_norm": 0.8716481328010559,
+      "learning_rate": 0.00018156566199855657,
+      "loss": 1.689,
+      "step": 9320
+    },
+    {
+      "epoch": 1.96555537999684,
+      "grad_norm": 0.8738270401954651,
+      "learning_rate": 0.00018152734818824658,
+      "loss": 1.7629,
+      "step": 9330
+    },
+    {
+      "epoch": 1.9676620845841892,
+      "grad_norm": 0.8747379779815674,
+      "learning_rate": 0.00018148899865496503,
+      "loss": 1.712,
+      "step": 9340
+    },
+    {
+      "epoch": 1.9697687891715385,
+      "grad_norm": 0.9044255018234253,
+      "learning_rate": 0.00018145061341551553,
+      "loss": 1.7433,
+      "step": 9350
+    },
+    {
+      "epoch": 1.9718754937588878,
+      "grad_norm": 0.8422555923461914,
+      "learning_rate": 0.00018141219248671745,
+      "loss": 1.7475,
+      "step": 9360
+    },
+    {
+      "epoch": 1.973982198346237,
+      "grad_norm": 0.9065391421318054,
+      "learning_rate": 0.00018137373588540578,
+      "loss": 1.777,
+      "step": 9370
+    },
+    {
+      "epoch": 1.9760889029335862,
+      "grad_norm": 0.8669577240943909,
+      "learning_rate": 0.00018133524362843104,
+      "loss": 1.7281,
+      "step": 9380
+    },
+    {
+      "epoch": 1.9781956075209353,
+      "grad_norm": 0.8608076572418213,
+      "learning_rate": 0.0001812967157326595,
+      "loss": 1.7726,
+      "step": 9390
+    },
+    {
+      "epoch": 1.9803023121082846,
+      "grad_norm": 0.8125661015510559,
+      "learning_rate": 0.00018125815221497294,
+      "loss": 1.7158,
+      "step": 9400
+    },
+    {
+      "epoch": 1.9824090166956339,
+      "grad_norm": 0.92203289270401,
+      "learning_rate": 0.00018121955309226886,
+      "loss": 1.7336,
+      "step": 9410
+    },
+    {
+      "epoch": 1.9845157212829831,
+      "grad_norm": 0.8154654502868652,
+      "learning_rate": 0.00018118091838146029,
+      "loss": 1.7337,
+      "step": 9420
+    },
+    {
+      "epoch": 1.9866224258703324,
+      "grad_norm": 0.8414307832717896,
+      "learning_rate": 0.00018114224809947583,
+      "loss": 1.7626,
+      "step": 9430
+    },
+    {
+      "epoch": 1.9887291304576817,
+      "grad_norm": 0.9456828832626343,
+      "learning_rate": 0.0001811035422632597,
+      "loss": 1.7152,
+      "step": 9440
+    },
+    {
+      "epoch": 1.9908358350450308,
+      "grad_norm": 0.8258214592933655,
+      "learning_rate": 0.00018106480088977172,
+      "loss": 1.7039,
+      "step": 9450
+    },
+    {
+      "epoch": 1.9929425396323799,
+      "grad_norm": 0.8530787229537964,
+      "learning_rate": 0.00018102602399598728,
+      "loss": 1.7288,
+      "step": 9460
+    },
+    {
+      "epoch": 1.9950492442197292,
+      "grad_norm": 0.8169214129447937,
+      "learning_rate": 0.00018098721159889728,
+      "loss": 1.7202,
+      "step": 9470
+    },
+    {
+      "epoch": 1.9971559488070785,
+      "grad_norm": 0.868748664855957,
+      "learning_rate": 0.00018094836371550824,
+      "loss": 1.711,
+      "step": 9480
+    },
+    {
+      "epoch": 1.9992626533944278,
+      "grad_norm": 0.7722933888435364,
+      "learning_rate": 0.00018090948036284215,
+      "loss": 1.6874,
+      "step": 9490
+    },
+    {
+      "epoch": 2.001369357981777,
+      "grad_norm": 0.7461708188056946,
+      "learning_rate": 0.0001808705615579367,
+      "loss": 1.6395,
+      "step": 9500
+    },
+    {
+      "epoch": 2.0034760625691264,
+      "grad_norm": 0.9301486015319824,
+      "learning_rate": 0.00018083160731784486,
+      "loss": 1.6459,
+      "step": 9510
+    },
+    {
+      "epoch": 2.0055827671564757,
+      "grad_norm": 0.834187924861908,
+      "learning_rate": 0.00018079261765963537,
+      "loss": 1.6873,
+      "step": 9520
+    },
+    {
+      "epoch": 2.0076894717438245,
+      "grad_norm": 0.8011589646339417,
+      "learning_rate": 0.0001807535926003924,
+      "loss": 1.6617,
+      "step": 9530
+    },
+    {
+      "epoch": 2.009796176331174,
+      "grad_norm": 0.8627424836158752,
+      "learning_rate": 0.00018071453215721554,
+      "loss": 1.7085,
+      "step": 9540
+    },
+    {
+      "epoch": 2.011902880918523,
+      "grad_norm": 0.9021561145782471,
+      "learning_rate": 0.00018067543634722006,
+      "loss": 1.6688,
+      "step": 9550
+    },
+    {
+      "epoch": 2.0140095855058724,
+      "grad_norm": 0.8814119100570679,
+      "learning_rate": 0.00018063630518753662,
+      "loss": 1.7258,
+      "step": 9560
+    },
+    {
+      "epoch": 2.0161162900932217,
+      "grad_norm": 0.9105542898178101,
+      "learning_rate": 0.0001805971386953113,
+      "loss": 1.6976,
+      "step": 9570
+    },
+    {
+      "epoch": 2.018222994680571,
+      "grad_norm": 0.8711789846420288,
+      "learning_rate": 0.00018055793688770587,
+      "loss": 1.7255,
+      "step": 9580
+    },
+    {
+      "epoch": 2.0203296992679203,
+      "grad_norm": 0.8914005160331726,
+      "learning_rate": 0.00018051869978189731,
+      "loss": 1.7591,
+      "step": 9590
+    },
+    {
+      "epoch": 2.0224364038552696,
+      "grad_norm": 0.874055802822113,
+      "learning_rate": 0.00018047942739507836,
+      "loss": 1.626,
+      "step": 9600
+    },
+    {
+      "epoch": 2.0245431084426184,
+      "grad_norm": 0.823403537273407,
+      "learning_rate": 0.00018044011974445697,
+      "loss": 1.7155,
+      "step": 9610
+    },
+    {
+      "epoch": 2.0266498130299677,
+      "grad_norm": 0.859565258026123,
+      "learning_rate": 0.00018040077684725666,
+      "loss": 1.7357,
+      "step": 9620
+    },
+    {
+      "epoch": 2.028756517617317,
+      "grad_norm": 0.8823413252830505,
+      "learning_rate": 0.0001803613987207163,
+      "loss": 1.673,
+      "step": 9630
+    },
+    {
+      "epoch": 2.0308632222046663,
+      "grad_norm": 0.903207540512085,
+      "learning_rate": 0.00018032198538209043,
+      "loss": 1.708,
+      "step": 9640
+    },
+    {
+      "epoch": 2.0329699267920156,
+      "grad_norm": 0.9172302484512329,
+      "learning_rate": 0.0001802825368486487,
+      "loss": 1.7111,
+      "step": 9650
+    },
+    {
+      "epoch": 2.035076631379365,
+      "grad_norm": 0.8453816175460815,
+      "learning_rate": 0.00018024305313767646,
+      "loss": 1.6574,
+      "step": 9660
+    },
+    {
+      "epoch": 2.037183335966714,
+      "grad_norm": 0.8865382671356201,
+      "learning_rate": 0.00018020353426647428,
+      "loss": 1.7316,
+      "step": 9670
+    },
+    {
+      "epoch": 2.0392900405540635,
+      "grad_norm": 0.9413029551506042,
+      "learning_rate": 0.0001801639802523582,
+      "loss": 1.7115,
+      "step": 9680
+    },
+    {
+      "epoch": 2.0413967451414123,
+      "grad_norm": 0.8472778797149658,
+      "learning_rate": 0.00018012439111265974,
+      "loss": 1.6621,
+      "step": 9690
+    },
+    {
+      "epoch": 2.0435034497287616,
+      "grad_norm": 0.8548828363418579,
+      "learning_rate": 0.00018008476686472564,
+      "loss": 1.7178,
+      "step": 9700
+    },
+    {
+      "epoch": 2.045610154316111,
+      "grad_norm": 0.9240100979804993,
+      "learning_rate": 0.0001800451075259182,
+      "loss": 1.6855,
+      "step": 9710
+    },
+    {
+      "epoch": 2.0477168589034602,
+      "grad_norm": 0.8424657583236694,
+      "learning_rate": 0.000180005413113615,
+      "loss": 1.6294,
+      "step": 9720
+    },
+    {
+      "epoch": 2.0498235634908095,
+      "grad_norm": 0.8709080815315247,
+      "learning_rate": 0.00017996568364520897,
+      "loss": 1.68,
+      "step": 9730
+    },
+    {
+      "epoch": 2.051930268078159,
+      "grad_norm": 0.8857349753379822,
+      "learning_rate": 0.00017992591913810845,
+      "loss": 1.7878,
+      "step": 9740
+    },
+    {
+      "epoch": 2.054036972665508,
+      "grad_norm": 0.928957462310791,
+      "learning_rate": 0.00017988611960973713,
+      "loss": 1.6697,
+      "step": 9750
+    },
+    {
+      "epoch": 2.0561436772528574,
+      "grad_norm": 0.8724343776702881,
+      "learning_rate": 0.00017984628507753406,
+      "loss": 1.6534,
+      "step": 9760
+    },
+    {
+      "epoch": 2.0582503818402063,
+      "grad_norm": 0.9669876098632812,
+      "learning_rate": 0.00017980641555895356,
+      "loss": 1.6382,
+      "step": 9770
+    },
+    {
+      "epoch": 2.0603570864275556,
+      "grad_norm": 0.853915274143219,
+      "learning_rate": 0.00017976651107146533,
+      "loss": 1.7274,
+      "step": 9780
+    },
+    {
+      "epoch": 2.062463791014905,
+      "grad_norm": 0.8353849053382874,
+      "learning_rate": 0.00017972657163255442,
+      "loss": 1.6684,
+      "step": 9790
+    },
+    {
+      "epoch": 2.064570495602254,
+      "grad_norm": 0.8487059473991394,
+      "learning_rate": 0.00017968659725972112,
+      "loss": 1.7395,
+      "step": 9800
+    },
+    {
+      "epoch": 2.0666772001896034,
+      "grad_norm": 0.8349198698997498,
+      "learning_rate": 0.00017964658797048108,
+      "loss": 1.7004,
+      "step": 9810
+    },
+    {
+      "epoch": 2.0687839047769527,
+      "grad_norm": 0.866244375705719,
+      "learning_rate": 0.0001796065437823652,
+      "loss": 1.6692,
+      "step": 9820
+    },
+    {
+      "epoch": 2.070890609364302,
+      "grad_norm": 0.9268345236778259,
+      "learning_rate": 0.0001795664647129198,
+      "loss": 1.7057,
+      "step": 9830
+    },
+    {
+      "epoch": 2.0729973139516513,
+      "grad_norm": 0.803345263004303,
+      "learning_rate": 0.0001795263507797063,
+      "loss": 1.6856,
+      "step": 9840
+    },
+    {
+      "epoch": 2.075104018539,
+      "grad_norm": 0.8338049054145813,
+      "learning_rate": 0.00017948620200030152,
+      "loss": 1.7012,
+      "step": 9850
+    },
+    {
+      "epoch": 2.0772107231263495,
+      "grad_norm": 0.9227036833763123,
+      "learning_rate": 0.00017944601839229755,
+      "loss": 1.6416,
+      "step": 9860
+    },
+    {
+      "epoch": 2.0793174277136988,
+      "grad_norm": 1.008952021598816,
+      "learning_rate": 0.00017940579997330165,
+      "loss": 1.6402,
+      "step": 9870
+    },
+    {
+      "epoch": 2.081424132301048,
+      "grad_norm": 0.9596250653266907,
+      "learning_rate": 0.0001793655467609364,
+      "loss": 1.6992,
+      "step": 9880
+    },
+    {
+      "epoch": 2.0835308368883974,
+      "grad_norm": 0.8340634107589722,
+      "learning_rate": 0.00017932525877283964,
+      "loss": 1.7393,
+      "step": 9890
+    },
+    {
+      "epoch": 2.0856375414757466,
+      "grad_norm": 0.8354076743125916,
+      "learning_rate": 0.00017928493602666445,
+      "loss": 1.7162,
+      "step": 9900
+    },
+    {
+      "epoch": 2.087744246063096,
+      "grad_norm": 0.8993865847587585,
+      "learning_rate": 0.00017924457854007902,
+      "loss": 1.6699,
+      "step": 9910
+    },
+    {
+      "epoch": 2.0898509506504452,
+      "grad_norm": 0.9259373545646667,
+      "learning_rate": 0.00017920418633076698,
+      "loss": 1.6892,
+      "step": 9920
+    },
+    {
+      "epoch": 2.091957655237794,
+      "grad_norm": 0.9740672707557678,
+      "learning_rate": 0.0001791637594164269,
+      "loss": 1.7243,
+      "step": 9930
+    },
+    {
+      "epoch": 2.0940643598251434,
+      "grad_norm": 0.8578957319259644,
+      "learning_rate": 0.00017912329781477287,
+      "loss": 1.6387,
+      "step": 9940
+    },
+    {
+      "epoch": 2.0961710644124927,
+      "grad_norm": 0.8971062302589417,
+      "learning_rate": 0.00017908280154353392,
+      "loss": 1.7315,
+      "step": 9950
+    },
+    {
+      "epoch": 2.098277768999842,
+      "grad_norm": 0.8996395468711853,
+      "learning_rate": 0.00017904227062045437,
+      "loss": 1.7068,
+      "step": 9960
+    },
+    {
+      "epoch": 2.1003844735871913,
+      "grad_norm": 0.858382523059845,
+      "learning_rate": 0.00017900170506329372,
+      "loss": 1.7447,
+      "step": 9970
+    },
+    {
+      "epoch": 2.1024911781745406,
+      "grad_norm": 0.8814343810081482,
+      "learning_rate": 0.00017896110488982672,
+      "loss": 1.6968,
+      "step": 9980
+    },
+    {
+      "epoch": 2.10459788276189,
+      "grad_norm": 0.9278027415275574,
+      "learning_rate": 0.00017892047011784312,
+      "loss": 1.7095,
+      "step": 9990
+    },
+    {
+      "epoch": 2.106704587349239,
+      "grad_norm": 0.8749867081642151,
+      "learning_rate": 0.00017887980076514796,
+      "loss": 1.7037,
+      "step": 10000
+    },
+    {
+      "epoch": 2.108811291936588,
+      "grad_norm": 0.8612695932388306,
+      "learning_rate": 0.0001788390968495614,
+      "loss": 1.6536,
+      "step": 10010
+    },
+    {
+      "epoch": 2.1109179965239373,
+      "grad_norm": 0.8968421220779419,
+      "learning_rate": 0.00017879835838891875,
+      "loss": 1.6926,
+      "step": 10020
+    },
+    {
+      "epoch": 2.1130247011112866,
+      "grad_norm": 0.9752348065376282,
+      "learning_rate": 0.00017875758540107043,
+      "loss": 1.6765,
+      "step": 10030
+    },
+    {
+      "epoch": 2.115131405698636,
+      "grad_norm": 0.8976007699966431,
+      "learning_rate": 0.00017871677790388203,
+      "loss": 1.7059,
+      "step": 10040
+    },
+    {
+      "epoch": 2.117238110285985,
+      "grad_norm": 0.8147952556610107,
+      "learning_rate": 0.00017867593591523422,
+      "loss": 1.7046,
+      "step": 10050
+    },
+    {
+      "epoch": 2.1193448148733345,
+      "grad_norm": 0.9791306853294373,
+      "learning_rate": 0.00017863505945302279,
+      "loss": 1.6575,
+      "step": 10060
+    },
+    {
+      "epoch": 2.1214515194606838,
+      "grad_norm": 0.8902132511138916,
+      "learning_rate": 0.0001785941485351587,
+      "loss": 1.6564,
+      "step": 10070
+    },
+    {
+      "epoch": 2.123558224048033,
+      "grad_norm": 0.8692643642425537,
+      "learning_rate": 0.00017855320317956784,
+      "loss": 1.7202,
+      "step": 10080
+    },
+    {
+      "epoch": 2.125664928635382,
+      "grad_norm": 0.8463397026062012,
+      "learning_rate": 0.00017851222340419144,
+      "loss": 1.7316,
+      "step": 10090
+    },
+    {
+      "epoch": 2.127771633222731,
+      "grad_norm": 0.9453892707824707,
+      "learning_rate": 0.00017847120922698562,
+      "loss": 1.6316,
+      "step": 10100
+    },
+    {
+      "epoch": 2.1298783378100805,
+      "grad_norm": 0.8649837970733643,
+      "learning_rate": 0.00017843016066592158,
+      "loss": 1.5966,
+      "step": 10110
+    },
+    {
+      "epoch": 2.13198504239743,
+      "grad_norm": 0.8016608357429504,
+      "learning_rate": 0.0001783890777389857,
+      "loss": 1.7382,
+      "step": 10120
+    },
+    {
+      "epoch": 2.134091746984779,
+      "grad_norm": 0.879031777381897,
+      "learning_rate": 0.00017834796046417933,
+      "loss": 1.6988,
+      "step": 10130
+    },
+    {
+      "epoch": 2.1361984515721284,
+      "grad_norm": 0.9154124855995178,
+      "learning_rate": 0.00017830680885951887,
+      "loss": 1.6508,
+      "step": 10140
+    },
+    {
+      "epoch": 2.1383051561594777,
+      "grad_norm": 0.8642693758010864,
+      "learning_rate": 0.00017826562294303585,
+      "loss": 1.7181,
+      "step": 10150
+    },
+    {
+      "epoch": 2.140411860746827,
+      "grad_norm": 0.9341524243354797,
+      "learning_rate": 0.00017822440273277673,
+      "loss": 1.6984,
+      "step": 10160
+    },
+    {
+      "epoch": 2.142518565334176,
+      "grad_norm": 0.8300173282623291,
+      "learning_rate": 0.000178183148246803,
+      "loss": 1.7304,
+      "step": 10170
+    },
+    {
+      "epoch": 2.144625269921525,
+      "grad_norm": 0.9484438896179199,
+      "learning_rate": 0.00017814185950319126,
+      "loss": 1.722,
+      "step": 10180
+    },
+    {
+      "epoch": 2.1467319745088744,
+      "grad_norm": 0.9994838833808899,
+      "learning_rate": 0.00017810053652003304,
+      "loss": 1.7375,
+      "step": 10190
+    },
+    {
+      "epoch": 2.1488386790962237,
+      "grad_norm": 0.9783572554588318,
+      "learning_rate": 0.00017805917931543492,
+      "loss": 1.6866,
+      "step": 10200
+    },
+    {
+      "epoch": 2.150945383683573,
+      "grad_norm": 1.022603154182434,
+      "learning_rate": 0.00017801778790751843,
+      "loss": 1.715,
+      "step": 10210
+    },
+    {
+      "epoch": 2.1530520882709223,
+      "grad_norm": 0.9454113245010376,
+      "learning_rate": 0.00017797636231442016,
+      "loss": 1.738,
+      "step": 10220
+    },
+    {
+      "epoch": 2.1551587928582716,
+      "grad_norm": 0.9445421099662781,
+      "learning_rate": 0.00017793490255429157,
+      "loss": 1.6642,
+      "step": 10230
+    },
+    {
+      "epoch": 2.1572654974456205,
+      "grad_norm": 0.9507448077201843,
+      "learning_rate": 0.00017789340864529917,
+      "loss": 1.6549,
+      "step": 10240
+    },
+    {
+      "epoch": 2.1593722020329698,
+      "grad_norm": 0.9919808506965637,
+      "learning_rate": 0.00017785188060562442,
+      "loss": 1.6856,
+      "step": 10250
+    },
+    {
+      "epoch": 2.161478906620319,
+      "grad_norm": 0.8635092973709106,
+      "learning_rate": 0.00017781031845346375,
+      "loss": 1.669,
+      "step": 10260
+    },
+    {
+      "epoch": 2.1635856112076683,
+      "grad_norm": 0.9286298155784607,
+      "learning_rate": 0.00017776872220702847,
+      "loss": 1.7009,
+      "step": 10270
+    },
+    {
+      "epoch": 2.1656923157950176,
+      "grad_norm": 0.8796800374984741,
+      "learning_rate": 0.0001777270918845449,
+      "loss": 1.6913,
+      "step": 10280
+    },
+    {
+      "epoch": 2.167799020382367,
+      "grad_norm": 0.9373218417167664,
+      "learning_rate": 0.00017768542750425426,
+      "loss": 1.7442,
+      "step": 10290
+    },
+    {
+      "epoch": 2.1699057249697162,
+      "grad_norm": 0.8646446466445923,
+      "learning_rate": 0.00017764372908441275,
+      "loss": 1.7052,
+      "step": 10300
+    },
+    {
+      "epoch": 2.1720124295570655,
+      "grad_norm": 0.9008393287658691,
+      "learning_rate": 0.00017760199664329136,
+      "loss": 1.7049,
+      "step": 10310
+    },
+    {
+      "epoch": 2.174119134144415,
+      "grad_norm": 0.9161930084228516,
+      "learning_rate": 0.00017756023019917607,
+      "loss": 1.6693,
+      "step": 10320
+    },
+    {
+      "epoch": 2.1762258387317637,
+      "grad_norm": 0.9321724772453308,
+      "learning_rate": 0.0001775184297703678,
+      "loss": 1.6694,
+      "step": 10330
+    },
+    {
+      "epoch": 2.178332543319113,
+      "grad_norm": 1.1692795753479004,
+      "learning_rate": 0.0001774765953751823,
+      "loss": 1.7455,
+      "step": 10340
+    },
+    {
+      "epoch": 2.1804392479064623,
+      "grad_norm": 0.8697549104690552,
+      "learning_rate": 0.00017743472703195015,
+      "loss": 1.6826,
+      "step": 10350
+    },
+    {
+      "epoch": 2.1825459524938116,
+      "grad_norm": 0.9908336400985718,
+      "learning_rate": 0.000177392824759017,
+      "loss": 1.7432,
+      "step": 10360
+    },
+    {
+      "epoch": 2.184652657081161,
+      "grad_norm": 0.9399738311767578,
+      "learning_rate": 0.0001773508885747431,
+      "loss": 1.6717,
+      "step": 10370
+    },
+    {
+      "epoch": 2.18675936166851,
+      "grad_norm": 1.0489552021026611,
+      "learning_rate": 0.00017730891849750377,
+      "loss": 1.7272,
+      "step": 10380
+    },
+    {
+      "epoch": 2.1888660662558594,
+      "grad_norm": 0.9378235340118408,
+      "learning_rate": 0.00017726691454568908,
+      "loss": 1.6757,
+      "step": 10390
+    },
+    {
+      "epoch": 2.1909727708432083,
+      "grad_norm": 1.0289270877838135,
+      "learning_rate": 0.000177224876737704,
+      "loss": 1.6845,
+      "step": 10400
+    },
+    {
+      "epoch": 2.1930794754305576,
+      "grad_norm": 0.8981749415397644,
+      "learning_rate": 0.00017718280509196828,
+      "loss": 1.7542,
+      "step": 10410
+    },
+    {
+      "epoch": 2.195186180017907,
+      "grad_norm": 0.8597977161407471,
+      "learning_rate": 0.00017714069962691657,
+      "loss": 1.664,
+      "step": 10420
+    },
+    {
+      "epoch": 2.197292884605256,
+      "grad_norm": 0.8526979088783264,
+      "learning_rate": 0.0001770985603609982,
+      "loss": 1.7156,
+      "step": 10430
+    },
+    {
+      "epoch": 2.1993995891926055,
+      "grad_norm": 0.9204460382461548,
+      "learning_rate": 0.0001770563873126775,
+      "loss": 1.6903,
+      "step": 10440
+    },
+    {
+      "epoch": 2.2015062937799548,
+      "grad_norm": 0.8693051934242249,
+      "learning_rate": 0.00017701418050043342,
+      "loss": 1.7246,
+      "step": 10450
+    },
+    {
+      "epoch": 2.203612998367304,
+      "grad_norm": 0.8942760825157166,
+      "learning_rate": 0.00017697193994275983,
+      "loss": 1.7206,
+      "step": 10460
+    },
+    {
+      "epoch": 2.2057197029546534,
+      "grad_norm": 0.8524116277694702,
+      "learning_rate": 0.00017692966565816532,
+      "loss": 1.7519,
+      "step": 10470
+    },
+    {
+      "epoch": 2.2078264075420027,
+      "grad_norm": 0.9318793416023254,
+      "learning_rate": 0.00017688735766517333,
+      "loss": 1.7078,
+      "step": 10480
+    },
+    {
+      "epoch": 2.2099331121293515,
+      "grad_norm": 0.8922185301780701,
+      "learning_rate": 0.00017684501598232198,
+      "loss": 1.7328,
+      "step": 10490
+    },
+    {
+      "epoch": 2.212039816716701,
+      "grad_norm": 0.8484925627708435,
+      "learning_rate": 0.0001768026406281642,
+      "loss": 1.7063,
+      "step": 10500
+    },
+    {
+      "epoch": 2.21414652130405,
+      "grad_norm": 0.907193660736084,
+      "learning_rate": 0.00017676023162126772,
+      "loss": 1.714,
+      "step": 10510
+    },
+    {
+      "epoch": 2.2162532258913994,
+      "grad_norm": 0.9571147561073303,
+      "learning_rate": 0.00017671778898021488,
+      "loss": 1.6853,
+      "step": 10520
+    },
+    {
+      "epoch": 2.2183599304787487,
+      "grad_norm": 0.8519482016563416,
+      "learning_rate": 0.0001766753127236029,
+      "loss": 1.6397,
+      "step": 10530
+    },
+    {
+      "epoch": 2.220466635066098,
+      "grad_norm": 0.8464012742042542,
+      "learning_rate": 0.00017663280287004364,
+      "loss": 1.787,
+      "step": 10540
+    },
+    {
+      "epoch": 2.2225733396534473,
+      "grad_norm": 1.021557092666626,
+      "learning_rate": 0.00017659025943816373,
+      "loss": 1.6816,
+      "step": 10550
+    },
+    {
+      "epoch": 2.224680044240796,
+      "grad_norm": 0.9452894926071167,
+      "learning_rate": 0.00017654768244660448,
+      "loss": 1.6548,
+      "step": 10560
+    },
+    {
+      "epoch": 2.2267867488281454,
+      "grad_norm": 0.8741638660430908,
+      "learning_rate": 0.00017650507191402194,
+      "loss": 1.6928,
+      "step": 10570
+    },
+    {
+      "epoch": 2.2288934534154947,
+      "grad_norm": 0.9022242426872253,
+      "learning_rate": 0.00017646242785908682,
+      "loss": 1.6819,
+      "step": 10580
+    },
+    {
+      "epoch": 2.231000158002844,
+      "grad_norm": 0.9004740118980408,
+      "learning_rate": 0.00017641975030048454,
+      "loss": 1.693,
+      "step": 10590
+    },
+    {
+      "epoch": 2.2331068625901933,
+      "grad_norm": 0.9306367635726929,
+      "learning_rate": 0.0001763770392569152,
+      "loss": 1.7281,
+      "step": 10600
+    },
+    {
+      "epoch": 2.2352135671775426,
+      "grad_norm": 0.9229249358177185,
+      "learning_rate": 0.00017633429474709356,
+      "loss": 1.6821,
+      "step": 10610
+    },
+    {
+      "epoch": 2.237320271764892,
+      "grad_norm": 0.9667083024978638,
+      "learning_rate": 0.00017629151678974907,
+      "loss": 1.6875,
+      "step": 10620
+    },
+    {
+      "epoch": 2.239426976352241,
+      "grad_norm": 0.8459573984146118,
+      "learning_rate": 0.0001762487054036258,
+      "loss": 1.7336,
+      "step": 10630
+    },
+    {
+      "epoch": 2.24153368093959,
+      "grad_norm": 0.9185715913772583,
+      "learning_rate": 0.00017620586060748252,
+      "loss": 1.6614,
+      "step": 10640
+    },
+    {
+      "epoch": 2.2436403855269393,
+      "grad_norm": 0.9300093650817871,
+      "learning_rate": 0.00017616298242009251,
+      "loss": 1.6556,
+      "step": 10650
+    },
+    {
+      "epoch": 2.2457470901142886,
+      "grad_norm": 0.8929982781410217,
+      "learning_rate": 0.0001761200708602439,
+      "loss": 1.6979,
+      "step": 10660
+    },
+    {
+      "epoch": 2.247853794701638,
+      "grad_norm": 0.9404104351997375,
+      "learning_rate": 0.00017607712594673922,
+      "loss": 1.7029,
+      "step": 10670
+    },
+    {
+      "epoch": 2.2499604992889872,
+      "grad_norm": 0.9714049696922302,
+      "learning_rate": 0.00017603414769839577,
+      "loss": 1.7188,
+      "step": 10680
+    },
+    {
+      "epoch": 2.2520672038763365,
+      "grad_norm": 0.8535948991775513,
+      "learning_rate": 0.00017599113613404538,
+      "loss": 1.6941,
+      "step": 10690
+    },
+    {
+      "epoch": 2.254173908463686,
+      "grad_norm": 0.9558704495429993,
+      "learning_rate": 0.0001759480912725345,
+      "loss": 1.6855,
+      "step": 10700
+    },
+    {
+      "epoch": 2.256280613051035,
+      "grad_norm": 0.9051923155784607,
+      "learning_rate": 0.00017590501313272415,
+      "loss": 1.7355,
+      "step": 10710
+    },
+    {
+      "epoch": 2.258387317638384,
+      "grad_norm": 0.8770729899406433,
+      "learning_rate": 0.00017586190173348996,
+      "loss": 1.7381,
+      "step": 10720
+    },
+    {
+      "epoch": 2.2604940222257333,
+      "grad_norm": 0.8900496363639832,
+      "learning_rate": 0.00017581875709372216,
+      "loss": 1.7573,
+      "step": 10730
+    },
+    {
+      "epoch": 2.2626007268130826,
+      "grad_norm": 0.8546925187110901,
+      "learning_rate": 0.00017577557923232546,
+      "loss": 1.7518,
+      "step": 10740
+    },
+    {
+      "epoch": 2.264707431400432,
+      "grad_norm": 1.025599718093872,
+      "learning_rate": 0.0001757323681682192,
+      "loss": 1.6861,
+      "step": 10750
+    },
+    {
+      "epoch": 2.266814135987781,
+      "grad_norm": 0.8811286687850952,
+      "learning_rate": 0.0001756891239203372,
+      "loss": 1.7308,
+      "step": 10760
+    },
+    {
+      "epoch": 2.2689208405751304,
+      "grad_norm": 0.9262977242469788,
+      "learning_rate": 0.00017564584650762793,
+      "loss": 1.6927,
+      "step": 10770
+    },
+    {
+      "epoch": 2.2710275451624797,
+      "grad_norm": 1.1650166511535645,
+      "learning_rate": 0.00017560253594905425,
+      "loss": 1.7301,
+      "step": 10780
+    },
+    {
+      "epoch": 2.2731342497498286,
+      "grad_norm": 0.8260765671730042,
+      "learning_rate": 0.0001755591922635937,
+      "loss": 1.6494,
+      "step": 10790
+    },
+    {
+      "epoch": 2.275240954337178,
+      "grad_norm": 0.9326931238174438,
+      "learning_rate": 0.00017551581547023819,
+      "loss": 1.7466,
+      "step": 10800
+    },
+    {
+      "epoch": 2.277347658924527,
+      "grad_norm": 0.8884145021438599,
+      "learning_rate": 0.0001754724055879942,
+      "loss": 1.7261,
+      "step": 10810
+    },
+    {
+      "epoch": 2.2794543635118765,
+      "grad_norm": 0.907815158367157,
+      "learning_rate": 0.00017542896263588275,
+      "loss": 1.6978,
+      "step": 10820
+    },
+    {
+      "epoch": 2.2815610680992258,
+      "grad_norm": 0.9290655255317688,
+      "learning_rate": 0.0001753854866329393,
+      "loss": 1.6516,
+      "step": 10830
+    },
+    {
+      "epoch": 2.283667772686575,
+      "grad_norm": 0.9227153062820435,
+      "learning_rate": 0.00017534197759821378,
+      "loss": 1.6579,
+      "step": 10840
+    },
+    {
+      "epoch": 2.2857744772739244,
+      "grad_norm": 0.8803461194038391,
+      "learning_rate": 0.00017529843555077066,
+      "loss": 1.6537,
+      "step": 10850
+    },
+    {
+      "epoch": 2.2878811818612736,
+      "grad_norm": 1.0009804964065552,
+      "learning_rate": 0.00017525486050968875,
+      "loss": 1.7348,
+      "step": 10860
+    },
+    {
+      "epoch": 2.289987886448623,
+      "grad_norm": 0.8984347581863403,
+      "learning_rate": 0.00017521125249406145,
+      "loss": 1.6616,
+      "step": 10870
+    },
+    {
+      "epoch": 2.292094591035972,
+      "grad_norm": 0.8679141402244568,
+      "learning_rate": 0.00017516761152299657,
+      "loss": 1.6898,
+      "step": 10880
+    },
+    {
+      "epoch": 2.294201295623321,
+      "grad_norm": 1.022753357887268,
+      "learning_rate": 0.00017512393761561632,
+      "loss": 1.6664,
+      "step": 10890
+    },
+    {
+      "epoch": 2.2963080002106704,
+      "grad_norm": 0.919381856918335,
+      "learning_rate": 0.00017508023079105736,
+      "loss": 1.7005,
+      "step": 10900
+    },
+    {
+      "epoch": 2.2984147047980197,
+      "grad_norm": 0.9226942658424377,
+      "learning_rate": 0.0001750364910684708,
+      "loss": 1.7123,
+      "step": 10910
+    },
+    {
+      "epoch": 2.300521409385369,
+      "grad_norm": 0.9635823369026184,
+      "learning_rate": 0.00017499271846702213,
+      "loss": 1.7071,
+      "step": 10920
+    },
+    {
+      "epoch": 2.3026281139727183,
+      "grad_norm": 0.9143548011779785,
+      "learning_rate": 0.00017494891300589131,
+      "loss": 1.6471,
+      "step": 10930
+    },
+    {
+      "epoch": 2.3047348185600676,
+      "grad_norm": 0.8815865516662598,
+      "learning_rate": 0.00017490507470427257,
+      "loss": 1.7108,
+      "step": 10940
+    },
+    {
+      "epoch": 2.3068415231474164,
+      "grad_norm": 0.8886958956718445,
+      "learning_rate": 0.0001748612035813747,
+      "loss": 1.6857,
+      "step": 10950
+    },
+    {
+      "epoch": 2.3089482277347657,
+      "grad_norm": 0.9607595801353455,
+      "learning_rate": 0.00017481729965642065,
+      "loss": 1.8113,
+      "step": 10960
+    },
+    {
+      "epoch": 2.311054932322115,
+      "grad_norm": 0.8694586157798767,
+      "learning_rate": 0.00017477336294864805,
+      "loss": 1.7077,
+      "step": 10970
+    },
+    {
+      "epoch": 2.3131616369094643,
+      "grad_norm": 0.931908130645752,
+      "learning_rate": 0.00017472939347730856,
+      "loss": 1.7296,
+      "step": 10980
+    },
+    {
+      "epoch": 2.3152683414968136,
+      "grad_norm": 0.8658048510551453,
+      "learning_rate": 0.00017468539126166846,
+      "loss": 1.7228,
+      "step": 10990
+    },
+    {
+      "epoch": 2.317375046084163,
+      "grad_norm": 0.8600513935089111,
+      "learning_rate": 0.00017464135632100825,
+      "loss": 1.7335,
+      "step": 11000
+    },
+    {
+      "epoch": 2.319481750671512,
+      "grad_norm": 0.8735826015472412,
+      "learning_rate": 0.00017459728867462275,
+      "loss": 1.6597,
+      "step": 11010
+    },
+    {
+      "epoch": 2.3215884552588615,
+      "grad_norm": 0.9487072229385376,
+      "learning_rate": 0.00017455318834182118,
+      "loss": 1.7637,
+      "step": 11020
+    },
+    {
+      "epoch": 2.323695159846211,
+      "grad_norm": 0.9656599164009094,
+      "learning_rate": 0.00017450905534192708,
+      "loss": 1.705,
+      "step": 11030
+    },
+    {
+      "epoch": 2.3258018644335596,
+      "grad_norm": 0.8537636995315552,
+      "learning_rate": 0.0001744648896942782,
+      "loss": 1.6545,
+      "step": 11040
+    },
+    {
+      "epoch": 2.327908569020909,
+      "grad_norm": 0.8958091139793396,
+      "learning_rate": 0.0001744206914182268,
+      "loss": 1.7148,
+      "step": 11050
+    },
+    {
+      "epoch": 2.330015273608258,
+      "grad_norm": 0.8656923770904541,
+      "learning_rate": 0.0001743764605331392,
+      "loss": 1.7047,
+      "step": 11060
+    },
+    {
+      "epoch": 2.3321219781956075,
+      "grad_norm": 0.8860281109809875,
+      "learning_rate": 0.00017433219705839616,
+      "loss": 1.7217,
+      "step": 11070
+    },
+    {
+      "epoch": 2.334228682782957,
+      "grad_norm": 0.9574180841445923,
+      "learning_rate": 0.0001742879010133927,
+      "loss": 1.7557,
+      "step": 11080
+    },
+    {
+      "epoch": 2.336335387370306,
+      "grad_norm": 0.9499590992927551,
+      "learning_rate": 0.00017424357241753807,
+      "loss": 1.7174,
+      "step": 11090
+    },
+    {
+      "epoch": 2.3384420919576554,
+      "grad_norm": 0.8961427211761475,
+      "learning_rate": 0.00017419921129025576,
+      "loss": 1.762,
+      "step": 11100
+    },
+    {
+      "epoch": 2.3405487965450043,
+      "grad_norm": 0.910950243473053,
+      "learning_rate": 0.00017415481765098364,
+      "loss": 1.7514,
+      "step": 11110
+    },
+    {
+      "epoch": 2.3426555011323535,
+      "grad_norm": 0.9184181690216064,
+      "learning_rate": 0.0001741103915191737,
+      "loss": 1.6957,
+      "step": 11120
+    },
+    {
+      "epoch": 2.344762205719703,
+      "grad_norm": 0.9001920819282532,
+      "learning_rate": 0.00017406593291429217,
+      "loss": 1.6939,
+      "step": 11130
+    },
+    {
+      "epoch": 2.346868910307052,
+      "grad_norm": 0.8733050227165222,
+      "learning_rate": 0.00017402144185581965,
+      "loss": 1.7023,
+      "step": 11140
+    },
+    {
+      "epoch": 2.3489756148944014,
+      "grad_norm": 0.8722548484802246,
+      "learning_rate": 0.0001739769183632508,
+      "loss": 1.7007,
+      "step": 11150
+    },
+    {
+      "epoch": 2.3510823194817507,
+      "grad_norm": 0.8651793599128723,
+      "learning_rate": 0.0001739323624560945,
+      "loss": 1.7203,
+      "step": 11160
+    },
+    {
+      "epoch": 2.3531890240691,
+      "grad_norm": 0.960211992263794,
+      "learning_rate": 0.000173887774153874,
+      "loss": 1.7257,
+      "step": 11170
+    },
+    {
+      "epoch": 2.3552957286564493,
+      "grad_norm": 0.8777878284454346,
+      "learning_rate": 0.00017384315347612655,
+      "loss": 1.667,
+      "step": 11180
+    },
+    {
+      "epoch": 2.3574024332437986,
+      "grad_norm": 0.9975689649581909,
+      "learning_rate": 0.00017379850044240368,
+      "loss": 1.6877,
+      "step": 11190
+    },
+    {
+      "epoch": 2.3595091378311475,
+      "grad_norm": 0.9246053695678711,
+      "learning_rate": 0.00017375381507227108,
+      "loss": 1.6909,
+      "step": 11200
+    },
+    {
+      "epoch": 2.3616158424184968,
+      "grad_norm": 0.8531624674797058,
+      "learning_rate": 0.00017370909738530864,
+      "loss": 1.6837,
+      "step": 11210
+    },
+    {
+      "epoch": 2.363722547005846,
+      "grad_norm": 0.8880084753036499,
+      "learning_rate": 0.00017366434740111037,
+      "loss": 1.7014,
+      "step": 11220
+    },
+    {
+      "epoch": 2.3658292515931953,
+      "grad_norm": 0.8885220289230347,
+      "learning_rate": 0.0001736195651392844,
+      "loss": 1.7163,
+      "step": 11230
+    },
+    {
+      "epoch": 2.3679359561805446,
+      "grad_norm": 0.9168665409088135,
+      "learning_rate": 0.0001735747506194531,
+      "loss": 1.7306,
+      "step": 11240
+    },
+    {
+      "epoch": 2.370042660767894,
+      "grad_norm": 0.8922172784805298,
+      "learning_rate": 0.00017352990386125292,
+      "loss": 1.714,
+      "step": 11250
+    },
+    {
+      "epoch": 2.3721493653552432,
+      "grad_norm": 0.8480457067489624,
+      "learning_rate": 0.0001734850248843344,
+      "loss": 1.6654,
+      "step": 11260
+    },
+    {
+      "epoch": 2.374256069942592,
+      "grad_norm": 0.862450897693634,
+      "learning_rate": 0.00017344011370836227,
+      "loss": 1.645,
+      "step": 11270
+    },
+    {
+      "epoch": 2.3763627745299414,
+      "grad_norm": 0.9275450706481934,
+      "learning_rate": 0.00017339517035301532,
+      "loss": 1.7187,
+      "step": 11280
+    },
+    {
+      "epoch": 2.3784694791172907,
+      "grad_norm": 0.9133449792861938,
+      "learning_rate": 0.00017335019483798644,
+      "loss": 1.7186,
+      "step": 11290
+    },
+    {
+      "epoch": 2.38057618370464,
+      "grad_norm": 0.915347695350647,
+      "learning_rate": 0.00017330518718298264,
+      "loss": 1.7369,
+      "step": 11300
+    },
+    {
+      "epoch": 2.3826828882919893,
+      "grad_norm": 1.0750889778137207,
+      "learning_rate": 0.000173260147407725,
+      "loss": 1.6625,
+      "step": 11310
+    },
+    {
+      "epoch": 2.3847895928793386,
+      "grad_norm": 0.9652765393257141,
+      "learning_rate": 0.00017321507553194867,
+      "loss": 1.6807,
+      "step": 11320
+    },
+    {
+      "epoch": 2.386896297466688,
+      "grad_norm": 0.9385799169540405,
+      "learning_rate": 0.00017316997157540288,
+      "loss": 1.7278,
+      "step": 11330
+    },
+    {
+      "epoch": 2.389003002054037,
+      "grad_norm": 0.8610515594482422,
+      "learning_rate": 0.00017312483555785086,
+      "loss": 1.6791,
+      "step": 11340
+    },
+    {
+      "epoch": 2.3911097066413864,
+      "grad_norm": 0.9047316908836365,
+      "learning_rate": 0.00017307966749907,
+      "loss": 1.7269,
+      "step": 11350
+    },
+    {
+      "epoch": 2.3932164112287353,
+      "grad_norm": 0.7869266867637634,
+      "learning_rate": 0.0001730344674188516,
+      "loss": 1.6686,
+      "step": 11360
+    },
+    {
+      "epoch": 2.3953231158160846,
+      "grad_norm": 0.8471513390541077,
+      "learning_rate": 0.00017298923533700107,
+      "loss": 1.7153,
+      "step": 11370
+    },
+    {
+      "epoch": 2.397429820403434,
+      "grad_norm": 1.0041714906692505,
+      "learning_rate": 0.00017294397127333785,
+      "loss": 1.6957,
+      "step": 11380
+    },
+    {
+      "epoch": 2.399536524990783,
+      "grad_norm": 0.8737631440162659,
+      "learning_rate": 0.00017289867524769537,
+      "loss": 1.7366,
+      "step": 11390
+    },
+    {
+      "epoch": 2.4016432295781325,
+      "grad_norm": 0.9573279619216919,
+      "learning_rate": 0.000172853347279921,
+      "loss": 1.7048,
+      "step": 11400
+    },
+    {
+      "epoch": 2.4037499341654818,
+      "grad_norm": 0.8812615275382996,
+      "learning_rate": 0.00017280798738987624,
+      "loss": 1.7328,
+      "step": 11410
+    },
+    {
+      "epoch": 2.405856638752831,
+      "grad_norm": 0.8517493605613708,
+      "learning_rate": 0.0001727625955974365,
+      "loss": 1.6838,
+      "step": 11420
+    },
+    {
+      "epoch": 2.40796334334018,
+      "grad_norm": 0.9157832264900208,
+      "learning_rate": 0.00017271717192249116,
+      "loss": 1.6942,
+      "step": 11430
+    },
+    {
+      "epoch": 2.410070047927529,
+      "grad_norm": 0.9208922982215881,
+      "learning_rate": 0.00017267171638494358,
+      "loss": 1.6941,
+      "step": 11440
+    },
+    {
+      "epoch": 2.4121767525148785,
+      "grad_norm": 0.896136462688446,
+      "learning_rate": 0.00017262622900471105,
+      "loss": 1.714,
+      "step": 11450
+    },
+    {
+      "epoch": 2.414283457102228,
+      "grad_norm": 0.8526272177696228,
+      "learning_rate": 0.00017258070980172494,
+      "loss": 1.6841,
+      "step": 11460
+    },
+    {
+      "epoch": 2.416390161689577,
+      "grad_norm": 0.9494383335113525,
+      "learning_rate": 0.00017253515879593043,
+      "loss": 1.7294,
+      "step": 11470
+    },
+    {
+      "epoch": 2.4184968662769264,
+      "grad_norm": 0.907871663570404,
+      "learning_rate": 0.00017248957600728664,
+      "loss": 1.7119,
+      "step": 11480
+    },
+    {
+      "epoch": 2.4206035708642757,
+      "grad_norm": 1.0190140008926392,
+      "learning_rate": 0.00017244396145576672,
+      "loss": 1.6932,
+      "step": 11490
+    },
+    {
+      "epoch": 2.422710275451625,
+      "grad_norm": 0.8717211484909058,
+      "learning_rate": 0.0001723983151613576,
+      "loss": 1.6928,
+      "step": 11500
+    },
+    {
+      "epoch": 2.4248169800389743,
+      "grad_norm": 0.8648276329040527,
+      "learning_rate": 0.0001723526371440603,
+      "loss": 1.7075,
+      "step": 11510
+    },
+    {
+      "epoch": 2.426923684626323,
+      "grad_norm": 0.8947681784629822,
+      "learning_rate": 0.0001723069274238895,
+      "loss": 1.6637,
+      "step": 11520
+    },
+    {
+      "epoch": 2.4290303892136724,
+      "grad_norm": 0.8753421306610107,
+      "learning_rate": 0.000172261186020874,
+      "loss": 1.7371,
+      "step": 11530
+    },
+    {
+      "epoch": 2.4311370938010217,
+      "grad_norm": 0.8403404951095581,
+      "learning_rate": 0.00017221541295505636,
+      "loss": 1.6949,
+      "step": 11540
+    },
+    {
+      "epoch": 2.433243798388371,
+      "grad_norm": 0.9043484926223755,
+      "learning_rate": 0.00017216960824649303,
+      "loss": 1.6836,
+      "step": 11550
+    },
+    {
+      "epoch": 2.4353505029757203,
+      "grad_norm": 0.8930955529212952,
+      "learning_rate": 0.00017212377191525434,
+      "loss": 1.7813,
+      "step": 11560
+    },
+    {
+      "epoch": 2.4374572075630696,
+      "grad_norm": 0.9738860130310059,
+      "learning_rate": 0.00017207790398142446,
+      "loss": 1.6875,
+      "step": 11570
+    },
+    {
+      "epoch": 2.439563912150419,
+      "grad_norm": 0.9277392029762268,
+      "learning_rate": 0.0001720320044651014,
+      "loss": 1.6789,
+      "step": 11580
+    },
+    {
+      "epoch": 2.4416706167377678,
+      "grad_norm": 0.8932106494903564,
+      "learning_rate": 0.0001719860733863971,
+      "loss": 1.7109,
+      "step": 11590
+    },
+    {
+      "epoch": 2.443777321325117,
+      "grad_norm": 0.9118972420692444,
+      "learning_rate": 0.00017194011076543717,
+      "loss": 1.7243,
+      "step": 11600
+    },
+    {
+      "epoch": 2.4458840259124663,
+      "grad_norm": 0.8895049691200256,
+      "learning_rate": 0.0001718941166223612,
+      "loss": 1.7096,
+      "step": 11610
+    },
+    {
+      "epoch": 2.4479907304998156,
+      "grad_norm": 0.8548287153244019,
+      "learning_rate": 0.00017184809097732246,
+      "loss": 1.6847,
+      "step": 11620
+    },
+    {
+      "epoch": 2.450097435087165,
+      "grad_norm": 0.8783696293830872,
+      "learning_rate": 0.00017180203385048812,
+      "loss": 1.7226,
+      "step": 11630
+    },
+    {
+      "epoch": 2.4522041396745142,
+      "grad_norm": 0.8651697039604187,
+      "learning_rate": 0.00017175594526203905,
+      "loss": 1.7217,
+      "step": 11640
+    },
+    {
+      "epoch": 2.4543108442618635,
+      "grad_norm": 0.9524984359741211,
+      "learning_rate": 0.00017170982523217004,
+      "loss": 1.7337,
+      "step": 11650
+    },
+    {
+      "epoch": 2.4564175488492124,
+      "grad_norm": 0.8341417908668518,
+      "learning_rate": 0.00017166367378108953,
+      "loss": 1.7415,
+      "step": 11660
+    },
+    {
+      "epoch": 2.4585242534365617,
+      "grad_norm": 0.9668021202087402,
+      "learning_rate": 0.00017161749092901984,
+      "loss": 1.6845,
+      "step": 11670
+    },
+    {
+      "epoch": 2.460630958023911,
+      "grad_norm": 1.0953508615493774,
+      "learning_rate": 0.00017157127669619688,
+      "loss": 1.7007,
+      "step": 11680
+    },
+    {
+      "epoch": 2.4627376626112603,
+      "grad_norm": 0.9357819557189941,
+      "learning_rate": 0.00017152503110287048,
+      "loss": 1.7105,
+      "step": 11690
+    },
+    {
+      "epoch": 2.4648443671986096,
+      "grad_norm": 1.1564433574676514,
+      "learning_rate": 0.00017147875416930416,
+      "loss": 1.6699,
+      "step": 11700
+    },
+    {
+      "epoch": 2.466951071785959,
+      "grad_norm": 0.9135658740997314,
+      "learning_rate": 0.00017143244591577515,
+      "loss": 1.6882,
+      "step": 11710
+    },
+    {
+      "epoch": 2.469057776373308,
+      "grad_norm": 0.9228572249412537,
+      "learning_rate": 0.00017138610636257436,
+      "loss": 1.6808,
+      "step": 11720
+    },
+    {
+      "epoch": 2.4711644809606574,
+      "grad_norm": 0.914090096950531,
+      "learning_rate": 0.00017133973553000654,
+      "loss": 1.6233,
+      "step": 11730
+    },
+    {
+      "epoch": 2.4732711855480067,
+      "grad_norm": 0.9056432843208313,
+      "learning_rate": 0.00017129333343839003,
+      "loss": 1.7001,
+      "step": 11740
+    },
+    {
+      "epoch": 2.4753778901353556,
+      "grad_norm": 0.9140064120292664,
+      "learning_rate": 0.00017124690010805692,
+      "loss": 1.7155,
+      "step": 11750
+    },
+    {
+      "epoch": 2.477484594722705,
+      "grad_norm": 0.8333889245986938,
+      "learning_rate": 0.00017120043555935298,
+      "loss": 1.6708,
+      "step": 11760
+    },
+    {
+      "epoch": 2.479591299310054,
+      "grad_norm": 0.8906539678573608,
+      "learning_rate": 0.00017115393981263768,
+      "loss": 1.694,
+      "step": 11770
+    },
+    {
+      "epoch": 2.4816980038974035,
+      "grad_norm": 0.8971114158630371,
+      "learning_rate": 0.0001711074128882841,
+      "loss": 1.7009,
+      "step": 11780
+    },
+    {
+      "epoch": 2.4838047084847528,
+      "grad_norm": 0.856102466583252,
+      "learning_rate": 0.00017106085480667903,
+      "loss": 1.7115,
+      "step": 11790
+    },
+    {
+      "epoch": 2.485911413072102,
+      "grad_norm": 0.9299646019935608,
+      "learning_rate": 0.00017101426558822292,
+      "loss": 1.7067,
+      "step": 11800
+    },
+    {
+      "epoch": 2.4880181176594514,
+      "grad_norm": 0.9112943410873413,
+      "learning_rate": 0.00017096764525332986,
+      "loss": 1.6979,
+      "step": 11810
+    },
+    {
+      "epoch": 2.4901248222468,
+      "grad_norm": 0.917447566986084,
+      "learning_rate": 0.00017092099382242748,
+      "loss": 1.7305,
+      "step": 11820
+    },
+    {
+      "epoch": 2.4922315268341495,
+      "grad_norm": 0.9493505358695984,
+      "learning_rate": 0.0001708743113159572,
+      "loss": 1.6831,
+      "step": 11830
+    },
+    {
+      "epoch": 2.494338231421499,
+      "grad_norm": 0.9228153824806213,
+      "learning_rate": 0.0001708275977543739,
+      "loss": 1.6944,
+      "step": 11840
+    },
+    {
+      "epoch": 2.496444936008848,
+      "grad_norm": 0.8501965999603271,
+      "learning_rate": 0.0001707808531581462,
+      "loss": 1.7237,
+      "step": 11850
+    },
+    {
+      "epoch": 2.4985516405961974,
+      "grad_norm": 0.9229296445846558,
+      "learning_rate": 0.00017073407754775622,
+      "loss": 1.7281,
+      "step": 11860
+    },
+    {
+      "epoch": 2.5006583451835467,
+      "grad_norm": 0.9164074659347534,
+      "learning_rate": 0.0001706872709436997,
+      "loss": 1.7504,
+      "step": 11870
+    },
+    {
+      "epoch": 2.502765049770896,
+      "grad_norm": 0.9984068274497986,
+      "learning_rate": 0.00017064043336648599,
+      "loss": 1.7848,
+      "step": 11880
+    },
+    {
+      "epoch": 2.5048717543582453,
+      "grad_norm": 0.9881888031959534,
+      "learning_rate": 0.00017059356483663796,
+      "loss": 1.6548,
+      "step": 11890
+    },
+    {
+      "epoch": 2.5069784589455946,
+      "grad_norm": 0.8894940614700317,
+      "learning_rate": 0.00017054666537469213,
+      "loss": 1.6994,
+      "step": 11900
+    },
+    {
+      "epoch": 2.5090851635329434,
+      "grad_norm": 0.93770432472229,
+      "learning_rate": 0.00017049973500119845,
+      "loss": 1.6519,
+      "step": 11910
+    },
+    {
+      "epoch": 2.5111918681202927,
+      "grad_norm": 0.9530473351478577,
+      "learning_rate": 0.00017045277373672047,
+      "loss": 1.7667,
+      "step": 11920
+    },
+    {
+      "epoch": 2.513298572707642,
+      "grad_norm": 1.0557433366775513,
+      "learning_rate": 0.00017040578160183536,
+      "loss": 1.6852,
+      "step": 11930
+    },
+    {
+      "epoch": 2.5154052772949913,
+      "grad_norm": 0.9705979228019714,
+      "learning_rate": 0.0001703587586171337,
+      "loss": 1.7019,
+      "step": 11940
+    },
+    {
+      "epoch": 2.5175119818823406,
+      "grad_norm": 0.9046064019203186,
+      "learning_rate": 0.0001703117048032196,
+      "loss": 1.7381,
+      "step": 11950
+    },
+    {
+      "epoch": 2.51961868646969,
+      "grad_norm": 0.9303194880485535,
+      "learning_rate": 0.0001702646201807107,
+      "loss": 1.692,
+      "step": 11960
+    },
+    {
+      "epoch": 2.521725391057039,
+      "grad_norm": 0.903297483921051,
+      "learning_rate": 0.0001702175047702382,
+      "loss": 1.6716,
+      "step": 11970
+    },
+    {
+      "epoch": 2.523832095644388,
+      "grad_norm": 0.8236584067344666,
+      "learning_rate": 0.00017017035859244673,
+      "loss": 1.7018,
+      "step": 11980
+    },
+    {
+      "epoch": 2.525938800231738,
+      "grad_norm": 0.8487248420715332,
+      "learning_rate": 0.00017012318166799437,
+      "loss": 1.7035,
+      "step": 11990
+    },
+    {
+      "epoch": 2.5280455048190866,
+      "grad_norm": 0.8535507917404175,
+      "learning_rate": 0.00017007597401755276,
+      "loss": 1.6756,
+      "step": 12000
+    },
+    {
+      "epoch": 2.530152209406436,
+      "grad_norm": 0.8676731586456299,
+      "learning_rate": 0.00017002873566180688,
+      "loss": 1.7124,
+      "step": 12010
+    },
+    {
+      "epoch": 2.532258913993785,
+      "grad_norm": 0.9095235466957092,
+      "learning_rate": 0.0001699814666214553,
+      "loss": 1.6922,
+      "step": 12020
+    },
+    {
+      "epoch": 2.5343656185811345,
+      "grad_norm": 0.8557278513908386,
+      "learning_rate": 0.00016993416691720998,
+      "loss": 1.6995,
+      "step": 12030
+    },
+    {
+      "epoch": 2.536472323168484,
+      "grad_norm": 0.9567210674285889,
+      "learning_rate": 0.00016988683656979624,
+      "loss": 1.6733,
+      "step": 12040
+    },
+    {
+      "epoch": 2.5385790277558327,
+      "grad_norm": 0.8367411494255066,
+      "learning_rate": 0.00016983947559995297,
+      "loss": 1.7573,
+      "step": 12050
+    },
+    {
+      "epoch": 2.5406857323431824,
+      "grad_norm": 0.9761552810668945,
+      "learning_rate": 0.00016979208402843237,
+      "loss": 1.7031,
+      "step": 12060
+    },
+    {
+      "epoch": 2.5427924369305313,
+      "grad_norm": 0.9026761651039124,
+      "learning_rate": 0.0001697446618760001,
+      "loss": 1.6968,
+      "step": 12070
+    },
+    {
+      "epoch": 2.5448991415178805,
+      "grad_norm": 0.8959203362464905,
+      "learning_rate": 0.00016969720916343515,
+      "loss": 1.6517,
+      "step": 12080
+    },
+    {
+      "epoch": 2.54700584610523,
+      "grad_norm": 0.8353654742240906,
+      "learning_rate": 0.00016964972591153,
+      "loss": 1.6856,
+      "step": 12090
+    },
+    {
+      "epoch": 2.549112550692579,
+      "grad_norm": 1.0780528783798218,
+      "learning_rate": 0.00016960221214109045,
+      "loss": 1.6545,
+      "step": 12100
+    },
+    {
+      "epoch": 2.5512192552799284,
+      "grad_norm": 0.8808447122573853,
+      "learning_rate": 0.00016955466787293576,
+      "loss": 1.6997,
+      "step": 12110
+    },
+    {
+      "epoch": 2.5533259598672777,
+      "grad_norm": 0.9192379117012024,
+      "learning_rate": 0.00016950709312789833,
+      "loss": 1.7434,
+      "step": 12120
+    },
+    {
+      "epoch": 2.555432664454627,
+      "grad_norm": 0.9900842905044556,
+      "learning_rate": 0.00016945948792682417,
+      "loss": 1.7091,
+      "step": 12130
+    },
+    {
+      "epoch": 2.557539369041976,
+      "grad_norm": 0.9383872747421265,
+      "learning_rate": 0.0001694118522905725,
+      "loss": 1.6993,
+      "step": 12140
+    },
+    {
+      "epoch": 2.559646073629325,
+      "grad_norm": 0.9453622102737427,
+      "learning_rate": 0.00016936418624001592,
+      "loss": 1.7373,
+      "step": 12150
+    },
+    {
+      "epoch": 2.5617527782166745,
+      "grad_norm": 0.8687297701835632,
+      "learning_rate": 0.0001693164897960403,
+      "loss": 1.7023,
+      "step": 12160
+    },
+    {
+      "epoch": 2.5638594828040238,
+      "grad_norm": 0.8873741030693054,
+      "learning_rate": 0.00016926876297954492,
+      "loss": 1.6679,
+      "step": 12170
+    },
+    {
+      "epoch": 2.565966187391373,
+      "grad_norm": 0.9118444919586182,
+      "learning_rate": 0.00016922100581144228,
+      "loss": 1.702,
+      "step": 12180
+    },
+    {
+      "epoch": 2.5680728919787223,
+      "grad_norm": 0.9042273759841919,
+      "learning_rate": 0.00016917321831265826,
+      "loss": 1.7259,
+      "step": 12190
+    },
+    {
+      "epoch": 2.5701795965660716,
+      "grad_norm": 0.9207640886306763,
+      "learning_rate": 0.00016912540050413195,
+      "loss": 1.6712,
+      "step": 12200
+    },
+    {
+      "epoch": 2.5722863011534205,
+      "grad_norm": 0.9605659246444702,
+      "learning_rate": 0.00016907755240681577,
+      "loss": 1.7248,
+      "step": 12210
+    },
+    {
+      "epoch": 2.5743930057407702,
+      "grad_norm": 0.9034086465835571,
+      "learning_rate": 0.00016902967404167538,
+      "loss": 1.6718,
+      "step": 12220
+    },
+    {
+      "epoch": 2.576499710328119,
+      "grad_norm": 0.9003251791000366,
+      "learning_rate": 0.00016898176542968975,
+      "loss": 1.7107,
+      "step": 12230
+    },
+    {
+      "epoch": 2.5786064149154684,
+      "grad_norm": 0.9571187496185303,
+      "learning_rate": 0.00016893382659185105,
+      "loss": 1.671,
+      "step": 12240
+    },
+    {
+      "epoch": 2.5807131195028177,
+      "grad_norm": 0.9620792269706726,
+      "learning_rate": 0.00016888585754916476,
+      "loss": 1.7082,
+      "step": 12250
+    },
+    {
+      "epoch": 2.582819824090167,
+      "grad_norm": 0.9372028112411499,
+      "learning_rate": 0.0001688378583226495,
+      "loss": 1.7293,
+      "step": 12260
+    },
+    {
+      "epoch": 2.5849265286775163,
+      "grad_norm": 0.9726297855377197,
+      "learning_rate": 0.00016878982893333717,
+      "loss": 1.7104,
+      "step": 12270
+    },
+    {
+      "epoch": 2.5870332332648656,
+      "grad_norm": 1.0220943689346313,
+      "learning_rate": 0.00016874176940227296,
+      "loss": 1.7373,
+      "step": 12280
+    },
+    {
+      "epoch": 2.589139937852215,
+      "grad_norm": 0.8310849666595459,
+      "learning_rate": 0.0001686936797505151,
+      "loss": 1.6991,
+      "step": 12290
+    },
+    {
+      "epoch": 2.5912466424395637,
+      "grad_norm": 0.8683304190635681,
+      "learning_rate": 0.00016864555999913518,
+      "loss": 1.7168,
+      "step": 12300
+    },
+    {
+      "epoch": 2.593353347026913,
+      "grad_norm": 0.8827438354492188,
+      "learning_rate": 0.00016859741016921786,
+      "loss": 1.702,
+      "step": 12310
+    },
+    {
+      "epoch": 2.5954600516142623,
+      "grad_norm": 0.91736900806427,
+      "learning_rate": 0.00016854923028186111,
+      "loss": 1.6877,
+      "step": 12320
+    },
+    {
+      "epoch": 2.5975667562016116,
+      "grad_norm": 0.8782617449760437,
+      "learning_rate": 0.00016850102035817588,
+      "loss": 1.7195,
+      "step": 12330
+    },
+    {
+      "epoch": 2.599673460788961,
+      "grad_norm": 0.9509816765785217,
+      "learning_rate": 0.0001684527804192865,
+      "loss": 1.7482,
+      "step": 12340
+    },
+    {
+      "epoch": 2.60178016537631,
+      "grad_norm": 0.9325737357139587,
+      "learning_rate": 0.00016840451048633025,
+      "loss": 1.6947,
+      "step": 12350
+    },
+    {
+      "epoch": 2.6038868699636595,
+      "grad_norm": 0.9182351231575012,
+      "learning_rate": 0.0001683562105804577,
+      "loss": 1.6918,
+      "step": 12360
+    },
+    {
+      "epoch": 2.6059935745510083,
+      "grad_norm": 0.919793426990509,
+      "learning_rate": 0.00016830788072283247,
+      "loss": 1.7496,
+      "step": 12370
+    },
+    {
+      "epoch": 2.608100279138358,
+      "grad_norm": 0.8799266219139099,
+      "learning_rate": 0.00016825952093463135,
+      "loss": 1.6664,
+      "step": 12380
+    },
+    {
+      "epoch": 2.610206983725707,
+      "grad_norm": 0.9467419385910034,
+      "learning_rate": 0.00016821113123704424,
+      "loss": 1.7167,
+      "step": 12390
+    },
+    {
+      "epoch": 2.612313688313056,
+      "grad_norm": 0.9151462316513062,
+      "learning_rate": 0.00016816271165127412,
+      "loss": 1.7149,
+      "step": 12400
+    },
+    {
+      "epoch": 2.6144203929004055,
+      "grad_norm": 0.895194947719574,
+      "learning_rate": 0.00016811426219853702,
+      "loss": 1.7368,
+      "step": 12410
+    },
+    {
+      "epoch": 2.616527097487755,
+      "grad_norm": 1.095924735069275,
+      "learning_rate": 0.00016806578290006225,
+      "loss": 1.6619,
+      "step": 12420
+    },
+    {
+      "epoch": 2.618633802075104,
+      "grad_norm": 0.8870509266853333,
+      "learning_rate": 0.00016801727377709194,
+      "loss": 1.6918,
+      "step": 12430
+    },
+    {
+      "epoch": 2.6207405066624534,
+      "grad_norm": 0.967845618724823,
+      "learning_rate": 0.00016796873485088148,
+      "loss": 1.7309,
+      "step": 12440
+    },
+    {
+      "epoch": 2.6228472112498027,
+      "grad_norm": 1.0131421089172363,
+      "learning_rate": 0.00016792016614269924,
+      "loss": 1.6849,
+      "step": 12450
+    },
+    {
+      "epoch": 2.6249539158371515,
+      "grad_norm": 0.9242698550224304,
+      "learning_rate": 0.00016787156767382659,
+      "loss": 1.6903,
+      "step": 12460
+    },
+    {
+      "epoch": 2.627060620424501,
+      "grad_norm": 0.8336485624313354,
+      "learning_rate": 0.00016782293946555806,
+      "loss": 1.7431,
+      "step": 12470
+    },
+    {
+      "epoch": 2.62916732501185,
+      "grad_norm": 0.8880660533905029,
+      "learning_rate": 0.0001677742815392012,
+      "loss": 1.6785,
+      "step": 12480
+    },
+    {
+      "epoch": 2.6312740295991994,
+      "grad_norm": 0.918901264667511,
+      "learning_rate": 0.00016772559391607642,
+      "loss": 1.6933,
+      "step": 12490
+    },
+    {
+      "epoch": 2.6333807341865487,
+      "grad_norm": 1.0502138137817383,
+      "learning_rate": 0.00016767687661751733,
+      "loss": 1.7067,
+      "step": 12500
+    },
+    {
+      "epoch": 2.635487438773898,
+      "grad_norm": 0.842281699180603,
+      "learning_rate": 0.00016762812966487044,
+      "loss": 1.7439,
+      "step": 12510
+    },
+    {
+      "epoch": 2.6375941433612473,
+      "grad_norm": 0.9271451234817505,
+      "learning_rate": 0.0001675793530794953,
+      "loss": 1.7082,
+      "step": 12520
+    },
+    {
+      "epoch": 2.639700847948596,
+      "grad_norm": 0.9502926468849182,
+      "learning_rate": 0.0001675305468827644,
+      "loss": 1.7357,
+      "step": 12530
+    },
+    {
+      "epoch": 2.641807552535946,
+      "grad_norm": 0.9139789342880249,
+      "learning_rate": 0.00016748171109606328,
+      "loss": 1.6743,
+      "step": 12540
+    },
+    {
+      "epoch": 2.6439142571232948,
+      "grad_norm": 0.8416628837585449,
+      "learning_rate": 0.00016743284574079033,
+      "loss": 1.7216,
+      "step": 12550
+    },
+    {
+      "epoch": 2.646020961710644,
+      "grad_norm": 0.8748312592506409,
+      "learning_rate": 0.000167383950838357,
+      "loss": 1.678,
+      "step": 12560
+    },
+    {
+      "epoch": 2.6481276662979933,
+      "grad_norm": 0.8777117729187012,
+      "learning_rate": 0.00016733502641018766,
+      "loss": 1.682,
+      "step": 12570
+    },
+    {
+      "epoch": 2.6502343708853426,
+      "grad_norm": 0.9002516269683838,
+      "learning_rate": 0.00016728607247771957,
+      "loss": 1.7076,
+      "step": 12580
+    },
+    {
+      "epoch": 2.652341075472692,
+      "grad_norm": 0.9774075746536255,
+      "learning_rate": 0.000167237089062403,
+      "loss": 1.7174,
+      "step": 12590
+    },
+    {
+      "epoch": 2.6544477800600412,
+      "grad_norm": 0.8962463140487671,
+      "learning_rate": 0.00016718807618570106,
+      "loss": 1.7087,
+      "step": 12600
+    },
+    {
+      "epoch": 2.6565544846473905,
+      "grad_norm": 0.9497388601303101,
+      "learning_rate": 0.00016713903386908984,
+      "loss": 1.7252,
+      "step": 12610
+    },
+    {
+      "epoch": 2.6586611892347394,
+      "grad_norm": 0.907230794429779,
+      "learning_rate": 0.00016708996213405826,
+      "loss": 1.7184,
+      "step": 12620
+    },
+    {
+      "epoch": 2.6607678938220887,
+      "grad_norm": 0.8758313059806824,
+      "learning_rate": 0.00016704086100210815,
+      "loss": 1.7283,
+      "step": 12630
+    },
+    {
+      "epoch": 2.662874598409438,
+      "grad_norm": 0.9444327354431152,
+      "learning_rate": 0.00016699173049475425,
+      "loss": 1.7191,
+      "step": 12640
+    },
+    {
+      "epoch": 2.6649813029967873,
+      "grad_norm": 0.967747151851654,
+      "learning_rate": 0.00016694257063352416,
+      "loss": 1.7209,
+      "step": 12650
+    },
+    {
+      "epoch": 2.6670880075841366,
+      "grad_norm": 0.8999049067497253,
+      "learning_rate": 0.00016689338143995833,
+      "loss": 1.6764,
+      "step": 12660
+    },
+    {
+      "epoch": 2.669194712171486,
+      "grad_norm": 0.9095073342323303,
+      "learning_rate": 0.0001668441629356101,
+      "loss": 1.6978,
+      "step": 12670
+    },
+    {
+      "epoch": 2.671301416758835,
+      "grad_norm": 0.9646686315536499,
+      "learning_rate": 0.00016679491514204556,
+      "loss": 1.6882,
+      "step": 12680
+    },
+    {
+      "epoch": 2.673408121346184,
+      "grad_norm": 0.9153634309768677,
+      "learning_rate": 0.00016674563808084377,
+      "loss": 1.747,
+      "step": 12690
+    },
+    {
+      "epoch": 2.6755148259335337,
+      "grad_norm": 0.9942224621772766,
+      "learning_rate": 0.00016669633177359647,
+      "loss": 1.6708,
+      "step": 12700
+    },
+    {
+      "epoch": 2.6776215305208826,
+      "grad_norm": 0.9104297757148743,
+      "learning_rate": 0.0001666469962419083,
+      "loss": 1.6708,
+      "step": 12710
+    },
+    {
+      "epoch": 2.679728235108232,
+      "grad_norm": 0.8879741430282593,
+      "learning_rate": 0.00016659763150739677,
+      "loss": 1.7089,
+      "step": 12720
+    },
+    {
+      "epoch": 2.681834939695581,
+      "grad_norm": 0.9687016606330872,
+      "learning_rate": 0.00016654823759169199,
+      "loss": 1.7089,
+      "step": 12730
+    },
+    {
+      "epoch": 2.6839416442829305,
+      "grad_norm": 0.9552226662635803,
+      "learning_rate": 0.00016649881451643705,
+      "loss": 1.7382,
+      "step": 12740
+    },
+    {
+      "epoch": 2.6860483488702798,
+      "grad_norm": 0.9607418775558472,
+      "learning_rate": 0.0001664493623032877,
+      "loss": 1.7465,
+      "step": 12750
+    },
+    {
+      "epoch": 2.688155053457629,
+      "grad_norm": 1.0158402919769287,
+      "learning_rate": 0.00016639988097391252,
+      "loss": 1.7237,
+      "step": 12760
+    },
+    {
+      "epoch": 2.6902617580449784,
+      "grad_norm": 0.9217173457145691,
+      "learning_rate": 0.00016635037054999285,
+      "loss": 1.6807,
+      "step": 12770
+    },
+    {
+      "epoch": 2.692368462632327,
+      "grad_norm": 0.8327500820159912,
+      "learning_rate": 0.00016630083105322266,
+      "loss": 1.709,
+      "step": 12780
+    },
+    {
+      "epoch": 2.6944751672196765,
+      "grad_norm": 0.8606378436088562,
+      "learning_rate": 0.0001662512625053089,
+      "loss": 1.685,
+      "step": 12790
+    },
+    {
+      "epoch": 2.696581871807026,
+      "grad_norm": 1.037510871887207,
+      "learning_rate": 0.00016620166492797096,
+      "loss": 1.7485,
+      "step": 12800
+    },
+    {
+      "epoch": 2.698688576394375,
+      "grad_norm": 0.8210626244544983,
+      "learning_rate": 0.00016615203834294119,
+      "loss": 1.6761,
+      "step": 12810
+    },
+    {
+      "epoch": 2.7007952809817244,
+      "grad_norm": 0.9673810601234436,
+      "learning_rate": 0.00016610238277196446,
+      "loss": 1.7397,
+      "step": 12820
+    },
+    {
+      "epoch": 2.7029019855690737,
+      "grad_norm": 0.9024403691291809,
+      "learning_rate": 0.00016605269823679851,
+      "loss": 1.6665,
+      "step": 12830
+    },
+    {
+      "epoch": 2.705008690156423,
+      "grad_norm": 0.933525800704956,
+      "learning_rate": 0.00016600298475921365,
+      "loss": 1.7333,
+      "step": 12840
+    },
+    {
+      "epoch": 2.707115394743772,
+      "grad_norm": 0.9351423978805542,
+      "learning_rate": 0.00016595324236099294,
+      "loss": 1.7284,
+      "step": 12850
+    },
+    {
+      "epoch": 2.7092220993311216,
+      "grad_norm": 1.0248723030090332,
+      "learning_rate": 0.0001659034710639321,
+      "loss": 1.7172,
+      "step": 12860
+    },
+    {
+      "epoch": 2.7113288039184704,
+      "grad_norm": 0.8728225827217102,
+      "learning_rate": 0.00016585367088983946,
+      "loss": 1.6741,
+      "step": 12870
+    },
+    {
+      "epoch": 2.7134355085058197,
+      "grad_norm": 0.9945250153541565,
+      "learning_rate": 0.0001658038418605361,
+      "loss": 1.7471,
+      "step": 12880
+    },
+    {
+      "epoch": 2.715542213093169,
+      "grad_norm": 0.9469719529151917,
+      "learning_rate": 0.00016575398399785562,
+      "loss": 1.7206,
+      "step": 12890
+    },
+    {
+      "epoch": 2.7176489176805183,
+      "grad_norm": 0.8776702880859375,
+      "learning_rate": 0.00016570409732364437,
+      "loss": 1.7085,
+      "step": 12900
+    },
+    {
+      "epoch": 2.7197556222678676,
+      "grad_norm": 0.8826903104782104,
+      "learning_rate": 0.00016565418185976127,
+      "loss": 1.6886,
+      "step": 12910
+    },
+    {
+      "epoch": 2.7218623268552165,
+      "grad_norm": 0.8356149196624756,
+      "learning_rate": 0.00016560423762807783,
+      "loss": 1.7057,
+      "step": 12920
+    },
+    {
+      "epoch": 2.723969031442566,
+      "grad_norm": 0.8950599431991577,
+      "learning_rate": 0.00016555426465047823,
+      "loss": 1.7086,
+      "step": 12930
+    },
+    {
+      "epoch": 2.726075736029915,
+      "grad_norm": 0.9071326851844788,
+      "learning_rate": 0.00016550426294885919,
+      "loss": 1.6641,
+      "step": 12940
+    },
+    {
+      "epoch": 2.7281824406172643,
+      "grad_norm": 0.9181045293807983,
+      "learning_rate": 0.00016545423254513004,
+      "loss": 1.7222,
+      "step": 12950
+    },
+    {
+      "epoch": 2.7302891452046136,
+      "grad_norm": 0.9889811277389526,
+      "learning_rate": 0.0001654041734612127,
+      "loss": 1.6696,
+      "step": 12960
+    },
+    {
+      "epoch": 2.732395849791963,
+      "grad_norm": 0.9322980046272278,
+      "learning_rate": 0.00016535408571904164,
+      "loss": 1.6845,
+      "step": 12970
+    },
+    {
+      "epoch": 2.7345025543793122,
+      "grad_norm": 0.9056413769721985,
+      "learning_rate": 0.00016530396934056384,
+      "loss": 1.738,
+      "step": 12980
+    },
+    {
+      "epoch": 2.7366092589666615,
+      "grad_norm": 1.0681997537612915,
+      "learning_rate": 0.00016525382434773894,
+      "loss": 1.7147,
+      "step": 12990
+    },
+    {
+      "epoch": 2.738715963554011,
+      "grad_norm": 0.9294453263282776,
+      "learning_rate": 0.00016520365076253904,
+      "loss": 1.7205,
+      "step": 13000
+    },
+    {
+      "epoch": 2.7408226681413597,
+      "grad_norm": 0.950069010257721,
+      "learning_rate": 0.00016515344860694876,
+      "loss": 1.7498,
+      "step": 13010
+    },
+    {
+      "epoch": 2.742929372728709,
+      "grad_norm": 0.9262808561325073,
+      "learning_rate": 0.00016510321790296525,
+      "loss": 1.7285,
+      "step": 13020
+    },
+    {
+      "epoch": 2.7450360773160583,
+      "grad_norm": 0.865739643573761,
+      "learning_rate": 0.00016505295867259823,
+      "loss": 1.6656,
+      "step": 13030
+    },
+    {
+      "epoch": 2.7471427819034075,
+      "grad_norm": 1.1308355331420898,
+      "learning_rate": 0.0001650026709378698,
+      "loss": 1.7217,
+      "step": 13040
+    },
+    {
+      "epoch": 2.749249486490757,
+      "grad_norm": 0.882122814655304,
+      "learning_rate": 0.00016495235472081468,
+      "loss": 1.6862,
+      "step": 13050
+    },
+    {
+      "epoch": 2.751356191078106,
+      "grad_norm": 0.928149402141571,
+      "learning_rate": 0.00016490201004348,
+      "loss": 1.7385,
+      "step": 13060
+    },
+    {
+      "epoch": 2.7534628956654554,
+      "grad_norm": 0.8748145699501038,
+      "learning_rate": 0.00016485163692792534,
+      "loss": 1.7096,
+      "step": 13070
+    },
+    {
+      "epoch": 2.7555696002528043,
+      "grad_norm": 0.8686920404434204,
+      "learning_rate": 0.00016480123539622281,
+      "loss": 1.7108,
+      "step": 13080
+    },
+    {
+      "epoch": 2.757676304840154,
+      "grad_norm": 0.90756756067276,
+      "learning_rate": 0.00016475080547045687,
+      "loss": 1.6767,
+      "step": 13090
+    },
+    {
+      "epoch": 2.759783009427503,
+      "grad_norm": 0.9196756482124329,
+      "learning_rate": 0.00016470034717272456,
+      "loss": 1.7034,
+      "step": 13100
+    },
+    {
+      "epoch": 2.761889714014852,
+      "grad_norm": 0.9286429286003113,
+      "learning_rate": 0.0001646498605251352,
+      "loss": 1.6663,
+      "step": 13110
+    },
+    {
+      "epoch": 2.7639964186022015,
+      "grad_norm": 0.911522388458252,
+      "learning_rate": 0.00016459934554981066,
+      "loss": 1.7007,
+      "step": 13120
+    },
+    {
+      "epoch": 2.7661031231895508,
+      "grad_norm": 0.8197303414344788,
+      "learning_rate": 0.00016454880226888515,
+      "loss": 1.6557,
+      "step": 13130
+    },
+    {
+      "epoch": 2.7682098277769,
+      "grad_norm": 0.9045342206954956,
+      "learning_rate": 0.00016449823070450531,
+      "loss": 1.6377,
+      "step": 13140
+    },
+    {
+      "epoch": 2.7703165323642494,
+      "grad_norm": 0.9467363953590393,
+      "learning_rate": 0.0001644476308788302,
+      "loss": 1.6749,
+      "step": 13150
+    },
+    {
+      "epoch": 2.7724232369515986,
+      "grad_norm": 0.8628239035606384,
+      "learning_rate": 0.00016439700281403114,
+      "loss": 1.7064,
+      "step": 13160
+    },
+    {
+      "epoch": 2.7745299415389475,
+      "grad_norm": 0.8383623957633972,
+      "learning_rate": 0.00016434634653229199,
+      "loss": 1.6991,
+      "step": 13170
+    },
+    {
+      "epoch": 2.776636646126297,
+      "grad_norm": 0.870199978351593,
+      "learning_rate": 0.00016429566205580884,
+      "loss": 1.7157,
+      "step": 13180
+    },
+    {
+      "epoch": 2.778743350713646,
+      "grad_norm": 0.9087043404579163,
+      "learning_rate": 0.00016424494940679024,
+      "loss": 1.742,
+      "step": 13190
+    },
+    {
+      "epoch": 2.7808500553009954,
+      "grad_norm": 0.9809213280677795,
+      "learning_rate": 0.00016419420860745699,
+      "loss": 1.7031,
+      "step": 13200
+    },
+    {
+      "epoch": 2.7829567598883447,
+      "grad_norm": 0.9322521686553955,
+      "learning_rate": 0.0001641434396800423,
+      "loss": 1.7097,
+      "step": 13210
+    },
+    {
+      "epoch": 2.785063464475694,
+      "grad_norm": 0.8767238259315491,
+      "learning_rate": 0.00016409264264679164,
+      "loss": 1.641,
+      "step": 13220
+    },
+    {
+      "epoch": 2.7871701690630433,
+      "grad_norm": 0.9374739527702332,
+      "learning_rate": 0.00016404181752996289,
+      "loss": 1.6943,
+      "step": 13230
+    },
+    {
+      "epoch": 2.789276873650392,
+      "grad_norm": 0.9542452096939087,
+      "learning_rate": 0.00016399096435182613,
+      "loss": 1.6838,
+      "step": 13240
+    },
+    {
+      "epoch": 2.791383578237742,
+      "grad_norm": 0.9805099368095398,
+      "learning_rate": 0.00016394008313466376,
+      "loss": 1.7133,
+      "step": 13250
+    },
+    {
+      "epoch": 2.7934902828250907,
+      "grad_norm": 0.8735339045524597,
+      "learning_rate": 0.00016388917390077054,
+      "loss": 1.7007,
+      "step": 13260
+    },
+    {
+      "epoch": 2.79559698741244,
+      "grad_norm": 0.8826082944869995,
+      "learning_rate": 0.00016383823667245344,
+      "loss": 1.7285,
+      "step": 13270
+    },
+    {
+      "epoch": 2.7977036919997893,
+      "grad_norm": 1.036288857460022,
+      "learning_rate": 0.00016378727147203166,
+      "loss": 1.717,
+      "step": 13280
+    },
+    {
+      "epoch": 2.7998103965871386,
+      "grad_norm": 0.9070268869400024,
+      "learning_rate": 0.0001637362783218368,
+      "loss": 1.7483,
+      "step": 13290
+    },
+    {
+      "epoch": 2.801917101174488,
+      "grad_norm": 0.9468526244163513,
+      "learning_rate": 0.00016368525724421248,
+      "loss": 1.716,
+      "step": 13300
+    },
+    {
+      "epoch": 2.804023805761837,
+      "grad_norm": 0.9099157452583313,
+      "learning_rate": 0.00016363420826151482,
+      "loss": 1.7263,
+      "step": 13310
+    },
+    {
+      "epoch": 2.8061305103491865,
+      "grad_norm": 0.9246546626091003,
+      "learning_rate": 0.00016358313139611195,
+      "loss": 1.7236,
+      "step": 13320
+    },
+    {
+      "epoch": 2.8082372149365353,
+      "grad_norm": 0.9395228624343872,
+      "learning_rate": 0.00016353202667038433,
+      "loss": 1.7252,
+      "step": 13330
+    },
+    {
+      "epoch": 2.8103439195238846,
+      "grad_norm": 0.9023792743682861,
+      "learning_rate": 0.0001634808941067246,
+      "loss": 1.7201,
+      "step": 13340
+    },
+    {
+      "epoch": 2.812450624111234,
+      "grad_norm": 0.9175665974617004,
+      "learning_rate": 0.0001634297337275376,
+      "loss": 1.7183,
+      "step": 13350
+    },
+    {
+      "epoch": 2.814557328698583,
+      "grad_norm": 0.9023197293281555,
+      "learning_rate": 0.00016337854555524038,
+      "loss": 1.6694,
+      "step": 13360
+    },
+    {
+      "epoch": 2.8166640332859325,
+      "grad_norm": 0.9740870594978333,
+      "learning_rate": 0.0001633273296122621,
+      "loss": 1.6947,
+      "step": 13370
+    },
+    {
+      "epoch": 2.818770737873282,
+      "grad_norm": 0.822769820690155,
+      "learning_rate": 0.0001632760859210442,
+      "loss": 1.7328,
+      "step": 13380
+    },
+    {
+      "epoch": 2.820877442460631,
+      "grad_norm": 0.8751346468925476,
+      "learning_rate": 0.00016322481450404015,
+      "loss": 1.6763,
+      "step": 13390
+    },
+    {
+      "epoch": 2.82298414704798,
+      "grad_norm": 0.9259299635887146,
+      "learning_rate": 0.00016317351538371565,
+      "loss": 1.6463,
+      "step": 13400
+    },
+    {
+      "epoch": 2.8250908516353297,
+      "grad_norm": 0.9039295315742493,
+      "learning_rate": 0.0001631221885825485,
+      "loss": 1.6898,
+      "step": 13410
+    },
+    {
+      "epoch": 2.8271975562226785,
+      "grad_norm": 0.8728259801864624,
+      "learning_rate": 0.0001630708341230287,
+      "loss": 1.6981,
+      "step": 13420
+    },
+    {
+      "epoch": 2.829304260810028,
+      "grad_norm": 0.8956354856491089,
+      "learning_rate": 0.00016301945202765826,
+      "loss": 1.6968,
+      "step": 13430
+    },
+    {
+      "epoch": 2.831410965397377,
+      "grad_norm": 1.1208888292312622,
+      "learning_rate": 0.00016296804231895142,
+      "loss": 1.7184,
+      "step": 13440
+    },
+    {
+      "epoch": 2.8335176699847264,
+      "grad_norm": 0.9479233622550964,
+      "learning_rate": 0.00016291660501943438,
+      "loss": 1.7048,
+      "step": 13450
+    },
+    {
+      "epoch": 2.8356243745720757,
+      "grad_norm": 0.8727468848228455,
+      "learning_rate": 0.00016286514015164557,
+      "loss": 1.6782,
+      "step": 13460
+    },
+    {
+      "epoch": 2.837731079159425,
+      "grad_norm": 0.9350339770317078,
+      "learning_rate": 0.0001628136477381354,
+      "loss": 1.6597,
+      "step": 13470
+    },
+    {
+      "epoch": 2.8398377837467743,
+      "grad_norm": 0.8482766151428223,
+      "learning_rate": 0.00016276212780146637,
+      "loss": 1.7017,
+      "step": 13480
+    },
+    {
+      "epoch": 2.841944488334123,
+      "grad_norm": 0.9531964659690857,
+      "learning_rate": 0.00016271058036421314,
+      "loss": 1.7145,
+      "step": 13490
+    },
+    {
+      "epoch": 2.8440511929214725,
+      "grad_norm": 0.8633520007133484,
+      "learning_rate": 0.00016265900544896225,
+      "loss": 1.6757,
+      "step": 13500
+    },
+    {
+      "epoch": 2.8461578975088218,
+      "grad_norm": 0.8866189122200012,
+      "learning_rate": 0.00016260740307831237,
+      "loss": 1.6983,
+      "step": 13510
+    },
+    {
+      "epoch": 2.848264602096171,
+      "grad_norm": 0.8696046471595764,
+      "learning_rate": 0.00016255577327487425,
+      "loss": 1.6678,
+      "step": 13520
+    },
+    {
+      "epoch": 2.8503713066835203,
+      "grad_norm": 0.9444338083267212,
+      "learning_rate": 0.00016250411606127054,
+      "loss": 1.7161,
+      "step": 13530
+    },
+    {
+      "epoch": 2.8524780112708696,
+      "grad_norm": 0.8829008936882019,
+      "learning_rate": 0.00016245243146013603,
+      "loss": 1.7357,
+      "step": 13540
+    },
+    {
+      "epoch": 2.854584715858219,
+      "grad_norm": 0.9542901515960693,
+      "learning_rate": 0.00016240071949411738,
+      "loss": 1.7435,
+      "step": 13550
+    },
+    {
+      "epoch": 2.856691420445568,
+      "grad_norm": 0.9169602990150452,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 1.6806,
+      "step": 13560
+    },
+    {
+      "epoch": 2.8587981250329175,
+      "grad_norm": 0.9551498889923096,
+      "learning_rate": 0.00016229721355807465,
+      "loss": 1.6869,
+      "step": 13570
+    },
+    {
+      "epoch": 2.8609048296202664,
+      "grad_norm": 0.8909421563148499,
+      "learning_rate": 0.00016224541963340391,
+      "loss": 1.7237,
+      "step": 13580
+    },
+    {
+      "epoch": 2.8630115342076157,
+      "grad_norm": 1.0937942266464233,
+      "learning_rate": 0.00016219359843455577,
+      "loss": 1.6916,
+      "step": 13590
+    },
+    {
+      "epoch": 2.865118238794965,
+      "grad_norm": 0.9448012709617615,
+      "learning_rate": 0.0001621417499842368,
+      "loss": 1.6715,
+      "step": 13600
+    },
+    {
+      "epoch": 2.8672249433823143,
+      "grad_norm": 0.8611142039299011,
+      "learning_rate": 0.0001620898743051656,
+      "loss": 1.7111,
+      "step": 13610
+    },
+    {
+      "epoch": 2.8693316479696636,
+      "grad_norm": 0.8860224485397339,
+      "learning_rate": 0.0001620379714200725,
+      "loss": 1.7714,
+      "step": 13620
+    },
+    {
+      "epoch": 2.871438352557013,
+      "grad_norm": 0.9378777742385864,
+      "learning_rate": 0.0001619860413516999,
+      "loss": 1.6969,
+      "step": 13630
+    },
+    {
+      "epoch": 2.873545057144362,
+      "grad_norm": 0.8672656416893005,
+      "learning_rate": 0.00016193408412280217,
+      "loss": 1.6637,
+      "step": 13640
+    },
+    {
+      "epoch": 2.875651761731711,
+      "grad_norm": 0.9400450587272644,
+      "learning_rate": 0.00016188209975614542,
+      "loss": 1.6858,
+      "step": 13650
+    },
+    {
+      "epoch": 2.8777584663190603,
+      "grad_norm": 0.9799240231513977,
+      "learning_rate": 0.00016183008827450773,
+      "loss": 1.7139,
+      "step": 13660
+    },
+    {
+      "epoch": 2.8798651709064096,
+      "grad_norm": 0.8705218434333801,
+      "learning_rate": 0.00016177804970067907,
+      "loss": 1.72,
+      "step": 13670
+    },
+    {
+      "epoch": 2.881971875493759,
+      "grad_norm": 0.9402818083763123,
+      "learning_rate": 0.00016172598405746124,
+      "loss": 1.6982,
+      "step": 13680
+    },
+    {
+      "epoch": 2.884078580081108,
+      "grad_norm": 0.9194636344909668,
+      "learning_rate": 0.000161673891367668,
+      "loss": 1.7205,
+      "step": 13690
+    },
+    {
+      "epoch": 2.8861852846684575,
+      "grad_norm": 0.8392009139060974,
+      "learning_rate": 0.0001616217716541248,
+      "loss": 1.7389,
+      "step": 13700
+    },
+    {
+      "epoch": 2.8882919892558068,
+      "grad_norm": 0.9039947986602783,
+      "learning_rate": 0.00016156962493966908,
+      "loss": 1.7311,
+      "step": 13710
+    },
+    {
+      "epoch": 2.8903986938431556,
+      "grad_norm": 0.9231280088424683,
+      "learning_rate": 0.00016151745124715002,
+      "loss": 1.7134,
+      "step": 13720
+    },
+    {
+      "epoch": 2.8925053984305054,
+      "grad_norm": 0.9273865818977356,
+      "learning_rate": 0.00016146525059942865,
+      "loss": 1.6491,
+      "step": 13730
+    },
+    {
+      "epoch": 2.894612103017854,
+      "grad_norm": 0.8546093106269836,
+      "learning_rate": 0.00016141302301937786,
+      "loss": 1.6637,
+      "step": 13740
+    },
+    {
+      "epoch": 2.8967188076052035,
+      "grad_norm": 0.9186126589775085,
+      "learning_rate": 0.00016136076852988224,
+      "loss": 1.694,
+      "step": 13750
+    },
+    {
+      "epoch": 2.898825512192553,
+      "grad_norm": 0.919446587562561,
+      "learning_rate": 0.00016130848715383827,
+      "loss": 1.7034,
+      "step": 13760
+    },
+    {
+      "epoch": 2.900932216779902,
+      "grad_norm": 0.8708723187446594,
+      "learning_rate": 0.0001612561789141541,
+      "loss": 1.7133,
+      "step": 13770
+    },
+    {
+      "epoch": 2.9030389213672514,
+      "grad_norm": 0.8736213445663452,
+      "learning_rate": 0.0001612038438337498,
+      "loss": 1.7206,
+      "step": 13780
+    },
+    {
+      "epoch": 2.9051456259546002,
+      "grad_norm": 0.993733823299408,
+      "learning_rate": 0.00016115148193555706,
+      "loss": 1.7022,
+      "step": 13790
+    },
+    {
+      "epoch": 2.90725233054195,
+      "grad_norm": 0.8737487196922302,
+      "learning_rate": 0.0001610990932425194,
+      "loss": 1.7069,
+      "step": 13800
+    },
+    {
+      "epoch": 2.909359035129299,
+      "grad_norm": 0.91124427318573,
+      "learning_rate": 0.00016104667777759206,
+      "loss": 1.7307,
+      "step": 13810
+    },
+    {
+      "epoch": 2.911465739716648,
+      "grad_norm": 0.9024822115898132,
+      "learning_rate": 0.00016099423556374199,
+      "loss": 1.6838,
+      "step": 13820
+    },
+    {
+      "epoch": 2.9135724443039974,
+      "grad_norm": 0.9395371079444885,
+      "learning_rate": 0.00016094176662394792,
+      "loss": 1.7139,
+      "step": 13830
+    },
+    {
+      "epoch": 2.9156791488913467,
+      "grad_norm": 0.8873124122619629,
+      "learning_rate": 0.0001608892709812002,
+      "loss": 1.6932,
+      "step": 13840
+    },
+    {
+      "epoch": 2.917785853478696,
+      "grad_norm": 0.9905756115913391,
+      "learning_rate": 0.00016083674865850097,
+      "loss": 1.7101,
+      "step": 13850
+    },
+    {
+      "epoch": 2.9198925580660453,
+      "grad_norm": 0.8979195356369019,
+      "learning_rate": 0.00016078419967886402,
+      "loss": 1.6858,
+      "step": 13860
+    },
+    {
+      "epoch": 2.9219992626533946,
+      "grad_norm": 0.8918625712394714,
+      "learning_rate": 0.0001607316240653148,
+      "loss": 1.6297,
+      "step": 13870
+    },
+    {
+      "epoch": 2.9241059672407435,
+      "grad_norm": 0.9859835505485535,
+      "learning_rate": 0.00016067902184089048,
+      "loss": 1.6728,
+      "step": 13880
+    },
+    {
+      "epoch": 2.9262126718280927,
+      "grad_norm": 0.9373555779457092,
+      "learning_rate": 0.00016062639302863986,
+      "loss": 1.7194,
+      "step": 13890
+    },
+    {
+      "epoch": 2.928319376415442,
+      "grad_norm": 0.9152488112449646,
+      "learning_rate": 0.00016057373765162333,
+      "loss": 1.668,
+      "step": 13900
+    },
+    {
+      "epoch": 2.9304260810027913,
+      "grad_norm": 0.9642083644866943,
+      "learning_rate": 0.0001605210557329131,
+      "loss": 1.7473,
+      "step": 13910
+    },
+    {
+      "epoch": 2.9325327855901406,
+      "grad_norm": 0.9451881647109985,
+      "learning_rate": 0.0001604683472955928,
+      "loss": 1.6884,
+      "step": 13920
+    },
+    {
+      "epoch": 2.93463949017749,
+      "grad_norm": 0.9817609786987305,
+      "learning_rate": 0.00016041561236275777,
+      "loss": 1.7305,
+      "step": 13930
+    },
+    {
+      "epoch": 2.9367461947648392,
+      "grad_norm": 0.923049807548523,
+      "learning_rate": 0.00016036285095751503,
+      "loss": 1.7269,
+      "step": 13940
+    },
+    {
+      "epoch": 2.938852899352188,
+      "grad_norm": 0.9717063903808594,
+      "learning_rate": 0.00016031006310298306,
+      "loss": 1.7341,
+      "step": 13950
+    },
+    {
+      "epoch": 2.940959603939538,
+      "grad_norm": 0.901326060295105,
+      "learning_rate": 0.00016025724882229208,
+      "loss": 1.6908,
+      "step": 13960
+    },
+    {
+      "epoch": 2.9430663085268867,
+      "grad_norm": 0.8968400359153748,
+      "learning_rate": 0.0001602044081385837,
+      "loss": 1.6873,
+      "step": 13970
+    },
+    {
+      "epoch": 2.945173013114236,
+      "grad_norm": 0.8596915006637573,
+      "learning_rate": 0.00016015154107501133,
+      "loss": 1.7327,
+      "step": 13980
+    },
+    {
+      "epoch": 2.9472797177015853,
+      "grad_norm": 0.9334450960159302,
+      "learning_rate": 0.00016009864765473972,
+      "loss": 1.6833,
+      "step": 13990
+    },
+    {
+      "epoch": 2.9493864222889346,
+      "grad_norm": 0.8397333025932312,
+      "learning_rate": 0.00016004572790094535,
+      "loss": 1.7331,
+      "step": 14000
+    },
+    {
+      "epoch": 2.951493126876284,
+      "grad_norm": 0.8748630285263062,
+      "learning_rate": 0.00015999278183681604,
+      "loss": 1.7298,
+      "step": 14010
+    },
+    {
+      "epoch": 2.953599831463633,
+      "grad_norm": 0.9097732901573181,
+      "learning_rate": 0.0001599398094855514,
+      "loss": 1.7674,
+      "step": 14020
+    },
+    {
+      "epoch": 2.9557065360509824,
+      "grad_norm": 0.8727101683616638,
+      "learning_rate": 0.0001598868108703623,
+      "loss": 1.7557,
+      "step": 14030
+    },
+    {
+      "epoch": 2.9578132406383313,
+      "grad_norm": 0.8882416486740112,
+      "learning_rate": 0.00015983378601447127,
+      "loss": 1.6351,
+      "step": 14040
+    },
+    {
+      "epoch": 2.9599199452256806,
+      "grad_norm": 0.9257373213768005,
+      "learning_rate": 0.00015978073494111227,
+      "loss": 1.6756,
+      "step": 14050
+    },
+    {
+      "epoch": 2.96202664981303,
+      "grad_norm": 0.8525111675262451,
+      "learning_rate": 0.00015972765767353087,
+      "loss": 1.7364,
+      "step": 14060
+    },
+    {
+      "epoch": 2.964133354400379,
+      "grad_norm": 0.9729892611503601,
+      "learning_rate": 0.00015967455423498387,
+      "loss": 1.7332,
+      "step": 14070
+    },
+    {
+      "epoch": 2.9662400589877285,
+      "grad_norm": 0.9302247762680054,
+      "learning_rate": 0.00015962142464873985,
+      "loss": 1.7507,
+      "step": 14080
+    },
+    {
+      "epoch": 2.9683467635750778,
+      "grad_norm": 0.8983259201049805,
+      "learning_rate": 0.00015956826893807855,
+      "loss": 1.7273,
+      "step": 14090
+    },
+    {
+      "epoch": 2.970453468162427,
+      "grad_norm": 0.8971543908119202,
+      "learning_rate": 0.0001595150871262914,
+      "loss": 1.7365,
+      "step": 14100
+    },
+    {
+      "epoch": 2.972560172749776,
+      "grad_norm": 0.8822101354598999,
+      "learning_rate": 0.00015946187923668108,
+      "loss": 1.6777,
+      "step": 14110
+    },
+    {
+      "epoch": 2.9746668773371256,
+      "grad_norm": 0.9178860783576965,
+      "learning_rate": 0.00015940864529256186,
+      "loss": 1.6907,
+      "step": 14120
+    },
+    {
+      "epoch": 2.9767735819244745,
+      "grad_norm": 0.9445515275001526,
+      "learning_rate": 0.00015935538531725927,
+      "loss": 1.693,
+      "step": 14130
+    },
+    {
+      "epoch": 2.978880286511824,
+      "grad_norm": 0.8799481987953186,
+      "learning_rate": 0.00015930209933411036,
+      "loss": 1.6918,
+      "step": 14140
+    },
+    {
+      "epoch": 2.980986991099173,
+      "grad_norm": 0.8957604765892029,
+      "learning_rate": 0.00015924878736646352,
+      "loss": 1.7036,
+      "step": 14150
+    },
+    {
+      "epoch": 2.9830936956865224,
+      "grad_norm": 0.8179033398628235,
+      "learning_rate": 0.00015919544943767856,
+      "loss": 1.7331,
+      "step": 14160
+    },
+    {
+      "epoch": 2.9852004002738717,
+      "grad_norm": 1.4999958276748657,
+      "learning_rate": 0.00015914208557112663,
+      "loss": 1.6402,
+      "step": 14170
+    },
+    {
+      "epoch": 2.987307104861221,
+      "grad_norm": 0.9046642780303955,
+      "learning_rate": 0.00015908869579019027,
+      "loss": 1.7611,
+      "step": 14180
+    },
+    {
+      "epoch": 2.9894138094485703,
+      "grad_norm": 0.9739823341369629,
+      "learning_rate": 0.00015903528011826335,
+      "loss": 1.7253,
+      "step": 14190
+    },
+    {
+      "epoch": 2.991520514035919,
+      "grad_norm": 0.9123568534851074,
+      "learning_rate": 0.00015898183857875116,
+      "loss": 1.721,
+      "step": 14200
+    },
+    {
+      "epoch": 2.9936272186232684,
+      "grad_norm": 0.8985586166381836,
+      "learning_rate": 0.00015892837119507014,
+      "loss": 1.6525,
+      "step": 14210
+    },
+    {
+      "epoch": 2.9957339232106177,
+      "grad_norm": 0.9110245108604431,
+      "learning_rate": 0.00015887487799064838,
+      "loss": 1.6592,
+      "step": 14220
+    },
+    {
+      "epoch": 2.997840627797967,
+      "grad_norm": 0.9014399647712708,
+      "learning_rate": 0.00015882135898892492,
+      "loss": 1.6392,
+      "step": 14230
+    },
+    {
+      "epoch": 2.9999473323853163,
+      "grad_norm": 0.9020178914070129,
+      "learning_rate": 0.0001587678142133503,
+      "loss": 1.7085,
+      "step": 14240
+    },
+    {
+      "epoch": 3.0020540369726656,
+      "grad_norm": 1.0644667148590088,
+      "learning_rate": 0.0001587142436873864,
+      "loss": 1.6745,
+      "step": 14250
+    },
+    {
+      "epoch": 3.004160741560015,
+      "grad_norm": 0.9659833312034607,
+      "learning_rate": 0.00015866064743450618,
+      "loss": 1.6343,
+      "step": 14260
+    },
+    {
+      "epoch": 3.006267446147364,
+      "grad_norm": 0.8868164420127869,
+      "learning_rate": 0.00015860702547819413,
+      "loss": 1.6598,
+      "step": 14270
+    },
+    {
+      "epoch": 3.008374150734713,
+      "grad_norm": 0.9452905058860779,
+      "learning_rate": 0.00015855337784194577,
+      "loss": 1.7491,
+      "step": 14280
+    },
+    {
+      "epoch": 3.0104808553220623,
+      "grad_norm": 0.941245973110199,
+      "learning_rate": 0.00015849970454926804,
+      "loss": 1.664,
+      "step": 14290
+    },
+    {
+      "epoch": 3.0125875599094116,
+      "grad_norm": 0.9174633622169495,
+      "learning_rate": 0.000158446005623679,
+      "loss": 1.6228,
+      "step": 14300
+    },
+    {
+      "epoch": 3.014694264496761,
+      "grad_norm": 0.9531957507133484,
+      "learning_rate": 0.000158392281088708,
+      "loss": 1.6542,
+      "step": 14310
+    },
+    {
+      "epoch": 3.01680096908411,
+      "grad_norm": 0.9811040163040161,
+      "learning_rate": 0.00015833853096789566,
+      "loss": 1.6483,
+      "step": 14320
+    },
+    {
+      "epoch": 3.0189076736714595,
+      "grad_norm": 0.8868007659912109,
+      "learning_rate": 0.00015828475528479368,
+      "loss": 1.6262,
+      "step": 14330
+    },
+    {
+      "epoch": 3.021014378258809,
+      "grad_norm": 0.9833335280418396,
+      "learning_rate": 0.00015823095406296514,
+      "loss": 1.6906,
+      "step": 14340
+    },
+    {
+      "epoch": 3.023121082846158,
+      "grad_norm": 0.9299181699752808,
+      "learning_rate": 0.00015817712732598413,
+      "loss": 1.6371,
+      "step": 14350
+    },
+    {
+      "epoch": 3.025227787433507,
+      "grad_norm": 1.0234558582305908,
+      "learning_rate": 0.00015812327509743602,
+      "loss": 1.698,
+      "step": 14360
+    },
+    {
+      "epoch": 3.0273344920208562,
+      "grad_norm": 0.9303063750267029,
+      "learning_rate": 0.00015806939740091734,
+      "loss": 1.6533,
+      "step": 14370
+    },
+    {
+      "epoch": 3.0294411966082055,
+      "grad_norm": 0.8761981725692749,
+      "learning_rate": 0.00015801549426003577,
+      "loss": 1.6728,
+      "step": 14380
+    },
+    {
+      "epoch": 3.031547901195555,
+      "grad_norm": 0.8972175717353821,
+      "learning_rate": 0.00015796156569841013,
+      "loss": 1.6416,
+      "step": 14390
+    },
+    {
+      "epoch": 3.033654605782904,
+      "grad_norm": 1.033262848854065,
+      "learning_rate": 0.00015790761173967036,
+      "loss": 1.6115,
+      "step": 14400
+    },
+    {
+      "epoch": 3.0357613103702534,
+      "grad_norm": 0.9915304780006409,
+      "learning_rate": 0.00015785363240745757,
+      "loss": 1.685,
+      "step": 14410
+    },
+    {
+      "epoch": 3.0378680149576027,
+      "grad_norm": 0.9661949276924133,
+      "learning_rate": 0.00015779962772542402,
+      "loss": 1.6853,
+      "step": 14420
+    },
+    {
+      "epoch": 3.039974719544952,
+      "grad_norm": 1.1520849466323853,
+      "learning_rate": 0.00015774559771723298,
+      "loss": 1.6574,
+      "step": 14430
+    },
+    {
+      "epoch": 3.042081424132301,
+      "grad_norm": 0.8876193165779114,
+      "learning_rate": 0.00015769154240655885,
+      "loss": 1.6226,
+      "step": 14440
+    },
+    {
+      "epoch": 3.04418812871965,
+      "grad_norm": 1.0137007236480713,
+      "learning_rate": 0.0001576374618170872,
+      "loss": 1.6743,
+      "step": 14450
+    },
+    {
+      "epoch": 3.0462948333069995,
+      "grad_norm": 0.8431521058082581,
+      "learning_rate": 0.00015758335597251458,
+      "loss": 1.6573,
+      "step": 14460
+    },
+    {
+      "epoch": 3.0484015378943488,
+      "grad_norm": 1.0139840841293335,
+      "learning_rate": 0.00015752922489654857,
+      "loss": 1.6128,
+      "step": 14470
+    },
+    {
+      "epoch": 3.050508242481698,
+      "grad_norm": 1.04917311668396,
+      "learning_rate": 0.00015747506861290796,
+      "loss": 1.6697,
+      "step": 14480
+    },
+    {
+      "epoch": 3.0526149470690473,
+      "grad_norm": 0.9627788066864014,
+      "learning_rate": 0.00015742088714532247,
+      "loss": 1.6761,
+      "step": 14490
+    },
+    {
+      "epoch": 3.0547216516563966,
+      "grad_norm": 0.926226794719696,
+      "learning_rate": 0.0001573666805175329,
+      "loss": 1.6598,
+      "step": 14500
+    },
+    {
+      "epoch": 3.056828356243746,
+      "grad_norm": 0.99714595079422,
+      "learning_rate": 0.00015731244875329107,
+      "loss": 1.6662,
+      "step": 14510
+    },
+    {
+      "epoch": 3.058935060831095,
+      "grad_norm": 0.960364580154419,
+      "learning_rate": 0.00015725819187635968,
+      "loss": 1.6963,
+      "step": 14520
+    },
+    {
+      "epoch": 3.061041765418444,
+      "grad_norm": 0.8881476521492004,
+      "learning_rate": 0.00015720390991051268,
+      "loss": 1.6545,
+      "step": 14530
+    },
+    {
+      "epoch": 3.0631484700057934,
+      "grad_norm": 0.9406470656394958,
+      "learning_rate": 0.00015714960287953485,
+      "loss": 1.584,
+      "step": 14540
+    },
+    {
+      "epoch": 3.0652551745931427,
+      "grad_norm": 0.9198954105377197,
+      "learning_rate": 0.00015709527080722202,
+      "loss": 1.628,
+      "step": 14550
+    },
+    {
+      "epoch": 3.067361879180492,
+      "grad_norm": 0.9906532764434814,
+      "learning_rate": 0.0001570409137173809,
+      "loss": 1.6715,
+      "step": 14560
+    },
+    {
+      "epoch": 3.0694685837678413,
+      "grad_norm": 0.9512636661529541,
+      "learning_rate": 0.0001569865316338293,
+      "loss": 1.6115,
+      "step": 14570
+    },
+    {
+      "epoch": 3.0715752883551906,
+      "grad_norm": 0.9917583465576172,
+      "learning_rate": 0.00015693212458039584,
+      "loss": 1.7425,
+      "step": 14580
+    },
+    {
+      "epoch": 3.0736819929425394,
+      "grad_norm": 0.9491133093833923,
+      "learning_rate": 0.00015687769258092017,
+      "loss": 1.6637,
+      "step": 14590
+    },
+    {
+      "epoch": 3.0757886975298887,
+      "grad_norm": 0.9375881552696228,
+      "learning_rate": 0.00015682323565925285,
+      "loss": 1.6363,
+      "step": 14600
+    },
+    {
+      "epoch": 3.077895402117238,
+      "grad_norm": 1.0055121183395386,
+      "learning_rate": 0.00015676875383925534,
+      "loss": 1.6571,
+      "step": 14610
+    },
+    {
+      "epoch": 3.0800021067045873,
+      "grad_norm": 1.0039416551589966,
+      "learning_rate": 0.0001567142471448001,
+      "loss": 1.6895,
+      "step": 14620
+    },
+    {
+      "epoch": 3.0821088112919366,
+      "grad_norm": 0.9418062567710876,
+      "learning_rate": 0.00015665971559977035,
+      "loss": 1.6936,
+      "step": 14630
+    },
+    {
+      "epoch": 3.084215515879286,
+      "grad_norm": 0.9344249367713928,
+      "learning_rate": 0.00015660515922806027,
+      "loss": 1.6527,
+      "step": 14640
+    },
+    {
+      "epoch": 3.086322220466635,
+      "grad_norm": 0.9298831224441528,
+      "learning_rate": 0.00015655057805357493,
+      "loss": 1.6325,
+      "step": 14650
+    },
+    {
+      "epoch": 3.0884289250539845,
+      "grad_norm": 0.8685382604598999,
+      "learning_rate": 0.00015649597210023027,
+      "loss": 1.7013,
+      "step": 14660
+    },
+    {
+      "epoch": 3.0905356296413338,
+      "grad_norm": 0.920759916305542,
+      "learning_rate": 0.000156441341391953,
+      "loss": 1.6177,
+      "step": 14670
+    },
+    {
+      "epoch": 3.0926423342286826,
+      "grad_norm": 0.9808409810066223,
+      "learning_rate": 0.00015638668595268084,
+      "loss": 1.6857,
+      "step": 14680
+    },
+    {
+      "epoch": 3.094749038816032,
+      "grad_norm": 1.0194792747497559,
+      "learning_rate": 0.0001563320058063622,
+      "loss": 1.6346,
+      "step": 14690
+    },
+    {
+      "epoch": 3.096855743403381,
+      "grad_norm": 0.9400737881660461,
+      "learning_rate": 0.00015627730097695638,
+      "loss": 1.6783,
+      "step": 14700
+    },
+    {
+      "epoch": 3.0989624479907305,
+      "grad_norm": 0.9842483997344971,
+      "learning_rate": 0.00015622257148843348,
+      "loss": 1.6614,
+      "step": 14710
+    },
+    {
+      "epoch": 3.10106915257808,
+      "grad_norm": 1.000064492225647,
+      "learning_rate": 0.00015616781736477444,
+      "loss": 1.66,
+      "step": 14720
+    },
+    {
+      "epoch": 3.103175857165429,
+      "grad_norm": 0.9329800605773926,
+      "learning_rate": 0.0001561130386299709,
+      "loss": 1.5831,
+      "step": 14730
+    },
+    {
+      "epoch": 3.1052825617527784,
+      "grad_norm": 1.0243964195251465,
+      "learning_rate": 0.00015605823530802541,
+      "loss": 1.6751,
+      "step": 14740
+    },
+    {
+      "epoch": 3.1073892663401272,
+      "grad_norm": 0.9882740378379822,
+      "learning_rate": 0.0001560034074229512,
+      "loss": 1.6543,
+      "step": 14750
+    },
+    {
+      "epoch": 3.1094959709274765,
+      "grad_norm": 0.9630525708198547,
+      "learning_rate": 0.0001559485549987723,
+      "loss": 1.6822,
+      "step": 14760
+    },
+    {
+      "epoch": 3.111602675514826,
+      "grad_norm": 0.9518427848815918,
+      "learning_rate": 0.00015589367805952348,
+      "loss": 1.6336,
+      "step": 14770
+    },
+    {
+      "epoch": 3.113709380102175,
+      "grad_norm": 0.9057418704032898,
+      "learning_rate": 0.0001558387766292503,
+      "loss": 1.6206,
+      "step": 14780
+    },
+    {
+      "epoch": 3.1158160846895244,
+      "grad_norm": 0.9594529271125793,
+      "learning_rate": 0.00015578385073200895,
+      "loss": 1.6772,
+      "step": 14790
+    },
+    {
+      "epoch": 3.1179227892768737,
+      "grad_norm": 0.8876820802688599,
+      "learning_rate": 0.0001557289003918664,
+      "loss": 1.5913,
+      "step": 14800
+    },
+    {
+      "epoch": 3.120029493864223,
+      "grad_norm": 0.977289617061615,
+      "learning_rate": 0.00015567392563290038,
+      "loss": 1.6617,
+      "step": 14810
+    },
+    {
+      "epoch": 3.1221361984515723,
+      "grad_norm": 0.8738213777542114,
+      "learning_rate": 0.0001556189264791992,
+      "loss": 1.7153,
+      "step": 14820
+    },
+    {
+      "epoch": 3.124242903038921,
+      "grad_norm": 1.0732994079589844,
+      "learning_rate": 0.0001555639029548621,
+      "loss": 1.664,
+      "step": 14830
+    },
+    {
+      "epoch": 3.1263496076262705,
+      "grad_norm": 0.9305723309516907,
+      "learning_rate": 0.00015550885508399856,
+      "loss": 1.6656,
+      "step": 14840
+    },
+    {
+      "epoch": 3.1284563122136197,
+      "grad_norm": 1.1515586376190186,
+      "learning_rate": 0.00015545378289072922,
+      "loss": 1.71,
+      "step": 14850
+    },
+    {
+      "epoch": 3.130563016800969,
+      "grad_norm": 0.9483801126480103,
+      "learning_rate": 0.00015539868639918505,
+      "loss": 1.6762,
+      "step": 14860
+    },
+    {
+      "epoch": 3.1326697213883183,
+      "grad_norm": 1.0394940376281738,
+      "learning_rate": 0.00015534356563350784,
+      "loss": 1.6828,
+      "step": 14870
+    },
+    {
+      "epoch": 3.1347764259756676,
+      "grad_norm": 0.9815645217895508,
+      "learning_rate": 0.0001552884206178498,
+      "loss": 1.6447,
+      "step": 14880
+    },
+    {
+      "epoch": 3.136883130563017,
+      "grad_norm": 0.9685027599334717,
+      "learning_rate": 0.00015523325137637407,
+      "loss": 1.6087,
+      "step": 14890
+    },
+    {
+      "epoch": 3.1389898351503662,
+      "grad_norm": 0.9185184240341187,
+      "learning_rate": 0.0001551780579332542,
+      "loss": 1.5929,
+      "step": 14900
+    },
+    {
+      "epoch": 3.141096539737715,
+      "grad_norm": 0.9931472539901733,
+      "learning_rate": 0.00015512284031267437,
+      "loss": 1.6585,
+      "step": 14910
+    },
+    {
+      "epoch": 3.1432032443250644,
+      "grad_norm": 0.9517951607704163,
+      "learning_rate": 0.00015506759853882934,
+      "loss": 1.6431,
+      "step": 14920
+    },
+    {
+      "epoch": 3.1453099489124137,
+      "grad_norm": 0.8860810995101929,
+      "learning_rate": 0.00015501233263592455,
+      "loss": 1.697,
+      "step": 14930
+    },
+    {
+      "epoch": 3.147416653499763,
+      "grad_norm": 0.961342990398407,
+      "learning_rate": 0.00015495704262817597,
+      "loss": 1.6574,
+      "step": 14940
+    },
+    {
+      "epoch": 3.1495233580871123,
+      "grad_norm": 0.9692322611808777,
+      "learning_rate": 0.00015490172853981004,
+      "loss": 1.6623,
+      "step": 14950
+    },
+    {
+      "epoch": 3.1516300626744616,
+      "grad_norm": 0.9061542749404907,
+      "learning_rate": 0.00015484639039506387,
+      "loss": 1.6955,
+      "step": 14960
+    },
+    {
+      "epoch": 3.153736767261811,
+      "grad_norm": 0.8963515162467957,
+      "learning_rate": 0.00015479102821818507,
+      "loss": 1.6676,
+      "step": 14970
+    },
+    {
+      "epoch": 3.15584347184916,
+      "grad_norm": 0.9289860725402832,
+      "learning_rate": 0.00015473564203343174,
+      "loss": 1.6048,
+      "step": 14980
+    },
+    {
+      "epoch": 3.157950176436509,
+      "grad_norm": 1.0143063068389893,
+      "learning_rate": 0.00015468023186507256,
+      "loss": 1.717,
+      "step": 14990
+    },
+    {
+      "epoch": 3.1600568810238583,
+      "grad_norm": 1.0174888372421265,
+      "learning_rate": 0.0001546247977373867,
+      "loss": 1.6563,
+      "step": 15000
+    },
+    {
+      "epoch": 3.1621635856112076,
+      "grad_norm": 1.0211094617843628,
+      "learning_rate": 0.0001545693396746638,
+      "loss": 1.6338,
+      "step": 15010
+    },
+    {
+      "epoch": 3.164270290198557,
+      "grad_norm": 0.966232419013977,
+      "learning_rate": 0.000154513857701204,
+      "loss": 1.6749,
+      "step": 15020
+    },
+    {
+      "epoch": 3.166376994785906,
+      "grad_norm": 0.9910480976104736,
+      "learning_rate": 0.000154458351841318,
+      "loss": 1.6248,
+      "step": 15030
+    },
+    {
+      "epoch": 3.1684836993732555,
+      "grad_norm": 1.0164378881454468,
+      "learning_rate": 0.00015440282211932682,
+      "loss": 1.6117,
+      "step": 15040
+    },
+    {
+      "epoch": 3.1705904039606048,
+      "grad_norm": 0.9360641241073608,
+      "learning_rate": 0.00015434726855956206,
+      "loss": 1.6809,
+      "step": 15050
+    },
+    {
+      "epoch": 3.172697108547954,
+      "grad_norm": 1.1147947311401367,
+      "learning_rate": 0.00015429169118636566,
+      "loss": 1.6561,
+      "step": 15060
+    },
+    {
+      "epoch": 3.174803813135303,
+      "grad_norm": 0.9635496735572815,
+      "learning_rate": 0.00015423609002409008,
+      "loss": 1.6598,
+      "step": 15070
+    },
+    {
+      "epoch": 3.176910517722652,
+      "grad_norm": 0.9060547351837158,
+      "learning_rate": 0.00015418046509709817,
+      "loss": 1.6666,
+      "step": 15080
+    },
+    {
+      "epoch": 3.1790172223100015,
+      "grad_norm": 1.014722228050232,
+      "learning_rate": 0.00015412481642976318,
+      "loss": 1.7179,
+      "step": 15090
+    },
+    {
+      "epoch": 3.181123926897351,
+      "grad_norm": 1.006274700164795,
+      "learning_rate": 0.00015406914404646882,
+      "loss": 1.7212,
+      "step": 15100
+    },
+    {
+      "epoch": 3.1832306314847,
+      "grad_norm": 0.9190952777862549,
+      "learning_rate": 0.00015401344797160907,
+      "loss": 1.6335,
+      "step": 15110
+    },
+    {
+      "epoch": 3.1853373360720494,
+      "grad_norm": 0.9608885049819946,
+      "learning_rate": 0.00015395772822958845,
+      "loss": 1.6894,
+      "step": 15120
+    },
+    {
+      "epoch": 3.1874440406593987,
+      "grad_norm": 1.0324962139129639,
+      "learning_rate": 0.00015390198484482176,
+      "loss": 1.656,
+      "step": 15130
+    },
+    {
+      "epoch": 3.189550745246748,
+      "grad_norm": 0.9675622582435608,
+      "learning_rate": 0.00015384621784173414,
+      "loss": 1.7216,
+      "step": 15140
+    },
+    {
+      "epoch": 3.191657449834097,
+      "grad_norm": 0.9279264211654663,
+      "learning_rate": 0.0001537904272447611,
+      "loss": 1.6508,
+      "step": 15150
+    },
+    {
+      "epoch": 3.193764154421446,
+      "grad_norm": 0.9079473614692688,
+      "learning_rate": 0.00015373461307834852,
+      "loss": 1.6871,
+      "step": 15160
+    },
+    {
+      "epoch": 3.1958708590087954,
+      "grad_norm": 1.0070098638534546,
+      "learning_rate": 0.0001536787753669526,
+      "loss": 1.7162,
+      "step": 15170
+    },
+    {
+      "epoch": 3.1979775635961447,
+      "grad_norm": 0.9919424057006836,
+      "learning_rate": 0.00015362291413503984,
+      "loss": 1.6805,
+      "step": 15180
+    },
+    {
+      "epoch": 3.200084268183494,
+      "grad_norm": 1.0416815280914307,
+      "learning_rate": 0.000153567029407087,
+      "loss": 1.7119,
+      "step": 15190
+    },
+    {
+      "epoch": 3.2021909727708433,
+      "grad_norm": 1.0771572589874268,
+      "learning_rate": 0.00015351112120758122,
+      "loss": 1.5851,
+      "step": 15200
+    },
+    {
+      "epoch": 3.2042976773581926,
+      "grad_norm": 1.0703963041305542,
+      "learning_rate": 0.0001534551895610199,
+      "loss": 1.6739,
+      "step": 15210
+    },
+    {
+      "epoch": 3.206404381945542,
+      "grad_norm": 0.9871819019317627,
+      "learning_rate": 0.00015339923449191067,
+      "loss": 1.6918,
+      "step": 15220
+    },
+    {
+      "epoch": 3.2085110865328907,
+      "grad_norm": 1.037592887878418,
+      "learning_rate": 0.00015334325602477146,
+      "loss": 1.6564,
+      "step": 15230
+    },
+    {
+      "epoch": 3.21061779112024,
+      "grad_norm": 0.9865399599075317,
+      "learning_rate": 0.00015328725418413045,
+      "loss": 1.6338,
+      "step": 15240
+    },
+    {
+      "epoch": 3.2127244957075893,
+      "grad_norm": 0.9120628237724304,
+      "learning_rate": 0.00015323122899452607,
+      "loss": 1.6434,
+      "step": 15250
+    },
+    {
+      "epoch": 3.2148312002949386,
+      "grad_norm": 0.9064800143241882,
+      "learning_rate": 0.00015317518048050697,
+      "loss": 1.6631,
+      "step": 15260
+    },
+    {
+      "epoch": 3.216937904882288,
+      "grad_norm": 0.9162564873695374,
+      "learning_rate": 0.00015311910866663196,
+      "loss": 1.6954,
+      "step": 15270
+    },
+    {
+      "epoch": 3.219044609469637,
+      "grad_norm": 0.9941088557243347,
+      "learning_rate": 0.00015306301357747022,
+      "loss": 1.7128,
+      "step": 15280
+    },
+    {
+      "epoch": 3.2211513140569865,
+      "grad_norm": 0.984843909740448,
+      "learning_rate": 0.00015300689523760097,
+      "loss": 1.6372,
+      "step": 15290
+    },
+    {
+      "epoch": 3.2232580186443354,
+      "grad_norm": 1.007107138633728,
+      "learning_rate": 0.00015295075367161367,
+      "loss": 1.6644,
+      "step": 15300
+    },
+    {
+      "epoch": 3.2253647232316847,
+      "grad_norm": 1.0158659219741821,
+      "learning_rate": 0.00015289458890410798,
+      "loss": 1.6354,
+      "step": 15310
+    },
+    {
+      "epoch": 3.227471427819034,
+      "grad_norm": 0.9480316042900085,
+      "learning_rate": 0.00015283840095969367,
+      "loss": 1.6573,
+      "step": 15320
+    },
+    {
+      "epoch": 3.2295781324063833,
+      "grad_norm": 0.9743335247039795,
+      "learning_rate": 0.00015278218986299074,
+      "loss": 1.7287,
+      "step": 15330
+    },
+    {
+      "epoch": 3.2316848369937325,
+      "grad_norm": 0.9238772392272949,
+      "learning_rate": 0.00015272595563862933,
+      "loss": 1.662,
+      "step": 15340
+    },
+    {
+      "epoch": 3.233791541581082,
+      "grad_norm": 0.9363320469856262,
+      "learning_rate": 0.00015266969831124962,
+      "loss": 1.6218,
+      "step": 15350
+    },
+    {
+      "epoch": 3.235898246168431,
+      "grad_norm": 0.9849210381507874,
+      "learning_rate": 0.00015261341790550196,
+      "loss": 1.6596,
+      "step": 15360
+    },
+    {
+      "epoch": 3.2380049507557804,
+      "grad_norm": 0.9171252250671387,
+      "learning_rate": 0.00015255711444604693,
+      "loss": 1.6972,
+      "step": 15370
+    },
+    {
+      "epoch": 3.2401116553431297,
+      "grad_norm": 1.0775831937789917,
+      "learning_rate": 0.000152500787957555,
+      "loss": 1.6343,
+      "step": 15380
+    },
+    {
+      "epoch": 3.2422183599304786,
+      "grad_norm": 0.9412891864776611,
+      "learning_rate": 0.0001524444384647069,
+      "loss": 1.6965,
+      "step": 15390
+    },
+    {
+      "epoch": 3.244325064517828,
+      "grad_norm": 1.0591611862182617,
+      "learning_rate": 0.00015238806599219336,
+      "loss": 1.6702,
+      "step": 15400
+    },
+    {
+      "epoch": 3.246431769105177,
+      "grad_norm": 0.8861743211746216,
+      "learning_rate": 0.00015233167056471523,
+      "loss": 1.708,
+      "step": 15410
+    },
+    {
+      "epoch": 3.2485384736925265,
+      "grad_norm": 1.0437482595443726,
+      "learning_rate": 0.0001522752522069833,
+      "loss": 1.6861,
+      "step": 15420
+    },
+    {
+      "epoch": 3.2506451782798758,
+      "grad_norm": 0.9577218294143677,
+      "learning_rate": 0.00015221881094371861,
+      "loss": 1.6652,
+      "step": 15430
+    },
+    {
+      "epoch": 3.252751882867225,
+      "grad_norm": 1.0100152492523193,
+      "learning_rate": 0.00015216234679965205,
+      "loss": 1.7185,
+      "step": 15440
+    },
+    {
+      "epoch": 3.2548585874545743,
+      "grad_norm": 0.9381777048110962,
+      "learning_rate": 0.0001521058597995246,
+      "loss": 1.6786,
+      "step": 15450
+    },
+    {
+      "epoch": 3.256965292041923,
+      "grad_norm": 0.9847040772438049,
+      "learning_rate": 0.00015204934996808727,
+      "loss": 1.6359,
+      "step": 15460
+    },
+    {
+      "epoch": 3.2590719966292725,
+      "grad_norm": 1.0387279987335205,
+      "learning_rate": 0.00015199281733010116,
+      "loss": 1.639,
+      "step": 15470
+    },
+    {
+      "epoch": 3.261178701216622,
+      "grad_norm": 0.9832319617271423,
+      "learning_rate": 0.00015193626191033712,
+      "loss": 1.6587,
+      "step": 15480
+    },
+    {
+      "epoch": 3.263285405803971,
+      "grad_norm": 0.9799401760101318,
+      "learning_rate": 0.00015187968373357619,
+      "loss": 1.6814,
+      "step": 15490
+    },
+    {
+      "epoch": 3.2653921103913204,
+      "grad_norm": 1.0971506834030151,
+      "learning_rate": 0.00015182308282460935,
+      "loss": 1.6852,
+      "step": 15500
+    },
+    {
+      "epoch": 3.2674988149786697,
+      "grad_norm": 0.9791223406791687,
+      "learning_rate": 0.0001517664592082375,
+      "loss": 1.6598,
+      "step": 15510
+    },
+    {
+      "epoch": 3.269605519566019,
+      "grad_norm": 1.0151629447937012,
+      "learning_rate": 0.00015170981290927145,
+      "loss": 1.6781,
+      "step": 15520
+    },
+    {
+      "epoch": 3.2717122241533683,
+      "grad_norm": 0.9513714909553528,
+      "learning_rate": 0.00015165314395253212,
+      "loss": 1.6378,
+      "step": 15530
+    },
+    {
+      "epoch": 3.2738189287407176,
+      "grad_norm": 0.9968097805976868,
+      "learning_rate": 0.0001515964523628501,
+      "loss": 1.6473,
+      "step": 15540
+    },
+    {
+      "epoch": 3.2759256333280664,
+      "grad_norm": 1.050000548362732,
+      "learning_rate": 0.00015153973816506614,
+      "loss": 1.654,
+      "step": 15550
+    },
+    {
+      "epoch": 3.2780323379154157,
+      "grad_norm": 1.0254466533660889,
+      "learning_rate": 0.00015148300138403075,
+      "loss": 1.6328,
+      "step": 15560
+    },
+    {
+      "epoch": 3.280139042502765,
+      "grad_norm": 0.9959595203399658,
+      "learning_rate": 0.00015142624204460435,
+      "loss": 1.6786,
+      "step": 15570
+    },
+    {
+      "epoch": 3.2822457470901143,
+      "grad_norm": 1.058243989944458,
+      "learning_rate": 0.00015136946017165734,
+      "loss": 1.6118,
+      "step": 15580
+    },
+    {
+      "epoch": 3.2843524516774636,
+      "grad_norm": 1.0479722023010254,
+      "learning_rate": 0.0001513126557900699,
+      "loss": 1.6355,
+      "step": 15590
+    },
+    {
+      "epoch": 3.286459156264813,
+      "grad_norm": 0.9290971159934998,
+      "learning_rate": 0.00015125582892473204,
+      "loss": 1.6835,
+      "step": 15600
+    },
+    {
+      "epoch": 3.288565860852162,
+      "grad_norm": 0.9415523409843445,
+      "learning_rate": 0.0001511989796005438,
+      "loss": 1.6497,
+      "step": 15610
+    },
+    {
+      "epoch": 3.290672565439511,
+      "grad_norm": 0.9990530610084534,
+      "learning_rate": 0.00015114210784241482,
+      "loss": 1.6725,
+      "step": 15620
+    },
+    {
+      "epoch": 3.2927792700268603,
+      "grad_norm": 0.8826045393943787,
+      "learning_rate": 0.00015108521367526479,
+      "loss": 1.6639,
+      "step": 15630
+    },
+    {
+      "epoch": 3.2948859746142096,
+      "grad_norm": 0.9666330218315125,
+      "learning_rate": 0.00015102829712402302,
+      "loss": 1.7046,
+      "step": 15640
+    },
+    {
+      "epoch": 3.296992679201559,
+      "grad_norm": 0.9691329598426819,
+      "learning_rate": 0.00015097135821362883,
+      "loss": 1.714,
+      "step": 15650
+    },
+    {
+      "epoch": 3.299099383788908,
+      "grad_norm": 0.9856581687927246,
+      "learning_rate": 0.00015091439696903115,
+      "loss": 1.653,
+      "step": 15660
+    },
+    {
+      "epoch": 3.3012060883762575,
+      "grad_norm": 1.028440237045288,
+      "learning_rate": 0.00015085741341518886,
+      "loss": 1.6483,
+      "step": 15670
+    },
+    {
+      "epoch": 3.303312792963607,
+      "grad_norm": 1.0359004735946655,
+      "learning_rate": 0.00015080040757707046,
+      "loss": 1.7243,
+      "step": 15680
+    },
+    {
+      "epoch": 3.305419497550956,
+      "grad_norm": 1.0426095724105835,
+      "learning_rate": 0.00015074337947965435,
+      "loss": 1.6496,
+      "step": 15690
+    },
+    {
+      "epoch": 3.3075262021383054,
+      "grad_norm": 0.9464336633682251,
+      "learning_rate": 0.0001506863291479286,
+      "loss": 1.7339,
+      "step": 15700
+    },
+    {
+      "epoch": 3.3096329067256542,
+      "grad_norm": 0.9904953241348267,
+      "learning_rate": 0.00015062925660689106,
+      "loss": 1.6224,
+      "step": 15710
+    },
+    {
+      "epoch": 3.3117396113130035,
+      "grad_norm": 0.9366286396980286,
+      "learning_rate": 0.00015057216188154928,
+      "loss": 1.6543,
+      "step": 15720
+    },
+    {
+      "epoch": 3.313846315900353,
+      "grad_norm": 0.9428682923316956,
+      "learning_rate": 0.00015051504499692054,
+      "loss": 1.6566,
+      "step": 15730
+    },
+    {
+      "epoch": 3.315953020487702,
+      "grad_norm": 0.914440393447876,
+      "learning_rate": 0.0001504579059780319,
+      "loss": 1.6787,
+      "step": 15740
+    },
+    {
+      "epoch": 3.3180597250750514,
+      "grad_norm": 1.1147249937057495,
+      "learning_rate": 0.00015040074484992,
+      "loss": 1.6386,
+      "step": 15750
+    },
+    {
+      "epoch": 3.3201664296624007,
+      "grad_norm": 0.9929258227348328,
+      "learning_rate": 0.00015034356163763123,
+      "loss": 1.6538,
+      "step": 15760
+    },
+    {
+      "epoch": 3.32227313424975,
+      "grad_norm": 0.9763256907463074,
+      "learning_rate": 0.00015028635636622165,
+      "loss": 1.6431,
+      "step": 15770
+    },
+    {
+      "epoch": 3.324379838837099,
+      "grad_norm": 0.9423137903213501,
+      "learning_rate": 0.00015022912906075702,
+      "loss": 1.6739,
+      "step": 15780
+    },
+    {
+      "epoch": 3.326486543424448,
+      "grad_norm": 1.0271004438400269,
+      "learning_rate": 0.00015017187974631271,
+      "loss": 1.7341,
+      "step": 15790
+    },
+    {
+      "epoch": 3.3285932480117975,
+      "grad_norm": 1.0378199815750122,
+      "learning_rate": 0.00015011460844797372,
+      "loss": 1.7028,
+      "step": 15800
+    },
+    {
+      "epoch": 3.3306999525991468,
+      "grad_norm": 0.9106980562210083,
+      "learning_rate": 0.0001500573151908347,
+      "loss": 1.6631,
+      "step": 15810
+    },
+    {
+      "epoch": 3.332806657186496,
+      "grad_norm": 0.9558465480804443,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.7068,
+      "step": 15820
+    },
+    {
+      "epoch": 3.3349133617738453,
+      "grad_norm": 0.9833946824073792,
+      "learning_rate": 0.0001499426629005835,
+      "loss": 1.6597,
+      "step": 15830
+    },
+    {
+      "epoch": 3.3370200663611946,
+      "grad_norm": 0.9429462552070618,
+      "learning_rate": 0.00014988530391770856,
+      "loss": 1.6672,
+      "step": 15840
+    },
+    {
+      "epoch": 3.339126770948544,
+      "grad_norm": 1.01088285446167,
+      "learning_rate": 0.0001498279230765084,
+      "loss": 1.6842,
+      "step": 15850
+    },
+    {
+      "epoch": 3.3412334755358932,
+      "grad_norm": 0.9075558185577393,
+      "learning_rate": 0.00014977052040212566,
+      "loss": 1.6968,
+      "step": 15860
+    },
+    {
+      "epoch": 3.343340180123242,
+      "grad_norm": 1.0675263404846191,
+      "learning_rate": 0.00014971309591971252,
+      "loss": 1.6572,
+      "step": 15870
+    },
+    {
+      "epoch": 3.3454468847105914,
+      "grad_norm": 0.978440523147583,
+      "learning_rate": 0.00014965564965443079,
+      "loss": 1.7329,
+      "step": 15880
+    },
+    {
+      "epoch": 3.3475535892979407,
+      "grad_norm": 0.9912124276161194,
+      "learning_rate": 0.00014959818163145174,
+      "loss": 1.6253,
+      "step": 15890
+    },
+    {
+      "epoch": 3.34966029388529,
+      "grad_norm": 0.9947911500930786,
+      "learning_rate": 0.00014954069187595633,
+      "loss": 1.6586,
+      "step": 15900
+    },
+    {
+      "epoch": 3.3517669984726393,
+      "grad_norm": 0.9598615765571594,
+      "learning_rate": 0.00014948318041313482,
+      "loss": 1.6465,
+      "step": 15910
+    },
+    {
+      "epoch": 3.3538737030599886,
+      "grad_norm": 0.9510218501091003,
+      "learning_rate": 0.0001494256472681872,
+      "loss": 1.6517,
+      "step": 15920
+    },
+    {
+      "epoch": 3.355980407647338,
+      "grad_norm": 1.0019654035568237,
+      "learning_rate": 0.0001493680924663228,
+      "loss": 1.6578,
+      "step": 15930
+    },
+    {
+      "epoch": 3.3580871122346867,
+      "grad_norm": 0.9807612895965576,
+      "learning_rate": 0.00014931051603276054,
+      "loss": 1.6576,
+      "step": 15940
+    },
+    {
+      "epoch": 3.360193816822036,
+      "grad_norm": 0.9929197430610657,
+      "learning_rate": 0.00014925291799272876,
+      "loss": 1.6551,
+      "step": 15950
+    },
+    {
+      "epoch": 3.3623005214093853,
+      "grad_norm": 0.9978349208831787,
+      "learning_rate": 0.00014919529837146528,
+      "loss": 1.6679,
+      "step": 15960
+    },
+    {
+      "epoch": 3.3644072259967346,
+      "grad_norm": 0.8876854777336121,
+      "learning_rate": 0.00014913765719421743,
+      "loss": 1.7459,
+      "step": 15970
+    },
+    {
+      "epoch": 3.366513930584084,
+      "grad_norm": 1.0585439205169678,
+      "learning_rate": 0.00014907999448624188,
+      "loss": 1.6814,
+      "step": 15980
+    },
+    {
+      "epoch": 3.368620635171433,
+      "grad_norm": 1.0199978351593018,
+      "learning_rate": 0.00014902231027280486,
+      "loss": 1.6441,
+      "step": 15990
+    },
+    {
+      "epoch": 3.3707273397587825,
+      "grad_norm": 0.9279028177261353,
+      "learning_rate": 0.00014896460457918185,
+      "loss": 1.6734,
+      "step": 16000
+    },
+    {
+      "epoch": 3.3728340443461313,
+      "grad_norm": 0.9600446820259094,
+      "learning_rate": 0.00014890687743065794,
+      "loss": 1.7033,
+      "step": 16010
+    },
+    {
+      "epoch": 3.3749407489334806,
+      "grad_norm": 0.9573974609375,
+      "learning_rate": 0.0001488491288525275,
+      "loss": 1.706,
+      "step": 16020
+    },
+    {
+      "epoch": 3.37704745352083,
+      "grad_norm": 0.9946306347846985,
+      "learning_rate": 0.00014879135887009435,
+      "loss": 1.6575,
+      "step": 16030
+    },
+    {
+      "epoch": 3.379154158108179,
+      "grad_norm": 0.9504879117012024,
+      "learning_rate": 0.0001487335675086716,
+      "loss": 1.6634,
+      "step": 16040
+    },
+    {
+      "epoch": 3.3812608626955285,
+      "grad_norm": 1.0545927286148071,
+      "learning_rate": 0.0001486757547935818,
+      "loss": 1.7045,
+      "step": 16050
+    },
+    {
+      "epoch": 3.383367567282878,
+      "grad_norm": 1.1232430934906006,
+      "learning_rate": 0.00014861792075015687,
+      "loss": 1.6703,
+      "step": 16060
+    },
+    {
+      "epoch": 3.385474271870227,
+      "grad_norm": 0.9520963430404663,
+      "learning_rate": 0.00014856006540373806,
+      "loss": 1.7003,
+      "step": 16070
+    },
+    {
+      "epoch": 3.3875809764575764,
+      "grad_norm": 1.0433330535888672,
+      "learning_rate": 0.0001485021887796759,
+      "loss": 1.6963,
+      "step": 16080
+    },
+    {
+      "epoch": 3.3896876810449257,
+      "grad_norm": 1.0261874198913574,
+      "learning_rate": 0.0001484442909033303,
+      "loss": 1.6862,
+      "step": 16090
+    },
+    {
+      "epoch": 3.3917943856322745,
+      "grad_norm": 0.8804453015327454,
+      "learning_rate": 0.00014838637180007047,
+      "loss": 1.6284,
+      "step": 16100
+    },
+    {
+      "epoch": 3.393901090219624,
+      "grad_norm": 1.066070795059204,
+      "learning_rate": 0.0001483284314952749,
+      "loss": 1.6803,
+      "step": 16110
+    },
+    {
+      "epoch": 3.396007794806973,
+      "grad_norm": 0.9975647330284119,
+      "learning_rate": 0.0001482704700143314,
+      "loss": 1.7284,
+      "step": 16120
+    },
+    {
+      "epoch": 3.3981144993943224,
+      "grad_norm": 1.016730546951294,
+      "learning_rate": 0.0001482124873826371,
+      "loss": 1.6747,
+      "step": 16130
+    },
+    {
+      "epoch": 3.4002212039816717,
+      "grad_norm": 1.0822088718414307,
+      "learning_rate": 0.00014815448362559826,
+      "loss": 1.6058,
+      "step": 16140
+    },
+    {
+      "epoch": 3.402327908569021,
+      "grad_norm": 0.9539144039154053,
+      "learning_rate": 0.00014809645876863052,
+      "loss": 1.6734,
+      "step": 16150
+    },
+    {
+      "epoch": 3.4044346131563703,
+      "grad_norm": 0.995520830154419,
+      "learning_rate": 0.0001480384128371587,
+      "loss": 1.6552,
+      "step": 16160
+    },
+    {
+      "epoch": 3.406541317743719,
+      "grad_norm": 1.0045973062515259,
+      "learning_rate": 0.00014798034585661695,
+      "loss": 1.7047,
+      "step": 16170
+    },
+    {
+      "epoch": 3.4086480223310685,
+      "grad_norm": 0.9666810035705566,
+      "learning_rate": 0.0001479222578524485,
+      "loss": 1.6763,
+      "step": 16180
+    },
+    {
+      "epoch": 3.4107547269184177,
+      "grad_norm": 0.9643430113792419,
+      "learning_rate": 0.0001478641488501059,
+      "loss": 1.6404,
+      "step": 16190
+    },
+    {
+      "epoch": 3.412861431505767,
+      "grad_norm": 1.0544071197509766,
+      "learning_rate": 0.00014780601887505088,
+      "loss": 1.6564,
+      "step": 16200
+    },
+    {
+      "epoch": 3.4149681360931163,
+      "grad_norm": 0.9909730553627014,
+      "learning_rate": 0.00014774786795275429,
+      "loss": 1.6865,
+      "step": 16210
+    },
+    {
+      "epoch": 3.4170748406804656,
+      "grad_norm": 0.9278872609138489,
+      "learning_rate": 0.00014768969610869627,
+      "loss": 1.6599,
+      "step": 16220
+    },
+    {
+      "epoch": 3.419181545267815,
+      "grad_norm": 0.9268772602081299,
+      "learning_rate": 0.00014763150336836604,
+      "loss": 1.6472,
+      "step": 16230
+    },
+    {
+      "epoch": 3.421288249855164,
+      "grad_norm": 0.9475058317184448,
+      "learning_rate": 0.00014757328975726207,
+      "loss": 1.667,
+      "step": 16240
+    },
+    {
+      "epoch": 3.4233949544425135,
+      "grad_norm": 1.048485279083252,
+      "learning_rate": 0.0001475150553008918,
+      "loss": 1.7311,
+      "step": 16250
+    },
+    {
+      "epoch": 3.4255016590298624,
+      "grad_norm": 0.9860157370567322,
+      "learning_rate": 0.00014745680002477203,
+      "loss": 1.6184,
+      "step": 16260
+    },
+    {
+      "epoch": 3.4276083636172117,
+      "grad_norm": 0.9206960201263428,
+      "learning_rate": 0.00014739852395442854,
+      "loss": 1.6858,
+      "step": 16270
+    },
+    {
+      "epoch": 3.429715068204561,
+      "grad_norm": 0.9898627400398254,
+      "learning_rate": 0.0001473402271153962,
+      "loss": 1.6377,
+      "step": 16280
+    },
+    {
+      "epoch": 3.4318217727919103,
+      "grad_norm": 1.0353739261627197,
+      "learning_rate": 0.00014728190953321903,
+      "loss": 1.6448,
+      "step": 16290
+    },
+    {
+      "epoch": 3.4339284773792595,
+      "grad_norm": 0.9169580340385437,
+      "learning_rate": 0.00014722357123345023,
+      "loss": 1.657,
+      "step": 16300
+    },
+    {
+      "epoch": 3.436035181966609,
+      "grad_norm": 1.0277870893478394,
+      "learning_rate": 0.00014716521224165192,
+      "loss": 1.6797,
+      "step": 16310
+    },
+    {
+      "epoch": 3.438141886553958,
+      "grad_norm": 0.9552232623100281,
+      "learning_rate": 0.00014710683258339536,
+      "loss": 1.6266,
+      "step": 16320
+    },
+    {
+      "epoch": 3.440248591141307,
+      "grad_norm": 0.9857643246650696,
+      "learning_rate": 0.00014704843228426087,
+      "loss": 1.6921,
+      "step": 16330
+    },
+    {
+      "epoch": 3.4423552957286563,
+      "grad_norm": 1.0019571781158447,
+      "learning_rate": 0.00014699001136983782,
+      "loss": 1.6315,
+      "step": 16340
+    },
+    {
+      "epoch": 3.4444620003160056,
+      "grad_norm": 1.0217735767364502,
+      "learning_rate": 0.00014693156986572456,
+      "loss": 1.6518,
+      "step": 16350
+    },
+    {
+      "epoch": 3.446568704903355,
+      "grad_norm": 0.9069836735725403,
+      "learning_rate": 0.00014687310779752855,
+      "loss": 1.5829,
+      "step": 16360
+    },
+    {
+      "epoch": 3.448675409490704,
+      "grad_norm": 1.0425034761428833,
+      "learning_rate": 0.00014681462519086614,
+      "loss": 1.7103,
+      "step": 16370
+    },
+    {
+      "epoch": 3.4507821140780535,
+      "grad_norm": 0.999427318572998,
+      "learning_rate": 0.0001467561220713628,
+      "loss": 1.65,
+      "step": 16380
+    },
+    {
+      "epoch": 3.4528888186654028,
+      "grad_norm": 0.9969028234481812,
+      "learning_rate": 0.00014669759846465296,
+      "loss": 1.6209,
+      "step": 16390
+    },
+    {
+      "epoch": 3.454995523252752,
+      "grad_norm": 1.015284538269043,
+      "learning_rate": 0.00014663905439637995,
+      "loss": 1.689,
+      "step": 16400
+    },
+    {
+      "epoch": 3.4571022278401013,
+      "grad_norm": 0.9351761937141418,
+      "learning_rate": 0.00014658048989219614,
+      "loss": 1.6442,
+      "step": 16410
+    },
+    {
+      "epoch": 3.45920893242745,
+      "grad_norm": 1.002864122390747,
+      "learning_rate": 0.00014652190497776286,
+      "loss": 1.635,
+      "step": 16420
+    },
+    {
+      "epoch": 3.4613156370147995,
+      "grad_norm": 1.0125610828399658,
+      "learning_rate": 0.00014646329967875032,
+      "loss": 1.6677,
+      "step": 16430
+    },
+    {
+      "epoch": 3.463422341602149,
+      "grad_norm": 1.0910165309906006,
+      "learning_rate": 0.0001464046740208377,
+      "loss": 1.7004,
+      "step": 16440
+    },
+    {
+      "epoch": 3.465529046189498,
+      "grad_norm": 0.9675852060317993,
+      "learning_rate": 0.00014634602802971312,
+      "loss": 1.667,
+      "step": 16450
+    },
+    {
+      "epoch": 3.4676357507768474,
+      "grad_norm": 1.0512768030166626,
+      "learning_rate": 0.00014628736173107363,
+      "loss": 1.675,
+      "step": 16460
+    },
+    {
+      "epoch": 3.4697424553641967,
+      "grad_norm": 1.171718716621399,
+      "learning_rate": 0.00014622867515062503,
+      "loss": 1.6984,
+      "step": 16470
+    },
+    {
+      "epoch": 3.471849159951546,
+      "grad_norm": 1.0073360204696655,
+      "learning_rate": 0.0001461699683140822,
+      "loss": 1.6667,
+      "step": 16480
+    },
+    {
+      "epoch": 3.473955864538895,
+      "grad_norm": 0.9846724271774292,
+      "learning_rate": 0.00014611124124716882,
+      "loss": 1.6832,
+      "step": 16490
+    },
+    {
+      "epoch": 3.476062569126244,
+      "grad_norm": 0.9981314539909363,
+      "learning_rate": 0.00014605249397561736,
+      "loss": 1.6448,
+      "step": 16500
+    },
+    {
+      "epoch": 3.4781692737135934,
+      "grad_norm": 0.9036053419113159,
+      "learning_rate": 0.00014599372652516926,
+      "loss": 1.6785,
+      "step": 16510
+    },
+    {
+      "epoch": 3.4802759783009427,
+      "grad_norm": 0.9795652031898499,
+      "learning_rate": 0.00014593493892157473,
+      "loss": 1.6672,
+      "step": 16520
+    },
+    {
+      "epoch": 3.482382682888292,
+      "grad_norm": 1.0752781629562378,
+      "learning_rate": 0.00014587613119059284,
+      "loss": 1.6374,
+      "step": 16530
+    },
+    {
+      "epoch": 3.4844893874756413,
+      "grad_norm": 0.9755626916885376,
+      "learning_rate": 0.0001458173033579914,
+      "loss": 1.6375,
+      "step": 16540
+    },
+    {
+      "epoch": 3.4865960920629906,
+      "grad_norm": 1.0344021320343018,
+      "learning_rate": 0.00014575845544954725,
+      "loss": 1.5904,
+      "step": 16550
+    },
+    {
+      "epoch": 3.48870279665034,
+      "grad_norm": 1.0018855333328247,
+      "learning_rate": 0.00014569958749104575,
+      "loss": 1.7213,
+      "step": 16560
+    },
+    {
+      "epoch": 3.490809501237689,
+      "grad_norm": 1.0331135988235474,
+      "learning_rate": 0.00014564069950828118,
+      "loss": 1.6596,
+      "step": 16570
+    },
+    {
+      "epoch": 3.492916205825038,
+      "grad_norm": 1.0158442258834839,
+      "learning_rate": 0.0001455817915270566,
+      "loss": 1.7339,
+      "step": 16580
+    },
+    {
+      "epoch": 3.4950229104123873,
+      "grad_norm": 1.087720274925232,
+      "learning_rate": 0.0001455228635731839,
+      "loss": 1.6767,
+      "step": 16590
+    },
+    {
+      "epoch": 3.4971296149997366,
+      "grad_norm": 0.95531165599823,
+      "learning_rate": 0.00014546391567248353,
+      "loss": 1.6726,
+      "step": 16600
+    },
+    {
+      "epoch": 3.499236319587086,
+      "grad_norm": 1.0385819673538208,
+      "learning_rate": 0.00014540494785078478,
+      "loss": 1.6336,
+      "step": 16610
+    },
+    {
+      "epoch": 3.501343024174435,
+      "grad_norm": 1.0225718021392822,
+      "learning_rate": 0.00014534596013392575,
+      "loss": 1.6422,
+      "step": 16620
+    },
+    {
+      "epoch": 3.5034497287617845,
+      "grad_norm": 1.0323877334594727,
+      "learning_rate": 0.00014528695254775316,
+      "loss": 1.7083,
+      "step": 16630
+    },
+    {
+      "epoch": 3.505556433349134,
+      "grad_norm": 0.9760758280754089,
+      "learning_rate": 0.00014522792511812243,
+      "loss": 1.6618,
+      "step": 16640
+    },
+    {
+      "epoch": 3.5076631379364827,
+      "grad_norm": 0.9613909721374512,
+      "learning_rate": 0.00014516887787089774,
+      "loss": 1.7509,
+      "step": 16650
+    },
+    {
+      "epoch": 3.509769842523832,
+      "grad_norm": 1.0987319946289062,
+      "learning_rate": 0.00014510981083195188,
+      "loss": 1.6929,
+      "step": 16660
+    },
+    {
+      "epoch": 3.5118765471111812,
+      "grad_norm": 0.8977022767066956,
+      "learning_rate": 0.00014505072402716643,
+      "loss": 1.656,
+      "step": 16670
+    },
+    {
+      "epoch": 3.5139832516985305,
+      "grad_norm": 0.9773346781730652,
+      "learning_rate": 0.00014499161748243147,
+      "loss": 1.7439,
+      "step": 16680
+    },
+    {
+      "epoch": 3.51608995628588,
+      "grad_norm": 0.9943004250526428,
+      "learning_rate": 0.00014493249122364584,
+      "loss": 1.664,
+      "step": 16690
+    },
+    {
+      "epoch": 3.518196660873229,
+      "grad_norm": 1.1163723468780518,
+      "learning_rate": 0.00014487334527671697,
+      "loss": 1.6813,
+      "step": 16700
+    },
+    {
+      "epoch": 3.5203033654605784,
+      "grad_norm": 0.9387895464897156,
+      "learning_rate": 0.00014481417966756102,
+      "loss": 1.706,
+      "step": 16710
+    },
+    {
+      "epoch": 3.5224100700479273,
+      "grad_norm": 0.9459789991378784,
+      "learning_rate": 0.00014475499442210258,
+      "loss": 1.725,
+      "step": 16720
+    },
+    {
+      "epoch": 3.524516774635277,
+      "grad_norm": 0.9388308525085449,
+      "learning_rate": 0.00014469578956627496,
+      "loss": 1.6384,
+      "step": 16730
+    },
+    {
+      "epoch": 3.526623479222626,
+      "grad_norm": 1.0331875085830688,
+      "learning_rate": 0.0001446365651260201,
+      "loss": 1.6376,
+      "step": 16740
+    },
+    {
+      "epoch": 3.528730183809975,
+      "grad_norm": 0.9883841276168823,
+      "learning_rate": 0.00014457732112728848,
+      "loss": 1.6539,
+      "step": 16750
+    },
+    {
+      "epoch": 3.5308368883973245,
+      "grad_norm": 0.9518382549285889,
+      "learning_rate": 0.00014451805759603908,
+      "loss": 1.6514,
+      "step": 16760
+    },
+    {
+      "epoch": 3.5329435929846738,
+      "grad_norm": 1.0333259105682373,
+      "learning_rate": 0.00014445877455823946,
+      "loss": 1.7183,
+      "step": 16770
+    },
+    {
+      "epoch": 3.535050297572023,
+      "grad_norm": 0.9487828016281128,
+      "learning_rate": 0.0001443994720398659,
+      "loss": 1.6944,
+      "step": 16780
+    },
+    {
+      "epoch": 3.5371570021593723,
+      "grad_norm": 0.9641709327697754,
+      "learning_rate": 0.00014434015006690299,
+      "loss": 1.6249,
+      "step": 16790
+    },
+    {
+      "epoch": 3.5392637067467216,
+      "grad_norm": 0.936957597732544,
+      "learning_rate": 0.00014428080866534396,
+      "loss": 1.6875,
+      "step": 16800
+    },
+    {
+      "epoch": 3.5413704113340705,
+      "grad_norm": 0.9620563983917236,
+      "learning_rate": 0.0001442214478611905,
+      "loss": 1.6837,
+      "step": 16810
+    },
+    {
+      "epoch": 3.54347711592142,
+      "grad_norm": 0.9967496395111084,
+      "learning_rate": 0.00014416206768045288,
+      "loss": 1.6423,
+      "step": 16820
+    },
+    {
+      "epoch": 3.545583820508769,
+      "grad_norm": 0.9146559834480286,
+      "learning_rate": 0.0001441026681491498,
+      "loss": 1.6752,
+      "step": 16830
+    },
+    {
+      "epoch": 3.5476905250961184,
+      "grad_norm": 1.0962780714035034,
+      "learning_rate": 0.0001440432492933084,
+      "loss": 1.7516,
+      "step": 16840
+    },
+    {
+      "epoch": 3.5497972296834677,
+      "grad_norm": 1.0509783029556274,
+      "learning_rate": 0.00014398381113896438,
+      "loss": 1.6429,
+      "step": 16850
+    },
+    {
+      "epoch": 3.551903934270817,
+      "grad_norm": 0.9411798119544983,
+      "learning_rate": 0.00014392435371216185,
+      "loss": 1.6632,
+      "step": 16860
+    },
+    {
+      "epoch": 3.5540106388581663,
+      "grad_norm": 1.1193450689315796,
+      "learning_rate": 0.0001438648770389534,
+      "loss": 1.7557,
+      "step": 16870
+    },
+    {
+      "epoch": 3.556117343445515,
+      "grad_norm": 0.9980596303939819,
+      "learning_rate": 0.00014380538114539996,
+      "loss": 1.6619,
+      "step": 16880
+    },
+    {
+      "epoch": 3.558224048032865,
+      "grad_norm": 0.8729853630065918,
+      "learning_rate": 0.00014374586605757095,
+      "loss": 1.6667,
+      "step": 16890
+    },
+    {
+      "epoch": 3.5603307526202137,
+      "grad_norm": 0.9395302534103394,
+      "learning_rate": 0.00014368633180154424,
+      "loss": 1.6658,
+      "step": 16900
+    },
+    {
+      "epoch": 3.562437457207563,
+      "grad_norm": 0.9311736226081848,
+      "learning_rate": 0.00014362677840340602,
+      "loss": 1.6424,
+      "step": 16910
+    },
+    {
+      "epoch": 3.5645441617949123,
+      "grad_norm": 0.8994746208190918,
+      "learning_rate": 0.0001435672058892509,
+      "loss": 1.678,
+      "step": 16920
+    },
+    {
+      "epoch": 3.5666508663822616,
+      "grad_norm": 1.0269373655319214,
+      "learning_rate": 0.00014350761428518185,
+      "loss": 1.6777,
+      "step": 16930
+    },
+    {
+      "epoch": 3.568757570969611,
+      "grad_norm": 1.0147863626480103,
+      "learning_rate": 0.00014344800361731027,
+      "loss": 1.6976,
+      "step": 16940
+    },
+    {
+      "epoch": 3.57086427555696,
+      "grad_norm": 0.9315280318260193,
+      "learning_rate": 0.00014338837391175582,
+      "loss": 1.6918,
+      "step": 16950
+    },
+    {
+      "epoch": 3.5729709801443095,
+      "grad_norm": 0.9913371205329895,
+      "learning_rate": 0.0001433287251946466,
+      "loss": 1.7383,
+      "step": 16960
+    },
+    {
+      "epoch": 3.5750776847316583,
+      "grad_norm": 1.0025800466537476,
+      "learning_rate": 0.00014326905749211888,
+      "loss": 1.6843,
+      "step": 16970
+    },
+    {
+      "epoch": 3.5771843893190076,
+      "grad_norm": 1.001194715499878,
+      "learning_rate": 0.00014320937083031748,
+      "loss": 1.7082,
+      "step": 16980
+    },
+    {
+      "epoch": 3.579291093906357,
+      "grad_norm": 0.9992178082466125,
+      "learning_rate": 0.0001431496652353953,
+      "loss": 1.7006,
+      "step": 16990
+    },
+    {
+      "epoch": 3.581397798493706,
+      "grad_norm": 0.9966685175895691,
+      "learning_rate": 0.0001430899407335137,
+      "loss": 1.6599,
+      "step": 17000
+    },
+    {
+      "epoch": 3.5835045030810555,
+      "grad_norm": 0.9909036755561829,
+      "learning_rate": 0.00014303019735084226,
+      "loss": 1.699,
+      "step": 17010
+    },
+    {
+      "epoch": 3.585611207668405,
+      "grad_norm": 0.8726781010627747,
+      "learning_rate": 0.0001429704351135588,
+      "loss": 1.7063,
+      "step": 17020
+    },
+    {
+      "epoch": 3.587717912255754,
+      "grad_norm": 0.9716811180114746,
+      "learning_rate": 0.00014291065404784946,
+      "loss": 1.6646,
+      "step": 17030
+    },
+    {
+      "epoch": 3.589824616843103,
+      "grad_norm": 1.0037916898727417,
+      "learning_rate": 0.0001428508541799086,
+      "loss": 1.6589,
+      "step": 17040
+    },
+    {
+      "epoch": 3.5919313214304527,
+      "grad_norm": 1.0080369710922241,
+      "learning_rate": 0.00014279103553593885,
+      "loss": 1.7106,
+      "step": 17050
+    },
+    {
+      "epoch": 3.5940380260178015,
+      "grad_norm": 1.0465925931930542,
+      "learning_rate": 0.00014273119814215102,
+      "loss": 1.6131,
+      "step": 17060
+    },
+    {
+      "epoch": 3.596144730605151,
+      "grad_norm": 0.9277071356773376,
+      "learning_rate": 0.00014267134202476417,
+      "loss": 1.6802,
+      "step": 17070
+    },
+    {
+      "epoch": 3.5982514351925,
+      "grad_norm": 0.9744635820388794,
+      "learning_rate": 0.00014261146721000553,
+      "loss": 1.6739,
+      "step": 17080
+    },
+    {
+      "epoch": 3.6003581397798494,
+      "grad_norm": 0.9288362264633179,
+      "learning_rate": 0.00014255157372411058,
+      "loss": 1.6332,
+      "step": 17090
+    },
+    {
+      "epoch": 3.6024648443671987,
+      "grad_norm": 0.9643329381942749,
+      "learning_rate": 0.0001424916615933229,
+      "loss": 1.6615,
+      "step": 17100
+    },
+    {
+      "epoch": 3.604571548954548,
+      "grad_norm": 1.0060192346572876,
+      "learning_rate": 0.00014243173084389437,
+      "loss": 1.6714,
+      "step": 17110
+    },
+    {
+      "epoch": 3.6066782535418973,
+      "grad_norm": 0.9957178831100464,
+      "learning_rate": 0.00014237178150208486,
+      "loss": 1.6903,
+      "step": 17120
+    },
+    {
+      "epoch": 3.608784958129246,
+      "grad_norm": 0.9705983400344849,
+      "learning_rate": 0.00014231181359416247,
+      "loss": 1.6499,
+      "step": 17130
+    },
+    {
+      "epoch": 3.6108916627165955,
+      "grad_norm": 1.016700029373169,
+      "learning_rate": 0.0001422518271464035,
+      "loss": 1.7183,
+      "step": 17140
+    },
+    {
+      "epoch": 3.6129983673039447,
+      "grad_norm": 1.1801646947860718,
+      "learning_rate": 0.0001421918221850923,
+      "loss": 1.6735,
+      "step": 17150
+    },
+    {
+      "epoch": 3.615105071891294,
+      "grad_norm": 1.018476128578186,
+      "learning_rate": 0.00014213179873652127,
+      "loss": 1.6662,
+      "step": 17160
+    },
+    {
+      "epoch": 3.6172117764786433,
+      "grad_norm": 1.0025722980499268,
+      "learning_rate": 0.00014207175682699107,
+      "loss": 1.6927,
+      "step": 17170
+    },
+    {
+      "epoch": 3.6193184810659926,
+      "grad_norm": 0.9412466883659363,
+      "learning_rate": 0.00014201169648281027,
+      "loss": 1.7017,
+      "step": 17180
+    },
+    {
+      "epoch": 3.621425185653342,
+      "grad_norm": 0.997974693775177,
+      "learning_rate": 0.0001419516177302957,
+      "loss": 1.6749,
+      "step": 17190
+    },
+    {
+      "epoch": 3.623531890240691,
+      "grad_norm": 1.0293715000152588,
+      "learning_rate": 0.00014189152059577214,
+      "loss": 1.6701,
+      "step": 17200
+    },
+    {
+      "epoch": 3.62563859482804,
+      "grad_norm": 1.0058647394180298,
+      "learning_rate": 0.0001418314051055724,
+      "loss": 1.6771,
+      "step": 17210
+    },
+    {
+      "epoch": 3.6277452994153894,
+      "grad_norm": 0.9235060214996338,
+      "learning_rate": 0.00014177127128603745,
+      "loss": 1.6577,
+      "step": 17220
+    },
+    {
+      "epoch": 3.6298520040027387,
+      "grad_norm": 1.0182476043701172,
+      "learning_rate": 0.00014171111916351623,
+      "loss": 1.786,
+      "step": 17230
+    },
+    {
+      "epoch": 3.631958708590088,
+      "grad_norm": 1.0080841779708862,
+      "learning_rate": 0.00014165094876436563,
+      "loss": 1.6719,
+      "step": 17240
+    },
+    {
+      "epoch": 3.6340654131774373,
+      "grad_norm": 0.9524149894714355,
+      "learning_rate": 0.00014159076011495061,
+      "loss": 1.6081,
+      "step": 17250
+    },
+    {
+      "epoch": 3.6361721177647865,
+      "grad_norm": 0.9929611086845398,
+      "learning_rate": 0.00014153055324164418,
+      "loss": 1.6987,
+      "step": 17260
+    },
+    {
+      "epoch": 3.6382788223521354,
+      "grad_norm": 0.9325882196426392,
+      "learning_rate": 0.00014147032817082728,
+      "loss": 1.6635,
+      "step": 17270
+    },
+    {
+      "epoch": 3.640385526939485,
+      "grad_norm": 1.0150893926620483,
+      "learning_rate": 0.0001414100849288888,
+      "loss": 1.6721,
+      "step": 17280
+    },
+    {
+      "epoch": 3.642492231526834,
+      "grad_norm": 1.0248265266418457,
+      "learning_rate": 0.00014134982354222563,
+      "loss": 1.7073,
+      "step": 17290
+    },
+    {
+      "epoch": 3.6445989361141833,
+      "grad_norm": 1.0384207963943481,
+      "learning_rate": 0.0001412895440372426,
+      "loss": 1.5953,
+      "step": 17300
+    },
+    {
+      "epoch": 3.6467056407015326,
+      "grad_norm": 0.996492862701416,
+      "learning_rate": 0.00014122924644035249,
+      "loss": 1.6656,
+      "step": 17310
+    },
+    {
+      "epoch": 3.648812345288882,
+      "grad_norm": 0.9860630035400391,
+      "learning_rate": 0.000141168930777976,
+      "loss": 1.7256,
+      "step": 17320
+    },
+    {
+      "epoch": 3.650919049876231,
+      "grad_norm": 0.9836267232894897,
+      "learning_rate": 0.0001411085970765417,
+      "loss": 1.6836,
+      "step": 17330
+    },
+    {
+      "epoch": 3.6530257544635805,
+      "grad_norm": 0.9142735004425049,
+      "learning_rate": 0.00014104824536248614,
+      "loss": 1.6744,
+      "step": 17340
+    },
+    {
+      "epoch": 3.6551324590509298,
+      "grad_norm": 1.0227885246276855,
+      "learning_rate": 0.00014098787566225375,
+      "loss": 1.7145,
+      "step": 17350
+    },
+    {
+      "epoch": 3.6572391636382786,
+      "grad_norm": 1.0329092741012573,
+      "learning_rate": 0.00014092748800229683,
+      "loss": 1.6737,
+      "step": 17360
+    },
+    {
+      "epoch": 3.659345868225628,
+      "grad_norm": 1.0570374727249146,
+      "learning_rate": 0.00014086708240907542,
+      "loss": 1.6968,
+      "step": 17370
+    },
+    {
+      "epoch": 3.661452572812977,
+      "grad_norm": 0.9342901110649109,
+      "learning_rate": 0.0001408066589090577,
+      "loss": 1.6643,
+      "step": 17380
+    },
+    {
+      "epoch": 3.6635592774003265,
+      "grad_norm": 0.9996681809425354,
+      "learning_rate": 0.00014074621752871943,
+      "loss": 1.6771,
+      "step": 17390
+    },
+    {
+      "epoch": 3.665665981987676,
+      "grad_norm": 0.9607349634170532,
+      "learning_rate": 0.00014068575829454436,
+      "loss": 1.6859,
+      "step": 17400
+    },
+    {
+      "epoch": 3.667772686575025,
+      "grad_norm": 1.1797279119491577,
+      "learning_rate": 0.00014062528123302395,
+      "loss": 1.6771,
+      "step": 17410
+    },
+    {
+      "epoch": 3.6698793911623744,
+      "grad_norm": 1.010092854499817,
+      "learning_rate": 0.00014056478637065761,
+      "loss": 1.7489,
+      "step": 17420
+    },
+    {
+      "epoch": 3.6719860957497232,
+      "grad_norm": 0.9948980212211609,
+      "learning_rate": 0.0001405042737339524,
+      "loss": 1.6696,
+      "step": 17430
+    },
+    {
+      "epoch": 3.674092800337073,
+      "grad_norm": 1.0839941501617432,
+      "learning_rate": 0.00014044374334942333,
+      "loss": 1.7265,
+      "step": 17440
+    },
+    {
+      "epoch": 3.676199504924422,
+      "grad_norm": 0.9895996451377869,
+      "learning_rate": 0.00014038319524359297,
+      "loss": 1.694,
+      "step": 17450
+    },
+    {
+      "epoch": 3.678306209511771,
+      "grad_norm": 1.0752763748168945,
+      "learning_rate": 0.00014032262944299194,
+      "loss": 1.6722,
+      "step": 17460
+    },
+    {
+      "epoch": 3.6804129140991204,
+      "grad_norm": 1.049815058708191,
+      "learning_rate": 0.0001402620459741583,
+      "loss": 1.6526,
+      "step": 17470
+    },
+    {
+      "epoch": 3.6825196186864697,
+      "grad_norm": 0.9036969542503357,
+      "learning_rate": 0.00014020144486363812,
+      "loss": 1.686,
+      "step": 17480
+    },
+    {
+      "epoch": 3.684626323273819,
+      "grad_norm": 1.0472948551177979,
+      "learning_rate": 0.00014014082613798503,
+      "loss": 1.7052,
+      "step": 17490
+    },
+    {
+      "epoch": 3.6867330278611683,
+      "grad_norm": 1.0632350444793701,
+      "learning_rate": 0.00014008018982376044,
+      "loss": 1.7132,
+      "step": 17500
+    },
+    {
+      "epoch": 3.6888397324485176,
+      "grad_norm": 0.9786173105239868,
+      "learning_rate": 0.00014001953594753352,
+      "loss": 1.6874,
+      "step": 17510
+    },
+    {
+      "epoch": 3.6909464370358664,
+      "grad_norm": 0.977205753326416,
+      "learning_rate": 0.00013995886453588104,
+      "loss": 1.6544,
+      "step": 17520
+    },
+    {
+      "epoch": 3.6930531416232157,
+      "grad_norm": 0.9975076913833618,
+      "learning_rate": 0.00013989817561538742,
+      "loss": 1.6845,
+      "step": 17530
+    },
+    {
+      "epoch": 3.695159846210565,
+      "grad_norm": 1.027984857559204,
+      "learning_rate": 0.00013983746921264494,
+      "loss": 1.6706,
+      "step": 17540
+    },
+    {
+      "epoch": 3.6972665507979143,
+      "grad_norm": 0.9895164966583252,
+      "learning_rate": 0.00013977674535425337,
+      "loss": 1.6418,
+      "step": 17550
+    },
+    {
+      "epoch": 3.6993732553852636,
+      "grad_norm": 0.8872331380844116,
+      "learning_rate": 0.0001397160040668202,
+      "loss": 1.7271,
+      "step": 17560
+    },
+    {
+      "epoch": 3.701479959972613,
+      "grad_norm": 0.8885982036590576,
+      "learning_rate": 0.00013965524537696048,
+      "loss": 1.6964,
+      "step": 17570
+    },
+    {
+      "epoch": 3.703586664559962,
+      "grad_norm": 0.9368189573287964,
+      "learning_rate": 0.00013959446931129704,
+      "loss": 1.6421,
+      "step": 17580
+    },
+    {
+      "epoch": 3.705693369147311,
+      "grad_norm": 1.0521513223648071,
+      "learning_rate": 0.00013953367589646015,
+      "loss": 1.6955,
+      "step": 17590
+    },
+    {
+      "epoch": 3.707800073734661,
+      "grad_norm": 0.9745957255363464,
+      "learning_rate": 0.00013947286515908779,
+      "loss": 1.6645,
+      "step": 17600
+    },
+    {
+      "epoch": 3.7099067783220097,
+      "grad_norm": 0.8766047358512878,
+      "learning_rate": 0.00013941203712582553,
+      "loss": 1.6694,
+      "step": 17610
+    },
+    {
+      "epoch": 3.712013482909359,
+      "grad_norm": 1.053513526916504,
+      "learning_rate": 0.00013935119182332642,
+      "loss": 1.7104,
+      "step": 17620
+    },
+    {
+      "epoch": 3.7141201874967082,
+      "grad_norm": 0.9564160704612732,
+      "learning_rate": 0.0001392903292782512,
+      "loss": 1.6903,
+      "step": 17630
+    },
+    {
+      "epoch": 3.7162268920840575,
+      "grad_norm": 0.9763982892036438,
+      "learning_rate": 0.0001392294495172681,
+      "loss": 1.6762,
+      "step": 17640
+    },
+    {
+      "epoch": 3.718333596671407,
+      "grad_norm": 0.9978336095809937,
+      "learning_rate": 0.00013916855256705288,
+      "loss": 1.6788,
+      "step": 17650
+    },
+    {
+      "epoch": 3.720440301258756,
+      "grad_norm": 1.0024161338806152,
+      "learning_rate": 0.0001391076384542889,
+      "loss": 1.6639,
+      "step": 17660
+    },
+    {
+      "epoch": 3.7225470058461054,
+      "grad_norm": 1.06181800365448,
+      "learning_rate": 0.00013904670720566698,
+      "loss": 1.6124,
+      "step": 17670
+    },
+    {
+      "epoch": 3.7246537104334543,
+      "grad_norm": 0.9750447273254395,
+      "learning_rate": 0.00013898575884788543,
+      "loss": 1.6872,
+      "step": 17680
+    },
+    {
+      "epoch": 3.7267604150208036,
+      "grad_norm": 0.9319090247154236,
+      "learning_rate": 0.0001389247934076501,
+      "loss": 1.6294,
+      "step": 17690
+    },
+    {
+      "epoch": 3.728867119608153,
+      "grad_norm": 0.9692747592926025,
+      "learning_rate": 0.0001388638109116744,
+      "loss": 1.6471,
+      "step": 17700
+    },
+    {
+      "epoch": 3.730973824195502,
+      "grad_norm": 0.9356905817985535,
+      "learning_rate": 0.00013880281138667905,
+      "loss": 1.649,
+      "step": 17710
+    },
+    {
+      "epoch": 3.7330805287828515,
+      "grad_norm": 0.9623681902885437,
+      "learning_rate": 0.0001387417948593923,
+      "loss": 1.7005,
+      "step": 17720
+    },
+    {
+      "epoch": 3.7351872333702008,
+      "grad_norm": 1.0994905233383179,
+      "learning_rate": 0.0001386807613565499,
+      "loss": 1.7227,
+      "step": 17730
+    },
+    {
+      "epoch": 3.73729393795755,
+      "grad_norm": 1.081142783164978,
+      "learning_rate": 0.000138619710904895,
+      "loss": 1.6773,
+      "step": 17740
+    },
+    {
+      "epoch": 3.739400642544899,
+      "grad_norm": 0.921646237373352,
+      "learning_rate": 0.00013855864353117816,
+      "loss": 1.6094,
+      "step": 17750
+    },
+    {
+      "epoch": 3.7415073471322486,
+      "grad_norm": 0.930452287197113,
+      "learning_rate": 0.00013849755926215735,
+      "loss": 1.6182,
+      "step": 17760
+    },
+    {
+      "epoch": 3.7436140517195975,
+      "grad_norm": 0.939112663269043,
+      "learning_rate": 0.00013843645812459802,
+      "loss": 1.6962,
+      "step": 17770
+    },
+    {
+      "epoch": 3.745720756306947,
+      "grad_norm": 0.9898761510848999,
+      "learning_rate": 0.0001383753401452729,
+      "loss": 1.6638,
+      "step": 17780
+    },
+    {
+      "epoch": 3.747827460894296,
+      "grad_norm": 0.9990235567092896,
+      "learning_rate": 0.00013831420535096223,
+      "loss": 1.6496,
+      "step": 17790
+    },
+    {
+      "epoch": 3.7499341654816454,
+      "grad_norm": 0.9664174318313599,
+      "learning_rate": 0.00013825305376845347,
+      "loss": 1.6869,
+      "step": 17800
+    },
+    {
+      "epoch": 3.7520408700689947,
+      "grad_norm": 0.9983745813369751,
+      "learning_rate": 0.0001381918854245415,
+      "loss": 1.6858,
+      "step": 17810
+    },
+    {
+      "epoch": 3.754147574656344,
+      "grad_norm": 0.9060636162757874,
+      "learning_rate": 0.00013813070034602863,
+      "loss": 1.6385,
+      "step": 17820
+    },
+    {
+      "epoch": 3.7562542792436933,
+      "grad_norm": 0.9503934979438782,
+      "learning_rate": 0.00013806949855972434,
+      "loss": 1.6733,
+      "step": 17830
+    },
+    {
+      "epoch": 3.758360983831042,
+      "grad_norm": 0.9625155925750732,
+      "learning_rate": 0.00013800828009244559,
+      "loss": 1.6871,
+      "step": 17840
+    },
+    {
+      "epoch": 3.7604676884183914,
+      "grad_norm": 0.985414981842041,
+      "learning_rate": 0.00013794704497101655,
+      "loss": 1.6641,
+      "step": 17850
+    },
+    {
+      "epoch": 3.7625743930057407,
+      "grad_norm": 0.9473032355308533,
+      "learning_rate": 0.00013788579322226868,
+      "loss": 1.6459,
+      "step": 17860
+    },
+    {
+      "epoch": 3.76468109759309,
+      "grad_norm": 1.0287021398544312,
+      "learning_rate": 0.0001378245248730408,
+      "loss": 1.6515,
+      "step": 17870
+    },
+    {
+      "epoch": 3.7667878021804393,
+      "grad_norm": 0.9983483552932739,
+      "learning_rate": 0.00013776323995017898,
+      "loss": 1.6521,
+      "step": 17880
+    },
+    {
+      "epoch": 3.7688945067677886,
+      "grad_norm": 1.0303962230682373,
+      "learning_rate": 0.00013770193848053648,
+      "loss": 1.7067,
+      "step": 17890
+    },
+    {
+      "epoch": 3.771001211355138,
+      "grad_norm": 1.0121102333068848,
+      "learning_rate": 0.00013764062049097389,
+      "loss": 1.6628,
+      "step": 17900
+    },
+    {
+      "epoch": 3.7731079159424867,
+      "grad_norm": 0.9781060218811035,
+      "learning_rate": 0.000137579286008359,
+      "loss": 1.5913,
+      "step": 17910
+    },
+    {
+      "epoch": 3.7752146205298365,
+      "grad_norm": 1.03104829788208,
+      "learning_rate": 0.0001375179350595669,
+      "loss": 1.675,
+      "step": 17920
+    },
+    {
+      "epoch": 3.7773213251171853,
+      "grad_norm": 0.9380589127540588,
+      "learning_rate": 0.00013745656767147978,
+      "loss": 1.6743,
+      "step": 17930
+    },
+    {
+      "epoch": 3.7794280297045346,
+      "grad_norm": 0.9748852252960205,
+      "learning_rate": 0.00013739518387098705,
+      "loss": 1.6895,
+      "step": 17940
+    },
+    {
+      "epoch": 3.781534734291884,
+      "grad_norm": 0.973847508430481,
+      "learning_rate": 0.00013733378368498543,
+      "loss": 1.6514,
+      "step": 17950
+    },
+    {
+      "epoch": 3.783641438879233,
+      "grad_norm": 1.1825324296951294,
+      "learning_rate": 0.00013727236714037872,
+      "loss": 1.6661,
+      "step": 17960
+    },
+    {
+      "epoch": 3.7857481434665825,
+      "grad_norm": 0.9843605756759644,
+      "learning_rate": 0.0001372109342640779,
+      "loss": 1.6192,
+      "step": 17970
+    },
+    {
+      "epoch": 3.787854848053932,
+      "grad_norm": 1.0862230062484741,
+      "learning_rate": 0.0001371494850830011,
+      "loss": 1.6936,
+      "step": 17980
+    },
+    {
+      "epoch": 3.789961552641281,
+      "grad_norm": 0.9965257048606873,
+      "learning_rate": 0.0001370880196240736,
+      "loss": 1.6191,
+      "step": 17990
+    },
+    {
+      "epoch": 3.79206825722863,
+      "grad_norm": 1.045751690864563,
+      "learning_rate": 0.0001370265379142279,
+      "loss": 1.746,
+      "step": 18000
+    },
+    {
+      "epoch": 3.7941749618159792,
+      "grad_norm": 0.9319239854812622,
+      "learning_rate": 0.00013696503998040342,
+      "loss": 1.7011,
+      "step": 18010
+    },
+    {
+      "epoch": 3.7962816664033285,
+      "grad_norm": 1.0652892589569092,
+      "learning_rate": 0.0001369035258495469,
+      "loss": 1.6326,
+      "step": 18020
+    },
+    {
+      "epoch": 3.798388370990678,
+      "grad_norm": 0.9452232718467712,
+      "learning_rate": 0.00013684199554861207,
+      "loss": 1.7333,
+      "step": 18030
+    },
+    {
+      "epoch": 3.800495075578027,
+      "grad_norm": 0.9951366186141968,
+      "learning_rate": 0.00013678044910455975,
+      "loss": 1.7036,
+      "step": 18040
+    },
+    {
+      "epoch": 3.8026017801653764,
+      "grad_norm": 1.0448187589645386,
+      "learning_rate": 0.0001367188865443578,
+      "loss": 1.6969,
+      "step": 18050
+    },
+    {
+      "epoch": 3.8047084847527257,
+      "grad_norm": 0.952127993106842,
+      "learning_rate": 0.0001366573078949813,
+      "loss": 1.6977,
+      "step": 18060
+    },
+    {
+      "epoch": 3.8068151893400746,
+      "grad_norm": 0.9487131237983704,
+      "learning_rate": 0.0001365957131834122,
+      "loss": 1.6284,
+      "step": 18070
+    },
+    {
+      "epoch": 3.808921893927424,
+      "grad_norm": 1.0855556726455688,
+      "learning_rate": 0.00013653410243663952,
+      "loss": 1.7035,
+      "step": 18080
+    },
+    {
+      "epoch": 3.811028598514773,
+      "grad_norm": 0.9785754084587097,
+      "learning_rate": 0.00013647247568165938,
+      "loss": 1.67,
+      "step": 18090
+    },
+    {
+      "epoch": 3.8131353031021225,
+      "grad_norm": 0.981380045413971,
+      "learning_rate": 0.0001364108329454749,
+      "loss": 1.7031,
+      "step": 18100
+    },
+    {
+      "epoch": 3.8152420076894717,
+      "grad_norm": 0.9764424562454224,
+      "learning_rate": 0.00013634917425509616,
+      "loss": 1.7354,
+      "step": 18110
+    },
+    {
+      "epoch": 3.817348712276821,
+      "grad_norm": 0.8839856386184692,
+      "learning_rate": 0.00013628749963754026,
+      "loss": 1.7048,
+      "step": 18120
+    },
+    {
+      "epoch": 3.8194554168641703,
+      "grad_norm": 1.0101412534713745,
+      "learning_rate": 0.0001362258091198312,
+      "loss": 1.6833,
+      "step": 18130
+    },
+    {
+      "epoch": 3.821562121451519,
+      "grad_norm": 0.9509012699127197,
+      "learning_rate": 0.00013616410272900014,
+      "loss": 1.6583,
+      "step": 18140
+    },
+    {
+      "epoch": 3.823668826038869,
+      "grad_norm": 0.9501906037330627,
+      "learning_rate": 0.00013610238049208495,
+      "loss": 1.6526,
+      "step": 18150
+    },
+    {
+      "epoch": 3.825775530626218,
+      "grad_norm": 0.9449414610862732,
+      "learning_rate": 0.0001360406424361306,
+      "loss": 1.6726,
+      "step": 18160
+    },
+    {
+      "epoch": 3.827882235213567,
+      "grad_norm": 0.9675168991088867,
+      "learning_rate": 0.00013597888858818898,
+      "loss": 1.6955,
+      "step": 18170
+    },
+    {
+      "epoch": 3.8299889398009164,
+      "grad_norm": 0.9368403553962708,
+      "learning_rate": 0.0001359171189753189,
+      "loss": 1.665,
+      "step": 18180
+    },
+    {
+      "epoch": 3.8320956443882657,
+      "grad_norm": 1.0501976013183594,
+      "learning_rate": 0.00013585533362458599,
+      "loss": 1.6632,
+      "step": 18190
+    },
+    {
+      "epoch": 3.834202348975615,
+      "grad_norm": 0.9386237859725952,
+      "learning_rate": 0.00013579353256306287,
+      "loss": 1.6612,
+      "step": 18200
+    },
+    {
+      "epoch": 3.8363090535629643,
+      "grad_norm": 1.0792053937911987,
+      "learning_rate": 0.00013573171581782897,
+      "loss": 1.6528,
+      "step": 18210
+    },
+    {
+      "epoch": 3.8384157581503136,
+      "grad_norm": 0.9612410664558411,
+      "learning_rate": 0.00013566988341597068,
+      "loss": 1.6405,
+      "step": 18220
+    },
+    {
+      "epoch": 3.8405224627376624,
+      "grad_norm": 1.0245602130889893,
+      "learning_rate": 0.00013560803538458123,
+      "loss": 1.6902,
+      "step": 18230
+    },
+    {
+      "epoch": 3.8426291673250117,
+      "grad_norm": 0.8941143751144409,
+      "learning_rate": 0.00013554617175076062,
+      "loss": 1.629,
+      "step": 18240
+    },
+    {
+      "epoch": 3.844735871912361,
+      "grad_norm": 0.9566497206687927,
+      "learning_rate": 0.00013548429254161575,
+      "loss": 1.5743,
+      "step": 18250
+    },
+    {
+      "epoch": 3.8468425764997103,
+      "grad_norm": 1.0370187759399414,
+      "learning_rate": 0.00013542239778426034,
+      "loss": 1.6781,
+      "step": 18260
+    },
+    {
+      "epoch": 3.8489492810870596,
+      "grad_norm": 0.8610029220581055,
+      "learning_rate": 0.00013536048750581494,
+      "loss": 1.681,
+      "step": 18270
+    },
+    {
+      "epoch": 3.851055985674409,
+      "grad_norm": 0.8663182854652405,
+      "learning_rate": 0.00013529856173340684,
+      "loss": 1.6648,
+      "step": 18280
+    },
+    {
+      "epoch": 3.853162690261758,
+      "grad_norm": 0.9131945371627808,
+      "learning_rate": 0.00013523662049417015,
+      "loss": 1.6575,
+      "step": 18290
+    },
+    {
+      "epoch": 3.855269394849107,
+      "grad_norm": 0.9651349782943726,
+      "learning_rate": 0.0001351746638152458,
+      "loss": 1.6545,
+      "step": 18300
+    },
+    {
+      "epoch": 3.8573760994364568,
+      "grad_norm": 0.8891710042953491,
+      "learning_rate": 0.00013511269172378147,
+      "loss": 1.643,
+      "step": 18310
+    },
+    {
+      "epoch": 3.8594828040238056,
+      "grad_norm": 0.9134824872016907,
+      "learning_rate": 0.00013505070424693153,
+      "loss": 1.6766,
+      "step": 18320
+    },
+    {
+      "epoch": 3.861589508611155,
+      "grad_norm": 1.019539475440979,
+      "learning_rate": 0.00013498870141185712,
+      "loss": 1.6787,
+      "step": 18330
+    },
+    {
+      "epoch": 3.863696213198504,
+      "grad_norm": 1.0163251161575317,
+      "learning_rate": 0.00013492668324572614,
+      "loss": 1.7008,
+      "step": 18340
+    },
+    {
+      "epoch": 3.8658029177858535,
+      "grad_norm": 0.9731863141059875,
+      "learning_rate": 0.00013486464977571324,
+      "loss": 1.6363,
+      "step": 18350
+    },
+    {
+      "epoch": 3.867909622373203,
+      "grad_norm": 0.9799426198005676,
+      "learning_rate": 0.00013480260102899966,
+      "loss": 1.6885,
+      "step": 18360
+    },
+    {
+      "epoch": 3.870016326960552,
+      "grad_norm": 1.055290699005127,
+      "learning_rate": 0.00013474053703277342,
+      "loss": 1.7328,
+      "step": 18370
+    },
+    {
+      "epoch": 3.8721230315479014,
+      "grad_norm": 0.930711030960083,
+      "learning_rate": 0.00013467845781422924,
+      "loss": 1.6725,
+      "step": 18380
+    },
+    {
+      "epoch": 3.8742297361352502,
+      "grad_norm": 0.9686673283576965,
+      "learning_rate": 0.00013461636340056843,
+      "loss": 1.6906,
+      "step": 18390
+    },
+    {
+      "epoch": 3.8763364407225995,
+      "grad_norm": 0.9889995455741882,
+      "learning_rate": 0.000134554253818999,
+      "loss": 1.6348,
+      "step": 18400
+    },
+    {
+      "epoch": 3.878443145309949,
+      "grad_norm": 0.964756965637207,
+      "learning_rate": 0.00013449212909673563,
+      "loss": 1.6999,
+      "step": 18410
+    },
+    {
+      "epoch": 3.880549849897298,
+      "grad_norm": 0.945366621017456,
+      "learning_rate": 0.0001344299892609996,
+      "loss": 1.6821,
+      "step": 18420
+    },
+    {
+      "epoch": 3.8826565544846474,
+      "grad_norm": 0.9812168478965759,
+      "learning_rate": 0.0001343678343390188,
+      "loss": 1.6809,
+      "step": 18430
+    },
+    {
+      "epoch": 3.8847632590719967,
+      "grad_norm": 0.9222221970558167,
+      "learning_rate": 0.00013430566435802783,
+      "loss": 1.6884,
+      "step": 18440
+    },
+    {
+      "epoch": 3.886869963659346,
+      "grad_norm": 1.0188971757888794,
+      "learning_rate": 0.00013424347934526772,
+      "loss": 1.6886,
+      "step": 18450
+    },
+    {
+      "epoch": 3.888976668246695,
+      "grad_norm": 0.8854445815086365,
+      "learning_rate": 0.00013418127932798623,
+      "loss": 1.6784,
+      "step": 18460
+    },
+    {
+      "epoch": 3.8910833728340446,
+      "grad_norm": 0.9683605432510376,
+      "learning_rate": 0.00013411906433343765,
+      "loss": 1.6627,
+      "step": 18470
+    },
+    {
+      "epoch": 3.8931900774213934,
+      "grad_norm": 0.9435099363327026,
+      "learning_rate": 0.00013405683438888282,
+      "loss": 1.7143,
+      "step": 18480
+    },
+    {
+      "epoch": 3.8952967820087427,
+      "grad_norm": 0.8995447158813477,
+      "learning_rate": 0.0001339945895215891,
+      "loss": 1.6652,
+      "step": 18490
+    },
+    {
+      "epoch": 3.897403486596092,
+      "grad_norm": 0.9004771113395691,
+      "learning_rate": 0.00013393232975883042,
+      "loss": 1.6349,
+      "step": 18500
+    },
+    {
+      "epoch": 3.8995101911834413,
+      "grad_norm": 1.0811883211135864,
+      "learning_rate": 0.00013387005512788733,
+      "loss": 1.6849,
+      "step": 18510
+    },
+    {
+      "epoch": 3.9016168957707906,
+      "grad_norm": 0.9638143181800842,
+      "learning_rate": 0.00013380776565604676,
+      "loss": 1.6405,
+      "step": 18520
+    },
+    {
+      "epoch": 3.90372360035814,
+      "grad_norm": 0.9664076566696167,
+      "learning_rate": 0.00013374546137060212,
+      "loss": 1.6693,
+      "step": 18530
+    },
+    {
+      "epoch": 3.905830304945489,
+      "grad_norm": 1.1116750240325928,
+      "learning_rate": 0.00013368314229885347,
+      "loss": 1.6814,
+      "step": 18540
+    },
+    {
+      "epoch": 3.907937009532838,
+      "grad_norm": 1.125832438468933,
+      "learning_rate": 0.00013362080846810725,
+      "loss": 1.6283,
+      "step": 18550
+    },
+    {
+      "epoch": 3.9100437141201874,
+      "grad_norm": 0.9035454988479614,
+      "learning_rate": 0.00013355845990567635,
+      "loss": 1.6729,
+      "step": 18560
+    },
+    {
+      "epoch": 3.9121504187075367,
+      "grad_norm": 1.0243180990219116,
+      "learning_rate": 0.00013349609663888015,
+      "loss": 1.6674,
+      "step": 18570
+    },
+    {
+      "epoch": 3.914257123294886,
+      "grad_norm": 1.1027157306671143,
+      "learning_rate": 0.00013343371869504444,
+      "loss": 1.6408,
+      "step": 18580
+    },
+    {
+      "epoch": 3.9163638278822352,
+      "grad_norm": 0.9812912344932556,
+      "learning_rate": 0.00013337132610150148,
+      "loss": 1.7142,
+      "step": 18590
+    },
+    {
+      "epoch": 3.9184705324695845,
+      "grad_norm": 0.9783474802970886,
+      "learning_rate": 0.00013330891888559002,
+      "loss": 1.6257,
+      "step": 18600
+    },
+    {
+      "epoch": 3.920577237056934,
+      "grad_norm": 1.0197913646697998,
+      "learning_rate": 0.000133246497074655,
+      "loss": 1.6701,
+      "step": 18610
+    },
+    {
+      "epoch": 3.9226839416442827,
+      "grad_norm": 1.0612571239471436,
+      "learning_rate": 0.00013318406069604794,
+      "loss": 1.6635,
+      "step": 18620
+    },
+    {
+      "epoch": 3.9247906462316324,
+      "grad_norm": 0.922271728515625,
+      "learning_rate": 0.00013312160977712668,
+      "loss": 1.6573,
+      "step": 18630
+    },
+    {
+      "epoch": 3.9268973508189813,
+      "grad_norm": 0.9498322010040283,
+      "learning_rate": 0.00013305914434525552,
+      "loss": 1.6633,
+      "step": 18640
+    },
+    {
+      "epoch": 3.9290040554063306,
+      "grad_norm": 0.9405461549758911,
+      "learning_rate": 0.00013299666442780493,
+      "loss": 1.6647,
+      "step": 18650
+    },
+    {
+      "epoch": 3.93111075999368,
+      "grad_norm": 0.9539029002189636,
+      "learning_rate": 0.00013293417005215188,
+      "loss": 1.6587,
+      "step": 18660
+    },
+    {
+      "epoch": 3.933217464581029,
+      "grad_norm": 0.9618557691574097,
+      "learning_rate": 0.00013287166124567964,
+      "loss": 1.7109,
+      "step": 18670
+    },
+    {
+      "epoch": 3.9353241691683785,
+      "grad_norm": 0.9625663161277771,
+      "learning_rate": 0.0001328091380357778,
+      "loss": 1.6915,
+      "step": 18680
+    },
+    {
+      "epoch": 3.9374308737557278,
+      "grad_norm": 0.9376285672187805,
+      "learning_rate": 0.00013274660044984224,
+      "loss": 1.7005,
+      "step": 18690
+    },
+    {
+      "epoch": 3.939537578343077,
+      "grad_norm": 0.8864640593528748,
+      "learning_rate": 0.00013268404851527518,
+      "loss": 1.6187,
+      "step": 18700
+    },
+    {
+      "epoch": 3.941644282930426,
+      "grad_norm": 1.2145129442214966,
+      "learning_rate": 0.00013262148225948506,
+      "loss": 1.6385,
+      "step": 18710
+    },
+    {
+      "epoch": 3.943750987517775,
+      "grad_norm": 0.9888447523117065,
+      "learning_rate": 0.0001325589017098867,
+      "loss": 1.64,
+      "step": 18720
+    },
+    {
+      "epoch": 3.9458576921051245,
+      "grad_norm": 1.1795809268951416,
+      "learning_rate": 0.0001324963068939011,
+      "loss": 1.7333,
+      "step": 18730
+    },
+    {
+      "epoch": 3.947964396692474,
+      "grad_norm": 1.0691869258880615,
+      "learning_rate": 0.00013243369783895548,
+      "loss": 1.6156,
+      "step": 18740
+    },
+    {
+      "epoch": 3.950071101279823,
+      "grad_norm": 0.9917706251144409,
+      "learning_rate": 0.0001323710745724834,
+      "loss": 1.5666,
+      "step": 18750
+    },
+    {
+      "epoch": 3.9521778058671724,
+      "grad_norm": 0.9891237020492554,
+      "learning_rate": 0.00013230843712192463,
+      "loss": 1.7003,
+      "step": 18760
+    },
+    {
+      "epoch": 3.9542845104545217,
+      "grad_norm": 0.9454847574234009,
+      "learning_rate": 0.0001322457855147251,
+      "loss": 1.7411,
+      "step": 18770
+    },
+    {
+      "epoch": 3.9563912150418705,
+      "grad_norm": 0.9509596228599548,
+      "learning_rate": 0.00013218311977833687,
+      "loss": 1.6653,
+      "step": 18780
+    },
+    {
+      "epoch": 3.9584979196292203,
+      "grad_norm": 0.9767935276031494,
+      "learning_rate": 0.00013212043994021845,
+      "loss": 1.7047,
+      "step": 18790
+    },
+    {
+      "epoch": 3.960604624216569,
+      "grad_norm": 0.972263514995575,
+      "learning_rate": 0.00013205774602783428,
+      "loss": 1.6834,
+      "step": 18800
+    },
+    {
+      "epoch": 3.9627113288039184,
+      "grad_norm": 0.9423323273658752,
+      "learning_rate": 0.00013199503806865504,
+      "loss": 1.6837,
+      "step": 18810
+    },
+    {
+      "epoch": 3.9648180333912677,
+      "grad_norm": 0.9944238066673279,
+      "learning_rate": 0.00013193231609015763,
+      "loss": 1.6581,
+      "step": 18820
+    },
+    {
+      "epoch": 3.966924737978617,
+      "grad_norm": 1.011017918586731,
+      "learning_rate": 0.00013186958011982502,
+      "loss": 1.657,
+      "step": 18830
+    },
+    {
+      "epoch": 3.9690314425659663,
+      "grad_norm": 0.9588932991027832,
+      "learning_rate": 0.0001318068301851463,
+      "loss": 1.6204,
+      "step": 18840
+    },
+    {
+      "epoch": 3.9711381471533156,
+      "grad_norm": 0.9741507768630981,
+      "learning_rate": 0.00013174406631361675,
+      "loss": 1.663,
+      "step": 18850
+    },
+    {
+      "epoch": 3.973244851740665,
+      "grad_norm": 0.897010326385498,
+      "learning_rate": 0.00013168128853273772,
+      "loss": 1.6428,
+      "step": 18860
+    },
+    {
+      "epoch": 3.9753515563280137,
+      "grad_norm": 0.9145323634147644,
+      "learning_rate": 0.00013161849687001666,
+      "loss": 1.6265,
+      "step": 18870
+    },
+    {
+      "epoch": 3.977458260915363,
+      "grad_norm": 1.093653678894043,
+      "learning_rate": 0.00013155569135296703,
+      "loss": 1.6954,
+      "step": 18880
+    },
+    {
+      "epoch": 3.9795649655027123,
+      "grad_norm": 0.9764483571052551,
+      "learning_rate": 0.0001314928720091085,
+      "loss": 1.6673,
+      "step": 18890
+    },
+    {
+      "epoch": 3.9816716700900616,
+      "grad_norm": 0.9174908995628357,
+      "learning_rate": 0.00013143003886596669,
+      "loss": 1.688,
+      "step": 18900
+    },
+    {
+      "epoch": 3.983778374677411,
+      "grad_norm": 0.9659301042556763,
+      "learning_rate": 0.00013136719195107335,
+      "loss": 1.6748,
+      "step": 18910
+    },
+    {
+      "epoch": 3.98588507926476,
+      "grad_norm": 0.9585952162742615,
+      "learning_rate": 0.00013130433129196614,
+      "loss": 1.6963,
+      "step": 18920
+    },
+    {
+      "epoch": 3.9879917838521095,
+      "grad_norm": 0.9116827249526978,
+      "learning_rate": 0.00013124145691618884,
+      "loss": 1.6635,
+      "step": 18930
+    },
+    {
+      "epoch": 3.9900984884394584,
+      "grad_norm": 1.0654014348983765,
+      "learning_rate": 0.00013117856885129126,
+      "loss": 1.6963,
+      "step": 18940
+    },
+    {
+      "epoch": 3.9922051930268077,
+      "grad_norm": 1.0164047479629517,
+      "learning_rate": 0.00013111566712482913,
+      "loss": 1.7179,
+      "step": 18950
+    },
+    {
+      "epoch": 3.994311897614157,
+      "grad_norm": 0.9828543066978455,
+      "learning_rate": 0.0001310527517643642,
+      "loss": 1.6681,
+      "step": 18960
+    },
+    {
+      "epoch": 3.9964186022015062,
+      "grad_norm": 0.9909005761146545,
+      "learning_rate": 0.00013098982279746422,
+      "loss": 1.6163,
+      "step": 18970
+    },
+    {
+      "epoch": 3.9985253067888555,
+      "grad_norm": 0.9685956239700317,
+      "learning_rate": 0.00013092688025170284,
+      "loss": 1.6844,
+      "step": 18980
+    },
+    {
+      "epoch": 4.000632011376204,
+      "grad_norm": 0.9608546495437622,
+      "learning_rate": 0.00013086392415465972,
+      "loss": 1.6726,
+      "step": 18990
+    },
+    {
+      "epoch": 4.002738715963554,
+      "grad_norm": 0.9137300848960876,
+      "learning_rate": 0.0001308009545339205,
+      "loss": 1.6214,
+      "step": 19000
+    },
+    {
+      "epoch": 4.004845420550903,
+      "grad_norm": 1.1166579723358154,
+      "learning_rate": 0.00013073797141707657,
+      "loss": 1.5999,
+      "step": 19010
+    },
+    {
+      "epoch": 4.006952125138253,
+      "grad_norm": 0.9817916750907898,
+      "learning_rate": 0.00013067497483172538,
+      "loss": 1.6498,
+      "step": 19020
+    },
+    {
+      "epoch": 4.009058829725602,
+      "grad_norm": 0.9361011981964111,
+      "learning_rate": 0.0001306119648054703,
+      "loss": 1.629,
+      "step": 19030
+    },
+    {
+      "epoch": 4.011165534312951,
+      "grad_norm": 0.9715505242347717,
+      "learning_rate": 0.00013054894136592052,
+      "loss": 1.6419,
+      "step": 19040
+    },
+    {
+      "epoch": 4.0132722389003,
+      "grad_norm": 0.9041184782981873,
+      "learning_rate": 0.00013048590454069108,
+      "loss": 1.5842,
+      "step": 19050
+    },
+    {
+      "epoch": 4.015378943487649,
+      "grad_norm": 0.9362772107124329,
+      "learning_rate": 0.0001304228543574029,
+      "loss": 1.575,
+      "step": 19060
+    },
+    {
+      "epoch": 4.017485648074999,
+      "grad_norm": 1.1021233797073364,
+      "learning_rate": 0.00013035979084368292,
+      "loss": 1.6569,
+      "step": 19070
+    },
+    {
+      "epoch": 4.019592352662348,
+      "grad_norm": 0.9974276423454285,
+      "learning_rate": 0.00013029671402716366,
+      "loss": 1.6237,
+      "step": 19080
+    },
+    {
+      "epoch": 4.021699057249697,
+      "grad_norm": 0.9724931120872498,
+      "learning_rate": 0.00013023362393548363,
+      "loss": 1.68,
+      "step": 19090
+    },
+    {
+      "epoch": 4.023805761837046,
+      "grad_norm": 0.990185558795929,
+      "learning_rate": 0.0001301705205962871,
+      "loss": 1.6594,
+      "step": 19100
+    },
+    {
+      "epoch": 4.025912466424396,
+      "grad_norm": 1.051291584968567,
+      "learning_rate": 0.0001301074040372242,
+      "loss": 1.5814,
+      "step": 19110
+    },
+    {
+      "epoch": 4.028019171011745,
+      "grad_norm": 0.9896820783615112,
+      "learning_rate": 0.0001300442742859508,
+      "loss": 1.6006,
+      "step": 19120
+    },
+    {
+      "epoch": 4.0301258755990945,
+      "grad_norm": 1.0852347612380981,
+      "learning_rate": 0.00012998113137012855,
+      "loss": 1.6283,
+      "step": 19130
+    },
+    {
+      "epoch": 4.032232580186443,
+      "grad_norm": 1.0564210414886475,
+      "learning_rate": 0.00012991797531742492,
+      "loss": 1.636,
+      "step": 19140
+    },
+    {
+      "epoch": 4.034339284773792,
+      "grad_norm": 0.9386522769927979,
+      "learning_rate": 0.00012985480615551305,
+      "loss": 1.6024,
+      "step": 19150
+    },
+    {
+      "epoch": 4.036445989361142,
+      "grad_norm": 1.0262477397918701,
+      "learning_rate": 0.00012979162391207194,
+      "loss": 1.6168,
+      "step": 19160
+    },
+    {
+      "epoch": 4.038552693948491,
+      "grad_norm": 0.9269583225250244,
+      "learning_rate": 0.00012972842861478618,
+      "loss": 1.6511,
+      "step": 19170
+    },
+    {
+      "epoch": 4.0406593985358406,
+      "grad_norm": 1.0529104471206665,
+      "learning_rate": 0.00012966522029134623,
+      "loss": 1.6147,
+      "step": 19180
+    },
+    {
+      "epoch": 4.042766103123189,
+      "grad_norm": 1.0251187086105347,
+      "learning_rate": 0.00012960199896944815,
+      "loss": 1.5781,
+      "step": 19190
+    },
+    {
+      "epoch": 4.044872807710539,
+      "grad_norm": 1.0744245052337646,
+      "learning_rate": 0.00012953876467679373,
+      "loss": 1.6442,
+      "step": 19200
+    },
+    {
+      "epoch": 4.046979512297888,
+      "grad_norm": 0.9448406100273132,
+      "learning_rate": 0.00012947551744109043,
+      "loss": 1.6216,
+      "step": 19210
+    },
+    {
+      "epoch": 4.049086216885237,
+      "grad_norm": 0.9437319040298462,
+      "learning_rate": 0.00012941225729005143,
+      "loss": 1.6557,
+      "step": 19220
+    },
+    {
+      "epoch": 4.051192921472587,
+      "grad_norm": 1.092026710510254,
+      "learning_rate": 0.0001293489842513955,
+      "loss": 1.6463,
+      "step": 19230
+    },
+    {
+      "epoch": 4.053299626059935,
+      "grad_norm": 1.0943164825439453,
+      "learning_rate": 0.00012928569835284713,
+      "loss": 1.5869,
+      "step": 19240
+    },
+    {
+      "epoch": 4.055406330647285,
+      "grad_norm": 0.9766450524330139,
+      "learning_rate": 0.00012922239962213637,
+      "loss": 1.6009,
+      "step": 19250
+    },
+    {
+      "epoch": 4.057513035234634,
+      "grad_norm": 0.9823459982872009,
+      "learning_rate": 0.00012915908808699893,
+      "loss": 1.6072,
+      "step": 19260
+    },
+    {
+      "epoch": 4.059619739821984,
+      "grad_norm": 0.9854733347892761,
+      "learning_rate": 0.00012909576377517616,
+      "loss": 1.6373,
+      "step": 19270
+    },
+    {
+      "epoch": 4.061726444409333,
+      "grad_norm": 1.1235730648040771,
+      "learning_rate": 0.00012903242671441492,
+      "loss": 1.6886,
+      "step": 19280
+    },
+    {
+      "epoch": 4.063833148996682,
+      "grad_norm": 1.0431162118911743,
+      "learning_rate": 0.0001289690769324678,
+      "loss": 1.6286,
+      "step": 19290
+    },
+    {
+      "epoch": 4.065939853584031,
+      "grad_norm": 1.0117398500442505,
+      "learning_rate": 0.00012890571445709278,
+      "loss": 1.6423,
+      "step": 19300
+    },
+    {
+      "epoch": 4.06804655817138,
+      "grad_norm": 1.196179747581482,
+      "learning_rate": 0.00012884233931605358,
+      "loss": 1.64,
+      "step": 19310
+    },
+    {
+      "epoch": 4.07015326275873,
+      "grad_norm": 1.04131019115448,
+      "learning_rate": 0.00012877895153711935,
+      "loss": 1.6285,
+      "step": 19320
+    },
+    {
+      "epoch": 4.072259967346079,
+      "grad_norm": 0.9589956402778625,
+      "learning_rate": 0.00012871555114806483,
+      "loss": 1.6349,
+      "step": 19330
+    },
+    {
+      "epoch": 4.074366671933428,
+      "grad_norm": 0.9664011597633362,
+      "learning_rate": 0.00012865213817667023,
+      "loss": 1.585,
+      "step": 19340
+    },
+    {
+      "epoch": 4.076473376520777,
+      "grad_norm": 0.9522480964660645,
+      "learning_rate": 0.0001285887126507214,
+      "loss": 1.641,
+      "step": 19350
+    },
+    {
+      "epoch": 4.078580081108127,
+      "grad_norm": 1.0472582578659058,
+      "learning_rate": 0.00012852527459800953,
+      "loss": 1.676,
+      "step": 19360
+    },
+    {
+      "epoch": 4.080686785695476,
+      "grad_norm": 0.9799109697341919,
+      "learning_rate": 0.00012846182404633143,
+      "loss": 1.618,
+      "step": 19370
+    },
+    {
+      "epoch": 4.082793490282825,
+      "grad_norm": 0.9914186000823975,
+      "learning_rate": 0.00012839836102348926,
+      "loss": 1.6168,
+      "step": 19380
+    },
+    {
+      "epoch": 4.084900194870174,
+      "grad_norm": 0.9672318696975708,
+      "learning_rate": 0.0001283348855572908,
+      "loss": 1.6168,
+      "step": 19390
+    },
+    {
+      "epoch": 4.087006899457523,
+      "grad_norm": 1.0729410648345947,
+      "learning_rate": 0.00012827139767554915,
+      "loss": 1.6477,
+      "step": 19400
+    },
+    {
+      "epoch": 4.089113604044873,
+      "grad_norm": 1.1007764339447021,
+      "learning_rate": 0.00012820789740608293,
+      "loss": 1.5919,
+      "step": 19410
+    },
+    {
+      "epoch": 4.091220308632222,
+      "grad_norm": 0.9777454733848572,
+      "learning_rate": 0.0001281443847767161,
+      "loss": 1.5997,
+      "step": 19420
+    },
+    {
+      "epoch": 4.093327013219572,
+      "grad_norm": 1.0719486474990845,
+      "learning_rate": 0.00012808085981527815,
+      "loss": 1.6411,
+      "step": 19430
+    },
+    {
+      "epoch": 4.0954337178069204,
+      "grad_norm": 1.0390957593917847,
+      "learning_rate": 0.00012801732254960388,
+      "loss": 1.6416,
+      "step": 19440
+    },
+    {
+      "epoch": 4.09754042239427,
+      "grad_norm": 1.065137505531311,
+      "learning_rate": 0.00012795377300753357,
+      "loss": 1.7202,
+      "step": 19450
+    },
+    {
+      "epoch": 4.099647126981619,
+      "grad_norm": 1.0135070085525513,
+      "learning_rate": 0.00012789021121691274,
+      "loss": 1.6056,
+      "step": 19460
+    },
+    {
+      "epoch": 4.101753831568968,
+      "grad_norm": 0.9829651117324829,
+      "learning_rate": 0.00012782663720559246,
+      "loss": 1.5973,
+      "step": 19470
+    },
+    {
+      "epoch": 4.103860536156318,
+      "grad_norm": 1.0231988430023193,
+      "learning_rate": 0.00012776305100142897,
+      "loss": 1.6736,
+      "step": 19480
+    },
+    {
+      "epoch": 4.1059672407436665,
+      "grad_norm": 0.9512834548950195,
+      "learning_rate": 0.00012769945263228403,
+      "loss": 1.6567,
+      "step": 19490
+    },
+    {
+      "epoch": 4.108073945331016,
+      "grad_norm": 1.0609872341156006,
+      "learning_rate": 0.00012763584212602453,
+      "loss": 1.6434,
+      "step": 19500
+    },
+    {
+      "epoch": 4.110180649918365,
+      "grad_norm": 1.026944637298584,
+      "learning_rate": 0.0001275722195105229,
+      "loss": 1.6767,
+      "step": 19510
+    },
+    {
+      "epoch": 4.112287354505715,
+      "grad_norm": 1.0436118841171265,
+      "learning_rate": 0.00012750858481365673,
+      "loss": 1.6491,
+      "step": 19520
+    },
+    {
+      "epoch": 4.114394059093064,
+      "grad_norm": 1.051046371459961,
+      "learning_rate": 0.0001274449380633089,
+      "loss": 1.6575,
+      "step": 19530
+    },
+    {
+      "epoch": 4.1165007636804125,
+      "grad_norm": 1.0463584661483765,
+      "learning_rate": 0.00012738127928736765,
+      "loss": 1.6379,
+      "step": 19540
+    },
+    {
+      "epoch": 4.118607468267762,
+      "grad_norm": 1.15203058719635,
+      "learning_rate": 0.00012731760851372644,
+      "loss": 1.6481,
+      "step": 19550
+    },
+    {
+      "epoch": 4.120714172855111,
+      "grad_norm": 1.0916799306869507,
+      "learning_rate": 0.00012725392577028402,
+      "loss": 1.6209,
+      "step": 19560
+    },
+    {
+      "epoch": 4.122820877442461,
+      "grad_norm": 1.14989173412323,
+      "learning_rate": 0.00012719023108494435,
+      "loss": 1.6247,
+      "step": 19570
+    },
+    {
+      "epoch": 4.12492758202981,
+      "grad_norm": 0.949806272983551,
+      "learning_rate": 0.00012712652448561656,
+      "loss": 1.6374,
+      "step": 19580
+    },
+    {
+      "epoch": 4.127034286617159,
+      "grad_norm": 1.0403326749801636,
+      "learning_rate": 0.00012706280600021522,
+      "loss": 1.6005,
+      "step": 19590
+    },
+    {
+      "epoch": 4.129140991204508,
+      "grad_norm": 1.0589529275894165,
+      "learning_rate": 0.00012699907565665982,
+      "loss": 1.6739,
+      "step": 19600
+    },
+    {
+      "epoch": 4.131247695791858,
+      "grad_norm": 1.0490063428878784,
+      "learning_rate": 0.0001269353334828753,
+      "loss": 1.6425,
+      "step": 19610
+    },
+    {
+      "epoch": 4.133354400379207,
+      "grad_norm": 1.061720371246338,
+      "learning_rate": 0.0001268715795067916,
+      "loss": 1.6322,
+      "step": 19620
+    },
+    {
+      "epoch": 4.135461104966556,
+      "grad_norm": 0.9578418731689453,
+      "learning_rate": 0.0001268078137563439,
+      "loss": 1.6109,
+      "step": 19630
+    },
+    {
+      "epoch": 4.1375678095539055,
+      "grad_norm": 0.9411417245864868,
+      "learning_rate": 0.0001267440362594726,
+      "loss": 1.5956,
+      "step": 19640
+    },
+    {
+      "epoch": 4.139674514141254,
+      "grad_norm": 1.0242791175842285,
+      "learning_rate": 0.00012668024704412317,
+      "loss": 1.6445,
+      "step": 19650
+    },
+    {
+      "epoch": 4.141781218728604,
+      "grad_norm": 1.0523643493652344,
+      "learning_rate": 0.0001266164461382462,
+      "loss": 1.5977,
+      "step": 19660
+    },
+    {
+      "epoch": 4.143887923315953,
+      "grad_norm": 0.977243959903717,
+      "learning_rate": 0.00012655263356979747,
+      "loss": 1.6381,
+      "step": 19670
+    },
+    {
+      "epoch": 4.145994627903303,
+      "grad_norm": 1.057154893875122,
+      "learning_rate": 0.00012648880936673787,
+      "loss": 1.6399,
+      "step": 19680
+    },
+    {
+      "epoch": 4.1481013324906515,
+      "grad_norm": 1.0574363470077515,
+      "learning_rate": 0.00012642497355703326,
+      "loss": 1.6326,
+      "step": 19690
+    },
+    {
+      "epoch": 4.150208037078,
+      "grad_norm": 1.0001354217529297,
+      "learning_rate": 0.00012636112616865475,
+      "loss": 1.5812,
+      "step": 19700
+    },
+    {
+      "epoch": 4.15231474166535,
+      "grad_norm": 1.0256274938583374,
+      "learning_rate": 0.00012629726722957846,
+      "loss": 1.6748,
+      "step": 19710
+    },
+    {
+      "epoch": 4.154421446252699,
+      "grad_norm": 0.9991124272346497,
+      "learning_rate": 0.00012623339676778557,
+      "loss": 1.6085,
+      "step": 19720
+    },
+    {
+      "epoch": 4.156528150840049,
+      "grad_norm": 0.9248740673065186,
+      "learning_rate": 0.00012616951481126223,
+      "loss": 1.5815,
+      "step": 19730
+    },
+    {
+      "epoch": 4.1586348554273975,
+      "grad_norm": 1.117120385169983,
+      "learning_rate": 0.00012610562138799978,
+      "loss": 1.6614,
+      "step": 19740
+    },
+    {
+      "epoch": 4.160741560014747,
+      "grad_norm": 1.041090488433838,
+      "learning_rate": 0.00012604171652599448,
+      "loss": 1.606,
+      "step": 19750
+    },
+    {
+      "epoch": 4.162848264602096,
+      "grad_norm": 1.0659749507904053,
+      "learning_rate": 0.00012597780025324764,
+      "loss": 1.6511,
+      "step": 19760
+    },
+    {
+      "epoch": 4.164954969189445,
+      "grad_norm": 1.1362602710723877,
+      "learning_rate": 0.00012591387259776551,
+      "loss": 1.5923,
+      "step": 19770
+    },
+    {
+      "epoch": 4.167061673776795,
+      "grad_norm": 0.9490585327148438,
+      "learning_rate": 0.00012584993358755945,
+      "loss": 1.6317,
+      "step": 19780
+    },
+    {
+      "epoch": 4.169168378364144,
+      "grad_norm": 0.9721894860267639,
+      "learning_rate": 0.00012578598325064565,
+      "loss": 1.6217,
+      "step": 19790
+    },
+    {
+      "epoch": 4.171275082951493,
+      "grad_norm": 1.014675498008728,
+      "learning_rate": 0.00012572202161504543,
+      "loss": 1.6416,
+      "step": 19800
+    },
+    {
+      "epoch": 4.173381787538842,
+      "grad_norm": 0.959456205368042,
+      "learning_rate": 0.00012565804870878484,
+      "loss": 1.6154,
+      "step": 19810
+    },
+    {
+      "epoch": 4.175488492126192,
+      "grad_norm": 1.0517370700836182,
+      "learning_rate": 0.00012559406455989506,
+      "loss": 1.6004,
+      "step": 19820
+    },
+    {
+      "epoch": 4.177595196713541,
+      "grad_norm": 1.101207971572876,
+      "learning_rate": 0.00012553006919641214,
+      "loss": 1.6393,
+      "step": 19830
+    },
+    {
+      "epoch": 4.1797019013008905,
+      "grad_norm": 1.0225372314453125,
+      "learning_rate": 0.00012546606264637699,
+      "loss": 1.633,
+      "step": 19840
+    },
+    {
+      "epoch": 4.181808605888239,
+      "grad_norm": 0.9832279682159424,
+      "learning_rate": 0.00012540204493783553,
+      "loss": 1.6122,
+      "step": 19850
+    },
+    {
+      "epoch": 4.183915310475588,
+      "grad_norm": 1.0497034788131714,
+      "learning_rate": 0.00012533801609883842,
+      "loss": 1.6318,
+      "step": 19860
+    },
+    {
+      "epoch": 4.186022015062938,
+      "grad_norm": 0.9535748958587646,
+      "learning_rate": 0.00012527397615744138,
+      "loss": 1.6368,
+      "step": 19870
+    },
+    {
+      "epoch": 4.188128719650287,
+      "grad_norm": 1.0655570030212402,
+      "learning_rate": 0.0001252099251417048,
+      "loss": 1.6132,
+      "step": 19880
+    },
+    {
+      "epoch": 4.1902354242376365,
+      "grad_norm": 1.0989586114883423,
+      "learning_rate": 0.0001251458630796941,
+      "loss": 1.6372,
+      "step": 19890
+    },
+    {
+      "epoch": 4.192342128824985,
+      "grad_norm": 1.0104575157165527,
+      "learning_rate": 0.00012508178999947936,
+      "loss": 1.6516,
+      "step": 19900
+    },
+    {
+      "epoch": 4.194448833412335,
+      "grad_norm": 1.0405486822128296,
+      "learning_rate": 0.00012501770592913568,
+      "loss": 1.5935,
+      "step": 19910
+    },
+    {
+      "epoch": 4.196555537999684,
+      "grad_norm": 1.1122767925262451,
+      "learning_rate": 0.00012495361089674285,
+      "loss": 1.6198,
+      "step": 19920
+    },
+    {
+      "epoch": 4.198662242587034,
+      "grad_norm": 1.0726100206375122,
+      "learning_rate": 0.00012488950493038552,
+      "loss": 1.6371,
+      "step": 19930
+    },
+    {
+      "epoch": 4.2007689471743825,
+      "grad_norm": 1.0831425189971924,
+      "learning_rate": 0.000124825388058153,
+      "loss": 1.6321,
+      "step": 19940
+    },
+    {
+      "epoch": 4.202875651761731,
+      "grad_norm": 1.1081782579421997,
+      "learning_rate": 0.00012476126030813963,
+      "loss": 1.5934,
+      "step": 19950
+    },
+    {
+      "epoch": 4.204982356349081,
+      "grad_norm": 0.9888300895690918,
+      "learning_rate": 0.0001246971217084443,
+      "loss": 1.6427,
+      "step": 19960
+    },
+    {
+      "epoch": 4.20708906093643,
+      "grad_norm": 1.0104364156723022,
+      "learning_rate": 0.00012463297228717073,
+      "loss": 1.601,
+      "step": 19970
+    },
+    {
+      "epoch": 4.20919576552378,
+      "grad_norm": 1.0632632970809937,
+      "learning_rate": 0.00012456881207242732,
+      "loss": 1.5513,
+      "step": 19980
+    },
+    {
+      "epoch": 4.211302470111129,
+      "grad_norm": 1.0740721225738525,
+      "learning_rate": 0.0001245046410923274,
+      "loss": 1.69,
+      "step": 19990
+    },
+    {
+      "epoch": 4.213409174698478,
+      "grad_norm": 1.067266583442688,
+      "learning_rate": 0.00012444045937498873,
+      "loss": 1.69,
+      "step": 20000
+    },
+    {
+      "epoch": 4.215515879285827,
+      "grad_norm": 1.0238685607910156,
+      "learning_rate": 0.000124376266948534,
+      "loss": 1.5979,
+      "step": 20010
+    },
+    {
+      "epoch": 4.217622583873176,
+      "grad_norm": 1.1018869876861572,
+      "learning_rate": 0.00012431206384109044,
+      "loss": 1.6022,
+      "step": 20020
+    },
+    {
+      "epoch": 4.219729288460526,
+      "grad_norm": 1.0108250379562378,
+      "learning_rate": 0.00012424785008079015,
+      "loss": 1.6489,
+      "step": 20030
+    },
+    {
+      "epoch": 4.221835993047875,
+      "grad_norm": 0.9962803721427917,
+      "learning_rate": 0.00012418362569576965,
+      "loss": 1.6018,
+      "step": 20040
+    },
+    {
+      "epoch": 4.223942697635224,
+      "grad_norm": 0.9581331610679626,
+      "learning_rate": 0.00012411939071417034,
+      "loss": 1.6362,
+      "step": 20050
+    },
+    {
+      "epoch": 4.226049402222573,
+      "grad_norm": 1.0649348497390747,
+      "learning_rate": 0.00012405514516413807,
+      "loss": 1.6541,
+      "step": 20060
+    },
+    {
+      "epoch": 4.228156106809923,
+      "grad_norm": 1.1011024713516235,
+      "learning_rate": 0.0001239908890738235,
+      "loss": 1.6053,
+      "step": 20070
+    },
+    {
+      "epoch": 4.230262811397272,
+      "grad_norm": 1.0573521852493286,
+      "learning_rate": 0.0001239266224713818,
+      "loss": 1.6766,
+      "step": 20080
+    },
+    {
+      "epoch": 4.232369515984621,
+      "grad_norm": 1.1805315017700195,
+      "learning_rate": 0.00012386234538497282,
+      "loss": 1.6301,
+      "step": 20090
+    },
+    {
+      "epoch": 4.23447622057197,
+      "grad_norm": 0.9401804208755493,
+      "learning_rate": 0.00012379805784276082,
+      "loss": 1.6689,
+      "step": 20100
+    },
+    {
+      "epoch": 4.236582925159319,
+      "grad_norm": 0.9770873188972473,
+      "learning_rate": 0.0001237337598729149,
+      "loss": 1.6291,
+      "step": 20110
+    },
+    {
+      "epoch": 4.238689629746669,
+      "grad_norm": 1.0193746089935303,
+      "learning_rate": 0.00012366945150360861,
+      "loss": 1.6169,
+      "step": 20120
+    },
+    {
+      "epoch": 4.240796334334018,
+      "grad_norm": 1.0372521877288818,
+      "learning_rate": 0.00012360513276301997,
+      "loss": 1.6679,
+      "step": 20130
+    },
+    {
+      "epoch": 4.2429030389213676,
+      "grad_norm": 1.0554676055908203,
+      "learning_rate": 0.00012354080367933166,
+      "loss": 1.6073,
+      "step": 20140
+    },
+    {
+      "epoch": 4.245009743508716,
+      "grad_norm": 1.1003928184509277,
+      "learning_rate": 0.0001234764642807309,
+      "loss": 1.5924,
+      "step": 20150
+    },
+    {
+      "epoch": 4.247116448096066,
+      "grad_norm": 0.9821735620498657,
+      "learning_rate": 0.0001234121145954094,
+      "loss": 1.6366,
+      "step": 20160
+    },
+    {
+      "epoch": 4.249223152683415,
+      "grad_norm": 1.034835934638977,
+      "learning_rate": 0.0001233477546515633,
+      "loss": 1.6446,
+      "step": 20170
+    },
+    {
+      "epoch": 4.251329857270764,
+      "grad_norm": 1.0512527227401733,
+      "learning_rate": 0.00012328338447739333,
+      "loss": 1.6344,
+      "step": 20180
+    },
+    {
+      "epoch": 4.253436561858114,
+      "grad_norm": 0.9722573161125183,
+      "learning_rate": 0.00012321900410110464,
+      "loss": 1.5923,
+      "step": 20190
+    },
+    {
+      "epoch": 4.255543266445462,
+      "grad_norm": 1.1065928936004639,
+      "learning_rate": 0.000123154613550907,
+      "loss": 1.6165,
+      "step": 20200
+    },
+    {
+      "epoch": 4.257649971032812,
+      "grad_norm": 0.9826769232749939,
+      "learning_rate": 0.0001230902128550144,
+      "loss": 1.6689,
+      "step": 20210
+    },
+    {
+      "epoch": 4.259756675620161,
+      "grad_norm": 1.2589317560195923,
+      "learning_rate": 0.00012302580204164541,
+      "loss": 1.6341,
+      "step": 20220
+    },
+    {
+      "epoch": 4.261863380207511,
+      "grad_norm": 0.9650371670722961,
+      "learning_rate": 0.00012296138113902308,
+      "loss": 1.6348,
+      "step": 20230
+    },
+    {
+      "epoch": 4.26397008479486,
+      "grad_norm": 1.121001124382019,
+      "learning_rate": 0.00012289695017537485,
+      "loss": 1.6525,
+      "step": 20240
+    },
+    {
+      "epoch": 4.2660767893822085,
+      "grad_norm": 1.0059106349945068,
+      "learning_rate": 0.00012283250917893244,
+      "loss": 1.6566,
+      "step": 20250
+    },
+    {
+      "epoch": 4.268183493969558,
+      "grad_norm": 1.1646438837051392,
+      "learning_rate": 0.00012276805817793208,
+      "loss": 1.6315,
+      "step": 20260
+    },
+    {
+      "epoch": 4.270290198556907,
+      "grad_norm": 1.138887882232666,
+      "learning_rate": 0.00012270359720061445,
+      "loss": 1.6118,
+      "step": 20270
+    },
+    {
+      "epoch": 4.272396903144257,
+      "grad_norm": 1.1170200109481812,
+      "learning_rate": 0.0001226391262752245,
+      "loss": 1.6587,
+      "step": 20280
+    },
+    {
+      "epoch": 4.274503607731606,
+      "grad_norm": 1.041637897491455,
+      "learning_rate": 0.00012257464543001146,
+      "loss": 1.6728,
+      "step": 20290
+    },
+    {
+      "epoch": 4.276610312318955,
+      "grad_norm": 1.0330857038497925,
+      "learning_rate": 0.00012251015469322916,
+      "loss": 1.6058,
+      "step": 20300
+    },
+    {
+      "epoch": 4.278717016906304,
+      "grad_norm": 0.9935159087181091,
+      "learning_rate": 0.00012244565409313547,
+      "loss": 1.6765,
+      "step": 20310
+    },
+    {
+      "epoch": 4.280823721493654,
+      "grad_norm": 1.0498121976852417,
+      "learning_rate": 0.00012238114365799286,
+      "loss": 1.7234,
+      "step": 20320
+    },
+    {
+      "epoch": 4.282930426081003,
+      "grad_norm": 1.0286202430725098,
+      "learning_rate": 0.00012231662341606785,
+      "loss": 1.6389,
+      "step": 20330
+    },
+    {
+      "epoch": 4.285037130668352,
+      "grad_norm": 1.294057011604309,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 1.6428,
+      "step": 20340
+    },
+    {
+      "epoch": 4.287143835255701,
+      "grad_norm": 1.061030387878418,
+      "learning_rate": 0.00012218755362495887,
+      "loss": 1.6834,
+      "step": 20350
+    },
+    {
+      "epoch": 4.28925053984305,
+      "grad_norm": 1.0638394355773926,
+      "learning_rate": 0.00012212300413232962,
+      "loss": 1.6978,
+      "step": 20360
+    },
+    {
+      "epoch": 4.2913572444304,
+      "grad_norm": 1.0417766571044922,
+      "learning_rate": 0.0001220584449460274,
+      "loss": 1.6336,
+      "step": 20370
+    },
+    {
+      "epoch": 4.293463949017749,
+      "grad_norm": 0.9782829880714417,
+      "learning_rate": 0.0001219938760943403,
+      "loss": 1.6014,
+      "step": 20380
+    },
+    {
+      "epoch": 4.295570653605099,
+      "grad_norm": 1.0527657270431519,
+      "learning_rate": 0.00012192929760556049,
+      "loss": 1.6763,
+      "step": 20390
+    },
+    {
+      "epoch": 4.2976773581924474,
+      "grad_norm": 1.0444494485855103,
+      "learning_rate": 0.00012186470950798445,
+      "loss": 1.6582,
+      "step": 20400
+    },
+    {
+      "epoch": 4.299784062779796,
+      "grad_norm": 1.0852185487747192,
+      "learning_rate": 0.00012180011182991289,
+      "loss": 1.6798,
+      "step": 20410
+    },
+    {
+      "epoch": 4.301890767367146,
+      "grad_norm": 1.030380129814148,
+      "learning_rate": 0.00012173550459965062,
+      "loss": 1.6249,
+      "step": 20420
+    },
+    {
+      "epoch": 4.303997471954495,
+      "grad_norm": 0.9517531394958496,
+      "learning_rate": 0.00012167088784550673,
+      "loss": 1.6335,
+      "step": 20430
+    },
+    {
+      "epoch": 4.306104176541845,
+      "grad_norm": 1.0121312141418457,
+      "learning_rate": 0.00012160626159579447,
+      "loss": 1.6665,
+      "step": 20440
+    },
+    {
+      "epoch": 4.3082108811291935,
+      "grad_norm": 0.9951829314231873,
+      "learning_rate": 0.0001215416258788312,
+      "loss": 1.6113,
+      "step": 20450
+    },
+    {
+      "epoch": 4.310317585716543,
+      "grad_norm": 1.082280158996582,
+      "learning_rate": 0.00012147698072293842,
+      "loss": 1.6288,
+      "step": 20460
+    },
+    {
+      "epoch": 4.312424290303892,
+      "grad_norm": 0.967485785484314,
+      "learning_rate": 0.0001214123261564419,
+      "loss": 1.5958,
+      "step": 20470
+    },
+    {
+      "epoch": 4.314530994891241,
+      "grad_norm": 1.0343081951141357,
+      "learning_rate": 0.00012134766220767135,
+      "loss": 1.6265,
+      "step": 20480
+    },
+    {
+      "epoch": 4.316637699478591,
+      "grad_norm": 1.054532527923584,
+      "learning_rate": 0.00012128298890496072,
+      "loss": 1.6081,
+      "step": 20490
+    },
+    {
+      "epoch": 4.3187444040659395,
+      "grad_norm": 0.9487448334693909,
+      "learning_rate": 0.00012121830627664801,
+      "loss": 1.6646,
+      "step": 20500
+    },
+    {
+      "epoch": 4.320851108653289,
+      "grad_norm": 1.0977160930633545,
+      "learning_rate": 0.00012115361435107531,
+      "loss": 1.6541,
+      "step": 20510
+    },
+    {
+      "epoch": 4.322957813240638,
+      "grad_norm": 1.0766124725341797,
+      "learning_rate": 0.00012108891315658879,
+      "loss": 1.6211,
+      "step": 20520
+    },
+    {
+      "epoch": 4.325064517827988,
+      "grad_norm": 1.025858998298645,
+      "learning_rate": 0.00012102420272153869,
+      "loss": 1.6154,
+      "step": 20530
+    },
+    {
+      "epoch": 4.327171222415337,
+      "grad_norm": 0.9398080110549927,
+      "learning_rate": 0.00012095948307427925,
+      "loss": 1.6605,
+      "step": 20540
+    },
+    {
+      "epoch": 4.329277927002686,
+      "grad_norm": 1.111257791519165,
+      "learning_rate": 0.00012089475424316883,
+      "loss": 1.6486,
+      "step": 20550
+    },
+    {
+      "epoch": 4.331384631590035,
+      "grad_norm": 0.9591060876846313,
+      "learning_rate": 0.00012083001625656973,
+      "loss": 1.6059,
+      "step": 20560
+    },
+    {
+      "epoch": 4.333491336177384,
+      "grad_norm": 1.1178287267684937,
+      "learning_rate": 0.00012076526914284833,
+      "loss": 1.6159,
+      "step": 20570
+    },
+    {
+      "epoch": 4.335598040764734,
+      "grad_norm": 1.0105968713760376,
+      "learning_rate": 0.00012070051293037492,
+      "loss": 1.68,
+      "step": 20580
+    },
+    {
+      "epoch": 4.337704745352083,
+      "grad_norm": 1.0441944599151611,
+      "learning_rate": 0.00012063574764752394,
+      "loss": 1.6522,
+      "step": 20590
+    },
+    {
+      "epoch": 4.3398114499394325,
+      "grad_norm": 1.0669680833816528,
+      "learning_rate": 0.00012057097332267359,
+      "loss": 1.6637,
+      "step": 20600
+    },
+    {
+      "epoch": 4.341918154526781,
+      "grad_norm": 1.169826626777649,
+      "learning_rate": 0.00012050618998420624,
+      "loss": 1.6253,
+      "step": 20610
+    },
+    {
+      "epoch": 4.344024859114131,
+      "grad_norm": 1.0377858877182007,
+      "learning_rate": 0.00012044139766050802,
+      "loss": 1.593,
+      "step": 20620
+    },
+    {
+      "epoch": 4.34613156370148,
+      "grad_norm": 1.0214862823486328,
+      "learning_rate": 0.00012037659637996916,
+      "loss": 1.6451,
+      "step": 20630
+    },
+    {
+      "epoch": 4.34823826828883,
+      "grad_norm": 1.0646579265594482,
+      "learning_rate": 0.00012031178617098371,
+      "loss": 1.6613,
+      "step": 20640
+    },
+    {
+      "epoch": 4.3503449728761785,
+      "grad_norm": 1.0184786319732666,
+      "learning_rate": 0.00012024696706194967,
+      "loss": 1.6069,
+      "step": 20650
+    },
+    {
+      "epoch": 4.352451677463527,
+      "grad_norm": 1.0893901586532593,
+      "learning_rate": 0.00012018213908126889,
+      "loss": 1.6436,
+      "step": 20660
+    },
+    {
+      "epoch": 4.354558382050877,
+      "grad_norm": 0.9943939447402954,
+      "learning_rate": 0.00012011730225734723,
+      "loss": 1.6233,
+      "step": 20670
+    },
+    {
+      "epoch": 4.356665086638226,
+      "grad_norm": 1.5163390636444092,
+      "learning_rate": 0.00012005245661859434,
+      "loss": 1.6874,
+      "step": 20680
+    },
+    {
+      "epoch": 4.358771791225576,
+      "grad_norm": 1.077407717704773,
+      "learning_rate": 0.00011998760219342368,
+      "loss": 1.6384,
+      "step": 20690
+    },
+    {
+      "epoch": 4.3608784958129245,
+      "grad_norm": 1.0055080652236938,
+      "learning_rate": 0.00011992273901025269,
+      "loss": 1.6354,
+      "step": 20700
+    },
+    {
+      "epoch": 4.362985200400274,
+      "grad_norm": 1.019344449043274,
+      "learning_rate": 0.00011985786709750251,
+      "loss": 1.6234,
+      "step": 20710
+    },
+    {
+      "epoch": 4.365091904987623,
+      "grad_norm": 1.0513646602630615,
+      "learning_rate": 0.00011979298648359823,
+      "loss": 1.6259,
+      "step": 20720
+    },
+    {
+      "epoch": 4.367198609574972,
+      "grad_norm": 1.1279222965240479,
+      "learning_rate": 0.00011972809719696864,
+      "loss": 1.6621,
+      "step": 20730
+    },
+    {
+      "epoch": 4.369305314162322,
+      "grad_norm": 1.0146011114120483,
+      "learning_rate": 0.00011966319926604641,
+      "loss": 1.6617,
+      "step": 20740
+    },
+    {
+      "epoch": 4.371412018749671,
+      "grad_norm": 1.0017142295837402,
+      "learning_rate": 0.00011959829271926799,
+      "loss": 1.6856,
+      "step": 20750
+    },
+    {
+      "epoch": 4.37351872333702,
+      "grad_norm": 0.9613775014877319,
+      "learning_rate": 0.0001195333775850736,
+      "loss": 1.6276,
+      "step": 20760
+    },
+    {
+      "epoch": 4.375625427924369,
+      "grad_norm": 1.031233549118042,
+      "learning_rate": 0.00011946845389190715,
+      "loss": 1.6607,
+      "step": 20770
+    },
+    {
+      "epoch": 4.377732132511719,
+      "grad_norm": 0.905860960483551,
+      "learning_rate": 0.0001194035216682164,
+      "loss": 1.7105,
+      "step": 20780
+    },
+    {
+      "epoch": 4.379838837099068,
+      "grad_norm": 1.1338081359863281,
+      "learning_rate": 0.00011933858094245281,
+      "loss": 1.6139,
+      "step": 20790
+    },
+    {
+      "epoch": 4.381945541686417,
+      "grad_norm": 1.0127646923065186,
+      "learning_rate": 0.00011927363174307156,
+      "loss": 1.672,
+      "step": 20800
+    },
+    {
+      "epoch": 4.384052246273766,
+      "grad_norm": 0.9693343043327332,
+      "learning_rate": 0.00011920867409853154,
+      "loss": 1.5391,
+      "step": 20810
+    },
+    {
+      "epoch": 4.386158950861115,
+      "grad_norm": 1.008926510810852,
+      "learning_rate": 0.00011914370803729533,
+      "loss": 1.637,
+      "step": 20820
+    },
+    {
+      "epoch": 4.388265655448465,
+      "grad_norm": 1.1421765089035034,
+      "learning_rate": 0.00011907873358782926,
+      "loss": 1.6087,
+      "step": 20830
+    },
+    {
+      "epoch": 4.390372360035814,
+      "grad_norm": 1.052461862564087,
+      "learning_rate": 0.00011901375077860329,
+      "loss": 1.6072,
+      "step": 20840
+    },
+    {
+      "epoch": 4.3924790646231635,
+      "grad_norm": 1.0214227437973022,
+      "learning_rate": 0.00011894875963809098,
+      "loss": 1.6071,
+      "step": 20850
+    },
+    {
+      "epoch": 4.394585769210512,
+      "grad_norm": 1.00290846824646,
+      "learning_rate": 0.00011888376019476966,
+      "loss": 1.5995,
+      "step": 20860
+    },
+    {
+      "epoch": 4.396692473797862,
+      "grad_norm": 1.0803271532058716,
+      "learning_rate": 0.00011881875247712025,
+      "loss": 1.6456,
+      "step": 20870
+    },
+    {
+      "epoch": 4.398799178385211,
+      "grad_norm": 1.1548503637313843,
+      "learning_rate": 0.00011875373651362727,
+      "loss": 1.6417,
+      "step": 20880
+    },
+    {
+      "epoch": 4.40090588297256,
+      "grad_norm": 1.0493422746658325,
+      "learning_rate": 0.00011868871233277884,
+      "loss": 1.6475,
+      "step": 20890
+    },
+    {
+      "epoch": 4.4030125875599095,
+      "grad_norm": 0.9755949974060059,
+      "learning_rate": 0.00011862367996306673,
+      "loss": 1.6587,
+      "step": 20900
+    },
+    {
+      "epoch": 4.405119292147258,
+      "grad_norm": 1.0143030881881714,
+      "learning_rate": 0.00011855863943298631,
+      "loss": 1.6612,
+      "step": 20910
+    },
+    {
+      "epoch": 4.407225996734608,
+      "grad_norm": 1.1264982223510742,
+      "learning_rate": 0.0001184935907710365,
+      "loss": 1.649,
+      "step": 20920
+    },
+    {
+      "epoch": 4.409332701321957,
+      "grad_norm": 1.0828158855438232,
+      "learning_rate": 0.00011842853400571971,
+      "loss": 1.5933,
+      "step": 20930
+    },
+    {
+      "epoch": 4.411439405909307,
+      "grad_norm": 1.0389914512634277,
+      "learning_rate": 0.00011836346916554205,
+      "loss": 1.6446,
+      "step": 20940
+    },
+    {
+      "epoch": 4.413546110496656,
+      "grad_norm": 1.0088750123977661,
+      "learning_rate": 0.00011829839627901302,
+      "loss": 1.6015,
+      "step": 20950
+    },
+    {
+      "epoch": 4.415652815084005,
+      "grad_norm": 1.0506969690322876,
+      "learning_rate": 0.00011823331537464574,
+      "loss": 1.649,
+      "step": 20960
+    },
+    {
+      "epoch": 4.417759519671354,
+      "grad_norm": 1.0698243379592896,
+      "learning_rate": 0.00011816822648095687,
+      "loss": 1.627,
+      "step": 20970
+    },
+    {
+      "epoch": 4.419866224258703,
+      "grad_norm": 1.0903515815734863,
+      "learning_rate": 0.00011810312962646644,
+      "loss": 1.6891,
+      "step": 20980
+    },
+    {
+      "epoch": 4.421972928846053,
+      "grad_norm": 1.1337631940841675,
+      "learning_rate": 0.00011803802483969806,
+      "loss": 1.6733,
+      "step": 20990
+    },
+    {
+      "epoch": 4.424079633433402,
+      "grad_norm": 1.6732696294784546,
+      "learning_rate": 0.00011797291214917881,
+      "loss": 1.6171,
+      "step": 21000
+    },
+    {
+      "epoch": 4.426186338020751,
+      "grad_norm": 0.9726201891899109,
+      "learning_rate": 0.00011790779158343925,
+      "loss": 1.6015,
+      "step": 21010
+    },
+    {
+      "epoch": 4.4282930426081,
+      "grad_norm": 1.0177321434020996,
+      "learning_rate": 0.00011784266317101333,
+      "loss": 1.6375,
+      "step": 21020
+    },
+    {
+      "epoch": 4.43039974719545,
+      "grad_norm": 1.0415085554122925,
+      "learning_rate": 0.00011777752694043849,
+      "loss": 1.656,
+      "step": 21030
+    },
+    {
+      "epoch": 4.432506451782799,
+      "grad_norm": 1.0331918001174927,
+      "learning_rate": 0.00011771238292025558,
+      "loss": 1.6425,
+      "step": 21040
+    },
+    {
+      "epoch": 4.434613156370148,
+      "grad_norm": 0.9632758498191833,
+      "learning_rate": 0.00011764723113900886,
+      "loss": 1.6347,
+      "step": 21050
+    },
+    {
+      "epoch": 4.436719860957497,
+      "grad_norm": 1.1020656824111938,
+      "learning_rate": 0.00011758207162524598,
+      "loss": 1.657,
+      "step": 21060
+    },
+    {
+      "epoch": 4.438826565544846,
+      "grad_norm": 1.0980470180511475,
+      "learning_rate": 0.000117516904407518,
+      "loss": 1.674,
+      "step": 21070
+    },
+    {
+      "epoch": 4.440933270132196,
+      "grad_norm": 1.0126473903656006,
+      "learning_rate": 0.00011745172951437932,
+      "loss": 1.6464,
+      "step": 21080
+    },
+    {
+      "epoch": 4.443039974719545,
+      "grad_norm": 0.9396146535873413,
+      "learning_rate": 0.00011738654697438782,
+      "loss": 1.6033,
+      "step": 21090
+    },
+    {
+      "epoch": 4.4451466793068946,
+      "grad_norm": 1.0903655290603638,
+      "learning_rate": 0.00011732135681610452,
+      "loss": 1.6508,
+      "step": 21100
+    },
+    {
+      "epoch": 4.447253383894243,
+      "grad_norm": 1.0132126808166504,
+      "learning_rate": 0.00011725615906809397,
+      "loss": 1.6235,
+      "step": 21110
+    },
+    {
+      "epoch": 4.449360088481592,
+      "grad_norm": 1.0690926313400269,
+      "learning_rate": 0.00011719095375892396,
+      "loss": 1.6116,
+      "step": 21120
+    },
+    {
+      "epoch": 4.451466793068942,
+      "grad_norm": 1.099784255027771,
+      "learning_rate": 0.00011712574091716563,
+      "loss": 1.6192,
+      "step": 21130
+    },
+    {
+      "epoch": 4.453573497656291,
+      "grad_norm": 1.0641016960144043,
+      "learning_rate": 0.00011706052057139335,
+      "loss": 1.6692,
+      "step": 21140
+    },
+    {
+      "epoch": 4.455680202243641,
+      "grad_norm": 0.9757359027862549,
+      "learning_rate": 0.00011699529275018484,
+      "loss": 1.6685,
+      "step": 21150
+    },
+    {
+      "epoch": 4.457786906830989,
+      "grad_norm": 1.322670817375183,
+      "learning_rate": 0.0001169300574821211,
+      "loss": 1.6492,
+      "step": 21160
+    },
+    {
+      "epoch": 4.459893611418339,
+      "grad_norm": 1.0116708278656006,
+      "learning_rate": 0.0001168648147957864,
+      "loss": 1.619,
+      "step": 21170
+    },
+    {
+      "epoch": 4.462000316005688,
+      "grad_norm": 1.0692319869995117,
+      "learning_rate": 0.00011679956471976814,
+      "loss": 1.6236,
+      "step": 21180
+    },
+    {
+      "epoch": 4.464107020593037,
+      "grad_norm": 0.9994955658912659,
+      "learning_rate": 0.00011673430728265713,
+      "loss": 1.6552,
+      "step": 21190
+    },
+    {
+      "epoch": 4.466213725180387,
+      "grad_norm": 1.0054254531860352,
+      "learning_rate": 0.00011666904251304731,
+      "loss": 1.6309,
+      "step": 21200
+    },
+    {
+      "epoch": 4.4683204297677355,
+      "grad_norm": 1.0914092063903809,
+      "learning_rate": 0.00011660377043953588,
+      "loss": 1.6467,
+      "step": 21210
+    },
+    {
+      "epoch": 4.470427134355085,
+      "grad_norm": 0.9891067147254944,
+      "learning_rate": 0.00011653849109072314,
+      "loss": 1.6362,
+      "step": 21220
+    },
+    {
+      "epoch": 4.472533838942434,
+      "grad_norm": 1.1097562313079834,
+      "learning_rate": 0.00011647320449521268,
+      "loss": 1.6397,
+      "step": 21230
+    },
+    {
+      "epoch": 4.474640543529784,
+      "grad_norm": 1.0569071769714355,
+      "learning_rate": 0.0001164079106816113,
+      "loss": 1.6645,
+      "step": 21240
+    },
+    {
+      "epoch": 4.476747248117133,
+      "grad_norm": 1.0638412237167358,
+      "learning_rate": 0.00011634260967852882,
+      "loss": 1.6816,
+      "step": 21250
+    },
+    {
+      "epoch": 4.478853952704482,
+      "grad_norm": 1.0064091682434082,
+      "learning_rate": 0.00011627730151457829,
+      "loss": 1.6874,
+      "step": 21260
+    },
+    {
+      "epoch": 4.480960657291831,
+      "grad_norm": 1.0138236284255981,
+      "learning_rate": 0.00011621198621837593,
+      "loss": 1.6736,
+      "step": 21270
+    },
+    {
+      "epoch": 4.48306736187918,
+      "grad_norm": 1.0266871452331543,
+      "learning_rate": 0.00011614666381854107,
+      "loss": 1.638,
+      "step": 21280
+    },
+    {
+      "epoch": 4.48517406646653,
+      "grad_norm": 1.0543068647384644,
+      "learning_rate": 0.00011608133434369604,
+      "loss": 1.6258,
+      "step": 21290
+    },
+    {
+      "epoch": 4.487280771053879,
+      "grad_norm": 1.0614491701126099,
+      "learning_rate": 0.00011601599782246646,
+      "loss": 1.6601,
+      "step": 21300
+    },
+    {
+      "epoch": 4.489387475641228,
+      "grad_norm": 1.1936269998550415,
+      "learning_rate": 0.00011595065428348087,
+      "loss": 1.6566,
+      "step": 21310
+    },
+    {
+      "epoch": 4.491494180228577,
+      "grad_norm": 1.0331757068634033,
+      "learning_rate": 0.00011588530375537101,
+      "loss": 1.6704,
+      "step": 21320
+    },
+    {
+      "epoch": 4.493600884815927,
+      "grad_norm": 1.0366097688674927,
+      "learning_rate": 0.0001158199462667716,
+      "loss": 1.7132,
+      "step": 21330
+    },
+    {
+      "epoch": 4.495707589403276,
+      "grad_norm": 0.966823935508728,
+      "learning_rate": 0.00011575458184632044,
+      "loss": 1.5783,
+      "step": 21340
+    },
+    {
+      "epoch": 4.497814293990626,
+      "grad_norm": 1.0771065950393677,
+      "learning_rate": 0.00011568921052265836,
+      "loss": 1.6208,
+      "step": 21350
+    },
+    {
+      "epoch": 4.4999209985779745,
+      "grad_norm": 1.0702791213989258,
+      "learning_rate": 0.00011562383232442926,
+      "loss": 1.7674,
+      "step": 21360
+    },
+    {
+      "epoch": 4.502027703165323,
+      "grad_norm": 1.0116688013076782,
+      "learning_rate": 0.00011555844728027993,
+      "loss": 1.6573,
+      "step": 21370
+    },
+    {
+      "epoch": 4.504134407752673,
+      "grad_norm": 0.97417813539505,
+      "learning_rate": 0.00011549305541886032,
+      "loss": 1.6242,
+      "step": 21380
+    },
+    {
+      "epoch": 4.506241112340022,
+      "grad_norm": 1.091288447380066,
+      "learning_rate": 0.00011542765676882325,
+      "loss": 1.6407,
+      "step": 21390
+    },
+    {
+      "epoch": 4.508347816927372,
+      "grad_norm": 1.0396335124969482,
+      "learning_rate": 0.0001153622513588246,
+      "loss": 1.6791,
+      "step": 21400
+    },
+    {
+      "epoch": 4.5104545215147205,
+      "grad_norm": 1.0363614559173584,
+      "learning_rate": 0.0001152968392175231,
+      "loss": 1.6589,
+      "step": 21410
+    },
+    {
+      "epoch": 4.51256122610207,
+      "grad_norm": 0.9808732271194458,
+      "learning_rate": 0.0001152314203735805,
+      "loss": 1.6257,
+      "step": 21420
+    },
+    {
+      "epoch": 4.514667930689419,
+      "grad_norm": 1.0970127582550049,
+      "learning_rate": 0.00011516599485566153,
+      "loss": 1.6904,
+      "step": 21430
+    },
+    {
+      "epoch": 4.516774635276768,
+      "grad_norm": 1.1880615949630737,
+      "learning_rate": 0.00011510056269243379,
+      "loss": 1.6835,
+      "step": 21440
+    },
+    {
+      "epoch": 4.518881339864118,
+      "grad_norm": 1.0334802865982056,
+      "learning_rate": 0.00011503512391256776,
+      "loss": 1.6452,
+      "step": 21450
+    },
+    {
+      "epoch": 4.5209880444514665,
+      "grad_norm": 1.0974247455596924,
+      "learning_rate": 0.00011496967854473688,
+      "loss": 1.6913,
+      "step": 21460
+    },
+    {
+      "epoch": 4.523094749038816,
+      "grad_norm": 1.1479419469833374,
+      "learning_rate": 0.00011490422661761744,
+      "loss": 1.6741,
+      "step": 21470
+    },
+    {
+      "epoch": 4.525201453626165,
+      "grad_norm": 1.1152433156967163,
+      "learning_rate": 0.00011483876815988867,
+      "loss": 1.6518,
+      "step": 21480
+    },
+    {
+      "epoch": 4.527308158213515,
+      "grad_norm": 1.0902926921844482,
+      "learning_rate": 0.00011477330320023255,
+      "loss": 1.7099,
+      "step": 21490
+    },
+    {
+      "epoch": 4.529414862800864,
+      "grad_norm": 1.1328729391098022,
+      "learning_rate": 0.00011470783176733395,
+      "loss": 1.6352,
+      "step": 21500
+    },
+    {
+      "epoch": 4.5315215673882125,
+      "grad_norm": 1.0286601781845093,
+      "learning_rate": 0.00011464235388988067,
+      "loss": 1.7122,
+      "step": 21510
+    },
+    {
+      "epoch": 4.533628271975562,
+      "grad_norm": 1.0103448629379272,
+      "learning_rate": 0.00011457686959656322,
+      "loss": 1.6624,
+      "step": 21520
+    },
+    {
+      "epoch": 4.535734976562911,
+      "grad_norm": 0.9834364056587219,
+      "learning_rate": 0.00011451137891607495,
+      "loss": 1.6217,
+      "step": 21530
+    },
+    {
+      "epoch": 4.537841681150261,
+      "grad_norm": 1.0024162530899048,
+      "learning_rate": 0.00011444588187711205,
+      "loss": 1.6639,
+      "step": 21540
+    },
+    {
+      "epoch": 4.53994838573761,
+      "grad_norm": 1.019455909729004,
+      "learning_rate": 0.00011438037850837342,
+      "loss": 1.6285,
+      "step": 21550
+    },
+    {
+      "epoch": 4.5420550903249595,
+      "grad_norm": 1.0064702033996582,
+      "learning_rate": 0.00011431486883856082,
+      "loss": 1.6412,
+      "step": 21560
+    },
+    {
+      "epoch": 4.544161794912308,
+      "grad_norm": 1.1464800834655762,
+      "learning_rate": 0.0001142493528963787,
+      "loss": 1.6867,
+      "step": 21570
+    },
+    {
+      "epoch": 4.546268499499657,
+      "grad_norm": 0.978492021560669,
+      "learning_rate": 0.00011418383071053431,
+      "loss": 1.6104,
+      "step": 21580
+    },
+    {
+      "epoch": 4.548375204087007,
+      "grad_norm": 1.1636004447937012,
+      "learning_rate": 0.0001141183023097376,
+      "loss": 1.6889,
+      "step": 21590
+    },
+    {
+      "epoch": 4.550481908674356,
+      "grad_norm": 1.0960801839828491,
+      "learning_rate": 0.00011405276772270126,
+      "loss": 1.6693,
+      "step": 21600
+    },
+    {
+      "epoch": 4.5525886132617055,
+      "grad_norm": 1.0665209293365479,
+      "learning_rate": 0.0001139872269781407,
+      "loss": 1.6581,
+      "step": 21610
+    },
+    {
+      "epoch": 4.554695317849054,
+      "grad_norm": 1.0270514488220215,
+      "learning_rate": 0.00011392168010477398,
+      "loss": 1.6687,
+      "step": 21620
+    },
+    {
+      "epoch": 4.556802022436404,
+      "grad_norm": 1.0477863550186157,
+      "learning_rate": 0.0001138561271313219,
+      "loss": 1.6623,
+      "step": 21630
+    },
+    {
+      "epoch": 4.558908727023753,
+      "grad_norm": 1.1489893198013306,
+      "learning_rate": 0.00011379056808650794,
+      "loss": 1.6419,
+      "step": 21640
+    },
+    {
+      "epoch": 4.561015431611103,
+      "grad_norm": 0.995190441608429,
+      "learning_rate": 0.0001137250029990582,
+      "loss": 1.6129,
+      "step": 21650
+    },
+    {
+      "epoch": 4.5631221361984515,
+      "grad_norm": 1.0449438095092773,
+      "learning_rate": 0.0001136594318977014,
+      "loss": 1.6834,
+      "step": 21660
+    },
+    {
+      "epoch": 4.565228840785801,
+      "grad_norm": 0.9984011054039001,
+      "learning_rate": 0.00011359385481116897,
+      "loss": 1.6747,
+      "step": 21670
+    },
+    {
+      "epoch": 4.56733554537315,
+      "grad_norm": 1.0843603610992432,
+      "learning_rate": 0.00011352827176819496,
+      "loss": 1.6948,
+      "step": 21680
+    },
+    {
+      "epoch": 4.569442249960499,
+      "grad_norm": 1.0422046184539795,
+      "learning_rate": 0.00011346268279751595,
+      "loss": 1.6508,
+      "step": 21690
+    },
+    {
+      "epoch": 4.571548954547849,
+      "grad_norm": 0.9797605276107788,
+      "learning_rate": 0.00011339708792787119,
+      "loss": 1.5822,
+      "step": 21700
+    },
+    {
+      "epoch": 4.573655659135198,
+      "grad_norm": 0.9891635179519653,
+      "learning_rate": 0.00011333148718800248,
+      "loss": 1.6098,
+      "step": 21710
+    },
+    {
+      "epoch": 4.575762363722547,
+      "grad_norm": 1.0078808069229126,
+      "learning_rate": 0.0001132658806066542,
+      "loss": 1.5623,
+      "step": 21720
+    },
+    {
+      "epoch": 4.577869068309896,
+      "grad_norm": 0.9395504593849182,
+      "learning_rate": 0.00011320026821257333,
+      "loss": 1.6912,
+      "step": 21730
+    },
+    {
+      "epoch": 4.579975772897246,
+      "grad_norm": 0.9831303954124451,
+      "learning_rate": 0.00011313465003450931,
+      "loss": 1.6553,
+      "step": 21740
+    },
+    {
+      "epoch": 4.582082477484595,
+      "grad_norm": 0.9971016645431519,
+      "learning_rate": 0.00011306902610121419,
+      "loss": 1.651,
+      "step": 21750
+    },
+    {
+      "epoch": 4.584189182071944,
+      "grad_norm": 1.0432978868484497,
+      "learning_rate": 0.00011300339644144252,
+      "loss": 1.6727,
+      "step": 21760
+    },
+    {
+      "epoch": 4.586295886659293,
+      "grad_norm": 1.1212437152862549,
+      "learning_rate": 0.00011293776108395135,
+      "loss": 1.6057,
+      "step": 21770
+    },
+    {
+      "epoch": 4.588402591246642,
+      "grad_norm": 0.9745138883590698,
+      "learning_rate": 0.00011287212005750024,
+      "loss": 1.6231,
+      "step": 21780
+    },
+    {
+      "epoch": 4.590509295833992,
+      "grad_norm": 1.057394027709961,
+      "learning_rate": 0.00011280647339085118,
+      "loss": 1.5696,
+      "step": 21790
+    },
+    {
+      "epoch": 4.592616000421341,
+      "grad_norm": 1.1260936260223389,
+      "learning_rate": 0.00011274082111276876,
+      "loss": 1.6752,
+      "step": 21800
+    },
+    {
+      "epoch": 4.5947227050086905,
+      "grad_norm": 1.133716344833374,
+      "learning_rate": 0.00011267516325201985,
+      "loss": 1.6331,
+      "step": 21810
+    },
+    {
+      "epoch": 4.596829409596039,
+      "grad_norm": 1.1236519813537598,
+      "learning_rate": 0.00011260949983737398,
+      "loss": 1.6272,
+      "step": 21820
+    },
+    {
+      "epoch": 4.598936114183388,
+      "grad_norm": 0.9887076020240784,
+      "learning_rate": 0.00011254383089760285,
+      "loss": 1.6449,
+      "step": 21830
+    },
+    {
+      "epoch": 4.601042818770738,
+      "grad_norm": 1.0048518180847168,
+      "learning_rate": 0.00011247815646148087,
+      "loss": 1.6848,
+      "step": 21840
+    },
+    {
+      "epoch": 4.603149523358087,
+      "grad_norm": 0.9584482312202454,
+      "learning_rate": 0.00011241247655778464,
+      "loss": 1.6587,
+      "step": 21850
+    },
+    {
+      "epoch": 4.6052562279454365,
+      "grad_norm": 1.220842957496643,
+      "learning_rate": 0.00011234679121529323,
+      "loss": 1.64,
+      "step": 21860
+    },
+    {
+      "epoch": 4.607362932532785,
+      "grad_norm": 1.1299933195114136,
+      "learning_rate": 0.00011228110046278808,
+      "loss": 1.661,
+      "step": 21870
+    },
+    {
+      "epoch": 4.609469637120135,
+      "grad_norm": 1.0172736644744873,
+      "learning_rate": 0.00011221540432905309,
+      "loss": 1.6298,
+      "step": 21880
+    },
+    {
+      "epoch": 4.611576341707484,
+      "grad_norm": 1.0505467653274536,
+      "learning_rate": 0.00011214970284287435,
+      "loss": 1.668,
+      "step": 21890
+    },
+    {
+      "epoch": 4.613683046294833,
+      "grad_norm": 0.9825325608253479,
+      "learning_rate": 0.00011208399603304047,
+      "loss": 1.5831,
+      "step": 21900
+    },
+    {
+      "epoch": 4.615789750882183,
+      "grad_norm": 1.0815013647079468,
+      "learning_rate": 0.00011201828392834223,
+      "loss": 1.6227,
+      "step": 21910
+    },
+    {
+      "epoch": 4.617896455469531,
+      "grad_norm": 0.9960325360298157,
+      "learning_rate": 0.00011195256655757288,
+      "loss": 1.646,
+      "step": 21920
+    },
+    {
+      "epoch": 4.620003160056881,
+      "grad_norm": 1.1286170482635498,
+      "learning_rate": 0.00011188684394952789,
+      "loss": 1.6757,
+      "step": 21930
+    },
+    {
+      "epoch": 4.62210986464423,
+      "grad_norm": 0.9996662735939026,
+      "learning_rate": 0.00011182111613300501,
+      "loss": 1.6136,
+      "step": 21940
+    },
+    {
+      "epoch": 4.62421656923158,
+      "grad_norm": 1.0416985750198364,
+      "learning_rate": 0.00011175538313680431,
+      "loss": 1.655,
+      "step": 21950
+    },
+    {
+      "epoch": 4.626323273818929,
+      "grad_norm": 1.0647679567337036,
+      "learning_rate": 0.00011168964498972818,
+      "loss": 1.6109,
+      "step": 21960
+    },
+    {
+      "epoch": 4.628429978406278,
+      "grad_norm": 1.1595183610916138,
+      "learning_rate": 0.00011162390172058115,
+      "loss": 1.6105,
+      "step": 21970
+    },
+    {
+      "epoch": 4.630536682993627,
+      "grad_norm": 1.0993425846099854,
+      "learning_rate": 0.00011155815335817011,
+      "loss": 1.6102,
+      "step": 21980
+    },
+    {
+      "epoch": 4.632643387580977,
+      "grad_norm": 0.9971186518669128,
+      "learning_rate": 0.00011149239993130403,
+      "loss": 1.6735,
+      "step": 21990
+    },
+    {
+      "epoch": 4.634750092168326,
+      "grad_norm": 1.1622364521026611,
+      "learning_rate": 0.00011142664146879432,
+      "loss": 1.6381,
+      "step": 22000
+    },
+    {
+      "epoch": 4.636856796755675,
+      "grad_norm": 1.0967098474502563,
+      "learning_rate": 0.00011136087799945438,
+      "loss": 1.6526,
+      "step": 22010
+    },
+    {
+      "epoch": 4.638963501343024,
+      "grad_norm": 1.0782030820846558,
+      "learning_rate": 0.00011129510955209996,
+      "loss": 1.755,
+      "step": 22020
+    },
+    {
+      "epoch": 4.641070205930373,
+      "grad_norm": 0.965355634689331,
+      "learning_rate": 0.00011122933615554889,
+      "loss": 1.6236,
+      "step": 22030
+    },
+    {
+      "epoch": 4.643176910517723,
+      "grad_norm": 1.0064630508422852,
+      "learning_rate": 0.00011116355783862122,
+      "loss": 1.6899,
+      "step": 22040
+    },
+    {
+      "epoch": 4.645283615105072,
+      "grad_norm": 1.1502881050109863,
+      "learning_rate": 0.00011109777463013915,
+      "loss": 1.6342,
+      "step": 22050
+    },
+    {
+      "epoch": 4.647390319692422,
+      "grad_norm": 1.1655616760253906,
+      "learning_rate": 0.00011103198655892699,
+      "loss": 1.636,
+      "step": 22060
+    },
+    {
+      "epoch": 4.64949702427977,
+      "grad_norm": 1.063859462738037,
+      "learning_rate": 0.00011096619365381123,
+      "loss": 1.6424,
+      "step": 22070
+    },
+    {
+      "epoch": 4.651603728867119,
+      "grad_norm": 1.0649337768554688,
+      "learning_rate": 0.00011090039594362045,
+      "loss": 1.6466,
+      "step": 22080
+    },
+    {
+      "epoch": 4.653710433454469,
+      "grad_norm": 0.9440916180610657,
+      "learning_rate": 0.00011083459345718535,
+      "loss": 1.6441,
+      "step": 22090
+    },
+    {
+      "epoch": 4.655817138041818,
+      "grad_norm": 1.06228768825531,
+      "learning_rate": 0.00011076878622333868,
+      "loss": 1.6706,
+      "step": 22100
+    },
+    {
+      "epoch": 4.657923842629168,
+      "grad_norm": 0.9528666734695435,
+      "learning_rate": 0.00011070297427091534,
+      "loss": 1.6179,
+      "step": 22110
+    },
+    {
+      "epoch": 4.660030547216516,
+      "grad_norm": 1.1522499322891235,
+      "learning_rate": 0.00011063715762875225,
+      "loss": 1.6745,
+      "step": 22120
+    },
+    {
+      "epoch": 4.662137251803866,
+      "grad_norm": 1.049260139465332,
+      "learning_rate": 0.00011057133632568839,
+      "loss": 1.6675,
+      "step": 22130
+    },
+    {
+      "epoch": 4.664243956391215,
+      "grad_norm": 1.0696789026260376,
+      "learning_rate": 0.00011050551039056479,
+      "loss": 1.6355,
+      "step": 22140
+    },
+    {
+      "epoch": 4.666350660978564,
+      "grad_norm": 0.9847993850708008,
+      "learning_rate": 0.0001104396798522245,
+      "loss": 1.6592,
+      "step": 22150
+    },
+    {
+      "epoch": 4.668457365565914,
+      "grad_norm": 1.102442979812622,
+      "learning_rate": 0.0001103738447395126,
+      "loss": 1.6874,
+      "step": 22160
+    },
+    {
+      "epoch": 4.6705640701532625,
+      "grad_norm": 0.9978004097938538,
+      "learning_rate": 0.0001103080050812762,
+      "loss": 1.6361,
+      "step": 22170
+    },
+    {
+      "epoch": 4.672670774740612,
+      "grad_norm": 1.0541409254074097,
+      "learning_rate": 0.00011024216090636433,
+      "loss": 1.6731,
+      "step": 22180
+    },
+    {
+      "epoch": 4.674777479327961,
+      "grad_norm": 1.0716190338134766,
+      "learning_rate": 0.00011017631224362803,
+      "loss": 1.6109,
+      "step": 22190
+    },
+    {
+      "epoch": 4.676884183915311,
+      "grad_norm": 1.0343621969223022,
+      "learning_rate": 0.00011011045912192035,
+      "loss": 1.6403,
+      "step": 22200
+    },
+    {
+      "epoch": 4.67899088850266,
+      "grad_norm": 1.1150816679000854,
+      "learning_rate": 0.00011004460157009626,
+      "loss": 1.6232,
+      "step": 22210
+    },
+    {
+      "epoch": 4.6810975930900085,
+      "grad_norm": 0.9567576050758362,
+      "learning_rate": 0.00010997873961701266,
+      "loss": 1.6275,
+      "step": 22220
+    },
+    {
+      "epoch": 4.683204297677358,
+      "grad_norm": 1.0110678672790527,
+      "learning_rate": 0.00010991287329152838,
+      "loss": 1.6055,
+      "step": 22230
+    },
+    {
+      "epoch": 4.685311002264707,
+      "grad_norm": 1.0686061382293701,
+      "learning_rate": 0.00010984700262250418,
+      "loss": 1.6376,
+      "step": 22240
+    },
+    {
+      "epoch": 4.687417706852057,
+      "grad_norm": 1.069300651550293,
+      "learning_rate": 0.00010978112763880275,
+      "loss": 1.6952,
+      "step": 22250
+    },
+    {
+      "epoch": 4.689524411439406,
+      "grad_norm": 0.9756374955177307,
+      "learning_rate": 0.0001097152483692886,
+      "loss": 1.6478,
+      "step": 22260
+    },
+    {
+      "epoch": 4.691631116026755,
+      "grad_norm": 1.0548217296600342,
+      "learning_rate": 0.00010964936484282817,
+      "loss": 1.6522,
+      "step": 22270
+    },
+    {
+      "epoch": 4.693737820614104,
+      "grad_norm": 1.073805570602417,
+      "learning_rate": 0.00010958347708828976,
+      "loss": 1.6336,
+      "step": 22280
+    },
+    {
+      "epoch": 4.695844525201454,
+      "grad_norm": 1.0033069849014282,
+      "learning_rate": 0.00010951758513454351,
+      "loss": 1.6493,
+      "step": 22290
+    },
+    {
+      "epoch": 4.697951229788803,
+      "grad_norm": 1.0431855916976929,
+      "learning_rate": 0.00010945168901046139,
+      "loss": 1.622,
+      "step": 22300
+    },
+    {
+      "epoch": 4.700057934376153,
+      "grad_norm": 1.0515035390853882,
+      "learning_rate": 0.00010938578874491722,
+      "loss": 1.6085,
+      "step": 22310
+    },
+    {
+      "epoch": 4.7021646389635015,
+      "grad_norm": 0.9794484972953796,
+      "learning_rate": 0.00010931988436678666,
+      "loss": 1.664,
+      "step": 22320
+    },
+    {
+      "epoch": 4.70427134355085,
+      "grad_norm": 1.1576461791992188,
+      "learning_rate": 0.00010925397590494712,
+      "loss": 1.7116,
+      "step": 22330
+    },
+    {
+      "epoch": 4.7063780481382,
+      "grad_norm": 1.0061620473861694,
+      "learning_rate": 0.00010918806338827778,
+      "loss": 1.57,
+      "step": 22340
+    },
+    {
+      "epoch": 4.708484752725549,
+      "grad_norm": 1.0988759994506836,
+      "learning_rate": 0.00010912214684565967,
+      "loss": 1.6373,
+      "step": 22350
+    },
+    {
+      "epoch": 4.710591457312899,
+      "grad_norm": 1.1311630010604858,
+      "learning_rate": 0.00010905622630597558,
+      "loss": 1.6238,
+      "step": 22360
+    },
+    {
+      "epoch": 4.7126981619002475,
+      "grad_norm": 1.1295340061187744,
+      "learning_rate": 0.00010899030179810997,
+      "loss": 1.6584,
+      "step": 22370
+    },
+    {
+      "epoch": 4.714804866487597,
+      "grad_norm": 1.0294084548950195,
+      "learning_rate": 0.00010892437335094912,
+      "loss": 1.6811,
+      "step": 22380
+    },
+    {
+      "epoch": 4.716911571074946,
+      "grad_norm": 1.053384780883789,
+      "learning_rate": 0.00010885844099338094,
+      "loss": 1.6385,
+      "step": 22390
+    },
+    {
+      "epoch": 4.719018275662295,
+      "grad_norm": 1.0761774778366089,
+      "learning_rate": 0.00010879250475429523,
+      "loss": 1.6331,
+      "step": 22400
+    },
+    {
+      "epoch": 4.721124980249645,
+      "grad_norm": 1.0743370056152344,
+      "learning_rate": 0.00010872656466258328,
+      "loss": 1.6635,
+      "step": 22410
+    },
+    {
+      "epoch": 4.7232316848369935,
+      "grad_norm": 0.9481725096702576,
+      "learning_rate": 0.00010866062074713825,
+      "loss": 1.6565,
+      "step": 22420
+    },
+    {
+      "epoch": 4.725338389424343,
+      "grad_norm": 1.148763656616211,
+      "learning_rate": 0.00010859467303685482,
+      "loss": 1.6608,
+      "step": 22430
+    },
+    {
+      "epoch": 4.727445094011692,
+      "grad_norm": 1.1060622930526733,
+      "learning_rate": 0.00010852872156062946,
+      "loss": 1.6188,
+      "step": 22440
+    },
+    {
+      "epoch": 4.729551798599042,
+      "grad_norm": 1.1848762035369873,
+      "learning_rate": 0.00010846276634736021,
+      "loss": 1.5904,
+      "step": 22450
+    },
+    {
+      "epoch": 4.731658503186391,
+      "grad_norm": 0.9851653575897217,
+      "learning_rate": 0.00010839680742594678,
+      "loss": 1.6341,
+      "step": 22460
+    },
+    {
+      "epoch": 4.7337652077737395,
+      "grad_norm": 0.9424974322319031,
+      "learning_rate": 0.00010833084482529048,
+      "loss": 1.6531,
+      "step": 22470
+    },
+    {
+      "epoch": 4.735871912361089,
+      "grad_norm": 0.9752402901649475,
+      "learning_rate": 0.00010826487857429428,
+      "loss": 1.6471,
+      "step": 22480
+    },
+    {
+      "epoch": 4.737978616948438,
+      "grad_norm": 1.0955601930618286,
+      "learning_rate": 0.00010819890870186271,
+      "loss": 1.6078,
+      "step": 22490
+    },
+    {
+      "epoch": 4.740085321535788,
+      "grad_norm": 1.05659818649292,
+      "learning_rate": 0.00010813293523690191,
+      "loss": 1.6918,
+      "step": 22500
+    },
+    {
+      "epoch": 4.742192026123137,
+      "grad_norm": 1.0022358894348145,
+      "learning_rate": 0.00010806695820831954,
+      "loss": 1.6364,
+      "step": 22510
+    },
+    {
+      "epoch": 4.7442987307104865,
+      "grad_norm": 1.1091896295547485,
+      "learning_rate": 0.00010800097764502491,
+      "loss": 1.6751,
+      "step": 22520
+    },
+    {
+      "epoch": 4.746405435297835,
+      "grad_norm": 0.9929733276367188,
+      "learning_rate": 0.0001079349935759288,
+      "loss": 1.6447,
+      "step": 22530
+    },
+    {
+      "epoch": 4.748512139885184,
+      "grad_norm": 1.0926586389541626,
+      "learning_rate": 0.00010786900602994359,
+      "loss": 1.7089,
+      "step": 22540
+    },
+    {
+      "epoch": 4.750618844472534,
+      "grad_norm": 1.0213204622268677,
+      "learning_rate": 0.00010780301503598306,
+      "loss": 1.6099,
+      "step": 22550
+    },
+    {
+      "epoch": 4.752725549059883,
+      "grad_norm": 0.9645788669586182,
+      "learning_rate": 0.00010773702062296273,
+      "loss": 1.6107,
+      "step": 22560
+    },
+    {
+      "epoch": 4.7548322536472325,
+      "grad_norm": 1.0832529067993164,
+      "learning_rate": 0.00010767102281979939,
+      "loss": 1.6678,
+      "step": 22570
+    },
+    {
+      "epoch": 4.756938958234581,
+      "grad_norm": 0.9932439923286438,
+      "learning_rate": 0.00010760502165541144,
+      "loss": 1.6744,
+      "step": 22580
+    },
+    {
+      "epoch": 4.759045662821931,
+      "grad_norm": 1.130690574645996,
+      "learning_rate": 0.00010753901715871866,
+      "loss": 1.6977,
+      "step": 22590
+    },
+    {
+      "epoch": 4.76115236740928,
+      "grad_norm": 1.1236021518707275,
+      "learning_rate": 0.00010747300935864243,
+      "loss": 1.6445,
+      "step": 22600
+    },
+    {
+      "epoch": 4.763259071996629,
+      "grad_norm": 1.0128141641616821,
+      "learning_rate": 0.00010740699828410545,
+      "loss": 1.6873,
+      "step": 22610
+    },
+    {
+      "epoch": 4.7653657765839785,
+      "grad_norm": 0.9882675409317017,
+      "learning_rate": 0.00010734098396403192,
+      "loss": 1.5715,
+      "step": 22620
+    },
+    {
+      "epoch": 4.767472481171327,
+      "grad_norm": 1.1569823026657104,
+      "learning_rate": 0.0001072749664273474,
+      "loss": 1.6743,
+      "step": 22630
+    },
+    {
+      "epoch": 4.769579185758677,
+      "grad_norm": 1.044657588005066,
+      "learning_rate": 0.00010720894570297897,
+      "loss": 1.5966,
+      "step": 22640
+    },
+    {
+      "epoch": 4.771685890346026,
+      "grad_norm": 1.0096189975738525,
+      "learning_rate": 0.00010714292181985498,
+      "loss": 1.6633,
+      "step": 22650
+    },
+    {
+      "epoch": 4.773792594933376,
+      "grad_norm": 1.0168404579162598,
+      "learning_rate": 0.00010707689480690526,
+      "loss": 1.6276,
+      "step": 22660
+    },
+    {
+      "epoch": 4.775899299520725,
+      "grad_norm": 1.054799199104309,
+      "learning_rate": 0.00010701086469306096,
+      "loss": 1.6701,
+      "step": 22670
+    },
+    {
+      "epoch": 4.778006004108074,
+      "grad_norm": 0.9628526568412781,
+      "learning_rate": 0.00010694483150725458,
+      "loss": 1.6597,
+      "step": 22680
+    },
+    {
+      "epoch": 4.780112708695423,
+      "grad_norm": 1.0607578754425049,
+      "learning_rate": 0.00010687879527842007,
+      "loss": 1.6841,
+      "step": 22690
+    },
+    {
+      "epoch": 4.782219413282773,
+      "grad_norm": 1.0231002569198608,
+      "learning_rate": 0.00010681275603549252,
+      "loss": 1.6606,
+      "step": 22700
+    },
+    {
+      "epoch": 4.784326117870122,
+      "grad_norm": 1.0138951539993286,
+      "learning_rate": 0.00010674671380740851,
+      "loss": 1.6407,
+      "step": 22710
+    },
+    {
+      "epoch": 4.786432822457471,
+      "grad_norm": 1.051766276359558,
+      "learning_rate": 0.00010668066862310589,
+      "loss": 1.6696,
+      "step": 22720
+    },
+    {
+      "epoch": 4.78853952704482,
+      "grad_norm": 1.0689328908920288,
+      "learning_rate": 0.00010661462051152376,
+      "loss": 1.6823,
+      "step": 22730
+    },
+    {
+      "epoch": 4.790646231632169,
+      "grad_norm": 1.0172207355499268,
+      "learning_rate": 0.00010654856950160253,
+      "loss": 1.6111,
+      "step": 22740
+    },
+    {
+      "epoch": 4.792752936219519,
+      "grad_norm": 1.068111538887024,
+      "learning_rate": 0.00010648251562228386,
+      "loss": 1.6345,
+      "step": 22750
+    },
+    {
+      "epoch": 4.794859640806868,
+      "grad_norm": 1.1267240047454834,
+      "learning_rate": 0.00010641645890251075,
+      "loss": 1.6202,
+      "step": 22760
+    },
+    {
+      "epoch": 4.7969663453942175,
+      "grad_norm": 1.025101900100708,
+      "learning_rate": 0.00010635039937122733,
+      "loss": 1.5892,
+      "step": 22770
+    },
+    {
+      "epoch": 4.799073049981566,
+      "grad_norm": 1.0108810663223267,
+      "learning_rate": 0.00010628433705737901,
+      "loss": 1.6042,
+      "step": 22780
+    },
+    {
+      "epoch": 4.801179754568915,
+      "grad_norm": 1.0703437328338623,
+      "learning_rate": 0.00010621827198991245,
+      "loss": 1.6693,
+      "step": 22790
+    },
+    {
+      "epoch": 4.803286459156265,
+      "grad_norm": 1.0624349117279053,
+      "learning_rate": 0.00010615220419777548,
+      "loss": 1.6896,
+      "step": 22800
+    },
+    {
+      "epoch": 4.805393163743614,
+      "grad_norm": 1.0351321697235107,
+      "learning_rate": 0.00010608613370991711,
+      "loss": 1.6833,
+      "step": 22810
+    },
+    {
+      "epoch": 4.8074998683309635,
+      "grad_norm": 1.0743846893310547,
+      "learning_rate": 0.0001060200605552876,
+      "loss": 1.608,
+      "step": 22820
+    },
+    {
+      "epoch": 4.809606572918312,
+      "grad_norm": 1.0879895687103271,
+      "learning_rate": 0.00010595398476283827,
+      "loss": 1.6452,
+      "step": 22830
+    },
+    {
+      "epoch": 4.811713277505662,
+      "grad_norm": 1.1424341201782227,
+      "learning_rate": 0.00010588790636152168,
+      "loss": 1.7068,
+      "step": 22840
+    },
+    {
+      "epoch": 4.813819982093011,
+      "grad_norm": 1.0059975385665894,
+      "learning_rate": 0.00010582182538029157,
+      "loss": 1.6526,
+      "step": 22850
+    },
+    {
+      "epoch": 4.81592668668036,
+      "grad_norm": 0.9729864597320557,
+      "learning_rate": 0.00010575574184810269,
+      "loss": 1.6288,
+      "step": 22860
+    },
+    {
+      "epoch": 4.81803339126771,
+      "grad_norm": 1.1198335886001587,
+      "learning_rate": 0.00010568965579391097,
+      "loss": 1.589,
+      "step": 22870
+    },
+    {
+      "epoch": 4.820140095855058,
+      "grad_norm": 1.0577073097229004,
+      "learning_rate": 0.00010562356724667346,
+      "loss": 1.5819,
+      "step": 22880
+    },
+    {
+      "epoch": 4.822246800442408,
+      "grad_norm": 1.1223348379135132,
+      "learning_rate": 0.00010555747623534831,
+      "loss": 1.6413,
+      "step": 22890
+    },
+    {
+      "epoch": 4.824353505029757,
+      "grad_norm": 1.072959065437317,
+      "learning_rate": 0.00010549138278889468,
+      "loss": 1.6769,
+      "step": 22900
+    },
+    {
+      "epoch": 4.826460209617107,
+      "grad_norm": 1.004937767982483,
+      "learning_rate": 0.00010542528693627287,
+      "loss": 1.6091,
+      "step": 22910
+    },
+    {
+      "epoch": 4.828566914204456,
+      "grad_norm": 1.094373345375061,
+      "learning_rate": 0.0001053591887064442,
+      "loss": 1.6446,
+      "step": 22920
+    },
+    {
+      "epoch": 4.8306736187918045,
+      "grad_norm": 1.0144459009170532,
+      "learning_rate": 0.00010529308812837104,
+      "loss": 1.6605,
+      "step": 22930
+    },
+    {
+      "epoch": 4.832780323379154,
+      "grad_norm": 1.0409224033355713,
+      "learning_rate": 0.00010522698523101682,
+      "loss": 1.6474,
+      "step": 22940
+    },
+    {
+      "epoch": 4.834887027966503,
+      "grad_norm": 1.112221598625183,
+      "learning_rate": 0.0001051608800433459,
+      "loss": 1.7047,
+      "step": 22950
+    },
+    {
+      "epoch": 4.836993732553853,
+      "grad_norm": 1.0993469953536987,
+      "learning_rate": 0.00010509477259432372,
+      "loss": 1.6449,
+      "step": 22960
+    },
+    {
+      "epoch": 4.839100437141202,
+      "grad_norm": 1.0196629762649536,
+      "learning_rate": 0.00010502866291291668,
+      "loss": 1.7036,
+      "step": 22970
+    },
+    {
+      "epoch": 4.841207141728551,
+      "grad_norm": 1.0654680728912354,
+      "learning_rate": 0.00010496255102809223,
+      "loss": 1.6403,
+      "step": 22980
+    },
+    {
+      "epoch": 4.8433138463159,
+      "grad_norm": 0.9815942049026489,
+      "learning_rate": 0.00010489643696881864,
+      "loss": 1.6527,
+      "step": 22990
+    },
+    {
+      "epoch": 4.84542055090325,
+      "grad_norm": 1.2717540264129639,
+      "learning_rate": 0.00010483032076406528,
+      "loss": 1.6497,
+      "step": 23000
+    },
+    {
+      "epoch": 4.847527255490599,
+      "grad_norm": 1.1045358180999756,
+      "learning_rate": 0.00010476420244280232,
+      "loss": 1.677,
+      "step": 23010
+    },
+    {
+      "epoch": 4.849633960077949,
+      "grad_norm": 0.9338797926902771,
+      "learning_rate": 0.00010469808203400102,
+      "loss": 1.6032,
+      "step": 23020
+    },
+    {
+      "epoch": 4.851740664665297,
+      "grad_norm": 0.9959046840667725,
+      "learning_rate": 0.00010463195956663338,
+      "loss": 1.5908,
+      "step": 23030
+    },
+    {
+      "epoch": 4.853847369252646,
+      "grad_norm": 1.0763238668441772,
+      "learning_rate": 0.00010456583506967248,
+      "loss": 1.635,
+      "step": 23040
+    },
+    {
+      "epoch": 4.855954073839996,
+      "grad_norm": 1.1013065576553345,
+      "learning_rate": 0.00010449970857209213,
+      "loss": 1.6311,
+      "step": 23050
+    },
+    {
+      "epoch": 4.858060778427345,
+      "grad_norm": 1.1367875337600708,
+      "learning_rate": 0.00010443358010286714,
+      "loss": 1.6263,
+      "step": 23060
+    },
+    {
+      "epoch": 4.860167483014695,
+      "grad_norm": 1.066049337387085,
+      "learning_rate": 0.00010436744969097306,
+      "loss": 1.6411,
+      "step": 23070
+    },
+    {
+      "epoch": 4.862274187602043,
+      "grad_norm": 0.9799880385398865,
+      "learning_rate": 0.00010430131736538644,
+      "loss": 1.6492,
+      "step": 23080
+    },
+    {
+      "epoch": 4.864380892189393,
+      "grad_norm": 1.122512936592102,
+      "learning_rate": 0.00010423518315508452,
+      "loss": 1.6537,
+      "step": 23090
+    },
+    {
+      "epoch": 4.866487596776742,
+      "grad_norm": 0.9705328941345215,
+      "learning_rate": 0.00010416904708904548,
+      "loss": 1.6212,
+      "step": 23100
+    },
+    {
+      "epoch": 4.868594301364091,
+      "grad_norm": 1.0140635967254639,
+      "learning_rate": 0.00010410290919624824,
+      "loss": 1.6039,
+      "step": 23110
+    },
+    {
+      "epoch": 4.870701005951441,
+      "grad_norm": 1.1238670349121094,
+      "learning_rate": 0.00010403676950567258,
+      "loss": 1.6271,
+      "step": 23120
+    },
+    {
+      "epoch": 4.8728077105387895,
+      "grad_norm": 1.032712697982788,
+      "learning_rate": 0.000103970628046299,
+      "loss": 1.6567,
+      "step": 23130
+    },
+    {
+      "epoch": 4.874914415126139,
+      "grad_norm": 1.0045222043991089,
+      "learning_rate": 0.00010390448484710886,
+      "loss": 1.6499,
+      "step": 23140
+    },
+    {
+      "epoch": 4.877021119713488,
+      "grad_norm": 1.1023186445236206,
+      "learning_rate": 0.00010383833993708416,
+      "loss": 1.6053,
+      "step": 23150
+    },
+    {
+      "epoch": 4.879127824300838,
+      "grad_norm": 1.5840126276016235,
+      "learning_rate": 0.00010377219334520783,
+      "loss": 1.6386,
+      "step": 23160
+    },
+    {
+      "epoch": 4.881234528888187,
+      "grad_norm": 1.037644386291504,
+      "learning_rate": 0.00010370604510046331,
+      "loss": 1.6591,
+      "step": 23170
+    },
+    {
+      "epoch": 4.8833412334755355,
+      "grad_norm": 1.0414832830429077,
+      "learning_rate": 0.00010363989523183495,
+      "loss": 1.6327,
+      "step": 23180
+    },
+    {
+      "epoch": 4.885447938062885,
+      "grad_norm": 1.0389552116394043,
+      "learning_rate": 0.00010357374376830775,
+      "loss": 1.668,
+      "step": 23190
+    },
+    {
+      "epoch": 4.887554642650234,
+      "grad_norm": 1.1060019731521606,
+      "learning_rate": 0.00010350759073886739,
+      "loss": 1.6629,
+      "step": 23200
+    },
+    {
+      "epoch": 4.889661347237584,
+      "grad_norm": 0.971713662147522,
+      "learning_rate": 0.00010344143617250029,
+      "loss": 1.644,
+      "step": 23210
+    },
+    {
+      "epoch": 4.891768051824933,
+      "grad_norm": 1.0252668857574463,
+      "learning_rate": 0.00010337528009819344,
+      "loss": 1.6615,
+      "step": 23220
+    },
+    {
+      "epoch": 4.893874756412282,
+      "grad_norm": 0.9966990947723389,
+      "learning_rate": 0.00010330912254493456,
+      "loss": 1.6698,
+      "step": 23230
+    },
+    {
+      "epoch": 4.895981460999631,
+      "grad_norm": 1.0967049598693848,
+      "learning_rate": 0.00010324296354171207,
+      "loss": 1.6607,
+      "step": 23240
+    },
+    {
+      "epoch": 4.89808816558698,
+      "grad_norm": 1.0470962524414062,
+      "learning_rate": 0.00010317680311751496,
+      "loss": 1.6537,
+      "step": 23250
+    },
+    {
+      "epoch": 4.90019487017433,
+      "grad_norm": 1.0356905460357666,
+      "learning_rate": 0.00010311064130133279,
+      "loss": 1.6644,
+      "step": 23260
+    },
+    {
+      "epoch": 4.902301574761679,
+      "grad_norm": 1.1939691305160522,
+      "learning_rate": 0.00010304447812215582,
+      "loss": 1.6593,
+      "step": 23270
+    },
+    {
+      "epoch": 4.9044082793490285,
+      "grad_norm": 1.1555665731430054,
+      "learning_rate": 0.00010297831360897492,
+      "loss": 1.7027,
+      "step": 23280
+    },
+    {
+      "epoch": 4.906514983936377,
+      "grad_norm": 1.1872739791870117,
+      "learning_rate": 0.00010291214779078149,
+      "loss": 1.6771,
+      "step": 23290
+    },
+    {
+      "epoch": 4.908621688523727,
+      "grad_norm": 1.0362814664840698,
+      "learning_rate": 0.00010284598069656746,
+      "loss": 1.641,
+      "step": 23300
+    },
+    {
+      "epoch": 4.910728393111076,
+      "grad_norm": 1.0210367441177368,
+      "learning_rate": 0.00010277981235532541,
+      "loss": 1.6396,
+      "step": 23310
+    },
+    {
+      "epoch": 4.912835097698425,
+      "grad_norm": 1.0176624059677124,
+      "learning_rate": 0.00010271364279604842,
+      "loss": 1.6409,
+      "step": 23320
+    },
+    {
+      "epoch": 4.9149418022857745,
+      "grad_norm": 0.9932822585105896,
+      "learning_rate": 0.00010264747204773018,
+      "loss": 1.6659,
+      "step": 23330
+    },
+    {
+      "epoch": 4.917048506873123,
+      "grad_norm": 1.0647072792053223,
+      "learning_rate": 0.00010258130013936474,
+      "loss": 1.6686,
+      "step": 23340
+    },
+    {
+      "epoch": 4.919155211460473,
+      "grad_norm": 1.2044399976730347,
+      "learning_rate": 0.0001025151270999468,
+      "loss": 1.6833,
+      "step": 23350
+    },
+    {
+      "epoch": 4.921261916047822,
+      "grad_norm": 0.9959510564804077,
+      "learning_rate": 0.00010244895295847147,
+      "loss": 1.6623,
+      "step": 23360
+    },
+    {
+      "epoch": 4.923368620635172,
+      "grad_norm": 1.0841436386108398,
+      "learning_rate": 0.00010238277774393448,
+      "loss": 1.6428,
+      "step": 23370
+    },
+    {
+      "epoch": 4.9254753252225205,
+      "grad_norm": 1.029536247253418,
+      "learning_rate": 0.00010231660148533183,
+      "loss": 1.6339,
+      "step": 23380
+    },
+    {
+      "epoch": 4.92758202980987,
+      "grad_norm": 1.031693458557129,
+      "learning_rate": 0.0001022504242116601,
+      "loss": 1.6584,
+      "step": 23390
+    },
+    {
+      "epoch": 4.929688734397219,
+      "grad_norm": 1.041793942451477,
+      "learning_rate": 0.00010218424595191631,
+      "loss": 1.6233,
+      "step": 23400
+    },
+    {
+      "epoch": 4.931795438984569,
+      "grad_norm": 1.0688579082489014,
+      "learning_rate": 0.00010211806673509794,
+      "loss": 1.6484,
+      "step": 23410
+    },
+    {
+      "epoch": 4.933902143571918,
+      "grad_norm": 0.9542795419692993,
+      "learning_rate": 0.00010205188659020275,
+      "loss": 1.6047,
+      "step": 23420
+    },
+    {
+      "epoch": 4.9360088481592665,
+      "grad_norm": 0.9721381068229675,
+      "learning_rate": 0.00010198570554622909,
+      "loss": 1.7019,
+      "step": 23430
+    },
+    {
+      "epoch": 4.938115552746616,
+      "grad_norm": 1.0544655323028564,
+      "learning_rate": 0.00010191952363217557,
+      "loss": 1.6674,
+      "step": 23440
+    },
+    {
+      "epoch": 4.940222257333965,
+      "grad_norm": 0.9668622612953186,
+      "learning_rate": 0.00010185334087704124,
+      "loss": 1.5844,
+      "step": 23450
+    },
+    {
+      "epoch": 4.942328961921315,
+      "grad_norm": 0.9595541954040527,
+      "learning_rate": 0.00010178715730982549,
+      "loss": 1.6182,
+      "step": 23460
+    },
+    {
+      "epoch": 4.944435666508664,
+      "grad_norm": 1.1048792600631714,
+      "learning_rate": 0.00010172097295952811,
+      "loss": 1.6485,
+      "step": 23470
+    },
+    {
+      "epoch": 4.9465423710960135,
+      "grad_norm": 1.0720999240875244,
+      "learning_rate": 0.00010165478785514919,
+      "loss": 1.7186,
+      "step": 23480
+    },
+    {
+      "epoch": 4.948649075683362,
+      "grad_norm": 1.021140456199646,
+      "learning_rate": 0.00010158860202568916,
+      "loss": 1.6569,
+      "step": 23490
+    },
+    {
+      "epoch": 4.950755780270711,
+      "grad_norm": 1.0183557271957397,
+      "learning_rate": 0.00010152241550014881,
+      "loss": 1.6664,
+      "step": 23500
+    },
+    {
+      "epoch": 4.952862484858061,
+      "grad_norm": 1.0429364442825317,
+      "learning_rate": 0.00010145622830752912,
+      "loss": 1.6682,
+      "step": 23510
+    },
+    {
+      "epoch": 4.95496918944541,
+      "grad_norm": 1.113517165184021,
+      "learning_rate": 0.00010139004047683151,
+      "loss": 1.6062,
+      "step": 23520
+    },
+    {
+      "epoch": 4.9570758940327595,
+      "grad_norm": 1.080854058265686,
+      "learning_rate": 0.00010132385203705761,
+      "loss": 1.5665,
+      "step": 23530
+    },
+    {
+      "epoch": 4.959182598620108,
+      "grad_norm": 1.0921235084533691,
+      "learning_rate": 0.00010125766301720929,
+      "loss": 1.6353,
+      "step": 23540
+    },
+    {
+      "epoch": 4.961289303207458,
+      "grad_norm": 0.9792695641517639,
+      "learning_rate": 0.0001011914734462887,
+      "loss": 1.6395,
+      "step": 23550
+    },
+    {
+      "epoch": 4.963396007794807,
+      "grad_norm": 0.903640866279602,
+      "learning_rate": 0.00010112528335329823,
+      "loss": 1.5959,
+      "step": 23560
+    },
+    {
+      "epoch": 4.965502712382156,
+      "grad_norm": 1.024780511856079,
+      "learning_rate": 0.0001010590927672405,
+      "loss": 1.6883,
+      "step": 23570
+    },
+    {
+      "epoch": 4.9676094169695055,
+      "grad_norm": 0.9267536401748657,
+      "learning_rate": 0.00010099290171711841,
+      "loss": 1.6925,
+      "step": 23580
+    },
+    {
+      "epoch": 4.969716121556854,
+      "grad_norm": 1.0130085945129395,
+      "learning_rate": 0.00010092671023193491,
+      "loss": 1.6629,
+      "step": 23590
+    },
+    {
+      "epoch": 4.971822826144204,
+      "grad_norm": 1.0210721492767334,
+      "learning_rate": 0.00010086051834069328,
+      "loss": 1.6247,
+      "step": 23600
+    },
+    {
+      "epoch": 4.973929530731553,
+      "grad_norm": 1.1113983392715454,
+      "learning_rate": 0.00010079432607239692,
+      "loss": 1.6352,
+      "step": 23610
+    },
+    {
+      "epoch": 4.976036235318903,
+      "grad_norm": 1.0311906337738037,
+      "learning_rate": 0.00010072813345604941,
+      "loss": 1.6338,
+      "step": 23620
+    },
+    {
+      "epoch": 4.978142939906252,
+      "grad_norm": 1.0975770950317383,
+      "learning_rate": 0.00010066194052065446,
+      "loss": 1.6354,
+      "step": 23630
+    },
+    {
+      "epoch": 4.9802496444936,
+      "grad_norm": 1.080855131149292,
+      "learning_rate": 0.00010059574729521595,
+      "loss": 1.6562,
+      "step": 23640
+    },
+    {
+      "epoch": 4.98235634908095,
+      "grad_norm": 1.0517081022262573,
+      "learning_rate": 0.00010052955380873788,
+      "loss": 1.6971,
+      "step": 23650
+    },
+    {
+      "epoch": 4.984463053668299,
+      "grad_norm": 1.005123257637024,
+      "learning_rate": 0.00010046336009022435,
+      "loss": 1.6412,
+      "step": 23660
+    },
+    {
+      "epoch": 4.986569758255649,
+      "grad_norm": 1.0352355241775513,
+      "learning_rate": 0.00010039716616867957,
+      "loss": 1.6435,
+      "step": 23670
+    },
+    {
+      "epoch": 4.988676462842998,
+      "grad_norm": 1.0361087322235107,
+      "learning_rate": 0.00010033097207310785,
+      "loss": 1.647,
+      "step": 23680
+    },
+    {
+      "epoch": 4.990783167430347,
+      "grad_norm": 0.9913681745529175,
+      "learning_rate": 0.00010026477783251351,
+      "loss": 1.6346,
+      "step": 23690
+    },
+    {
+      "epoch": 4.992889872017696,
+      "grad_norm": 1.0833758115768433,
+      "learning_rate": 0.0001001985834759011,
+      "loss": 1.6415,
+      "step": 23700
+    },
+    {
+      "epoch": 4.994996576605046,
+      "grad_norm": 1.0982452630996704,
+      "learning_rate": 0.00010013238903227499,
+      "loss": 1.6706,
+      "step": 23710
+    },
+    {
+      "epoch": 4.997103281192395,
+      "grad_norm": 0.9679703116416931,
+      "learning_rate": 0.00010006619453063979,
+      "loss": 1.6587,
+      "step": 23720
+    },
+    {
+      "epoch": 4.9992099857797445,
+      "grad_norm": 1.0053178071975708,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 23730
+    },
+    {
+      "epoch": 5.001316690367093,
+      "grad_norm": 1.0724413394927979,
+      "learning_rate": 9.993380546936023e-05,
+      "loss": 1.6388,
+      "step": 23740
+    },
+    {
+      "epoch": 5.003423394954442,
+      "grad_norm": 1.3603681325912476,
+      "learning_rate": 9.986761096772502e-05,
+      "loss": 1.5901,
+      "step": 23750
+    },
+    {
+      "epoch": 5.005530099541792,
+      "grad_norm": 1.0234671831130981,
+      "learning_rate": 9.980141652409895e-05,
+      "loss": 1.667,
+      "step": 23760
+    },
+    {
+      "epoch": 5.007636804129141,
+      "grad_norm": 0.9571532607078552,
+      "learning_rate": 9.97352221674865e-05,
+      "loss": 1.5851,
+      "step": 23770
+    },
+    {
+      "epoch": 5.0097435087164905,
+      "grad_norm": 1.0448963642120361,
+      "learning_rate": 9.966902792689219e-05,
+      "loss": 1.557,
+      "step": 23780
+    },
+    {
+      "epoch": 5.011850213303839,
+      "grad_norm": 1.0232895612716675,
+      "learning_rate": 9.960283383132045e-05,
+      "loss": 1.6287,
+      "step": 23790
+    },
+    {
+      "epoch": 5.013956917891189,
+      "grad_norm": 1.0618042945861816,
+      "learning_rate": 9.953663990977568e-05,
+      "loss": 1.5744,
+      "step": 23800
+    },
+    {
+      "epoch": 5.016063622478538,
+      "grad_norm": 1.0534571409225464,
+      "learning_rate": 9.947044619126216e-05,
+      "loss": 1.694,
+      "step": 23810
+    },
+    {
+      "epoch": 5.018170327065887,
+      "grad_norm": 1.0805995464324951,
+      "learning_rate": 9.940425270478407e-05,
+      "loss": 1.6219,
+      "step": 23820
+    },
+    {
+      "epoch": 5.020277031653237,
+      "grad_norm": 0.9920625686645508,
+      "learning_rate": 9.933805947934555e-05,
+      "loss": 1.5763,
+      "step": 23830
+    },
+    {
+      "epoch": 5.022383736240585,
+      "grad_norm": 1.4843225479125977,
+      "learning_rate": 9.927186654395063e-05,
+      "loss": 1.6695,
+      "step": 23840
+    },
+    {
+      "epoch": 5.024490440827935,
+      "grad_norm": 1.0274052619934082,
+      "learning_rate": 9.920567392760312e-05,
+      "loss": 1.5961,
+      "step": 23850
+    },
+    {
+      "epoch": 5.026597145415284,
+      "grad_norm": 1.0526154041290283,
+      "learning_rate": 9.913948165930676e-05,
+      "loss": 1.6448,
+      "step": 23860
+    },
+    {
+      "epoch": 5.028703850002634,
+      "grad_norm": 1.0544629096984863,
+      "learning_rate": 9.907328976806511e-05,
+      "loss": 1.6435,
+      "step": 23870
+    },
+    {
+      "epoch": 5.030810554589983,
+      "grad_norm": 1.0605446100234985,
+      "learning_rate": 9.900709828288164e-05,
+      "loss": 1.6284,
+      "step": 23880
+    },
+    {
+      "epoch": 5.0329172591773315,
+      "grad_norm": 1.1349766254425049,
+      "learning_rate": 9.894090723275951e-05,
+      "loss": 1.6044,
+      "step": 23890
+    },
+    {
+      "epoch": 5.035023963764681,
+      "grad_norm": 1.1093242168426514,
+      "learning_rate": 9.88747166467018e-05,
+      "loss": 1.6237,
+      "step": 23900
+    },
+    {
+      "epoch": 5.03713066835203,
+      "grad_norm": 1.014807939529419,
+      "learning_rate": 9.880852655371134e-05,
+      "loss": 1.5961,
+      "step": 23910
+    },
+    {
+      "epoch": 5.03923737293938,
+      "grad_norm": 1.281245231628418,
+      "learning_rate": 9.874233698279075e-05,
+      "loss": 1.5518,
+      "step": 23920
+    },
+    {
+      "epoch": 5.041344077526729,
+      "grad_norm": 1.0977730751037598,
+      "learning_rate": 9.867614796294242e-05,
+      "loss": 1.6002,
+      "step": 23930
+    },
+    {
+      "epoch": 5.043450782114078,
+      "grad_norm": 1.0470744371414185,
+      "learning_rate": 9.860995952316851e-05,
+      "loss": 1.6183,
+      "step": 23940
+    },
+    {
+      "epoch": 5.045557486701427,
+      "grad_norm": 1.046972393989563,
+      "learning_rate": 9.854377169247089e-05,
+      "loss": 1.5417,
+      "step": 23950
+    },
+    {
+      "epoch": 5.047664191288777,
+      "grad_norm": 1.1596167087554932,
+      "learning_rate": 9.847758449985124e-05,
+      "loss": 1.6612,
+      "step": 23960
+    },
+    {
+      "epoch": 5.049770895876126,
+      "grad_norm": 1.0991291999816895,
+      "learning_rate": 9.841139797431087e-05,
+      "loss": 1.5882,
+      "step": 23970
+    },
+    {
+      "epoch": 5.051877600463475,
+      "grad_norm": 1.0760891437530518,
+      "learning_rate": 9.834521214485083e-05,
+      "loss": 1.5909,
+      "step": 23980
+    },
+    {
+      "epoch": 5.053984305050824,
+      "grad_norm": 0.985348105430603,
+      "learning_rate": 9.827902704047191e-05,
+      "loss": 1.587,
+      "step": 23990
+    },
+    {
+      "epoch": 5.056091009638173,
+      "grad_norm": 1.231652021408081,
+      "learning_rate": 9.821284269017455e-05,
+      "loss": 1.6468,
+      "step": 24000
+    },
+    {
+      "epoch": 5.058197714225523,
+      "grad_norm": 1.148422122001648,
+      "learning_rate": 9.81466591229588e-05,
+      "loss": 1.6027,
+      "step": 24010
+    },
+    {
+      "epoch": 5.060304418812872,
+      "grad_norm": 1.0643984079360962,
+      "learning_rate": 9.808047636782447e-05,
+      "loss": 1.6109,
+      "step": 24020
+    },
+    {
+      "epoch": 5.062411123400222,
+      "grad_norm": 1.1943291425704956,
+      "learning_rate": 9.801429445377094e-05,
+      "loss": 1.7016,
+      "step": 24030
+    },
+    {
+      "epoch": 5.06451782798757,
+      "grad_norm": 1.0921201705932617,
+      "learning_rate": 9.794811340979724e-05,
+      "loss": 1.6567,
+      "step": 24040
+    },
+    {
+      "epoch": 5.066624532574919,
+      "grad_norm": 1.1693758964538574,
+      "learning_rate": 9.788193326490212e-05,
+      "loss": 1.6591,
+      "step": 24050
+    },
+    {
+      "epoch": 5.068731237162269,
+      "grad_norm": 1.1490429639816284,
+      "learning_rate": 9.781575404808371e-05,
+      "loss": 1.6554,
+      "step": 24060
+    },
+    {
+      "epoch": 5.070837941749618,
+      "grad_norm": 1.0068416595458984,
+      "learning_rate": 9.774957578833993e-05,
+      "loss": 1.6122,
+      "step": 24070
+    },
+    {
+      "epoch": 5.072944646336968,
+      "grad_norm": 0.9988663196563721,
+      "learning_rate": 9.768339851466818e-05,
+      "loss": 1.604,
+      "step": 24080
+    },
+    {
+      "epoch": 5.0750513509243165,
+      "grad_norm": 1.1303640604019165,
+      "learning_rate": 9.761722225606557e-05,
+      "loss": 1.585,
+      "step": 24090
+    },
+    {
+      "epoch": 5.077158055511666,
+      "grad_norm": 1.080378770828247,
+      "learning_rate": 9.755104704152854e-05,
+      "loss": 1.6062,
+      "step": 24100
+    },
+    {
+      "epoch": 5.079264760099015,
+      "grad_norm": 1.0893718004226685,
+      "learning_rate": 9.748487290005324e-05,
+      "loss": 1.6053,
+      "step": 24110
+    },
+    {
+      "epoch": 5.081371464686365,
+      "grad_norm": 1.0815832614898682,
+      "learning_rate": 9.741869986063526e-05,
+      "loss": 1.6045,
+      "step": 24120
+    },
+    {
+      "epoch": 5.083478169273714,
+      "grad_norm": 1.0142110586166382,
+      "learning_rate": 9.735252795226987e-05,
+      "loss": 1.5406,
+      "step": 24130
+    },
+    {
+      "epoch": 5.0855848738610625,
+      "grad_norm": 1.080077886581421,
+      "learning_rate": 9.728635720395159e-05,
+      "loss": 1.6617,
+      "step": 24140
+    },
+    {
+      "epoch": 5.087691578448412,
+      "grad_norm": 1.115033507347107,
+      "learning_rate": 9.722018764467461e-05,
+      "loss": 1.6468,
+      "step": 24150
+    },
+    {
+      "epoch": 5.089798283035761,
+      "grad_norm": 1.0328733921051025,
+      "learning_rate": 9.715401930343254e-05,
+      "loss": 1.6028,
+      "step": 24160
+    },
+    {
+      "epoch": 5.091904987623111,
+      "grad_norm": 1.0720160007476807,
+      "learning_rate": 9.708785220921856e-05,
+      "loss": 1.5708,
+      "step": 24170
+    },
+    {
+      "epoch": 5.09401169221046,
+      "grad_norm": 1.17787504196167,
+      "learning_rate": 9.702168639102509e-05,
+      "loss": 1.6243,
+      "step": 24180
+    },
+    {
+      "epoch": 5.096118396797809,
+      "grad_norm": 1.1949994564056396,
+      "learning_rate": 9.695552187784419e-05,
+      "loss": 1.651,
+      "step": 24190
+    },
+    {
+      "epoch": 5.098225101385158,
+      "grad_norm": 0.9967186450958252,
+      "learning_rate": 9.68893586986672e-05,
+      "loss": 1.6253,
+      "step": 24200
+    },
+    {
+      "epoch": 5.100331805972507,
+      "grad_norm": 1.130844235420227,
+      "learning_rate": 9.682319688248509e-05,
+      "loss": 1.5494,
+      "step": 24210
+    },
+    {
+      "epoch": 5.102438510559857,
+      "grad_norm": 1.107001781463623,
+      "learning_rate": 9.675703645828794e-05,
+      "loss": 1.6392,
+      "step": 24220
+    },
+    {
+      "epoch": 5.104545215147206,
+      "grad_norm": 1.0460437536239624,
+      "learning_rate": 9.669087745506545e-05,
+      "loss": 1.5718,
+      "step": 24230
+    },
+    {
+      "epoch": 5.1066519197345555,
+      "grad_norm": 1.0819109678268433,
+      "learning_rate": 9.662471990180657e-05,
+      "loss": 1.6356,
+      "step": 24240
+    },
+    {
+      "epoch": 5.108758624321904,
+      "grad_norm": 1.12892746925354,
+      "learning_rate": 9.655856382749976e-05,
+      "loss": 1.6667,
+      "step": 24250
+    },
+    {
+      "epoch": 5.110865328909254,
+      "grad_norm": 1.0783275365829468,
+      "learning_rate": 9.649240926113262e-05,
+      "loss": 1.6301,
+      "step": 24260
+    },
+    {
+      "epoch": 5.112972033496603,
+      "grad_norm": 1.098510980606079,
+      "learning_rate": 9.642625623169226e-05,
+      "loss": 1.5735,
+      "step": 24270
+    },
+    {
+      "epoch": 5.115078738083952,
+      "grad_norm": 1.0497877597808838,
+      "learning_rate": 9.636010476816504e-05,
+      "loss": 1.5672,
+      "step": 24280
+    },
+    {
+      "epoch": 5.1171854426713015,
+      "grad_norm": 1.0647207498550415,
+      "learning_rate": 9.629395489953669e-05,
+      "loss": 1.5622,
+      "step": 24290
+    },
+    {
+      "epoch": 5.11929214725865,
+      "grad_norm": 1.110742449760437,
+      "learning_rate": 9.622780665479222e-05,
+      "loss": 1.5973,
+      "step": 24300
+    },
+    {
+      "epoch": 5.121398851846,
+      "grad_norm": 1.070984959602356,
+      "learning_rate": 9.616166006291585e-05,
+      "loss": 1.6043,
+      "step": 24310
+    },
+    {
+      "epoch": 5.123505556433349,
+      "grad_norm": 1.0085152387619019,
+      "learning_rate": 9.609551515289116e-05,
+      "loss": 1.6405,
+      "step": 24320
+    },
+    {
+      "epoch": 5.125612261020699,
+      "grad_norm": 1.1093374490737915,
+      "learning_rate": 9.602937195370099e-05,
+      "loss": 1.5763,
+      "step": 24330
+    },
+    {
+      "epoch": 5.1277189656080475,
+      "grad_norm": 1.0516998767852783,
+      "learning_rate": 9.596323049432746e-05,
+      "loss": 1.6104,
+      "step": 24340
+    },
+    {
+      "epoch": 5.129825670195397,
+      "grad_norm": 1.1160739660263062,
+      "learning_rate": 9.589709080375179e-05,
+      "loss": 1.5914,
+      "step": 24350
+    },
+    {
+      "epoch": 5.131932374782746,
+      "grad_norm": 1.3322851657867432,
+      "learning_rate": 9.583095291095453e-05,
+      "loss": 1.5964,
+      "step": 24360
+    },
+    {
+      "epoch": 5.134039079370095,
+      "grad_norm": 1.0118359327316284,
+      "learning_rate": 9.576481684491548e-05,
+      "loss": 1.607,
+      "step": 24370
+    },
+    {
+      "epoch": 5.136145783957445,
+      "grad_norm": 1.0594619512557983,
+      "learning_rate": 9.569868263461361e-05,
+      "loss": 1.5388,
+      "step": 24380
+    },
+    {
+      "epoch": 5.1382524885447936,
+      "grad_norm": 1.0311650037765503,
+      "learning_rate": 9.563255030902697e-05,
+      "loss": 1.594,
+      "step": 24390
+    },
+    {
+      "epoch": 5.140359193132143,
+      "grad_norm": 1.0667462348937988,
+      "learning_rate": 9.556641989713288e-05,
+      "loss": 1.5542,
+      "step": 24400
+    },
+    {
+      "epoch": 5.142465897719492,
+      "grad_norm": 1.0168761014938354,
+      "learning_rate": 9.550029142790786e-05,
+      "loss": 1.6031,
+      "step": 24410
+    },
+    {
+      "epoch": 5.144572602306842,
+      "grad_norm": 1.099887728691101,
+      "learning_rate": 9.543416493032757e-05,
+      "loss": 1.6045,
+      "step": 24420
+    },
+    {
+      "epoch": 5.146679306894191,
+      "grad_norm": 1.058099389076233,
+      "learning_rate": 9.536804043336664e-05,
+      "loss": 1.6458,
+      "step": 24430
+    },
+    {
+      "epoch": 5.14878601148154,
+      "grad_norm": 1.0442464351654053,
+      "learning_rate": 9.5301917965999e-05,
+      "loss": 1.5772,
+      "step": 24440
+    },
+    {
+      "epoch": 5.150892716068889,
+      "grad_norm": 1.1567966938018799,
+      "learning_rate": 9.523579755719769e-05,
+      "loss": 1.5836,
+      "step": 24450
+    },
+    {
+      "epoch": 5.152999420656238,
+      "grad_norm": 1.0755870342254639,
+      "learning_rate": 9.516967923593479e-05,
+      "loss": 1.5786,
+      "step": 24460
+    },
+    {
+      "epoch": 5.155106125243588,
+      "grad_norm": 1.0643812417984009,
+      "learning_rate": 9.51035630311814e-05,
+      "loss": 1.6168,
+      "step": 24470
+    },
+    {
+      "epoch": 5.157212829830937,
+      "grad_norm": 1.075704574584961,
+      "learning_rate": 9.503744897190778e-05,
+      "loss": 1.6217,
+      "step": 24480
+    },
+    {
+      "epoch": 5.1593195344182865,
+      "grad_norm": 1.0636301040649414,
+      "learning_rate": 9.497133708708331e-05,
+      "loss": 1.5837,
+      "step": 24490
+    },
+    {
+      "epoch": 5.161426239005635,
+      "grad_norm": 1.1529415845870972,
+      "learning_rate": 9.490522740567633e-05,
+      "loss": 1.6462,
+      "step": 24500
+    },
+    {
+      "epoch": 5.163532943592985,
+      "grad_norm": 1.1077044010162354,
+      "learning_rate": 9.483911995665414e-05,
+      "loss": 1.6843,
+      "step": 24510
+    },
+    {
+      "epoch": 5.165639648180334,
+      "grad_norm": 1.245644450187683,
+      "learning_rate": 9.477301476898322e-05,
+      "loss": 1.6067,
+      "step": 24520
+    },
+    {
+      "epoch": 5.167746352767683,
+      "grad_norm": 1.0524848699569702,
+      "learning_rate": 9.470691187162897e-05,
+      "loss": 1.5735,
+      "step": 24530
+    },
+    {
+      "epoch": 5.1698530573550325,
+      "grad_norm": 1.0744348764419556,
+      "learning_rate": 9.464081129355586e-05,
+      "loss": 1.621,
+      "step": 24540
+    },
+    {
+      "epoch": 5.171959761942381,
+      "grad_norm": 1.111391544342041,
+      "learning_rate": 9.457471306372716e-05,
+      "loss": 1.6196,
+      "step": 24550
+    },
+    {
+      "epoch": 5.174066466529731,
+      "grad_norm": 1.1415176391601562,
+      "learning_rate": 9.450861721110534e-05,
+      "loss": 1.5828,
+      "step": 24560
+    },
+    {
+      "epoch": 5.17617317111708,
+      "grad_norm": 1.121319055557251,
+      "learning_rate": 9.444252376465171e-05,
+      "loss": 1.6659,
+      "step": 24570
+    },
+    {
+      "epoch": 5.17827987570443,
+      "grad_norm": 1.061152458190918,
+      "learning_rate": 9.437643275332653e-05,
+      "loss": 1.6155,
+      "step": 24580
+    },
+    {
+      "epoch": 5.180386580291779,
+      "grad_norm": 1.0673737525939941,
+      "learning_rate": 9.431034420608906e-05,
+      "loss": 1.6082,
+      "step": 24590
+    },
+    {
+      "epoch": 5.182493284879127,
+      "grad_norm": 1.1684726476669312,
+      "learning_rate": 9.424425815189733e-05,
+      "loss": 1.6265,
+      "step": 24600
+    },
+    {
+      "epoch": 5.184599989466477,
+      "grad_norm": 1.05269455909729,
+      "learning_rate": 9.417817461970844e-05,
+      "loss": 1.6044,
+      "step": 24610
+    },
+    {
+      "epoch": 5.186706694053826,
+      "grad_norm": 1.1382789611816406,
+      "learning_rate": 9.41120936384783e-05,
+      "loss": 1.5781,
+      "step": 24620
+    },
+    {
+      "epoch": 5.188813398641176,
+      "grad_norm": 1.5912727117538452,
+      "learning_rate": 9.404601523716175e-05,
+      "loss": 1.6607,
+      "step": 24630
+    },
+    {
+      "epoch": 5.190920103228525,
+      "grad_norm": 1.1234204769134521,
+      "learning_rate": 9.397993944471244e-05,
+      "loss": 1.6193,
+      "step": 24640
+    },
+    {
+      "epoch": 5.193026807815874,
+      "grad_norm": 1.079758882522583,
+      "learning_rate": 9.39138662900829e-05,
+      "loss": 1.6121,
+      "step": 24650
+    },
+    {
+      "epoch": 5.195133512403223,
+      "grad_norm": 1.2727773189544678,
+      "learning_rate": 9.384779580222453e-05,
+      "loss": 1.6537,
+      "step": 24660
+    },
+    {
+      "epoch": 5.197240216990573,
+      "grad_norm": 1.0580523014068604,
+      "learning_rate": 9.378172801008757e-05,
+      "loss": 1.6081,
+      "step": 24670
+    },
+    {
+      "epoch": 5.199346921577922,
+      "grad_norm": 1.1032179594039917,
+      "learning_rate": 9.371566294262101e-05,
+      "loss": 1.6044,
+      "step": 24680
+    },
+    {
+      "epoch": 5.201453626165271,
+      "grad_norm": 1.1013224124908447,
+      "learning_rate": 9.364960062877268e-05,
+      "loss": 1.6322,
+      "step": 24690
+    },
+    {
+      "epoch": 5.20356033075262,
+      "grad_norm": 1.0431034564971924,
+      "learning_rate": 9.358354109748926e-05,
+      "loss": 1.5924,
+      "step": 24700
+    },
+    {
+      "epoch": 5.205667035339969,
+      "grad_norm": 1.073081374168396,
+      "learning_rate": 9.351748437771615e-05,
+      "loss": 1.5925,
+      "step": 24710
+    },
+    {
+      "epoch": 5.207773739927319,
+      "grad_norm": 1.0321084260940552,
+      "learning_rate": 9.345143049839749e-05,
+      "loss": 1.5903,
+      "step": 24720
+    },
+    {
+      "epoch": 5.209880444514668,
+      "grad_norm": 1.0614155530929565,
+      "learning_rate": 9.338537948847626e-05,
+      "loss": 1.5674,
+      "step": 24730
+    },
+    {
+      "epoch": 5.2119871491020175,
+      "grad_norm": 1.030477523803711,
+      "learning_rate": 9.331933137689412e-05,
+      "loss": 1.6419,
+      "step": 24740
+    },
+    {
+      "epoch": 5.214093853689366,
+      "grad_norm": 1.0811430215835571,
+      "learning_rate": 9.325328619259151e-05,
+      "loss": 1.6482,
+      "step": 24750
+    },
+    {
+      "epoch": 5.216200558276715,
+      "grad_norm": 1.0927019119262695,
+      "learning_rate": 9.31872439645075e-05,
+      "loss": 1.5877,
+      "step": 24760
+    },
+    {
+      "epoch": 5.218307262864065,
+      "grad_norm": 1.027279019355774,
+      "learning_rate": 9.312120472157997e-05,
+      "loss": 1.6179,
+      "step": 24770
+    },
+    {
+      "epoch": 5.220413967451414,
+      "grad_norm": 1.262536644935608,
+      "learning_rate": 9.305516849274541e-05,
+      "loss": 1.5903,
+      "step": 24780
+    },
+    {
+      "epoch": 5.222520672038764,
+      "grad_norm": 1.244111180305481,
+      "learning_rate": 9.298913530693907e-05,
+      "loss": 1.6425,
+      "step": 24790
+    },
+    {
+      "epoch": 5.224627376626112,
+      "grad_norm": 1.0479352474212646,
+      "learning_rate": 9.292310519309475e-05,
+      "loss": 1.5862,
+      "step": 24800
+    },
+    {
+      "epoch": 5.226734081213462,
+      "grad_norm": 1.1215065717697144,
+      "learning_rate": 9.285707818014502e-05,
+      "loss": 1.5962,
+      "step": 24810
+    },
+    {
+      "epoch": 5.228840785800811,
+      "grad_norm": 1.1024200916290283,
+      "learning_rate": 9.279105429702103e-05,
+      "loss": 1.6951,
+      "step": 24820
+    },
+    {
+      "epoch": 5.230947490388161,
+      "grad_norm": 1.1429402828216553,
+      "learning_rate": 9.272503357265261e-05,
+      "loss": 1.6295,
+      "step": 24830
+    },
+    {
+      "epoch": 5.23305419497551,
+      "grad_norm": 1.1070572137832642,
+      "learning_rate": 9.265901603596811e-05,
+      "loss": 1.5736,
+      "step": 24840
+    },
+    {
+      "epoch": 5.2351608995628585,
+      "grad_norm": 1.0643175840377808,
+      "learning_rate": 9.259300171589456e-05,
+      "loss": 1.5951,
+      "step": 24850
+    },
+    {
+      "epoch": 5.237267604150208,
+      "grad_norm": 1.0902659893035889,
+      "learning_rate": 9.252699064135758e-05,
+      "loss": 1.6062,
+      "step": 24860
+    },
+    {
+      "epoch": 5.239374308737557,
+      "grad_norm": 1.0354982614517212,
+      "learning_rate": 9.246098284128133e-05,
+      "loss": 1.5459,
+      "step": 24870
+    },
+    {
+      "epoch": 5.241481013324907,
+      "grad_norm": 1.082728385925293,
+      "learning_rate": 9.239497834458861e-05,
+      "loss": 1.6218,
+      "step": 24880
+    },
+    {
+      "epoch": 5.243587717912256,
+      "grad_norm": 1.2707469463348389,
+      "learning_rate": 9.232897718020064e-05,
+      "loss": 1.6518,
+      "step": 24890
+    },
+    {
+      "epoch": 5.245694422499605,
+      "grad_norm": 1.0314414501190186,
+      "learning_rate": 9.226297937703728e-05,
+      "loss": 1.6052,
+      "step": 24900
+    },
+    {
+      "epoch": 5.247801127086954,
+      "grad_norm": 1.0788559913635254,
+      "learning_rate": 9.219698496401693e-05,
+      "loss": 1.5714,
+      "step": 24910
+    },
+    {
+      "epoch": 5.249907831674303,
+      "grad_norm": 1.1269972324371338,
+      "learning_rate": 9.213099397005646e-05,
+      "loss": 1.5576,
+      "step": 24920
+    },
+    {
+      "epoch": 5.252014536261653,
+      "grad_norm": 1.141788125038147,
+      "learning_rate": 9.206500642407123e-05,
+      "loss": 1.6976,
+      "step": 24930
+    },
+    {
+      "epoch": 5.254121240849002,
+      "grad_norm": 1.159085750579834,
+      "learning_rate": 9.199902235497511e-05,
+      "loss": 1.6217,
+      "step": 24940
+    },
+    {
+      "epoch": 5.256227945436351,
+      "grad_norm": 1.0742684602737427,
+      "learning_rate": 9.193304179168046e-05,
+      "loss": 1.6252,
+      "step": 24950
+    },
+    {
+      "epoch": 5.2583346500237,
+      "grad_norm": 1.231498122215271,
+      "learning_rate": 9.186706476309812e-05,
+      "loss": 1.6454,
+      "step": 24960
+    },
+    {
+      "epoch": 5.26044135461105,
+      "grad_norm": 1.103709101676941,
+      "learning_rate": 9.180109129813731e-05,
+      "loss": 1.629,
+      "step": 24970
+    },
+    {
+      "epoch": 5.262548059198399,
+      "grad_norm": 1.1012301445007324,
+      "learning_rate": 9.173512142570573e-05,
+      "loss": 1.6015,
+      "step": 24980
+    },
+    {
+      "epoch": 5.264654763785748,
+      "grad_norm": 1.0851894617080688,
+      "learning_rate": 9.166915517470953e-05,
+      "loss": 1.6225,
+      "step": 24990
+    },
+    {
+      "epoch": 5.266761468373097,
+      "grad_norm": 1.2022123336791992,
+      "learning_rate": 9.160319257405326e-05,
+      "loss": 1.6095,
+      "step": 25000
+    },
+    {
+      "epoch": 5.268868172960446,
+      "grad_norm": 1.1617381572723389,
+      "learning_rate": 9.153723365263983e-05,
+      "loss": 1.6482,
+      "step": 25010
+    },
+    {
+      "epoch": 5.270974877547796,
+      "grad_norm": 1.0850499868392944,
+      "learning_rate": 9.147127843937055e-05,
+      "loss": 1.5923,
+      "step": 25020
+    },
+    {
+      "epoch": 5.273081582135145,
+      "grad_norm": 1.1439275741577148,
+      "learning_rate": 9.14053269631452e-05,
+      "loss": 1.5755,
+      "step": 25030
+    },
+    {
+      "epoch": 5.275188286722495,
+      "grad_norm": 1.04306960105896,
+      "learning_rate": 9.13393792528618e-05,
+      "loss": 1.637,
+      "step": 25040
+    },
+    {
+      "epoch": 5.2772949913098435,
+      "grad_norm": 1.1085509061813354,
+      "learning_rate": 9.127343533741673e-05,
+      "loss": 1.7139,
+      "step": 25050
+    },
+    {
+      "epoch": 5.279401695897193,
+      "grad_norm": 1.081034779548645,
+      "learning_rate": 9.12074952457048e-05,
+      "loss": 1.5843,
+      "step": 25060
+    },
+    {
+      "epoch": 5.281508400484542,
+      "grad_norm": 1.0647776126861572,
+      "learning_rate": 9.114155900661906e-05,
+      "loss": 1.6279,
+      "step": 25070
+    },
+    {
+      "epoch": 5.283615105071891,
+      "grad_norm": 1.064026117324829,
+      "learning_rate": 9.107562664905093e-05,
+      "loss": 1.6156,
+      "step": 25080
+    },
+    {
+      "epoch": 5.285721809659241,
+      "grad_norm": 1.2240235805511475,
+      "learning_rate": 9.100969820189006e-05,
+      "loss": 1.629,
+      "step": 25090
+    },
+    {
+      "epoch": 5.2878285142465895,
+      "grad_norm": 1.0848407745361328,
+      "learning_rate": 9.094377369402444e-05,
+      "loss": 1.6134,
+      "step": 25100
+    },
+    {
+      "epoch": 5.289935218833939,
+      "grad_norm": 1.1111202239990234,
+      "learning_rate": 9.087785315434034e-05,
+      "loss": 1.6009,
+      "step": 25110
+    },
+    {
+      "epoch": 5.292041923421288,
+      "grad_norm": 1.1308156251907349,
+      "learning_rate": 9.081193661172224e-05,
+      "loss": 1.5953,
+      "step": 25120
+    },
+    {
+      "epoch": 5.294148628008638,
+      "grad_norm": 1.0957039594650269,
+      "learning_rate": 9.074602409505293e-05,
+      "loss": 1.59,
+      "step": 25130
+    },
+    {
+      "epoch": 5.296255332595987,
+      "grad_norm": 1.0466959476470947,
+      "learning_rate": 9.068011563321336e-05,
+      "loss": 1.5666,
+      "step": 25140
+    },
+    {
+      "epoch": 5.298362037183336,
+      "grad_norm": 1.1054922342300415,
+      "learning_rate": 9.061421125508279e-05,
+      "loss": 1.6434,
+      "step": 25150
+    },
+    {
+      "epoch": 5.300468741770685,
+      "grad_norm": 1.0235182046890259,
+      "learning_rate": 9.054831098953863e-05,
+      "loss": 1.6244,
+      "step": 25160
+    },
+    {
+      "epoch": 5.302575446358034,
+      "grad_norm": 1.0452880859375,
+      "learning_rate": 9.048241486545653e-05,
+      "loss": 1.6336,
+      "step": 25170
+    },
+    {
+      "epoch": 5.304682150945384,
+      "grad_norm": 1.25437331199646,
+      "learning_rate": 9.041652291171028e-05,
+      "loss": 1.5979,
+      "step": 25180
+    },
+    {
+      "epoch": 5.306788855532733,
+      "grad_norm": 1.1870005130767822,
+      "learning_rate": 9.035063515717185e-05,
+      "loss": 1.582,
+      "step": 25190
+    },
+    {
+      "epoch": 5.3088955601200825,
+      "grad_norm": 1.1450788974761963,
+      "learning_rate": 9.028475163071141e-05,
+      "loss": 1.6209,
+      "step": 25200
+    },
+    {
+      "epoch": 5.311002264707431,
+      "grad_norm": 1.0536702871322632,
+      "learning_rate": 9.021887236119729e-05,
+      "loss": 1.6007,
+      "step": 25210
+    },
+    {
+      "epoch": 5.313108969294781,
+      "grad_norm": 1.4930493831634521,
+      "learning_rate": 9.015299737749585e-05,
+      "loss": 1.627,
+      "step": 25220
+    },
+    {
+      "epoch": 5.31521567388213,
+      "grad_norm": 1.0780115127563477,
+      "learning_rate": 9.008712670847164e-05,
+      "loss": 1.5879,
+      "step": 25230
+    },
+    {
+      "epoch": 5.317322378469479,
+      "grad_norm": 1.0802395343780518,
+      "learning_rate": 9.002126038298736e-05,
+      "loss": 1.6739,
+      "step": 25240
+    },
+    {
+      "epoch": 5.3194290830568285,
+      "grad_norm": 1.0295484066009521,
+      "learning_rate": 8.995539842990376e-05,
+      "loss": 1.6355,
+      "step": 25250
+    },
+    {
+      "epoch": 5.321535787644177,
+      "grad_norm": 1.0127300024032593,
+      "learning_rate": 8.988954087807968e-05,
+      "loss": 1.6036,
+      "step": 25260
+    },
+    {
+      "epoch": 5.323642492231527,
+      "grad_norm": 1.171979308128357,
+      "learning_rate": 8.9823687756372e-05,
+      "loss": 1.5417,
+      "step": 25270
+    },
+    {
+      "epoch": 5.325749196818876,
+      "grad_norm": 1.1540085077285767,
+      "learning_rate": 8.975783909363571e-05,
+      "loss": 1.6326,
+      "step": 25280
+    },
+    {
+      "epoch": 5.327855901406226,
+      "grad_norm": 1.2152167558670044,
+      "learning_rate": 8.969199491872384e-05,
+      "loss": 1.5994,
+      "step": 25290
+    },
+    {
+      "epoch": 5.3299626059935745,
+      "grad_norm": 1.049820899963379,
+      "learning_rate": 8.962615526048742e-05,
+      "loss": 1.6396,
+      "step": 25300
+    },
+    {
+      "epoch": 5.332069310580923,
+      "grad_norm": 1.0768330097198486,
+      "learning_rate": 8.956032014777552e-05,
+      "loss": 1.6109,
+      "step": 25310
+    },
+    {
+      "epoch": 5.334176015168273,
+      "grad_norm": 1.1724385023117065,
+      "learning_rate": 8.949448960943524e-05,
+      "loss": 1.6675,
+      "step": 25320
+    },
+    {
+      "epoch": 5.336282719755622,
+      "grad_norm": 1.080008625984192,
+      "learning_rate": 8.942866367431166e-05,
+      "loss": 1.6312,
+      "step": 25330
+    },
+    {
+      "epoch": 5.338389424342972,
+      "grad_norm": 1.121904730796814,
+      "learning_rate": 8.936284237124778e-05,
+      "loss": 1.6396,
+      "step": 25340
+    },
+    {
+      "epoch": 5.3404961289303206,
+      "grad_norm": 0.990679144859314,
+      "learning_rate": 8.929702572908468e-05,
+      "loss": 1.5981,
+      "step": 25350
+    },
+    {
+      "epoch": 5.34260283351767,
+      "grad_norm": 1.048987865447998,
+      "learning_rate": 8.923121377666134e-05,
+      "loss": 1.6284,
+      "step": 25360
+    },
+    {
+      "epoch": 5.344709538105019,
+      "grad_norm": 1.0290054082870483,
+      "learning_rate": 8.916540654281469e-05,
+      "loss": 1.6364,
+      "step": 25370
+    },
+    {
+      "epoch": 5.346816242692368,
+      "grad_norm": 1.119303584098816,
+      "learning_rate": 8.909960405637958e-05,
+      "loss": 1.6075,
+      "step": 25380
+    },
+    {
+      "epoch": 5.348922947279718,
+      "grad_norm": 1.2384835481643677,
+      "learning_rate": 8.90338063461888e-05,
+      "loss": 1.5558,
+      "step": 25390
+    },
+    {
+      "epoch": 5.351029651867067,
+      "grad_norm": 1.0883729457855225,
+      "learning_rate": 8.896801344107302e-05,
+      "loss": 1.6098,
+      "step": 25400
+    },
+    {
+      "epoch": 5.353136356454416,
+      "grad_norm": 1.1402106285095215,
+      "learning_rate": 8.890222536986085e-05,
+      "loss": 1.6391,
+      "step": 25410
+    },
+    {
+      "epoch": 5.355243061041765,
+      "grad_norm": 1.0882619619369507,
+      "learning_rate": 8.88364421613788e-05,
+      "loss": 1.6,
+      "step": 25420
+    },
+    {
+      "epoch": 5.357349765629115,
+      "grad_norm": 1.1555756330490112,
+      "learning_rate": 8.877066384445114e-05,
+      "loss": 1.6376,
+      "step": 25430
+    },
+    {
+      "epoch": 5.359456470216464,
+      "grad_norm": 1.139312505722046,
+      "learning_rate": 8.870489044790006e-05,
+      "loss": 1.6167,
+      "step": 25440
+    },
+    {
+      "epoch": 5.3615631748038135,
+      "grad_norm": 1.1492891311645508,
+      "learning_rate": 8.86391220005456e-05,
+      "loss": 1.6342,
+      "step": 25450
+    },
+    {
+      "epoch": 5.363669879391162,
+      "grad_norm": 1.0864403247833252,
+      "learning_rate": 8.857335853120572e-05,
+      "loss": 1.6052,
+      "step": 25460
+    },
+    {
+      "epoch": 5.365776583978511,
+      "grad_norm": 1.119722604751587,
+      "learning_rate": 8.8507600068696e-05,
+      "loss": 1.6132,
+      "step": 25470
+    },
+    {
+      "epoch": 5.367883288565861,
+      "grad_norm": 1.1324105262756348,
+      "learning_rate": 8.844184664182993e-05,
+      "loss": 1.6156,
+      "step": 25480
+    },
+    {
+      "epoch": 5.36998999315321,
+      "grad_norm": 1.0984848737716675,
+      "learning_rate": 8.837609827941885e-05,
+      "loss": 1.6018,
+      "step": 25490
+    },
+    {
+      "epoch": 5.3720966977405595,
+      "grad_norm": 1.1173757314682007,
+      "learning_rate": 8.831035501027186e-05,
+      "loss": 1.5513,
+      "step": 25500
+    },
+    {
+      "epoch": 5.374203402327908,
+      "grad_norm": 1.0673028230667114,
+      "learning_rate": 8.82446168631957e-05,
+      "loss": 1.5594,
+      "step": 25510
+    },
+    {
+      "epoch": 5.376310106915258,
+      "grad_norm": 1.1615711450576782,
+      "learning_rate": 8.8178883866995e-05,
+      "loss": 1.6259,
+      "step": 25520
+    },
+    {
+      "epoch": 5.378416811502607,
+      "grad_norm": 1.045413851737976,
+      "learning_rate": 8.811315605047212e-05,
+      "loss": 1.6354,
+      "step": 25530
+    },
+    {
+      "epoch": 5.380523516089957,
+      "grad_norm": 1.2000606060028076,
+      "learning_rate": 8.804743344242715e-05,
+      "loss": 1.6617,
+      "step": 25540
+    },
+    {
+      "epoch": 5.382630220677306,
+      "grad_norm": 1.1387240886688232,
+      "learning_rate": 8.798171607165778e-05,
+      "loss": 1.6394,
+      "step": 25550
+    },
+    {
+      "epoch": 5.384736925264654,
+      "grad_norm": 1.3192468881607056,
+      "learning_rate": 8.791600396695954e-05,
+      "loss": 1.5966,
+      "step": 25560
+    },
+    {
+      "epoch": 5.386843629852004,
+      "grad_norm": 1.1031746864318848,
+      "learning_rate": 8.785029715712564e-05,
+      "loss": 1.5806,
+      "step": 25570
+    },
+    {
+      "epoch": 5.388950334439353,
+      "grad_norm": 1.1289430856704712,
+      "learning_rate": 8.778459567094696e-05,
+      "loss": 1.5852,
+      "step": 25580
+    },
+    {
+      "epoch": 5.391057039026703,
+      "grad_norm": 1.1894451379776,
+      "learning_rate": 8.771889953721193e-05,
+      "loss": 1.6102,
+      "step": 25590
+    },
+    {
+      "epoch": 5.393163743614052,
+      "grad_norm": 1.0416464805603027,
+      "learning_rate": 8.765320878470679e-05,
+      "loss": 1.626,
+      "step": 25600
+    },
+    {
+      "epoch": 5.395270448201401,
+      "grad_norm": 1.0740944147109985,
+      "learning_rate": 8.758752344221537e-05,
+      "loss": 1.6292,
+      "step": 25610
+    },
+    {
+      "epoch": 5.39737715278875,
+      "grad_norm": 1.1314433813095093,
+      "learning_rate": 8.752184353851916e-05,
+      "loss": 1.5899,
+      "step": 25620
+    },
+    {
+      "epoch": 5.399483857376099,
+      "grad_norm": 1.1938408613204956,
+      "learning_rate": 8.745616910239716e-05,
+      "loss": 1.6401,
+      "step": 25630
+    },
+    {
+      "epoch": 5.401590561963449,
+      "grad_norm": 1.1085186004638672,
+      "learning_rate": 8.739050016262604e-05,
+      "loss": 1.6419,
+      "step": 25640
+    },
+    {
+      "epoch": 5.403697266550798,
+      "grad_norm": 1.1080095767974854,
+      "learning_rate": 8.732483674798013e-05,
+      "loss": 1.6605,
+      "step": 25650
+    },
+    {
+      "epoch": 5.405803971138147,
+      "grad_norm": 1.1335920095443726,
+      "learning_rate": 8.72591788872313e-05,
+      "loss": 1.6491,
+      "step": 25660
+    },
+    {
+      "epoch": 5.407910675725496,
+      "grad_norm": 1.1067054271697998,
+      "learning_rate": 8.719352660914884e-05,
+      "loss": 1.6226,
+      "step": 25670
+    },
+    {
+      "epoch": 5.410017380312846,
+      "grad_norm": 1.1139326095581055,
+      "learning_rate": 8.712787994249979e-05,
+      "loss": 1.6116,
+      "step": 25680
+    },
+    {
+      "epoch": 5.412124084900195,
+      "grad_norm": 1.1228965520858765,
+      "learning_rate": 8.706223891604866e-05,
+      "loss": 1.583,
+      "step": 25690
+    },
+    {
+      "epoch": 5.414230789487544,
+      "grad_norm": 1.0639370679855347,
+      "learning_rate": 8.699660355855748e-05,
+      "loss": 1.5674,
+      "step": 25700
+    },
+    {
+      "epoch": 5.416337494074893,
+      "grad_norm": 1.1020694971084595,
+      "learning_rate": 8.693097389878584e-05,
+      "loss": 1.5689,
+      "step": 25710
+    },
+    {
+      "epoch": 5.418444198662242,
+      "grad_norm": 1.1313263177871704,
+      "learning_rate": 8.686534996549071e-05,
+      "loss": 1.6155,
+      "step": 25720
+    },
+    {
+      "epoch": 5.420550903249592,
+      "grad_norm": 1.2440193891525269,
+      "learning_rate": 8.679973178742668e-05,
+      "loss": 1.6607,
+      "step": 25730
+    },
+    {
+      "epoch": 5.422657607836941,
+      "grad_norm": 1.121762752532959,
+      "learning_rate": 8.673411939334581e-05,
+      "loss": 1.5493,
+      "step": 25740
+    },
+    {
+      "epoch": 5.424764312424291,
+      "grad_norm": 1.0492937564849854,
+      "learning_rate": 8.666851281199757e-05,
+      "loss": 1.6558,
+      "step": 25750
+    },
+    {
+      "epoch": 5.426871017011639,
+      "grad_norm": 1.2097651958465576,
+      "learning_rate": 8.660291207212882e-05,
+      "loss": 1.6651,
+      "step": 25760
+    },
+    {
+      "epoch": 5.428977721598989,
+      "grad_norm": 1.1282267570495605,
+      "learning_rate": 8.653731720248406e-05,
+      "loss": 1.603,
+      "step": 25770
+    },
+    {
+      "epoch": 5.431084426186338,
+      "grad_norm": 1.0867459774017334,
+      "learning_rate": 8.647172823180505e-05,
+      "loss": 1.5858,
+      "step": 25780
+    },
+    {
+      "epoch": 5.433191130773687,
+      "grad_norm": 1.1194344758987427,
+      "learning_rate": 8.640614518883105e-05,
+      "loss": 1.6352,
+      "step": 25790
+    },
+    {
+      "epoch": 5.435297835361037,
+      "grad_norm": 1.0908195972442627,
+      "learning_rate": 8.634056810229862e-05,
+      "loss": 1.68,
+      "step": 25800
+    },
+    {
+      "epoch": 5.4374045399483855,
+      "grad_norm": 1.1710094213485718,
+      "learning_rate": 8.627499700094183e-05,
+      "loss": 1.5755,
+      "step": 25810
+    },
+    {
+      "epoch": 5.439511244535735,
+      "grad_norm": 1.1145213842391968,
+      "learning_rate": 8.620943191349207e-05,
+      "loss": 1.6015,
+      "step": 25820
+    },
+    {
+      "epoch": 5.441617949123084,
+      "grad_norm": 1.1294487714767456,
+      "learning_rate": 8.614387286867814e-05,
+      "loss": 1.65,
+      "step": 25830
+    },
+    {
+      "epoch": 5.443724653710434,
+      "grad_norm": 1.099099040031433,
+      "learning_rate": 8.607831989522604e-05,
+      "loss": 1.5673,
+      "step": 25840
+    },
+    {
+      "epoch": 5.445831358297783,
+      "grad_norm": 1.1329737901687622,
+      "learning_rate": 8.601277302185932e-05,
+      "loss": 1.6757,
+      "step": 25850
+    },
+    {
+      "epoch": 5.447938062885132,
+      "grad_norm": 1.0884069204330444,
+      "learning_rate": 8.594723227729875e-05,
+      "loss": 1.5949,
+      "step": 25860
+    },
+    {
+      "epoch": 5.450044767472481,
+      "grad_norm": 1.1422146558761597,
+      "learning_rate": 8.588169769026244e-05,
+      "loss": 1.649,
+      "step": 25870
+    },
+    {
+      "epoch": 5.45215147205983,
+      "grad_norm": 1.214929223060608,
+      "learning_rate": 8.581616928946571e-05,
+      "loss": 1.6687,
+      "step": 25880
+    },
+    {
+      "epoch": 5.45425817664718,
+      "grad_norm": 0.945561945438385,
+      "learning_rate": 8.575064710362131e-05,
+      "loss": 1.6441,
+      "step": 25890
+    },
+    {
+      "epoch": 5.456364881234529,
+      "grad_norm": 1.2264434099197388,
+      "learning_rate": 8.568513116143919e-05,
+      "loss": 1.6985,
+      "step": 25900
+    },
+    {
+      "epoch": 5.458471585821878,
+      "grad_norm": 1.0190399885177612,
+      "learning_rate": 8.561962149162662e-05,
+      "loss": 1.5734,
+      "step": 25910
+    },
+    {
+      "epoch": 5.460578290409227,
+      "grad_norm": 1.111526608467102,
+      "learning_rate": 8.555411812288798e-05,
+      "loss": 1.6245,
+      "step": 25920
+    },
+    {
+      "epoch": 5.462684994996577,
+      "grad_norm": 1.1148344278335571,
+      "learning_rate": 8.548862108392506e-05,
+      "loss": 1.5885,
+      "step": 25930
+    },
+    {
+      "epoch": 5.464791699583926,
+      "grad_norm": 1.1234092712402344,
+      "learning_rate": 8.542313040343679e-05,
+      "loss": 1.6301,
+      "step": 25940
+    },
+    {
+      "epoch": 5.466898404171275,
+      "grad_norm": 1.1449817419052124,
+      "learning_rate": 8.535764611011938e-05,
+      "loss": 1.6448,
+      "step": 25950
+    },
+    {
+      "epoch": 5.4690051087586244,
+      "grad_norm": 1.0495420694351196,
+      "learning_rate": 8.529216823266606e-05,
+      "loss": 1.581,
+      "step": 25960
+    },
+    {
+      "epoch": 5.471111813345973,
+      "grad_norm": 1.131170630455017,
+      "learning_rate": 8.522669679976749e-05,
+      "loss": 1.6223,
+      "step": 25970
+    },
+    {
+      "epoch": 5.473218517933323,
+      "grad_norm": 1.1759270429611206,
+      "learning_rate": 8.516123184011135e-05,
+      "loss": 1.5741,
+      "step": 25980
+    },
+    {
+      "epoch": 5.475325222520672,
+      "grad_norm": 1.2842549085617065,
+      "learning_rate": 8.509577338238255e-05,
+      "loss": 1.5891,
+      "step": 25990
+    },
+    {
+      "epoch": 5.477431927108022,
+      "grad_norm": 1.1138408184051514,
+      "learning_rate": 8.503032145526314e-05,
+      "loss": 1.6362,
+      "step": 26000
+    },
+    {
+      "epoch": 5.4795386316953705,
+      "grad_norm": 1.0257008075714111,
+      "learning_rate": 8.496487608743225e-05,
+      "loss": 1.6618,
+      "step": 26010
+    },
+    {
+      "epoch": 5.481645336282719,
+      "grad_norm": 1.2054387331008911,
+      "learning_rate": 8.489943730756622e-05,
+      "loss": 1.683,
+      "step": 26020
+    },
+    {
+      "epoch": 5.483752040870069,
+      "grad_norm": 1.1666181087493896,
+      "learning_rate": 8.483400514433845e-05,
+      "loss": 1.6326,
+      "step": 26030
+    },
+    {
+      "epoch": 5.485858745457418,
+      "grad_norm": 1.123583197593689,
+      "learning_rate": 8.47685796264195e-05,
+      "loss": 1.6099,
+      "step": 26040
+    },
+    {
+      "epoch": 5.487965450044768,
+      "grad_norm": 1.0301923751831055,
+      "learning_rate": 8.470316078247694e-05,
+      "loss": 1.6046,
+      "step": 26050
+    },
+    {
+      "epoch": 5.4900721546321165,
+      "grad_norm": 1.0988959074020386,
+      "learning_rate": 8.463774864117542e-05,
+      "loss": 1.607,
+      "step": 26060
+    },
+    {
+      "epoch": 5.492178859219466,
+      "grad_norm": 1.058692216873169,
+      "learning_rate": 8.457234323117675e-05,
+      "loss": 1.6333,
+      "step": 26070
+    },
+    {
+      "epoch": 5.494285563806815,
+      "grad_norm": 1.0051920413970947,
+      "learning_rate": 8.450694458113969e-05,
+      "loss": 1.6068,
+      "step": 26080
+    },
+    {
+      "epoch": 5.496392268394165,
+      "grad_norm": 1.1012892723083496,
+      "learning_rate": 8.444155271972008e-05,
+      "loss": 1.6615,
+      "step": 26090
+    },
+    {
+      "epoch": 5.498498972981514,
+      "grad_norm": 1.1324644088745117,
+      "learning_rate": 8.437616767557077e-05,
+      "loss": 1.6613,
+      "step": 26100
+    },
+    {
+      "epoch": 5.5006056775688625,
+      "grad_norm": 1.0465189218521118,
+      "learning_rate": 8.431078947734164e-05,
+      "loss": 1.6052,
+      "step": 26110
+    },
+    {
+      "epoch": 5.502712382156212,
+      "grad_norm": 1.173836350440979,
+      "learning_rate": 8.424541815367957e-05,
+      "loss": 1.579,
+      "step": 26120
+    },
+    {
+      "epoch": 5.504819086743561,
+      "grad_norm": 1.185910940170288,
+      "learning_rate": 8.418005373322841e-05,
+      "loss": 1.5884,
+      "step": 26130
+    },
+    {
+      "epoch": 5.506925791330911,
+      "grad_norm": 1.1530224084854126,
+      "learning_rate": 8.4114696244629e-05,
+      "loss": 1.5754,
+      "step": 26140
+    },
+    {
+      "epoch": 5.50903249591826,
+      "grad_norm": 1.2175930738449097,
+      "learning_rate": 8.404934571651913e-05,
+      "loss": 1.6492,
+      "step": 26150
+    },
+    {
+      "epoch": 5.5111392005056095,
+      "grad_norm": 1.1386138200759888,
+      "learning_rate": 8.398400217753357e-05,
+      "loss": 1.572,
+      "step": 26160
+    },
+    {
+      "epoch": 5.513245905092958,
+      "grad_norm": 1.1737759113311768,
+      "learning_rate": 8.391866565630397e-05,
+      "loss": 1.5755,
+      "step": 26170
+    },
+    {
+      "epoch": 5.515352609680308,
+      "grad_norm": 0.9875378608703613,
+      "learning_rate": 8.385333618145896e-05,
+      "loss": 1.5388,
+      "step": 26180
+    },
+    {
+      "epoch": 5.517459314267657,
+      "grad_norm": 1.1286633014678955,
+      "learning_rate": 8.378801378162407e-05,
+      "loss": 1.5541,
+      "step": 26190
+    },
+    {
+      "epoch": 5.519566018855006,
+      "grad_norm": 1.0399701595306396,
+      "learning_rate": 8.372269848542172e-05,
+      "loss": 1.6503,
+      "step": 26200
+    },
+    {
+      "epoch": 5.5216727234423555,
+      "grad_norm": 1.1287180185317993,
+      "learning_rate": 8.36573903214712e-05,
+      "loss": 1.6099,
+      "step": 26210
+    },
+    {
+      "epoch": 5.523779428029704,
+      "grad_norm": 1.062724232673645,
+      "learning_rate": 8.359208931838871e-05,
+      "loss": 1.6084,
+      "step": 26220
+    },
+    {
+      "epoch": 5.525886132617054,
+      "grad_norm": 1.0824880599975586,
+      "learning_rate": 8.35267955047873e-05,
+      "loss": 1.6435,
+      "step": 26230
+    },
+    {
+      "epoch": 5.527992837204403,
+      "grad_norm": 1.113713264465332,
+      "learning_rate": 8.346150890927688e-05,
+      "loss": 1.6137,
+      "step": 26240
+    },
+    {
+      "epoch": 5.530099541791753,
+      "grad_norm": 1.0756171941757202,
+      "learning_rate": 8.339622956046417e-05,
+      "loss": 1.6287,
+      "step": 26250
+    },
+    {
+      "epoch": 5.5322062463791015,
+      "grad_norm": 1.0738415718078613,
+      "learning_rate": 8.333095748695271e-05,
+      "loss": 1.5887,
+      "step": 26260
+    },
+    {
+      "epoch": 5.53431295096645,
+      "grad_norm": 1.0821735858917236,
+      "learning_rate": 8.326569271734287e-05,
+      "loss": 1.6155,
+      "step": 26270
+    },
+    {
+      "epoch": 5.5364196555538,
+      "grad_norm": 1.0803221464157104,
+      "learning_rate": 8.320043528023188e-05,
+      "loss": 1.6377,
+      "step": 26280
+    },
+    {
+      "epoch": 5.538526360141149,
+      "grad_norm": 1.025138020515442,
+      "learning_rate": 8.313518520421366e-05,
+      "loss": 1.6606,
+      "step": 26290
+    },
+    {
+      "epoch": 5.540633064728499,
+      "grad_norm": 1.1567028760910034,
+      "learning_rate": 8.306994251787892e-05,
+      "loss": 1.6837,
+      "step": 26300
+    },
+    {
+      "epoch": 5.5427397693158476,
+      "grad_norm": 1.1138674020767212,
+      "learning_rate": 8.300470724981517e-05,
+      "loss": 1.6178,
+      "step": 26310
+    },
+    {
+      "epoch": 5.544846473903197,
+      "grad_norm": 1.135926365852356,
+      "learning_rate": 8.293947942860666e-05,
+      "loss": 1.6356,
+      "step": 26320
+    },
+    {
+      "epoch": 5.546953178490546,
+      "grad_norm": 1.0830113887786865,
+      "learning_rate": 8.287425908283442e-05,
+      "loss": 1.567,
+      "step": 26330
+    },
+    {
+      "epoch": 5.549059883077895,
+      "grad_norm": 1.042052149772644,
+      "learning_rate": 8.280904624107606e-05,
+      "loss": 1.6557,
+      "step": 26340
+    },
+    {
+      "epoch": 5.551166587665245,
+      "grad_norm": 1.1980870962142944,
+      "learning_rate": 8.274384093190605e-05,
+      "loss": 1.638,
+      "step": 26350
+    },
+    {
+      "epoch": 5.553273292252594,
+      "grad_norm": 1.0648488998413086,
+      "learning_rate": 8.267864318389549e-05,
+      "loss": 1.6735,
+      "step": 26360
+    },
+    {
+      "epoch": 5.555379996839943,
+      "grad_norm": 1.1223556995391846,
+      "learning_rate": 8.261345302561223e-05,
+      "loss": 1.5817,
+      "step": 26370
+    },
+    {
+      "epoch": 5.557486701427292,
+      "grad_norm": 1.0523229837417603,
+      "learning_rate": 8.254827048562069e-05,
+      "loss": 1.6265,
+      "step": 26380
+    },
+    {
+      "epoch": 5.559593406014642,
+      "grad_norm": 1.073630928993225,
+      "learning_rate": 8.248309559248203e-05,
+      "loss": 1.6052,
+      "step": 26390
+    },
+    {
+      "epoch": 5.561700110601991,
+      "grad_norm": 1.099674940109253,
+      "learning_rate": 8.241792837475405e-05,
+      "loss": 1.6322,
+      "step": 26400
+    },
+    {
+      "epoch": 5.56380681518934,
+      "grad_norm": 1.1496034860610962,
+      "learning_rate": 8.235276886099119e-05,
+      "loss": 1.6381,
+      "step": 26410
+    },
+    {
+      "epoch": 5.565913519776689,
+      "grad_norm": 1.0635652542114258,
+      "learning_rate": 8.228761707974445e-05,
+      "loss": 1.6508,
+      "step": 26420
+    },
+    {
+      "epoch": 5.568020224364038,
+      "grad_norm": 1.1200566291809082,
+      "learning_rate": 8.222247305956153e-05,
+      "loss": 1.6321,
+      "step": 26430
+    },
+    {
+      "epoch": 5.570126928951388,
+      "grad_norm": 1.3193455934524536,
+      "learning_rate": 8.215733682898669e-05,
+      "loss": 1.5639,
+      "step": 26440
+    },
+    {
+      "epoch": 5.572233633538737,
+      "grad_norm": 1.0898486375808716,
+      "learning_rate": 8.209220841656078e-05,
+      "loss": 1.6149,
+      "step": 26450
+    },
+    {
+      "epoch": 5.5743403381260865,
+      "grad_norm": 1.1833173036575317,
+      "learning_rate": 8.202708785082121e-05,
+      "loss": 1.6015,
+      "step": 26460
+    },
+    {
+      "epoch": 5.576447042713435,
+      "grad_norm": 1.1096851825714111,
+      "learning_rate": 8.196197516030198e-05,
+      "loss": 1.6056,
+      "step": 26470
+    },
+    {
+      "epoch": 5.578553747300785,
+      "grad_norm": 1.1732834577560425,
+      "learning_rate": 8.18968703735336e-05,
+      "loss": 1.6239,
+      "step": 26480
+    },
+    {
+      "epoch": 5.580660451888134,
+      "grad_norm": 1.0716811418533325,
+      "learning_rate": 8.183177351904318e-05,
+      "loss": 1.6712,
+      "step": 26490
+    },
+    {
+      "epoch": 5.582767156475484,
+      "grad_norm": 1.0455117225646973,
+      "learning_rate": 8.176668462535427e-05,
+      "loss": 1.6137,
+      "step": 26500
+    },
+    {
+      "epoch": 5.584873861062833,
+      "grad_norm": 1.180620789527893,
+      "learning_rate": 8.1701603720987e-05,
+      "loss": 1.5832,
+      "step": 26510
+    },
+    {
+      "epoch": 5.586980565650181,
+      "grad_norm": 1.1026450395584106,
+      "learning_rate": 8.163653083445799e-05,
+      "loss": 1.6063,
+      "step": 26520
+    },
+    {
+      "epoch": 5.589087270237531,
+      "grad_norm": 1.145521879196167,
+      "learning_rate": 8.157146599428028e-05,
+      "loss": 1.6038,
+      "step": 26530
+    },
+    {
+      "epoch": 5.59119397482488,
+      "grad_norm": 1.016776204109192,
+      "learning_rate": 8.150640922896356e-05,
+      "loss": 1.5613,
+      "step": 26540
+    },
+    {
+      "epoch": 5.59330067941223,
+      "grad_norm": 1.0819216966629028,
+      "learning_rate": 8.144136056701371e-05,
+      "loss": 1.5235,
+      "step": 26550
+    },
+    {
+      "epoch": 5.595407383999579,
+      "grad_norm": 1.1226952075958252,
+      "learning_rate": 8.137632003693329e-05,
+      "loss": 1.6309,
+      "step": 26560
+    },
+    {
+      "epoch": 5.597514088586928,
+      "grad_norm": 1.149537205696106,
+      "learning_rate": 8.131128766722117e-05,
+      "loss": 1.5602,
+      "step": 26570
+    },
+    {
+      "epoch": 5.599620793174277,
+      "grad_norm": 1.0373797416687012,
+      "learning_rate": 8.124626348637279e-05,
+      "loss": 1.6514,
+      "step": 26580
+    },
+    {
+      "epoch": 5.601727497761626,
+      "grad_norm": 1.1900537014007568,
+      "learning_rate": 8.118124752287979e-05,
+      "loss": 1.647,
+      "step": 26590
+    },
+    {
+      "epoch": 5.603834202348976,
+      "grad_norm": 1.0474778413772583,
+      "learning_rate": 8.111623980523035e-05,
+      "loss": 1.6313,
+      "step": 26600
+    },
+    {
+      "epoch": 5.605940906936325,
+      "grad_norm": 1.0889447927474976,
+      "learning_rate": 8.105124036190901e-05,
+      "loss": 1.5919,
+      "step": 26610
+    },
+    {
+      "epoch": 5.608047611523674,
+      "grad_norm": 1.520419716835022,
+      "learning_rate": 8.098624922139676e-05,
+      "loss": 1.6039,
+      "step": 26620
+    },
+    {
+      "epoch": 5.610154316111023,
+      "grad_norm": 1.1650199890136719,
+      "learning_rate": 8.092126641217076e-05,
+      "loss": 1.6202,
+      "step": 26630
+    },
+    {
+      "epoch": 5.612261020698373,
+      "grad_norm": 1.0928066968917847,
+      "learning_rate": 8.085629196270469e-05,
+      "loss": 1.6322,
+      "step": 26640
+    },
+    {
+      "epoch": 5.614367725285722,
+      "grad_norm": 1.0415284633636475,
+      "learning_rate": 8.079132590146847e-05,
+      "loss": 1.61,
+      "step": 26650
+    },
+    {
+      "epoch": 5.616474429873071,
+      "grad_norm": 1.1244189739227295,
+      "learning_rate": 8.072636825692849e-05,
+      "loss": 1.5919,
+      "step": 26660
+    },
+    {
+      "epoch": 5.61858113446042,
+      "grad_norm": 1.1185684204101562,
+      "learning_rate": 8.066141905754723e-05,
+      "loss": 1.6128,
+      "step": 26670
+    },
+    {
+      "epoch": 5.620687839047769,
+      "grad_norm": 1.0240288972854614,
+      "learning_rate": 8.059647833178363e-05,
+      "loss": 1.6269,
+      "step": 26680
+    },
+    {
+      "epoch": 5.622794543635119,
+      "grad_norm": 1.125137448310852,
+      "learning_rate": 8.053154610809285e-05,
+      "loss": 1.6419,
+      "step": 26690
+    },
+    {
+      "epoch": 5.624901248222468,
+      "grad_norm": 1.057777762413025,
+      "learning_rate": 8.046662241492645e-05,
+      "loss": 1.6237,
+      "step": 26700
+    },
+    {
+      "epoch": 5.627007952809818,
+      "grad_norm": 1.1793187856674194,
+      "learning_rate": 8.040170728073202e-05,
+      "loss": 1.6258,
+      "step": 26710
+    },
+    {
+      "epoch": 5.629114657397166,
+      "grad_norm": 1.0160819292068481,
+      "learning_rate": 8.03368007339536e-05,
+      "loss": 1.6346,
+      "step": 26720
+    },
+    {
+      "epoch": 5.631221361984515,
+      "grad_norm": 1.032878041267395,
+      "learning_rate": 8.027190280303137e-05,
+      "loss": 1.6404,
+      "step": 26730
+    },
+    {
+      "epoch": 5.633328066571865,
+      "grad_norm": 1.0659453868865967,
+      "learning_rate": 8.020701351640182e-05,
+      "loss": 1.6393,
+      "step": 26740
+    },
+    {
+      "epoch": 5.635434771159214,
+      "grad_norm": 1.0905574560165405,
+      "learning_rate": 8.014213290249751e-05,
+      "loss": 1.5401,
+      "step": 26750
+    },
+    {
+      "epoch": 5.637541475746564,
+      "grad_norm": 1.096535563468933,
+      "learning_rate": 8.007726098974734e-05,
+      "loss": 1.5796,
+      "step": 26760
+    },
+    {
+      "epoch": 5.6396481803339125,
+      "grad_norm": 1.1352156400680542,
+      "learning_rate": 8.00123978065763e-05,
+      "loss": 1.6769,
+      "step": 26770
+    },
+    {
+      "epoch": 5.641754884921262,
+      "grad_norm": 1.0784434080123901,
+      "learning_rate": 7.994754338140569e-05,
+      "loss": 1.6259,
+      "step": 26780
+    },
+    {
+      "epoch": 5.643861589508611,
+      "grad_norm": 1.0838322639465332,
+      "learning_rate": 7.988269774265278e-05,
+      "loss": 1.6178,
+      "step": 26790
+    },
+    {
+      "epoch": 5.64596829409596,
+      "grad_norm": 1.0983805656433105,
+      "learning_rate": 7.981786091873112e-05,
+      "loss": 1.6762,
+      "step": 26800
+    },
+    {
+      "epoch": 5.64807499868331,
+      "grad_norm": 1.1972914934158325,
+      "learning_rate": 7.975303293805035e-05,
+      "loss": 1.6466,
+      "step": 26810
+    },
+    {
+      "epoch": 5.6501817032706585,
+      "grad_norm": 1.0667072534561157,
+      "learning_rate": 7.96882138290163e-05,
+      "loss": 1.6487,
+      "step": 26820
+    },
+    {
+      "epoch": 5.652288407858008,
+      "grad_norm": 1.1564662456512451,
+      "learning_rate": 7.962340362003089e-05,
+      "loss": 1.5964,
+      "step": 26830
+    },
+    {
+      "epoch": 5.654395112445357,
+      "grad_norm": 1.1387529373168945,
+      "learning_rate": 7.9558602339492e-05,
+      "loss": 1.5973,
+      "step": 26840
+    },
+    {
+      "epoch": 5.656501817032707,
+      "grad_norm": 1.087855577468872,
+      "learning_rate": 7.949381001579378e-05,
+      "loss": 1.5811,
+      "step": 26850
+    },
+    {
+      "epoch": 5.658608521620056,
+      "grad_norm": 1.3254786729812622,
+      "learning_rate": 7.942902667732639e-05,
+      "loss": 1.5901,
+      "step": 26860
+    },
+    {
+      "epoch": 5.660715226207405,
+      "grad_norm": 1.0886156558990479,
+      "learning_rate": 7.936425235247611e-05,
+      "loss": 1.6013,
+      "step": 26870
+    },
+    {
+      "epoch": 5.662821930794754,
+      "grad_norm": 1.041137456893921,
+      "learning_rate": 7.929948706962508e-05,
+      "loss": 1.5307,
+      "step": 26880
+    },
+    {
+      "epoch": 5.664928635382104,
+      "grad_norm": 1.093061089515686,
+      "learning_rate": 7.92347308571517e-05,
+      "loss": 1.5949,
+      "step": 26890
+    },
+    {
+      "epoch": 5.667035339969453,
+      "grad_norm": 1.2626030445098877,
+      "learning_rate": 7.916998374343028e-05,
+      "loss": 1.6477,
+      "step": 26900
+    },
+    {
+      "epoch": 5.669142044556802,
+      "grad_norm": 1.0964906215667725,
+      "learning_rate": 7.910524575683122e-05,
+      "loss": 1.6141,
+      "step": 26910
+    },
+    {
+      "epoch": 5.6712487491441514,
+      "grad_norm": 1.0444769859313965,
+      "learning_rate": 7.904051692572079e-05,
+      "loss": 1.6259,
+      "step": 26920
+    },
+    {
+      "epoch": 5.6733554537315,
+      "grad_norm": 1.0373730659484863,
+      "learning_rate": 7.897579727846133e-05,
+      "loss": 1.6066,
+      "step": 26930
+    },
+    {
+      "epoch": 5.67546215831885,
+      "grad_norm": 1.141251802444458,
+      "learning_rate": 7.891108684341121e-05,
+      "loss": 1.6742,
+      "step": 26940
+    },
+    {
+      "epoch": 5.677568862906199,
+      "grad_norm": 1.1581555604934692,
+      "learning_rate": 7.884638564892472e-05,
+      "loss": 1.6523,
+      "step": 26950
+    },
+    {
+      "epoch": 5.679675567493549,
+      "grad_norm": 1.0533156394958496,
+      "learning_rate": 7.8781693723352e-05,
+      "loss": 1.6443,
+      "step": 26960
+    },
+    {
+      "epoch": 5.6817822720808975,
+      "grad_norm": 1.1277610063552856,
+      "learning_rate": 7.871701109503929e-05,
+      "loss": 1.62,
+      "step": 26970
+    },
+    {
+      "epoch": 5.683888976668246,
+      "grad_norm": 1.0494487285614014,
+      "learning_rate": 7.865233779232866e-05,
+      "loss": 1.6113,
+      "step": 26980
+    },
+    {
+      "epoch": 5.685995681255596,
+      "grad_norm": 1.0661777257919312,
+      "learning_rate": 7.858767384355815e-05,
+      "loss": 1.5868,
+      "step": 26990
+    },
+    {
+      "epoch": 5.688102385842945,
+      "grad_norm": 1.1448099613189697,
+      "learning_rate": 7.852301927706159e-05,
+      "loss": 1.6307,
+      "step": 27000
+    },
+    {
+      "epoch": 5.690209090430295,
+      "grad_norm": 1.0377132892608643,
+      "learning_rate": 7.845837412116883e-05,
+      "loss": 1.5962,
+      "step": 27010
+    },
+    {
+      "epoch": 5.6923157950176435,
+      "grad_norm": 1.0388470888137817,
+      "learning_rate": 7.839373840420554e-05,
+      "loss": 1.5914,
+      "step": 27020
+    },
+    {
+      "epoch": 5.694422499604993,
+      "grad_norm": 1.1136940717697144,
+      "learning_rate": 7.83291121544933e-05,
+      "loss": 1.6221,
+      "step": 27030
+    },
+    {
+      "epoch": 5.696529204192342,
+      "grad_norm": 1.113476037979126,
+      "learning_rate": 7.826449540034939e-05,
+      "loss": 1.6005,
+      "step": 27040
+    },
+    {
+      "epoch": 5.698635908779691,
+      "grad_norm": 1.1359562873840332,
+      "learning_rate": 7.819988817008713e-05,
+      "loss": 1.6283,
+      "step": 27050
+    },
+    {
+      "epoch": 5.700742613367041,
+      "grad_norm": 1.0916094779968262,
+      "learning_rate": 7.813529049201556e-05,
+      "loss": 1.5927,
+      "step": 27060
+    },
+    {
+      "epoch": 5.7028493179543895,
+      "grad_norm": 1.1190940141677856,
+      "learning_rate": 7.807070239443957e-05,
+      "loss": 1.6316,
+      "step": 27070
+    },
+    {
+      "epoch": 5.704956022541739,
+      "grad_norm": 1.1429566144943237,
+      "learning_rate": 7.800612390565974e-05,
+      "loss": 1.6386,
+      "step": 27080
+    },
+    {
+      "epoch": 5.707062727129088,
+      "grad_norm": 1.0969349145889282,
+      "learning_rate": 7.794155505397261e-05,
+      "loss": 1.572,
+      "step": 27090
+    },
+    {
+      "epoch": 5.709169431716438,
+      "grad_norm": 1.0643476247787476,
+      "learning_rate": 7.787699586767042e-05,
+      "loss": 1.5452,
+      "step": 27100
+    },
+    {
+      "epoch": 5.711276136303787,
+      "grad_norm": 1.1677190065383911,
+      "learning_rate": 7.781244637504114e-05,
+      "loss": 1.6587,
+      "step": 27110
+    },
+    {
+      "epoch": 5.713382840891136,
+      "grad_norm": 1.0919948816299438,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 1.6646,
+      "step": 27120
+    },
+    {
+      "epoch": 5.715489545478485,
+      "grad_norm": 1.1836225986480713,
+      "learning_rate": 7.768337658393216e-05,
+      "loss": 1.6362,
+      "step": 27130
+    },
+    {
+      "epoch": 5.717596250065834,
+      "grad_norm": 1.017246961593628,
+      "learning_rate": 7.761885634200717e-05,
+      "loss": 1.6277,
+      "step": 27140
+    },
+    {
+      "epoch": 5.719702954653184,
+      "grad_norm": 1.0970447063446045,
+      "learning_rate": 7.755434590686452e-05,
+      "loss": 1.5896,
+      "step": 27150
+    },
+    {
+      "epoch": 5.721809659240533,
+      "grad_norm": 1.150378704071045,
+      "learning_rate": 7.748984530677089e-05,
+      "loss": 1.6005,
+      "step": 27160
+    },
+    {
+      "epoch": 5.7239163638278825,
+      "grad_norm": 1.0526460409164429,
+      "learning_rate": 7.742535456998853e-05,
+      "loss": 1.6235,
+      "step": 27170
+    },
+    {
+      "epoch": 5.726023068415231,
+      "grad_norm": 1.0322532653808594,
+      "learning_rate": 7.736087372477554e-05,
+      "loss": 1.6019,
+      "step": 27180
+    },
+    {
+      "epoch": 5.728129773002581,
+      "grad_norm": 1.119554042816162,
+      "learning_rate": 7.729640279938555e-05,
+      "loss": 1.5993,
+      "step": 27190
+    },
+    {
+      "epoch": 5.73023647758993,
+      "grad_norm": 1.0483386516571045,
+      "learning_rate": 7.723194182206792e-05,
+      "loss": 1.6072,
+      "step": 27200
+    },
+    {
+      "epoch": 5.73234318217728,
+      "grad_norm": 1.065516710281372,
+      "learning_rate": 7.71674908210676e-05,
+      "loss": 1.6462,
+      "step": 27210
+    },
+    {
+      "epoch": 5.7344498867646285,
+      "grad_norm": 1.101328730583191,
+      "learning_rate": 7.710304982462519e-05,
+      "loss": 1.6684,
+      "step": 27220
+    },
+    {
+      "epoch": 5.736556591351977,
+      "grad_norm": 1.0765429735183716,
+      "learning_rate": 7.70386188609769e-05,
+      "loss": 1.6649,
+      "step": 27230
+    },
+    {
+      "epoch": 5.738663295939327,
+      "grad_norm": 1.154428243637085,
+      "learning_rate": 7.69741979583546e-05,
+      "loss": 1.6245,
+      "step": 27240
+    },
+    {
+      "epoch": 5.740770000526676,
+      "grad_norm": 1.134865164756775,
+      "learning_rate": 7.690978714498563e-05,
+      "loss": 1.5566,
+      "step": 27250
+    },
+    {
+      "epoch": 5.742876705114026,
+      "grad_norm": 1.0900312662124634,
+      "learning_rate": 7.684538644909302e-05,
+      "loss": 1.6537,
+      "step": 27260
+    },
+    {
+      "epoch": 5.7449834097013746,
+      "grad_norm": 1.1284072399139404,
+      "learning_rate": 7.678099589889534e-05,
+      "loss": 1.6081,
+      "step": 27270
+    },
+    {
+      "epoch": 5.747090114288724,
+      "grad_norm": 1.0693503618240356,
+      "learning_rate": 7.671661552260671e-05,
+      "loss": 1.5738,
+      "step": 27280
+    },
+    {
+      "epoch": 5.749196818876073,
+      "grad_norm": 1.5038498640060425,
+      "learning_rate": 7.665224534843673e-05,
+      "loss": 1.6403,
+      "step": 27290
+    },
+    {
+      "epoch": 5.751303523463422,
+      "grad_norm": 1.0032192468643188,
+      "learning_rate": 7.658788540459062e-05,
+      "loss": 1.6093,
+      "step": 27300
+    },
+    {
+      "epoch": 5.753410228050772,
+      "grad_norm": 1.174879550933838,
+      "learning_rate": 7.652353571926908e-05,
+      "loss": 1.5718,
+      "step": 27310
+    },
+    {
+      "epoch": 5.755516932638121,
+      "grad_norm": 1.0580400228500366,
+      "learning_rate": 7.645919632066833e-05,
+      "loss": 1.6227,
+      "step": 27320
+    },
+    {
+      "epoch": 5.75762363722547,
+      "grad_norm": 1.119176983833313,
+      "learning_rate": 7.639486723698006e-05,
+      "loss": 1.6271,
+      "step": 27330
+    },
+    {
+      "epoch": 5.759730341812819,
+      "grad_norm": 1.1590195894241333,
+      "learning_rate": 7.633054849639144e-05,
+      "loss": 1.6028,
+      "step": 27340
+    },
+    {
+      "epoch": 5.761837046400169,
+      "grad_norm": 1.1386713981628418,
+      "learning_rate": 7.62662401270851e-05,
+      "loss": 1.6229,
+      "step": 27350
+    },
+    {
+      "epoch": 5.763943750987518,
+      "grad_norm": 1.172624111175537,
+      "learning_rate": 7.620194215723919e-05,
+      "loss": 1.564,
+      "step": 27360
+    },
+    {
+      "epoch": 5.766050455574867,
+      "grad_norm": 1.1533126831054688,
+      "learning_rate": 7.613765461502724e-05,
+      "loss": 1.5977,
+      "step": 27370
+    },
+    {
+      "epoch": 5.768157160162216,
+      "grad_norm": 1.1233863830566406,
+      "learning_rate": 7.60733775286182e-05,
+      "loss": 1.616,
+      "step": 27380
+    },
+    {
+      "epoch": 5.770263864749565,
+      "grad_norm": 1.0440213680267334,
+      "learning_rate": 7.600911092617651e-05,
+      "loss": 1.5581,
+      "step": 27390
+    },
+    {
+      "epoch": 5.772370569336915,
+      "grad_norm": 1.2219536304473877,
+      "learning_rate": 7.594485483586193e-05,
+      "loss": 1.651,
+      "step": 27400
+    },
+    {
+      "epoch": 5.774477273924264,
+      "grad_norm": 1.1317471265792847,
+      "learning_rate": 7.588060928582971e-05,
+      "loss": 1.5883,
+      "step": 27410
+    },
+    {
+      "epoch": 5.7765839785116135,
+      "grad_norm": 1.0306445360183716,
+      "learning_rate": 7.581637430423037e-05,
+      "loss": 1.602,
+      "step": 27420
+    },
+    {
+      "epoch": 5.778690683098962,
+      "grad_norm": 1.0400179624557495,
+      "learning_rate": 7.575214991920987e-05,
+      "loss": 1.6392,
+      "step": 27430
+    },
+    {
+      "epoch": 5.780797387686311,
+      "grad_norm": 1.022694706916809,
+      "learning_rate": 7.568793615890954e-05,
+      "loss": 1.6382,
+      "step": 27440
+    },
+    {
+      "epoch": 5.782904092273661,
+      "grad_norm": 1.1518008708953857,
+      "learning_rate": 7.562373305146604e-05,
+      "loss": 1.6253,
+      "step": 27450
+    },
+    {
+      "epoch": 5.78501079686101,
+      "grad_norm": 1.0711225271224976,
+      "learning_rate": 7.55595406250113e-05,
+      "loss": 1.5917,
+      "step": 27460
+    },
+    {
+      "epoch": 5.78711750144836,
+      "grad_norm": 1.1197304725646973,
+      "learning_rate": 7.549535890767263e-05,
+      "loss": 1.6342,
+      "step": 27470
+    },
+    {
+      "epoch": 5.789224206035708,
+      "grad_norm": 1.1032601594924927,
+      "learning_rate": 7.543118792757266e-05,
+      "loss": 1.6164,
+      "step": 27480
+    },
+    {
+      "epoch": 5.791330910623058,
+      "grad_norm": 1.1401535272598267,
+      "learning_rate": 7.536702771282932e-05,
+      "loss": 1.5832,
+      "step": 27490
+    },
+    {
+      "epoch": 5.793437615210407,
+      "grad_norm": 1.1442997455596924,
+      "learning_rate": 7.530287829155574e-05,
+      "loss": 1.6285,
+      "step": 27500
+    },
+    {
+      "epoch": 5.795544319797756,
+      "grad_norm": 1.0326517820358276,
+      "learning_rate": 7.523873969186039e-05,
+      "loss": 1.5764,
+      "step": 27510
+    },
+    {
+      "epoch": 5.797651024385106,
+      "grad_norm": 1.0888991355895996,
+      "learning_rate": 7.517461194184699e-05,
+      "loss": 1.6155,
+      "step": 27520
+    },
+    {
+      "epoch": 5.7997577289724545,
+      "grad_norm": 1.1346875429153442,
+      "learning_rate": 7.511049506961454e-05,
+      "loss": 1.6698,
+      "step": 27530
+    },
+    {
+      "epoch": 5.801864433559804,
+      "grad_norm": 1.1853289604187012,
+      "learning_rate": 7.504638910325717e-05,
+      "loss": 1.6589,
+      "step": 27540
+    },
+    {
+      "epoch": 5.803971138147153,
+      "grad_norm": 1.0693249702453613,
+      "learning_rate": 7.498229407086432e-05,
+      "loss": 1.5959,
+      "step": 27550
+    },
+    {
+      "epoch": 5.806077842734503,
+      "grad_norm": 1.0473322868347168,
+      "learning_rate": 7.491821000052064e-05,
+      "loss": 1.6197,
+      "step": 27560
+    },
+    {
+      "epoch": 5.808184547321852,
+      "grad_norm": 1.1137909889221191,
+      "learning_rate": 7.485413692030596e-05,
+      "loss": 1.6887,
+      "step": 27570
+    },
+    {
+      "epoch": 5.810291251909201,
+      "grad_norm": 1.1461670398712158,
+      "learning_rate": 7.479007485829523e-05,
+      "loss": 1.611,
+      "step": 27580
+    },
+    {
+      "epoch": 5.81239795649655,
+      "grad_norm": 1.1173228025436401,
+      "learning_rate": 7.472602384255864e-05,
+      "loss": 1.5538,
+      "step": 27590
+    },
+    {
+      "epoch": 5.8145046610839,
+      "grad_norm": 1.0433090925216675,
+      "learning_rate": 7.466198390116158e-05,
+      "loss": 1.6569,
+      "step": 27600
+    },
+    {
+      "epoch": 5.816611365671249,
+      "grad_norm": 1.139930248260498,
+      "learning_rate": 7.45979550621645e-05,
+      "loss": 1.591,
+      "step": 27610
+    },
+    {
+      "epoch": 5.818718070258598,
+      "grad_norm": 1.1254889965057373,
+      "learning_rate": 7.453393735362302e-05,
+      "loss": 1.61,
+      "step": 27620
+    },
+    {
+      "epoch": 5.820824774845947,
+      "grad_norm": 1.256567358970642,
+      "learning_rate": 7.446993080358789e-05,
+      "loss": 1.6197,
+      "step": 27630
+    },
+    {
+      "epoch": 5.822931479433296,
+      "grad_norm": 1.20447838306427,
+      "learning_rate": 7.440593544010495e-05,
+      "loss": 1.5921,
+      "step": 27640
+    },
+    {
+      "epoch": 5.825038184020646,
+      "grad_norm": 1.09376859664917,
+      "learning_rate": 7.434195129121518e-05,
+      "loss": 1.642,
+      "step": 27650
+    },
+    {
+      "epoch": 5.827144888607995,
+      "grad_norm": 1.2998801469802856,
+      "learning_rate": 7.427797838495463e-05,
+      "loss": 1.606,
+      "step": 27660
+    },
+    {
+      "epoch": 5.829251593195345,
+      "grad_norm": 1.0475540161132812,
+      "learning_rate": 7.421401674935435e-05,
+      "loss": 1.6485,
+      "step": 27670
+    },
+    {
+      "epoch": 5.831358297782693,
+      "grad_norm": 1.0838693380355835,
+      "learning_rate": 7.415006641244057e-05,
+      "loss": 1.6273,
+      "step": 27680
+    },
+    {
+      "epoch": 5.833465002370042,
+      "grad_norm": 1.039505958557129,
+      "learning_rate": 7.408612740223448e-05,
+      "loss": 1.5898,
+      "step": 27690
+    },
+    {
+      "epoch": 5.835571706957392,
+      "grad_norm": 1.1330510377883911,
+      "learning_rate": 7.40221997467524e-05,
+      "loss": 1.6115,
+      "step": 27700
+    },
+    {
+      "epoch": 5.837678411544741,
+      "grad_norm": 1.0775504112243652,
+      "learning_rate": 7.395828347400555e-05,
+      "loss": 1.6676,
+      "step": 27710
+    },
+    {
+      "epoch": 5.839785116132091,
+      "grad_norm": 1.0811083316802979,
+      "learning_rate": 7.389437861200024e-05,
+      "loss": 1.6371,
+      "step": 27720
+    },
+    {
+      "epoch": 5.8418918207194395,
+      "grad_norm": 1.005247712135315,
+      "learning_rate": 7.383048518873778e-05,
+      "loss": 1.6141,
+      "step": 27730
+    },
+    {
+      "epoch": 5.843998525306789,
+      "grad_norm": 1.088276982307434,
+      "learning_rate": 7.376660323221449e-05,
+      "loss": 1.6159,
+      "step": 27740
+    },
+    {
+      "epoch": 5.846105229894138,
+      "grad_norm": 1.114585518836975,
+      "learning_rate": 7.370273277042156e-05,
+      "loss": 1.6081,
+      "step": 27750
+    },
+    {
+      "epoch": 5.848211934481487,
+      "grad_norm": 1.0871803760528564,
+      "learning_rate": 7.363887383134527e-05,
+      "loss": 1.6395,
+      "step": 27760
+    },
+    {
+      "epoch": 5.850318639068837,
+      "grad_norm": 1.176871657371521,
+      "learning_rate": 7.357502644296677e-05,
+      "loss": 1.601,
+      "step": 27770
+    },
+    {
+      "epoch": 5.8524253436561855,
+      "grad_norm": 1.0000860691070557,
+      "learning_rate": 7.35111906332622e-05,
+      "loss": 1.6131,
+      "step": 27780
+    },
+    {
+      "epoch": 5.854532048243535,
+      "grad_norm": 1.133975863456726,
+      "learning_rate": 7.344736643020256e-05,
+      "loss": 1.5859,
+      "step": 27790
+    },
+    {
+      "epoch": 5.856638752830884,
+      "grad_norm": 1.1784876585006714,
+      "learning_rate": 7.338355386175382e-05,
+      "loss": 1.6133,
+      "step": 27800
+    },
+    {
+      "epoch": 5.858745457418234,
+      "grad_norm": 1.029212474822998,
+      "learning_rate": 7.331975295587687e-05,
+      "loss": 1.6396,
+      "step": 27810
+    },
+    {
+      "epoch": 5.860852162005583,
+      "grad_norm": 1.0330170392990112,
+      "learning_rate": 7.325596374052743e-05,
+      "loss": 1.618,
+      "step": 27820
+    },
+    {
+      "epoch": 5.8629588665929315,
+      "grad_norm": 1.0614304542541504,
+      "learning_rate": 7.319218624365613e-05,
+      "loss": 1.6479,
+      "step": 27830
+    },
+    {
+      "epoch": 5.865065571180281,
+      "grad_norm": 1.1751939058303833,
+      "learning_rate": 7.312842049320844e-05,
+      "loss": 1.6193,
+      "step": 27840
+    },
+    {
+      "epoch": 5.86717227576763,
+      "grad_norm": 1.1354511976242065,
+      "learning_rate": 7.306466651712474e-05,
+      "loss": 1.5616,
+      "step": 27850
+    },
+    {
+      "epoch": 5.86927898035498,
+      "grad_norm": 1.0025415420532227,
+      "learning_rate": 7.30009243433402e-05,
+      "loss": 1.5796,
+      "step": 27860
+    },
+    {
+      "epoch": 5.871385684942329,
+      "grad_norm": 1.099372148513794,
+      "learning_rate": 7.293719399978482e-05,
+      "loss": 1.621,
+      "step": 27870
+    },
+    {
+      "epoch": 5.8734923895296784,
+      "grad_norm": 1.2273434400558472,
+      "learning_rate": 7.287347551438344e-05,
+      "loss": 1.6476,
+      "step": 27880
+    },
+    {
+      "epoch": 5.875599094117027,
+      "grad_norm": 1.0725783109664917,
+      "learning_rate": 7.280976891505569e-05,
+      "loss": 1.6296,
+      "step": 27890
+    },
+    {
+      "epoch": 5.877705798704377,
+      "grad_norm": 1.0418058633804321,
+      "learning_rate": 7.2746074229716e-05,
+      "loss": 1.5929,
+      "step": 27900
+    },
+    {
+      "epoch": 5.879812503291726,
+      "grad_norm": 1.028952717781067,
+      "learning_rate": 7.268239148627355e-05,
+      "loss": 1.6189,
+      "step": 27910
+    },
+    {
+      "epoch": 5.881919207879076,
+      "grad_norm": 1.1607900857925415,
+      "learning_rate": 7.261872071263236e-05,
+      "loss": 1.6088,
+      "step": 27920
+    },
+    {
+      "epoch": 5.8840259124664245,
+      "grad_norm": 1.19911789894104,
+      "learning_rate": 7.25550619366911e-05,
+      "loss": 1.5934,
+      "step": 27930
+    },
+    {
+      "epoch": 5.886132617053773,
+      "grad_norm": 1.0657827854156494,
+      "learning_rate": 7.249141518634327e-05,
+      "loss": 1.6062,
+      "step": 27940
+    },
+    {
+      "epoch": 5.888239321641123,
+      "grad_norm": 1.0900195837020874,
+      "learning_rate": 7.242778048947713e-05,
+      "loss": 1.6429,
+      "step": 27950
+    },
+    {
+      "epoch": 5.890346026228472,
+      "grad_norm": 1.1822017431259155,
+      "learning_rate": 7.236415787397548e-05,
+      "loss": 1.5911,
+      "step": 27960
+    },
+    {
+      "epoch": 5.892452730815822,
+      "grad_norm": 1.1533838510513306,
+      "learning_rate": 7.230054736771601e-05,
+      "loss": 1.619,
+      "step": 27970
+    },
+    {
+      "epoch": 5.8945594354031705,
+      "grad_norm": 0.9922227263450623,
+      "learning_rate": 7.223694899857103e-05,
+      "loss": 1.6274,
+      "step": 27980
+    },
+    {
+      "epoch": 5.89666613999052,
+      "grad_norm": 1.097437858581543,
+      "learning_rate": 7.217336279440761e-05,
+      "loss": 1.6355,
+      "step": 27990
+    },
+    {
+      "epoch": 5.898772844577869,
+      "grad_norm": 1.1534727811813354,
+      "learning_rate": 7.210978878308729e-05,
+      "loss": 1.629,
+      "step": 28000
+    },
+    {
+      "epoch": 5.900879549165218,
+      "grad_norm": 1.0968070030212402,
+      "learning_rate": 7.204622699246646e-05,
+      "loss": 1.5983,
+      "step": 28010
+    },
+    {
+      "epoch": 5.902986253752568,
+      "grad_norm": 1.0842479467391968,
+      "learning_rate": 7.198267745039612e-05,
+      "loss": 1.6278,
+      "step": 28020
+    },
+    {
+      "epoch": 5.9050929583399165,
+      "grad_norm": 1.136145830154419,
+      "learning_rate": 7.19191401847219e-05,
+      "loss": 1.5965,
+      "step": 28030
+    },
+    {
+      "epoch": 5.907199662927266,
+      "grad_norm": 1.0902760028839111,
+      "learning_rate": 7.185561522328395e-05,
+      "loss": 1.6764,
+      "step": 28040
+    },
+    {
+      "epoch": 5.909306367514615,
+      "grad_norm": 1.0746238231658936,
+      "learning_rate": 7.179210259391709e-05,
+      "loss": 1.6268,
+      "step": 28050
+    },
+    {
+      "epoch": 5.911413072101965,
+      "grad_norm": 0.9946722984313965,
+      "learning_rate": 7.172860232445084e-05,
+      "loss": 1.6565,
+      "step": 28060
+    },
+    {
+      "epoch": 5.913519776689314,
+      "grad_norm": 1.1601204872131348,
+      "learning_rate": 7.166511444270924e-05,
+      "loss": 1.6128,
+      "step": 28070
+    },
+    {
+      "epoch": 5.915626481276663,
+      "grad_norm": 1.1329623460769653,
+      "learning_rate": 7.160163897651075e-05,
+      "loss": 1.6341,
+      "step": 28080
+    },
+    {
+      "epoch": 5.917733185864012,
+      "grad_norm": 1.0780909061431885,
+      "learning_rate": 7.153817595366858e-05,
+      "loss": 1.6146,
+      "step": 28090
+    },
+    {
+      "epoch": 5.919839890451361,
+      "grad_norm": 1.1066025495529175,
+      "learning_rate": 7.147472540199046e-05,
+      "loss": 1.613,
+      "step": 28100
+    },
+    {
+      "epoch": 5.921946595038711,
+      "grad_norm": 1.0269134044647217,
+      "learning_rate": 7.141128734927863e-05,
+      "loss": 1.6617,
+      "step": 28110
+    },
+    {
+      "epoch": 5.92405329962606,
+      "grad_norm": 1.249660849571228,
+      "learning_rate": 7.134786182332978e-05,
+      "loss": 1.6327,
+      "step": 28120
+    },
+    {
+      "epoch": 5.9261600042134095,
+      "grad_norm": 0.9970906972885132,
+      "learning_rate": 7.128444885193518e-05,
+      "loss": 1.6346,
+      "step": 28130
+    },
+    {
+      "epoch": 5.928266708800758,
+      "grad_norm": 1.132813572883606,
+      "learning_rate": 7.122104846288064e-05,
+      "loss": 1.6349,
+      "step": 28140
+    },
+    {
+      "epoch": 5.930373413388107,
+      "grad_norm": 1.0950262546539307,
+      "learning_rate": 7.115766068394645e-05,
+      "loss": 1.6456,
+      "step": 28150
+    },
+    {
+      "epoch": 5.932480117975457,
+      "grad_norm": 1.1574265956878662,
+      "learning_rate": 7.109428554290725e-05,
+      "loss": 1.6806,
+      "step": 28160
+    },
+    {
+      "epoch": 5.934586822562806,
+      "grad_norm": 1.1822057962417603,
+      "learning_rate": 7.103092306753222e-05,
+      "loss": 1.6143,
+      "step": 28170
+    },
+    {
+      "epoch": 5.9366935271501555,
+      "grad_norm": 1.03468656539917,
+      "learning_rate": 7.096757328558506e-05,
+      "loss": 1.6355,
+      "step": 28180
+    },
+    {
+      "epoch": 5.938800231737504,
+      "grad_norm": 1.0852649211883545,
+      "learning_rate": 7.090423622482389e-05,
+      "loss": 1.6394,
+      "step": 28190
+    },
+    {
+      "epoch": 5.940906936324854,
+      "grad_norm": 1.0834746360778809,
+      "learning_rate": 7.08409119130011e-05,
+      "loss": 1.6541,
+      "step": 28200
+    },
+    {
+      "epoch": 5.943013640912203,
+      "grad_norm": 1.068300724029541,
+      "learning_rate": 7.077760037786365e-05,
+      "loss": 1.5601,
+      "step": 28210
+    },
+    {
+      "epoch": 5.945120345499553,
+      "grad_norm": 1.1912277936935425,
+      "learning_rate": 7.071430164715288e-05,
+      "loss": 1.6005,
+      "step": 28220
+    },
+    {
+      "epoch": 5.947227050086902,
+      "grad_norm": 1.164495825767517,
+      "learning_rate": 7.065101574860449e-05,
+      "loss": 1.5994,
+      "step": 28230
+    },
+    {
+      "epoch": 5.949333754674251,
+      "grad_norm": 1.1123220920562744,
+      "learning_rate": 7.058774270994862e-05,
+      "loss": 1.6172,
+      "step": 28240
+    },
+    {
+      "epoch": 5.9514404592616,
+      "grad_norm": 1.1080528497695923,
+      "learning_rate": 7.052448255890957e-05,
+      "loss": 1.6239,
+      "step": 28250
+    },
+    {
+      "epoch": 5.953547163848949,
+      "grad_norm": 1.1013644933700562,
+      "learning_rate": 7.04612353232063e-05,
+      "loss": 1.6834,
+      "step": 28260
+    },
+    {
+      "epoch": 5.955653868436299,
+      "grad_norm": 1.1154600381851196,
+      "learning_rate": 7.039800103055186e-05,
+      "loss": 1.5973,
+      "step": 28270
+    },
+    {
+      "epoch": 5.957760573023648,
+      "grad_norm": 1.1243196725845337,
+      "learning_rate": 7.033477970865381e-05,
+      "loss": 1.6159,
+      "step": 28280
+    },
+    {
+      "epoch": 5.959867277610997,
+      "grad_norm": 1.254236102104187,
+      "learning_rate": 7.027157138521383e-05,
+      "loss": 1.6784,
+      "step": 28290
+    },
+    {
+      "epoch": 5.961973982198346,
+      "grad_norm": 1.1903520822525024,
+      "learning_rate": 7.02083760879281e-05,
+      "loss": 1.6246,
+      "step": 28300
+    },
+    {
+      "epoch": 5.964080686785696,
+      "grad_norm": 1.1204966306686401,
+      "learning_rate": 7.014519384448696e-05,
+      "loss": 1.6253,
+      "step": 28310
+    },
+    {
+      "epoch": 5.966187391373045,
+      "grad_norm": 1.0842779874801636,
+      "learning_rate": 7.008202468257514e-05,
+      "loss": 1.6459,
+      "step": 28320
+    },
+    {
+      "epoch": 5.968294095960394,
+      "grad_norm": 1.201533317565918,
+      "learning_rate": 7.001886862987147e-05,
+      "loss": 1.5967,
+      "step": 28330
+    },
+    {
+      "epoch": 5.970400800547743,
+      "grad_norm": 1.0753484964370728,
+      "learning_rate": 6.995572571404923e-05,
+      "loss": 1.6426,
+      "step": 28340
+    },
+    {
+      "epoch": 5.972507505135092,
+      "grad_norm": 1.1155014038085938,
+      "learning_rate": 6.989259596277582e-05,
+      "loss": 1.6384,
+      "step": 28350
+    },
+    {
+      "epoch": 5.974614209722442,
+      "grad_norm": 1.0992554426193237,
+      "learning_rate": 6.982947940371293e-05,
+      "loss": 1.6077,
+      "step": 28360
+    },
+    {
+      "epoch": 5.976720914309791,
+      "grad_norm": 1.2092218399047852,
+      "learning_rate": 6.97663760645164e-05,
+      "loss": 1.6299,
+      "step": 28370
+    },
+    {
+      "epoch": 5.9788276188971405,
+      "grad_norm": 1.1124134063720703,
+      "learning_rate": 6.970328597283637e-05,
+      "loss": 1.629,
+      "step": 28380
+    },
+    {
+      "epoch": 5.980934323484489,
+      "grad_norm": 1.0230249166488647,
+      "learning_rate": 6.964020915631711e-05,
+      "loss": 1.6255,
+      "step": 28390
+    },
+    {
+      "epoch": 5.983041028071838,
+      "grad_norm": 1.0409824848175049,
+      "learning_rate": 6.957714564259712e-05,
+      "loss": 1.5886,
+      "step": 28400
+    },
+    {
+      "epoch": 5.985147732659188,
+      "grad_norm": 1.25199294090271,
+      "learning_rate": 6.951409545930895e-05,
+      "loss": 1.6144,
+      "step": 28410
+    },
+    {
+      "epoch": 5.987254437246537,
+      "grad_norm": 1.1676099300384521,
+      "learning_rate": 6.945105863407951e-05,
+      "loss": 1.5898,
+      "step": 28420
+    },
+    {
+      "epoch": 5.989361141833887,
+      "grad_norm": 1.007283329963684,
+      "learning_rate": 6.93880351945297e-05,
+      "loss": 1.6614,
+      "step": 28430
+    },
+    {
+      "epoch": 5.991467846421235,
+      "grad_norm": 1.2138159275054932,
+      "learning_rate": 6.932502516827461e-05,
+      "loss": 1.6746,
+      "step": 28440
+    },
+    {
+      "epoch": 5.993574551008585,
+      "grad_norm": 1.0807868242263794,
+      "learning_rate": 6.926202858292345e-05,
+      "loss": 1.6564,
+      "step": 28450
+    },
+    {
+      "epoch": 5.995681255595934,
+      "grad_norm": 1.0263831615447998,
+      "learning_rate": 6.919904546607954e-05,
+      "loss": 1.5506,
+      "step": 28460
+    },
+    {
+      "epoch": 5.997787960183283,
+      "grad_norm": 1.1143971681594849,
+      "learning_rate": 6.913607584534026e-05,
+      "loss": 1.6399,
+      "step": 28470
+    },
+    {
+      "epoch": 5.999894664770633,
+      "grad_norm": 1.2648664712905884,
+      "learning_rate": 6.907311974829716e-05,
+      "loss": 1.5995,
+      "step": 28480
+    },
+    {
+      "epoch": 6.0020013693579815,
+      "grad_norm": 1.0600049495697021,
+      "learning_rate": 6.901017720253583e-05,
+      "loss": 1.5734,
+      "step": 28490
+    },
+    {
+      "epoch": 6.004108073945331,
+      "grad_norm": 1.108899712562561,
+      "learning_rate": 6.894724823563583e-05,
+      "loss": 1.5201,
+      "step": 28500
+    },
+    {
+      "epoch": 6.00621477853268,
+      "grad_norm": 1.150876760482788,
+      "learning_rate": 6.888433287517088e-05,
+      "loss": 1.6381,
+      "step": 28510
+    },
+    {
+      "epoch": 6.00832148312003,
+      "grad_norm": 1.0762908458709717,
+      "learning_rate": 6.882143114870876e-05,
+      "loss": 1.5533,
+      "step": 28520
+    },
+    {
+      "epoch": 6.010428187707379,
+      "grad_norm": 1.0255169868469238,
+      "learning_rate": 6.875854308381118e-05,
+      "loss": 1.6005,
+      "step": 28530
+    },
+    {
+      "epoch": 6.012534892294728,
+      "grad_norm": 1.126582145690918,
+      "learning_rate": 6.869566870803388e-05,
+      "loss": 1.5961,
+      "step": 28540
+    },
+    {
+      "epoch": 6.014641596882077,
+      "grad_norm": 1.1807374954223633,
+      "learning_rate": 6.863280804892668e-05,
+      "loss": 1.5693,
+      "step": 28550
+    },
+    {
+      "epoch": 6.016748301469426,
+      "grad_norm": 1.0278984308242798,
+      "learning_rate": 6.85699611340333e-05,
+      "loss": 1.6046,
+      "step": 28560
+    },
+    {
+      "epoch": 6.018855006056776,
+      "grad_norm": 1.2131428718566895,
+      "learning_rate": 6.850712799089151e-05,
+      "loss": 1.527,
+      "step": 28570
+    },
+    {
+      "epoch": 6.020961710644125,
+      "grad_norm": 1.1807914972305298,
+      "learning_rate": 6.844430864703298e-05,
+      "loss": 1.6019,
+      "step": 28580
+    },
+    {
+      "epoch": 6.023068415231474,
+      "grad_norm": 1.1616885662078857,
+      "learning_rate": 6.838150312998338e-05,
+      "loss": 1.6801,
+      "step": 28590
+    },
+    {
+      "epoch": 6.025175119818823,
+      "grad_norm": 1.0600801706314087,
+      "learning_rate": 6.831871146726228e-05,
+      "loss": 1.5839,
+      "step": 28600
+    },
+    {
+      "epoch": 6.027281824406173,
+      "grad_norm": 1.0691717863082886,
+      "learning_rate": 6.825593368638327e-05,
+      "loss": 1.6165,
+      "step": 28610
+    },
+    {
+      "epoch": 6.029388528993522,
+      "grad_norm": 1.1156972646713257,
+      "learning_rate": 6.819316981485372e-05,
+      "loss": 1.5783,
+      "step": 28620
+    },
+    {
+      "epoch": 6.031495233580871,
+      "grad_norm": 1.1417192220687866,
+      "learning_rate": 6.813041988017501e-05,
+      "loss": 1.6113,
+      "step": 28630
+    },
+    {
+      "epoch": 6.03360193816822,
+      "grad_norm": 1.1809382438659668,
+      "learning_rate": 6.806768390984237e-05,
+      "loss": 1.6015,
+      "step": 28640
+    },
+    {
+      "epoch": 6.035708642755569,
+      "grad_norm": 1.2206168174743652,
+      "learning_rate": 6.800496193134498e-05,
+      "loss": 1.5658,
+      "step": 28650
+    },
+    {
+      "epoch": 6.037815347342919,
+      "grad_norm": 1.0852173566818237,
+      "learning_rate": 6.794225397216575e-05,
+      "loss": 1.5586,
+      "step": 28660
+    },
+    {
+      "epoch": 6.039922051930268,
+      "grad_norm": 1.072217583656311,
+      "learning_rate": 6.787956005978156e-05,
+      "loss": 1.5642,
+      "step": 28670
+    },
+    {
+      "epoch": 6.042028756517618,
+      "grad_norm": 1.183638334274292,
+      "learning_rate": 6.781688022166311e-05,
+      "loss": 1.6155,
+      "step": 28680
+    },
+    {
+      "epoch": 6.0441354611049665,
+      "grad_norm": 1.0388456583023071,
+      "learning_rate": 6.775421448527496e-05,
+      "loss": 1.5514,
+      "step": 28690
+    },
+    {
+      "epoch": 6.046242165692316,
+      "grad_norm": 1.0287905931472778,
+      "learning_rate": 6.76915628780754e-05,
+      "loss": 1.5245,
+      "step": 28700
+    },
+    {
+      "epoch": 6.048348870279665,
+      "grad_norm": 1.1689507961273193,
+      "learning_rate": 6.76289254275166e-05,
+      "loss": 1.5968,
+      "step": 28710
+    },
+    {
+      "epoch": 6.050455574867014,
+      "grad_norm": 1.067297101020813,
+      "learning_rate": 6.756630216104454e-05,
+      "loss": 1.6348,
+      "step": 28720
+    },
+    {
+      "epoch": 6.052562279454364,
+      "grad_norm": 1.0096595287322998,
+      "learning_rate": 6.750369310609894e-05,
+      "loss": 1.5733,
+      "step": 28730
+    },
+    {
+      "epoch": 6.0546689840417125,
+      "grad_norm": 1.132238745689392,
+      "learning_rate": 6.744109829011332e-05,
+      "loss": 1.5642,
+      "step": 28740
+    },
+    {
+      "epoch": 6.056775688629062,
+      "grad_norm": 1.286741018295288,
+      "learning_rate": 6.737851774051495e-05,
+      "loss": 1.6235,
+      "step": 28750
+    },
+    {
+      "epoch": 6.058882393216411,
+      "grad_norm": 1.1896122694015503,
+      "learning_rate": 6.731595148472485e-05,
+      "loss": 1.5754,
+      "step": 28760
+    },
+    {
+      "epoch": 6.060989097803761,
+      "grad_norm": 1.0230028629302979,
+      "learning_rate": 6.725339955015777e-05,
+      "loss": 1.5619,
+      "step": 28770
+    },
+    {
+      "epoch": 6.06309580239111,
+      "grad_norm": 1.1299915313720703,
+      "learning_rate": 6.719086196422225e-05,
+      "loss": 1.6199,
+      "step": 28780
+    },
+    {
+      "epoch": 6.0652025069784585,
+      "grad_norm": 1.1828107833862305,
+      "learning_rate": 6.712833875432038e-05,
+      "loss": 1.6065,
+      "step": 28790
+    },
+    {
+      "epoch": 6.067309211565808,
+      "grad_norm": 1.1760649681091309,
+      "learning_rate": 6.706582994784814e-05,
+      "loss": 1.5737,
+      "step": 28800
+    },
+    {
+      "epoch": 6.069415916153157,
+      "grad_norm": 1.1444995403289795,
+      "learning_rate": 6.700333557219511e-05,
+      "loss": 1.5816,
+      "step": 28810
+    },
+    {
+      "epoch": 6.071522620740507,
+      "grad_norm": 1.1875180006027222,
+      "learning_rate": 6.694085565474453e-05,
+      "loss": 1.5932,
+      "step": 28820
+    },
+    {
+      "epoch": 6.073629325327856,
+      "grad_norm": 1.281719446182251,
+      "learning_rate": 6.687839022287332e-05,
+      "loss": 1.6174,
+      "step": 28830
+    },
+    {
+      "epoch": 6.0757360299152054,
+      "grad_norm": 1.2580798864364624,
+      "learning_rate": 6.681593930395209e-05,
+      "loss": 1.5437,
+      "step": 28840
+    },
+    {
+      "epoch": 6.077842734502554,
+      "grad_norm": 1.21761155128479,
+      "learning_rate": 6.675350292534504e-05,
+      "loss": 1.6229,
+      "step": 28850
+    },
+    {
+      "epoch": 6.079949439089904,
+      "grad_norm": 1.111264944076538,
+      "learning_rate": 6.669108111441003e-05,
+      "loss": 1.5935,
+      "step": 28860
+    },
+    {
+      "epoch": 6.082056143677253,
+      "grad_norm": 1.094279408454895,
+      "learning_rate": 6.662867389849851e-05,
+      "loss": 1.5806,
+      "step": 28870
+    },
+    {
+      "epoch": 6.084162848264602,
+      "grad_norm": 1.3840709924697876,
+      "learning_rate": 6.656628130495558e-05,
+      "loss": 1.5542,
+      "step": 28880
+    },
+    {
+      "epoch": 6.0862695528519515,
+      "grad_norm": 1.1345621347427368,
+      "learning_rate": 6.650390336111989e-05,
+      "loss": 1.6027,
+      "step": 28890
+    },
+    {
+      "epoch": 6.0883762574393,
+      "grad_norm": 1.1738914251327515,
+      "learning_rate": 6.644154009432369e-05,
+      "loss": 1.5833,
+      "step": 28900
+    },
+    {
+      "epoch": 6.09048296202665,
+      "grad_norm": 1.157949686050415,
+      "learning_rate": 6.637919153189279e-05,
+      "loss": 1.5879,
+      "step": 28910
+    },
+    {
+      "epoch": 6.092589666613999,
+      "grad_norm": 1.1940367221832275,
+      "learning_rate": 6.631685770114654e-05,
+      "loss": 1.5931,
+      "step": 28920
+    },
+    {
+      "epoch": 6.094696371201349,
+      "grad_norm": 1.1197195053100586,
+      "learning_rate": 6.625453862939789e-05,
+      "loss": 1.6045,
+      "step": 28930
+    },
+    {
+      "epoch": 6.0968030757886975,
+      "grad_norm": 1.4872801303863525,
+      "learning_rate": 6.619223434395329e-05,
+      "loss": 1.5962,
+      "step": 28940
+    },
+    {
+      "epoch": 6.098909780376046,
+      "grad_norm": 1.1432135105133057,
+      "learning_rate": 6.61299448721127e-05,
+      "loss": 1.5848,
+      "step": 28950
+    },
+    {
+      "epoch": 6.101016484963396,
+      "grad_norm": 1.1360822916030884,
+      "learning_rate": 6.606767024116957e-05,
+      "loss": 1.5676,
+      "step": 28960
+    },
+    {
+      "epoch": 6.103123189550745,
+      "grad_norm": 1.1027390956878662,
+      "learning_rate": 6.600541047841093e-05,
+      "loss": 1.6078,
+      "step": 28970
+    },
+    {
+      "epoch": 6.105229894138095,
+      "grad_norm": 1.1606420278549194,
+      "learning_rate": 6.594316561111724e-05,
+      "loss": 1.5931,
+      "step": 28980
+    },
+    {
+      "epoch": 6.1073365987254435,
+      "grad_norm": 1.0428427457809448,
+      "learning_rate": 6.588093566656238e-05,
+      "loss": 1.5535,
+      "step": 28990
+    },
+    {
+      "epoch": 6.109443303312793,
+      "grad_norm": 1.0354915857315063,
+      "learning_rate": 6.581872067201378e-05,
+      "loss": 1.595,
+      "step": 29000
+    },
+    {
+      "epoch": 6.111550007900142,
+      "grad_norm": 1.0693248510360718,
+      "learning_rate": 6.57565206547323e-05,
+      "loss": 1.5438,
+      "step": 29010
+    },
+    {
+      "epoch": 6.113656712487492,
+      "grad_norm": 1.1684085130691528,
+      "learning_rate": 6.569433564197222e-05,
+      "loss": 1.6196,
+      "step": 29020
+    },
+    {
+      "epoch": 6.115763417074841,
+      "grad_norm": 1.12764310836792,
+      "learning_rate": 6.563216566098121e-05,
+      "loss": 1.501,
+      "step": 29030
+    },
+    {
+      "epoch": 6.11787012166219,
+      "grad_norm": 1.1445949077606201,
+      "learning_rate": 6.557001073900044e-05,
+      "loss": 1.5407,
+      "step": 29040
+    },
+    {
+      "epoch": 6.119976826249539,
+      "grad_norm": 1.206442952156067,
+      "learning_rate": 6.55078709032644e-05,
+      "loss": 1.6259,
+      "step": 29050
+    },
+    {
+      "epoch": 6.122083530836888,
+      "grad_norm": 1.1316516399383545,
+      "learning_rate": 6.544574618100102e-05,
+      "loss": 1.6372,
+      "step": 29060
+    },
+    {
+      "epoch": 6.124190235424238,
+      "grad_norm": 1.1655209064483643,
+      "learning_rate": 6.538363659943162e-05,
+      "loss": 1.6235,
+      "step": 29070
+    },
+    {
+      "epoch": 6.126296940011587,
+      "grad_norm": 1.2157286405563354,
+      "learning_rate": 6.53215421857708e-05,
+      "loss": 1.5969,
+      "step": 29080
+    },
+    {
+      "epoch": 6.1284036445989365,
+      "grad_norm": 1.1558120250701904,
+      "learning_rate": 6.525946296722659e-05,
+      "loss": 1.5945,
+      "step": 29090
+    },
+    {
+      "epoch": 6.130510349186285,
+      "grad_norm": 1.1660462617874146,
+      "learning_rate": 6.519739897100034e-05,
+      "loss": 1.6076,
+      "step": 29100
+    },
+    {
+      "epoch": 6.132617053773634,
+      "grad_norm": 1.1626811027526855,
+      "learning_rate": 6.51353502242868e-05,
+      "loss": 1.5999,
+      "step": 29110
+    },
+    {
+      "epoch": 6.134723758360984,
+      "grad_norm": 1.0529296398162842,
+      "learning_rate": 6.507331675427387e-05,
+      "loss": 1.5767,
+      "step": 29120
+    },
+    {
+      "epoch": 6.136830462948333,
+      "grad_norm": 1.1551191806793213,
+      "learning_rate": 6.50112985881429e-05,
+      "loss": 1.6153,
+      "step": 29130
+    },
+    {
+      "epoch": 6.1389371675356825,
+      "grad_norm": 1.0937594175338745,
+      "learning_rate": 6.494929575306848e-05,
+      "loss": 1.5513,
+      "step": 29140
+    },
+    {
+      "epoch": 6.141043872123031,
+      "grad_norm": 1.1500444412231445,
+      "learning_rate": 6.488730827621856e-05,
+      "loss": 1.5965,
+      "step": 29150
+    },
+    {
+      "epoch": 6.143150576710381,
+      "grad_norm": 1.208426594734192,
+      "learning_rate": 6.482533618475422e-05,
+      "loss": 1.4839,
+      "step": 29160
+    },
+    {
+      "epoch": 6.14525728129773,
+      "grad_norm": 1.1101106405258179,
+      "learning_rate": 6.476337950582987e-05,
+      "loss": 1.5582,
+      "step": 29170
+    },
+    {
+      "epoch": 6.147363985885079,
+      "grad_norm": 1.1369212865829468,
+      "learning_rate": 6.470143826659317e-05,
+      "loss": 1.6388,
+      "step": 29180
+    },
+    {
+      "epoch": 6.149470690472429,
+      "grad_norm": 1.0529881715774536,
+      "learning_rate": 6.46395124941851e-05,
+      "loss": 1.6363,
+      "step": 29190
+    },
+    {
+      "epoch": 6.151577395059777,
+      "grad_norm": 1.0460222959518433,
+      "learning_rate": 6.457760221573968e-05,
+      "loss": 1.5687,
+      "step": 29200
+    },
+    {
+      "epoch": 6.153684099647127,
+      "grad_norm": 1.164514183998108,
+      "learning_rate": 6.451570745838426e-05,
+      "loss": 1.5877,
+      "step": 29210
+    },
+    {
+      "epoch": 6.155790804234476,
+      "grad_norm": 1.1551470756530762,
+      "learning_rate": 6.445382824923938e-05,
+      "loss": 1.618,
+      "step": 29220
+    },
+    {
+      "epoch": 6.157897508821826,
+      "grad_norm": 1.1310980319976807,
+      "learning_rate": 6.43919646154188e-05,
+      "loss": 1.5745,
+      "step": 29230
+    },
+    {
+      "epoch": 6.160004213409175,
+      "grad_norm": 1.1333801746368408,
+      "learning_rate": 6.433011658402933e-05,
+      "loss": 1.5899,
+      "step": 29240
+    },
+    {
+      "epoch": 6.162110917996524,
+      "grad_norm": 1.0840401649475098,
+      "learning_rate": 6.426828418217104e-05,
+      "loss": 1.5188,
+      "step": 29250
+    },
+    {
+      "epoch": 6.164217622583873,
+      "grad_norm": 1.150305986404419,
+      "learning_rate": 6.420646743693714e-05,
+      "loss": 1.5772,
+      "step": 29260
+    },
+    {
+      "epoch": 6.166324327171222,
+      "grad_norm": 1.1294101476669312,
+      "learning_rate": 6.414466637541405e-05,
+      "loss": 1.5676,
+      "step": 29270
+    },
+    {
+      "epoch": 6.168431031758572,
+      "grad_norm": 1.1302536725997925,
+      "learning_rate": 6.408288102468113e-05,
+      "loss": 1.6208,
+      "step": 29280
+    },
+    {
+      "epoch": 6.170537736345921,
+      "grad_norm": 1.1266132593154907,
+      "learning_rate": 6.402111141181101e-05,
+      "loss": 1.6378,
+      "step": 29290
+    },
+    {
+      "epoch": 6.17264444093327,
+      "grad_norm": 1.1256543397903442,
+      "learning_rate": 6.39593575638694e-05,
+      "loss": 1.6542,
+      "step": 29300
+    },
+    {
+      "epoch": 6.174751145520619,
+      "grad_norm": 1.1547867059707642,
+      "learning_rate": 6.38976195079151e-05,
+      "loss": 1.6013,
+      "step": 29310
+    },
+    {
+      "epoch": 6.176857850107969,
+      "grad_norm": 1.083693504333496,
+      "learning_rate": 6.38358972709999e-05,
+      "loss": 1.6854,
+      "step": 29320
+    },
+    {
+      "epoch": 6.178964554695318,
+      "grad_norm": 1.1603076457977295,
+      "learning_rate": 6.377419088016881e-05,
+      "loss": 1.6129,
+      "step": 29330
+    },
+    {
+      "epoch": 6.1810712592826675,
+      "grad_norm": 1.0910979509353638,
+      "learning_rate": 6.371250036245976e-05,
+      "loss": 1.5678,
+      "step": 29340
+    },
+    {
+      "epoch": 6.183177963870016,
+      "grad_norm": 1.075230598449707,
+      "learning_rate": 6.365082574490384e-05,
+      "loss": 1.5757,
+      "step": 29350
+    },
+    {
+      "epoch": 6.185284668457365,
+      "grad_norm": 1.2227821350097656,
+      "learning_rate": 6.358916705452514e-05,
+      "loss": 1.5971,
+      "step": 29360
+    },
+    {
+      "epoch": 6.187391373044715,
+      "grad_norm": 1.148966908454895,
+      "learning_rate": 6.352752431834063e-05,
+      "loss": 1.5922,
+      "step": 29370
+    },
+    {
+      "epoch": 6.189498077632064,
+      "grad_norm": 1.168792963027954,
+      "learning_rate": 6.34658975633605e-05,
+      "loss": 1.5615,
+      "step": 29380
+    },
+    {
+      "epoch": 6.191604782219414,
+      "grad_norm": 1.1150977611541748,
+      "learning_rate": 6.340428681658783e-05,
+      "loss": 1.6151,
+      "step": 29390
+    },
+    {
+      "epoch": 6.193711486806762,
+      "grad_norm": 1.0817590951919556,
+      "learning_rate": 6.334269210501875e-05,
+      "loss": 1.6724,
+      "step": 29400
+    },
+    {
+      "epoch": 6.195818191394112,
+      "grad_norm": 1.126131296157837,
+      "learning_rate": 6.328111345564221e-05,
+      "loss": 1.5664,
+      "step": 29410
+    },
+    {
+      "epoch": 6.197924895981461,
+      "grad_norm": 1.2539746761322021,
+      "learning_rate": 6.321955089544029e-05,
+      "loss": 1.5707,
+      "step": 29420
+    },
+    {
+      "epoch": 6.20003160056881,
+      "grad_norm": 1.1847485303878784,
+      "learning_rate": 6.315800445138796e-05,
+      "loss": 1.5636,
+      "step": 29430
+    },
+    {
+      "epoch": 6.20213830515616,
+      "grad_norm": 1.1532986164093018,
+      "learning_rate": 6.309647415045315e-05,
+      "loss": 1.5578,
+      "step": 29440
+    },
+    {
+      "epoch": 6.2042450097435085,
+      "grad_norm": 1.2231844663619995,
+      "learning_rate": 6.30349600195966e-05,
+      "loss": 1.5815,
+      "step": 29450
+    },
+    {
+      "epoch": 6.206351714330858,
+      "grad_norm": 1.1287641525268555,
+      "learning_rate": 6.297346208577213e-05,
+      "loss": 1.5534,
+      "step": 29460
+    },
+    {
+      "epoch": 6.208458418918207,
+      "grad_norm": 1.108749270439148,
+      "learning_rate": 6.291198037592639e-05,
+      "loss": 1.5468,
+      "step": 29470
+    },
+    {
+      "epoch": 6.210565123505557,
+      "grad_norm": 1.1339576244354248,
+      "learning_rate": 6.285051491699896e-05,
+      "loss": 1.5976,
+      "step": 29480
+    },
+    {
+      "epoch": 6.212671828092906,
+      "grad_norm": 1.1239781379699707,
+      "learning_rate": 6.278906573592213e-05,
+      "loss": 1.6005,
+      "step": 29490
+    },
+    {
+      "epoch": 6.2147785326802545,
+      "grad_norm": 1.057753562927246,
+      "learning_rate": 6.272763285962129e-05,
+      "loss": 1.5794,
+      "step": 29500
+    },
+    {
+      "epoch": 6.216885237267604,
+      "grad_norm": 1.1971989870071411,
+      "learning_rate": 6.266621631501457e-05,
+      "loss": 1.574,
+      "step": 29510
+    },
+    {
+      "epoch": 6.218991941854953,
+      "grad_norm": 1.1916954517364502,
+      "learning_rate": 6.260481612901299e-05,
+      "loss": 1.5978,
+      "step": 29520
+    },
+    {
+      "epoch": 6.221098646442303,
+      "grad_norm": 1.0900837182998657,
+      "learning_rate": 6.254343232852027e-05,
+      "loss": 1.5698,
+      "step": 29530
+    },
+    {
+      "epoch": 6.223205351029652,
+      "grad_norm": 1.1096452474594116,
+      "learning_rate": 6.248206494043313e-05,
+      "loss": 1.5639,
+      "step": 29540
+    },
+    {
+      "epoch": 6.225312055617001,
+      "grad_norm": 1.263198733329773,
+      "learning_rate": 6.2420713991641e-05,
+      "loss": 1.5872,
+      "step": 29550
+    },
+    {
+      "epoch": 6.22741876020435,
+      "grad_norm": 1.0684478282928467,
+      "learning_rate": 6.235937950902615e-05,
+      "loss": 1.6383,
+      "step": 29560
+    },
+    {
+      "epoch": 6.2295254647917,
+      "grad_norm": 1.1642452478408813,
+      "learning_rate": 6.229806151946353e-05,
+      "loss": 1.6017,
+      "step": 29570
+    },
+    {
+      "epoch": 6.231632169379049,
+      "grad_norm": 1.056420922279358,
+      "learning_rate": 6.223676004982105e-05,
+      "loss": 1.5695,
+      "step": 29580
+    },
+    {
+      "epoch": 6.233738873966398,
+      "grad_norm": 1.1780263185501099,
+      "learning_rate": 6.217547512695919e-05,
+      "loss": 1.6149,
+      "step": 29590
+    },
+    {
+      "epoch": 6.235845578553747,
+      "grad_norm": 1.1113979816436768,
+      "learning_rate": 6.211420677773131e-05,
+      "loss": 1.6249,
+      "step": 29600
+    },
+    {
+      "epoch": 6.237952283141096,
+      "grad_norm": 1.2237204313278198,
+      "learning_rate": 6.205295502898348e-05,
+      "loss": 1.6159,
+      "step": 29610
+    },
+    {
+      "epoch": 6.240058987728446,
+      "grad_norm": 1.144103765487671,
+      "learning_rate": 6.199171990755441e-05,
+      "loss": 1.6249,
+      "step": 29620
+    },
+    {
+      "epoch": 6.242165692315795,
+      "grad_norm": 1.0986157655715942,
+      "learning_rate": 6.193050144027565e-05,
+      "loss": 1.5917,
+      "step": 29630
+    },
+    {
+      "epoch": 6.244272396903145,
+      "grad_norm": 1.1357136964797974,
+      "learning_rate": 6.18692996539714e-05,
+      "loss": 1.634,
+      "step": 29640
+    },
+    {
+      "epoch": 6.2463791014904935,
+      "grad_norm": 1.0918197631835938,
+      "learning_rate": 6.180811457545852e-05,
+      "loss": 1.5904,
+      "step": 29650
+    },
+    {
+      "epoch": 6.248485806077842,
+      "grad_norm": 1.162324070930481,
+      "learning_rate": 6.174694623154658e-05,
+      "loss": 1.6446,
+      "step": 29660
+    },
+    {
+      "epoch": 6.250592510665192,
+      "grad_norm": 1.066695213317871,
+      "learning_rate": 6.168579464903779e-05,
+      "loss": 1.592,
+      "step": 29670
+    },
+    {
+      "epoch": 6.252699215252541,
+      "grad_norm": 1.1125965118408203,
+      "learning_rate": 6.16246598547271e-05,
+      "loss": 1.5864,
+      "step": 29680
+    },
+    {
+      "epoch": 6.254805919839891,
+      "grad_norm": 1.1239444017410278,
+      "learning_rate": 6.1563541875402e-05,
+      "loss": 1.5655,
+      "step": 29690
+    },
+    {
+      "epoch": 6.2569126244272395,
+      "grad_norm": 1.1455018520355225,
+      "learning_rate": 6.150244073784266e-05,
+      "loss": 1.6228,
+      "step": 29700
+    },
+    {
+      "epoch": 6.259019329014589,
+      "grad_norm": 1.1525976657867432,
+      "learning_rate": 6.144135646882188e-05,
+      "loss": 1.6001,
+      "step": 29710
+    },
+    {
+      "epoch": 6.261126033601938,
+      "grad_norm": 1.305464506149292,
+      "learning_rate": 6.138028909510503e-05,
+      "loss": 1.5917,
+      "step": 29720
+    },
+    {
+      "epoch": 6.263232738189288,
+      "grad_norm": 1.2044486999511719,
+      "learning_rate": 6.131923864345012e-05,
+      "loss": 1.6158,
+      "step": 29730
+    },
+    {
+      "epoch": 6.265339442776637,
+      "grad_norm": 1.2145870923995972,
+      "learning_rate": 6.125820514060772e-05,
+      "loss": 1.6551,
+      "step": 29740
+    },
+    {
+      "epoch": 6.2674461473639855,
+      "grad_norm": 1.1999444961547852,
+      "learning_rate": 6.119718861332098e-05,
+      "loss": 1.6351,
+      "step": 29750
+    },
+    {
+      "epoch": 6.269552851951335,
+      "grad_norm": 1.1321429014205933,
+      "learning_rate": 6.113618908832561e-05,
+      "loss": 1.6331,
+      "step": 29760
+    },
+    {
+      "epoch": 6.271659556538684,
+      "grad_norm": 1.1454631090164185,
+      "learning_rate": 6.107520659234988e-05,
+      "loss": 1.5784,
+      "step": 29770
+    },
+    {
+      "epoch": 6.273766261126034,
+      "grad_norm": 1.2279682159423828,
+      "learning_rate": 6.101424115211458e-05,
+      "loss": 1.5881,
+      "step": 29780
+    },
+    {
+      "epoch": 6.275872965713383,
+      "grad_norm": 1.215173363685608,
+      "learning_rate": 6.095329279433304e-05,
+      "loss": 1.5792,
+      "step": 29790
+    },
+    {
+      "epoch": 6.2779796703007325,
+      "grad_norm": 1.0732548236846924,
+      "learning_rate": 6.089236154571109e-05,
+      "loss": 1.6326,
+      "step": 29800
+    },
+    {
+      "epoch": 6.280086374888081,
+      "grad_norm": 1.3970751762390137,
+      "learning_rate": 6.083144743294713e-05,
+      "loss": 1.5841,
+      "step": 29810
+    },
+    {
+      "epoch": 6.28219307947543,
+      "grad_norm": 1.1148728132247925,
+      "learning_rate": 6.0770550482731924e-05,
+      "loss": 1.6009,
+      "step": 29820
+    },
+    {
+      "epoch": 6.28429978406278,
+      "grad_norm": 1.209598422050476,
+      "learning_rate": 6.070967072174881e-05,
+      "loss": 1.5648,
+      "step": 29830
+    },
+    {
+      "epoch": 6.286406488650129,
+      "grad_norm": 1.1604822874069214,
+      "learning_rate": 6.0648808176673586e-05,
+      "loss": 1.6543,
+      "step": 29840
+    },
+    {
+      "epoch": 6.2885131932374785,
+      "grad_norm": 1.1224958896636963,
+      "learning_rate": 6.058796287417451e-05,
+      "loss": 1.622,
+      "step": 29850
+    },
+    {
+      "epoch": 6.290619897824827,
+      "grad_norm": 1.227657675743103,
+      "learning_rate": 6.0527134840912224e-05,
+      "loss": 1.633,
+      "step": 29860
+    },
+    {
+      "epoch": 6.292726602412177,
+      "grad_norm": 1.1823822259902954,
+      "learning_rate": 6.046632410353987e-05,
+      "loss": 1.5963,
+      "step": 29870
+    },
+    {
+      "epoch": 6.294833306999526,
+      "grad_norm": 1.0884348154067993,
+      "learning_rate": 6.0405530688702986e-05,
+      "loss": 1.5579,
+      "step": 29880
+    },
+    {
+      "epoch": 6.296940011586875,
+      "grad_norm": 1.2381103038787842,
+      "learning_rate": 6.034475462303952e-05,
+      "loss": 1.5499,
+      "step": 29890
+    },
+    {
+      "epoch": 6.2990467161742245,
+      "grad_norm": 1.1996432542800903,
+      "learning_rate": 6.028399593317984e-05,
+      "loss": 1.6038,
+      "step": 29900
+    },
+    {
+      "epoch": 6.301153420761573,
+      "grad_norm": 1.2356693744659424,
+      "learning_rate": 6.022325464574665e-05,
+      "loss": 1.5827,
+      "step": 29910
+    },
+    {
+      "epoch": 6.303260125348923,
+      "grad_norm": 1.1580696105957031,
+      "learning_rate": 6.016253078735508e-05,
+      "loss": 1.5349,
+      "step": 29920
+    },
+    {
+      "epoch": 6.305366829936272,
+      "grad_norm": 1.0700081586837769,
+      "learning_rate": 6.010182438461258e-05,
+      "loss": 1.5857,
+      "step": 29930
+    },
+    {
+      "epoch": 6.307473534523622,
+      "grad_norm": 1.2955549955368042,
+      "learning_rate": 6.0041135464119024e-05,
+      "loss": 1.578,
+      "step": 29940
+    },
+    {
+      "epoch": 6.3095802391109705,
+      "grad_norm": 1.116965889930725,
+      "learning_rate": 5.998046405246651e-05,
+      "loss": 1.5308,
+      "step": 29950
+    },
+    {
+      "epoch": 6.31168694369832,
+      "grad_norm": 1.232913851737976,
+      "learning_rate": 5.991981017623955e-05,
+      "loss": 1.5701,
+      "step": 29960
+    },
+    {
+      "epoch": 6.313793648285669,
+      "grad_norm": 1.167012333869934,
+      "learning_rate": 5.9859173862014985e-05,
+      "loss": 1.645,
+      "step": 29970
+    },
+    {
+      "epoch": 6.315900352873018,
+      "grad_norm": 1.1298688650131226,
+      "learning_rate": 5.979855513636192e-05,
+      "loss": 1.6271,
+      "step": 29980
+    },
+    {
+      "epoch": 6.318007057460368,
+      "grad_norm": 1.1685657501220703,
+      "learning_rate": 5.9737954025841725e-05,
+      "loss": 1.5824,
+      "step": 29990
+    },
+    {
+      "epoch": 6.320113762047717,
+      "grad_norm": 1.1890677213668823,
+      "learning_rate": 5.9677370557008104e-05,
+      "loss": 1.6214,
+      "step": 30000
+    },
+    {
+      "epoch": 6.322220466635066,
+      "grad_norm": 1.2406339645385742,
+      "learning_rate": 5.961680475640703e-05,
+      "loss": 1.5818,
+      "step": 30010
+    },
+    {
+      "epoch": 6.324327171222415,
+      "grad_norm": 1.1883368492126465,
+      "learning_rate": 5.955625665057672e-05,
+      "loss": 1.6185,
+      "step": 30020
+    },
+    {
+      "epoch": 6.326433875809765,
+      "grad_norm": 1.1744321584701538,
+      "learning_rate": 5.9495726266047605e-05,
+      "loss": 1.588,
+      "step": 30030
+    },
+    {
+      "epoch": 6.328540580397114,
+      "grad_norm": 1.2040607929229736,
+      "learning_rate": 5.9435213629342416e-05,
+      "loss": 1.5988,
+      "step": 30040
+    },
+    {
+      "epoch": 6.3306472849844635,
+      "grad_norm": 1.179335117340088,
+      "learning_rate": 5.9374718766976043e-05,
+      "loss": 1.6026,
+      "step": 30050
+    },
+    {
+      "epoch": 6.332753989571812,
+      "grad_norm": 1.173584222793579,
+      "learning_rate": 5.9314241705455674e-05,
+      "loss": 1.5836,
+      "step": 30060
+    },
+    {
+      "epoch": 6.334860694159161,
+      "grad_norm": 1.1390973329544067,
+      "learning_rate": 5.9253782471280596e-05,
+      "loss": 1.6209,
+      "step": 30070
+    },
+    {
+      "epoch": 6.336967398746511,
+      "grad_norm": 1.0759764909744263,
+      "learning_rate": 5.919334109094232e-05,
+      "loss": 1.6303,
+      "step": 30080
+    },
+    {
+      "epoch": 6.33907410333386,
+      "grad_norm": 1.2041103839874268,
+      "learning_rate": 5.9132917590924564e-05,
+      "loss": 1.6344,
+      "step": 30090
+    },
+    {
+      "epoch": 6.3411808079212095,
+      "grad_norm": 1.1497009992599487,
+      "learning_rate": 5.9072511997703226e-05,
+      "loss": 1.5925,
+      "step": 30100
+    },
+    {
+      "epoch": 6.343287512508558,
+      "grad_norm": 1.1539160013198853,
+      "learning_rate": 5.901212433774625e-05,
+      "loss": 1.5589,
+      "step": 30110
+    },
+    {
+      "epoch": 6.345394217095908,
+      "grad_norm": 1.163013219833374,
+      "learning_rate": 5.895175463751385e-05,
+      "loss": 1.5567,
+      "step": 30120
+    },
+    {
+      "epoch": 6.347500921683257,
+      "grad_norm": 1.2151119709014893,
+      "learning_rate": 5.889140292345831e-05,
+      "loss": 1.5016,
+      "step": 30130
+    },
+    {
+      "epoch": 6.349607626270606,
+      "grad_norm": 1.0841346979141235,
+      "learning_rate": 5.883106922202405e-05,
+      "loss": 1.5996,
+      "step": 30140
+    },
+    {
+      "epoch": 6.351714330857956,
+      "grad_norm": 1.169562578201294,
+      "learning_rate": 5.877075355964754e-05,
+      "loss": 1.5329,
+      "step": 30150
+    },
+    {
+      "epoch": 6.353821035445304,
+      "grad_norm": 1.2437294721603394,
+      "learning_rate": 5.871045596275742e-05,
+      "loss": 1.5844,
+      "step": 30160
+    },
+    {
+      "epoch": 6.355927740032654,
+      "grad_norm": 1.1044965982437134,
+      "learning_rate": 5.86501764577744e-05,
+      "loss": 1.586,
+      "step": 30170
+    },
+    {
+      "epoch": 6.358034444620003,
+      "grad_norm": 1.2414278984069824,
+      "learning_rate": 5.858991507111122e-05,
+      "loss": 1.6109,
+      "step": 30180
+    },
+    {
+      "epoch": 6.360141149207353,
+      "grad_norm": 1.134537935256958,
+      "learning_rate": 5.852967182917276e-05,
+      "loss": 1.5789,
+      "step": 30190
+    },
+    {
+      "epoch": 6.362247853794702,
+      "grad_norm": 1.0381525754928589,
+      "learning_rate": 5.846944675835584e-05,
+      "loss": 1.596,
+      "step": 30200
+    },
+    {
+      "epoch": 6.36435455838205,
+      "grad_norm": 1.9543304443359375,
+      "learning_rate": 5.84092398850494e-05,
+      "loss": 1.6616,
+      "step": 30210
+    },
+    {
+      "epoch": 6.3664612629694,
+      "grad_norm": 1.0826011896133423,
+      "learning_rate": 5.834905123563441e-05,
+      "loss": 1.5841,
+      "step": 30220
+    },
+    {
+      "epoch": 6.368567967556749,
+      "grad_norm": 1.1161965131759644,
+      "learning_rate": 5.828888083648382e-05,
+      "loss": 1.6423,
+      "step": 30230
+    },
+    {
+      "epoch": 6.370674672144099,
+      "grad_norm": 1.0613067150115967,
+      "learning_rate": 5.8228728713962543e-05,
+      "loss": 1.5799,
+      "step": 30240
+    },
+    {
+      "epoch": 6.372781376731448,
+      "grad_norm": 1.1326632499694824,
+      "learning_rate": 5.81685948944276e-05,
+      "loss": 1.5887,
+      "step": 30250
+    },
+    {
+      "epoch": 6.374888081318797,
+      "grad_norm": 1.1356877088546753,
+      "learning_rate": 5.8108479404227857e-05,
+      "loss": 1.5807,
+      "step": 30260
+    },
+    {
+      "epoch": 6.376994785906146,
+      "grad_norm": 1.159637212753296,
+      "learning_rate": 5.8048382269704305e-05,
+      "loss": 1.6054,
+      "step": 30270
+    },
+    {
+      "epoch": 6.379101490493496,
+      "grad_norm": 1.1319040060043335,
+      "learning_rate": 5.798830351718975e-05,
+      "loss": 1.5625,
+      "step": 30280
+    },
+    {
+      "epoch": 6.381208195080845,
+      "grad_norm": 1.3133894205093384,
+      "learning_rate": 5.7928243173008956e-05,
+      "loss": 1.5916,
+      "step": 30290
+    },
+    {
+      "epoch": 6.383314899668194,
+      "grad_norm": 1.1023930311203003,
+      "learning_rate": 5.786820126347876e-05,
+      "loss": 1.5471,
+      "step": 30300
+    },
+    {
+      "epoch": 6.385421604255543,
+      "grad_norm": 1.0441666841506958,
+      "learning_rate": 5.780817781490777e-05,
+      "loss": 1.5525,
+      "step": 30310
+    },
+    {
+      "epoch": 6.387528308842892,
+      "grad_norm": 1.1849002838134766,
+      "learning_rate": 5.7748172853596504e-05,
+      "loss": 1.6018,
+      "step": 30320
+    },
+    {
+      "epoch": 6.389635013430242,
+      "grad_norm": 1.1467089653015137,
+      "learning_rate": 5.768818640583755e-05,
+      "loss": 1.5891,
+      "step": 30330
+    },
+    {
+      "epoch": 6.391741718017591,
+      "grad_norm": 1.1927987337112427,
+      "learning_rate": 5.762821849791515e-05,
+      "loss": 1.6288,
+      "step": 30340
+    },
+    {
+      "epoch": 6.393848422604941,
+      "grad_norm": 1.230624794960022,
+      "learning_rate": 5.7568269156105656e-05,
+      "loss": 1.5764,
+      "step": 30350
+    },
+    {
+      "epoch": 6.395955127192289,
+      "grad_norm": 1.03030264377594,
+      "learning_rate": 5.750833840667711e-05,
+      "loss": 1.6052,
+      "step": 30360
+    },
+    {
+      "epoch": 6.398061831779639,
+      "grad_norm": 1.1722559928894043,
+      "learning_rate": 5.744842627588942e-05,
+      "loss": 1.5977,
+      "step": 30370
+    },
+    {
+      "epoch": 6.400168536366988,
+      "grad_norm": 1.084792137145996,
+      "learning_rate": 5.7388532789994476e-05,
+      "loss": 1.5983,
+      "step": 30380
+    },
+    {
+      "epoch": 6.402275240954337,
+      "grad_norm": 1.12361478805542,
+      "learning_rate": 5.7328657975235864e-05,
+      "loss": 1.5713,
+      "step": 30390
+    },
+    {
+      "epoch": 6.404381945541687,
+      "grad_norm": 1.1321178674697876,
+      "learning_rate": 5.7268801857848974e-05,
+      "loss": 1.6019,
+      "step": 30400
+    },
+    {
+      "epoch": 6.4064886501290355,
+      "grad_norm": 1.0965027809143066,
+      "learning_rate": 5.7208964464061165e-05,
+      "loss": 1.5976,
+      "step": 30410
+    },
+    {
+      "epoch": 6.408595354716385,
+      "grad_norm": 1.1020747423171997,
+      "learning_rate": 5.7149145820091385e-05,
+      "loss": 1.568,
+      "step": 30420
+    },
+    {
+      "epoch": 6.410702059303734,
+      "grad_norm": 1.088662028312683,
+      "learning_rate": 5.7089345952150555e-05,
+      "loss": 1.6088,
+      "step": 30430
+    },
+    {
+      "epoch": 6.412808763891084,
+      "grad_norm": 1.1830599308013916,
+      "learning_rate": 5.7029564886441245e-05,
+      "loss": 1.5691,
+      "step": 30440
+    },
+    {
+      "epoch": 6.414915468478433,
+      "grad_norm": 1.2777645587921143,
+      "learning_rate": 5.696980264915777e-05,
+      "loss": 1.6209,
+      "step": 30450
+    },
+    {
+      "epoch": 6.4170221730657815,
+      "grad_norm": 1.0497311353683472,
+      "learning_rate": 5.691005926648631e-05,
+      "loss": 1.5402,
+      "step": 30460
+    },
+    {
+      "epoch": 6.419128877653131,
+      "grad_norm": 1.1205629110336304,
+      "learning_rate": 5.685033476460471e-05,
+      "loss": 1.6328,
+      "step": 30470
+    },
+    {
+      "epoch": 6.42123558224048,
+      "grad_norm": 1.097438097000122,
+      "learning_rate": 5.6790629169682564e-05,
+      "loss": 1.586,
+      "step": 30480
+    },
+    {
+      "epoch": 6.42334228682783,
+      "grad_norm": 1.1904405355453491,
+      "learning_rate": 5.673094250788115e-05,
+      "loss": 1.5376,
+      "step": 30490
+    },
+    {
+      "epoch": 6.425448991415179,
+      "grad_norm": 1.2280725240707397,
+      "learning_rate": 5.6671274805353434e-05,
+      "loss": 1.601,
+      "step": 30500
+    },
+    {
+      "epoch": 6.427555696002528,
+      "grad_norm": 1.2686785459518433,
+      "learning_rate": 5.6611626088244194e-05,
+      "loss": 1.6357,
+      "step": 30510
+    },
+    {
+      "epoch": 6.429662400589877,
+      "grad_norm": 1.2746089696884155,
+      "learning_rate": 5.6551996382689776e-05,
+      "loss": 1.6072,
+      "step": 30520
+    },
+    {
+      "epoch": 6.431769105177226,
+      "grad_norm": 1.127000093460083,
+      "learning_rate": 5.649238571481815e-05,
+      "loss": 1.6258,
+      "step": 30530
+    },
+    {
+      "epoch": 6.433875809764576,
+      "grad_norm": 1.1913783550262451,
+      "learning_rate": 5.6432794110749134e-05,
+      "loss": 1.5714,
+      "step": 30540
+    },
+    {
+      "epoch": 6.435982514351925,
+      "grad_norm": 1.0561548471450806,
+      "learning_rate": 5.6373221596594e-05,
+      "loss": 1.528,
+      "step": 30550
+    },
+    {
+      "epoch": 6.438089218939274,
+      "grad_norm": 1.189866304397583,
+      "learning_rate": 5.631366819845578e-05,
+      "loss": 1.5789,
+      "step": 30560
+    },
+    {
+      "epoch": 6.440195923526623,
+      "grad_norm": 1.1098788976669312,
+      "learning_rate": 5.625413394242907e-05,
+      "loss": 1.5776,
+      "step": 30570
+    },
+    {
+      "epoch": 6.442302628113973,
+      "grad_norm": 1.271070122718811,
+      "learning_rate": 5.6194618854600057e-05,
+      "loss": 1.5436,
+      "step": 30580
+    },
+    {
+      "epoch": 6.444409332701322,
+      "grad_norm": 1.1668862104415894,
+      "learning_rate": 5.613512296104663e-05,
+      "loss": 1.6552,
+      "step": 30590
+    },
+    {
+      "epoch": 6.446516037288671,
+      "grad_norm": 1.2380750179290771,
+      "learning_rate": 5.607564628783817e-05,
+      "loss": 1.5845,
+      "step": 30600
+    },
+    {
+      "epoch": 6.4486227418760205,
+      "grad_norm": 1.224056601524353,
+      "learning_rate": 5.601618886103561e-05,
+      "loss": 1.6456,
+      "step": 30610
+    },
+    {
+      "epoch": 6.450729446463369,
+      "grad_norm": 1.0602716207504272,
+      "learning_rate": 5.595675070669162e-05,
+      "loss": 1.5599,
+      "step": 30620
+    },
+    {
+      "epoch": 6.452836151050719,
+      "grad_norm": 1.1630531549453735,
+      "learning_rate": 5.589733185085022e-05,
+      "loss": 1.5894,
+      "step": 30630
+    },
+    {
+      "epoch": 6.454942855638068,
+      "grad_norm": 1.0980303287506104,
+      "learning_rate": 5.583793231954713e-05,
+      "loss": 1.6006,
+      "step": 30640
+    },
+    {
+      "epoch": 6.457049560225418,
+      "grad_norm": 1.1695753335952759,
+      "learning_rate": 5.577855213880951e-05,
+      "loss": 1.598,
+      "step": 30650
+    },
+    {
+      "epoch": 6.4591562648127665,
+      "grad_norm": 1.0704690217971802,
+      "learning_rate": 5.571919133465605e-05,
+      "loss": 1.6008,
+      "step": 30660
+    },
+    {
+      "epoch": 6.461262969400116,
+      "grad_norm": 1.102163314819336,
+      "learning_rate": 5.565984993309703e-05,
+      "loss": 1.5416,
+      "step": 30670
+    },
+    {
+      "epoch": 6.463369673987465,
+      "grad_norm": 1.143849492073059,
+      "learning_rate": 5.560052796013413e-05,
+      "loss": 1.5771,
+      "step": 30680
+    },
+    {
+      "epoch": 6.465476378574814,
+      "grad_norm": 1.03353750705719,
+      "learning_rate": 5.5541225441760524e-05,
+      "loss": 1.6104,
+      "step": 30690
+    },
+    {
+      "epoch": 6.467583083162164,
+      "grad_norm": 1.1685138940811157,
+      "learning_rate": 5.5481942403960986e-05,
+      "loss": 1.6455,
+      "step": 30700
+    },
+    {
+      "epoch": 6.4696897877495125,
+      "grad_norm": 1.1921008825302124,
+      "learning_rate": 5.542267887271155e-05,
+      "loss": 1.6081,
+      "step": 30710
+    },
+    {
+      "epoch": 6.471796492336862,
+      "grad_norm": 1.1776456832885742,
+      "learning_rate": 5.5363434873979903e-05,
+      "loss": 1.6246,
+      "step": 30720
+    },
+    {
+      "epoch": 6.473903196924211,
+      "grad_norm": 1.1411466598510742,
+      "learning_rate": 5.530421043372507e-05,
+      "loss": 1.6324,
+      "step": 30730
+    },
+    {
+      "epoch": 6.476009901511561,
+      "grad_norm": 1.1445951461791992,
+      "learning_rate": 5.524500557789745e-05,
+      "loss": 1.6065,
+      "step": 30740
+    },
+    {
+      "epoch": 6.47811660609891,
+      "grad_norm": 1.1429086923599243,
+      "learning_rate": 5.518582033243902e-05,
+      "loss": 1.5875,
+      "step": 30750
+    },
+    {
+      "epoch": 6.4802233106862595,
+      "grad_norm": 1.1539630889892578,
+      "learning_rate": 5.512665472328302e-05,
+      "loss": 1.6545,
+      "step": 30760
+    },
+    {
+      "epoch": 6.482330015273608,
+      "grad_norm": 1.1232472658157349,
+      "learning_rate": 5.506750877635418e-05,
+      "loss": 1.5867,
+      "step": 30770
+    },
+    {
+      "epoch": 6.484436719860957,
+      "grad_norm": 1.2769269943237305,
+      "learning_rate": 5.500838251756857e-05,
+      "loss": 1.6165,
+      "step": 30780
+    },
+    {
+      "epoch": 6.486543424448307,
+      "grad_norm": 1.1701322793960571,
+      "learning_rate": 5.4949275972833594e-05,
+      "loss": 1.5832,
+      "step": 30790
+    },
+    {
+      "epoch": 6.488650129035656,
+      "grad_norm": 1.1360645294189453,
+      "learning_rate": 5.489018916804813e-05,
+      "loss": 1.6098,
+      "step": 30800
+    },
+    {
+      "epoch": 6.4907568336230055,
+      "grad_norm": 1.127044677734375,
+      "learning_rate": 5.4831122129102307e-05,
+      "loss": 1.6443,
+      "step": 30810
+    },
+    {
+      "epoch": 6.492863538210354,
+      "grad_norm": 1.1451194286346436,
+      "learning_rate": 5.477207488187759e-05,
+      "loss": 1.6208,
+      "step": 30820
+    },
+    {
+      "epoch": 6.494970242797704,
+      "grad_norm": 1.1845287084579468,
+      "learning_rate": 5.471304745224689e-05,
+      "loss": 1.6055,
+      "step": 30830
+    },
+    {
+      "epoch": 6.497076947385053,
+      "grad_norm": 1.208879828453064,
+      "learning_rate": 5.465403986607426e-05,
+      "loss": 1.6061,
+      "step": 30840
+    },
+    {
+      "epoch": 6.499183651972402,
+      "grad_norm": 1.2800372838974,
+      "learning_rate": 5.4595052149215246e-05,
+      "loss": 1.5853,
+      "step": 30850
+    },
+    {
+      "epoch": 6.5012903565597515,
+      "grad_norm": 1.1612744331359863,
+      "learning_rate": 5.4536084327516535e-05,
+      "loss": 1.6532,
+      "step": 30860
+    },
+    {
+      "epoch": 6.5033970611471,
+      "grad_norm": 1.0994377136230469,
+      "learning_rate": 5.447713642681612e-05,
+      "loss": 1.6062,
+      "step": 30870
+    },
+    {
+      "epoch": 6.50550376573445,
+      "grad_norm": 1.1329749822616577,
+      "learning_rate": 5.441820847294339e-05,
+      "loss": 1.5962,
+      "step": 30880
+    },
+    {
+      "epoch": 6.507610470321799,
+      "grad_norm": 1.0713906288146973,
+      "learning_rate": 5.435930049171885e-05,
+      "loss": 1.5756,
+      "step": 30890
+    },
+    {
+      "epoch": 6.509717174909149,
+      "grad_norm": 1.1019048690795898,
+      "learning_rate": 5.430041250895428e-05,
+      "loss": 1.5566,
+      "step": 30900
+    },
+    {
+      "epoch": 6.5118238794964975,
+      "grad_norm": 1.2118427753448486,
+      "learning_rate": 5.424154455045278e-05,
+      "loss": 1.5868,
+      "step": 30910
+    },
+    {
+      "epoch": 6.513930584083846,
+      "grad_norm": 1.2621134519577026,
+      "learning_rate": 5.418269664200857e-05,
+      "loss": 1.5591,
+      "step": 30920
+    },
+    {
+      "epoch": 6.516037288671196,
+      "grad_norm": 1.1320388317108154,
+      "learning_rate": 5.4123868809407206e-05,
+      "loss": 1.6229,
+      "step": 30930
+    },
+    {
+      "epoch": 6.518143993258545,
+      "grad_norm": 1.1174076795578003,
+      "learning_rate": 5.4065061078425315e-05,
+      "loss": 1.6151,
+      "step": 30940
+    },
+    {
+      "epoch": 6.520250697845895,
+      "grad_norm": 1.1697906255722046,
+      "learning_rate": 5.400627347483076e-05,
+      "loss": 1.6131,
+      "step": 30950
+    },
+    {
+      "epoch": 6.522357402433244,
+      "grad_norm": 1.0805909633636475,
+      "learning_rate": 5.3947506024382665e-05,
+      "loss": 1.5527,
+      "step": 30960
+    },
+    {
+      "epoch": 6.524464107020593,
+      "grad_norm": 1.077239990234375,
+      "learning_rate": 5.388875875283124e-05,
+      "loss": 1.6083,
+      "step": 30970
+    },
+    {
+      "epoch": 6.526570811607942,
+      "grad_norm": 1.1366991996765137,
+      "learning_rate": 5.3830031685917803e-05,
+      "loss": 1.5768,
+      "step": 30980
+    },
+    {
+      "epoch": 6.528677516195291,
+      "grad_norm": 1.2108395099639893,
+      "learning_rate": 5.377132484937499e-05,
+      "loss": 1.6281,
+      "step": 30990
+    },
+    {
+      "epoch": 6.530784220782641,
+      "grad_norm": 1.1738390922546387,
+      "learning_rate": 5.3712638268926397e-05,
+      "loss": 1.6515,
+      "step": 31000
+    },
+    {
+      "epoch": 6.53289092536999,
+      "grad_norm": 1.0865222215652466,
+      "learning_rate": 5.365397197028685e-05,
+      "loss": 1.5456,
+      "step": 31010
+    },
+    {
+      "epoch": 6.534997629957339,
+      "grad_norm": 1.1172924041748047,
+      "learning_rate": 5.359532597916233e-05,
+      "loss": 1.5758,
+      "step": 31020
+    },
+    {
+      "epoch": 6.537104334544688,
+      "grad_norm": 1.1861040592193604,
+      "learning_rate": 5.35367003212497e-05,
+      "loss": 1.5852,
+      "step": 31030
+    },
+    {
+      "epoch": 6.539211039132038,
+      "grad_norm": 1.1553261280059814,
+      "learning_rate": 5.3478095022237175e-05,
+      "loss": 1.5824,
+      "step": 31040
+    },
+    {
+      "epoch": 6.541317743719387,
+      "grad_norm": 1.1746402978897095,
+      "learning_rate": 5.341951010780386e-05,
+      "loss": 1.5837,
+      "step": 31050
+    },
+    {
+      "epoch": 6.5434244483067365,
+      "grad_norm": 1.233399748802185,
+      "learning_rate": 5.336094560362006e-05,
+      "loss": 1.5422,
+      "step": 31060
+    },
+    {
+      "epoch": 6.545531152894085,
+      "grad_norm": 1.1761796474456787,
+      "learning_rate": 5.330240153534707e-05,
+      "loss": 1.5942,
+      "step": 31070
+    },
+    {
+      "epoch": 6.547637857481435,
+      "grad_norm": 1.077040195465088,
+      "learning_rate": 5.324387792863719e-05,
+      "loss": 1.5745,
+      "step": 31080
+    },
+    {
+      "epoch": 6.549744562068784,
+      "grad_norm": 1.251773476600647,
+      "learning_rate": 5.3185374809133837e-05,
+      "loss": 1.6185,
+      "step": 31090
+    },
+    {
+      "epoch": 6.551851266656133,
+      "grad_norm": 1.1122722625732422,
+      "learning_rate": 5.312689220247151e-05,
+      "loss": 1.5961,
+      "step": 31100
+    },
+    {
+      "epoch": 6.553957971243483,
+      "grad_norm": 1.1091094017028809,
+      "learning_rate": 5.306843013427545e-05,
+      "loss": 1.6158,
+      "step": 31110
+    },
+    {
+      "epoch": 6.556064675830831,
+      "grad_norm": 1.1521689891815186,
+      "learning_rate": 5.300998863016222e-05,
+      "loss": 1.6193,
+      "step": 31120
+    },
+    {
+      "epoch": 6.558171380418181,
+      "grad_norm": 1.0843865871429443,
+      "learning_rate": 5.2951567715739126e-05,
+      "loss": 1.5983,
+      "step": 31130
+    },
+    {
+      "epoch": 6.56027808500553,
+      "grad_norm": 1.2138440608978271,
+      "learning_rate": 5.289316741660466e-05,
+      "loss": 1.5899,
+      "step": 31140
+    },
+    {
+      "epoch": 6.56238478959288,
+      "grad_norm": 1.0254400968551636,
+      "learning_rate": 5.283478775834811e-05,
+      "loss": 1.5816,
+      "step": 31150
+    },
+    {
+      "epoch": 6.564491494180229,
+      "grad_norm": 1.0957165956497192,
+      "learning_rate": 5.277642876654978e-05,
+      "loss": 1.6152,
+      "step": 31160
+    },
+    {
+      "epoch": 6.566598198767577,
+      "grad_norm": 1.1528323888778687,
+      "learning_rate": 5.271809046678094e-05,
+      "loss": 1.6044,
+      "step": 31170
+    },
+    {
+      "epoch": 6.568704903354927,
+      "grad_norm": 1.1122441291809082,
+      "learning_rate": 5.265977288460386e-05,
+      "loss": 1.5604,
+      "step": 31180
+    },
+    {
+      "epoch": 6.570811607942276,
+      "grad_norm": 1.1756147146224976,
+      "learning_rate": 5.2601476045571506e-05,
+      "loss": 1.6111,
+      "step": 31190
+    },
+    {
+      "epoch": 6.572918312529626,
+      "grad_norm": 1.2942123413085938,
+      "learning_rate": 5.254319997522796e-05,
+      "loss": 1.61,
+      "step": 31200
+    },
+    {
+      "epoch": 6.575025017116975,
+      "grad_norm": 1.1158119440078735,
+      "learning_rate": 5.2484944699108194e-05,
+      "loss": 1.6006,
+      "step": 31210
+    },
+    {
+      "epoch": 6.577131721704324,
+      "grad_norm": 1.1101962327957153,
+      "learning_rate": 5.242671024273798e-05,
+      "loss": 1.6298,
+      "step": 31220
+    },
+    {
+      "epoch": 6.579238426291673,
+      "grad_norm": 1.0920320749282837,
+      "learning_rate": 5.236849663163399e-05,
+      "loss": 1.6053,
+      "step": 31230
+    },
+    {
+      "epoch": 6.581345130879022,
+      "grad_norm": 1.1226073503494263,
+      "learning_rate": 5.231030389130375e-05,
+      "loss": 1.6229,
+      "step": 31240
+    },
+    {
+      "epoch": 6.583451835466372,
+      "grad_norm": 1.1309608221054077,
+      "learning_rate": 5.2252132047245704e-05,
+      "loss": 1.5947,
+      "step": 31250
+    },
+    {
+      "epoch": 6.585558540053721,
+      "grad_norm": 1.190983533859253,
+      "learning_rate": 5.21939811249492e-05,
+      "loss": 1.6526,
+      "step": 31260
+    },
+    {
+      "epoch": 6.58766524464107,
+      "grad_norm": 1.2079119682312012,
+      "learning_rate": 5.2135851149894124e-05,
+      "loss": 1.6248,
+      "step": 31270
+    },
+    {
+      "epoch": 6.589771949228419,
+      "grad_norm": 1.0407904386520386,
+      "learning_rate": 5.20777421475515e-05,
+      "loss": 1.5605,
+      "step": 31280
+    },
+    {
+      "epoch": 6.591878653815769,
+      "grad_norm": 1.154831886291504,
+      "learning_rate": 5.201965414338308e-05,
+      "loss": 1.6149,
+      "step": 31290
+    },
+    {
+      "epoch": 6.593985358403118,
+      "grad_norm": 1.1011749505996704,
+      "learning_rate": 5.196158716284128e-05,
+      "loss": 1.5944,
+      "step": 31300
+    },
+    {
+      "epoch": 6.596092062990467,
+      "grad_norm": 1.1772890090942383,
+      "learning_rate": 5.190354123136954e-05,
+      "loss": 1.604,
+      "step": 31310
+    },
+    {
+      "epoch": 6.598198767577816,
+      "grad_norm": 1.1958460807800293,
+      "learning_rate": 5.1845516374401784e-05,
+      "loss": 1.5671,
+      "step": 31320
+    },
+    {
+      "epoch": 6.600305472165165,
+      "grad_norm": 1.1239957809448242,
+      "learning_rate": 5.178751261736292e-05,
+      "loss": 1.5596,
+      "step": 31330
+    },
+    {
+      "epoch": 6.602412176752515,
+      "grad_norm": 1.0745184421539307,
+      "learning_rate": 5.1729529985668604e-05,
+      "loss": 1.5746,
+      "step": 31340
+    },
+    {
+      "epoch": 6.604518881339864,
+      "grad_norm": 1.1670557260513306,
+      "learning_rate": 5.1671568504725135e-05,
+      "loss": 1.5606,
+      "step": 31350
+    },
+    {
+      "epoch": 6.606625585927214,
+      "grad_norm": 1.1216731071472168,
+      "learning_rate": 5.1613628199929544e-05,
+      "loss": 1.6375,
+      "step": 31360
+    },
+    {
+      "epoch": 6.6087322905145625,
+      "grad_norm": 1.1803592443466187,
+      "learning_rate": 5.1555709096669725e-05,
+      "loss": 1.6,
+      "step": 31370
+    },
+    {
+      "epoch": 6.610838995101912,
+      "grad_norm": 1.1835579872131348,
+      "learning_rate": 5.14978112203241e-05,
+      "loss": 1.6252,
+      "step": 31380
+    },
+    {
+      "epoch": 6.612945699689261,
+      "grad_norm": 1.231615424156189,
+      "learning_rate": 5.1439934596261994e-05,
+      "loss": 1.6111,
+      "step": 31390
+    },
+    {
+      "epoch": 6.615052404276611,
+      "grad_norm": 1.2632482051849365,
+      "learning_rate": 5.138207924984313e-05,
+      "loss": 1.6047,
+      "step": 31400
+    },
+    {
+      "epoch": 6.61715910886396,
+      "grad_norm": 1.2507926225662231,
+      "learning_rate": 5.1324245206418184e-05,
+      "loss": 1.646,
+      "step": 31410
+    },
+    {
+      "epoch": 6.6192658134513085,
+      "grad_norm": 1.119822382926941,
+      "learning_rate": 5.126643249132843e-05,
+      "loss": 1.5353,
+      "step": 31420
+    },
+    {
+      "epoch": 6.621372518038658,
+      "grad_norm": 1.2157727479934692,
+      "learning_rate": 5.120864112990569e-05,
+      "loss": 1.6133,
+      "step": 31430
+    },
+    {
+      "epoch": 6.623479222626007,
+      "grad_norm": 1.1493059396743774,
+      "learning_rate": 5.11508711474725e-05,
+      "loss": 1.6899,
+      "step": 31440
+    },
+    {
+      "epoch": 6.625585927213357,
+      "grad_norm": 1.209083080291748,
+      "learning_rate": 5.109312256934208e-05,
+      "loss": 1.5788,
+      "step": 31450
+    },
+    {
+      "epoch": 6.627692631800706,
+      "grad_norm": 1.1497858762741089,
+      "learning_rate": 5.103539542081814e-05,
+      "loss": 1.5901,
+      "step": 31460
+    },
+    {
+      "epoch": 6.629799336388055,
+      "grad_norm": 1.1102197170257568,
+      "learning_rate": 5.097768972719522e-05,
+      "loss": 1.6064,
+      "step": 31470
+    },
+    {
+      "epoch": 6.631906040975404,
+      "grad_norm": 1.158963918685913,
+      "learning_rate": 5.092000551375814e-05,
+      "loss": 1.5992,
+      "step": 31480
+    },
+    {
+      "epoch": 6.634012745562753,
+      "grad_norm": 1.145154356956482,
+      "learning_rate": 5.086234280578257e-05,
+      "loss": 1.5399,
+      "step": 31490
+    },
+    {
+      "epoch": 6.636119450150103,
+      "grad_norm": 1.1013636589050293,
+      "learning_rate": 5.080470162853472e-05,
+      "loss": 1.6289,
+      "step": 31500
+    },
+    {
+      "epoch": 6.638226154737452,
+      "grad_norm": 1.1045435667037964,
+      "learning_rate": 5.0747082007271275e-05,
+      "loss": 1.5798,
+      "step": 31510
+    },
+    {
+      "epoch": 6.640332859324801,
+      "grad_norm": 1.1121597290039062,
+      "learning_rate": 5.068948396723947e-05,
+      "loss": 1.5345,
+      "step": 31520
+    },
+    {
+      "epoch": 6.64243956391215,
+      "grad_norm": 1.1220855712890625,
+      "learning_rate": 5.063190753367721e-05,
+      "loss": 1.6155,
+      "step": 31530
+    },
+    {
+      "epoch": 6.6445462684995,
+      "grad_norm": 1.2580904960632324,
+      "learning_rate": 5.0574352731812814e-05,
+      "loss": 1.5651,
+      "step": 31540
+    },
+    {
+      "epoch": 6.646652973086849,
+      "grad_norm": 1.054614782333374,
+      "learning_rate": 5.051681958686518e-05,
+      "loss": 1.6003,
+      "step": 31550
+    },
+    {
+      "epoch": 6.648759677674198,
+      "grad_norm": 1.0666719675064087,
+      "learning_rate": 5.0459308124043715e-05,
+      "loss": 1.5902,
+      "step": 31560
+    },
+    {
+      "epoch": 6.6508663822615475,
+      "grad_norm": 1.1629925966262817,
+      "learning_rate": 5.040181836854825e-05,
+      "loss": 1.659,
+      "step": 31570
+    },
+    {
+      "epoch": 6.652973086848896,
+      "grad_norm": 5.463768005371094,
+      "learning_rate": 5.0344350345569244e-05,
+      "loss": 1.5968,
+      "step": 31580
+    },
+    {
+      "epoch": 6.655079791436246,
+      "grad_norm": 1.0833951234817505,
+      "learning_rate": 5.028690408028748e-05,
+      "loss": 1.6454,
+      "step": 31590
+    },
+    {
+      "epoch": 6.657186496023595,
+      "grad_norm": 1.0549983978271484,
+      "learning_rate": 5.022947959787435e-05,
+      "loss": 1.6268,
+      "step": 31600
+    },
+    {
+      "epoch": 6.659293200610945,
+      "grad_norm": 1.0697112083435059,
+      "learning_rate": 5.0172076923491604e-05,
+      "loss": 1.546,
+      "step": 31610
+    },
+    {
+      "epoch": 6.6613999051982935,
+      "grad_norm": 1.2887377738952637,
+      "learning_rate": 5.0114696082291425e-05,
+      "loss": 1.5367,
+      "step": 31620
+    },
+    {
+      "epoch": 6.663506609785642,
+      "grad_norm": 1.2095539569854736,
+      "learning_rate": 5.0057337099416556e-05,
+      "loss": 1.6219,
+      "step": 31630
+    },
+    {
+      "epoch": 6.665613314372992,
+      "grad_norm": 1.0067800283432007,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 1.5961,
+      "step": 31640
+    },
+    {
+      "epoch": 6.667720018960341,
+      "grad_norm": 1.167069673538208,
+      "learning_rate": 4.9942684809165284e-05,
+      "loss": 1.5955,
+      "step": 31650
+    },
+    {
+      "epoch": 6.669826723547691,
+      "grad_norm": 1.0418312549591064,
+      "learning_rate": 4.9885391552026304e-05,
+      "loss": 1.5472,
+      "step": 31660
+    },
+    {
+      "epoch": 6.6719334281350395,
+      "grad_norm": 1.0489190816879272,
+      "learning_rate": 4.9828120253687296e-05,
+      "loss": 1.5671,
+      "step": 31670
+    },
+    {
+      "epoch": 6.674040132722389,
+      "grad_norm": 1.3005365133285522,
+      "learning_rate": 4.9770870939242986e-05,
+      "loss": 1.6017,
+      "step": 31680
+    },
+    {
+      "epoch": 6.676146837309738,
+      "grad_norm": 1.2087478637695312,
+      "learning_rate": 4.971364363377837e-05,
+      "loss": 1.5988,
+      "step": 31690
+    },
+    {
+      "epoch": 6.678253541897088,
+      "grad_norm": 1.241007924079895,
+      "learning_rate": 4.9656438362368784e-05,
+      "loss": 1.5843,
+      "step": 31700
+    },
+    {
+      "epoch": 6.680360246484437,
+      "grad_norm": 1.1099778413772583,
+      "learning_rate": 4.959925515008002e-05,
+      "loss": 1.656,
+      "step": 31710
+    },
+    {
+      "epoch": 6.6824669510717865,
+      "grad_norm": 1.1651355028152466,
+      "learning_rate": 4.954209402196813e-05,
+      "loss": 1.5808,
+      "step": 31720
+    },
+    {
+      "epoch": 6.684573655659135,
+      "grad_norm": 1.303068995475769,
+      "learning_rate": 4.948495500307945e-05,
+      "loss": 1.5864,
+      "step": 31730
+    },
+    {
+      "epoch": 6.686680360246484,
+      "grad_norm": 1.1450005769729614,
+      "learning_rate": 4.942783811845074e-05,
+      "loss": 1.6218,
+      "step": 31740
+    },
+    {
+      "epoch": 6.688787064833834,
+      "grad_norm": 1.2237937450408936,
+      "learning_rate": 4.937074339310894e-05,
+      "loss": 1.6114,
+      "step": 31750
+    },
+    {
+      "epoch": 6.690893769421183,
+      "grad_norm": 1.1926014423370361,
+      "learning_rate": 4.931367085207142e-05,
+      "loss": 1.6181,
+      "step": 31760
+    },
+    {
+      "epoch": 6.6930004740085325,
+      "grad_norm": 1.0968434810638428,
+      "learning_rate": 4.9256620520345675e-05,
+      "loss": 1.5742,
+      "step": 31770
+    },
+    {
+      "epoch": 6.695107178595881,
+      "grad_norm": 1.1037589311599731,
+      "learning_rate": 4.919959242292954e-05,
+      "loss": 1.5769,
+      "step": 31780
+    },
+    {
+      "epoch": 6.697213883183231,
+      "grad_norm": 1.150773286819458,
+      "learning_rate": 4.9142586584811165e-05,
+      "loss": 1.5985,
+      "step": 31790
+    },
+    {
+      "epoch": 6.69932058777058,
+      "grad_norm": 1.1506338119506836,
+      "learning_rate": 4.908560303096887e-05,
+      "loss": 1.5338,
+      "step": 31800
+    },
+    {
+      "epoch": 6.701427292357929,
+      "grad_norm": 1.2946168184280396,
+      "learning_rate": 4.90286417863712e-05,
+      "loss": 1.5999,
+      "step": 31810
+    },
+    {
+      "epoch": 6.7035339969452785,
+      "grad_norm": 1.1435717344284058,
+      "learning_rate": 4.8971702875977e-05,
+      "loss": 1.6167,
+      "step": 31820
+    },
+    {
+      "epoch": 6.705640701532627,
+      "grad_norm": 1.048068881034851,
+      "learning_rate": 4.891478632473524e-05,
+      "loss": 1.5388,
+      "step": 31830
+    },
+    {
+      "epoch": 6.707747406119977,
+      "grad_norm": 1.086499571800232,
+      "learning_rate": 4.88578921575852e-05,
+      "loss": 1.6239,
+      "step": 31840
+    },
+    {
+      "epoch": 6.709854110707326,
+      "grad_norm": 1.2164744138717651,
+      "learning_rate": 4.880102039945624e-05,
+      "loss": 1.646,
+      "step": 31850
+    },
+    {
+      "epoch": 6.711960815294676,
+      "grad_norm": 1.0961980819702148,
+      "learning_rate": 4.874417107526795e-05,
+      "loss": 1.5828,
+      "step": 31860
+    },
+    {
+      "epoch": 6.7140675198820245,
+      "grad_norm": 1.094759225845337,
+      "learning_rate": 4.868734420993014e-05,
+      "loss": 1.6037,
+      "step": 31870
+    },
+    {
+      "epoch": 6.716174224469373,
+      "grad_norm": 1.0369664430618286,
+      "learning_rate": 4.863053982834266e-05,
+      "loss": 1.5872,
+      "step": 31880
+    },
+    {
+      "epoch": 6.718280929056723,
+      "grad_norm": 1.1314085721969604,
+      "learning_rate": 4.857375795539566e-05,
+      "loss": 1.5673,
+      "step": 31890
+    },
+    {
+      "epoch": 6.720387633644072,
+      "grad_norm": 1.1117151975631714,
+      "learning_rate": 4.85169986159693e-05,
+      "loss": 1.6008,
+      "step": 31900
+    },
+    {
+      "epoch": 6.722494338231422,
+      "grad_norm": 1.1076699495315552,
+      "learning_rate": 4.8460261834933875e-05,
+      "loss": 1.6054,
+      "step": 31910
+    },
+    {
+      "epoch": 6.724601042818771,
+      "grad_norm": 1.090351939201355,
+      "learning_rate": 4.840354763714991e-05,
+      "loss": 1.5609,
+      "step": 31920
+    },
+    {
+      "epoch": 6.72670774740612,
+      "grad_norm": 1.0670236349105835,
+      "learning_rate": 4.834685604746794e-05,
+      "loss": 1.5798,
+      "step": 31930
+    },
+    {
+      "epoch": 6.728814451993469,
+      "grad_norm": 1.1509038209915161,
+      "learning_rate": 4.829018709072854e-05,
+      "loss": 1.645,
+      "step": 31940
+    },
+    {
+      "epoch": 6.730921156580818,
+      "grad_norm": 1.1492196321487427,
+      "learning_rate": 4.823354079176253e-05,
+      "loss": 1.5266,
+      "step": 31950
+    },
+    {
+      "epoch": 6.733027861168168,
+      "grad_norm": 1.3231983184814453,
+      "learning_rate": 4.8176917175390656e-05,
+      "loss": 1.5827,
+      "step": 31960
+    },
+    {
+      "epoch": 6.735134565755517,
+      "grad_norm": 1.2020155191421509,
+      "learning_rate": 4.812031626642382e-05,
+      "loss": 1.6198,
+      "step": 31970
+    },
+    {
+      "epoch": 6.737241270342866,
+      "grad_norm": 1.140032172203064,
+      "learning_rate": 4.8063738089662926e-05,
+      "loss": 1.5159,
+      "step": 31980
+    },
+    {
+      "epoch": 6.739347974930215,
+      "grad_norm": 1.1477890014648438,
+      "learning_rate": 4.800718266989888e-05,
+      "loss": 1.6024,
+      "step": 31990
+    },
+    {
+      "epoch": 6.741454679517565,
+      "grad_norm": 1.2770425081253052,
+      "learning_rate": 4.795065003191272e-05,
+      "loss": 1.5814,
+      "step": 32000
+    },
+    {
+      "epoch": 6.743561384104914,
+      "grad_norm": 1.1991952657699585,
+      "learning_rate": 4.7894140200475435e-05,
+      "loss": 1.5764,
+      "step": 32010
+    },
+    {
+      "epoch": 6.745668088692263,
+      "grad_norm": 1.2961416244506836,
+      "learning_rate": 4.7837653200347974e-05,
+      "loss": 1.5989,
+      "step": 32020
+    },
+    {
+      "epoch": 6.747774793279612,
+      "grad_norm": 1.1892178058624268,
+      "learning_rate": 4.7781189056281415e-05,
+      "loss": 1.6455,
+      "step": 32030
+    },
+    {
+      "epoch": 6.749881497866961,
+      "grad_norm": 1.1718902587890625,
+      "learning_rate": 4.772474779301669e-05,
+      "loss": 1.5826,
+      "step": 32040
+    },
+    {
+      "epoch": 6.751988202454311,
+      "grad_norm": 1.1832486391067505,
+      "learning_rate": 4.766832943528481e-05,
+      "loss": 1.6021,
+      "step": 32050
+    },
+    {
+      "epoch": 6.75409490704166,
+      "grad_norm": 1.1368720531463623,
+      "learning_rate": 4.7611934007806666e-05,
+      "loss": 1.5537,
+      "step": 32060
+    },
+    {
+      "epoch": 6.75620161162901,
+      "grad_norm": 1.0927391052246094,
+      "learning_rate": 4.755556153529311e-05,
+      "loss": 1.617,
+      "step": 32070
+    },
+    {
+      "epoch": 6.758308316216358,
+      "grad_norm": 1.3450697660446167,
+      "learning_rate": 4.749921204244503e-05,
+      "loss": 1.6012,
+      "step": 32080
+    },
+    {
+      "epoch": 6.760415020803708,
+      "grad_norm": 1.1041196584701538,
+      "learning_rate": 4.744288555395313e-05,
+      "loss": 1.6142,
+      "step": 32090
+    },
+    {
+      "epoch": 6.762521725391057,
+      "grad_norm": 1.079490065574646,
+      "learning_rate": 4.738658209449805e-05,
+      "loss": 1.6022,
+      "step": 32100
+    },
+    {
+      "epoch": 6.764628429978407,
+      "grad_norm": 1.116524338722229,
+      "learning_rate": 4.7330301688750434e-05,
+      "loss": 1.5029,
+      "step": 32110
+    },
+    {
+      "epoch": 6.766735134565756,
+      "grad_norm": 1.13874351978302,
+      "learning_rate": 4.72740443613707e-05,
+      "loss": 1.6044,
+      "step": 32120
+    },
+    {
+      "epoch": 6.7688418391531044,
+      "grad_norm": 1.1345837116241455,
+      "learning_rate": 4.7217810137009274e-05,
+      "loss": 1.5789,
+      "step": 32130
+    },
+    {
+      "epoch": 6.770948543740454,
+      "grad_norm": 1.2009543180465698,
+      "learning_rate": 4.716159904030637e-05,
+      "loss": 1.5548,
+      "step": 32140
+    },
+    {
+      "epoch": 6.773055248327803,
+      "grad_norm": 1.1193791627883911,
+      "learning_rate": 4.710541109589205e-05,
+      "loss": 1.612,
+      "step": 32150
+    },
+    {
+      "epoch": 6.775161952915153,
+      "grad_norm": 1.2925307750701904,
+      "learning_rate": 4.704924632838636e-05,
+      "loss": 1.604,
+      "step": 32160
+    },
+    {
+      "epoch": 6.777268657502502,
+      "grad_norm": 1.1835635900497437,
+      "learning_rate": 4.699310476239904e-05,
+      "loss": 1.6282,
+      "step": 32170
+    },
+    {
+      "epoch": 6.779375362089851,
+      "grad_norm": 1.1469124555587769,
+      "learning_rate": 4.693698642252979e-05,
+      "loss": 1.5664,
+      "step": 32180
+    },
+    {
+      "epoch": 6.7814820666772,
+      "grad_norm": 1.1878842115402222,
+      "learning_rate": 4.688089133336805e-05,
+      "loss": 1.6328,
+      "step": 32190
+    },
+    {
+      "epoch": 6.783588771264549,
+      "grad_norm": 1.113439679145813,
+      "learning_rate": 4.6824819519493057e-05,
+      "loss": 1.5955,
+      "step": 32200
+    },
+    {
+      "epoch": 6.785695475851899,
+      "grad_norm": 1.1598047018051147,
+      "learning_rate": 4.676877100547392e-05,
+      "loss": 1.5929,
+      "step": 32210
+    },
+    {
+      "epoch": 6.787802180439248,
+      "grad_norm": 1.0954216718673706,
+      "learning_rate": 4.671274581586958e-05,
+      "loss": 1.5691,
+      "step": 32220
+    },
+    {
+      "epoch": 6.789908885026597,
+      "grad_norm": 1.2070858478546143,
+      "learning_rate": 4.665674397522856e-05,
+      "loss": 1.6073,
+      "step": 32230
+    },
+    {
+      "epoch": 6.792015589613946,
+      "grad_norm": 1.273911476135254,
+      "learning_rate": 4.660076550808936e-05,
+      "loss": 1.5836,
+      "step": 32240
+    },
+    {
+      "epoch": 6.794122294201296,
+      "grad_norm": 1.1469873189926147,
+      "learning_rate": 4.654481043898011e-05,
+      "loss": 1.6065,
+      "step": 32250
+    },
+    {
+      "epoch": 6.796228998788645,
+      "grad_norm": 1.175328254699707,
+      "learning_rate": 4.648887879241879e-05,
+      "loss": 1.5451,
+      "step": 32260
+    },
+    {
+      "epoch": 6.798335703375994,
+      "grad_norm": 1.2798956632614136,
+      "learning_rate": 4.6432970592913026e-05,
+      "loss": 1.5988,
+      "step": 32270
+    },
+    {
+      "epoch": 6.800442407963343,
+      "grad_norm": 1.1209592819213867,
+      "learning_rate": 4.637708586496018e-05,
+      "loss": 1.5729,
+      "step": 32280
+    },
+    {
+      "epoch": 6.802549112550692,
+      "grad_norm": 1.106839656829834,
+      "learning_rate": 4.6321224633047365e-05,
+      "loss": 1.5888,
+      "step": 32290
+    },
+    {
+      "epoch": 6.804655817138042,
+      "grad_norm": 1.256089687347412,
+      "learning_rate": 4.6265386921651496e-05,
+      "loss": 1.6005,
+      "step": 32300
+    },
+    {
+      "epoch": 6.806762521725391,
+      "grad_norm": 1.0970877408981323,
+      "learning_rate": 4.6209572755238905e-05,
+      "loss": 1.5179,
+      "step": 32310
+    },
+    {
+      "epoch": 6.808869226312741,
+      "grad_norm": 1.0892575979232788,
+      "learning_rate": 4.61537821582659e-05,
+      "loss": 1.6311,
+      "step": 32320
+    },
+    {
+      "epoch": 6.8109759309000895,
+      "grad_norm": 1.0509310960769653,
+      "learning_rate": 4.609801515517825e-05,
+      "loss": 1.5343,
+      "step": 32330
+    },
+    {
+      "epoch": 6.813082635487438,
+      "grad_norm": 1.2261005640029907,
+      "learning_rate": 4.604227177041156e-05,
+      "loss": 1.6152,
+      "step": 32340
+    },
+    {
+      "epoch": 6.815189340074788,
+      "grad_norm": 1.4062305688858032,
+      "learning_rate": 4.598655202839096e-05,
+      "loss": 1.6324,
+      "step": 32350
+    },
+    {
+      "epoch": 6.817296044662137,
+      "grad_norm": 1.1476296186447144,
+      "learning_rate": 4.5930855953531214e-05,
+      "loss": 1.6057,
+      "step": 32360
+    },
+    {
+      "epoch": 6.819402749249487,
+      "grad_norm": 1.1502786874771118,
+      "learning_rate": 4.5875183570236815e-05,
+      "loss": 1.6032,
+      "step": 32370
+    },
+    {
+      "epoch": 6.8215094538368355,
+      "grad_norm": 1.0922356843948364,
+      "learning_rate": 4.581953490290188e-05,
+      "loss": 1.5566,
+      "step": 32380
+    },
+    {
+      "epoch": 6.823616158424185,
+      "grad_norm": 1.1090188026428223,
+      "learning_rate": 4.576390997590996e-05,
+      "loss": 1.6288,
+      "step": 32390
+    },
+    {
+      "epoch": 6.825722863011534,
+      "grad_norm": 1.0930324792861938,
+      "learning_rate": 4.570830881363439e-05,
+      "loss": 1.6241,
+      "step": 32400
+    },
+    {
+      "epoch": 6.827829567598884,
+      "grad_norm": 1.1673924922943115,
+      "learning_rate": 4.5652731440437965e-05,
+      "loss": 1.6006,
+      "step": 32410
+    },
+    {
+      "epoch": 6.829936272186233,
+      "grad_norm": 1.2876216173171997,
+      "learning_rate": 4.559717788067316e-05,
+      "loss": 1.5865,
+      "step": 32420
+    },
+    {
+      "epoch": 6.832042976773582,
+      "grad_norm": 1.2427315711975098,
+      "learning_rate": 4.554164815868204e-05,
+      "loss": 1.5477,
+      "step": 32430
+    },
+    {
+      "epoch": 6.834149681360931,
+      "grad_norm": 1.2781567573547363,
+      "learning_rate": 4.5486142298795995e-05,
+      "loss": 1.6095,
+      "step": 32440
+    },
+    {
+      "epoch": 6.83625638594828,
+      "grad_norm": 1.0563685894012451,
+      "learning_rate": 4.54306603253362e-05,
+      "loss": 1.5988,
+      "step": 32450
+    },
+    {
+      "epoch": 6.83836309053563,
+      "grad_norm": 1.2373830080032349,
+      "learning_rate": 4.537520226261333e-05,
+      "loss": 1.5737,
+      "step": 32460
+    },
+    {
+      "epoch": 6.840469795122979,
+      "grad_norm": 1.2991282939910889,
+      "learning_rate": 4.531976813492747e-05,
+      "loss": 1.6012,
+      "step": 32470
+    },
+    {
+      "epoch": 6.842576499710328,
+      "grad_norm": 1.3080326318740845,
+      "learning_rate": 4.5264357966568306e-05,
+      "loss": 1.5993,
+      "step": 32480
+    },
+    {
+      "epoch": 6.844683204297677,
+      "grad_norm": 1.129129409790039,
+      "learning_rate": 4.5208971781814955e-05,
+      "loss": 1.5381,
+      "step": 32490
+    },
+    {
+      "epoch": 6.846789908885027,
+      "grad_norm": 1.1796236038208008,
+      "learning_rate": 4.515360960493612e-05,
+      "loss": 1.5905,
+      "step": 32500
+    },
+    {
+      "epoch": 6.848896613472376,
+      "grad_norm": 1.5889365673065186,
+      "learning_rate": 4.509827146019e-05,
+      "loss": 1.5976,
+      "step": 32510
+    },
+    {
+      "epoch": 6.851003318059725,
+      "grad_norm": 1.1488497257232666,
+      "learning_rate": 4.5042957371824057e-05,
+      "loss": 1.6156,
+      "step": 32520
+    },
+    {
+      "epoch": 6.8531100226470745,
+      "grad_norm": 1.2863367795944214,
+      "learning_rate": 4.498766736407543e-05,
+      "loss": 1.5985,
+      "step": 32530
+    },
+    {
+      "epoch": 6.855216727234423,
+      "grad_norm": 1.0283994674682617,
+      "learning_rate": 4.493240146117066e-05,
+      "loss": 1.5567,
+      "step": 32540
+    },
+    {
+      "epoch": 6.857323431821773,
+      "grad_norm": 1.2300434112548828,
+      "learning_rate": 4.487715968732568e-05,
+      "loss": 1.5876,
+      "step": 32550
+    },
+    {
+      "epoch": 6.859430136409122,
+      "grad_norm": 1.172739863395691,
+      "learning_rate": 4.482194206674585e-05,
+      "loss": 1.5768,
+      "step": 32560
+    },
+    {
+      "epoch": 6.861536840996472,
+      "grad_norm": 1.1315293312072754,
+      "learning_rate": 4.476674862362593e-05,
+      "loss": 1.6282,
+      "step": 32570
+    },
+    {
+      "epoch": 6.8636435455838205,
+      "grad_norm": 1.1859827041625977,
+      "learning_rate": 4.471157938215017e-05,
+      "loss": 1.6081,
+      "step": 32580
+    },
+    {
+      "epoch": 6.865750250171169,
+      "grad_norm": 1.262385368347168,
+      "learning_rate": 4.465643436649224e-05,
+      "loss": 1.6119,
+      "step": 32590
+    },
+    {
+      "epoch": 6.867856954758519,
+      "grad_norm": 1.146864891052246,
+      "learning_rate": 4.460131360081496e-05,
+      "loss": 1.6026,
+      "step": 32600
+    },
+    {
+      "epoch": 6.869963659345868,
+      "grad_norm": 1.1739929914474487,
+      "learning_rate": 4.454621710927077e-05,
+      "loss": 1.6119,
+      "step": 32610
+    },
+    {
+      "epoch": 6.872070363933218,
+      "grad_norm": 1.1527085304260254,
+      "learning_rate": 4.4491144916001425e-05,
+      "loss": 1.6168,
+      "step": 32620
+    },
+    {
+      "epoch": 6.8741770685205665,
+      "grad_norm": 1.122193455696106,
+      "learning_rate": 4.443609704513797e-05,
+      "loss": 1.6375,
+      "step": 32630
+    },
+    {
+      "epoch": 6.876283773107916,
+      "grad_norm": 1.2170180082321167,
+      "learning_rate": 4.438107352080076e-05,
+      "loss": 1.6198,
+      "step": 32640
+    },
+    {
+      "epoch": 6.878390477695265,
+      "grad_norm": 1.1985820531845093,
+      "learning_rate": 4.4326074367099646e-05,
+      "loss": 1.5268,
+      "step": 32650
+    },
+    {
+      "epoch": 6.880497182282614,
+      "grad_norm": 1.2520807981491089,
+      "learning_rate": 4.42710996081336e-05,
+      "loss": 1.6396,
+      "step": 32660
+    },
+    {
+      "epoch": 6.882603886869964,
+      "grad_norm": 1.1947911977767944,
+      "learning_rate": 4.421614926799108e-05,
+      "loss": 1.5729,
+      "step": 32670
+    },
+    {
+      "epoch": 6.884710591457313,
+      "grad_norm": 1.089781403541565,
+      "learning_rate": 4.4161223370749746e-05,
+      "loss": 1.5976,
+      "step": 32680
+    },
+    {
+      "epoch": 6.886817296044662,
+      "grad_norm": 1.0784406661987305,
+      "learning_rate": 4.4106321940476516e-05,
+      "loss": 1.5902,
+      "step": 32690
+    },
+    {
+      "epoch": 6.888924000632011,
+      "grad_norm": 1.2805781364440918,
+      "learning_rate": 4.405144500122772e-05,
+      "loss": 1.6525,
+      "step": 32700
+    },
+    {
+      "epoch": 6.891030705219361,
+      "grad_norm": 1.1548502445220947,
+      "learning_rate": 4.399659257704879e-05,
+      "loss": 1.6348,
+      "step": 32710
+    },
+    {
+      "epoch": 6.89313740980671,
+      "grad_norm": 1.220477819442749,
+      "learning_rate": 4.3941764691974596e-05,
+      "loss": 1.57,
+      "step": 32720
+    },
+    {
+      "epoch": 6.895244114394059,
+      "grad_norm": 1.1432626247406006,
+      "learning_rate": 4.388696137002911e-05,
+      "loss": 1.5568,
+      "step": 32730
+    },
+    {
+      "epoch": 6.897350818981408,
+      "grad_norm": 1.245499849319458,
+      "learning_rate": 4.383218263522556e-05,
+      "loss": 1.6399,
+      "step": 32740
+    },
+    {
+      "epoch": 6.899457523568757,
+      "grad_norm": 1.1206386089324951,
+      "learning_rate": 4.377742851156652e-05,
+      "loss": 1.5825,
+      "step": 32750
+    },
+    {
+      "epoch": 6.901564228156107,
+      "grad_norm": 1.1021487712860107,
+      "learning_rate": 4.372269902304363e-05,
+      "loss": 1.6142,
+      "step": 32760
+    },
+    {
+      "epoch": 6.903670932743456,
+      "grad_norm": 1.198765754699707,
+      "learning_rate": 4.3667994193637796e-05,
+      "loss": 1.6203,
+      "step": 32770
+    },
+    {
+      "epoch": 6.9057776373308055,
+      "grad_norm": 1.1611610651016235,
+      "learning_rate": 4.3613314047319167e-05,
+      "loss": 1.6046,
+      "step": 32780
+    },
+    {
+      "epoch": 6.907884341918154,
+      "grad_norm": 1.2177382707595825,
+      "learning_rate": 4.355865860804698e-05,
+      "loss": 1.5561,
+      "step": 32790
+    },
+    {
+      "epoch": 6.909991046505504,
+      "grad_norm": 1.2026368379592896,
+      "learning_rate": 4.350402789976975e-05,
+      "loss": 1.6029,
+      "step": 32800
+    },
+    {
+      "epoch": 6.912097751092853,
+      "grad_norm": 1.1011264324188232,
+      "learning_rate": 4.3449421946425096e-05,
+      "loss": 1.5106,
+      "step": 32810
+    },
+    {
+      "epoch": 6.914204455680203,
+      "grad_norm": 1.1586644649505615,
+      "learning_rate": 4.339484077193974e-05,
+      "loss": 1.5585,
+      "step": 32820
+    },
+    {
+      "epoch": 6.9163111602675515,
+      "grad_norm": 1.0997034311294556,
+      "learning_rate": 4.3340284400229666e-05,
+      "loss": 1.6231,
+      "step": 32830
+    },
+    {
+      "epoch": 6.9184178648549,
+      "grad_norm": 1.154043197631836,
+      "learning_rate": 4.328575285519994e-05,
+      "loss": 1.6233,
+      "step": 32840
+    },
+    {
+      "epoch": 6.92052456944225,
+      "grad_norm": 1.0947542190551758,
+      "learning_rate": 4.323124616074464e-05,
+      "loss": 1.5705,
+      "step": 32850
+    },
+    {
+      "epoch": 6.922631274029599,
+      "grad_norm": 1.0261285305023193,
+      "learning_rate": 4.3176764340747177e-05,
+      "loss": 1.6055,
+      "step": 32860
+    },
+    {
+      "epoch": 6.924737978616949,
+      "grad_norm": 1.13263738155365,
+      "learning_rate": 4.312230741907984e-05,
+      "loss": 1.5702,
+      "step": 32870
+    },
+    {
+      "epoch": 6.926844683204298,
+      "grad_norm": 1.1369476318359375,
+      "learning_rate": 4.3067875419604184e-05,
+      "loss": 1.6287,
+      "step": 32880
+    },
+    {
+      "epoch": 6.928951387791647,
+      "grad_norm": 1.2614531517028809,
+      "learning_rate": 4.301346836617074e-05,
+      "loss": 1.6592,
+      "step": 32890
+    },
+    {
+      "epoch": 6.931058092378996,
+      "grad_norm": 1.1027562618255615,
+      "learning_rate": 4.29590862826191e-05,
+      "loss": 1.6164,
+      "step": 32900
+    },
+    {
+      "epoch": 6.933164796966345,
+      "grad_norm": 1.1705682277679443,
+      "learning_rate": 4.2904729192778006e-05,
+      "loss": 1.5558,
+      "step": 32910
+    },
+    {
+      "epoch": 6.935271501553695,
+      "grad_norm": 1.180618405342102,
+      "learning_rate": 4.285039712046517e-05,
+      "loss": 1.5074,
+      "step": 32920
+    },
+    {
+      "epoch": 6.937378206141044,
+      "grad_norm": 1.125746250152588,
+      "learning_rate": 4.279609008948732e-05,
+      "loss": 1.5831,
+      "step": 32930
+    },
+    {
+      "epoch": 6.939484910728393,
+      "grad_norm": 1.092772126197815,
+      "learning_rate": 4.2741808123640335e-05,
+      "loss": 1.6018,
+      "step": 32940
+    },
+    {
+      "epoch": 6.941591615315742,
+      "grad_norm": 1.1880598068237305,
+      "learning_rate": 4.2687551246708965e-05,
+      "loss": 1.6156,
+      "step": 32950
+    },
+    {
+      "epoch": 6.943698319903092,
+      "grad_norm": 1.1571134328842163,
+      "learning_rate": 4.26333194824671e-05,
+      "loss": 1.6002,
+      "step": 32960
+    },
+    {
+      "epoch": 6.945805024490441,
+      "grad_norm": 1.2684969902038574,
+      "learning_rate": 4.257911285467754e-05,
+      "loss": 1.6134,
+      "step": 32970
+    },
+    {
+      "epoch": 6.94791172907779,
+      "grad_norm": 1.1462793350219727,
+      "learning_rate": 4.252493138709204e-05,
+      "loss": 1.5812,
+      "step": 32980
+    },
+    {
+      "epoch": 6.950018433665139,
+      "grad_norm": 1.157240629196167,
+      "learning_rate": 4.2470775103451446e-05,
+      "loss": 1.634,
+      "step": 32990
+    },
+    {
+      "epoch": 6.952125138252488,
+      "grad_norm": 1.1271111965179443,
+      "learning_rate": 4.241664402748544e-05,
+      "loss": 1.5717,
+      "step": 33000
+    },
+    {
+      "epoch": 6.954231842839838,
+      "grad_norm": 1.1427202224731445,
+      "learning_rate": 4.236253818291281e-05,
+      "loss": 1.6627,
+      "step": 33010
+    },
+    {
+      "epoch": 6.956338547427187,
+      "grad_norm": 1.1169521808624268,
+      "learning_rate": 4.230845759344116e-05,
+      "loss": 1.589,
+      "step": 33020
+    },
+    {
+      "epoch": 6.958445252014537,
+      "grad_norm": 1.222920298576355,
+      "learning_rate": 4.2254402282767034e-05,
+      "loss": 1.6033,
+      "step": 33030
+    },
+    {
+      "epoch": 6.960551956601885,
+      "grad_norm": 1.2239333391189575,
+      "learning_rate": 4.2200372274576e-05,
+      "loss": 1.5999,
+      "step": 33040
+    },
+    {
+      "epoch": 6.962658661189234,
+      "grad_norm": 1.1619553565979004,
+      "learning_rate": 4.2146367592542444e-05,
+      "loss": 1.6088,
+      "step": 33050
+    },
+    {
+      "epoch": 6.964765365776584,
+      "grad_norm": 1.230373501777649,
+      "learning_rate": 4.209238826032965e-05,
+      "loss": 1.5548,
+      "step": 33060
+    },
+    {
+      "epoch": 6.966872070363933,
+      "grad_norm": 1.0696500539779663,
+      "learning_rate": 4.203843430158991e-05,
+      "loss": 1.6181,
+      "step": 33070
+    },
+    {
+      "epoch": 6.968978774951283,
+      "grad_norm": 1.0919783115386963,
+      "learning_rate": 4.198450573996423e-05,
+      "loss": 1.6078,
+      "step": 33080
+    },
+    {
+      "epoch": 6.9710854795386314,
+      "grad_norm": 1.0691747665405273,
+      "learning_rate": 4.1930602599082666e-05,
+      "loss": 1.6057,
+      "step": 33090
+    },
+    {
+      "epoch": 6.973192184125981,
+      "grad_norm": 1.1408584117889404,
+      "learning_rate": 4.1876724902564004e-05,
+      "loss": 1.607,
+      "step": 33100
+    },
+    {
+      "epoch": 6.97529888871333,
+      "grad_norm": 1.2705100774765015,
+      "learning_rate": 4.182287267401587e-05,
+      "loss": 1.5754,
+      "step": 33110
+    },
+    {
+      "epoch": 6.97740559330068,
+      "grad_norm": 1.117083191871643,
+      "learning_rate": 4.1769045937034876e-05,
+      "loss": 1.5889,
+      "step": 33120
+    },
+    {
+      "epoch": 6.979512297888029,
+      "grad_norm": 1.1689233779907227,
+      "learning_rate": 4.171524471520633e-05,
+      "loss": 1.5913,
+      "step": 33130
+    },
+    {
+      "epoch": 6.981619002475378,
+      "grad_norm": 1.0697065591812134,
+      "learning_rate": 4.166146903210436e-05,
+      "loss": 1.5788,
+      "step": 33140
+    },
+    {
+      "epoch": 6.983725707062727,
+      "grad_norm": 1.1633777618408203,
+      "learning_rate": 4.1607718911292025e-05,
+      "loss": 1.5982,
+      "step": 33150
+    },
+    {
+      "epoch": 6.985832411650076,
+      "grad_norm": 1.1077229976654053,
+      "learning_rate": 4.1553994376321023e-05,
+      "loss": 1.5866,
+      "step": 33160
+    },
+    {
+      "epoch": 6.987939116237426,
+      "grad_norm": 1.1923149824142456,
+      "learning_rate": 4.1500295450731994e-05,
+      "loss": 1.6038,
+      "step": 33170
+    },
+    {
+      "epoch": 6.990045820824775,
+      "grad_norm": 1.184695839881897,
+      "learning_rate": 4.144662215805426e-05,
+      "loss": 1.5184,
+      "step": 33180
+    },
+    {
+      "epoch": 6.992152525412124,
+      "grad_norm": 1.1566641330718994,
+      "learning_rate": 4.139297452180588e-05,
+      "loss": 1.6186,
+      "step": 33190
+    },
+    {
+      "epoch": 6.994259229999473,
+      "grad_norm": 1.1046571731567383,
+      "learning_rate": 4.133935256549383e-05,
+      "loss": 1.6273,
+      "step": 33200
+    },
+    {
+      "epoch": 6.996365934586823,
+      "grad_norm": 1.2560962438583374,
+      "learning_rate": 4.1285756312613654e-05,
+      "loss": 1.5725,
+      "step": 33210
+    },
+    {
+      "epoch": 6.998472639174172,
+      "grad_norm": 1.1830743551254272,
+      "learning_rate": 4.1232185786649704e-05,
+      "loss": 1.5894,
+      "step": 33220
+    },
+    {
+      "epoch": 7.000579343761521,
+      "grad_norm": 1.2797951698303223,
+      "learning_rate": 4.1178641011075116e-05,
+      "loss": 1.5852,
+      "step": 33230
+    },
+    {
+      "epoch": 7.00268604834887,
+      "grad_norm": 1.0687123537063599,
+      "learning_rate": 4.1125122009351634e-05,
+      "loss": 1.5228,
+      "step": 33240
+    },
+    {
+      "epoch": 7.004792752936219,
+      "grad_norm": 1.206784963607788,
+      "learning_rate": 4.107162880492984e-05,
+      "loss": 1.5989,
+      "step": 33250
+    },
+    {
+      "epoch": 7.006899457523569,
+      "grad_norm": 1.2578917741775513,
+      "learning_rate": 4.1018161421248905e-05,
+      "loss": 1.5553,
+      "step": 33260
+    },
+    {
+      "epoch": 7.009006162110918,
+      "grad_norm": 1.4243639707565308,
+      "learning_rate": 4.096471988173667e-05,
+      "loss": 1.5727,
+      "step": 33270
+    },
+    {
+      "epoch": 7.011112866698268,
+      "grad_norm": 1.162205696105957,
+      "learning_rate": 4.0911304209809776e-05,
+      "loss": 1.525,
+      "step": 33280
+    },
+    {
+      "epoch": 7.0132195712856165,
+      "grad_norm": 1.0842784643173218,
+      "learning_rate": 4.08579144288734e-05,
+      "loss": 1.5675,
+      "step": 33290
+    },
+    {
+      "epoch": 7.015326275872965,
+      "grad_norm": 1.1000008583068848,
+      "learning_rate": 4.080455056232147e-05,
+      "loss": 1.5373,
+      "step": 33300
+    },
+    {
+      "epoch": 7.017432980460315,
+      "grad_norm": 1.184962511062622,
+      "learning_rate": 4.075121263353653e-05,
+      "loss": 1.532,
+      "step": 33310
+    },
+    {
+      "epoch": 7.019539685047664,
+      "grad_norm": 1.0851813554763794,
+      "learning_rate": 4.069790066588967e-05,
+      "loss": 1.5424,
+      "step": 33320
+    },
+    {
+      "epoch": 7.021646389635014,
+      "grad_norm": 1.2001564502716064,
+      "learning_rate": 4.064461468274077e-05,
+      "loss": 1.5278,
+      "step": 33330
+    },
+    {
+      "epoch": 7.0237530942223625,
+      "grad_norm": 1.15140962600708,
+      "learning_rate": 4.05913547074382e-05,
+      "loss": 1.5639,
+      "step": 33340
+    },
+    {
+      "epoch": 7.025859798809712,
+      "grad_norm": 1.1785743236541748,
+      "learning_rate": 4.053812076331893e-05,
+      "loss": 1.5807,
+      "step": 33350
+    },
+    {
+      "epoch": 7.027966503397061,
+      "grad_norm": 1.2136023044586182,
+      "learning_rate": 4.048491287370863e-05,
+      "loss": 1.5895,
+      "step": 33360
+    },
+    {
+      "epoch": 7.030073207984411,
+      "grad_norm": 1.275848150253296,
+      "learning_rate": 4.043173106192145e-05,
+      "loss": 1.5946,
+      "step": 33370
+    },
+    {
+      "epoch": 7.03217991257176,
+      "grad_norm": 1.157455325126648,
+      "learning_rate": 4.0378575351260184e-05,
+      "loss": 1.5566,
+      "step": 33380
+    },
+    {
+      "epoch": 7.0342866171591085,
+      "grad_norm": 1.1953786611557007,
+      "learning_rate": 4.0325445765016145e-05,
+      "loss": 1.5714,
+      "step": 33390
+    },
+    {
+      "epoch": 7.036393321746458,
+      "grad_norm": 1.1416175365447998,
+      "learning_rate": 4.0272342326469157e-05,
+      "loss": 1.6163,
+      "step": 33400
+    },
+    {
+      "epoch": 7.038500026333807,
+      "grad_norm": 1.0863345861434937,
+      "learning_rate": 4.021926505888774e-05,
+      "loss": 1.5635,
+      "step": 33410
+    },
+    {
+      "epoch": 7.040606730921157,
+      "grad_norm": 1.1222482919692993,
+      "learning_rate": 4.016621398552877e-05,
+      "loss": 1.5723,
+      "step": 33420
+    },
+    {
+      "epoch": 7.042713435508506,
+      "grad_norm": 1.2037105560302734,
+      "learning_rate": 4.011318912963772e-05,
+      "loss": 1.5243,
+      "step": 33430
+    },
+    {
+      "epoch": 7.044820140095855,
+      "grad_norm": 1.1907230615615845,
+      "learning_rate": 4.006019051444864e-05,
+      "loss": 1.5661,
+      "step": 33440
+    },
+    {
+      "epoch": 7.046926844683204,
+      "grad_norm": 1.17070734500885,
+      "learning_rate": 4.000721816318395e-05,
+      "loss": 1.5788,
+      "step": 33450
+    },
+    {
+      "epoch": 7.049033549270553,
+      "grad_norm": 1.2719948291778564,
+      "learning_rate": 3.995427209905469e-05,
+      "loss": 1.5891,
+      "step": 33460
+    },
+    {
+      "epoch": 7.051140253857903,
+      "grad_norm": 1.1980891227722168,
+      "learning_rate": 3.99013523452603e-05,
+      "loss": 1.569,
+      "step": 33470
+    },
+    {
+      "epoch": 7.053246958445252,
+      "grad_norm": 1.1968605518341064,
+      "learning_rate": 3.9848458924988684e-05,
+      "loss": 1.5259,
+      "step": 33480
+    },
+    {
+      "epoch": 7.0553536630326015,
+      "grad_norm": 1.1652711629867554,
+      "learning_rate": 3.9795591861416316e-05,
+      "loss": 1.5396,
+      "step": 33490
+    },
+    {
+      "epoch": 7.05746036761995,
+      "grad_norm": 1.15365469455719,
+      "learning_rate": 3.974275117770798e-05,
+      "loss": 1.5551,
+      "step": 33500
+    },
+    {
+      "epoch": 7.0595670722073,
+      "grad_norm": 1.1745774745941162,
+      "learning_rate": 3.9689936897016944e-05,
+      "loss": 1.5099,
+      "step": 33510
+    },
+    {
+      "epoch": 7.061673776794649,
+      "grad_norm": 1.1286360025405884,
+      "learning_rate": 3.963714904248501e-05,
+      "loss": 1.5246,
+      "step": 33520
+    },
+    {
+      "epoch": 7.063780481381999,
+      "grad_norm": 1.1611193418502808,
+      "learning_rate": 3.958438763724224e-05,
+      "loss": 1.5001,
+      "step": 33530
+    },
+    {
+      "epoch": 7.0658871859693475,
+      "grad_norm": 1.140989065170288,
+      "learning_rate": 3.953165270440721e-05,
+      "loss": 1.5921,
+      "step": 33540
+    },
+    {
+      "epoch": 7.067993890556696,
+      "grad_norm": 1.2701148986816406,
+      "learning_rate": 3.947894426708696e-05,
+      "loss": 1.5792,
+      "step": 33550
+    },
+    {
+      "epoch": 7.070100595144046,
+      "grad_norm": 1.1872613430023193,
+      "learning_rate": 3.942626234837668e-05,
+      "loss": 1.5927,
+      "step": 33560
+    },
+    {
+      "epoch": 7.072207299731395,
+      "grad_norm": 1.215997338294983,
+      "learning_rate": 3.937360697136019e-05,
+      "loss": 1.5151,
+      "step": 33570
+    },
+    {
+      "epoch": 7.074314004318745,
+      "grad_norm": 1.2439547777175903,
+      "learning_rate": 3.9320978159109533e-05,
+      "loss": 1.5755,
+      "step": 33580
+    },
+    {
+      "epoch": 7.0764207089060935,
+      "grad_norm": 1.2074193954467773,
+      "learning_rate": 3.926837593468522e-05,
+      "loss": 1.5721,
+      "step": 33590
+    },
+    {
+      "epoch": 7.078527413493443,
+      "grad_norm": 1.0745258331298828,
+      "learning_rate": 3.921580032113602e-05,
+      "loss": 1.5437,
+      "step": 33600
+    },
+    {
+      "epoch": 7.080634118080792,
+      "grad_norm": 1.3252061605453491,
+      "learning_rate": 3.916325134149904e-05,
+      "loss": 1.5626,
+      "step": 33610
+    },
+    {
+      "epoch": 7.082740822668141,
+      "grad_norm": 1.164564609527588,
+      "learning_rate": 3.9110729018799785e-05,
+      "loss": 1.5461,
+      "step": 33620
+    },
+    {
+      "epoch": 7.084847527255491,
+      "grad_norm": 1.2133631706237793,
+      "learning_rate": 3.905823337605213e-05,
+      "loss": 1.5484,
+      "step": 33630
+    },
+    {
+      "epoch": 7.08695423184284,
+      "grad_norm": 1.2363920211791992,
+      "learning_rate": 3.900576443625803e-05,
+      "loss": 1.5937,
+      "step": 33640
+    },
+    {
+      "epoch": 7.089060936430189,
+      "grad_norm": 1.2874761819839478,
+      "learning_rate": 3.895332222240794e-05,
+      "loss": 1.4771,
+      "step": 33650
+    },
+    {
+      "epoch": 7.091167641017538,
+      "grad_norm": 1.1343390941619873,
+      "learning_rate": 3.8900906757480614e-05,
+      "loss": 1.5485,
+      "step": 33660
+    },
+    {
+      "epoch": 7.093274345604888,
+      "grad_norm": 1.2385504245758057,
+      "learning_rate": 3.884851806444296e-05,
+      "loss": 1.5506,
+      "step": 33670
+    },
+    {
+      "epoch": 7.095381050192237,
+      "grad_norm": 1.160474181175232,
+      "learning_rate": 3.879615616625024e-05,
+      "loss": 1.5962,
+      "step": 33680
+    },
+    {
+      "epoch": 7.097487754779586,
+      "grad_norm": 1.2261831760406494,
+      "learning_rate": 3.874382108584591e-05,
+      "loss": 1.5721,
+      "step": 33690
+    },
+    {
+      "epoch": 7.099594459366935,
+      "grad_norm": 1.2461113929748535,
+      "learning_rate": 3.8691512846161737e-05,
+      "loss": 1.5832,
+      "step": 33700
+    },
+    {
+      "epoch": 7.101701163954284,
+      "grad_norm": 1.687401294708252,
+      "learning_rate": 3.86392314701178e-05,
+      "loss": 1.5928,
+      "step": 33710
+    },
+    {
+      "epoch": 7.103807868541634,
+      "grad_norm": 1.188767433166504,
+      "learning_rate": 3.858697698062217e-05,
+      "loss": 1.6336,
+      "step": 33720
+    },
+    {
+      "epoch": 7.105914573128983,
+      "grad_norm": 1.3257564306259155,
+      "learning_rate": 3.8534749400571337e-05,
+      "loss": 1.5285,
+      "step": 33730
+    },
+    {
+      "epoch": 7.1080212777163325,
+      "grad_norm": 1.227262258529663,
+      "learning_rate": 3.848254875285e-05,
+      "loss": 1.5814,
+      "step": 33740
+    },
+    {
+      "epoch": 7.110127982303681,
+      "grad_norm": 1.3911917209625244,
+      "learning_rate": 3.843037506033096e-05,
+      "loss": 1.6014,
+      "step": 33750
+    },
+    {
+      "epoch": 7.112234686891031,
+      "grad_norm": 1.1162062883377075,
+      "learning_rate": 3.8378228345875246e-05,
+      "loss": 1.5378,
+      "step": 33760
+    },
+    {
+      "epoch": 7.11434139147838,
+      "grad_norm": 1.1487606763839722,
+      "learning_rate": 3.832610863233204e-05,
+      "loss": 1.5834,
+      "step": 33770
+    },
+    {
+      "epoch": 7.116448096065729,
+      "grad_norm": 1.1570956707000732,
+      "learning_rate": 3.8274015942538745e-05,
+      "loss": 1.5618,
+      "step": 33780
+    },
+    {
+      "epoch": 7.1185548006530786,
+      "grad_norm": 1.1539541482925415,
+      "learning_rate": 3.822195029932095e-05,
+      "loss": 1.5992,
+      "step": 33790
+    },
+    {
+      "epoch": 7.120661505240427,
+      "grad_norm": 1.1285444498062134,
+      "learning_rate": 3.8169911725492303e-05,
+      "loss": 1.5704,
+      "step": 33800
+    },
+    {
+      "epoch": 7.122768209827777,
+      "grad_norm": 1.1300140619277954,
+      "learning_rate": 3.8117900243854595e-05,
+      "loss": 1.528,
+      "step": 33810
+    },
+    {
+      "epoch": 7.124874914415126,
+      "grad_norm": 1.538737416267395,
+      "learning_rate": 3.806591587719784e-05,
+      "loss": 1.5414,
+      "step": 33820
+    },
+    {
+      "epoch": 7.126981619002476,
+      "grad_norm": 1.2989195585250854,
+      "learning_rate": 3.8013958648300075e-05,
+      "loss": 1.5495,
+      "step": 33830
+    },
+    {
+      "epoch": 7.129088323589825,
+      "grad_norm": 1.1311976909637451,
+      "learning_rate": 3.7962028579927555e-05,
+      "loss": 1.5997,
+      "step": 33840
+    },
+    {
+      "epoch": 7.131195028177173,
+      "grad_norm": 1.0657979249954224,
+      "learning_rate": 3.7910125694834445e-05,
+      "loss": 1.5286,
+      "step": 33850
+    },
+    {
+      "epoch": 7.133301732764523,
+      "grad_norm": 1.199328899383545,
+      "learning_rate": 3.7858250015763174e-05,
+      "loss": 1.5864,
+      "step": 33860
+    },
+    {
+      "epoch": 7.135408437351872,
+      "grad_norm": 1.2691328525543213,
+      "learning_rate": 3.780640156544424e-05,
+      "loss": 1.5788,
+      "step": 33870
+    },
+    {
+      "epoch": 7.137515141939222,
+      "grad_norm": 1.2796211242675781,
+      "learning_rate": 3.7754580366596115e-05,
+      "loss": 1.6114,
+      "step": 33880
+    },
+    {
+      "epoch": 7.139621846526571,
+      "grad_norm": 1.1912226676940918,
+      "learning_rate": 3.7702786441925355e-05,
+      "loss": 1.5474,
+      "step": 33890
+    },
+    {
+      "epoch": 7.14172855111392,
+      "grad_norm": 1.2463699579238892,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 1.5949,
+      "step": 33900
+    },
+    {
+      "epoch": 7.143835255701269,
+      "grad_norm": 1.1586321592330933,
+      "learning_rate": 3.7599280505882604e-05,
+      "loss": 1.5561,
+      "step": 33910
+    },
+    {
+      "epoch": 7.145941960288619,
+      "grad_norm": 1.2323464155197144,
+      "learning_rate": 3.7547568539864017e-05,
+      "loss": 1.5535,
+      "step": 33920
+    },
+    {
+      "epoch": 7.148048664875968,
+      "grad_norm": 1.402441143989563,
+      "learning_rate": 3.749588393872947e-05,
+      "loss": 1.5405,
+      "step": 33930
+    },
+    {
+      "epoch": 7.150155369463317,
+      "grad_norm": 1.1518326997756958,
+      "learning_rate": 3.7444226725125764e-05,
+      "loss": 1.5759,
+      "step": 33940
+    },
+    {
+      "epoch": 7.152262074050666,
+      "grad_norm": 1.2248179912567139,
+      "learning_rate": 3.739259692168764e-05,
+      "loss": 1.5951,
+      "step": 33950
+    },
+    {
+      "epoch": 7.154368778638015,
+      "grad_norm": 1.2795392274856567,
+      "learning_rate": 3.734099455103779e-05,
+      "loss": 1.5979,
+      "step": 33960
+    },
+    {
+      "epoch": 7.156475483225365,
+      "grad_norm": 1.1059311628341675,
+      "learning_rate": 3.728941963578687e-05,
+      "loss": 1.557,
+      "step": 33970
+    },
+    {
+      "epoch": 7.158582187812714,
+      "grad_norm": 1.275423288345337,
+      "learning_rate": 3.723787219853363e-05,
+      "loss": 1.549,
+      "step": 33980
+    },
+    {
+      "epoch": 7.160688892400064,
+      "grad_norm": 1.1001471281051636,
+      "learning_rate": 3.71863522618646e-05,
+      "loss": 1.5314,
+      "step": 33990
+    },
+    {
+      "epoch": 7.162795596987412,
+      "grad_norm": 1.242405652999878,
+      "learning_rate": 3.7134859848354485e-05,
+      "loss": 1.5806,
+      "step": 34000
+    },
+    {
+      "epoch": 7.164902301574761,
+      "grad_norm": 1.2491405010223389,
+      "learning_rate": 3.708339498056565e-05,
+      "loss": 1.543,
+      "step": 34010
+    },
+    {
+      "epoch": 7.167009006162111,
+      "grad_norm": 1.1283265352249146,
+      "learning_rate": 3.7031957681048604e-05,
+      "loss": 1.5682,
+      "step": 34020
+    },
+    {
+      "epoch": 7.16911571074946,
+      "grad_norm": 1.109413743019104,
+      "learning_rate": 3.698054797234175e-05,
+      "loss": 1.5197,
+      "step": 34030
+    },
+    {
+      "epoch": 7.17122241533681,
+      "grad_norm": 1.24946129322052,
+      "learning_rate": 3.6929165876971337e-05,
+      "loss": 1.5934,
+      "step": 34040
+    },
+    {
+      "epoch": 7.1733291199241584,
+      "grad_norm": 1.0502756834030151,
+      "learning_rate": 3.68778114174515e-05,
+      "loss": 1.5576,
+      "step": 34050
+    },
+    {
+      "epoch": 7.175435824511508,
+      "grad_norm": 1.1192634105682373,
+      "learning_rate": 3.682648461628439e-05,
+      "loss": 1.5288,
+      "step": 34060
+    },
+    {
+      "epoch": 7.177542529098857,
+      "grad_norm": 1.2234610319137573,
+      "learning_rate": 3.677518549595986e-05,
+      "loss": 1.5302,
+      "step": 34070
+    },
+    {
+      "epoch": 7.179649233686207,
+      "grad_norm": 1.157472014427185,
+      "learning_rate": 3.6723914078955825e-05,
+      "loss": 1.5859,
+      "step": 34080
+    },
+    {
+      "epoch": 7.181755938273556,
+      "grad_norm": 1.1958365440368652,
+      "learning_rate": 3.667267038773791e-05,
+      "loss": 1.5783,
+      "step": 34090
+    },
+    {
+      "epoch": 7.1838626428609045,
+      "grad_norm": 1.1906516551971436,
+      "learning_rate": 3.662145444475963e-05,
+      "loss": 1.5748,
+      "step": 34100
+    },
+    {
+      "epoch": 7.185969347448254,
+      "grad_norm": 1.313653588294983,
+      "learning_rate": 3.65702662724624e-05,
+      "loss": 1.5847,
+      "step": 34110
+    },
+    {
+      "epoch": 7.188076052035603,
+      "grad_norm": 1.2378566265106201,
+      "learning_rate": 3.65191058932754e-05,
+      "loss": 1.5802,
+      "step": 34120
+    },
+    {
+      "epoch": 7.190182756622953,
+      "grad_norm": 1.146131992340088,
+      "learning_rate": 3.646797332961569e-05,
+      "loss": 1.5321,
+      "step": 34130
+    },
+    {
+      "epoch": 7.192289461210302,
+      "grad_norm": 1.2050758600234985,
+      "learning_rate": 3.64168686038881e-05,
+      "loss": 1.6303,
+      "step": 34140
+    },
+    {
+      "epoch": 7.194396165797651,
+      "grad_norm": 1.172650933265686,
+      "learning_rate": 3.636579173848521e-05,
+      "loss": 1.5502,
+      "step": 34150
+    },
+    {
+      "epoch": 7.196502870385,
+      "grad_norm": 1.1262861490249634,
+      "learning_rate": 3.631474275578754e-05,
+      "loss": 1.5185,
+      "step": 34160
+    },
+    {
+      "epoch": 7.198609574972349,
+      "grad_norm": 1.1936100721359253,
+      "learning_rate": 3.626372167816326e-05,
+      "loss": 1.5495,
+      "step": 34170
+    },
+    {
+      "epoch": 7.200716279559699,
+      "grad_norm": 1.0827867984771729,
+      "learning_rate": 3.6212728527968345e-05,
+      "loss": 1.5777,
+      "step": 34180
+    },
+    {
+      "epoch": 7.202822984147048,
+      "grad_norm": 1.1810598373413086,
+      "learning_rate": 3.616176332754659e-05,
+      "loss": 1.5491,
+      "step": 34190
+    },
+    {
+      "epoch": 7.204929688734397,
+      "grad_norm": 1.2283319234848022,
+      "learning_rate": 3.6110826099229453e-05,
+      "loss": 1.5434,
+      "step": 34200
+    },
+    {
+      "epoch": 7.207036393321746,
+      "grad_norm": 1.2398056983947754,
+      "learning_rate": 3.605991686533625e-05,
+      "loss": 1.5411,
+      "step": 34210
+    },
+    {
+      "epoch": 7.209143097909096,
+      "grad_norm": 1.17624831199646,
+      "learning_rate": 3.6009035648173914e-05,
+      "loss": 1.5901,
+      "step": 34220
+    },
+    {
+      "epoch": 7.211249802496445,
+      "grad_norm": 1.192043662071228,
+      "learning_rate": 3.595818247003713e-05,
+      "loss": 1.5786,
+      "step": 34230
+    },
+    {
+      "epoch": 7.213356507083795,
+      "grad_norm": 1.3036718368530273,
+      "learning_rate": 3.590735735320837e-05,
+      "loss": 1.5868,
+      "step": 34240
+    },
+    {
+      "epoch": 7.2154632116711435,
+      "grad_norm": 1.1599690914154053,
+      "learning_rate": 3.5856560319957747e-05,
+      "loss": 1.5652,
+      "step": 34250
+    },
+    {
+      "epoch": 7.217569916258492,
+      "grad_norm": 1.0903196334838867,
+      "learning_rate": 3.580579139254303e-05,
+      "loss": 1.6047,
+      "step": 34260
+    },
+    {
+      "epoch": 7.219676620845842,
+      "grad_norm": 1.1610807180404663,
+      "learning_rate": 3.57550505932098e-05,
+      "loss": 1.5608,
+      "step": 34270
+    },
+    {
+      "epoch": 7.221783325433191,
+      "grad_norm": 1.1886292695999146,
+      "learning_rate": 3.570433794419117e-05,
+      "loss": 1.5595,
+      "step": 34280
+    },
+    {
+      "epoch": 7.223890030020541,
+      "grad_norm": 1.1972240209579468,
+      "learning_rate": 3.565365346770805e-05,
+      "loss": 1.5889,
+      "step": 34290
+    },
+    {
+      "epoch": 7.2259967346078895,
+      "grad_norm": 1.13959538936615,
+      "learning_rate": 3.560299718596889e-05,
+      "loss": 1.5306,
+      "step": 34300
+    },
+    {
+      "epoch": 7.228103439195239,
+      "grad_norm": 1.218733549118042,
+      "learning_rate": 3.5552369121169815e-05,
+      "loss": 1.5694,
+      "step": 34310
+    },
+    {
+      "epoch": 7.230210143782588,
+      "grad_norm": 1.1074182987213135,
+      "learning_rate": 3.550176929549468e-05,
+      "loss": 1.5485,
+      "step": 34320
+    },
+    {
+      "epoch": 7.232316848369937,
+      "grad_norm": 1.186241626739502,
+      "learning_rate": 3.545119773111486e-05,
+      "loss": 1.5769,
+      "step": 34330
+    },
+    {
+      "epoch": 7.234423552957287,
+      "grad_norm": 1.160736083984375,
+      "learning_rate": 3.540065445018933e-05,
+      "loss": 1.5437,
+      "step": 34340
+    },
+    {
+      "epoch": 7.2365302575446355,
+      "grad_norm": 1.1813569068908691,
+      "learning_rate": 3.535013947486481e-05,
+      "loss": 1.5968,
+      "step": 34350
+    },
+    {
+      "epoch": 7.238636962131985,
+      "grad_norm": 1.1769226789474487,
+      "learning_rate": 3.5299652827275455e-05,
+      "loss": 1.5811,
+      "step": 34360
+    },
+    {
+      "epoch": 7.240743666719334,
+      "grad_norm": 1.207960605621338,
+      "learning_rate": 3.5249194529543137e-05,
+      "loss": 1.5773,
+      "step": 34370
+    },
+    {
+      "epoch": 7.242850371306684,
+      "grad_norm": 1.3144844770431519,
+      "learning_rate": 3.5198764603777235e-05,
+      "loss": 1.613,
+      "step": 34380
+    },
+    {
+      "epoch": 7.244957075894033,
+      "grad_norm": 1.3451526165008545,
+      "learning_rate": 3.5148363072074666e-05,
+      "loss": 1.5226,
+      "step": 34390
+    },
+    {
+      "epoch": 7.247063780481382,
+      "grad_norm": 1.1638214588165283,
+      "learning_rate": 3.509798995652002e-05,
+      "loss": 1.5835,
+      "step": 34400
+    },
+    {
+      "epoch": 7.249170485068731,
+      "grad_norm": 1.2865285873413086,
+      "learning_rate": 3.50476452791853e-05,
+      "loss": 1.5573,
+      "step": 34410
+    },
+    {
+      "epoch": 7.25127718965608,
+      "grad_norm": 1.2313289642333984,
+      "learning_rate": 3.49973290621302e-05,
+      "loss": 1.5818,
+      "step": 34420
+    },
+    {
+      "epoch": 7.25338389424343,
+      "grad_norm": 1.1138957738876343,
+      "learning_rate": 3.494704132740181e-05,
+      "loss": 1.6223,
+      "step": 34430
+    },
+    {
+      "epoch": 7.255490598830779,
+      "grad_norm": 1.151701807975769,
+      "learning_rate": 3.489678209703475e-05,
+      "loss": 1.6069,
+      "step": 34440
+    },
+    {
+      "epoch": 7.2575973034181285,
+      "grad_norm": 1.2936291694641113,
+      "learning_rate": 3.4846551393051265e-05,
+      "loss": 1.541,
+      "step": 34450
+    },
+    {
+      "epoch": 7.259704008005477,
+      "grad_norm": 1.155973196029663,
+      "learning_rate": 3.4796349237461e-05,
+      "loss": 1.5876,
+      "step": 34460
+    },
+    {
+      "epoch": 7.261810712592827,
+      "grad_norm": 1.261993408203125,
+      "learning_rate": 3.4746175652261056e-05,
+      "loss": 1.5971,
+      "step": 34470
+    },
+    {
+      "epoch": 7.263917417180176,
+      "grad_norm": 1.2483882904052734,
+      "learning_rate": 3.469603065943617e-05,
+      "loss": 1.5946,
+      "step": 34480
+    },
+    {
+      "epoch": 7.266024121767525,
+      "grad_norm": 1.229948878288269,
+      "learning_rate": 3.464591428095838e-05,
+      "loss": 1.6399,
+      "step": 34490
+    },
+    {
+      "epoch": 7.2681308263548745,
+      "grad_norm": 1.2871240377426147,
+      "learning_rate": 3.459582653878731e-05,
+      "loss": 1.5594,
+      "step": 34500
+    },
+    {
+      "epoch": 7.270237530942223,
+      "grad_norm": 1.1843843460083008,
+      "learning_rate": 3.4545767454869995e-05,
+      "loss": 1.6033,
+      "step": 34510
+    },
+    {
+      "epoch": 7.272344235529573,
+      "grad_norm": 1.0694409608840942,
+      "learning_rate": 3.449573705114082e-05,
+      "loss": 1.5933,
+      "step": 34520
+    },
+    {
+      "epoch": 7.274450940116922,
+      "grad_norm": 1.301433801651001,
+      "learning_rate": 3.44457353495218e-05,
+      "loss": 1.5651,
+      "step": 34530
+    },
+    {
+      "epoch": 7.276557644704272,
+      "grad_norm": 1.12484872341156,
+      "learning_rate": 3.43957623719222e-05,
+      "loss": 1.6104,
+      "step": 34540
+    },
+    {
+      "epoch": 7.2786643492916205,
+      "grad_norm": 1.188636064529419,
+      "learning_rate": 3.434581814023875e-05,
+      "loss": 1.5948,
+      "step": 34550
+    },
+    {
+      "epoch": 7.28077105387897,
+      "grad_norm": 1.1812471151351929,
+      "learning_rate": 3.429590267635565e-05,
+      "loss": 1.5563,
+      "step": 34560
+    },
+    {
+      "epoch": 7.282877758466319,
+      "grad_norm": 1.2685455083847046,
+      "learning_rate": 3.4246016002144377e-05,
+      "loss": 1.5661,
+      "step": 34570
+    },
+    {
+      "epoch": 7.284984463053668,
+      "grad_norm": 1.173599123954773,
+      "learning_rate": 3.4196158139463915e-05,
+      "loss": 1.6113,
+      "step": 34580
+    },
+    {
+      "epoch": 7.287091167641018,
+      "grad_norm": 1.1645420789718628,
+      "learning_rate": 3.414632911016056e-05,
+      "loss": 1.5558,
+      "step": 34590
+    },
+    {
+      "epoch": 7.289197872228367,
+      "grad_norm": 1.149623990058899,
+      "learning_rate": 3.4096528936067905e-05,
+      "loss": 1.5926,
+      "step": 34600
+    },
+    {
+      "epoch": 7.291304576815716,
+      "grad_norm": 1.3477739095687866,
+      "learning_rate": 3.4046757639007066e-05,
+      "loss": 1.5735,
+      "step": 34610
+    },
+    {
+      "epoch": 7.293411281403065,
+      "grad_norm": 1.1365127563476562,
+      "learning_rate": 3.399701524078635e-05,
+      "loss": 1.5733,
+      "step": 34620
+    },
+    {
+      "epoch": 7.295517985990415,
+      "grad_norm": 1.2671220302581787,
+      "learning_rate": 3.394730176320151e-05,
+      "loss": 1.5625,
+      "step": 34630
+    },
+    {
+      "epoch": 7.297624690577764,
+      "grad_norm": 1.344285249710083,
+      "learning_rate": 3.389761722803557e-05,
+      "loss": 1.5495,
+      "step": 34640
+    },
+    {
+      "epoch": 7.299731395165113,
+      "grad_norm": 1.2595552206039429,
+      "learning_rate": 3.3847961657058845e-05,
+      "loss": 1.5918,
+      "step": 34650
+    },
+    {
+      "epoch": 7.301838099752462,
+      "grad_norm": 1.2583255767822266,
+      "learning_rate": 3.379833507202903e-05,
+      "loss": 1.591,
+      "step": 34660
+    },
+    {
+      "epoch": 7.303944804339811,
+      "grad_norm": 1.2452776432037354,
+      "learning_rate": 3.3748737494691153e-05,
+      "loss": 1.534,
+      "step": 34670
+    },
+    {
+      "epoch": 7.306051508927161,
+      "grad_norm": 1.0824054479599,
+      "learning_rate": 3.369916894677733e-05,
+      "loss": 1.5869,
+      "step": 34680
+    },
+    {
+      "epoch": 7.30815821351451,
+      "grad_norm": 1.1286417245864868,
+      "learning_rate": 3.3649629450007195e-05,
+      "loss": 1.6053,
+      "step": 34690
+    },
+    {
+      "epoch": 7.3102649181018595,
+      "grad_norm": 1.2368204593658447,
+      "learning_rate": 3.360011902608747e-05,
+      "loss": 1.5804,
+      "step": 34700
+    },
+    {
+      "epoch": 7.312371622689208,
+      "grad_norm": 1.3171190023422241,
+      "learning_rate": 3.355063769671232e-05,
+      "loss": 1.5864,
+      "step": 34710
+    },
+    {
+      "epoch": 7.314478327276557,
+      "grad_norm": 1.1727089881896973,
+      "learning_rate": 3.3501185483562994e-05,
+      "loss": 1.6236,
+      "step": 34720
+    },
+    {
+      "epoch": 7.316585031863907,
+      "grad_norm": 1.1134991645812988,
+      "learning_rate": 3.345176240830803e-05,
+      "loss": 1.5589,
+      "step": 34730
+    },
+    {
+      "epoch": 7.318691736451256,
+      "grad_norm": 1.2899798154830933,
+      "learning_rate": 3.340236849260324e-05,
+      "loss": 1.591,
+      "step": 34740
+    },
+    {
+      "epoch": 7.3207984410386056,
+      "grad_norm": 1.2112675905227661,
+      "learning_rate": 3.335300375809173e-05,
+      "loss": 1.6022,
+      "step": 34750
+    },
+    {
+      "epoch": 7.322905145625954,
+      "grad_norm": 1.1530252695083618,
+      "learning_rate": 3.330366822640356e-05,
+      "loss": 1.5894,
+      "step": 34760
+    },
+    {
+      "epoch": 7.325011850213304,
+      "grad_norm": 1.0778065919876099,
+      "learning_rate": 3.325436191915628e-05,
+      "loss": 1.6197,
+      "step": 34770
+    },
+    {
+      "epoch": 7.327118554800653,
+      "grad_norm": 1.2357008457183838,
+      "learning_rate": 3.320508485795445e-05,
+      "loss": 1.5383,
+      "step": 34780
+    },
+    {
+      "epoch": 7.329225259388003,
+      "grad_norm": 1.2149114608764648,
+      "learning_rate": 3.315583706438994e-05,
+      "loss": 1.5355,
+      "step": 34790
+    },
+    {
+      "epoch": 7.331331963975352,
+      "grad_norm": 1.1368036270141602,
+      "learning_rate": 3.31066185600417e-05,
+      "loss": 1.581,
+      "step": 34800
+    },
+    {
+      "epoch": 7.3334386685627,
+      "grad_norm": 1.1959999799728394,
+      "learning_rate": 3.305742936647586e-05,
+      "loss": 1.6053,
+      "step": 34810
+    },
+    {
+      "epoch": 7.33554537315005,
+      "grad_norm": 1.2553191184997559,
+      "learning_rate": 3.300826950524575e-05,
+      "loss": 1.5595,
+      "step": 34820
+    },
+    {
+      "epoch": 7.337652077737399,
+      "grad_norm": 1.2087421417236328,
+      "learning_rate": 3.2959138997891905e-05,
+      "loss": 1.5864,
+      "step": 34830
+    },
+    {
+      "epoch": 7.339758782324749,
+      "grad_norm": 1.136260986328125,
+      "learning_rate": 3.291003786594178e-05,
+      "loss": 1.5502,
+      "step": 34840
+    },
+    {
+      "epoch": 7.341865486912098,
+      "grad_norm": 1.2127354145050049,
+      "learning_rate": 3.28609661309102e-05,
+      "loss": 1.5129,
+      "step": 34850
+    },
+    {
+      "epoch": 7.343972191499447,
+      "grad_norm": 1.2038772106170654,
+      "learning_rate": 3.281192381429894e-05,
+      "loss": 1.5622,
+      "step": 34860
+    },
+    {
+      "epoch": 7.346078896086796,
+      "grad_norm": 1.3606103658676147,
+      "learning_rate": 3.276291093759701e-05,
+      "loss": 1.5531,
+      "step": 34870
+    },
+    {
+      "epoch": 7.348185600674145,
+      "grad_norm": 1.1576495170593262,
+      "learning_rate": 3.2713927522280453e-05,
+      "loss": 1.6216,
+      "step": 34880
+    },
+    {
+      "epoch": 7.350292305261495,
+      "grad_norm": 1.1671663522720337,
+      "learning_rate": 3.2664973589812364e-05,
+      "loss": 1.5243,
+      "step": 34890
+    },
+    {
+      "epoch": 7.352399009848844,
+      "grad_norm": 1.3078503608703613,
+      "learning_rate": 3.2616049161643005e-05,
+      "loss": 1.543,
+      "step": 34900
+    },
+    {
+      "epoch": 7.354505714436193,
+      "grad_norm": 1.1224507093429565,
+      "learning_rate": 3.256715425920969e-05,
+      "loss": 1.5786,
+      "step": 34910
+    },
+    {
+      "epoch": 7.356612419023542,
+      "grad_norm": 1.191428780555725,
+      "learning_rate": 3.251828890393677e-05,
+      "loss": 1.5707,
+      "step": 34920
+    },
+    {
+      "epoch": 7.358719123610892,
+      "grad_norm": 1.117315411567688,
+      "learning_rate": 3.246945311723564e-05,
+      "loss": 1.5758,
+      "step": 34930
+    },
+    {
+      "epoch": 7.360825828198241,
+      "grad_norm": 1.1541494131088257,
+      "learning_rate": 3.2420646920504726e-05,
+      "loss": 1.6113,
+      "step": 34940
+    },
+    {
+      "epoch": 7.362932532785591,
+      "grad_norm": 1.2011319398880005,
+      "learning_rate": 3.237187033512956e-05,
+      "loss": 1.5299,
+      "step": 34950
+    },
+    {
+      "epoch": 7.365039237372939,
+      "grad_norm": 1.2775424718856812,
+      "learning_rate": 3.232312338248271e-05,
+      "loss": 1.598,
+      "step": 34960
+    },
+    {
+      "epoch": 7.367145941960288,
+      "grad_norm": 1.1767414808273315,
+      "learning_rate": 3.2274406083923605e-05,
+      "loss": 1.5929,
+      "step": 34970
+    },
+    {
+      "epoch": 7.369252646547638,
+      "grad_norm": 1.212234377861023,
+      "learning_rate": 3.222571846079881e-05,
+      "loss": 1.5344,
+      "step": 34980
+    },
+    {
+      "epoch": 7.371359351134987,
+      "grad_norm": 1.1913750171661377,
+      "learning_rate": 3.217706053444193e-05,
+      "loss": 1.5611,
+      "step": 34990
+    },
+    {
+      "epoch": 7.373466055722337,
+      "grad_norm": 1.183814287185669,
+      "learning_rate": 3.212843232617343e-05,
+      "loss": 1.5926,
+      "step": 35000
+    },
+    {
+      "epoch": 7.3755727603096854,
+      "grad_norm": 1.2367053031921387,
+      "learning_rate": 3.207983385730081e-05,
+      "loss": 1.5913,
+      "step": 35010
+    },
+    {
+      "epoch": 7.377679464897035,
+      "grad_norm": 1.2764780521392822,
+      "learning_rate": 3.203126514911854e-05,
+      "loss": 1.5774,
+      "step": 35020
+    },
+    {
+      "epoch": 7.379786169484384,
+      "grad_norm": 1.2614119052886963,
+      "learning_rate": 3.198272622290804e-05,
+      "loss": 1.5315,
+      "step": 35030
+    },
+    {
+      "epoch": 7.381892874071733,
+      "grad_norm": 1.1370763778686523,
+      "learning_rate": 3.193421709993779e-05,
+      "loss": 1.5631,
+      "step": 35040
+    },
+    {
+      "epoch": 7.383999578659083,
+      "grad_norm": 1.2041585445404053,
+      "learning_rate": 3.188573780146298e-05,
+      "loss": 1.5932,
+      "step": 35050
+    },
+    {
+      "epoch": 7.3861062832464315,
+      "grad_norm": 1.2313627004623413,
+      "learning_rate": 3.1837288348725905e-05,
+      "loss": 1.6218,
+      "step": 35060
+    },
+    {
+      "epoch": 7.388212987833781,
+      "grad_norm": 1.1100727319717407,
+      "learning_rate": 3.178886876295578e-05,
+      "loss": 1.5867,
+      "step": 35070
+    },
+    {
+      "epoch": 7.39031969242113,
+      "grad_norm": 1.1173222064971924,
+      "learning_rate": 3.1740479065368665e-05,
+      "loss": 1.5747,
+      "step": 35080
+    },
+    {
+      "epoch": 7.39242639700848,
+      "grad_norm": 1.1595535278320312,
+      "learning_rate": 3.1692119277167564e-05,
+      "loss": 1.5614,
+      "step": 35090
+    },
+    {
+      "epoch": 7.394533101595829,
+      "grad_norm": 1.1855976581573486,
+      "learning_rate": 3.1643789419542324e-05,
+      "loss": 1.5725,
+      "step": 35100
+    },
+    {
+      "epoch": 7.3966398061831775,
+      "grad_norm": 1.3628852367401123,
+      "learning_rate": 3.159548951366975e-05,
+      "loss": 1.5792,
+      "step": 35110
+    },
+    {
+      "epoch": 7.398746510770527,
+      "grad_norm": 1.256475806236267,
+      "learning_rate": 3.154721958071356e-05,
+      "loss": 1.5672,
+      "step": 35120
+    },
+    {
+      "epoch": 7.400853215357876,
+      "grad_norm": 1.1298019886016846,
+      "learning_rate": 3.149897964182413e-05,
+      "loss": 1.5738,
+      "step": 35130
+    },
+    {
+      "epoch": 7.402959919945226,
+      "grad_norm": 1.2207309007644653,
+      "learning_rate": 3.145076971813891e-05,
+      "loss": 1.571,
+      "step": 35140
+    },
+    {
+      "epoch": 7.405066624532575,
+      "grad_norm": 1.2245393991470337,
+      "learning_rate": 3.140258983078214e-05,
+      "loss": 1.5812,
+      "step": 35150
+    },
+    {
+      "epoch": 7.407173329119924,
+      "grad_norm": 1.2725415229797363,
+      "learning_rate": 3.135444000086485e-05,
+      "loss": 1.6163,
+      "step": 35160
+    },
+    {
+      "epoch": 7.409280033707273,
+      "grad_norm": 1.2918919324874878,
+      "learning_rate": 3.130632024948491e-05,
+      "loss": 1.5951,
+      "step": 35170
+    },
+    {
+      "epoch": 7.411386738294623,
+      "grad_norm": 1.187846302986145,
+      "learning_rate": 3.1258230597727075e-05,
+      "loss": 1.5676,
+      "step": 35180
+    },
+    {
+      "epoch": 7.413493442881972,
+      "grad_norm": 1.155684232711792,
+      "learning_rate": 3.121017106666283e-05,
+      "loss": 1.5967,
+      "step": 35190
+    },
+    {
+      "epoch": 7.415600147469321,
+      "grad_norm": 1.173157811164856,
+      "learning_rate": 3.116214167735053e-05,
+      "loss": 1.5996,
+      "step": 35200
+    },
+    {
+      "epoch": 7.4177068520566705,
+      "grad_norm": 1.1492563486099243,
+      "learning_rate": 3.1114142450835294e-05,
+      "loss": 1.5971,
+      "step": 35210
+    },
+    {
+      "epoch": 7.419813556644019,
+      "grad_norm": 1.1434016227722168,
+      "learning_rate": 3.1066173408148955e-05,
+      "loss": 1.5462,
+      "step": 35220
+    },
+    {
+      "epoch": 7.421920261231369,
+      "grad_norm": 1.1040714979171753,
+      "learning_rate": 3.101823457031028e-05,
+      "loss": 1.5812,
+      "step": 35230
+    },
+    {
+      "epoch": 7.424026965818718,
+      "grad_norm": 1.2135083675384521,
+      "learning_rate": 3.097032595832462e-05,
+      "loss": 1.6262,
+      "step": 35240
+    },
+    {
+      "epoch": 7.426133670406068,
+      "grad_norm": 1.2643121480941772,
+      "learning_rate": 3.092244759318424e-05,
+      "loss": 1.6215,
+      "step": 35250
+    },
+    {
+      "epoch": 7.4282403749934165,
+      "grad_norm": 1.105613112449646,
+      "learning_rate": 3.087459949586807e-05,
+      "loss": 1.6266,
+      "step": 35260
+    },
+    {
+      "epoch": 7.430347079580766,
+      "grad_norm": 1.2127161026000977,
+      "learning_rate": 3.082678168734175e-05,
+      "loss": 1.5836,
+      "step": 35270
+    },
+    {
+      "epoch": 7.432453784168115,
+      "grad_norm": 1.2094069719314575,
+      "learning_rate": 3.077899418855772e-05,
+      "loss": 1.6044,
+      "step": 35280
+    },
+    {
+      "epoch": 7.434560488755464,
+      "grad_norm": 1.115659236907959,
+      "learning_rate": 3.07312370204551e-05,
+      "loss": 1.5297,
+      "step": 35290
+    },
+    {
+      "epoch": 7.436667193342814,
+      "grad_norm": 1.180795669555664,
+      "learning_rate": 3.068351020395971e-05,
+      "loss": 1.5198,
+      "step": 35300
+    },
+    {
+      "epoch": 7.4387738979301625,
+      "grad_norm": 1.097103238105774,
+      "learning_rate": 3.063581375998412e-05,
+      "loss": 1.6051,
+      "step": 35310
+    },
+    {
+      "epoch": 7.440880602517512,
+      "grad_norm": 1.161124587059021,
+      "learning_rate": 3.0588147709427506e-05,
+      "loss": 1.5481,
+      "step": 35320
+    },
+    {
+      "epoch": 7.442987307104861,
+      "grad_norm": 1.1073428392410278,
+      "learning_rate": 3.054051207317585e-05,
+      "loss": 1.5336,
+      "step": 35330
+    },
+    {
+      "epoch": 7.445094011692211,
+      "grad_norm": 1.1355111598968506,
+      "learning_rate": 3.04929068721017e-05,
+      "loss": 1.5453,
+      "step": 35340
+    },
+    {
+      "epoch": 7.44720071627956,
+      "grad_norm": 1.302330732345581,
+      "learning_rate": 3.0445332127064275e-05,
+      "loss": 1.5737,
+      "step": 35350
+    },
+    {
+      "epoch": 7.449307420866909,
+      "grad_norm": 1.3458727598190308,
+      "learning_rate": 3.0397787858909542e-05,
+      "loss": 1.5543,
+      "step": 35360
+    },
+    {
+      "epoch": 7.451414125454258,
+      "grad_norm": 1.2425020933151245,
+      "learning_rate": 3.0350274088470022e-05,
+      "loss": 1.5909,
+      "step": 35370
+    },
+    {
+      "epoch": 7.453520830041607,
+      "grad_norm": 1.2361297607421875,
+      "learning_rate": 3.0302790836564853e-05,
+      "loss": 1.6337,
+      "step": 35380
+    },
+    {
+      "epoch": 7.455627534628957,
+      "grad_norm": 1.1364221572875977,
+      "learning_rate": 3.0255338123999934e-05,
+      "loss": 1.5521,
+      "step": 35390
+    },
+    {
+      "epoch": 7.457734239216306,
+      "grad_norm": 1.3635613918304443,
+      "learning_rate": 3.0207915971567624e-05,
+      "loss": 1.6241,
+      "step": 35400
+    },
+    {
+      "epoch": 7.4598409438036555,
+      "grad_norm": 1.1741420030593872,
+      "learning_rate": 3.016052440004703e-05,
+      "loss": 1.63,
+      "step": 35410
+    },
+    {
+      "epoch": 7.461947648391004,
+      "grad_norm": 1.2724038362503052,
+      "learning_rate": 3.0113163430203772e-05,
+      "loss": 1.5863,
+      "step": 35420
+    },
+    {
+      "epoch": 7.464054352978353,
+      "grad_norm": 1.2235639095306396,
+      "learning_rate": 3.006583308279003e-05,
+      "loss": 1.5091,
+      "step": 35430
+    },
+    {
+      "epoch": 7.466161057565703,
+      "grad_norm": 1.1760095357894897,
+      "learning_rate": 3.001853337854471e-05,
+      "loss": 1.5496,
+      "step": 35440
+    },
+    {
+      "epoch": 7.468267762153052,
+      "grad_norm": 1.1114704608917236,
+      "learning_rate": 2.9971264338193138e-05,
+      "loss": 1.5444,
+      "step": 35450
+    },
+    {
+      "epoch": 7.4703744667404015,
+      "grad_norm": 1.2096513509750366,
+      "learning_rate": 2.992402598244727e-05,
+      "loss": 1.5935,
+      "step": 35460
+    },
+    {
+      "epoch": 7.47248117132775,
+      "grad_norm": 1.2605581283569336,
+      "learning_rate": 2.987681833200565e-05,
+      "loss": 1.5468,
+      "step": 35470
+    },
+    {
+      "epoch": 7.4745878759151,
+      "grad_norm": 1.1911935806274414,
+      "learning_rate": 2.9829641407553276e-05,
+      "loss": 1.6422,
+      "step": 35480
+    },
+    {
+      "epoch": 7.476694580502449,
+      "grad_norm": 1.1851458549499512,
+      "learning_rate": 2.9782495229761808e-05,
+      "loss": 1.571,
+      "step": 35490
+    },
+    {
+      "epoch": 7.478801285089799,
+      "grad_norm": 1.1616430282592773,
+      "learning_rate": 2.973537981928932e-05,
+      "loss": 1.5879,
+      "step": 35500
+    },
+    {
+      "epoch": 7.4809079896771475,
+      "grad_norm": 1.434672474861145,
+      "learning_rate": 2.9688295196780436e-05,
+      "loss": 1.5506,
+      "step": 35510
+    },
+    {
+      "epoch": 7.483014694264496,
+      "grad_norm": 1.253493309020996,
+      "learning_rate": 2.9641241382866348e-05,
+      "loss": 1.5878,
+      "step": 35520
+    },
+    {
+      "epoch": 7.485121398851846,
+      "grad_norm": 1.2330995798110962,
+      "learning_rate": 2.959421839816464e-05,
+      "loss": 1.5484,
+      "step": 35530
+    },
+    {
+      "epoch": 7.487228103439195,
+      "grad_norm": 1.020727515220642,
+      "learning_rate": 2.9547226263279526e-05,
+      "loss": 1.5971,
+      "step": 35540
+    },
+    {
+      "epoch": 7.489334808026545,
+      "grad_norm": 1.3436671495437622,
+      "learning_rate": 2.9500264998801584e-05,
+      "loss": 1.5988,
+      "step": 35550
+    },
+    {
+      "epoch": 7.491441512613894,
+      "grad_norm": 1.1431214809417725,
+      "learning_rate": 2.945333462530788e-05,
+      "loss": 1.5691,
+      "step": 35560
+    },
+    {
+      "epoch": 7.493548217201243,
+      "grad_norm": 1.2193323373794556,
+      "learning_rate": 2.9406435163362033e-05,
+      "loss": 1.5398,
+      "step": 35570
+    },
+    {
+      "epoch": 7.495654921788592,
+      "grad_norm": 1.1207623481750488,
+      "learning_rate": 2.9359566633514037e-05,
+      "loss": 1.5464,
+      "step": 35580
+    },
+    {
+      "epoch": 7.497761626375942,
+      "grad_norm": 1.1370456218719482,
+      "learning_rate": 2.9312729056300302e-05,
+      "loss": 1.5847,
+      "step": 35590
+    },
+    {
+      "epoch": 7.499868330963291,
+      "grad_norm": 1.2836565971374512,
+      "learning_rate": 2.926592245224381e-05,
+      "loss": 1.5317,
+      "step": 35600
+    },
+    {
+      "epoch": 7.50197503555064,
+      "grad_norm": 1.125777244567871,
+      "learning_rate": 2.9219146841853807e-05,
+      "loss": 1.5978,
+      "step": 35610
+    },
+    {
+      "epoch": 7.504081740137989,
+      "grad_norm": 1.1385287046432495,
+      "learning_rate": 2.9172402245626108e-05,
+      "loss": 1.5838,
+      "step": 35620
+    },
+    {
+      "epoch": 7.506188444725338,
+      "grad_norm": 1.2350115776062012,
+      "learning_rate": 2.912568868404284e-05,
+      "loss": 1.5388,
+      "step": 35630
+    },
+    {
+      "epoch": 7.508295149312688,
+      "grad_norm": 1.2484310865402222,
+      "learning_rate": 2.907900617757252e-05,
+      "loss": 1.5478,
+      "step": 35640
+    },
+    {
+      "epoch": 7.510401853900037,
+      "grad_norm": 1.2130504846572876,
+      "learning_rate": 2.9032354746670176e-05,
+      "loss": 1.5439,
+      "step": 35650
+    },
+    {
+      "epoch": 7.5125085584873865,
+      "grad_norm": 1.2337069511413574,
+      "learning_rate": 2.8985734411777097e-05,
+      "loss": 1.5881,
+      "step": 35660
+    },
+    {
+      "epoch": 7.514615263074735,
+      "grad_norm": 1.1304142475128174,
+      "learning_rate": 2.893914519332097e-05,
+      "loss": 1.5423,
+      "step": 35670
+    },
+    {
+      "epoch": 7.516721967662084,
+      "grad_norm": 1.271000623703003,
+      "learning_rate": 2.8892587111715917e-05,
+      "loss": 1.628,
+      "step": 35680
+    },
+    {
+      "epoch": 7.518828672249434,
+      "grad_norm": 1.1860854625701904,
+      "learning_rate": 2.8846060187362335e-05,
+      "loss": 1.5807,
+      "step": 35690
+    },
+    {
+      "epoch": 7.520935376836783,
+      "grad_norm": 1.1863863468170166,
+      "learning_rate": 2.879956444064703e-05,
+      "loss": 1.5932,
+      "step": 35700
+    },
+    {
+      "epoch": 7.5230420814241326,
+      "grad_norm": 1.2513294219970703,
+      "learning_rate": 2.8753099891943113e-05,
+      "loss": 1.5891,
+      "step": 35710
+    },
+    {
+      "epoch": 7.525148786011481,
+      "grad_norm": 1.1214491128921509,
+      "learning_rate": 2.8706666561609998e-05,
+      "loss": 1.5867,
+      "step": 35720
+    },
+    {
+      "epoch": 7.527255490598831,
+      "grad_norm": 1.2936240434646606,
+      "learning_rate": 2.8660264469993502e-05,
+      "loss": 1.5993,
+      "step": 35730
+    },
+    {
+      "epoch": 7.52936219518618,
+      "grad_norm": 1.1555366516113281,
+      "learning_rate": 2.8613893637425647e-05,
+      "loss": 1.575,
+      "step": 35740
+    },
+    {
+      "epoch": 7.531468899773529,
+      "grad_norm": 1.1356788873672485,
+      "learning_rate": 2.856755408422489e-05,
+      "loss": 1.5665,
+      "step": 35750
+    },
+    {
+      "epoch": 7.533575604360879,
+      "grad_norm": 1.0721943378448486,
+      "learning_rate": 2.8521245830695864e-05,
+      "loss": 1.565,
+      "step": 35760
+    },
+    {
+      "epoch": 7.535682308948227,
+      "grad_norm": 1.1919306516647339,
+      "learning_rate": 2.847496889712952e-05,
+      "loss": 1.5273,
+      "step": 35770
+    },
+    {
+      "epoch": 7.537789013535577,
+      "grad_norm": 1.2383922338485718,
+      "learning_rate": 2.842872330380314e-05,
+      "loss": 1.5732,
+      "step": 35780
+    },
+    {
+      "epoch": 7.539895718122926,
+      "grad_norm": 1.244070291519165,
+      "learning_rate": 2.83825090709802e-05,
+      "loss": 1.6233,
+      "step": 35790
+    },
+    {
+      "epoch": 7.542002422710276,
+      "grad_norm": 1.201996922492981,
+      "learning_rate": 2.8336326218910457e-05,
+      "loss": 1.5927,
+      "step": 35800
+    },
+    {
+      "epoch": 7.544109127297625,
+      "grad_norm": 1.1840076446533203,
+      "learning_rate": 2.829017476782997e-05,
+      "loss": 1.5434,
+      "step": 35810
+    },
+    {
+      "epoch": 7.5462158318849735,
+      "grad_norm": 1.3222793340682983,
+      "learning_rate": 2.8244054737960935e-05,
+      "loss": 1.5198,
+      "step": 35820
+    },
+    {
+      "epoch": 7.548322536472323,
+      "grad_norm": 1.1209752559661865,
+      "learning_rate": 2.819796614951191e-05,
+      "loss": 1.5988,
+      "step": 35830
+    },
+    {
+      "epoch": 7.550429241059672,
+      "grad_norm": 1.3570189476013184,
+      "learning_rate": 2.815190902267757e-05,
+      "loss": 1.5611,
+      "step": 35840
+    },
+    {
+      "epoch": 7.552535945647022,
+      "grad_norm": 1.2108376026153564,
+      "learning_rate": 2.810588337763881e-05,
+      "loss": 1.5437,
+      "step": 35850
+    },
+    {
+      "epoch": 7.554642650234371,
+      "grad_norm": 1.14738929271698,
+      "learning_rate": 2.805988923456283e-05,
+      "loss": 1.5708,
+      "step": 35860
+    },
+    {
+      "epoch": 7.55674935482172,
+      "grad_norm": 1.3338896036148071,
+      "learning_rate": 2.8013926613602936e-05,
+      "loss": 1.5849,
+      "step": 35870
+    },
+    {
+      "epoch": 7.558856059409069,
+      "grad_norm": 1.2129565477371216,
+      "learning_rate": 2.7967995534898596e-05,
+      "loss": 1.5874,
+      "step": 35880
+    },
+    {
+      "epoch": 7.560962763996419,
+      "grad_norm": 1.1787198781967163,
+      "learning_rate": 2.7922096018575572e-05,
+      "loss": 1.5445,
+      "step": 35890
+    },
+    {
+      "epoch": 7.563069468583768,
+      "grad_norm": 1.1483491659164429,
+      "learning_rate": 2.787622808474567e-05,
+      "loss": 1.5665,
+      "step": 35900
+    },
+    {
+      "epoch": 7.565176173171118,
+      "grad_norm": 1.1970431804656982,
+      "learning_rate": 2.783039175350699e-05,
+      "loss": 1.6202,
+      "step": 35910
+    },
+    {
+      "epoch": 7.567282877758466,
+      "grad_norm": 1.138543963432312,
+      "learning_rate": 2.7784587044943666e-05,
+      "loss": 1.6148,
+      "step": 35920
+    },
+    {
+      "epoch": 7.569389582345815,
+      "grad_norm": 1.1422301530838013,
+      "learning_rate": 2.7738813979126e-05,
+      "loss": 1.5764,
+      "step": 35930
+    },
+    {
+      "epoch": 7.571496286933165,
+      "grad_norm": 1.1119393110275269,
+      "learning_rate": 2.7693072576110514e-05,
+      "loss": 1.6223,
+      "step": 35940
+    },
+    {
+      "epoch": 7.573602991520514,
+      "grad_norm": 1.2076855897903442,
+      "learning_rate": 2.764736285593975e-05,
+      "loss": 1.5864,
+      "step": 35950
+    },
+    {
+      "epoch": 7.575709696107864,
+      "grad_norm": 1.2735531330108643,
+      "learning_rate": 2.7601684838642405e-05,
+      "loss": 1.6377,
+      "step": 35960
+    },
+    {
+      "epoch": 7.5778164006952125,
+      "grad_norm": 1.0892045497894287,
+      "learning_rate": 2.755603854423332e-05,
+      "loss": 1.5968,
+      "step": 35970
+    },
+    {
+      "epoch": 7.579923105282562,
+      "grad_norm": 1.2122892141342163,
+      "learning_rate": 2.7510423992713374e-05,
+      "loss": 1.6211,
+      "step": 35980
+    },
+    {
+      "epoch": 7.582029809869911,
+      "grad_norm": 1.1663323640823364,
+      "learning_rate": 2.7464841204069614e-05,
+      "loss": 1.5842,
+      "step": 35990
+    },
+    {
+      "epoch": 7.58413651445726,
+      "grad_norm": 1.1494828462600708,
+      "learning_rate": 2.7419290198275095e-05,
+      "loss": 1.6125,
+      "step": 36000
+    },
+    {
+      "epoch": 7.58624321904461,
+      "grad_norm": 1.1589758396148682,
+      "learning_rate": 2.737377099528895e-05,
+      "loss": 1.6178,
+      "step": 36010
+    },
+    {
+      "epoch": 7.5883499236319585,
+      "grad_norm": 1.177195429801941,
+      "learning_rate": 2.7328283615056472e-05,
+      "loss": 1.535,
+      "step": 36020
+    },
+    {
+      "epoch": 7.590456628219308,
+      "grad_norm": 1.2202266454696655,
+      "learning_rate": 2.728282807750886e-05,
+      "loss": 1.5805,
+      "step": 36030
+    },
+    {
+      "epoch": 7.592563332806657,
+      "grad_norm": 1.1336926221847534,
+      "learning_rate": 2.7237404402563517e-05,
+      "loss": 1.5508,
+      "step": 36040
+    },
+    {
+      "epoch": 7.594670037394007,
+      "grad_norm": 1.181666612625122,
+      "learning_rate": 2.7192012610123774e-05,
+      "loss": 1.6056,
+      "step": 36050
+    },
+    {
+      "epoch": 7.596776741981356,
+      "grad_norm": 1.1373564004898071,
+      "learning_rate": 2.7146652720079003e-05,
+      "loss": 1.5999,
+      "step": 36060
+    },
+    {
+      "epoch": 7.5988834465687045,
+      "grad_norm": 1.1596566438674927,
+      "learning_rate": 2.7101324752304635e-05,
+      "loss": 1.5928,
+      "step": 36070
+    },
+    {
+      "epoch": 7.600990151156054,
+      "grad_norm": 1.2005223035812378,
+      "learning_rate": 2.7056028726662176e-05,
+      "loss": 1.5748,
+      "step": 36080
+    },
+    {
+      "epoch": 7.603096855743403,
+      "grad_norm": 1.1738914251327515,
+      "learning_rate": 2.7010764662998933e-05,
+      "loss": 1.6077,
+      "step": 36090
+    },
+    {
+      "epoch": 7.605203560330753,
+      "grad_norm": 1.196463704109192,
+      "learning_rate": 2.69655325811484e-05,
+      "loss": 1.5577,
+      "step": 36100
+    },
+    {
+      "epoch": 7.607310264918102,
+      "grad_norm": 1.3594914674758911,
+      "learning_rate": 2.6920332500930025e-05,
+      "loss": 1.5649,
+      "step": 36110
+    },
+    {
+      "epoch": 7.609416969505451,
+      "grad_norm": 1.21880042552948,
+      "learning_rate": 2.6875164442149147e-05,
+      "loss": 1.5571,
+      "step": 36120
+    },
+    {
+      "epoch": 7.6115236740928,
+      "grad_norm": 1.2229676246643066,
+      "learning_rate": 2.6830028424597165e-05,
+      "loss": 1.5707,
+      "step": 36130
+    },
+    {
+      "epoch": 7.613630378680149,
+      "grad_norm": 1.1788535118103027,
+      "learning_rate": 2.6784924468051342e-05,
+      "loss": 1.5671,
+      "step": 36140
+    },
+    {
+      "epoch": 7.615737083267499,
+      "grad_norm": 1.1653674840927124,
+      "learning_rate": 2.6739852592274995e-05,
+      "loss": 1.5574,
+      "step": 36150
+    },
+    {
+      "epoch": 7.617843787854848,
+      "grad_norm": 1.0854195356369019,
+      "learning_rate": 2.669481281701739e-05,
+      "loss": 1.5698,
+      "step": 36160
+    },
+    {
+      "epoch": 7.6199504924421975,
+      "grad_norm": 1.1866564750671387,
+      "learning_rate": 2.6649805162013585e-05,
+      "loss": 1.554,
+      "step": 36170
+    },
+    {
+      "epoch": 7.622057197029546,
+      "grad_norm": 1.1668610572814941,
+      "learning_rate": 2.6604829646984686e-05,
+      "loss": 1.5912,
+      "step": 36180
+    },
+    {
+      "epoch": 7.624163901616896,
+      "grad_norm": 1.1470420360565186,
+      "learning_rate": 2.6559886291637748e-05,
+      "loss": 1.584,
+      "step": 36190
+    },
+    {
+      "epoch": 7.626270606204245,
+      "grad_norm": 1.1592053174972534,
+      "learning_rate": 2.651497511566562e-05,
+      "loss": 1.5544,
+      "step": 36200
+    },
+    {
+      "epoch": 7.628377310791594,
+      "grad_norm": 1.1775624752044678,
+      "learning_rate": 2.6470096138747126e-05,
+      "loss": 1.5315,
+      "step": 36210
+    },
+    {
+      "epoch": 7.6304840153789435,
+      "grad_norm": 1.2434234619140625,
+      "learning_rate": 2.6425249380546912e-05,
+      "loss": 1.569,
+      "step": 36220
+    },
+    {
+      "epoch": 7.632590719966292,
+      "grad_norm": 1.1641994714736938,
+      "learning_rate": 2.6380434860715598e-05,
+      "loss": 1.6168,
+      "step": 36230
+    },
+    {
+      "epoch": 7.634697424553642,
+      "grad_norm": 1.2335554361343384,
+      "learning_rate": 2.6335652598889683e-05,
+      "loss": 1.5694,
+      "step": 36240
+    },
+    {
+      "epoch": 7.636804129140991,
+      "grad_norm": 1.1290301084518433,
+      "learning_rate": 2.629090261469138e-05,
+      "loss": 1.5447,
+      "step": 36250
+    },
+    {
+      "epoch": 7.638910833728341,
+      "grad_norm": 1.1775555610656738,
+      "learning_rate": 2.624618492772891e-05,
+      "loss": 1.6394,
+      "step": 36260
+    },
+    {
+      "epoch": 7.6410175383156895,
+      "grad_norm": 1.2943990230560303,
+      "learning_rate": 2.620149955759633e-05,
+      "loss": 1.6198,
+      "step": 36270
+    },
+    {
+      "epoch": 7.643124242903039,
+      "grad_norm": 1.2637001276016235,
+      "learning_rate": 2.615684652387348e-05,
+      "loss": 1.5505,
+      "step": 36280
+    },
+    {
+      "epoch": 7.645230947490388,
+      "grad_norm": 1.195572018623352,
+      "learning_rate": 2.6112225846126038e-05,
+      "loss": 1.5403,
+      "step": 36290
+    },
+    {
+      "epoch": 7.647337652077738,
+      "grad_norm": 1.2487926483154297,
+      "learning_rate": 2.60676375439055e-05,
+      "loss": 1.5534,
+      "step": 36300
+    },
+    {
+      "epoch": 7.649444356665087,
+      "grad_norm": 1.304735541343689,
+      "learning_rate": 2.6023081636749225e-05,
+      "loss": 1.5638,
+      "step": 36310
+    },
+    {
+      "epoch": 7.651551061252436,
+      "grad_norm": 1.1997491121292114,
+      "learning_rate": 2.5978558144180363e-05,
+      "loss": 1.612,
+      "step": 36320
+    },
+    {
+      "epoch": 7.653657765839785,
+      "grad_norm": 1.226371169090271,
+      "learning_rate": 2.5934067085707834e-05,
+      "loss": 1.541,
+      "step": 36330
+    },
+    {
+      "epoch": 7.655764470427134,
+      "grad_norm": 1.1558582782745361,
+      "learning_rate": 2.5889608480826322e-05,
+      "loss": 1.6072,
+      "step": 36340
+    },
+    {
+      "epoch": 7.657871175014484,
+      "grad_norm": 1.1929916143417358,
+      "learning_rate": 2.5845182349016384e-05,
+      "loss": 1.5986,
+      "step": 36350
+    },
+    {
+      "epoch": 7.659977879601833,
+      "grad_norm": 1.2270394563674927,
+      "learning_rate": 2.5800788709744227e-05,
+      "loss": 1.5288,
+      "step": 36360
+    },
+    {
+      "epoch": 7.6620845841891825,
+      "grad_norm": 1.158097267150879,
+      "learning_rate": 2.5756427582461996e-05,
+      "loss": 1.5654,
+      "step": 36370
+    },
+    {
+      "epoch": 7.664191288776531,
+      "grad_norm": 1.2660669088363647,
+      "learning_rate": 2.5712098986607326e-05,
+      "loss": 1.5996,
+      "step": 36380
+    },
+    {
+      "epoch": 7.66629799336388,
+      "grad_norm": 1.1697864532470703,
+      "learning_rate": 2.5667802941603834e-05,
+      "loss": 1.5389,
+      "step": 36390
+    },
+    {
+      "epoch": 7.66840469795123,
+      "grad_norm": 1.2377358675003052,
+      "learning_rate": 2.5623539466860813e-05,
+      "loss": 1.5828,
+      "step": 36400
+    },
+    {
+      "epoch": 7.670511402538579,
+      "grad_norm": 1.2005665302276611,
+      "learning_rate": 2.5579308581773232e-05,
+      "loss": 1.6106,
+      "step": 36410
+    },
+    {
+      "epoch": 7.6726181071259285,
+      "grad_norm": 1.2478543519973755,
+      "learning_rate": 2.5535110305721776e-05,
+      "loss": 1.5887,
+      "step": 36420
+    },
+    {
+      "epoch": 7.674724811713277,
+      "grad_norm": 1.2146416902542114,
+      "learning_rate": 2.5490944658072945e-05,
+      "loss": 1.5917,
+      "step": 36430
+    },
+    {
+      "epoch": 7.676831516300627,
+      "grad_norm": 1.2361540794372559,
+      "learning_rate": 2.5446811658178815e-05,
+      "loss": 1.6649,
+      "step": 36440
+    },
+    {
+      "epoch": 7.678938220887976,
+      "grad_norm": 1.107953429222107,
+      "learning_rate": 2.540271132537729e-05,
+      "loss": 1.6139,
+      "step": 36450
+    },
+    {
+      "epoch": 7.681044925475325,
+      "grad_norm": 1.2849297523498535,
+      "learning_rate": 2.5358643678991788e-05,
+      "loss": 1.562,
+      "step": 36460
+    },
+    {
+      "epoch": 7.6831516300626745,
+      "grad_norm": 1.2075804471969604,
+      "learning_rate": 2.5314608738331537e-05,
+      "loss": 1.5133,
+      "step": 36470
+    },
+    {
+      "epoch": 7.685258334650023,
+      "grad_norm": 1.1946347951889038,
+      "learning_rate": 2.5270606522691443e-05,
+      "loss": 1.6608,
+      "step": 36480
+    },
+    {
+      "epoch": 7.687365039237373,
+      "grad_norm": 1.1127058267593384,
+      "learning_rate": 2.5226637051351987e-05,
+      "loss": 1.5892,
+      "step": 36490
+    },
+    {
+      "epoch": 7.689471743824722,
+      "grad_norm": 1.2100321054458618,
+      "learning_rate": 2.5182700343579334e-05,
+      "loss": 1.4847,
+      "step": 36500
+    },
+    {
+      "epoch": 7.691578448412072,
+      "grad_norm": 1.1425443887710571,
+      "learning_rate": 2.5138796418625343e-05,
+      "loss": 1.5488,
+      "step": 36510
+    },
+    {
+      "epoch": 7.693685152999421,
+      "grad_norm": 1.0502097606658936,
+      "learning_rate": 2.5094925295727423e-05,
+      "loss": 1.5681,
+      "step": 36520
+    },
+    {
+      "epoch": 7.695791857586769,
+      "grad_norm": 1.2372586727142334,
+      "learning_rate": 2.5051086994108743e-05,
+      "loss": 1.5978,
+      "step": 36530
+    },
+    {
+      "epoch": 7.697898562174119,
+      "grad_norm": 1.2679331302642822,
+      "learning_rate": 2.500728153297788e-05,
+      "loss": 1.5652,
+      "step": 36540
+    },
+    {
+      "epoch": 7.700005266761468,
+      "grad_norm": 1.316920280456543,
+      "learning_rate": 2.49635089315292e-05,
+      "loss": 1.5933,
+      "step": 36550
+    },
+    {
+      "epoch": 7.702111971348818,
+      "grad_norm": 1.1949541568756104,
+      "learning_rate": 2.4919769208942655e-05,
+      "loss": 1.5761,
+      "step": 36560
+    },
+    {
+      "epoch": 7.704218675936167,
+      "grad_norm": 1.2706875801086426,
+      "learning_rate": 2.4876062384383714e-05,
+      "loss": 1.5492,
+      "step": 36570
+    },
+    {
+      "epoch": 7.706325380523516,
+      "grad_norm": 1.247161865234375,
+      "learning_rate": 2.4832388477003443e-05,
+      "loss": 1.5714,
+      "step": 36580
+    },
+    {
+      "epoch": 7.708432085110865,
+      "grad_norm": 1.1570487022399902,
+      "learning_rate": 2.478874750593856e-05,
+      "loss": 1.6159,
+      "step": 36590
+    },
+    {
+      "epoch": 7.710538789698215,
+      "grad_norm": 1.2316917181015015,
+      "learning_rate": 2.4745139490311254e-05,
+      "loss": 1.5746,
+      "step": 36600
+    },
+    {
+      "epoch": 7.712645494285564,
+      "grad_norm": 1.1851727962493896,
+      "learning_rate": 2.4701564449229374e-05,
+      "loss": 1.6075,
+      "step": 36610
+    },
+    {
+      "epoch": 7.7147521988729135,
+      "grad_norm": 1.2722426652908325,
+      "learning_rate": 2.465802240178624e-05,
+      "loss": 1.5341,
+      "step": 36620
+    },
+    {
+      "epoch": 7.716858903460262,
+      "grad_norm": 1.2204933166503906,
+      "learning_rate": 2.46145133670607e-05,
+      "loss": 1.5937,
+      "step": 36630
+    },
+    {
+      "epoch": 7.718965608047611,
+      "grad_norm": 1.2491124868392944,
+      "learning_rate": 2.4571037364117255e-05,
+      "loss": 1.5956,
+      "step": 36640
+    },
+    {
+      "epoch": 7.721072312634961,
+      "grad_norm": 1.1790046691894531,
+      "learning_rate": 2.4527594412005793e-05,
+      "loss": 1.6315,
+      "step": 36650
+    },
+    {
+      "epoch": 7.72317901722231,
+      "grad_norm": 1.1278231143951416,
+      "learning_rate": 2.4484184529761834e-05,
+      "loss": 1.57,
+      "step": 36660
+    },
+    {
+      "epoch": 7.72528572180966,
+      "grad_norm": 1.13362455368042,
+      "learning_rate": 2.4440807736406335e-05,
+      "loss": 1.5596,
+      "step": 36670
+    },
+    {
+      "epoch": 7.727392426397008,
+      "grad_norm": 1.2489476203918457,
+      "learning_rate": 2.439746405094575e-05,
+      "loss": 1.5568,
+      "step": 36680
+    },
+    {
+      "epoch": 7.729499130984358,
+      "grad_norm": 1.2021193504333496,
+      "learning_rate": 2.43541534923721e-05,
+      "loss": 1.5935,
+      "step": 36690
+    },
+    {
+      "epoch": 7.731605835571707,
+      "grad_norm": 1.2708065509796143,
+      "learning_rate": 2.4310876079662824e-05,
+      "loss": 1.527,
+      "step": 36700
+    },
+    {
+      "epoch": 7.733712540159056,
+      "grad_norm": 1.2618707418441772,
+      "learning_rate": 2.4267631831780824e-05,
+      "loss": 1.5606,
+      "step": 36710
+    },
+    {
+      "epoch": 7.735819244746406,
+      "grad_norm": 1.1836202144622803,
+      "learning_rate": 2.4224420767674562e-05,
+      "loss": 1.6017,
+      "step": 36720
+    },
+    {
+      "epoch": 7.737925949333754,
+      "grad_norm": 1.3585106134414673,
+      "learning_rate": 2.4181242906277833e-05,
+      "loss": 1.5522,
+      "step": 36730
+    },
+    {
+      "epoch": 7.740032653921104,
+      "grad_norm": 1.2638752460479736,
+      "learning_rate": 2.4138098266510033e-05,
+      "loss": 1.6293,
+      "step": 36740
+    },
+    {
+      "epoch": 7.742139358508453,
+      "grad_norm": 1.2558181285858154,
+      "learning_rate": 2.409498686727587e-05,
+      "loss": 1.5783,
+      "step": 36750
+    },
+    {
+      "epoch": 7.744246063095803,
+      "grad_norm": 1.243468999862671,
+      "learning_rate": 2.405190872746551e-05,
+      "loss": 1.554,
+      "step": 36760
+    },
+    {
+      "epoch": 7.746352767683152,
+      "grad_norm": 1.2900869846343994,
+      "learning_rate": 2.4008863865954635e-05,
+      "loss": 1.6041,
+      "step": 36770
+    },
+    {
+      "epoch": 7.7484594722705005,
+      "grad_norm": 1.216533899307251,
+      "learning_rate": 2.3965852301604254e-05,
+      "loss": 1.5808,
+      "step": 36780
+    },
+    {
+      "epoch": 7.75056617685785,
+      "grad_norm": 1.1279761791229248,
+      "learning_rate": 2.392287405326078e-05,
+      "loss": 1.5403,
+      "step": 36790
+    },
+    {
+      "epoch": 7.752672881445199,
+      "grad_norm": 1.2638577222824097,
+      "learning_rate": 2.387992913975613e-05,
+      "loss": 1.5787,
+      "step": 36800
+    },
+    {
+      "epoch": 7.754779586032549,
+      "grad_norm": 1.2621861696243286,
+      "learning_rate": 2.3837017579907472e-05,
+      "loss": 1.6167,
+      "step": 36810
+    },
+    {
+      "epoch": 7.756886290619898,
+      "grad_norm": 1.2948744297027588,
+      "learning_rate": 2.379413939251751e-05,
+      "loss": 1.6138,
+      "step": 36820
+    },
+    {
+      "epoch": 7.758992995207247,
+      "grad_norm": 1.2331513166427612,
+      "learning_rate": 2.375129459637422e-05,
+      "loss": 1.6155,
+      "step": 36830
+    },
+    {
+      "epoch": 7.761099699794596,
+      "grad_norm": 1.1364085674285889,
+      "learning_rate": 2.370848321025093e-05,
+      "loss": 1.5179,
+      "step": 36840
+    },
+    {
+      "epoch": 7.763206404381945,
+      "grad_norm": 1.110427737236023,
+      "learning_rate": 2.3665705252906443e-05,
+      "loss": 1.6095,
+      "step": 36850
+    },
+    {
+      "epoch": 7.765313108969295,
+      "grad_norm": 1.229262113571167,
+      "learning_rate": 2.3622960743084798e-05,
+      "loss": 1.5977,
+      "step": 36860
+    },
+    {
+      "epoch": 7.767419813556644,
+      "grad_norm": 1.1048126220703125,
+      "learning_rate": 2.3580249699515467e-05,
+      "loss": 1.5216,
+      "step": 36870
+    },
+    {
+      "epoch": 7.769526518143993,
+      "grad_norm": 1.1986953020095825,
+      "learning_rate": 2.353757214091321e-05,
+      "loss": 1.556,
+      "step": 36880
+    },
+    {
+      "epoch": 7.771633222731342,
+      "grad_norm": 1.2192281484603882,
+      "learning_rate": 2.3494928085978073e-05,
+      "loss": 1.5995,
+      "step": 36890
+    },
+    {
+      "epoch": 7.773739927318692,
+      "grad_norm": 1.184574842453003,
+      "learning_rate": 2.345231755339554e-05,
+      "loss": 1.5295,
+      "step": 36900
+    },
+    {
+      "epoch": 7.775846631906041,
+      "grad_norm": 1.2875125408172607,
+      "learning_rate": 2.3409740561836313e-05,
+      "loss": 1.5371,
+      "step": 36910
+    },
+    {
+      "epoch": 7.777953336493391,
+      "grad_norm": 1.20929753780365,
+      "learning_rate": 2.3367197129956376e-05,
+      "loss": 1.5668,
+      "step": 36920
+    },
+    {
+      "epoch": 7.7800600410807395,
+      "grad_norm": 1.1370899677276611,
+      "learning_rate": 2.332468727639713e-05,
+      "loss": 1.5638,
+      "step": 36930
+    },
+    {
+      "epoch": 7.782166745668089,
+      "grad_norm": 1.2598953247070312,
+      "learning_rate": 2.3282211019785127e-05,
+      "loss": 1.5746,
+      "step": 36940
+    },
+    {
+      "epoch": 7.784273450255438,
+      "grad_norm": 1.2978923320770264,
+      "learning_rate": 2.3239768378732307e-05,
+      "loss": 1.5687,
+      "step": 36950
+    },
+    {
+      "epoch": 7.786380154842787,
+      "grad_norm": 1.1977964639663696,
+      "learning_rate": 2.3197359371835802e-05,
+      "loss": 1.552,
+      "step": 36960
+    },
+    {
+      "epoch": 7.788486859430137,
+      "grad_norm": 1.1075884103775024,
+      "learning_rate": 2.3154984017678015e-05,
+      "loss": 1.569,
+      "step": 36970
+    },
+    {
+      "epoch": 7.7905935640174855,
+      "grad_norm": 1.2914838790893555,
+      "learning_rate": 2.3112642334826684e-05,
+      "loss": 1.5667,
+      "step": 36980
+    },
+    {
+      "epoch": 7.792700268604835,
+      "grad_norm": 1.172547698020935,
+      "learning_rate": 2.30703343418347e-05,
+      "loss": 1.5901,
+      "step": 36990
+    },
+    {
+      "epoch": 7.794806973192184,
+      "grad_norm": 1.1503307819366455,
+      "learning_rate": 2.3028060057240187e-05,
+      "loss": 1.5928,
+      "step": 37000
+    },
+    {
+      "epoch": 7.796913677779534,
+      "grad_norm": 1.278609275817871,
+      "learning_rate": 2.298581949956662e-05,
+      "loss": 1.5382,
+      "step": 37010
+    },
+    {
+      "epoch": 7.799020382366883,
+      "grad_norm": 1.2486605644226074,
+      "learning_rate": 2.2943612687322525e-05,
+      "loss": 1.5521,
+      "step": 37020
+    },
+    {
+      "epoch": 7.8011270869542315,
+      "grad_norm": 1.2521138191223145,
+      "learning_rate": 2.290143963900181e-05,
+      "loss": 1.5802,
+      "step": 37030
+    },
+    {
+      "epoch": 7.803233791541581,
+      "grad_norm": 1.1833715438842773,
+      "learning_rate": 2.285930037308347e-05,
+      "loss": 1.5383,
+      "step": 37040
+    },
+    {
+      "epoch": 7.80534049612893,
+      "grad_norm": 1.2181299924850464,
+      "learning_rate": 2.2817194908031712e-05,
+      "loss": 1.5618,
+      "step": 37050
+    },
+    {
+      "epoch": 7.80744720071628,
+      "grad_norm": 1.2383207082748413,
+      "learning_rate": 2.2775123262296005e-05,
+      "loss": 1.6328,
+      "step": 37060
+    },
+    {
+      "epoch": 7.809553905303629,
+      "grad_norm": 1.2487322092056274,
+      "learning_rate": 2.2733085454310942e-05,
+      "loss": 1.5988,
+      "step": 37070
+    },
+    {
+      "epoch": 7.811660609890978,
+      "grad_norm": 1.2942522764205933,
+      "learning_rate": 2.2691081502496246e-05,
+      "loss": 1.5875,
+      "step": 37080
+    },
+    {
+      "epoch": 7.813767314478327,
+      "grad_norm": 1.1791707277297974,
+      "learning_rate": 2.264911142525693e-05,
+      "loss": 1.6108,
+      "step": 37090
+    },
+    {
+      "epoch": 7.815874019065676,
+      "grad_norm": 1.1863888502120972,
+      "learning_rate": 2.2607175240983026e-05,
+      "loss": 1.6131,
+      "step": 37100
+    },
+    {
+      "epoch": 7.817980723653026,
+      "grad_norm": 1.1202061176300049,
+      "learning_rate": 2.2565272968049844e-05,
+      "loss": 1.5871,
+      "step": 37110
+    },
+    {
+      "epoch": 7.820087428240375,
+      "grad_norm": 1.1665284633636475,
+      "learning_rate": 2.2523404624817736e-05,
+      "loss": 1.6289,
+      "step": 37120
+    },
+    {
+      "epoch": 7.8221941328277245,
+      "grad_norm": 1.2799750566482544,
+      "learning_rate": 2.2481570229632197e-05,
+      "loss": 1.5694,
+      "step": 37130
+    },
+    {
+      "epoch": 7.824300837415073,
+      "grad_norm": 1.236337423324585,
+      "learning_rate": 2.243976980082394e-05,
+      "loss": 1.5345,
+      "step": 37140
+    },
+    {
+      "epoch": 7.826407542002423,
+      "grad_norm": 1.1478630304336548,
+      "learning_rate": 2.2398003356708654e-05,
+      "loss": 1.5972,
+      "step": 37150
+    },
+    {
+      "epoch": 7.828514246589772,
+      "grad_norm": 1.2333014011383057,
+      "learning_rate": 2.2356270915587274e-05,
+      "loss": 1.5873,
+      "step": 37160
+    },
+    {
+      "epoch": 7.830620951177121,
+      "grad_norm": 1.4515944719314575,
+      "learning_rate": 2.2314572495745746e-05,
+      "loss": 1.5801,
+      "step": 37170
+    },
+    {
+      "epoch": 7.8327276557644705,
+      "grad_norm": 1.263717532157898,
+      "learning_rate": 2.2272908115455104e-05,
+      "loss": 1.593,
+      "step": 37180
+    },
+    {
+      "epoch": 7.834834360351819,
+      "grad_norm": 1.1389946937561035,
+      "learning_rate": 2.2231277792971515e-05,
+      "loss": 1.6318,
+      "step": 37190
+    },
+    {
+      "epoch": 7.836941064939169,
+      "grad_norm": 1.3768473863601685,
+      "learning_rate": 2.218968154653629e-05,
+      "loss": 1.5971,
+      "step": 37200
+    },
+    {
+      "epoch": 7.839047769526518,
+      "grad_norm": 1.1650984287261963,
+      "learning_rate": 2.2148119394375577e-05,
+      "loss": 1.612,
+      "step": 37210
+    },
+    {
+      "epoch": 7.841154474113868,
+      "grad_norm": 1.6648105382919312,
+      "learning_rate": 2.2106591354700845e-05,
+      "loss": 1.5907,
+      "step": 37220
+    },
+    {
+      "epoch": 7.8432611787012165,
+      "grad_norm": 1.2210161685943604,
+      "learning_rate": 2.2065097445708437e-05,
+      "loss": 1.5393,
+      "step": 37230
+    },
+    {
+      "epoch": 7.845367883288565,
+      "grad_norm": 1.1931644678115845,
+      "learning_rate": 2.2023637685579856e-05,
+      "loss": 1.628,
+      "step": 37240
+    },
+    {
+      "epoch": 7.847474587875915,
+      "grad_norm": 1.2531075477600098,
+      "learning_rate": 2.198221209248158e-05,
+      "loss": 1.5395,
+      "step": 37250
+    },
+    {
+      "epoch": 7.849581292463264,
+      "grad_norm": 1.1961562633514404,
+      "learning_rate": 2.194082068456509e-05,
+      "loss": 1.5847,
+      "step": 37260
+    },
+    {
+      "epoch": 7.851687997050614,
+      "grad_norm": 1.3024967908859253,
+      "learning_rate": 2.1899463479966952e-05,
+      "loss": 1.5841,
+      "step": 37270
+    },
+    {
+      "epoch": 7.853794701637963,
+      "grad_norm": 1.9943350553512573,
+      "learning_rate": 2.1858140496808777e-05,
+      "loss": 1.6065,
+      "step": 37280
+    },
+    {
+      "epoch": 7.855901406225312,
+      "grad_norm": 1.188504934310913,
+      "learning_rate": 2.181685175319702e-05,
+      "loss": 1.5932,
+      "step": 37290
+    },
+    {
+      "epoch": 7.858008110812661,
+      "grad_norm": 1.0779452323913574,
+      "learning_rate": 2.1775597267223323e-05,
+      "loss": 1.5936,
+      "step": 37300
+    },
+    {
+      "epoch": 7.860114815400011,
+      "grad_norm": 1.3473560810089111,
+      "learning_rate": 2.1734377056964172e-05,
+      "loss": 1.5284,
+      "step": 37310
+    },
+    {
+      "epoch": 7.86222151998736,
+      "grad_norm": 1.1941745281219482,
+      "learning_rate": 2.169319114048114e-05,
+      "loss": 1.6274,
+      "step": 37320
+    },
+    {
+      "epoch": 7.8643282245747095,
+      "grad_norm": 1.1974396705627441,
+      "learning_rate": 2.1652039535820712e-05,
+      "loss": 1.533,
+      "step": 37330
+    },
+    {
+      "epoch": 7.866434929162058,
+      "grad_norm": 1.1820552349090576,
+      "learning_rate": 2.161092226101432e-05,
+      "loss": 1.5985,
+      "step": 37340
+    },
+    {
+      "epoch": 7.868541633749407,
+      "grad_norm": 1.3820282220840454,
+      "learning_rate": 2.1569839334078422e-05,
+      "loss": 1.5649,
+      "step": 37350
+    },
+    {
+      "epoch": 7.870648338336757,
+      "grad_norm": 1.108791708946228,
+      "learning_rate": 2.152879077301443e-05,
+      "loss": 1.5098,
+      "step": 37360
+    },
+    {
+      "epoch": 7.872755042924106,
+      "grad_norm": 1.1899380683898926,
+      "learning_rate": 2.1487776595808575e-05,
+      "loss": 1.5486,
+      "step": 37370
+    },
+    {
+      "epoch": 7.8748617475114555,
+      "grad_norm": 1.275118112564087,
+      "learning_rate": 2.1446796820432167e-05,
+      "loss": 1.5725,
+      "step": 37380
+    },
+    {
+      "epoch": 7.876968452098804,
+      "grad_norm": 1.3126546144485474,
+      "learning_rate": 2.140585146484133e-05,
+      "loss": 1.6219,
+      "step": 37390
+    },
+    {
+      "epoch": 7.879075156686154,
+      "grad_norm": 1.2064414024353027,
+      "learning_rate": 2.136494054697722e-05,
+      "loss": 1.5985,
+      "step": 37400
+    },
+    {
+      "epoch": 7.881181861273503,
+      "grad_norm": 1.1565742492675781,
+      "learning_rate": 2.1324064084765815e-05,
+      "loss": 1.615,
+      "step": 37410
+    },
+    {
+      "epoch": 7.883288565860852,
+      "grad_norm": 1.1512229442596436,
+      "learning_rate": 2.1283222096117982e-05,
+      "loss": 1.5886,
+      "step": 37420
+    },
+    {
+      "epoch": 7.8853952704482015,
+      "grad_norm": 1.209027886390686,
+      "learning_rate": 2.124241459892955e-05,
+      "loss": 1.5634,
+      "step": 37430
+    },
+    {
+      "epoch": 7.88750197503555,
+      "grad_norm": 1.07988703250885,
+      "learning_rate": 2.1201641611081246e-05,
+      "loss": 1.546,
+      "step": 37440
+    },
+    {
+      "epoch": 7.8896086796229,
+      "grad_norm": 1.1130520105361938,
+      "learning_rate": 2.1160903150438605e-05,
+      "loss": 1.5706,
+      "step": 37450
+    },
+    {
+      "epoch": 7.891715384210249,
+      "grad_norm": 1.0994175672531128,
+      "learning_rate": 2.1120199234852067e-05,
+      "loss": 1.5346,
+      "step": 37460
+    },
+    {
+      "epoch": 7.893822088797599,
+      "grad_norm": 1.1334426403045654,
+      "learning_rate": 2.10795298821569e-05,
+      "loss": 1.5728,
+      "step": 37470
+    },
+    {
+      "epoch": 7.895928793384948,
+      "grad_norm": 1.1012332439422607,
+      "learning_rate": 2.1038895110173283e-05,
+      "loss": 1.5863,
+      "step": 37480
+    },
+    {
+      "epoch": 7.898035497972296,
+      "grad_norm": 1.2324810028076172,
+      "learning_rate": 2.0998294936706288e-05,
+      "loss": 1.632,
+      "step": 37490
+    },
+    {
+      "epoch": 7.900142202559646,
+      "grad_norm": 1.2288521528244019,
+      "learning_rate": 2.0957729379545655e-05,
+      "loss": 1.6035,
+      "step": 37500
+    },
+    {
+      "epoch": 7.902248907146995,
+      "grad_norm": 1.2216746807098389,
+      "learning_rate": 2.091719845646609e-05,
+      "loss": 1.61,
+      "step": 37510
+    },
+    {
+      "epoch": 7.904355611734345,
+      "grad_norm": 1.077216386795044,
+      "learning_rate": 2.0876702185227137e-05,
+      "loss": 1.5732,
+      "step": 37520
+    },
+    {
+      "epoch": 7.906462316321694,
+      "grad_norm": 1.1874003410339355,
+      "learning_rate": 2.0836240583573098e-05,
+      "loss": 1.5926,
+      "step": 37530
+    },
+    {
+      "epoch": 7.908569020909043,
+      "grad_norm": 1.2038264274597168,
+      "learning_rate": 2.079581366923308e-05,
+      "loss": 1.5913,
+      "step": 37540
+    },
+    {
+      "epoch": 7.910675725496392,
+      "grad_norm": 1.289731740951538,
+      "learning_rate": 2.0755421459920986e-05,
+      "loss": 1.5404,
+      "step": 37550
+    },
+    {
+      "epoch": 7.912782430083741,
+      "grad_norm": 1.265437126159668,
+      "learning_rate": 2.0715063973335568e-05,
+      "loss": 1.616,
+      "step": 37560
+    },
+    {
+      "epoch": 7.914889134671091,
+      "grad_norm": 1.1244462728500366,
+      "learning_rate": 2.067474122716039e-05,
+      "loss": 1.563,
+      "step": 37570
+    },
+    {
+      "epoch": 7.91699583925844,
+      "grad_norm": 1.1906565427780151,
+      "learning_rate": 2.0634453239063623e-05,
+      "loss": 1.5503,
+      "step": 37580
+    },
+    {
+      "epoch": 7.919102543845789,
+      "grad_norm": 1.2545281648635864,
+      "learning_rate": 2.0594200026698363e-05,
+      "loss": 1.5598,
+      "step": 37590
+    },
+    {
+      "epoch": 7.921209248433138,
+      "grad_norm": 1.2519073486328125,
+      "learning_rate": 2.0553981607702478e-05,
+      "loss": 1.5408,
+      "step": 37600
+    },
+    {
+      "epoch": 7.923315953020488,
+      "grad_norm": 1.3093730211257935,
+      "learning_rate": 2.051379799969849e-05,
+      "loss": 1.5862,
+      "step": 37610
+    },
+    {
+      "epoch": 7.925422657607837,
+      "grad_norm": 1.108938455581665,
+      "learning_rate": 2.04736492202937e-05,
+      "loss": 1.6347,
+      "step": 37620
+    },
+    {
+      "epoch": 7.927529362195187,
+      "grad_norm": 1.1940797567367554,
+      "learning_rate": 2.0433535287080217e-05,
+      "loss": 1.5676,
+      "step": 37630
+    },
+    {
+      "epoch": 7.929636066782535,
+      "grad_norm": 1.052841067314148,
+      "learning_rate": 2.0393456217634775e-05,
+      "loss": 1.5539,
+      "step": 37640
+    },
+    {
+      "epoch": 7.931742771369885,
+      "grad_norm": 1.1526647806167603,
+      "learning_rate": 2.035341202951897e-05,
+      "loss": 1.5962,
+      "step": 37650
+    },
+    {
+      "epoch": 7.933849475957234,
+      "grad_norm": 1.2988382577896118,
+      "learning_rate": 2.0313402740278908e-05,
+      "loss": 1.5831,
+      "step": 37660
+    },
+    {
+      "epoch": 7.935956180544583,
+      "grad_norm": 1.2303566932678223,
+      "learning_rate": 2.027342836744559e-05,
+      "loss": 1.6096,
+      "step": 37670
+    },
+    {
+      "epoch": 7.938062885131933,
+      "grad_norm": 1.2298696041107178,
+      "learning_rate": 2.0233488928534673e-05,
+      "loss": 1.5679,
+      "step": 37680
+    },
+    {
+      "epoch": 7.940169589719281,
+      "grad_norm": 1.2016994953155518,
+      "learning_rate": 2.0193584441046443e-05,
+      "loss": 1.5947,
+      "step": 37690
+    },
+    {
+      "epoch": 7.942276294306631,
+      "grad_norm": 1.2675684690475464,
+      "learning_rate": 2.015371492246596e-05,
+      "loss": 1.5852,
+      "step": 37700
+    },
+    {
+      "epoch": 7.94438299889398,
+      "grad_norm": 1.2211737632751465,
+      "learning_rate": 2.0113880390262884e-05,
+      "loss": 1.5966,
+      "step": 37710
+    },
+    {
+      "epoch": 7.94648970348133,
+      "grad_norm": 1.0325324535369873,
+      "learning_rate": 2.0074080861891565e-05,
+      "loss": 1.5497,
+      "step": 37720
+    },
+    {
+      "epoch": 7.948596408068679,
+      "grad_norm": 1.1817600727081299,
+      "learning_rate": 2.0034316354791062e-05,
+      "loss": 1.5462,
+      "step": 37730
+    },
+    {
+      "epoch": 7.9507031126560275,
+      "grad_norm": 1.1029366254806519,
+      "learning_rate": 1.9994586886385046e-05,
+      "loss": 1.5305,
+      "step": 37740
+    },
+    {
+      "epoch": 7.952809817243377,
+      "grad_norm": 1.2298988103866577,
+      "learning_rate": 1.995489247408181e-05,
+      "loss": 1.5816,
+      "step": 37750
+    },
+    {
+      "epoch": 7.954916521830726,
+      "grad_norm": 1.2545249462127686,
+      "learning_rate": 1.991523313527437e-05,
+      "loss": 1.6065,
+      "step": 37760
+    },
+    {
+      "epoch": 7.957023226418076,
+      "grad_norm": 1.2715802192687988,
+      "learning_rate": 1.987560888734027e-05,
+      "loss": 1.6085,
+      "step": 37770
+    },
+    {
+      "epoch": 7.959129931005425,
+      "grad_norm": 1.146997332572937,
+      "learning_rate": 1.98360197476418e-05,
+      "loss": 1.5374,
+      "step": 37780
+    },
+    {
+      "epoch": 7.961236635592774,
+      "grad_norm": 1.3229591846466064,
+      "learning_rate": 1.979646573352574e-05,
+      "loss": 1.6448,
+      "step": 37790
+    },
+    {
+      "epoch": 7.963343340180123,
+      "grad_norm": 1.1811490058898926,
+      "learning_rate": 1.9756946862323535e-05,
+      "loss": 1.5583,
+      "step": 37800
+    },
+    {
+      "epoch": 7.965450044767472,
+      "grad_norm": 1.248812198638916,
+      "learning_rate": 1.971746315135129e-05,
+      "loss": 1.5434,
+      "step": 37810
+    },
+    {
+      "epoch": 7.967556749354822,
+      "grad_norm": 1.3636667728424072,
+      "learning_rate": 1.9678014617909602e-05,
+      "loss": 1.6521,
+      "step": 37820
+    },
+    {
+      "epoch": 7.969663453942171,
+      "grad_norm": 1.1896461248397827,
+      "learning_rate": 1.9638601279283684e-05,
+      "loss": 1.6116,
+      "step": 37830
+    },
+    {
+      "epoch": 7.97177015852952,
+      "grad_norm": 1.2454185485839844,
+      "learning_rate": 1.9599223152743395e-05,
+      "loss": 1.5729,
+      "step": 37840
+    },
+    {
+      "epoch": 7.973876863116869,
+      "grad_norm": 1.169144630432129,
+      "learning_rate": 1.955988025554305e-05,
+      "loss": 1.5811,
+      "step": 37850
+    },
+    {
+      "epoch": 7.975983567704219,
+      "grad_norm": 1.1805130243301392,
+      "learning_rate": 1.9520572604921672e-05,
+      "loss": 1.6041,
+      "step": 37860
+    },
+    {
+      "epoch": 7.978090272291568,
+      "grad_norm": 1.1419270038604736,
+      "learning_rate": 1.9481300218102692e-05,
+      "loss": 1.5693,
+      "step": 37870
+    },
+    {
+      "epoch": 7.980196976878917,
+      "grad_norm": 1.2116881608963013,
+      "learning_rate": 1.9442063112294163e-05,
+      "loss": 1.5413,
+      "step": 37880
+    },
+    {
+      "epoch": 7.9823036814662665,
+      "grad_norm": 1.3711943626403809,
+      "learning_rate": 1.9402861304688712e-05,
+      "loss": 1.5626,
+      "step": 37890
+    },
+    {
+      "epoch": 7.984410386053615,
+      "grad_norm": 1.208719253540039,
+      "learning_rate": 1.936369481246344e-05,
+      "loss": 1.5836,
+      "step": 37900
+    },
+    {
+      "epoch": 7.986517090640965,
+      "grad_norm": 1.2907274961471558,
+      "learning_rate": 1.9324563652779947e-05,
+      "loss": 1.5959,
+      "step": 37910
+    },
+    {
+      "epoch": 7.988623795228314,
+      "grad_norm": 1.100823163986206,
+      "learning_rate": 1.9285467842784467e-05,
+      "loss": 1.597,
+      "step": 37920
+    },
+    {
+      "epoch": 7.990730499815664,
+      "grad_norm": 1.3616795539855957,
+      "learning_rate": 1.9246407399607625e-05,
+      "loss": 1.5545,
+      "step": 37930
+    },
+    {
+      "epoch": 7.9928372044030125,
+      "grad_norm": 1.2126030921936035,
+      "learning_rate": 1.9207382340364634e-05,
+      "loss": 1.5463,
+      "step": 37940
+    },
+    {
+      "epoch": 7.994943908990361,
+      "grad_norm": 1.2499946355819702,
+      "learning_rate": 1.9168392682155157e-05,
+      "loss": 1.5681,
+      "step": 37950
+    },
+    {
+      "epoch": 7.997050613577711,
+      "grad_norm": 1.1559704542160034,
+      "learning_rate": 1.912943844206333e-05,
+      "loss": 1.6395,
+      "step": 37960
+    },
+    {
+      "epoch": 7.99915731816506,
+      "grad_norm": 1.2376368045806885,
+      "learning_rate": 1.9090519637157846e-05,
+      "loss": 1.575,
+      "step": 37970
+    },
+    {
+      "epoch": 8.001264022752409,
+      "grad_norm": 1.3157291412353516,
+      "learning_rate": 1.9051636284491757e-05,
+      "loss": 1.6129,
+      "step": 37980
+    },
+    {
+      "epoch": 8.003370727339759,
+      "grad_norm": 1.1666113138198853,
+      "learning_rate": 1.901278840110272e-05,
+      "loss": 1.5028,
+      "step": 37990
+    },
+    {
+      "epoch": 8.005477431927108,
+      "grad_norm": 1.3203294277191162,
+      "learning_rate": 1.897397600401274e-05,
+      "loss": 1.5457,
+      "step": 38000
+    },
+    {
+      "epoch": 8.007584136514458,
+      "grad_norm": 1.1368193626403809,
+      "learning_rate": 1.8935199110228275e-05,
+      "loss": 1.5735,
+      "step": 38010
+    },
+    {
+      "epoch": 8.009690841101806,
+      "grad_norm": 1.1702295541763306,
+      "learning_rate": 1.8896457736740313e-05,
+      "loss": 1.5446,
+      "step": 38020
+    },
+    {
+      "epoch": 8.011797545689156,
+      "grad_norm": 1.217568039894104,
+      "learning_rate": 1.88577519005242e-05,
+      "loss": 1.5859,
+      "step": 38030
+    },
+    {
+      "epoch": 8.013904250276505,
+      "grad_norm": 1.2242965698242188,
+      "learning_rate": 1.8819081618539723e-05,
+      "loss": 1.6145,
+      "step": 38040
+    },
+    {
+      "epoch": 8.016010954863853,
+      "grad_norm": 1.059897541999817,
+      "learning_rate": 1.8780446907731142e-05,
+      "loss": 1.5619,
+      "step": 38050
+    },
+    {
+      "epoch": 8.018117659451203,
+      "grad_norm": 1.1061480045318604,
+      "learning_rate": 1.8741847785027045e-05,
+      "loss": 1.5779,
+      "step": 38060
+    },
+    {
+      "epoch": 8.020224364038553,
+      "grad_norm": 1.2350695133209229,
+      "learning_rate": 1.8703284267340516e-05,
+      "loss": 1.6109,
+      "step": 38070
+    },
+    {
+      "epoch": 8.022331068625903,
+      "grad_norm": 1.1782333850860596,
+      "learning_rate": 1.866475637156898e-05,
+      "loss": 1.5727,
+      "step": 38080
+    },
+    {
+      "epoch": 8.02443777321325,
+      "grad_norm": 1.1965060234069824,
+      "learning_rate": 1.862626411459424e-05,
+      "loss": 1.6016,
+      "step": 38090
+    },
+    {
+      "epoch": 8.0265444778006,
+      "grad_norm": 1.1553235054016113,
+      "learning_rate": 1.858780751328255e-05,
+      "loss": 1.6156,
+      "step": 38100
+    },
+    {
+      "epoch": 8.02865118238795,
+      "grad_norm": 1.1190050840377808,
+      "learning_rate": 1.8549386584484495e-05,
+      "loss": 1.568,
+      "step": 38110
+    },
+    {
+      "epoch": 8.030757886975298,
+      "grad_norm": 1.1927499771118164,
+      "learning_rate": 1.8511001345034994e-05,
+      "loss": 1.5496,
+      "step": 38120
+    },
+    {
+      "epoch": 8.032864591562648,
+      "grad_norm": 1.1628167629241943,
+      "learning_rate": 1.8472651811753428e-05,
+      "loss": 1.6048,
+      "step": 38130
+    },
+    {
+      "epoch": 8.034971296149997,
+      "grad_norm": 1.1890445947647095,
+      "learning_rate": 1.843433800144343e-05,
+      "loss": 1.5735,
+      "step": 38140
+    },
+    {
+      "epoch": 8.037078000737347,
+      "grad_norm": 1.137284755706787,
+      "learning_rate": 1.839605993089307e-05,
+      "loss": 1.5657,
+      "step": 38150
+    },
+    {
+      "epoch": 8.039184705324695,
+      "grad_norm": 1.2434216737747192,
+      "learning_rate": 1.8357817616874694e-05,
+      "loss": 1.5726,
+      "step": 38160
+    },
+    {
+      "epoch": 8.041291409912045,
+      "grad_norm": 1.249247670173645,
+      "learning_rate": 1.831961107614496e-05,
+      "loss": 1.5715,
+      "step": 38170
+    },
+    {
+      "epoch": 8.043398114499395,
+      "grad_norm": 1.3489915132522583,
+      "learning_rate": 1.8281440325444953e-05,
+      "loss": 1.5904,
+      "step": 38180
+    },
+    {
+      "epoch": 8.045504819086744,
+      "grad_norm": 1.2363272905349731,
+      "learning_rate": 1.8243305381499976e-05,
+      "loss": 1.555,
+      "step": 38190
+    },
+    {
+      "epoch": 8.047611523674092,
+      "grad_norm": 1.1866079568862915,
+      "learning_rate": 1.820520626101967e-05,
+      "loss": 1.5552,
+      "step": 38200
+    },
+    {
+      "epoch": 8.049718228261442,
+      "grad_norm": 1.2174458503723145,
+      "learning_rate": 1.816714298069804e-05,
+      "loss": 1.5627,
+      "step": 38210
+    },
+    {
+      "epoch": 8.051824932848792,
+      "grad_norm": 1.2444413900375366,
+      "learning_rate": 1.8129115557213262e-05,
+      "loss": 1.6285,
+      "step": 38220
+    },
+    {
+      "epoch": 8.05393163743614,
+      "grad_norm": 1.2618112564086914,
+      "learning_rate": 1.8091124007227945e-05,
+      "loss": 1.5893,
+      "step": 38230
+    },
+    {
+      "epoch": 8.05603834202349,
+      "grad_norm": 1.2796944379806519,
+      "learning_rate": 1.8053168347388884e-05,
+      "loss": 1.5575,
+      "step": 38240
+    },
+    {
+      "epoch": 8.05814504661084,
+      "grad_norm": 1.1825329065322876,
+      "learning_rate": 1.801524859432714e-05,
+      "loss": 1.5461,
+      "step": 38250
+    },
+    {
+      "epoch": 8.060251751198189,
+      "grad_norm": 1.2693184614181519,
+      "learning_rate": 1.7977364764658122e-05,
+      "loss": 1.5468,
+      "step": 38260
+    },
+    {
+      "epoch": 8.062358455785537,
+      "grad_norm": 1.3873083591461182,
+      "learning_rate": 1.79395168749814e-05,
+      "loss": 1.6055,
+      "step": 38270
+    },
+    {
+      "epoch": 8.064465160372887,
+      "grad_norm": 1.1636654138565063,
+      "learning_rate": 1.7901704941880914e-05,
+      "loss": 1.5814,
+      "step": 38280
+    },
+    {
+      "epoch": 8.066571864960236,
+      "grad_norm": 1.1651959419250488,
+      "learning_rate": 1.7863928981924726e-05,
+      "loss": 1.5662,
+      "step": 38290
+    },
+    {
+      "epoch": 8.068678569547584,
+      "grad_norm": 1.3011173009872437,
+      "learning_rate": 1.7826189011665184e-05,
+      "loss": 1.5698,
+      "step": 38300
+    },
+    {
+      "epoch": 8.070785274134934,
+      "grad_norm": 1.2538598775863647,
+      "learning_rate": 1.7788485047638925e-05,
+      "loss": 1.5508,
+      "step": 38310
+    },
+    {
+      "epoch": 8.072891978722284,
+      "grad_norm": 1.1562957763671875,
+      "learning_rate": 1.7750817106366714e-05,
+      "loss": 1.5399,
+      "step": 38320
+    },
+    {
+      "epoch": 8.074998683309634,
+      "grad_norm": 1.2713629007339478,
+      "learning_rate": 1.7713185204353567e-05,
+      "loss": 1.5962,
+      "step": 38330
+    },
+    {
+      "epoch": 8.077105387896982,
+      "grad_norm": 1.2110227346420288,
+      "learning_rate": 1.7675589358088763e-05,
+      "loss": 1.5827,
+      "step": 38340
+    },
+    {
+      "epoch": 8.079212092484331,
+      "grad_norm": 1.2139581441879272,
+      "learning_rate": 1.763802958404568e-05,
+      "loss": 1.5289,
+      "step": 38350
+    },
+    {
+      "epoch": 8.081318797071681,
+      "grad_norm": 1.1466100215911865,
+      "learning_rate": 1.7600505898681997e-05,
+      "loss": 1.5669,
+      "step": 38360
+    },
+    {
+      "epoch": 8.083425501659029,
+      "grad_norm": 1.122790813446045,
+      "learning_rate": 1.7563018318439496e-05,
+      "loss": 1.519,
+      "step": 38370
+    },
+    {
+      "epoch": 8.085532206246379,
+      "grad_norm": 1.6603208780288696,
+      "learning_rate": 1.7525566859744168e-05,
+      "loss": 1.609,
+      "step": 38380
+    },
+    {
+      "epoch": 8.087638910833729,
+      "grad_norm": 1.1294156312942505,
+      "learning_rate": 1.7488151539006203e-05,
+      "loss": 1.5371,
+      "step": 38390
+    },
+    {
+      "epoch": 8.089745615421078,
+      "grad_norm": 1.2689809799194336,
+      "learning_rate": 1.745077237261994e-05,
+      "loss": 1.5928,
+      "step": 38400
+    },
+    {
+      "epoch": 8.091852320008426,
+      "grad_norm": 1.1921470165252686,
+      "learning_rate": 1.7413429376963808e-05,
+      "loss": 1.5385,
+      "step": 38410
+    },
+    {
+      "epoch": 8.093959024595776,
+      "grad_norm": 1.2434748411178589,
+      "learning_rate": 1.7376122568400532e-05,
+      "loss": 1.5235,
+      "step": 38420
+    },
+    {
+      "epoch": 8.096065729183126,
+      "grad_norm": 1.590907335281372,
+      "learning_rate": 1.7338851963276825e-05,
+      "loss": 1.5797,
+      "step": 38430
+    },
+    {
+      "epoch": 8.098172433770474,
+      "grad_norm": 1.129142165184021,
+      "learning_rate": 1.730161757792367e-05,
+      "loss": 1.5699,
+      "step": 38440
+    },
+    {
+      "epoch": 8.100279138357823,
+      "grad_norm": 1.270219326019287,
+      "learning_rate": 1.7264419428656075e-05,
+      "loss": 1.5919,
+      "step": 38450
+    },
+    {
+      "epoch": 8.102385842945173,
+      "grad_norm": 1.2105395793914795,
+      "learning_rate": 1.7227257531773223e-05,
+      "loss": 1.5551,
+      "step": 38460
+    },
+    {
+      "epoch": 8.104492547532523,
+      "grad_norm": 1.2144361734390259,
+      "learning_rate": 1.7190131903558425e-05,
+      "loss": 1.5649,
+      "step": 38470
+    },
+    {
+      "epoch": 8.10659925211987,
+      "grad_norm": 1.217391014099121,
+      "learning_rate": 1.7153042560279065e-05,
+      "loss": 1.5267,
+      "step": 38480
+    },
+    {
+      "epoch": 8.10870595670722,
+      "grad_norm": 1.2412052154541016,
+      "learning_rate": 1.7115989518186615e-05,
+      "loss": 1.528,
+      "step": 38490
+    },
+    {
+      "epoch": 8.11081266129457,
+      "grad_norm": 1.7172247171401978,
+      "learning_rate": 1.707897279351671e-05,
+      "loss": 1.5828,
+      "step": 38500
+    },
+    {
+      "epoch": 8.11291936588192,
+      "grad_norm": 1.2050105333328247,
+      "learning_rate": 1.7041992402488994e-05,
+      "loss": 1.5875,
+      "step": 38510
+    },
+    {
+      "epoch": 8.115026070469268,
+      "grad_norm": 1.7654755115509033,
+      "learning_rate": 1.7005048361307262e-05,
+      "loss": 1.5203,
+      "step": 38520
+    },
+    {
+      "epoch": 8.117132775056618,
+      "grad_norm": 1.2826247215270996,
+      "learning_rate": 1.6968140686159328e-05,
+      "loss": 1.5569,
+      "step": 38530
+    },
+    {
+      "epoch": 8.119239479643968,
+      "grad_norm": 1.1842323541641235,
+      "learning_rate": 1.693126939321705e-05,
+      "loss": 1.538,
+      "step": 38540
+    },
+    {
+      "epoch": 8.121346184231315,
+      "grad_norm": 1.1324920654296875,
+      "learning_rate": 1.6894434498636446e-05,
+      "loss": 1.5795,
+      "step": 38550
+    },
+    {
+      "epoch": 8.123452888818665,
+      "grad_norm": 1.240378975868225,
+      "learning_rate": 1.6857636018557466e-05,
+      "loss": 1.5545,
+      "step": 38560
+    },
+    {
+      "epoch": 8.125559593406015,
+      "grad_norm": 1.0717569589614868,
+      "learning_rate": 1.682087396910422e-05,
+      "loss": 1.568,
+      "step": 38570
+    },
+    {
+      "epoch": 8.127666297993365,
+      "grad_norm": 1.266964077949524,
+      "learning_rate": 1.6784148366384754e-05,
+      "loss": 1.5935,
+      "step": 38580
+    },
+    {
+      "epoch": 8.129773002580713,
+      "grad_norm": 1.1756432056427002,
+      "learning_rate": 1.674745922649118e-05,
+      "loss": 1.5398,
+      "step": 38590
+    },
+    {
+      "epoch": 8.131879707168062,
+      "grad_norm": 1.2381880283355713,
+      "learning_rate": 1.671080656549965e-05,
+      "loss": 1.5288,
+      "step": 38600
+    },
+    {
+      "epoch": 8.133986411755412,
+      "grad_norm": 1.2864570617675781,
+      "learning_rate": 1.667419039947037e-05,
+      "loss": 1.5336,
+      "step": 38610
+    },
+    {
+      "epoch": 8.13609311634276,
+      "grad_norm": 1.2409577369689941,
+      "learning_rate": 1.6637610744447407e-05,
+      "loss": 1.5452,
+      "step": 38620
+    },
+    {
+      "epoch": 8.13819982093011,
+      "grad_norm": 1.2823350429534912,
+      "learning_rate": 1.6601067616458987e-05,
+      "loss": 1.5658,
+      "step": 38630
+    },
+    {
+      "epoch": 8.14030652551746,
+      "grad_norm": 1.2008086442947388,
+      "learning_rate": 1.656456103151728e-05,
+      "loss": 1.5276,
+      "step": 38640
+    },
+    {
+      "epoch": 8.14241323010481,
+      "grad_norm": 1.245653748512268,
+      "learning_rate": 1.6528091005618428e-05,
+      "loss": 1.5227,
+      "step": 38650
+    },
+    {
+      "epoch": 8.144519934692157,
+      "grad_norm": 1.184909701347351,
+      "learning_rate": 1.6491657554742557e-05,
+      "loss": 1.5415,
+      "step": 38660
+    },
+    {
+      "epoch": 8.146626639279507,
+      "grad_norm": 1.199066400527954,
+      "learning_rate": 1.6455260694853736e-05,
+      "loss": 1.5847,
+      "step": 38670
+    },
+    {
+      "epoch": 8.148733343866857,
+      "grad_norm": 1.2071822881698608,
+      "learning_rate": 1.6418900441900087e-05,
+      "loss": 1.5406,
+      "step": 38680
+    },
+    {
+      "epoch": 8.150840048454205,
+      "grad_norm": 1.2048323154449463,
+      "learning_rate": 1.6382576811813655e-05,
+      "loss": 1.5376,
+      "step": 38690
+    },
+    {
+      "epoch": 8.152946753041554,
+      "grad_norm": 1.1268657445907593,
+      "learning_rate": 1.6346289820510363e-05,
+      "loss": 1.5304,
+      "step": 38700
+    },
+    {
+      "epoch": 8.155053457628904,
+      "grad_norm": 1.2604269981384277,
+      "learning_rate": 1.631003948389016e-05,
+      "loss": 1.5056,
+      "step": 38710
+    },
+    {
+      "epoch": 8.157160162216254,
+      "grad_norm": 1.1815440654754639,
+      "learning_rate": 1.6273825817836963e-05,
+      "loss": 1.6137,
+      "step": 38720
+    },
+    {
+      "epoch": 8.159266866803602,
+      "grad_norm": 1.1286557912826538,
+      "learning_rate": 1.6237648838218532e-05,
+      "loss": 1.5052,
+      "step": 38730
+    },
+    {
+      "epoch": 8.161373571390952,
+      "grad_norm": 1.3425495624542236,
+      "learning_rate": 1.6201508560886602e-05,
+      "loss": 1.5563,
+      "step": 38740
+    },
+    {
+      "epoch": 8.163480275978301,
+      "grad_norm": 1.1465646028518677,
+      "learning_rate": 1.6165405001676793e-05,
+      "loss": 1.5663,
+      "step": 38750
+    },
+    {
+      "epoch": 8.16558698056565,
+      "grad_norm": 1.232774019241333,
+      "learning_rate": 1.612933817640868e-05,
+      "loss": 1.5836,
+      "step": 38760
+    },
+    {
+      "epoch": 8.167693685152999,
+      "grad_norm": 1.1636687517166138,
+      "learning_rate": 1.6093308100885774e-05,
+      "loss": 1.5692,
+      "step": 38770
+    },
+    {
+      "epoch": 8.169800389740349,
+      "grad_norm": 1.1775188446044922,
+      "learning_rate": 1.605731479089534e-05,
+      "loss": 1.5056,
+      "step": 38780
+    },
+    {
+      "epoch": 8.171907094327699,
+      "grad_norm": 1.2304803133010864,
+      "learning_rate": 1.6021358262208665e-05,
+      "loss": 1.5485,
+      "step": 38790
+    },
+    {
+      "epoch": 8.174013798915047,
+      "grad_norm": 1.2564809322357178,
+      "learning_rate": 1.5985438530580908e-05,
+      "loss": 1.5348,
+      "step": 38800
+    },
+    {
+      "epoch": 8.176120503502396,
+      "grad_norm": 1.2680792808532715,
+      "learning_rate": 1.5949555611751044e-05,
+      "loss": 1.5888,
+      "step": 38810
+    },
+    {
+      "epoch": 8.178227208089746,
+      "grad_norm": 1.1961019039154053,
+      "learning_rate": 1.5913709521441988e-05,
+      "loss": 1.5454,
+      "step": 38820
+    },
+    {
+      "epoch": 8.180333912677096,
+      "grad_norm": 1.2921595573425293,
+      "learning_rate": 1.5877900275360412e-05,
+      "loss": 1.57,
+      "step": 38830
+    },
+    {
+      "epoch": 8.182440617264444,
+      "grad_norm": 1.3242322206497192,
+      "learning_rate": 1.5842127889196956e-05,
+      "loss": 1.601,
+      "step": 38840
+    },
+    {
+      "epoch": 8.184547321851793,
+      "grad_norm": 1.1462351083755493,
+      "learning_rate": 1.580639237862608e-05,
+      "loss": 1.6022,
+      "step": 38850
+    },
+    {
+      "epoch": 8.186654026439143,
+      "grad_norm": 1.270314335823059,
+      "learning_rate": 1.5770693759306055e-05,
+      "loss": 1.6073,
+      "step": 38860
+    },
+    {
+      "epoch": 8.188760731026491,
+      "grad_norm": 1.2723679542541504,
+      "learning_rate": 1.5735032046878973e-05,
+      "loss": 1.5688,
+      "step": 38870
+    },
+    {
+      "epoch": 8.190867435613841,
+      "grad_norm": 1.1393686532974243,
+      "learning_rate": 1.5699407256970833e-05,
+      "loss": 1.5085,
+      "step": 38880
+    },
+    {
+      "epoch": 8.19297414020119,
+      "grad_norm": 1.1237406730651855,
+      "learning_rate": 1.5663819405191373e-05,
+      "loss": 1.5975,
+      "step": 38890
+    },
+    {
+      "epoch": 8.19508084478854,
+      "grad_norm": 1.1238974332809448,
+      "learning_rate": 1.5628268507134224e-05,
+      "loss": 1.517,
+      "step": 38900
+    },
+    {
+      "epoch": 8.197187549375888,
+      "grad_norm": 1.300145149230957,
+      "learning_rate": 1.5592754578376724e-05,
+      "loss": 1.6047,
+      "step": 38910
+    },
+    {
+      "epoch": 8.199294253963238,
+      "grad_norm": 1.1321696043014526,
+      "learning_rate": 1.5557277634480083e-05,
+      "loss": 1.5334,
+      "step": 38920
+    },
+    {
+      "epoch": 8.201400958550588,
+      "grad_norm": 1.1901246309280396,
+      "learning_rate": 1.5521837690989338e-05,
+      "loss": 1.5728,
+      "step": 38930
+    },
+    {
+      "epoch": 8.203507663137936,
+      "grad_norm": 1.2294230461120605,
+      "learning_rate": 1.5486434763433222e-05,
+      "loss": 1.5832,
+      "step": 38940
+    },
+    {
+      "epoch": 8.205614367725286,
+      "grad_norm": 1.2560628652572632,
+      "learning_rate": 1.5451068867324293e-05,
+      "loss": 1.5975,
+      "step": 38950
+    },
+    {
+      "epoch": 8.207721072312635,
+      "grad_norm": 1.312564492225647,
+      "learning_rate": 1.541574001815892e-05,
+      "loss": 1.568,
+      "step": 38960
+    },
+    {
+      "epoch": 8.209827776899985,
+      "grad_norm": 1.3545145988464355,
+      "learning_rate": 1.5380448231417144e-05,
+      "loss": 1.6047,
+      "step": 38970
+    },
+    {
+      "epoch": 8.211934481487333,
+      "grad_norm": 1.142449975013733,
+      "learning_rate": 1.5345193522562917e-05,
+      "loss": 1.5515,
+      "step": 38980
+    },
+    {
+      "epoch": 8.214041186074683,
+      "grad_norm": 1.1972579956054688,
+      "learning_rate": 1.530997590704375e-05,
+      "loss": 1.585,
+      "step": 38990
+    },
+    {
+      "epoch": 8.216147890662032,
+      "grad_norm": 1.2056550979614258,
+      "learning_rate": 1.527479540029104e-05,
+      "loss": 1.5195,
+      "step": 39000
+    },
+    {
+      "epoch": 8.21825459524938,
+      "grad_norm": 1.3925740718841553,
+      "learning_rate": 1.5239652017719919e-05,
+      "loss": 1.6004,
+      "step": 39010
+    },
+    {
+      "epoch": 8.22036129983673,
+      "grad_norm": 1.4021625518798828,
+      "learning_rate": 1.5204545774729207e-05,
+      "loss": 1.5881,
+      "step": 39020
+    },
+    {
+      "epoch": 8.22246800442408,
+      "grad_norm": 1.195132851600647,
+      "learning_rate": 1.5169476686701423e-05,
+      "loss": 1.5715,
+      "step": 39030
+    },
+    {
+      "epoch": 8.22457470901143,
+      "grad_norm": 1.2492402791976929,
+      "learning_rate": 1.5134444769002909e-05,
+      "loss": 1.4855,
+      "step": 39040
+    },
+    {
+      "epoch": 8.226681413598778,
+      "grad_norm": 1.1475170850753784,
+      "learning_rate": 1.5099450036983598e-05,
+      "loss": 1.5519,
+      "step": 39050
+    },
+    {
+      "epoch": 8.228788118186127,
+      "grad_norm": 1.213303565979004,
+      "learning_rate": 1.5064492505977234e-05,
+      "loss": 1.5203,
+      "step": 39060
+    },
+    {
+      "epoch": 8.230894822773477,
+      "grad_norm": 1.260926604270935,
+      "learning_rate": 1.5029572191301211e-05,
+      "loss": 1.5361,
+      "step": 39070
+    },
+    {
+      "epoch": 8.233001527360825,
+      "grad_norm": 1.3445188999176025,
+      "learning_rate": 1.4994689108256576e-05,
+      "loss": 1.5805,
+      "step": 39080
+    },
+    {
+      "epoch": 8.235108231948175,
+      "grad_norm": 1.1416704654693604,
+      "learning_rate": 1.4959843272128172e-05,
+      "loss": 1.5933,
+      "step": 39090
+    },
+    {
+      "epoch": 8.237214936535525,
+      "grad_norm": 1.2642247676849365,
+      "learning_rate": 1.4925034698184393e-05,
+      "loss": 1.5494,
+      "step": 39100
+    },
+    {
+      "epoch": 8.239321641122874,
+      "grad_norm": 1.2801305055618286,
+      "learning_rate": 1.4890263401677429e-05,
+      "loss": 1.6004,
+      "step": 39110
+    },
+    {
+      "epoch": 8.241428345710222,
+      "grad_norm": 1.2009177207946777,
+      "learning_rate": 1.4855529397843038e-05,
+      "loss": 1.5567,
+      "step": 39120
+    },
+    {
+      "epoch": 8.243535050297572,
+      "grad_norm": 1.2630966901779175,
+      "learning_rate": 1.4820832701900667e-05,
+      "loss": 1.5823,
+      "step": 39130
+    },
+    {
+      "epoch": 8.245641754884922,
+      "grad_norm": 1.2626142501831055,
+      "learning_rate": 1.4786173329053466e-05,
+      "loss": 1.567,
+      "step": 39140
+    },
+    {
+      "epoch": 8.24774845947227,
+      "grad_norm": 1.112261414527893,
+      "learning_rate": 1.4751551294488154e-05,
+      "loss": 1.5831,
+      "step": 39150
+    },
+    {
+      "epoch": 8.24985516405962,
+      "grad_norm": 1.3133232593536377,
+      "learning_rate": 1.4716966613375116e-05,
+      "loss": 1.5637,
+      "step": 39160
+    },
+    {
+      "epoch": 8.25196186864697,
+      "grad_norm": 1.27349853515625,
+      "learning_rate": 1.4682419300868423e-05,
+      "loss": 1.5465,
+      "step": 39170
+    },
+    {
+      "epoch": 8.254068573234319,
+      "grad_norm": 1.2491436004638672,
+      "learning_rate": 1.4647909372105672e-05,
+      "loss": 1.5928,
+      "step": 39180
+    },
+    {
+      "epoch": 8.256175277821667,
+      "grad_norm": 1.2191392183303833,
+      "learning_rate": 1.4613436842208183e-05,
+      "loss": 1.5809,
+      "step": 39190
+    },
+    {
+      "epoch": 8.258281982409017,
+      "grad_norm": 1.204507827758789,
+      "learning_rate": 1.4579001726280828e-05,
+      "loss": 1.5881,
+      "step": 39200
+    },
+    {
+      "epoch": 8.260388686996366,
+      "grad_norm": 1.1342881917953491,
+      "learning_rate": 1.454460403941207e-05,
+      "loss": 1.5041,
+      "step": 39210
+    },
+    {
+      "epoch": 8.262495391583716,
+      "grad_norm": 1.216679573059082,
+      "learning_rate": 1.451024379667404e-05,
+      "loss": 1.5633,
+      "step": 39220
+    },
+    {
+      "epoch": 8.264602096171064,
+      "grad_norm": 1.2390002012252808,
+      "learning_rate": 1.4475921013122406e-05,
+      "loss": 1.5611,
+      "step": 39230
+    },
+    {
+      "epoch": 8.266708800758414,
+      "grad_norm": 1.191501259803772,
+      "learning_rate": 1.4441635703796408e-05,
+      "loss": 1.5184,
+      "step": 39240
+    },
+    {
+      "epoch": 8.268815505345763,
+      "grad_norm": 1.177857756614685,
+      "learning_rate": 1.4407387883718959e-05,
+      "loss": 1.604,
+      "step": 39250
+    },
+    {
+      "epoch": 8.270922209933111,
+      "grad_norm": 1.2561581134796143,
+      "learning_rate": 1.4373177567896413e-05,
+      "loss": 1.5755,
+      "step": 39260
+    },
+    {
+      "epoch": 8.273028914520461,
+      "grad_norm": 1.167371153831482,
+      "learning_rate": 1.433900477131882e-05,
+      "loss": 1.5556,
+      "step": 39270
+    },
+    {
+      "epoch": 8.275135619107811,
+      "grad_norm": 1.2407907247543335,
+      "learning_rate": 1.4304869508959707e-05,
+      "loss": 1.5155,
+      "step": 39280
+    },
+    {
+      "epoch": 8.27724232369516,
+      "grad_norm": 1.125529408454895,
+      "learning_rate": 1.427077179577615e-05,
+      "loss": 1.5944,
+      "step": 39290
+    },
+    {
+      "epoch": 8.279349028282509,
+      "grad_norm": 1.4028643369674683,
+      "learning_rate": 1.4236711646708844e-05,
+      "loss": 1.5687,
+      "step": 39300
+    },
+    {
+      "epoch": 8.281455732869858,
+      "grad_norm": 1.199676275253296,
+      "learning_rate": 1.4202689076681962e-05,
+      "loss": 1.5419,
+      "step": 39310
+    },
+    {
+      "epoch": 8.283562437457208,
+      "grad_norm": 1.2607035636901855,
+      "learning_rate": 1.4168704100603214e-05,
+      "loss": 1.5766,
+      "step": 39320
+    },
+    {
+      "epoch": 8.285669142044556,
+      "grad_norm": 1.2903956174850464,
+      "learning_rate": 1.4134756733363886e-05,
+      "loss": 1.5419,
+      "step": 39330
+    },
+    {
+      "epoch": 8.287775846631906,
+      "grad_norm": 1.4275097846984863,
+      "learning_rate": 1.41008469898387e-05,
+      "loss": 1.5228,
+      "step": 39340
+    },
+    {
+      "epoch": 8.289882551219256,
+      "grad_norm": 1.247402310371399,
+      "learning_rate": 1.4066974884886008e-05,
+      "loss": 1.6018,
+      "step": 39350
+    },
+    {
+      "epoch": 8.291989255806605,
+      "grad_norm": 1.2947194576263428,
+      "learning_rate": 1.4033140433347569e-05,
+      "loss": 1.5944,
+      "step": 39360
+    },
+    {
+      "epoch": 8.294095960393953,
+      "grad_norm": 1.195644736289978,
+      "learning_rate": 1.3999343650048669e-05,
+      "loss": 1.5578,
+      "step": 39370
+    },
+    {
+      "epoch": 8.296202664981303,
+      "grad_norm": 1.1885980367660522,
+      "learning_rate": 1.396558454979814e-05,
+      "loss": 1.5829,
+      "step": 39380
+    },
+    {
+      "epoch": 8.298309369568653,
+      "grad_norm": 1.2197813987731934,
+      "learning_rate": 1.3931863147388202e-05,
+      "loss": 1.5968,
+      "step": 39390
+    },
+    {
+      "epoch": 8.300416074156,
+      "grad_norm": 1.19045090675354,
+      "learning_rate": 1.3898179457594684e-05,
+      "loss": 1.549,
+      "step": 39400
+    },
+    {
+      "epoch": 8.30252277874335,
+      "grad_norm": 1.2025426626205444,
+      "learning_rate": 1.386453349517679e-05,
+      "loss": 1.5336,
+      "step": 39410
+    },
+    {
+      "epoch": 8.3046294833307,
+      "grad_norm": 1.3740406036376953,
+      "learning_rate": 1.3830925274877216e-05,
+      "loss": 1.5438,
+      "step": 39420
+    },
+    {
+      "epoch": 8.30673618791805,
+      "grad_norm": 1.0788160562515259,
+      "learning_rate": 1.3797354811422158e-05,
+      "loss": 1.511,
+      "step": 39430
+    },
+    {
+      "epoch": 8.308842892505398,
+      "grad_norm": 1.213222861289978,
+      "learning_rate": 1.376382211952123e-05,
+      "loss": 1.5702,
+      "step": 39440
+    },
+    {
+      "epoch": 8.310949597092748,
+      "grad_norm": 1.2587624788284302,
+      "learning_rate": 1.3730327213867478e-05,
+      "loss": 1.5513,
+      "step": 39450
+    },
+    {
+      "epoch": 8.313056301680097,
+      "grad_norm": 1.2077735662460327,
+      "learning_rate": 1.369687010913746e-05,
+      "loss": 1.5716,
+      "step": 39460
+    },
+    {
+      "epoch": 8.315163006267445,
+      "grad_norm": 1.3706378936767578,
+      "learning_rate": 1.3663450819991107e-05,
+      "loss": 1.5903,
+      "step": 39470
+    },
+    {
+      "epoch": 8.317269710854795,
+      "grad_norm": 1.298016905784607,
+      "learning_rate": 1.363006936107183e-05,
+      "loss": 1.5508,
+      "step": 39480
+    },
+    {
+      "epoch": 8.319376415442145,
+      "grad_norm": 1.2610794305801392,
+      "learning_rate": 1.3596725747006411e-05,
+      "loss": 1.5297,
+      "step": 39490
+    },
+    {
+      "epoch": 8.321483120029495,
+      "grad_norm": 1.2800568342208862,
+      "learning_rate": 1.3563419992405068e-05,
+      "loss": 1.5699,
+      "step": 39500
+    },
+    {
+      "epoch": 8.323589824616842,
+      "grad_norm": 1.2344589233398438,
+      "learning_rate": 1.3530152111861483e-05,
+      "loss": 1.5382,
+      "step": 39510
+    },
+    {
+      "epoch": 8.325696529204192,
+      "grad_norm": 1.2370203733444214,
+      "learning_rate": 1.349692211995266e-05,
+      "loss": 1.5622,
+      "step": 39520
+    },
+    {
+      "epoch": 8.327803233791542,
+      "grad_norm": 1.173630714416504,
+      "learning_rate": 1.346373003123903e-05,
+      "loss": 1.6422,
+      "step": 39530
+    },
+    {
+      "epoch": 8.32990993837889,
+      "grad_norm": 1.1007499694824219,
+      "learning_rate": 1.343057586026446e-05,
+      "loss": 1.5922,
+      "step": 39540
+    },
+    {
+      "epoch": 8.33201664296624,
+      "grad_norm": 1.3459322452545166,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 1.5034,
+      "step": 39550
+    },
+    {
+      "epoch": 8.33412334755359,
+      "grad_norm": 1.2300962209701538,
+      "learning_rate": 1.3364381329624687e-05,
+      "loss": 1.5769,
+      "step": 39560
+    },
+    {
+      "epoch": 8.33623005214094,
+      "grad_norm": 1.2691370248794556,
+      "learning_rate": 1.333134099896406e-05,
+      "loss": 1.6051,
+      "step": 39570
+    },
+    {
+      "epoch": 8.338336756728287,
+      "grad_norm": 1.3355259895324707,
+      "learning_rate": 1.3298338644051578e-05,
+      "loss": 1.5746,
+      "step": 39580
+    },
+    {
+      "epoch": 8.340443461315637,
+      "grad_norm": 1.2271134853363037,
+      "learning_rate": 1.3265374279347975e-05,
+      "loss": 1.5573,
+      "step": 39590
+    },
+    {
+      "epoch": 8.342550165902987,
+      "grad_norm": 1.389338731765747,
+      "learning_rate": 1.3232447919297274e-05,
+      "loss": 1.564,
+      "step": 39600
+    },
+    {
+      "epoch": 8.344656870490336,
+      "grad_norm": 1.3343056440353394,
+      "learning_rate": 1.3199559578326858e-05,
+      "loss": 1.5654,
+      "step": 39610
+    },
+    {
+      "epoch": 8.346763575077684,
+      "grad_norm": 1.2581366300582886,
+      "learning_rate": 1.3166709270847511e-05,
+      "loss": 1.5501,
+      "step": 39620
+    },
+    {
+      "epoch": 8.348870279665034,
+      "grad_norm": 1.626578688621521,
+      "learning_rate": 1.313389701125325e-05,
+      "loss": 1.5701,
+      "step": 39630
+    },
+    {
+      "epoch": 8.350976984252384,
+      "grad_norm": 1.2747974395751953,
+      "learning_rate": 1.3101122813921529e-05,
+      "loss": 1.5525,
+      "step": 39640
+    },
+    {
+      "epoch": 8.353083688839732,
+      "grad_norm": 1.28239905834198,
+      "learning_rate": 1.3068386693213053e-05,
+      "loss": 1.5239,
+      "step": 39650
+    },
+    {
+      "epoch": 8.355190393427081,
+      "grad_norm": 1.2086161375045776,
+      "learning_rate": 1.3035688663471834e-05,
+      "loss": 1.5576,
+      "step": 39660
+    },
+    {
+      "epoch": 8.357297098014431,
+      "grad_norm": 1.1600311994552612,
+      "learning_rate": 1.3003028739025258e-05,
+      "loss": 1.4787,
+      "step": 39670
+    },
+    {
+      "epoch": 8.359403802601781,
+      "grad_norm": 1.2429834604263306,
+      "learning_rate": 1.2970406934183954e-05,
+      "loss": 1.5686,
+      "step": 39680
+    },
+    {
+      "epoch": 8.361510507189129,
+      "grad_norm": 1.1623550653457642,
+      "learning_rate": 1.29378232632419e-05,
+      "loss": 1.5772,
+      "step": 39690
+    },
+    {
+      "epoch": 8.363617211776479,
+      "grad_norm": 1.1314207315444946,
+      "learning_rate": 1.2905277740476318e-05,
+      "loss": 1.5337,
+      "step": 39700
+    },
+    {
+      "epoch": 8.365723916363828,
+      "grad_norm": 1.2326688766479492,
+      "learning_rate": 1.2872770380147703e-05,
+      "loss": 1.4982,
+      "step": 39710
+    },
+    {
+      "epoch": 8.367830620951176,
+      "grad_norm": 1.2618603706359863,
+      "learning_rate": 1.2840301196499893e-05,
+      "loss": 1.5631,
+      "step": 39720
+    },
+    {
+      "epoch": 8.369937325538526,
+      "grad_norm": 1.1323174238204956,
+      "learning_rate": 1.2807870203760009e-05,
+      "loss": 1.5235,
+      "step": 39730
+    },
+    {
+      "epoch": 8.372044030125876,
+      "grad_norm": 1.1603468656539917,
+      "learning_rate": 1.2775477416138294e-05,
+      "loss": 1.5585,
+      "step": 39740
+    },
+    {
+      "epoch": 8.374150734713226,
+      "grad_norm": 1.1536343097686768,
+      "learning_rate": 1.2743122847828415e-05,
+      "loss": 1.5427,
+      "step": 39750
+    },
+    {
+      "epoch": 8.376257439300574,
+      "grad_norm": 1.1194301843643188,
+      "learning_rate": 1.271080651300719e-05,
+      "loss": 1.5408,
+      "step": 39760
+    },
+    {
+      "epoch": 8.378364143887923,
+      "grad_norm": 1.503873586654663,
+      "learning_rate": 1.2678528425834758e-05,
+      "loss": 1.6087,
+      "step": 39770
+    },
+    {
+      "epoch": 8.380470848475273,
+      "grad_norm": 1.1918821334838867,
+      "learning_rate": 1.2646288600454448e-05,
+      "loss": 1.5488,
+      "step": 39780
+    },
+    {
+      "epoch": 8.382577553062621,
+      "grad_norm": 1.1944622993469238,
+      "learning_rate": 1.2614087050992796e-05,
+      "loss": 1.6118,
+      "step": 39790
+    },
+    {
+      "epoch": 8.38468425764997,
+      "grad_norm": 1.2867789268493652,
+      "learning_rate": 1.2581923791559647e-05,
+      "loss": 1.5397,
+      "step": 39800
+    },
+    {
+      "epoch": 8.38679096223732,
+      "grad_norm": 1.3712007999420166,
+      "learning_rate": 1.2549798836248072e-05,
+      "loss": 1.5778,
+      "step": 39810
+    },
+    {
+      "epoch": 8.38889766682467,
+      "grad_norm": 1.2475578784942627,
+      "learning_rate": 1.2517712199134224e-05,
+      "loss": 1.5816,
+      "step": 39820
+    },
+    {
+      "epoch": 8.391004371412018,
+      "grad_norm": 1.3664524555206299,
+      "learning_rate": 1.2485663894277611e-05,
+      "loss": 1.5837,
+      "step": 39830
+    },
+    {
+      "epoch": 8.393111075999368,
+      "grad_norm": 1.4213124513626099,
+      "learning_rate": 1.2453653935720867e-05,
+      "loss": 1.5526,
+      "step": 39840
+    },
+    {
+      "epoch": 8.395217780586718,
+      "grad_norm": 1.2256556749343872,
+      "learning_rate": 1.2421682337489882e-05,
+      "loss": 1.5103,
+      "step": 39850
+    },
+    {
+      "epoch": 8.397324485174067,
+      "grad_norm": 1.2881014347076416,
+      "learning_rate": 1.2389749113593684e-05,
+      "loss": 1.5409,
+      "step": 39860
+    },
+    {
+      "epoch": 8.399431189761415,
+      "grad_norm": 1.249493956565857,
+      "learning_rate": 1.2357854278024484e-05,
+      "loss": 1.5371,
+      "step": 39870
+    },
+    {
+      "epoch": 8.401537894348765,
+      "grad_norm": 1.1821091175079346,
+      "learning_rate": 1.2325997844757719e-05,
+      "loss": 1.5334,
+      "step": 39880
+    },
+    {
+      "epoch": 8.403644598936115,
+      "grad_norm": 1.125915288925171,
+      "learning_rate": 1.2294179827752007e-05,
+      "loss": 1.6042,
+      "step": 39890
+    },
+    {
+      "epoch": 8.405751303523463,
+      "grad_norm": 1.2114832401275635,
+      "learning_rate": 1.2262400240949023e-05,
+      "loss": 1.5729,
+      "step": 39900
+    },
+    {
+      "epoch": 8.407858008110813,
+      "grad_norm": 1.1864620447158813,
+      "learning_rate": 1.2230659098273744e-05,
+      "loss": 1.5487,
+      "step": 39910
+    },
+    {
+      "epoch": 8.409964712698162,
+      "grad_norm": 1.3160470724105835,
+      "learning_rate": 1.2198956413634199e-05,
+      "loss": 1.5549,
+      "step": 39920
+    },
+    {
+      "epoch": 8.412071417285512,
+      "grad_norm": 1.2356914281845093,
+      "learning_rate": 1.216729220092162e-05,
+      "loss": 1.5617,
+      "step": 39930
+    },
+    {
+      "epoch": 8.41417812187286,
+      "grad_norm": 1.2617237567901611,
+      "learning_rate": 1.213566647401041e-05,
+      "loss": 1.6353,
+      "step": 39940
+    },
+    {
+      "epoch": 8.41628482646021,
+      "grad_norm": 1.1665171384811401,
+      "learning_rate": 1.2104079246757993e-05,
+      "loss": 1.5639,
+      "step": 39950
+    },
+    {
+      "epoch": 8.41839153104756,
+      "grad_norm": 1.424302101135254,
+      "learning_rate": 1.2072530533005012e-05,
+      "loss": 1.5733,
+      "step": 39960
+    },
+    {
+      "epoch": 8.420498235634907,
+      "grad_norm": 1.1204131841659546,
+      "learning_rate": 1.2041020346575272e-05,
+      "loss": 1.5616,
+      "step": 39970
+    },
+    {
+      "epoch": 8.422604940222257,
+      "grad_norm": 1.3583896160125732,
+      "learning_rate": 1.2009548701275598e-05,
+      "loss": 1.5992,
+      "step": 39980
+    },
+    {
+      "epoch": 8.424711644809607,
+      "grad_norm": 1.1708359718322754,
+      "learning_rate": 1.197811561089598e-05,
+      "loss": 1.5822,
+      "step": 39990
+    },
+    {
+      "epoch": 8.426818349396957,
+      "grad_norm": 1.2146726846694946,
+      "learning_rate": 1.1946721089209479e-05,
+      "loss": 1.5704,
+      "step": 40000
+    },
+    {
+      "epoch": 8.428925053984305,
+      "grad_norm": 1.3206424713134766,
+      "learning_rate": 1.1915365149972324e-05,
+      "loss": 1.4784,
+      "step": 40010
+    },
+    {
+      "epoch": 8.431031758571654,
+      "grad_norm": 1.2263844013214111,
+      "learning_rate": 1.1884047806923815e-05,
+      "loss": 1.5193,
+      "step": 40020
+    },
+    {
+      "epoch": 8.433138463159004,
+      "grad_norm": 1.1511719226837158,
+      "learning_rate": 1.1852769073786263e-05,
+      "loss": 1.5587,
+      "step": 40030
+    },
+    {
+      "epoch": 8.435245167746352,
+      "grad_norm": 1.2120052576065063,
+      "learning_rate": 1.182152896426515e-05,
+      "loss": 1.5952,
+      "step": 40040
+    },
+    {
+      "epoch": 8.437351872333702,
+      "grad_norm": 1.214147925376892,
+      "learning_rate": 1.1790327492049025e-05,
+      "loss": 1.6125,
+      "step": 40050
+    },
+    {
+      "epoch": 8.439458576921052,
+      "grad_norm": 1.2446225881576538,
+      "learning_rate": 1.1759164670809486e-05,
+      "loss": 1.5809,
+      "step": 40060
+    },
+    {
+      "epoch": 8.441565281508401,
+      "grad_norm": 1.2550058364868164,
+      "learning_rate": 1.1728040514201144e-05,
+      "loss": 1.5168,
+      "step": 40070
+    },
+    {
+      "epoch": 8.44367198609575,
+      "grad_norm": 1.2683697938919067,
+      "learning_rate": 1.169695503586179e-05,
+      "loss": 1.5667,
+      "step": 40080
+    },
+    {
+      "epoch": 8.445778690683099,
+      "grad_norm": 1.1972013711929321,
+      "learning_rate": 1.1665908249412161e-05,
+      "loss": 1.5599,
+      "step": 40090
+    },
+    {
+      "epoch": 8.447885395270449,
+      "grad_norm": 1.1732401847839355,
+      "learning_rate": 1.163490016845611e-05,
+      "loss": 1.5032,
+      "step": 40100
+    },
+    {
+      "epoch": 8.449992099857797,
+      "grad_norm": 1.2353123426437378,
+      "learning_rate": 1.1603930806580444e-05,
+      "loss": 1.6133,
+      "step": 40110
+    },
+    {
+      "epoch": 8.452098804445146,
+      "grad_norm": 1.1491025686264038,
+      "learning_rate": 1.1573000177355086e-05,
+      "loss": 1.5902,
+      "step": 40120
+    },
+    {
+      "epoch": 8.454205509032496,
+      "grad_norm": 1.2281687259674072,
+      "learning_rate": 1.1542108294332998e-05,
+      "loss": 1.5905,
+      "step": 40130
+    },
+    {
+      "epoch": 8.456312213619846,
+      "grad_norm": 1.223631501197815,
+      "learning_rate": 1.1511255171050084e-05,
+      "loss": 1.5575,
+      "step": 40140
+    },
+    {
+      "epoch": 8.458418918207194,
+      "grad_norm": 1.2047030925750732,
+      "learning_rate": 1.1480440821025296e-05,
+      "loss": 1.5834,
+      "step": 40150
+    },
+    {
+      "epoch": 8.460525622794544,
+      "grad_norm": 1.2944790124893188,
+      "learning_rate": 1.1449665257760656e-05,
+      "loss": 1.5372,
+      "step": 40160
+    },
+    {
+      "epoch": 8.462632327381893,
+      "grad_norm": 1.3119227886199951,
+      "learning_rate": 1.1418928494741087e-05,
+      "loss": 1.5558,
+      "step": 40170
+    },
+    {
+      "epoch": 8.464739031969241,
+      "grad_norm": 1.5002557039260864,
+      "learning_rate": 1.1388230545434653e-05,
+      "loss": 1.5156,
+      "step": 40180
+    },
+    {
+      "epoch": 8.466845736556591,
+      "grad_norm": 1.2909941673278809,
+      "learning_rate": 1.1357571423292213e-05,
+      "loss": 1.6109,
+      "step": 40190
+    },
+    {
+      "epoch": 8.46895244114394,
+      "grad_norm": 1.1943180561065674,
+      "learning_rate": 1.1326951141747788e-05,
+      "loss": 1.548,
+      "step": 40200
+    },
+    {
+      "epoch": 8.47105914573129,
+      "grad_norm": 1.3649312257766724,
+      "learning_rate": 1.1296369714218324e-05,
+      "loss": 1.5584,
+      "step": 40210
+    },
+    {
+      "epoch": 8.473165850318638,
+      "grad_norm": 1.2651944160461426,
+      "learning_rate": 1.1265827154103703e-05,
+      "loss": 1.6027,
+      "step": 40220
+    },
+    {
+      "epoch": 8.475272554905988,
+      "grad_norm": 1.1884119510650635,
+      "learning_rate": 1.1235323474786841e-05,
+      "loss": 1.554,
+      "step": 40230
+    },
+    {
+      "epoch": 8.477379259493338,
+      "grad_norm": 1.2143009901046753,
+      "learning_rate": 1.120485868963358e-05,
+      "loss": 1.5674,
+      "step": 40240
+    },
+    {
+      "epoch": 8.479485964080688,
+      "grad_norm": 1.1663596630096436,
+      "learning_rate": 1.1174432811992685e-05,
+      "loss": 1.5234,
+      "step": 40250
+    },
+    {
+      "epoch": 8.481592668668036,
+      "grad_norm": 1.1769542694091797,
+      "learning_rate": 1.1144045855195973e-05,
+      "loss": 1.5517,
+      "step": 40260
+    },
+    {
+      "epoch": 8.483699373255385,
+      "grad_norm": 1.2329270839691162,
+      "learning_rate": 1.1113697832558101e-05,
+      "loss": 1.5428,
+      "step": 40270
+    },
+    {
+      "epoch": 8.485806077842735,
+      "grad_norm": 1.2631030082702637,
+      "learning_rate": 1.1083388757376712e-05,
+      "loss": 1.5605,
+      "step": 40280
+    },
+    {
+      "epoch": 8.487912782430083,
+      "grad_norm": 1.1870076656341553,
+      "learning_rate": 1.1053118642932425e-05,
+      "loss": 1.5147,
+      "step": 40290
+    },
+    {
+      "epoch": 8.490019487017433,
+      "grad_norm": 1.1924930810928345,
+      "learning_rate": 1.1022887502488688e-05,
+      "loss": 1.5811,
+      "step": 40300
+    },
+    {
+      "epoch": 8.492126191604783,
+      "grad_norm": 1.269508719444275,
+      "learning_rate": 1.0992695349291981e-05,
+      "loss": 1.5839,
+      "step": 40310
+    },
+    {
+      "epoch": 8.494232896192132,
+      "grad_norm": 1.3270164728164673,
+      "learning_rate": 1.0962542196571634e-05,
+      "loss": 1.5849,
+      "step": 40320
+    },
+    {
+      "epoch": 8.49633960077948,
+      "grad_norm": 1.3483185768127441,
+      "learning_rate": 1.0932428057539879e-05,
+      "loss": 1.545,
+      "step": 40330
+    },
+    {
+      "epoch": 8.49844630536683,
+      "grad_norm": 1.2481123208999634,
+      "learning_rate": 1.0902352945391903e-05,
+      "loss": 1.5465,
+      "step": 40340
+    },
+    {
+      "epoch": 8.50055300995418,
+      "grad_norm": 1.2228620052337646,
+      "learning_rate": 1.0872316873305766e-05,
+      "loss": 1.5942,
+      "step": 40350
+    },
+    {
+      "epoch": 8.502659714541528,
+      "grad_norm": 1.1880698204040527,
+      "learning_rate": 1.0842319854442395e-05,
+      "loss": 1.533,
+      "step": 40360
+    },
+    {
+      "epoch": 8.504766419128877,
+      "grad_norm": 1.247296929359436,
+      "learning_rate": 1.0812361901945678e-05,
+      "loss": 1.5614,
+      "step": 40370
+    },
+    {
+      "epoch": 8.506873123716227,
+      "grad_norm": 1.2996197938919067,
+      "learning_rate": 1.078244302894229e-05,
+      "loss": 1.5818,
+      "step": 40380
+    },
+    {
+      "epoch": 8.508979828303577,
+      "grad_norm": 1.2891303300857544,
+      "learning_rate": 1.0752563248541891e-05,
+      "loss": 1.5511,
+      "step": 40390
+    },
+    {
+      "epoch": 8.511086532890925,
+      "grad_norm": 1.3689333200454712,
+      "learning_rate": 1.0722722573836907e-05,
+      "loss": 1.611,
+      "step": 40400
+    },
+    {
+      "epoch": 8.513193237478275,
+      "grad_norm": 1.157731533050537,
+      "learning_rate": 1.069292101790268e-05,
+      "loss": 1.524,
+      "step": 40410
+    },
+    {
+      "epoch": 8.515299942065624,
+      "grad_norm": 1.2392263412475586,
+      "learning_rate": 1.0663158593797428e-05,
+      "loss": 1.5202,
+      "step": 40420
+    },
+    {
+      "epoch": 8.517406646652972,
+      "grad_norm": 1.198107123374939,
+      "learning_rate": 1.063343531456219e-05,
+      "loss": 1.5084,
+      "step": 40430
+    },
+    {
+      "epoch": 8.519513351240322,
+      "grad_norm": 1.1941255331039429,
+      "learning_rate": 1.0603751193220846e-05,
+      "loss": 1.5902,
+      "step": 40440
+    },
+    {
+      "epoch": 8.521620055827672,
+      "grad_norm": 1.1136577129364014,
+      "learning_rate": 1.0574106242780179e-05,
+      "loss": 1.5638,
+      "step": 40450
+    },
+    {
+      "epoch": 8.523726760415022,
+      "grad_norm": 1.2597277164459229,
+      "learning_rate": 1.0544500476229713e-05,
+      "loss": 1.5807,
+      "step": 40460
+    },
+    {
+      "epoch": 8.52583346500237,
+      "grad_norm": 1.1556949615478516,
+      "learning_rate": 1.0514933906541901e-05,
+      "loss": 1.5527,
+      "step": 40470
+    },
+    {
+      "epoch": 8.52794016958972,
+      "grad_norm": 1.279253363609314,
+      "learning_rate": 1.048540654667195e-05,
+      "loss": 1.5874,
+      "step": 40480
+    },
+    {
+      "epoch": 8.530046874177069,
+      "grad_norm": 1.2556148767471313,
+      "learning_rate": 1.0455918409557908e-05,
+      "loss": 1.5802,
+      "step": 40490
+    },
+    {
+      "epoch": 8.532153578764417,
+      "grad_norm": 1.2259470224380493,
+      "learning_rate": 1.0426469508120662e-05,
+      "loss": 1.5214,
+      "step": 40500
+    },
+    {
+      "epoch": 8.534260283351767,
+      "grad_norm": 1.1735920906066895,
+      "learning_rate": 1.0397059855263858e-05,
+      "loss": 1.6112,
+      "step": 40510
+    },
+    {
+      "epoch": 8.536366987939116,
+      "grad_norm": 1.2973119020462036,
+      "learning_rate": 1.0367689463874008e-05,
+      "loss": 1.5514,
+      "step": 40520
+    },
+    {
+      "epoch": 8.538473692526466,
+      "grad_norm": 1.1571524143218994,
+      "learning_rate": 1.0338358346820353e-05,
+      "loss": 1.548,
+      "step": 40530
+    },
+    {
+      "epoch": 8.540580397113814,
+      "grad_norm": 1.2462031841278076,
+      "learning_rate": 1.0309066516954958e-05,
+      "loss": 1.5325,
+      "step": 40540
+    },
+    {
+      "epoch": 8.542687101701164,
+      "grad_norm": 1.3027243614196777,
+      "learning_rate": 1.0279813987112696e-05,
+      "loss": 1.6096,
+      "step": 40550
+    },
+    {
+      "epoch": 8.544793806288514,
+      "grad_norm": 1.1657319068908691,
+      "learning_rate": 1.0250600770111185e-05,
+      "loss": 1.5217,
+      "step": 40560
+    },
+    {
+      "epoch": 8.546900510875862,
+      "grad_norm": 1.2053802013397217,
+      "learning_rate": 1.0221426878750805e-05,
+      "loss": 1.5174,
+      "step": 40570
+    },
+    {
+      "epoch": 8.549007215463211,
+      "grad_norm": 1.1766663789749146,
+      "learning_rate": 1.0192292325814756e-05,
+      "loss": 1.5205,
+      "step": 40580
+    },
+    {
+      "epoch": 8.551113920050561,
+      "grad_norm": 1.3153703212738037,
+      "learning_rate": 1.0163197124068957e-05,
+      "loss": 1.5878,
+      "step": 40590
+    },
+    {
+      "epoch": 8.55322062463791,
+      "grad_norm": 1.141251564025879,
+      "learning_rate": 1.013414128626211e-05,
+      "loss": 1.5941,
+      "step": 40600
+    },
+    {
+      "epoch": 8.555327329225259,
+      "grad_norm": 1.2303365468978882,
+      "learning_rate": 1.0105124825125666e-05,
+      "loss": 1.5336,
+      "step": 40610
+    },
+    {
+      "epoch": 8.557434033812608,
+      "grad_norm": 1.178717017173767,
+      "learning_rate": 1.0076147753373789e-05,
+      "loss": 1.5571,
+      "step": 40620
+    },
+    {
+      "epoch": 8.559540738399958,
+      "grad_norm": 1.3122880458831787,
+      "learning_rate": 1.004721008370344e-05,
+      "loss": 1.6165,
+      "step": 40630
+    },
+    {
+      "epoch": 8.561647442987308,
+      "grad_norm": 1.1219425201416016,
+      "learning_rate": 1.0018311828794268e-05,
+      "loss": 1.5958,
+      "step": 40640
+    },
+    {
+      "epoch": 8.563754147574656,
+      "grad_norm": 1.148552656173706,
+      "learning_rate": 9.989453001308657e-06,
+      "loss": 1.5789,
+      "step": 40650
+    },
+    {
+      "epoch": 8.565860852162006,
+      "grad_norm": 1.2436506748199463,
+      "learning_rate": 9.960633613891756e-06,
+      "loss": 1.5607,
+      "step": 40660
+    },
+    {
+      "epoch": 8.567967556749355,
+      "grad_norm": 1.113144874572754,
+      "learning_rate": 9.931853679171377e-06,
+      "loss": 1.583,
+      "step": 40670
+    },
+    {
+      "epoch": 8.570074261336703,
+      "grad_norm": 1.1895748376846313,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 1.6034,
+      "step": 40680
+    },
+    {
+      "epoch": 8.572180965924053,
+      "grad_norm": 1.2008453607559204,
+      "learning_rate": 9.874412218245155e-06,
+      "loss": 1.5242,
+      "step": 40690
+    },
+    {
+      "epoch": 8.574287670511403,
+      "grad_norm": 1.219973087310791,
+      "learning_rate": 9.845750717208502e-06,
+      "loss": 1.6285,
+      "step": 40700
+    },
+    {
+      "epoch": 8.576394375098753,
+      "grad_norm": 1.3711133003234863,
+      "learning_rate": 9.817128719206825e-06,
+      "loss": 1.5516,
+      "step": 40710
+    },
+    {
+      "epoch": 8.5785010796861,
+      "grad_norm": 1.0981019735336304,
+      "learning_rate": 9.788546236781459e-06,
+      "loss": 1.5154,
+      "step": 40720
+    },
+    {
+      "epoch": 8.58060778427345,
+      "grad_norm": 1.314899206161499,
+      "learning_rate": 9.760003282456409e-06,
+      "loss": 1.585,
+      "step": 40730
+    },
+    {
+      "epoch": 8.5827144888608,
+      "grad_norm": 1.2251036167144775,
+      "learning_rate": 9.731499868738447e-06,
+      "loss": 1.5973,
+      "step": 40740
+    },
+    {
+      "epoch": 8.584821193448148,
+      "grad_norm": 1.1139544248580933,
+      "learning_rate": 9.703036008116895e-06,
+      "loss": 1.5602,
+      "step": 40750
+    },
+    {
+      "epoch": 8.586927898035498,
+      "grad_norm": 1.4896215200424194,
+      "learning_rate": 9.674611713063864e-06,
+      "loss": 1.5935,
+      "step": 40760
+    },
+    {
+      "epoch": 8.589034602622847,
+      "grad_norm": 1.1655323505401611,
+      "learning_rate": 9.646226996034048e-06,
+      "loss": 1.5746,
+      "step": 40770
+    },
+    {
+      "epoch": 8.591141307210197,
+      "grad_norm": 1.2914824485778809,
+      "learning_rate": 9.617881869464807e-06,
+      "loss": 1.5852,
+      "step": 40780
+    },
+    {
+      "epoch": 8.593248011797545,
+      "grad_norm": 1.2509586811065674,
+      "learning_rate": 9.589576345776218e-06,
+      "loss": 1.5507,
+      "step": 40790
+    },
+    {
+      "epoch": 8.595354716384895,
+      "grad_norm": 1.3798872232437134,
+      "learning_rate": 9.561310437370907e-06,
+      "loss": 1.6163,
+      "step": 40800
+    },
+    {
+      "epoch": 8.597461420972245,
+      "grad_norm": 1.440629005432129,
+      "learning_rate": 9.533084156634242e-06,
+      "loss": 1.5344,
+      "step": 40810
+    },
+    {
+      "epoch": 8.599568125559593,
+      "grad_norm": 1.2847709655761719,
+      "learning_rate": 9.504897515934153e-06,
+      "loss": 1.5853,
+      "step": 40820
+    },
+    {
+      "epoch": 8.601674830146942,
+      "grad_norm": 1.203344702720642,
+      "learning_rate": 9.476750527621214e-06,
+      "loss": 1.5734,
+      "step": 40830
+    },
+    {
+      "epoch": 8.603781534734292,
+      "grad_norm": 1.2557659149169922,
+      "learning_rate": 9.448643204028662e-06,
+      "loss": 1.5918,
+      "step": 40840
+    },
+    {
+      "epoch": 8.605888239321642,
+      "grad_norm": 1.1673754453659058,
+      "learning_rate": 9.420575557472333e-06,
+      "loss": 1.5612,
+      "step": 40850
+    },
+    {
+      "epoch": 8.60799494390899,
+      "grad_norm": 1.145298957824707,
+      "learning_rate": 9.392547600250634e-06,
+      "loss": 1.5774,
+      "step": 40860
+    },
+    {
+      "epoch": 8.61010164849634,
+      "grad_norm": 1.214869737625122,
+      "learning_rate": 9.364559344644663e-06,
+      "loss": 1.5249,
+      "step": 40870
+    },
+    {
+      "epoch": 8.61220835308369,
+      "grad_norm": 1.2023444175720215,
+      "learning_rate": 9.336610802918044e-06,
+      "loss": 1.5852,
+      "step": 40880
+    },
+    {
+      "epoch": 8.614315057671039,
+      "grad_norm": 1.2107261419296265,
+      "learning_rate": 9.308701987317081e-06,
+      "loss": 1.5125,
+      "step": 40890
+    },
+    {
+      "epoch": 8.616421762258387,
+      "grad_norm": 1.2665753364562988,
+      "learning_rate": 9.280832910070591e-06,
+      "loss": 1.5619,
+      "step": 40900
+    },
+    {
+      "epoch": 8.618528466845737,
+      "grad_norm": 1.1741652488708496,
+      "learning_rate": 9.253003583390008e-06,
+      "loss": 1.5431,
+      "step": 40910
+    },
+    {
+      "epoch": 8.620635171433086,
+      "grad_norm": 1.2381701469421387,
+      "learning_rate": 9.225214019469385e-06,
+      "loss": 1.5154,
+      "step": 40920
+    },
+    {
+      "epoch": 8.622741876020434,
+      "grad_norm": 1.2759102582931519,
+      "learning_rate": 9.197464230485298e-06,
+      "loss": 1.566,
+      "step": 40930
+    },
+    {
+      "epoch": 8.624848580607784,
+      "grad_norm": 1.186509370803833,
+      "learning_rate": 9.169754228596905e-06,
+      "loss": 1.5888,
+      "step": 40940
+    },
+    {
+      "epoch": 8.626955285195134,
+      "grad_norm": 1.198004126548767,
+      "learning_rate": 9.142084025945984e-06,
+      "loss": 1.5688,
+      "step": 40950
+    },
+    {
+      "epoch": 8.629061989782482,
+      "grad_norm": 1.2855273485183716,
+      "learning_rate": 9.114453634656783e-06,
+      "loss": 1.5654,
+      "step": 40960
+    },
+    {
+      "epoch": 8.631168694369832,
+      "grad_norm": 1.3223928213119507,
+      "learning_rate": 9.086863066836203e-06,
+      "loss": 1.5425,
+      "step": 40970
+    },
+    {
+      "epoch": 8.633275398957181,
+      "grad_norm": 1.2067506313323975,
+      "learning_rate": 9.059312334573633e-06,
+      "loss": 1.5451,
+      "step": 40980
+    },
+    {
+      "epoch": 8.635382103544531,
+      "grad_norm": 1.1481016874313354,
+      "learning_rate": 9.031801449940991e-06,
+      "loss": 1.578,
+      "step": 40990
+    },
+    {
+      "epoch": 8.637488808131879,
+      "grad_norm": 1.2925342321395874,
+      "learning_rate": 9.004330424992813e-06,
+      "loss": 1.5271,
+      "step": 41000
+    },
+    {
+      "epoch": 8.639595512719229,
+      "grad_norm": 1.3120172023773193,
+      "learning_rate": 8.976899271766092e-06,
+      "loss": 1.5395,
+      "step": 41010
+    },
+    {
+      "epoch": 8.641702217306579,
+      "grad_norm": 1.2305859327316284,
+      "learning_rate": 8.949508002280382e-06,
+      "loss": 1.5323,
+      "step": 41020
+    },
+    {
+      "epoch": 8.643808921893928,
+      "grad_norm": 1.2865490913391113,
+      "learning_rate": 8.922156628537792e-06,
+      "loss": 1.56,
+      "step": 41030
+    },
+    {
+      "epoch": 8.645915626481276,
+      "grad_norm": 1.1152595281600952,
+      "learning_rate": 8.89484516252287e-06,
+      "loss": 1.544,
+      "step": 41040
+    },
+    {
+      "epoch": 8.648022331068626,
+      "grad_norm": 1.2578926086425781,
+      "learning_rate": 8.867573616202751e-06,
+      "loss": 1.5684,
+      "step": 41050
+    },
+    {
+      "epoch": 8.650129035655976,
+      "grad_norm": 1.2378671169281006,
+      "learning_rate": 8.840342001527091e-06,
+      "loss": 1.5555,
+      "step": 41060
+    },
+    {
+      "epoch": 8.652235740243324,
+      "grad_norm": 1.2699627876281738,
+      "learning_rate": 8.813150330427945e-06,
+      "loss": 1.5894,
+      "step": 41070
+    },
+    {
+      "epoch": 8.654342444830673,
+      "grad_norm": 1.2434086799621582,
+      "learning_rate": 8.785998614819957e-06,
+      "loss": 1.4924,
+      "step": 41080
+    },
+    {
+      "epoch": 8.656449149418023,
+      "grad_norm": 1.194634199142456,
+      "learning_rate": 8.758886866600257e-06,
+      "loss": 1.5675,
+      "step": 41090
+    },
+    {
+      "epoch": 8.658555854005373,
+      "grad_norm": 1.182447910308838,
+      "learning_rate": 8.731815097648433e-06,
+      "loss": 1.5356,
+      "step": 41100
+    },
+    {
+      "epoch": 8.66066255859272,
+      "grad_norm": 1.409224033355713,
+      "learning_rate": 8.70478331982656e-06,
+      "loss": 1.5601,
+      "step": 41110
+    },
+    {
+      "epoch": 8.66276926318007,
+      "grad_norm": 1.3378043174743652,
+      "learning_rate": 8.677791544979174e-06,
+      "loss": 1.6083,
+      "step": 41120
+    },
+    {
+      "epoch": 8.66487596776742,
+      "grad_norm": 1.2825632095336914,
+      "learning_rate": 8.65083978493334e-06,
+      "loss": 1.5947,
+      "step": 41130
+    },
+    {
+      "epoch": 8.666982672354768,
+      "grad_norm": 1.223988652229309,
+      "learning_rate": 8.623928051498575e-06,
+      "loss": 1.593,
+      "step": 41140
+    },
+    {
+      "epoch": 8.669089376942118,
+      "grad_norm": 1.2797350883483887,
+      "learning_rate": 8.597056356466771e-06,
+      "loss": 1.5612,
+      "step": 41150
+    },
+    {
+      "epoch": 8.671196081529468,
+      "grad_norm": 1.1674673557281494,
+      "learning_rate": 8.570224711612385e-06,
+      "loss": 1.5863,
+      "step": 41160
+    },
+    {
+      "epoch": 8.673302786116817,
+      "grad_norm": 1.359400987625122,
+      "learning_rate": 8.543433128692291e-06,
+      "loss": 1.5858,
+      "step": 41170
+    },
+    {
+      "epoch": 8.675409490704165,
+      "grad_norm": 1.2248836755752563,
+      "learning_rate": 8.516681619445788e-06,
+      "loss": 1.5338,
+      "step": 41180
+    },
+    {
+      "epoch": 8.677516195291515,
+      "grad_norm": 1.2979825735092163,
+      "learning_rate": 8.489970195594632e-06,
+      "loss": 1.6042,
+      "step": 41190
+    },
+    {
+      "epoch": 8.679622899878865,
+      "grad_norm": 1.1553622484207153,
+      "learning_rate": 8.463298868842972e-06,
+      "loss": 1.5417,
+      "step": 41200
+    },
+    {
+      "epoch": 8.681729604466213,
+      "grad_norm": 1.1689449548721313,
+      "learning_rate": 8.436667650877472e-06,
+      "loss": 1.5669,
+      "step": 41210
+    },
+    {
+      "epoch": 8.683836309053563,
+      "grad_norm": 1.2584521770477295,
+      "learning_rate": 8.410076553367208e-06,
+      "loss": 1.5273,
+      "step": 41220
+    },
+    {
+      "epoch": 8.685943013640912,
+      "grad_norm": 1.209145188331604,
+      "learning_rate": 8.383525587963558e-06,
+      "loss": 1.5444,
+      "step": 41230
+    },
+    {
+      "epoch": 8.688049718228262,
+      "grad_norm": 1.187502145767212,
+      "learning_rate": 8.357014766300441e-06,
+      "loss": 1.5391,
+      "step": 41240
+    },
+    {
+      "epoch": 8.69015642281561,
+      "grad_norm": 1.2135961055755615,
+      "learning_rate": 8.330544099994187e-06,
+      "loss": 1.5121,
+      "step": 41250
+    },
+    {
+      "epoch": 8.69226312740296,
+      "grad_norm": 1.060791254043579,
+      "learning_rate": 8.304113600643438e-06,
+      "loss": 1.5645,
+      "step": 41260
+    },
+    {
+      "epoch": 8.69436983199031,
+      "grad_norm": 1.342658519744873,
+      "learning_rate": 8.27772327982932e-06,
+      "loss": 1.5468,
+      "step": 41270
+    },
+    {
+      "epoch": 8.69647653657766,
+      "grad_norm": 1.1672006845474243,
+      "learning_rate": 8.251373149115293e-06,
+      "loss": 1.5637,
+      "step": 41280
+    },
+    {
+      "epoch": 8.698583241165007,
+      "grad_norm": 1.3452717065811157,
+      "learning_rate": 8.225063220047246e-06,
+      "loss": 1.5927,
+      "step": 41290
+    },
+    {
+      "epoch": 8.700689945752357,
+      "grad_norm": 1.2547844648361206,
+      "learning_rate": 8.19879350415349e-06,
+      "loss": 1.5255,
+      "step": 41300
+    },
+    {
+      "epoch": 8.702796650339707,
+      "grad_norm": 1.1102582216262817,
+      "learning_rate": 8.172564012944595e-06,
+      "loss": 1.4948,
+      "step": 41310
+    },
+    {
+      "epoch": 8.704903354927055,
+      "grad_norm": 1.167216420173645,
+      "learning_rate": 8.146374757913622e-06,
+      "loss": 1.5291,
+      "step": 41320
+    },
+    {
+      "epoch": 8.707010059514404,
+      "grad_norm": 1.2935529947280884,
+      "learning_rate": 8.12022575053597e-06,
+      "loss": 1.582,
+      "step": 41330
+    },
+    {
+      "epoch": 8.709116764101754,
+      "grad_norm": 1.1825650930404663,
+      "learning_rate": 8.094117002269363e-06,
+      "loss": 1.6038,
+      "step": 41340
+    },
+    {
+      "epoch": 8.711223468689102,
+      "grad_norm": 1.3403667211532593,
+      "learning_rate": 8.068048524553961e-06,
+      "loss": 1.5006,
+      "step": 41350
+    },
+    {
+      "epoch": 8.713330173276452,
+      "grad_norm": 1.3244026899337769,
+      "learning_rate": 8.042020328812161e-06,
+      "loss": 1.5853,
+      "step": 41360
+    },
+    {
+      "epoch": 8.715436877863802,
+      "grad_norm": 1.2212400436401367,
+      "learning_rate": 8.016032426448817e-06,
+      "loss": 1.4957,
+      "step": 41370
+    },
+    {
+      "epoch": 8.717543582451151,
+      "grad_norm": 1.154133915901184,
+      "learning_rate": 7.990084828851108e-06,
+      "loss": 1.5267,
+      "step": 41380
+    },
+    {
+      "epoch": 8.7196502870385,
+      "grad_norm": 1.243924617767334,
+      "learning_rate": 7.96417754738853e-06,
+      "loss": 1.6473,
+      "step": 41390
+    },
+    {
+      "epoch": 8.721756991625849,
+      "grad_norm": 1.189622402191162,
+      "learning_rate": 7.938310593412879e-06,
+      "loss": 1.5284,
+      "step": 41400
+    },
+    {
+      "epoch": 8.723863696213199,
+      "grad_norm": 1.4922682046890259,
+      "learning_rate": 7.912483978258367e-06,
+      "loss": 1.6112,
+      "step": 41410
+    },
+    {
+      "epoch": 8.725970400800549,
+      "grad_norm": 1.2274786233901978,
+      "learning_rate": 7.886697713241453e-06,
+      "loss": 1.5704,
+      "step": 41420
+    },
+    {
+      "epoch": 8.728077105387896,
+      "grad_norm": 1.1772549152374268,
+      "learning_rate": 7.860951809660989e-06,
+      "loss": 1.552,
+      "step": 41430
+    },
+    {
+      "epoch": 8.730183809975246,
+      "grad_norm": 1.0843391418457031,
+      "learning_rate": 7.835246278798037e-06,
+      "loss": 1.5429,
+      "step": 41440
+    },
+    {
+      "epoch": 8.732290514562596,
+      "grad_norm": 1.07290780544281,
+      "learning_rate": 7.809581131916066e-06,
+      "loss": 1.5829,
+      "step": 41450
+    },
+    {
+      "epoch": 8.734397219149944,
+      "grad_norm": 1.1702159643173218,
+      "learning_rate": 7.783956380260837e-06,
+      "loss": 1.5034,
+      "step": 41460
+    },
+    {
+      "epoch": 8.736503923737294,
+      "grad_norm": 1.2474020719528198,
+      "learning_rate": 7.758372035060357e-06,
+      "loss": 1.5376,
+      "step": 41470
+    },
+    {
+      "epoch": 8.738610628324643,
+      "grad_norm": 1.2100849151611328,
+      "learning_rate": 7.73282810752497e-06,
+      "loss": 1.5916,
+      "step": 41480
+    },
+    {
+      "epoch": 8.740717332911993,
+      "grad_norm": 1.3087376356124878,
+      "learning_rate": 7.70732460884731e-06,
+      "loss": 1.5164,
+      "step": 41490
+    },
+    {
+      "epoch": 8.742824037499341,
+      "grad_norm": 1.180748701095581,
+      "learning_rate": 7.681861550202252e-06,
+      "loss": 1.5888,
+      "step": 41500
+    },
+    {
+      "epoch": 8.74493074208669,
+      "grad_norm": 1.3653792142868042,
+      "learning_rate": 7.656438942747058e-06,
+      "loss": 1.5426,
+      "step": 41510
+    },
+    {
+      "epoch": 8.74703744667404,
+      "grad_norm": 1.3318077325820923,
+      "learning_rate": 7.631056797621106e-06,
+      "loss": 1.6475,
+      "step": 41520
+    },
+    {
+      "epoch": 8.749144151261389,
+      "grad_norm": 1.1052581071853638,
+      "learning_rate": 7.605715125946178e-06,
+      "loss": 1.4625,
+      "step": 41530
+    },
+    {
+      "epoch": 8.751250855848738,
+      "grad_norm": 1.2538378238677979,
+      "learning_rate": 7.5804139388262915e-06,
+      "loss": 1.5603,
+      "step": 41540
+    },
+    {
+      "epoch": 8.753357560436088,
+      "grad_norm": 1.4633102416992188,
+      "learning_rate": 7.5551532473476795e-06,
+      "loss": 1.5634,
+      "step": 41550
+    },
+    {
+      "epoch": 8.755464265023438,
+      "grad_norm": 1.199808120727539,
+      "learning_rate": 7.529933062578864e-06,
+      "loss": 1.5058,
+      "step": 41560
+    },
+    {
+      "epoch": 8.757570969610786,
+      "grad_norm": 1.220557689666748,
+      "learning_rate": 7.504753395570629e-06,
+      "loss": 1.574,
+      "step": 41570
+    },
+    {
+      "epoch": 8.759677674198135,
+      "grad_norm": 1.2284119129180908,
+      "learning_rate": 7.479614257355971e-06,
+      "loss": 1.4977,
+      "step": 41580
+    },
+    {
+      "epoch": 8.761784378785485,
+      "grad_norm": 1.1737325191497803,
+      "learning_rate": 7.454515658950167e-06,
+      "loss": 1.5178,
+      "step": 41590
+    },
+    {
+      "epoch": 8.763891083372833,
+      "grad_norm": 1.3059794902801514,
+      "learning_rate": 7.42945761135071e-06,
+      "loss": 1.6303,
+      "step": 41600
+    },
+    {
+      "epoch": 8.765997787960183,
+      "grad_norm": 1.3335444927215576,
+      "learning_rate": 7.404440125537293e-06,
+      "loss": 1.501,
+      "step": 41610
+    },
+    {
+      "epoch": 8.768104492547533,
+      "grad_norm": 1.2088760137557983,
+      "learning_rate": 7.379463212471915e-06,
+      "loss": 1.5628,
+      "step": 41620
+    },
+    {
+      "epoch": 8.770211197134882,
+      "grad_norm": 1.393725037574768,
+      "learning_rate": 7.354526883098711e-06,
+      "loss": 1.572,
+      "step": 41630
+    },
+    {
+      "epoch": 8.77231790172223,
+      "grad_norm": 1.21114182472229,
+      "learning_rate": 7.329631148344118e-06,
+      "loss": 1.5486,
+      "step": 41640
+    },
+    {
+      "epoch": 8.77442460630958,
+      "grad_norm": 1.2755677700042725,
+      "learning_rate": 7.30477601911671e-06,
+      "loss": 1.5395,
+      "step": 41650
+    },
+    {
+      "epoch": 8.77653131089693,
+      "grad_norm": 1.1830251216888428,
+      "learning_rate": 7.279961506307287e-06,
+      "loss": 1.5716,
+      "step": 41660
+    },
+    {
+      "epoch": 8.77863801548428,
+      "grad_norm": 1.2919485569000244,
+      "learning_rate": 7.255187620788894e-06,
+      "loss": 1.5423,
+      "step": 41670
+    },
+    {
+      "epoch": 8.780744720071628,
+      "grad_norm": 1.2150613069534302,
+      "learning_rate": 7.230454373416739e-06,
+      "loss": 1.6086,
+      "step": 41680
+    },
+    {
+      "epoch": 8.782851424658977,
+      "grad_norm": 1.2287827730178833,
+      "learning_rate": 7.205761775028197e-06,
+      "loss": 1.6237,
+      "step": 41690
+    },
+    {
+      "epoch": 8.784958129246327,
+      "grad_norm": 1.1854701042175293,
+      "learning_rate": 7.181109836442912e-06,
+      "loss": 1.6204,
+      "step": 41700
+    },
+    {
+      "epoch": 8.787064833833675,
+      "grad_norm": 1.306795358657837,
+      "learning_rate": 7.156498568462633e-06,
+      "loss": 1.5797,
+      "step": 41710
+    },
+    {
+      "epoch": 8.789171538421025,
+      "grad_norm": 1.2896075248718262,
+      "learning_rate": 7.1319279818713445e-06,
+      "loss": 1.5644,
+      "step": 41720
+    },
+    {
+      "epoch": 8.791278243008374,
+      "grad_norm": 1.2205928564071655,
+      "learning_rate": 7.1073980874351575e-06,
+      "loss": 1.5979,
+      "step": 41730
+    },
+    {
+      "epoch": 8.793384947595724,
+      "grad_norm": 1.1458282470703125,
+      "learning_rate": 7.082908895902374e-06,
+      "loss": 1.5617,
+      "step": 41740
+    },
+    {
+      "epoch": 8.795491652183072,
+      "grad_norm": 1.1576645374298096,
+      "learning_rate": 7.058460418003488e-06,
+      "loss": 1.5693,
+      "step": 41750
+    },
+    {
+      "epoch": 8.797598356770422,
+      "grad_norm": 1.2268850803375244,
+      "learning_rate": 7.034052664451118e-06,
+      "loss": 1.5878,
+      "step": 41760
+    },
+    {
+      "epoch": 8.799705061357772,
+      "grad_norm": 1.1965655088424683,
+      "learning_rate": 7.009685645940023e-06,
+      "loss": 1.5876,
+      "step": 41770
+    },
+    {
+      "epoch": 8.80181176594512,
+      "grad_norm": 1.2396894693374634,
+      "learning_rate": 6.985359373147182e-06,
+      "loss": 1.5395,
+      "step": 41780
+    },
+    {
+      "epoch": 8.80391847053247,
+      "grad_norm": 1.2615655660629272,
+      "learning_rate": 6.961073856731648e-06,
+      "loss": 1.5189,
+      "step": 41790
+    },
+    {
+      "epoch": 8.806025175119819,
+      "grad_norm": 1.1825131177902222,
+      "learning_rate": 6.936829107334664e-06,
+      "loss": 1.5358,
+      "step": 41800
+    },
+    {
+      "epoch": 8.808131879707169,
+      "grad_norm": 1.2486509084701538,
+      "learning_rate": 6.9126251355795864e-06,
+      "loss": 1.571,
+      "step": 41810
+    },
+    {
+      "epoch": 8.810238584294517,
+      "grad_norm": 1.2613776922225952,
+      "learning_rate": 6.88846195207189e-06,
+      "loss": 1.5789,
+      "step": 41820
+    },
+    {
+      "epoch": 8.812345288881867,
+      "grad_norm": 1.1759439706802368,
+      "learning_rate": 6.864339567399225e-06,
+      "loss": 1.5835,
+      "step": 41830
+    },
+    {
+      "epoch": 8.814451993469216,
+      "grad_norm": 1.1128981113433838,
+      "learning_rate": 6.840257992131316e-06,
+      "loss": 1.5402,
+      "step": 41840
+    },
+    {
+      "epoch": 8.816558698056564,
+      "grad_norm": 1.1860219240188599,
+      "learning_rate": 6.816217236820032e-06,
+      "loss": 1.5533,
+      "step": 41850
+    },
+    {
+      "epoch": 8.818665402643914,
+      "grad_norm": 1.1567201614379883,
+      "learning_rate": 6.7922173119993606e-06,
+      "loss": 1.5627,
+      "step": 41860
+    },
+    {
+      "epoch": 8.820772107231264,
+      "grad_norm": 1.254657506942749,
+      "learning_rate": 6.768258228185353e-06,
+      "loss": 1.5441,
+      "step": 41870
+    },
+    {
+      "epoch": 8.822878811818613,
+      "grad_norm": 1.3093767166137695,
+      "learning_rate": 6.7443399958762584e-06,
+      "loss": 1.5616,
+      "step": 41880
+    },
+    {
+      "epoch": 8.824985516405961,
+      "grad_norm": 1.260617733001709,
+      "learning_rate": 6.720462625552326e-06,
+      "loss": 1.5495,
+      "step": 41890
+    },
+    {
+      "epoch": 8.827092220993311,
+      "grad_norm": 1.1566492319107056,
+      "learning_rate": 6.6966261276759424e-06,
+      "loss": 1.5999,
+      "step": 41900
+    },
+    {
+      "epoch": 8.82919892558066,
+      "grad_norm": 1.1674546003341675,
+      "learning_rate": 6.672830512691608e-06,
+      "loss": 1.5354,
+      "step": 41910
+    },
+    {
+      "epoch": 8.83130563016801,
+      "grad_norm": 1.2879528999328613,
+      "learning_rate": 6.64907579102586e-06,
+      "loss": 1.5497,
+      "step": 41920
+    },
+    {
+      "epoch": 8.833412334755359,
+      "grad_norm": 1.2599519491195679,
+      "learning_rate": 6.625361973087363e-06,
+      "loss": 1.5339,
+      "step": 41930
+    },
+    {
+      "epoch": 8.835519039342708,
+      "grad_norm": 1.1604090929031372,
+      "learning_rate": 6.6016890692668364e-06,
+      "loss": 1.5772,
+      "step": 41940
+    },
+    {
+      "epoch": 8.837625743930058,
+      "grad_norm": 1.200273036956787,
+      "learning_rate": 6.578057089937062e-06,
+      "loss": 1.505,
+      "step": 41950
+    },
+    {
+      "epoch": 8.839732448517406,
+      "grad_norm": 1.5028703212738037,
+      "learning_rate": 6.554466045452923e-06,
+      "loss": 1.5707,
+      "step": 41960
+    },
+    {
+      "epoch": 8.841839153104756,
+      "grad_norm": 1.3208073377609253,
+      "learning_rate": 6.530915946151339e-06,
+      "loss": 1.5414,
+      "step": 41970
+    },
+    {
+      "epoch": 8.843945857692106,
+      "grad_norm": 1.1587634086608887,
+      "learning_rate": 6.507406802351268e-06,
+      "loss": 1.5389,
+      "step": 41980
+    },
+    {
+      "epoch": 8.846052562279453,
+      "grad_norm": 1.2481507062911987,
+      "learning_rate": 6.4839386243538025e-06,
+      "loss": 1.5671,
+      "step": 41990
+    },
+    {
+      "epoch": 8.848159266866803,
+      "grad_norm": 1.1359668970108032,
+      "learning_rate": 6.460511422441984e-06,
+      "loss": 1.5504,
+      "step": 42000
+    },
+    {
+      "epoch": 8.850265971454153,
+      "grad_norm": 1.2089869976043701,
+      "learning_rate": 6.4371252068809786e-06,
+      "loss": 1.5816,
+      "step": 42010
+    },
+    {
+      "epoch": 8.852372676041503,
+      "grad_norm": 1.1908295154571533,
+      "learning_rate": 6.413779987917956e-06,
+      "loss": 1.6369,
+      "step": 42020
+    },
+    {
+      "epoch": 8.85447938062885,
+      "grad_norm": 1.216638207435608,
+      "learning_rate": 6.390475775782101e-06,
+      "loss": 1.5828,
+      "step": 42030
+    },
+    {
+      "epoch": 8.8565860852162,
+      "grad_norm": 1.127992868423462,
+      "learning_rate": 6.367212580684712e-06,
+      "loss": 1.5546,
+      "step": 42040
+    },
+    {
+      "epoch": 8.85869278980355,
+      "grad_norm": 1.1806529760360718,
+      "learning_rate": 6.343990412819023e-06,
+      "loss": 1.528,
+      "step": 42050
+    },
+    {
+      "epoch": 8.8607994943909,
+      "grad_norm": 1.1873255968093872,
+      "learning_rate": 6.320809282360319e-06,
+      "loss": 1.5194,
+      "step": 42060
+    },
+    {
+      "epoch": 8.862906198978248,
+      "grad_norm": 1.7294059991836548,
+      "learning_rate": 6.297669199465961e-06,
+      "loss": 1.5223,
+      "step": 42070
+    },
+    {
+      "epoch": 8.865012903565598,
+      "grad_norm": 1.1746385097503662,
+      "learning_rate": 6.274570174275218e-06,
+      "loss": 1.4832,
+      "step": 42080
+    },
+    {
+      "epoch": 8.867119608152947,
+      "grad_norm": 1.351119041442871,
+      "learning_rate": 6.2515122169094835e-06,
+      "loss": 1.6232,
+      "step": 42090
+    },
+    {
+      "epoch": 8.869226312740295,
+      "grad_norm": 1.2291535139083862,
+      "learning_rate": 6.2284953374720736e-06,
+      "loss": 1.5923,
+      "step": 42100
+    },
+    {
+      "epoch": 8.871333017327645,
+      "grad_norm": 1.1676898002624512,
+      "learning_rate": 6.205519546048322e-06,
+      "loss": 1.593,
+      "step": 42110
+    },
+    {
+      "epoch": 8.873439721914995,
+      "grad_norm": 1.1755423545837402,
+      "learning_rate": 6.1825848527055865e-06,
+      "loss": 1.5588,
+      "step": 42120
+    },
+    {
+      "epoch": 8.875546426502344,
+      "grad_norm": 1.2773000001907349,
+      "learning_rate": 6.159691267493206e-06,
+      "loss": 1.5397,
+      "step": 42130
+    },
+    {
+      "epoch": 8.877653131089692,
+      "grad_norm": 1.2411293983459473,
+      "learning_rate": 6.136838800442457e-06,
+      "loss": 1.5132,
+      "step": 42140
+    },
+    {
+      "epoch": 8.879759835677042,
+      "grad_norm": 1.338410496711731,
+      "learning_rate": 6.114027461566696e-06,
+      "loss": 1.5821,
+      "step": 42150
+    },
+    {
+      "epoch": 8.881866540264392,
+      "grad_norm": 1.1013096570968628,
+      "learning_rate": 6.091257260861172e-06,
+      "loss": 1.5418,
+      "step": 42160
+    },
+    {
+      "epoch": 8.88397324485174,
+      "grad_norm": 1.2717993259429932,
+      "learning_rate": 6.068528208303148e-06,
+      "loss": 1.6138,
+      "step": 42170
+    },
+    {
+      "epoch": 8.88607994943909,
+      "grad_norm": 1.244744896888733,
+      "learning_rate": 6.04584031385188e-06,
+      "loss": 1.5054,
+      "step": 42180
+    },
+    {
+      "epoch": 8.88818665402644,
+      "grad_norm": 1.3655146360397339,
+      "learning_rate": 6.0231935874484945e-06,
+      "loss": 1.5829,
+      "step": 42190
+    },
+    {
+      "epoch": 8.890293358613789,
+      "grad_norm": 1.2557286024093628,
+      "learning_rate": 6.000588039016208e-06,
+      "loss": 1.5387,
+      "step": 42200
+    },
+    {
+      "epoch": 8.892400063201137,
+      "grad_norm": 1.334020972251892,
+      "learning_rate": 5.978023678460099e-06,
+      "loss": 1.568,
+      "step": 42210
+    },
+    {
+      "epoch": 8.894506767788487,
+      "grad_norm": 1.276594638824463,
+      "learning_rate": 5.9555005156672335e-06,
+      "loss": 1.5658,
+      "step": 42220
+    },
+    {
+      "epoch": 8.896613472375837,
+      "grad_norm": 1.1930742263793945,
+      "learning_rate": 5.933018560506643e-06,
+      "loss": 1.5815,
+      "step": 42230
+    },
+    {
+      "epoch": 8.898720176963185,
+      "grad_norm": 1.10635507106781,
+      "learning_rate": 5.910577822829233e-06,
+      "loss": 1.5378,
+      "step": 42240
+    },
+    {
+      "epoch": 8.900826881550534,
+      "grad_norm": 1.0961418151855469,
+      "learning_rate": 5.88817831246794e-06,
+      "loss": 1.5494,
+      "step": 42250
+    },
+    {
+      "epoch": 8.902933586137884,
+      "grad_norm": 1.249894142150879,
+      "learning_rate": 5.865820039237624e-06,
+      "loss": 1.5608,
+      "step": 42260
+    },
+    {
+      "epoch": 8.905040290725234,
+      "grad_norm": 1.2282543182373047,
+      "learning_rate": 5.843503012934959e-06,
+      "loss": 1.5222,
+      "step": 42270
+    },
+    {
+      "epoch": 8.907146995312582,
+      "grad_norm": 1.2692981958389282,
+      "learning_rate": 5.821227243338712e-06,
+      "loss": 1.5812,
+      "step": 42280
+    },
+    {
+      "epoch": 8.909253699899931,
+      "grad_norm": 1.2558562755584717,
+      "learning_rate": 5.798992740209441e-06,
+      "loss": 1.5812,
+      "step": 42290
+    },
+    {
+      "epoch": 8.911360404487281,
+      "grad_norm": 1.233802318572998,
+      "learning_rate": 5.77679951328971e-06,
+      "loss": 1.5638,
+      "step": 42300
+    },
+    {
+      "epoch": 8.913467109074631,
+      "grad_norm": 1.3241041898727417,
+      "learning_rate": 5.75464757230395e-06,
+      "loss": 1.5635,
+      "step": 42310
+    },
+    {
+      "epoch": 8.915573813661979,
+      "grad_norm": 1.1713545322418213,
+      "learning_rate": 5.732536926958487e-06,
+      "loss": 1.5177,
+      "step": 42320
+    },
+    {
+      "epoch": 8.917680518249329,
+      "grad_norm": 1.1677508354187012,
+      "learning_rate": 5.710467586941615e-06,
+      "loss": 1.5437,
+      "step": 42330
+    },
+    {
+      "epoch": 8.919787222836678,
+      "grad_norm": 1.3364185094833374,
+      "learning_rate": 5.6884395619235085e-06,
+      "loss": 1.5431,
+      "step": 42340
+    },
+    {
+      "epoch": 8.921893927424026,
+      "grad_norm": 1.236284852027893,
+      "learning_rate": 5.66645286155616e-06,
+      "loss": 1.5379,
+      "step": 42350
+    },
+    {
+      "epoch": 8.924000632011376,
+      "grad_norm": 1.1496565341949463,
+      "learning_rate": 5.644507495473572e-06,
+      "loss": 1.5879,
+      "step": 42360
+    },
+    {
+      "epoch": 8.926107336598726,
+      "grad_norm": 1.2142751216888428,
+      "learning_rate": 5.622603473291543e-06,
+      "loss": 1.5205,
+      "step": 42370
+    },
+    {
+      "epoch": 8.928214041186074,
+      "grad_norm": 1.2585910558700562,
+      "learning_rate": 5.600740804607829e-06,
+      "loss": 1.5507,
+      "step": 42380
+    },
+    {
+      "epoch": 8.930320745773423,
+      "grad_norm": 1.2263305187225342,
+      "learning_rate": 5.5789194990020225e-06,
+      "loss": 1.512,
+      "step": 42390
+    },
+    {
+      "epoch": 8.932427450360773,
+      "grad_norm": 1.1774612665176392,
+      "learning_rate": 5.557139566035574e-06,
+      "loss": 1.5541,
+      "step": 42400
+    },
+    {
+      "epoch": 8.934534154948123,
+      "grad_norm": 1.2883888483047485,
+      "learning_rate": 5.535401015251851e-06,
+      "loss": 1.5749,
+      "step": 42410
+    },
+    {
+      "epoch": 8.936640859535471,
+      "grad_norm": 1.1521271467208862,
+      "learning_rate": 5.5137038561761115e-06,
+      "loss": 1.5371,
+      "step": 42420
+    },
+    {
+      "epoch": 8.93874756412282,
+      "grad_norm": 1.2090059518814087,
+      "learning_rate": 5.49204809831535e-06,
+      "loss": 1.5231,
+      "step": 42430
+    },
+    {
+      "epoch": 8.94085426871017,
+      "grad_norm": 1.3226497173309326,
+      "learning_rate": 5.470433751158577e-06,
+      "loss": 1.5888,
+      "step": 42440
+    },
+    {
+      "epoch": 8.94296097329752,
+      "grad_norm": 1.1438806056976318,
+      "learning_rate": 5.4488608241765494e-06,
+      "loss": 1.546,
+      "step": 42450
+    },
+    {
+      "epoch": 8.945067677884868,
+      "grad_norm": 1.3889515399932861,
+      "learning_rate": 5.427329326821906e-06,
+      "loss": 1.6614,
+      "step": 42460
+    },
+    {
+      "epoch": 8.947174382472218,
+      "grad_norm": 1.2498170137405396,
+      "learning_rate": 5.4058392685292005e-06,
+      "loss": 1.6041,
+      "step": 42470
+    },
+    {
+      "epoch": 8.949281087059568,
+      "grad_norm": 1.1849504709243774,
+      "learning_rate": 5.3843906587146886e-06,
+      "loss": 1.5202,
+      "step": 42480
+    },
+    {
+      "epoch": 8.951387791646916,
+      "grad_norm": 1.1745749711990356,
+      "learning_rate": 5.362983506776564e-06,
+      "loss": 1.5924,
+      "step": 42490
+    },
+    {
+      "epoch": 8.953494496234265,
+      "grad_norm": 1.1476824283599854,
+      "learning_rate": 5.341617822094869e-06,
+      "loss": 1.5793,
+      "step": 42500
+    },
+    {
+      "epoch": 8.955601200821615,
+      "grad_norm": 1.2073520421981812,
+      "learning_rate": 5.320293614031413e-06,
+      "loss": 1.6405,
+      "step": 42510
+    },
+    {
+      "epoch": 8.957707905408965,
+      "grad_norm": 1.2965010404586792,
+      "learning_rate": 5.299010891929856e-06,
+      "loss": 1.5816,
+      "step": 42520
+    },
+    {
+      "epoch": 8.959814609996313,
+      "grad_norm": 1.189401626586914,
+      "learning_rate": 5.277769665115695e-06,
+      "loss": 1.5539,
+      "step": 42530
+    },
+    {
+      "epoch": 8.961921314583662,
+      "grad_norm": 1.3214972019195557,
+      "learning_rate": 5.256569942896217e-06,
+      "loss": 1.5813,
+      "step": 42540
+    },
+    {
+      "epoch": 8.964028019171012,
+      "grad_norm": 1.2079973220825195,
+      "learning_rate": 5.23541173456058e-06,
+      "loss": 1.5224,
+      "step": 42550
+    },
+    {
+      "epoch": 8.96613472375836,
+      "grad_norm": 1.121533989906311,
+      "learning_rate": 5.214295049379658e-06,
+      "loss": 1.5912,
+      "step": 42560
+    },
+    {
+      "epoch": 8.96824142834571,
+      "grad_norm": 1.1464295387268066,
+      "learning_rate": 5.193219896606194e-06,
+      "loss": 1.537,
+      "step": 42570
+    },
+    {
+      "epoch": 8.97034813293306,
+      "grad_norm": 1.2208278179168701,
+      "learning_rate": 5.172186285474756e-06,
+      "loss": 1.5689,
+      "step": 42580
+    },
+    {
+      "epoch": 8.97245483752041,
+      "grad_norm": 1.1732256412506104,
+      "learning_rate": 5.15119422520165e-06,
+      "loss": 1.6011,
+      "step": 42590
+    },
+    {
+      "epoch": 8.974561542107757,
+      "grad_norm": 1.1097040176391602,
+      "learning_rate": 5.130243724984995e-06,
+      "loss": 1.5775,
+      "step": 42600
+    },
+    {
+      "epoch": 8.976668246695107,
+      "grad_norm": 1.2230794429779053,
+      "learning_rate": 5.1093347940047274e-06,
+      "loss": 1.5617,
+      "step": 42610
+    },
+    {
+      "epoch": 8.978774951282457,
+      "grad_norm": 1.1923304796218872,
+      "learning_rate": 5.0884674414225284e-06,
+      "loss": 1.5653,
+      "step": 42620
+    },
+    {
+      "epoch": 8.980881655869805,
+      "grad_norm": 1.2577967643737793,
+      "learning_rate": 5.067641676381918e-06,
+      "loss": 1.6404,
+      "step": 42630
+    },
+    {
+      "epoch": 8.982988360457155,
+      "grad_norm": 1.5352810621261597,
+      "learning_rate": 5.046857508008085e-06,
+      "loss": 1.5895,
+      "step": 42640
+    },
+    {
+      "epoch": 8.985095065044504,
+      "grad_norm": 1.1417043209075928,
+      "learning_rate": 5.026114945408123e-06,
+      "loss": 1.565,
+      "step": 42650
+    },
+    {
+      "epoch": 8.987201769631854,
+      "grad_norm": 1.1386953592300415,
+      "learning_rate": 5.005413997670816e-06,
+      "loss": 1.5226,
+      "step": 42660
+    },
+    {
+      "epoch": 8.989308474219202,
+      "grad_norm": 1.1961097717285156,
+      "learning_rate": 4.984754673866732e-06,
+      "loss": 1.5532,
+      "step": 42670
+    },
+    {
+      "epoch": 8.991415178806552,
+      "grad_norm": 1.2846159934997559,
+      "learning_rate": 4.964136983048184e-06,
+      "loss": 1.5413,
+      "step": 42680
+    },
+    {
+      "epoch": 8.993521883393901,
+      "grad_norm": 1.1944465637207031,
+      "learning_rate": 4.9435609342493025e-06,
+      "loss": 1.5316,
+      "step": 42690
+    },
+    {
+      "epoch": 8.995628587981251,
+      "grad_norm": 1.21571946144104,
+      "learning_rate": 4.923026536485875e-06,
+      "loss": 1.5421,
+      "step": 42700
+    },
+    {
+      "epoch": 8.9977352925686,
+      "grad_norm": 1.240976333618164,
+      "learning_rate": 4.902533798755548e-06,
+      "loss": 1.5161,
+      "step": 42710
+    },
+    {
+      "epoch": 8.999841997155949,
+      "grad_norm": 1.2830826044082642,
+      "learning_rate": 4.8820827300376075e-06,
+      "loss": 1.5955,
+      "step": 42720
+    },
+    {
+      "epoch": 9.001948701743299,
+      "grad_norm": 1.0986697673797607,
+      "learning_rate": 4.861673339293149e-06,
+      "loss": 1.5327,
+      "step": 42730
+    },
+    {
+      "epoch": 9.004055406330647,
+      "grad_norm": 1.1212446689605713,
+      "learning_rate": 4.84130563546501e-06,
+      "loss": 1.5482,
+      "step": 42740
+    },
+    {
+      "epoch": 9.006162110917996,
+      "grad_norm": 1.2350519895553589,
+      "learning_rate": 4.820979627477706e-06,
+      "loss": 1.5517,
+      "step": 42750
+    },
+    {
+      "epoch": 9.008268815505346,
+      "grad_norm": 1.2882555723190308,
+      "learning_rate": 4.800695324237547e-06,
+      "loss": 1.5526,
+      "step": 42760
+    },
+    {
+      "epoch": 9.010375520092696,
+      "grad_norm": 1.3195393085479736,
+      "learning_rate": 4.780452734632524e-06,
+      "loss": 1.5865,
+      "step": 42770
+    },
+    {
+      "epoch": 9.012482224680044,
+      "grad_norm": 1.2144135236740112,
+      "learning_rate": 4.760251867532362e-06,
+      "loss": 1.5783,
+      "step": 42780
+    },
+    {
+      "epoch": 9.014588929267394,
+      "grad_norm": 1.2432124614715576,
+      "learning_rate": 4.7400927317885256e-06,
+      "loss": 1.5697,
+      "step": 42790
+    },
+    {
+      "epoch": 9.016695633854743,
+      "grad_norm": 1.3771353960037231,
+      "learning_rate": 4.7199753362341614e-06,
+      "loss": 1.5572,
+      "step": 42800
+    },
+    {
+      "epoch": 9.018802338442091,
+      "grad_norm": 1.1509369611740112,
+      "learning_rate": 4.699899689684129e-06,
+      "loss": 1.5553,
+      "step": 42810
+    },
+    {
+      "epoch": 9.020909043029441,
+      "grad_norm": 1.2895727157592773,
+      "learning_rate": 4.679865800935046e-06,
+      "loss": 1.5799,
+      "step": 42820
+    },
+    {
+      "epoch": 9.02301574761679,
+      "grad_norm": 1.110680341720581,
+      "learning_rate": 4.659873678765158e-06,
+      "loss": 1.5675,
+      "step": 42830
+    },
+    {
+      "epoch": 9.02512245220414,
+      "grad_norm": 1.3074404001235962,
+      "learning_rate": 4.639923331934471e-06,
+      "loss": 1.5556,
+      "step": 42840
+    },
+    {
+      "epoch": 9.027229156791488,
+      "grad_norm": 1.2942652702331543,
+      "learning_rate": 4.620014769184644e-06,
+      "loss": 1.6094,
+      "step": 42850
+    },
+    {
+      "epoch": 9.029335861378838,
+      "grad_norm": 1.2963061332702637,
+      "learning_rate": 4.600147999239035e-06,
+      "loss": 1.6269,
+      "step": 42860
+    },
+    {
+      "epoch": 9.031442565966188,
+      "grad_norm": 1.1793369054794312,
+      "learning_rate": 4.5803230308027356e-06,
+      "loss": 1.5271,
+      "step": 42870
+    },
+    {
+      "epoch": 9.033549270553536,
+      "grad_norm": 1.3445175886154175,
+      "learning_rate": 4.560539872562463e-06,
+      "loss": 1.5342,
+      "step": 42880
+    },
+    {
+      "epoch": 9.035655975140886,
+      "grad_norm": 1.1091607809066772,
+      "learning_rate": 4.540798533186619e-06,
+      "loss": 1.5355,
+      "step": 42890
+    },
+    {
+      "epoch": 9.037762679728235,
+      "grad_norm": 1.1904314756393433,
+      "learning_rate": 4.521099021325336e-06,
+      "loss": 1.5208,
+      "step": 42900
+    },
+    {
+      "epoch": 9.039869384315585,
+      "grad_norm": 1.192204475402832,
+      "learning_rate": 4.501441345610347e-06,
+      "loss": 1.4896,
+      "step": 42910
+    },
+    {
+      "epoch": 9.041976088902933,
+      "grad_norm": 1.1834639310836792,
+      "learning_rate": 4.481825514655114e-06,
+      "loss": 1.5564,
+      "step": 42920
+    },
+    {
+      "epoch": 9.044082793490283,
+      "grad_norm": 1.3617198467254639,
+      "learning_rate": 4.462251537054718e-06,
+      "loss": 1.5556,
+      "step": 42930
+    },
+    {
+      "epoch": 9.046189498077633,
+      "grad_norm": 1.2874000072479248,
+      "learning_rate": 4.442719421385922e-06,
+      "loss": 1.5975,
+      "step": 42940
+    },
+    {
+      "epoch": 9.04829620266498,
+      "grad_norm": 1.1991859674453735,
+      "learning_rate": 4.423229176207167e-06,
+      "loss": 1.6017,
+      "step": 42950
+    },
+    {
+      "epoch": 9.05040290725233,
+      "grad_norm": 1.1703438758850098,
+      "learning_rate": 4.403780810058511e-06,
+      "loss": 1.5776,
+      "step": 42960
+    },
+    {
+      "epoch": 9.05250961183968,
+      "grad_norm": 1.2019577026367188,
+      "learning_rate": 4.38437433146166e-06,
+      "loss": 1.5743,
+      "step": 42970
+    },
+    {
+      "epoch": 9.05461631642703,
+      "grad_norm": 1.319351315498352,
+      "learning_rate": 4.365009748920012e-06,
+      "loss": 1.558,
+      "step": 42980
+    },
+    {
+      "epoch": 9.056723021014378,
+      "grad_norm": 1.1495599746704102,
+      "learning_rate": 4.345687070918559e-06,
+      "loss": 1.576,
+      "step": 42990
+    },
+    {
+      "epoch": 9.058829725601727,
+      "grad_norm": 1.2733286619186401,
+      "learning_rate": 4.326406305923958e-06,
+      "loss": 1.5244,
+      "step": 43000
+    },
+    {
+      "epoch": 9.060936430189077,
+      "grad_norm": 1.318764090538025,
+      "learning_rate": 4.307167462384498e-06,
+      "loss": 1.5456,
+      "step": 43010
+    },
+    {
+      "epoch": 9.063043134776425,
+      "grad_norm": 1.2200381755828857,
+      "learning_rate": 4.287970548730069e-06,
+      "loss": 1.5125,
+      "step": 43020
+    },
+    {
+      "epoch": 9.065149839363775,
+      "grad_norm": 1.2284795045852661,
+      "learning_rate": 4.268815573372242e-06,
+      "loss": 1.5087,
+      "step": 43030
+    },
+    {
+      "epoch": 9.067256543951125,
+      "grad_norm": 1.171825647354126,
+      "learning_rate": 4.249702544704171e-06,
+      "loss": 1.5662,
+      "step": 43040
+    },
+    {
+      "epoch": 9.069363248538474,
+      "grad_norm": 1.3023587465286255,
+      "learning_rate": 4.230631471100655e-06,
+      "loss": 1.5747,
+      "step": 43050
+    },
+    {
+      "epoch": 9.071469953125822,
+      "grad_norm": 1.1990822553634644,
+      "learning_rate": 4.211602360918099e-06,
+      "loss": 1.53,
+      "step": 43060
+    },
+    {
+      "epoch": 9.073576657713172,
+      "grad_norm": 1.2179408073425293,
+      "learning_rate": 4.192615222494489e-06,
+      "loss": 1.5735,
+      "step": 43070
+    },
+    {
+      "epoch": 9.075683362300522,
+      "grad_norm": 1.1778931617736816,
+      "learning_rate": 4.173670064149482e-06,
+      "loss": 1.5302,
+      "step": 43080
+    },
+    {
+      "epoch": 9.077790066887871,
+      "grad_norm": 1.3378112316131592,
+      "learning_rate": 4.154766894184292e-06,
+      "loss": 1.5586,
+      "step": 43090
+    },
+    {
+      "epoch": 9.07989677147522,
+      "grad_norm": 1.2669785022735596,
+      "learning_rate": 4.1359057208817605e-06,
+      "loss": 1.5415,
+      "step": 43100
+    },
+    {
+      "epoch": 9.08200347606257,
+      "grad_norm": 1.32623291015625,
+      "learning_rate": 4.117086552506322e-06,
+      "loss": 1.5398,
+      "step": 43110
+    },
+    {
+      "epoch": 9.084110180649919,
+      "grad_norm": 1.186213493347168,
+      "learning_rate": 4.098309397303978e-06,
+      "loss": 1.5541,
+      "step": 43120
+    },
+    {
+      "epoch": 9.086216885237267,
+      "grad_norm": 1.2825602293014526,
+      "learning_rate": 4.079574263502384e-06,
+      "loss": 1.5482,
+      "step": 43130
+    },
+    {
+      "epoch": 9.088323589824617,
+      "grad_norm": 1.2611690759658813,
+      "learning_rate": 4.060881159310725e-06,
+      "loss": 1.5587,
+      "step": 43140
+    },
+    {
+      "epoch": 9.090430294411966,
+      "grad_norm": 1.2552509307861328,
+      "learning_rate": 4.042230092919774e-06,
+      "loss": 1.5142,
+      "step": 43150
+    },
+    {
+      "epoch": 9.092536998999316,
+      "grad_norm": 1.2151157855987549,
+      "learning_rate": 4.023621072501926e-06,
+      "loss": 1.5874,
+      "step": 43160
+    },
+    {
+      "epoch": 9.094643703586664,
+      "grad_norm": 1.1732696294784546,
+      "learning_rate": 4.005054106211104e-06,
+      "loss": 1.5516,
+      "step": 43170
+    },
+    {
+      "epoch": 9.096750408174014,
+      "grad_norm": 1.32643461227417,
+      "learning_rate": 3.986529202182832e-06,
+      "loss": 1.6017,
+      "step": 43180
+    },
+    {
+      "epoch": 9.098857112761364,
+      "grad_norm": 1.1873167753219604,
+      "learning_rate": 3.968046368534217e-06,
+      "loss": 1.5591,
+      "step": 43190
+    },
+    {
+      "epoch": 9.100963817348712,
+      "grad_norm": 1.229538917541504,
+      "learning_rate": 3.949605613363882e-06,
+      "loss": 1.5554,
+      "step": 43200
+    },
+    {
+      "epoch": 9.103070521936061,
+      "grad_norm": 1.1299848556518555,
+      "learning_rate": 3.931206944752064e-06,
+      "loss": 1.5277,
+      "step": 43210
+    },
+    {
+      "epoch": 9.105177226523411,
+      "grad_norm": 1.3684306144714355,
+      "learning_rate": 3.912850370760534e-06,
+      "loss": 1.5113,
+      "step": 43220
+    },
+    {
+      "epoch": 9.10728393111076,
+      "grad_norm": 1.313065528869629,
+      "learning_rate": 3.894535899432606e-06,
+      "loss": 1.4936,
+      "step": 43230
+    },
+    {
+      "epoch": 9.109390635698109,
+      "grad_norm": 1.2317758798599243,
+      "learning_rate": 3.87626353879319e-06,
+      "loss": 1.5658,
+      "step": 43240
+    },
+    {
+      "epoch": 9.111497340285458,
+      "grad_norm": 1.3811818361282349,
+      "learning_rate": 3.8580332968486955e-06,
+      "loss": 1.6049,
+      "step": 43250
+    },
+    {
+      "epoch": 9.113604044872808,
+      "grad_norm": 1.2241859436035156,
+      "learning_rate": 3.839845181587098e-06,
+      "loss": 1.5565,
+      "step": 43260
+    },
+    {
+      "epoch": 9.115710749460156,
+      "grad_norm": 1.2962020635604858,
+      "learning_rate": 3.821699200977924e-06,
+      "loss": 1.551,
+      "step": 43270
+    },
+    {
+      "epoch": 9.117817454047506,
+      "grad_norm": 1.173917293548584,
+      "learning_rate": 3.8035953629722234e-06,
+      "loss": 1.5632,
+      "step": 43280
+    },
+    {
+      "epoch": 9.119924158634856,
+      "grad_norm": 1.336701512336731,
+      "learning_rate": 3.785533675502584e-06,
+      "loss": 1.5325,
+      "step": 43290
+    },
+    {
+      "epoch": 9.122030863222205,
+      "grad_norm": 1.2437199354171753,
+      "learning_rate": 3.76751414648312e-06,
+      "loss": 1.5471,
+      "step": 43300
+    },
+    {
+      "epoch": 9.124137567809553,
+      "grad_norm": 1.1865382194519043,
+      "learning_rate": 3.749536783809482e-06,
+      "loss": 1.5753,
+      "step": 43310
+    },
+    {
+      "epoch": 9.126244272396903,
+      "grad_norm": 1.3648736476898193,
+      "learning_rate": 3.7316015953588467e-06,
+      "loss": 1.5323,
+      "step": 43320
+    },
+    {
+      "epoch": 9.128350976984253,
+      "grad_norm": 1.2792103290557861,
+      "learning_rate": 3.7137085889898947e-06,
+      "loss": 1.5369,
+      "step": 43330
+    },
+    {
+      "epoch": 9.1304576815716,
+      "grad_norm": 1.1069854497909546,
+      "learning_rate": 3.6958577725428433e-06,
+      "loss": 1.5356,
+      "step": 43340
+    },
+    {
+      "epoch": 9.13256438615895,
+      "grad_norm": 1.3729944229125977,
+      "learning_rate": 3.6780491538394025e-06,
+      "loss": 1.5163,
+      "step": 43350
+    },
+    {
+      "epoch": 9.1346710907463,
+      "grad_norm": 1.1662119626998901,
+      "learning_rate": 3.6602827406828076e-06,
+      "loss": 1.5907,
+      "step": 43360
+    },
+    {
+      "epoch": 9.13677779533365,
+      "grad_norm": 1.2316755056381226,
+      "learning_rate": 3.64255854085781e-06,
+      "loss": 1.6138,
+      "step": 43370
+    },
+    {
+      "epoch": 9.138884499920998,
+      "grad_norm": 1.2182281017303467,
+      "learning_rate": 3.6248765621306414e-06,
+      "loss": 1.5445,
+      "step": 43380
+    },
+    {
+      "epoch": 9.140991204508348,
+      "grad_norm": 1.2118333578109741,
+      "learning_rate": 3.6072368122490265e-06,
+      "loss": 1.6002,
+      "step": 43390
+    },
+    {
+      "epoch": 9.143097909095697,
+      "grad_norm": 1.1032582521438599,
+      "learning_rate": 3.5896392989422377e-06,
+      "loss": 1.5135,
+      "step": 43400
+    },
+    {
+      "epoch": 9.145204613683047,
+      "grad_norm": 1.1986793279647827,
+      "learning_rate": 3.5720840299209747e-06,
+      "loss": 1.5815,
+      "step": 43410
+    },
+    {
+      "epoch": 9.147311318270395,
+      "grad_norm": 1.3122923374176025,
+      "learning_rate": 3.5545710128774835e-06,
+      "loss": 1.5405,
+      "step": 43420
+    },
+    {
+      "epoch": 9.149418022857745,
+      "grad_norm": 1.2399784326553345,
+      "learning_rate": 3.5371002554854593e-06,
+      "loss": 1.5949,
+      "step": 43430
+    },
+    {
+      "epoch": 9.151524727445095,
+      "grad_norm": 1.2310932874679565,
+      "learning_rate": 3.519671765400079e-06,
+      "loss": 1.5763,
+      "step": 43440
+    },
+    {
+      "epoch": 9.153631432032443,
+      "grad_norm": 1.2371530532836914,
+      "learning_rate": 3.502285550258044e-06,
+      "loss": 1.5614,
+      "step": 43450
+    },
+    {
+      "epoch": 9.155738136619792,
+      "grad_norm": 1.3835184574127197,
+      "learning_rate": 3.484941617677473e-06,
+      "loss": 1.6102,
+      "step": 43460
+    },
+    {
+      "epoch": 9.157844841207142,
+      "grad_norm": 1.3356512784957886,
+      "learning_rate": 3.467639975257997e-06,
+      "loss": 1.5885,
+      "step": 43470
+    },
+    {
+      "epoch": 9.159951545794492,
+      "grad_norm": 1.2307041883468628,
+      "learning_rate": 3.4503806305807074e-06,
+      "loss": 1.5764,
+      "step": 43480
+    },
+    {
+      "epoch": 9.16205825038184,
+      "grad_norm": 1.1840221881866455,
+      "learning_rate": 3.4331635912081437e-06,
+      "loss": 1.5634,
+      "step": 43490
+    },
+    {
+      "epoch": 9.16416495496919,
+      "grad_norm": 1.4013233184814453,
+      "learning_rate": 3.4159888646843495e-06,
+      "loss": 1.5445,
+      "step": 43500
+    },
+    {
+      "epoch": 9.16627165955654,
+      "grad_norm": 1.3255256414413452,
+      "learning_rate": 3.3988564585347937e-06,
+      "loss": 1.63,
+      "step": 43510
+    },
+    {
+      "epoch": 9.168378364143887,
+      "grad_norm": 1.1528620719909668,
+      "learning_rate": 3.3817663802663935e-06,
+      "loss": 1.5377,
+      "step": 43520
+    },
+    {
+      "epoch": 9.170485068731237,
+      "grad_norm": 1.2030662298202515,
+      "learning_rate": 3.364718637367548e-06,
+      "loss": 1.5362,
+      "step": 43530
+    },
+    {
+      "epoch": 9.172591773318587,
+      "grad_norm": 1.131676435470581,
+      "learning_rate": 3.3477132373081254e-06,
+      "loss": 1.5372,
+      "step": 43540
+    },
+    {
+      "epoch": 9.174698477905936,
+      "grad_norm": 1.3713462352752686,
+      "learning_rate": 3.3307501875393556e-06,
+      "loss": 1.5488,
+      "step": 43550
+    },
+    {
+      "epoch": 9.176805182493284,
+      "grad_norm": 1.1891109943389893,
+      "learning_rate": 3.313829495493992e-06,
+      "loss": 1.5883,
+      "step": 43560
+    },
+    {
+      "epoch": 9.178911887080634,
+      "grad_norm": 1.176886796951294,
+      "learning_rate": 3.2969511685862042e-06,
+      "loss": 1.5753,
+      "step": 43570
+    },
+    {
+      "epoch": 9.181018591667984,
+      "grad_norm": 1.270033597946167,
+      "learning_rate": 3.2801152142115766e-06,
+      "loss": 1.55,
+      "step": 43580
+    },
+    {
+      "epoch": 9.183125296255332,
+      "grad_norm": 1.247550368309021,
+      "learning_rate": 3.2633216397471966e-06,
+      "loss": 1.5918,
+      "step": 43590
+    },
+    {
+      "epoch": 9.185232000842682,
+      "grad_norm": 1.1211265325546265,
+      "learning_rate": 3.2465704525514785e-06,
+      "loss": 1.5428,
+      "step": 43600
+    },
+    {
+      "epoch": 9.187338705430031,
+      "grad_norm": 1.260231614112854,
+      "learning_rate": 3.2298616599643285e-06,
+      "loss": 1.5487,
+      "step": 43610
+    },
+    {
+      "epoch": 9.189445410017381,
+      "grad_norm": 1.412537932395935,
+      "learning_rate": 3.2131952693070898e-06,
+      "loss": 1.571,
+      "step": 43620
+    },
+    {
+      "epoch": 9.191552114604729,
+      "grad_norm": 1.1907163858413696,
+      "learning_rate": 3.196571287882488e-06,
+      "loss": 1.5942,
+      "step": 43630
+    },
+    {
+      "epoch": 9.193658819192079,
+      "grad_norm": 1.2359150648117065,
+      "learning_rate": 3.1799897229746857e-06,
+      "loss": 1.5795,
+      "step": 43640
+    },
+    {
+      "epoch": 9.195765523779428,
+      "grad_norm": 1.2746754884719849,
+      "learning_rate": 3.1634505818492256e-06,
+      "loss": 1.5895,
+      "step": 43650
+    },
+    {
+      "epoch": 9.197872228366776,
+      "grad_norm": 1.262993574142456,
+      "learning_rate": 3.146953871753122e-06,
+      "loss": 1.5722,
+      "step": 43660
+    },
+    {
+      "epoch": 9.199978932954126,
+      "grad_norm": 1.1132774353027344,
+      "learning_rate": 3.1304995999147713e-06,
+      "loss": 1.5522,
+      "step": 43670
+    },
+    {
+      "epoch": 9.202085637541476,
+      "grad_norm": 1.0957084894180298,
+      "learning_rate": 3.1140877735439387e-06,
+      "loss": 1.57,
+      "step": 43680
+    },
+    {
+      "epoch": 9.204192342128826,
+      "grad_norm": 1.1807489395141602,
+      "learning_rate": 3.0977183998318282e-06,
+      "loss": 1.5659,
+      "step": 43690
+    },
+    {
+      "epoch": 9.206299046716174,
+      "grad_norm": 1.3260751962661743,
+      "learning_rate": 3.0813914859510572e-06,
+      "loss": 1.5896,
+      "step": 43700
+    },
+    {
+      "epoch": 9.208405751303523,
+      "grad_norm": 1.2571806907653809,
+      "learning_rate": 3.0651070390556034e-06,
+      "loss": 1.5722,
+      "step": 43710
+    },
+    {
+      "epoch": 9.210512455890873,
+      "grad_norm": 1.2674102783203125,
+      "learning_rate": 3.048865066280848e-06,
+      "loss": 1.5034,
+      "step": 43720
+    },
+    {
+      "epoch": 9.212619160478223,
+      "grad_norm": 1.1594977378845215,
+      "learning_rate": 3.032665574743543e-06,
+      "loss": 1.5583,
+      "step": 43730
+    },
+    {
+      "epoch": 9.21472586506557,
+      "grad_norm": 1.2206066846847534,
+      "learning_rate": 3.0165085715418763e-06,
+      "loss": 1.5349,
+      "step": 43740
+    },
+    {
+      "epoch": 9.21683256965292,
+      "grad_norm": 1.282946228981018,
+      "learning_rate": 3.000394063755396e-06,
+      "loss": 1.552,
+      "step": 43750
+    },
+    {
+      "epoch": 9.21893927424027,
+      "grad_norm": 1.2245168685913086,
+      "learning_rate": 2.984322058444977e-06,
+      "loss": 1.5567,
+      "step": 43760
+    },
+    {
+      "epoch": 9.221045978827618,
+      "grad_norm": 1.1592695713043213,
+      "learning_rate": 2.9682925626529522e-06,
+      "loss": 1.5166,
+      "step": 43770
+    },
+    {
+      "epoch": 9.223152683414968,
+      "grad_norm": 1.0854359865188599,
+      "learning_rate": 2.952305583402981e-06,
+      "loss": 1.5572,
+      "step": 43780
+    },
+    {
+      "epoch": 9.225259388002318,
+      "grad_norm": 1.132101058959961,
+      "learning_rate": 2.9363611277001156e-06,
+      "loss": 1.5624,
+      "step": 43790
+    },
+    {
+      "epoch": 9.227366092589667,
+      "grad_norm": 1.1339458227157593,
+      "learning_rate": 2.9204592025307566e-06,
+      "loss": 1.4836,
+      "step": 43800
+    },
+    {
+      "epoch": 9.229472797177015,
+      "grad_norm": 1.1902509927749634,
+      "learning_rate": 2.904599814862663e-06,
+      "loss": 1.5311,
+      "step": 43810
+    },
+    {
+      "epoch": 9.231579501764365,
+      "grad_norm": 1.4040547609329224,
+      "learning_rate": 2.8887829716449876e-06,
+      "loss": 1.5296,
+      "step": 43820
+    },
+    {
+      "epoch": 9.233686206351715,
+      "grad_norm": 1.332128882408142,
+      "learning_rate": 2.87300867980822e-06,
+      "loss": 1.5542,
+      "step": 43830
+    },
+    {
+      "epoch": 9.235792910939063,
+      "grad_norm": 1.20232355594635,
+      "learning_rate": 2.857276946264198e-06,
+      "loss": 1.5255,
+      "step": 43840
+    },
+    {
+      "epoch": 9.237899615526413,
+      "grad_norm": 1.3149586915969849,
+      "learning_rate": 2.8415877779061182e-06,
+      "loss": 1.5336,
+      "step": 43850
+    },
+    {
+      "epoch": 9.240006320113762,
+      "grad_norm": 1.2837882041931152,
+      "learning_rate": 2.8259411816085492e-06,
+      "loss": 1.5487,
+      "step": 43860
+    },
+    {
+      "epoch": 9.242113024701112,
+      "grad_norm": 1.4165587425231934,
+      "learning_rate": 2.810337164227361e-06,
+      "loss": 1.5605,
+      "step": 43870
+    },
+    {
+      "epoch": 9.24421972928846,
+      "grad_norm": 1.1678107976913452,
+      "learning_rate": 2.79477573259983e-06,
+      "loss": 1.5262,
+      "step": 43880
+    },
+    {
+      "epoch": 9.24632643387581,
+      "grad_norm": 1.2908071279525757,
+      "learning_rate": 2.7792568935444796e-06,
+      "loss": 1.6054,
+      "step": 43890
+    },
+    {
+      "epoch": 9.24843313846316,
+      "grad_norm": 1.2241575717926025,
+      "learning_rate": 2.7637806538612586e-06,
+      "loss": 1.5645,
+      "step": 43900
+    },
+    {
+      "epoch": 9.250539843050507,
+      "grad_norm": 1.18755304813385,
+      "learning_rate": 2.748347020331421e-06,
+      "loss": 1.5223,
+      "step": 43910
+    },
+    {
+      "epoch": 9.252646547637857,
+      "grad_norm": 1.396857738494873,
+      "learning_rate": 2.732955999717546e-06,
+      "loss": 1.5947,
+      "step": 43920
+    },
+    {
+      "epoch": 9.254753252225207,
+      "grad_norm": 1.3757117986679077,
+      "learning_rate": 2.717607598763505e-06,
+      "loss": 1.5703,
+      "step": 43930
+    },
+    {
+      "epoch": 9.256859956812557,
+      "grad_norm": 1.3316391706466675,
+      "learning_rate": 2.702301824194586e-06,
+      "loss": 1.4793,
+      "step": 43940
+    },
+    {
+      "epoch": 9.258966661399905,
+      "grad_norm": 1.4265036582946777,
+      "learning_rate": 2.687038682717302e-06,
+      "loss": 1.5784,
+      "step": 43950
+    },
+    {
+      "epoch": 9.261073365987254,
+      "grad_norm": 1.159681797027588,
+      "learning_rate": 2.6718181810195696e-06,
+      "loss": 1.5113,
+      "step": 43960
+    },
+    {
+      "epoch": 9.263180070574604,
+      "grad_norm": 1.3730366230010986,
+      "learning_rate": 2.656640325770543e-06,
+      "loss": 1.4718,
+      "step": 43970
+    },
+    {
+      "epoch": 9.265286775161952,
+      "grad_norm": 1.3480654954910278,
+      "learning_rate": 2.6415051236207355e-06,
+      "loss": 1.5478,
+      "step": 43980
+    },
+    {
+      "epoch": 9.267393479749302,
+      "grad_norm": 1.2406569719314575,
+      "learning_rate": 2.626412581201987e-06,
+      "loss": 1.5258,
+      "step": 43990
+    },
+    {
+      "epoch": 9.269500184336652,
+      "grad_norm": 1.192407250404358,
+      "learning_rate": 2.6113627051273957e-06,
+      "loss": 1.591,
+      "step": 44000
+    },
+    {
+      "epoch": 9.271606888924001,
+      "grad_norm": 1.2141858339309692,
+      "learning_rate": 2.5963555019913988e-06,
+      "loss": 1.6122,
+      "step": 44010
+    },
+    {
+      "epoch": 9.27371359351135,
+      "grad_norm": 1.2186552286148071,
+      "learning_rate": 2.5813909783697354e-06,
+      "loss": 1.5491,
+      "step": 44020
+    },
+    {
+      "epoch": 9.275820298098699,
+      "grad_norm": 1.2834081649780273,
+      "learning_rate": 2.5664691408194165e-06,
+      "loss": 1.5743,
+      "step": 44030
+    },
+    {
+      "epoch": 9.277927002686049,
+      "grad_norm": 1.265353798866272,
+      "learning_rate": 2.551589995878789e-06,
+      "loss": 1.5876,
+      "step": 44040
+    },
+    {
+      "epoch": 9.280033707273397,
+      "grad_norm": 1.170989751815796,
+      "learning_rate": 2.5367535500674724e-06,
+      "loss": 1.5179,
+      "step": 44050
+    },
+    {
+      "epoch": 9.282140411860746,
+      "grad_norm": 1.162787914276123,
+      "learning_rate": 2.521959809886343e-06,
+      "loss": 1.5376,
+      "step": 44060
+    },
+    {
+      "epoch": 9.284247116448096,
+      "grad_norm": 1.1817100048065186,
+      "learning_rate": 2.5072087818176382e-06,
+      "loss": 1.5615,
+      "step": 44070
+    },
+    {
+      "epoch": 9.286353821035446,
+      "grad_norm": 1.217679500579834,
+      "learning_rate": 2.4925004723248323e-06,
+      "loss": 1.5643,
+      "step": 44080
+    },
+    {
+      "epoch": 9.288460525622794,
+      "grad_norm": 1.2032785415649414,
+      "learning_rate": 2.477834887852659e-06,
+      "loss": 1.5034,
+      "step": 44090
+    },
+    {
+      "epoch": 9.290567230210144,
+      "grad_norm": 1.3014835119247437,
+      "learning_rate": 2.4632120348272003e-06,
+      "loss": 1.4977,
+      "step": 44100
+    },
+    {
+      "epoch": 9.292673934797493,
+      "grad_norm": 1.2489004135131836,
+      "learning_rate": 2.4486319196557416e-06,
+      "loss": 1.5721,
+      "step": 44110
+    },
+    {
+      "epoch": 9.294780639384843,
+      "grad_norm": 1.165895700454712,
+      "learning_rate": 2.434094548726917e-06,
+      "loss": 1.5625,
+      "step": 44120
+    },
+    {
+      "epoch": 9.296887343972191,
+      "grad_norm": 1.2659752368927002,
+      "learning_rate": 2.419599928410554e-06,
+      "loss": 1.5215,
+      "step": 44130
+    },
+    {
+      "epoch": 9.29899404855954,
+      "grad_norm": 1.2900304794311523,
+      "learning_rate": 2.405148065057794e-06,
+      "loss": 1.6137,
+      "step": 44140
+    },
+    {
+      "epoch": 9.30110075314689,
+      "grad_norm": 1.1113080978393555,
+      "learning_rate": 2.390738965001038e-06,
+      "loss": 1.5197,
+      "step": 44150
+    },
+    {
+      "epoch": 9.303207457734239,
+      "grad_norm": 1.139173984527588,
+      "learning_rate": 2.376372634553936e-06,
+      "loss": 1.5136,
+      "step": 44160
+    },
+    {
+      "epoch": 9.305314162321588,
+      "grad_norm": 1.209246039390564,
+      "learning_rate": 2.3620490800114304e-06,
+      "loss": 1.605,
+      "step": 44170
+    },
+    {
+      "epoch": 9.307420866908938,
+      "grad_norm": 1.2354110479354858,
+      "learning_rate": 2.347768307649667e-06,
+      "loss": 1.5638,
+      "step": 44180
+    },
+    {
+      "epoch": 9.309527571496288,
+      "grad_norm": 1.3547914028167725,
+      "learning_rate": 2.3335303237260853e-06,
+      "loss": 1.6327,
+      "step": 44190
+    },
+    {
+      "epoch": 9.311634276083636,
+      "grad_norm": 1.3228518962860107,
+      "learning_rate": 2.3193351344793835e-06,
+      "loss": 1.5426,
+      "step": 44200
+    },
+    {
+      "epoch": 9.313740980670985,
+      "grad_norm": 1.3366025686264038,
+      "learning_rate": 2.3051827461294638e-06,
+      "loss": 1.5076,
+      "step": 44210
+    },
+    {
+      "epoch": 9.315847685258335,
+      "grad_norm": 1.2990881204605103,
+      "learning_rate": 2.291073164877511e-06,
+      "loss": 1.5914,
+      "step": 44220
+    },
+    {
+      "epoch": 9.317954389845683,
+      "grad_norm": 1.2209333181381226,
+      "learning_rate": 2.2770063969059562e-06,
+      "loss": 1.5379,
+      "step": 44230
+    },
+    {
+      "epoch": 9.320061094433033,
+      "grad_norm": 1.2527897357940674,
+      "learning_rate": 2.2629824483784366e-06,
+      "loss": 1.5866,
+      "step": 44240
+    },
+    {
+      "epoch": 9.322167799020383,
+      "grad_norm": 1.1806588172912598,
+      "learning_rate": 2.24900132543987e-06,
+      "loss": 1.545,
+      "step": 44250
+    },
+    {
+      "epoch": 9.324274503607732,
+      "grad_norm": 1.2967365980148315,
+      "learning_rate": 2.235063034216378e-06,
+      "loss": 1.5978,
+      "step": 44260
+    },
+    {
+      "epoch": 9.32638120819508,
+      "grad_norm": 1.301858901977539,
+      "learning_rate": 2.2211675808153198e-06,
+      "loss": 1.5622,
+      "step": 44270
+    },
+    {
+      "epoch": 9.32848791278243,
+      "grad_norm": 1.3768510818481445,
+      "learning_rate": 2.207314971325292e-06,
+      "loss": 1.5269,
+      "step": 44280
+    },
+    {
+      "epoch": 9.33059461736978,
+      "grad_norm": 1.2430403232574463,
+      "learning_rate": 2.193505211816127e-06,
+      "loss": 1.5703,
+      "step": 44290
+    },
+    {
+      "epoch": 9.332701321957128,
+      "grad_norm": 1.2545567750930786,
+      "learning_rate": 2.1797383083388412e-06,
+      "loss": 1.4986,
+      "step": 44300
+    },
+    {
+      "epoch": 9.334808026544477,
+      "grad_norm": 1.3329411745071411,
+      "learning_rate": 2.166014266925731e-06,
+      "loss": 1.5672,
+      "step": 44310
+    },
+    {
+      "epoch": 9.336914731131827,
+      "grad_norm": 1.41471266746521,
+      "learning_rate": 2.152333093590264e-06,
+      "loss": 1.561,
+      "step": 44320
+    },
+    {
+      "epoch": 9.339021435719177,
+      "grad_norm": 1.2551592588424683,
+      "learning_rate": 2.1386947943271562e-06,
+      "loss": 1.5539,
+      "step": 44330
+    },
+    {
+      "epoch": 9.341128140306525,
+      "grad_norm": 1.2264280319213867,
+      "learning_rate": 2.125099375112316e-06,
+      "loss": 1.5668,
+      "step": 44340
+    },
+    {
+      "epoch": 9.343234844893875,
+      "grad_norm": 1.3428797721862793,
+      "learning_rate": 2.1115468419028672e-06,
+      "loss": 1.5565,
+      "step": 44350
+    },
+    {
+      "epoch": 9.345341549481224,
+      "grad_norm": 1.0985000133514404,
+      "learning_rate": 2.0980372006371487e-06,
+      "loss": 1.5038,
+      "step": 44360
+    },
+    {
+      "epoch": 9.347448254068572,
+      "grad_norm": 1.2095003128051758,
+      "learning_rate": 2.0845704572347025e-06,
+      "loss": 1.5727,
+      "step": 44370
+    },
+    {
+      "epoch": 9.349554958655922,
+      "grad_norm": 1.3696715831756592,
+      "learning_rate": 2.0711466175962756e-06,
+      "loss": 1.5711,
+      "step": 44380
+    },
+    {
+      "epoch": 9.351661663243272,
+      "grad_norm": 1.1304073333740234,
+      "learning_rate": 2.0577656876038076e-06,
+      "loss": 1.5606,
+      "step": 44390
+    },
+    {
+      "epoch": 9.353768367830622,
+      "grad_norm": 1.2599557638168335,
+      "learning_rate": 2.0444276731204415e-06,
+      "loss": 1.5672,
+      "step": 44400
+    },
+    {
+      "epoch": 9.35587507241797,
+      "grad_norm": 1.1687496900558472,
+      "learning_rate": 2.0311325799905356e-06,
+      "loss": 1.5357,
+      "step": 44410
+    },
+    {
+      "epoch": 9.35798177700532,
+      "grad_norm": 1.28921377658844,
+      "learning_rate": 2.0178804140396078e-06,
+      "loss": 1.5935,
+      "step": 44420
+    },
+    {
+      "epoch": 9.360088481592669,
+      "grad_norm": 1.1254773139953613,
+      "learning_rate": 2.004671181074369e-06,
+      "loss": 1.5027,
+      "step": 44430
+    },
+    {
+      "epoch": 9.362195186180017,
+      "grad_norm": 1.3228946924209595,
+      "learning_rate": 1.9915048868827558e-06,
+      "loss": 1.5286,
+      "step": 44440
+    },
+    {
+      "epoch": 9.364301890767367,
+      "grad_norm": 1.2278170585632324,
+      "learning_rate": 1.9783815372338423e-06,
+      "loss": 1.5514,
+      "step": 44450
+    },
+    {
+      "epoch": 9.366408595354716,
+      "grad_norm": 1.287775993347168,
+      "learning_rate": 1.9653011378779283e-06,
+      "loss": 1.5499,
+      "step": 44460
+    },
+    {
+      "epoch": 9.368515299942066,
+      "grad_norm": 1.283502221107483,
+      "learning_rate": 1.9522636945464635e-06,
+      "loss": 1.5364,
+      "step": 44470
+    },
+    {
+      "epoch": 9.370622004529414,
+      "grad_norm": 1.1731789112091064,
+      "learning_rate": 1.9392692129520882e-06,
+      "loss": 1.5227,
+      "step": 44480
+    },
+    {
+      "epoch": 9.372728709116764,
+      "grad_norm": 1.3029754161834717,
+      "learning_rate": 1.9263176987886043e-06,
+      "loss": 1.5858,
+      "step": 44490
+    },
+    {
+      "epoch": 9.374835413704114,
+      "grad_norm": 1.1961798667907715,
+      "learning_rate": 1.9134091577310274e-06,
+      "loss": 1.588,
+      "step": 44500
+    },
+    {
+      "epoch": 9.376942118291463,
+      "grad_norm": 1.2847782373428345,
+      "learning_rate": 1.9005435954354667e-06,
+      "loss": 1.5501,
+      "step": 44510
+    },
+    {
+      "epoch": 9.379048822878811,
+      "grad_norm": 1.1511763334274292,
+      "learning_rate": 1.88772101753929e-06,
+      "loss": 1.5782,
+      "step": 44520
+    },
+    {
+      "epoch": 9.381155527466161,
+      "grad_norm": 1.1530286073684692,
+      "learning_rate": 1.8749414296609702e-06,
+      "loss": 1.5413,
+      "step": 44530
+    },
+    {
+      "epoch": 9.38326223205351,
+      "grad_norm": 1.1921340227127075,
+      "learning_rate": 1.862204837400161e-06,
+      "loss": 1.5548,
+      "step": 44540
+    },
+    {
+      "epoch": 9.385368936640859,
+      "grad_norm": 1.3235081434249878,
+      "learning_rate": 1.8495112463376874e-06,
+      "loss": 1.5859,
+      "step": 44550
+    },
+    {
+      "epoch": 9.387475641228209,
+      "grad_norm": 1.2756280899047852,
+      "learning_rate": 1.836860662035489e-06,
+      "loss": 1.5467,
+      "step": 44560
+    },
+    {
+      "epoch": 9.389582345815558,
+      "grad_norm": 1.198059320449829,
+      "learning_rate": 1.8242530900367315e-06,
+      "loss": 1.6058,
+      "step": 44570
+    },
+    {
+      "epoch": 9.391689050402908,
+      "grad_norm": 1.1972780227661133,
+      "learning_rate": 1.8116885358656744e-06,
+      "loss": 1.5333,
+      "step": 44580
+    },
+    {
+      "epoch": 9.393795754990256,
+      "grad_norm": 1.2168101072311401,
+      "learning_rate": 1.7991670050277354e-06,
+      "loss": 1.5402,
+      "step": 44590
+    },
+    {
+      "epoch": 9.395902459577606,
+      "grad_norm": 1.1930609941482544,
+      "learning_rate": 1.786688503009537e-06,
+      "loss": 1.5392,
+      "step": 44600
+    },
+    {
+      "epoch": 9.398009164164955,
+      "grad_norm": 1.1218000650405884,
+      "learning_rate": 1.7742530352787612e-06,
+      "loss": 1.4655,
+      "step": 44610
+    },
+    {
+      "epoch": 9.400115868752303,
+      "grad_norm": 1.3308579921722412,
+      "learning_rate": 1.7618606072842936e-06,
+      "loss": 1.5094,
+      "step": 44620
+    },
+    {
+      "epoch": 9.402222573339653,
+      "grad_norm": 1.2352577447891235,
+      "learning_rate": 1.7495112244561573e-06,
+      "loss": 1.5612,
+      "step": 44630
+    },
+    {
+      "epoch": 9.404329277927003,
+      "grad_norm": 1.2200725078582764,
+      "learning_rate": 1.7372048922054906e-06,
+      "loss": 1.5206,
+      "step": 44640
+    },
+    {
+      "epoch": 9.406435982514353,
+      "grad_norm": 1.4431538581848145,
+      "learning_rate": 1.7249416159245691e-06,
+      "loss": 1.5482,
+      "step": 44650
+    },
+    {
+      "epoch": 9.4085426871017,
+      "grad_norm": 1.1892200708389282,
+      "learning_rate": 1.7127214009868385e-06,
+      "loss": 1.64,
+      "step": 44660
+    },
+    {
+      "epoch": 9.41064939168905,
+      "grad_norm": 1.2224235534667969,
+      "learning_rate": 1.7005442527468163e-06,
+      "loss": 1.5128,
+      "step": 44670
+    },
+    {
+      "epoch": 9.4127560962764,
+      "grad_norm": 1.1963303089141846,
+      "learning_rate": 1.6884101765402116e-06,
+      "loss": 1.5092,
+      "step": 44680
+    },
+    {
+      "epoch": 9.414862800863748,
+      "grad_norm": 1.2295386791229248,
+      "learning_rate": 1.676319177683816e-06,
+      "loss": 1.5728,
+      "step": 44690
+    },
+    {
+      "epoch": 9.416969505451098,
+      "grad_norm": 1.390763759613037,
+      "learning_rate": 1.6642712614755695e-06,
+      "loss": 1.5631,
+      "step": 44700
+    },
+    {
+      "epoch": 9.419076210038448,
+      "grad_norm": 1.2150217294692993,
+      "learning_rate": 1.6522664331945382e-06,
+      "loss": 1.5224,
+      "step": 44710
+    },
+    {
+      "epoch": 9.421182914625797,
+      "grad_norm": 1.1678669452667236,
+      "learning_rate": 1.6403046981008807e-06,
+      "loss": 1.6289,
+      "step": 44720
+    },
+    {
+      "epoch": 9.423289619213145,
+      "grad_norm": 1.193841814994812,
+      "learning_rate": 1.6283860614358936e-06,
+      "loss": 1.6028,
+      "step": 44730
+    },
+    {
+      "epoch": 9.425396323800495,
+      "grad_norm": 1.2163934707641602,
+      "learning_rate": 1.6165105284219772e-06,
+      "loss": 1.5197,
+      "step": 44740
+    },
+    {
+      "epoch": 9.427503028387845,
+      "grad_norm": 1.2479548454284668,
+      "learning_rate": 1.6046781042626802e-06,
+      "loss": 1.5534,
+      "step": 44750
+    },
+    {
+      "epoch": 9.429609732975194,
+      "grad_norm": 1.2134898900985718,
+      "learning_rate": 1.5928887941426107e-06,
+      "loss": 1.5745,
+      "step": 44760
+    },
+    {
+      "epoch": 9.431716437562542,
+      "grad_norm": 1.2324718236923218,
+      "learning_rate": 1.5811426032275146e-06,
+      "loss": 1.5941,
+      "step": 44770
+    },
+    {
+      "epoch": 9.433823142149892,
+      "grad_norm": 1.2660291194915771,
+      "learning_rate": 1.5694395366642411e-06,
+      "loss": 1.5639,
+      "step": 44780
+    },
+    {
+      "epoch": 9.435929846737242,
+      "grad_norm": 1.1077207326889038,
+      "learning_rate": 1.5577795995807554e-06,
+      "loss": 1.5146,
+      "step": 44790
+    },
+    {
+      "epoch": 9.43803655132459,
+      "grad_norm": 1.2963979244232178,
+      "learning_rate": 1.5461627970860814e-06,
+      "loss": 1.5483,
+      "step": 44800
+    },
+    {
+      "epoch": 9.44014325591194,
+      "grad_norm": 1.2777230739593506,
+      "learning_rate": 1.5345891342704033e-06,
+      "loss": 1.5602,
+      "step": 44810
+    },
+    {
+      "epoch": 9.44224996049929,
+      "grad_norm": 1.2303513288497925,
+      "learning_rate": 1.523058616204942e-06,
+      "loss": 1.5415,
+      "step": 44820
+    },
+    {
+      "epoch": 9.444356665086639,
+      "grad_norm": 1.3114386796951294,
+      "learning_rate": 1.5115712479420786e-06,
+      "loss": 1.6028,
+      "step": 44830
+    },
+    {
+      "epoch": 9.446463369673987,
+      "grad_norm": 1.1598353385925293,
+      "learning_rate": 1.5001270345152307e-06,
+      "loss": 1.5337,
+      "step": 44840
+    },
+    {
+      "epoch": 9.448570074261337,
+      "grad_norm": 1.2327178716659546,
+      "learning_rate": 1.4887259809389208e-06,
+      "loss": 1.5464,
+      "step": 44850
+    },
+    {
+      "epoch": 9.450676778848687,
+      "grad_norm": 1.2036960124969482,
+      "learning_rate": 1.4773680922087863e-06,
+      "loss": 1.5555,
+      "step": 44860
+    },
+    {
+      "epoch": 9.452783483436034,
+      "grad_norm": 1.258427381515503,
+      "learning_rate": 1.4660533733015236e-06,
+      "loss": 1.5747,
+      "step": 44870
+    },
+    {
+      "epoch": 9.454890188023384,
+      "grad_norm": 1.2606942653656006,
+      "learning_rate": 1.4547818291749115e-06,
+      "loss": 1.5928,
+      "step": 44880
+    },
+    {
+      "epoch": 9.456996892610734,
+      "grad_norm": 1.2755640745162964,
+      "learning_rate": 1.4435534647678438e-06,
+      "loss": 1.5394,
+      "step": 44890
+    },
+    {
+      "epoch": 9.459103597198084,
+      "grad_norm": 1.2024848461151123,
+      "learning_rate": 1.4323682850002406e-06,
+      "loss": 1.568,
+      "step": 44900
+    },
+    {
+      "epoch": 9.461210301785432,
+      "grad_norm": 1.3687775135040283,
+      "learning_rate": 1.4212262947731703e-06,
+      "loss": 1.5108,
+      "step": 44910
+    },
+    {
+      "epoch": 9.463317006372781,
+      "grad_norm": 1.1984235048294067,
+      "learning_rate": 1.4101274989687053e-06,
+      "loss": 1.5219,
+      "step": 44920
+    },
+    {
+      "epoch": 9.465423710960131,
+      "grad_norm": 1.1227729320526123,
+      "learning_rate": 1.399071902450022e-06,
+      "loss": 1.6081,
+      "step": 44930
+    },
+    {
+      "epoch": 9.467530415547479,
+      "grad_norm": 1.3621063232421875,
+      "learning_rate": 1.3880595100613792e-06,
+      "loss": 1.5531,
+      "step": 44940
+    },
+    {
+      "epoch": 9.469637120134829,
+      "grad_norm": 1.2998682260513306,
+      "learning_rate": 1.3770903266281054e-06,
+      "loss": 1.5531,
+      "step": 44950
+    },
+    {
+      "epoch": 9.471743824722179,
+      "grad_norm": 1.3776503801345825,
+      "learning_rate": 1.3661643569565785e-06,
+      "loss": 1.5866,
+      "step": 44960
+    },
+    {
+      "epoch": 9.473850529309528,
+      "grad_norm": 1.2252628803253174,
+      "learning_rate": 1.3552816058342354e-06,
+      "loss": 1.5095,
+      "step": 44970
+    },
+    {
+      "epoch": 9.475957233896876,
+      "grad_norm": 1.3023484945297241,
+      "learning_rate": 1.3444420780296063e-06,
+      "loss": 1.5863,
+      "step": 44980
+    },
+    {
+      "epoch": 9.478063938484226,
+      "grad_norm": 1.2645686864852905,
+      "learning_rate": 1.3336457782922474e-06,
+      "loss": 1.5368,
+      "step": 44990
+    },
+    {
+      "epoch": 9.480170643071576,
+      "grad_norm": 1.1360158920288086,
+      "learning_rate": 1.3228927113528189e-06,
+      "loss": 1.5091,
+      "step": 45000
+    },
+    {
+      "epoch": 9.482277347658924,
+      "grad_norm": 1.1841646432876587,
+      "learning_rate": 1.3121828819229743e-06,
+      "loss": 1.5503,
+      "step": 45010
+    },
+    {
+      "epoch": 9.484384052246273,
+      "grad_norm": 1.1105971336364746,
+      "learning_rate": 1.3015162946954706e-06,
+      "loss": 1.5561,
+      "step": 45020
+    },
+    {
+      "epoch": 9.486490756833623,
+      "grad_norm": 1.1991256475448608,
+      "learning_rate": 1.290892954344125e-06,
+      "loss": 1.5536,
+      "step": 45030
+    },
+    {
+      "epoch": 9.488597461420973,
+      "grad_norm": 1.1073957681655884,
+      "learning_rate": 1.2803128655237694e-06,
+      "loss": 1.5014,
+      "step": 45040
+    },
+    {
+      "epoch": 9.490704166008321,
+      "grad_norm": 1.2259361743927002,
+      "learning_rate": 1.2697760328702734e-06,
+      "loss": 1.5671,
+      "step": 45050
+    },
+    {
+      "epoch": 9.49281087059567,
+      "grad_norm": 1.1295838356018066,
+      "learning_rate": 1.2592824610006215e-06,
+      "loss": 1.5262,
+      "step": 45060
+    },
+    {
+      "epoch": 9.49491757518302,
+      "grad_norm": 1.258636236190796,
+      "learning_rate": 1.248832154512769e-06,
+      "loss": 1.55,
+      "step": 45070
+    },
+    {
+      "epoch": 9.497024279770368,
+      "grad_norm": 1.2501273155212402,
+      "learning_rate": 1.2384251179857643e-06,
+      "loss": 1.5668,
+      "step": 45080
+    },
+    {
+      "epoch": 9.499130984357718,
+      "grad_norm": 1.2161610126495361,
+      "learning_rate": 1.2280613559796595e-06,
+      "loss": 1.5267,
+      "step": 45090
+    },
+    {
+      "epoch": 9.501237688945068,
+      "grad_norm": 1.2681390047073364,
+      "learning_rate": 1.2177408730355554e-06,
+      "loss": 1.5289,
+      "step": 45100
+    },
+    {
+      "epoch": 9.503344393532418,
+      "grad_norm": 1.1886988878250122,
+      "learning_rate": 1.2074636736756129e-06,
+      "loss": 1.5128,
+      "step": 45110
+    },
+    {
+      "epoch": 9.505451098119766,
+      "grad_norm": 1.2830315828323364,
+      "learning_rate": 1.1972297624030072e-06,
+      "loss": 1.5249,
+      "step": 45120
+    },
+    {
+      "epoch": 9.507557802707115,
+      "grad_norm": 1.2447185516357422,
+      "learning_rate": 1.18703914370194e-06,
+      "loss": 1.5655,
+      "step": 45130
+    },
+    {
+      "epoch": 9.509664507294465,
+      "grad_norm": 1.3017278909683228,
+      "learning_rate": 1.1768918220376624e-06,
+      "loss": 1.5557,
+      "step": 45140
+    },
+    {
+      "epoch": 9.511771211881815,
+      "grad_norm": 1.3122484683990479,
+      "learning_rate": 1.1667878018564171e-06,
+      "loss": 1.5859,
+      "step": 45150
+    },
+    {
+      "epoch": 9.513877916469163,
+      "grad_norm": 1.339020848274231,
+      "learning_rate": 1.1567270875855407e-06,
+      "loss": 1.5386,
+      "step": 45160
+    },
+    {
+      "epoch": 9.515984621056512,
+      "grad_norm": 1.227014422416687,
+      "learning_rate": 1.1467096836333069e-06,
+      "loss": 1.6457,
+      "step": 45170
+    },
+    {
+      "epoch": 9.518091325643862,
+      "grad_norm": 1.3103618621826172,
+      "learning_rate": 1.1367355943890823e-06,
+      "loss": 1.5938,
+      "step": 45180
+    },
+    {
+      "epoch": 9.52019803023121,
+      "grad_norm": 1.1503431797027588,
+      "learning_rate": 1.1268048242232375e-06,
+      "loss": 1.5428,
+      "step": 45190
+    },
+    {
+      "epoch": 9.52230473481856,
+      "grad_norm": 1.1854219436645508,
+      "learning_rate": 1.1169173774871478e-06,
+      "loss": 1.5368,
+      "step": 45200
+    },
+    {
+      "epoch": 9.52441143940591,
+      "grad_norm": 1.2122063636779785,
+      "learning_rate": 1.1070732585132026e-06,
+      "loss": 1.5124,
+      "step": 45210
+    },
+    {
+      "epoch": 9.52651814399326,
+      "grad_norm": 1.2711586952209473,
+      "learning_rate": 1.0972724716148187e-06,
+      "loss": 1.5631,
+      "step": 45220
+    },
+    {
+      "epoch": 9.528624848580607,
+      "grad_norm": 1.3009141683578491,
+      "learning_rate": 1.0875150210864271e-06,
+      "loss": 1.5276,
+      "step": 45230
+    },
+    {
+      "epoch": 9.530731553167957,
+      "grad_norm": 1.1954138278961182,
+      "learning_rate": 1.0778009112034748e-06,
+      "loss": 1.4954,
+      "step": 45240
+    },
+    {
+      "epoch": 9.532838257755307,
+      "grad_norm": 1.1968759298324585,
+      "learning_rate": 1.0681301462223903e-06,
+      "loss": 1.5634,
+      "step": 45250
+    },
+    {
+      "epoch": 9.534944962342655,
+      "grad_norm": 1.2859160900115967,
+      "learning_rate": 1.0585027303806393e-06,
+      "loss": 1.5997,
+      "step": 45260
+    },
+    {
+      "epoch": 9.537051666930005,
+      "grad_norm": 1.1200820207595825,
+      "learning_rate": 1.0489186678966812e-06,
+      "loss": 1.5602,
+      "step": 45270
+    },
+    {
+      "epoch": 9.539158371517354,
+      "grad_norm": 1.3090044260025024,
+      "learning_rate": 1.0393779629699786e-06,
+      "loss": 1.6024,
+      "step": 45280
+    },
+    {
+      "epoch": 9.541265076104704,
+      "grad_norm": 1.2200546264648438,
+      "learning_rate": 1.0298806197809984e-06,
+      "loss": 1.5606,
+      "step": 45290
+    },
+    {
+      "epoch": 9.543371780692052,
+      "grad_norm": 1.2394194602966309,
+      "learning_rate": 1.0204266424912123e-06,
+      "loss": 1.5037,
+      "step": 45300
+    },
+    {
+      "epoch": 9.545478485279402,
+      "grad_norm": 1.1983662843704224,
+      "learning_rate": 1.011016035243073e-06,
+      "loss": 1.5588,
+      "step": 45310
+    },
+    {
+      "epoch": 9.547585189866751,
+      "grad_norm": 1.2025030851364136,
+      "learning_rate": 1.001648802160049e-06,
+      "loss": 1.5663,
+      "step": 45320
+    },
+    {
+      "epoch": 9.5496918944541,
+      "grad_norm": 1.2647790908813477,
+      "learning_rate": 9.923249473466012e-07,
+      "loss": 1.5482,
+      "step": 45330
+    },
+    {
+      "epoch": 9.55179859904145,
+      "grad_norm": 1.2974402904510498,
+      "learning_rate": 9.830444748881728e-07,
+      "loss": 1.5217,
+      "step": 45340
+    },
+    {
+      "epoch": 9.553905303628799,
+      "grad_norm": 1.124445915222168,
+      "learning_rate": 9.738073888512e-07,
+      "loss": 1.4856,
+      "step": 45350
+    },
+    {
+      "epoch": 9.556012008216149,
+      "grad_norm": 1.7168290615081787,
+      "learning_rate": 9.64613693283123e-07,
+      "loss": 1.5335,
+      "step": 45360
+    },
+    {
+      "epoch": 9.558118712803497,
+      "grad_norm": 1.2465564012527466,
+      "learning_rate": 9.554633922123412e-07,
+      "loss": 1.5456,
+      "step": 45370
+    },
+    {
+      "epoch": 9.560225417390846,
+      "grad_norm": 1.1842801570892334,
+      "learning_rate": 9.463564896482813e-07,
+      "loss": 1.5513,
+      "step": 45380
+    },
+    {
+      "epoch": 9.562332121978196,
+      "grad_norm": 1.1776254177093506,
+      "learning_rate": 9.372929895813065e-07,
+      "loss": 1.5909,
+      "step": 45390
+    },
+    {
+      "epoch": 9.564438826565546,
+      "grad_norm": 1.1699087619781494,
+      "learning_rate": 9.282728959827958e-07,
+      "loss": 1.5508,
+      "step": 45400
+    },
+    {
+      "epoch": 9.566545531152894,
+      "grad_norm": 1.3377196788787842,
+      "learning_rate": 9.192962128050986e-07,
+      "loss": 1.5649,
+      "step": 45410
+    },
+    {
+      "epoch": 9.568652235740243,
+      "grad_norm": 1.380293369293213,
+      "learning_rate": 9.103629439815354e-07,
+      "loss": 1.5433,
+      "step": 45420
+    },
+    {
+      "epoch": 9.570758940327593,
+      "grad_norm": 1.2865245342254639,
+      "learning_rate": 9.014730934264192e-07,
+      "loss": 1.5992,
+      "step": 45430
+    },
+    {
+      "epoch": 9.572865644914941,
+      "grad_norm": 1.338066816329956,
+      "learning_rate": 8.926266650350234e-07,
+      "loss": 1.5397,
+      "step": 45440
+    },
+    {
+      "epoch": 9.574972349502291,
+      "grad_norm": 1.3003287315368652,
+      "learning_rate": 8.838236626836138e-07,
+      "loss": 1.6043,
+      "step": 45450
+    },
+    {
+      "epoch": 9.57707905408964,
+      "grad_norm": 1.2428746223449707,
+      "learning_rate": 8.750640902294161e-07,
+      "loss": 1.6224,
+      "step": 45460
+    },
+    {
+      "epoch": 9.579185758676989,
+      "grad_norm": 1.2704057693481445,
+      "learning_rate": 8.663479515106043e-07,
+      "loss": 1.5504,
+      "step": 45470
+    },
+    {
+      "epoch": 9.581292463264338,
+      "grad_norm": 1.1228466033935547,
+      "learning_rate": 8.57675250346368e-07,
+      "loss": 1.563,
+      "step": 45480
+    },
+    {
+      "epoch": 9.583399167851688,
+      "grad_norm": 1.1902949810028076,
+      "learning_rate": 8.490459905368342e-07,
+      "loss": 1.4802,
+      "step": 45490
+    },
+    {
+      "epoch": 9.585505872439038,
+      "grad_norm": 1.2982428073883057,
+      "learning_rate": 8.404601758630892e-07,
+      "loss": 1.5255,
+      "step": 45500
+    },
+    {
+      "epoch": 9.587612577026386,
+      "grad_norm": 1.2793209552764893,
+      "learning_rate": 8.319178100872016e-07,
+      "loss": 1.5295,
+      "step": 45510
+    },
+    {
+      "epoch": 9.589719281613736,
+      "grad_norm": 1.2290621995925903,
+      "learning_rate": 8.234188969521883e-07,
+      "loss": 1.5463,
+      "step": 45520
+    },
+    {
+      "epoch": 9.591825986201085,
+      "grad_norm": 1.2028400897979736,
+      "learning_rate": 8.149634401820372e-07,
+      "loss": 1.5438,
+      "step": 45530
+    },
+    {
+      "epoch": 9.593932690788435,
+      "grad_norm": 1.3339173793792725,
+      "learning_rate": 8.065514434816845e-07,
+      "loss": 1.5446,
+      "step": 45540
+    },
+    {
+      "epoch": 9.596039395375783,
+      "grad_norm": 1.1351807117462158,
+      "learning_rate": 7.981829105370375e-07,
+      "loss": 1.5374,
+      "step": 45550
+    },
+    {
+      "epoch": 9.598146099963133,
+      "grad_norm": 1.3664101362228394,
+      "learning_rate": 7.898578450149407e-07,
+      "loss": 1.586,
+      "step": 45560
+    },
+    {
+      "epoch": 9.600252804550482,
+      "grad_norm": 1.2555339336395264,
+      "learning_rate": 7.815762505632096e-07,
+      "loss": 1.5537,
+      "step": 45570
+    },
+    {
+      "epoch": 9.60235950913783,
+      "grad_norm": 1.0828977823257446,
+      "learning_rate": 7.733381308105969e-07,
+      "loss": 1.5487,
+      "step": 45580
+    },
+    {
+      "epoch": 9.60446621372518,
+      "grad_norm": 1.2509876489639282,
+      "learning_rate": 7.651434893668152e-07,
+      "loss": 1.5715,
+      "step": 45590
+    },
+    {
+      "epoch": 9.60657291831253,
+      "grad_norm": 1.2072515487670898,
+      "learning_rate": 7.569923298225146e-07,
+      "loss": 1.4886,
+      "step": 45600
+    },
+    {
+      "epoch": 9.60867962289988,
+      "grad_norm": 1.249969482421875,
+      "learning_rate": 7.488846557493267e-07,
+      "loss": 1.5317,
+      "step": 45610
+    },
+    {
+      "epoch": 9.610786327487228,
+      "grad_norm": 1.304600477218628,
+      "learning_rate": 7.408204706997879e-07,
+      "loss": 1.6078,
+      "step": 45620
+    },
+    {
+      "epoch": 9.612893032074577,
+      "grad_norm": 1.2158945798873901,
+      "learning_rate": 7.327997782073936e-07,
+      "loss": 1.5877,
+      "step": 45630
+    },
+    {
+      "epoch": 9.614999736661927,
+      "grad_norm": 1.1745023727416992,
+      "learning_rate": 7.248225817865884e-07,
+      "loss": 1.5379,
+      "step": 45640
+    },
+    {
+      "epoch": 9.617106441249275,
+      "grad_norm": 1.2478928565979004,
+      "learning_rate": 7.168888849327426e-07,
+      "loss": 1.5642,
+      "step": 45650
+    },
+    {
+      "epoch": 9.619213145836625,
+      "grad_norm": 1.1718308925628662,
+      "learning_rate": 7.08998691122198e-07,
+      "loss": 1.5113,
+      "step": 45660
+    },
+    {
+      "epoch": 9.621319850423975,
+      "grad_norm": 1.2330695390701294,
+      "learning_rate": 7.01152003812211e-07,
+      "loss": 1.5686,
+      "step": 45670
+    },
+    {
+      "epoch": 9.623426555011324,
+      "grad_norm": 1.255374789237976,
+      "learning_rate": 6.933488264409538e-07,
+      "loss": 1.5387,
+      "step": 45680
+    },
+    {
+      "epoch": 9.625533259598672,
+      "grad_norm": 1.1764744520187378,
+      "learning_rate": 6.855891624275801e-07,
+      "loss": 1.4828,
+      "step": 45690
+    },
+    {
+      "epoch": 9.627639964186022,
+      "grad_norm": 1.2803027629852295,
+      "learning_rate": 6.77873015172148e-07,
+      "loss": 1.5569,
+      "step": 45700
+    },
+    {
+      "epoch": 9.629746668773372,
+      "grad_norm": 1.295817255973816,
+      "learning_rate": 6.702003880556418e-07,
+      "loss": 1.5269,
+      "step": 45710
+    },
+    {
+      "epoch": 9.63185337336072,
+      "grad_norm": 1.1505221128463745,
+      "learning_rate": 6.625712844400056e-07,
+      "loss": 1.5517,
+      "step": 45720
+    },
+    {
+      "epoch": 9.63396007794807,
+      "grad_norm": 1.2030140161514282,
+      "learning_rate": 6.549857076680987e-07,
+      "loss": 1.5361,
+      "step": 45730
+    },
+    {
+      "epoch": 9.63606678253542,
+      "grad_norm": 1.2046858072280884,
+      "learning_rate": 6.474436610636958e-07,
+      "loss": 1.5525,
+      "step": 45740
+    },
+    {
+      "epoch": 9.638173487122769,
+      "grad_norm": 1.2719230651855469,
+      "learning_rate": 6.399451479315088e-07,
+      "loss": 1.5241,
+      "step": 45750
+    },
+    {
+      "epoch": 9.640280191710117,
+      "grad_norm": 1.1040761470794678,
+      "learning_rate": 6.324901715571651e-07,
+      "loss": 1.5709,
+      "step": 45760
+    },
+    {
+      "epoch": 9.642386896297467,
+      "grad_norm": 1.313246250152588,
+      "learning_rate": 6.250787352072518e-07,
+      "loss": 1.5177,
+      "step": 45770
+    },
+    {
+      "epoch": 9.644493600884816,
+      "grad_norm": 1.1665434837341309,
+      "learning_rate": 6.177108421292266e-07,
+      "loss": 1.4843,
+      "step": 45780
+    },
+    {
+      "epoch": 9.646600305472166,
+      "grad_norm": 1.19307541847229,
+      "learning_rate": 6.103864955514849e-07,
+      "loss": 1.4744,
+      "step": 45790
+    },
+    {
+      "epoch": 9.648707010059514,
+      "grad_norm": 1.35244882106781,
+      "learning_rate": 6.031056986833705e-07,
+      "loss": 1.5368,
+      "step": 45800
+    },
+    {
+      "epoch": 9.650813714646864,
+      "grad_norm": 1.1607619524002075,
+      "learning_rate": 5.958684547151095e-07,
+      "loss": 1.6192,
+      "step": 45810
+    },
+    {
+      "epoch": 9.652920419234214,
+      "grad_norm": 1.2676018476486206,
+      "learning_rate": 5.886747668178538e-07,
+      "loss": 1.5227,
+      "step": 45820
+    },
+    {
+      "epoch": 9.655027123821561,
+      "grad_norm": 1.2928742170333862,
+      "learning_rate": 5.815246381436934e-07,
+      "loss": 1.518,
+      "step": 45830
+    },
+    {
+      "epoch": 9.657133828408911,
+      "grad_norm": 1.1910942792892456,
+      "learning_rate": 5.744180718255776e-07,
+      "loss": 1.5501,
+      "step": 45840
+    },
+    {
+      "epoch": 9.659240532996261,
+      "grad_norm": 1.215248465538025,
+      "learning_rate": 5.673550709774267e-07,
+      "loss": 1.6009,
+      "step": 45850
+    },
+    {
+      "epoch": 9.661347237583609,
+      "grad_norm": 1.1708420515060425,
+      "learning_rate": 5.603356386940429e-07,
+      "loss": 1.5504,
+      "step": 45860
+    },
+    {
+      "epoch": 9.663453942170959,
+      "grad_norm": 1.3606902360916138,
+      "learning_rate": 5.533597780511435e-07,
+      "loss": 1.5371,
+      "step": 45870
+    },
+    {
+      "epoch": 9.665560646758308,
+      "grad_norm": 1.2347360849380493,
+      "learning_rate": 5.464274921053503e-07,
+      "loss": 1.5587,
+      "step": 45880
+    },
+    {
+      "epoch": 9.667667351345658,
+      "grad_norm": 1.3127825260162354,
+      "learning_rate": 5.395387838942001e-07,
+      "loss": 1.5692,
+      "step": 45890
+    },
+    {
+      "epoch": 9.669774055933006,
+      "grad_norm": 1.1706701517105103,
+      "learning_rate": 5.326936564361118e-07,
+      "loss": 1.5369,
+      "step": 45900
+    },
+    {
+      "epoch": 9.671880760520356,
+      "grad_norm": 1.190574049949646,
+      "learning_rate": 5.258921127304528e-07,
+      "loss": 1.4852,
+      "step": 45910
+    },
+    {
+      "epoch": 9.673987465107706,
+      "grad_norm": 1.3065412044525146,
+      "learning_rate": 5.191341557574392e-07,
+      "loss": 1.5335,
+      "step": 45920
+    },
+    {
+      "epoch": 9.676094169695055,
+      "grad_norm": 1.227949857711792,
+      "learning_rate": 5.124197884782356e-07,
+      "loss": 1.5388,
+      "step": 45930
+    },
+    {
+      "epoch": 9.678200874282403,
+      "grad_norm": 1.2584184408187866,
+      "learning_rate": 5.057490138348775e-07,
+      "loss": 1.5407,
+      "step": 45940
+    },
+    {
+      "epoch": 9.680307578869753,
+      "grad_norm": 1.395398736000061,
+      "learning_rate": 4.991218347503157e-07,
+      "loss": 1.5738,
+      "step": 45950
+    },
+    {
+      "epoch": 9.682414283457103,
+      "grad_norm": 1.1984232664108276,
+      "learning_rate": 4.92538254128383e-07,
+      "loss": 1.5051,
+      "step": 45960
+    },
+    {
+      "epoch": 9.68452098804445,
+      "grad_norm": 1.2473891973495483,
+      "learning_rate": 4.859982748538272e-07,
+      "loss": 1.6031,
+      "step": 45970
+    },
+    {
+      "epoch": 9.6866276926318,
+      "grad_norm": 1.2355321645736694,
+      "learning_rate": 4.795018997922784e-07,
+      "loss": 1.5476,
+      "step": 45980
+    },
+    {
+      "epoch": 9.68873439721915,
+      "grad_norm": 1.1943268775939941,
+      "learning_rate": 4.7304913179025965e-07,
+      "loss": 1.566,
+      "step": 45990
+    },
+    {
+      "epoch": 9.6908411018065,
+      "grad_norm": 1.1601346731185913,
+      "learning_rate": 4.666399736751981e-07,
+      "loss": 1.5289,
+      "step": 46000
+    },
+    {
+      "epoch": 9.692947806393848,
+      "grad_norm": 1.346374750137329,
+      "learning_rate": 4.602744282554028e-07,
+      "loss": 1.5774,
+      "step": 46010
+    },
+    {
+      "epoch": 9.695054510981198,
+      "grad_norm": 1.2781177759170532,
+      "learning_rate": 4.5395249832007604e-07,
+      "loss": 1.5595,
+      "step": 46020
+    },
+    {
+      "epoch": 9.697161215568547,
+      "grad_norm": 1.2155574560165405,
+      "learning_rate": 4.4767418663930196e-07,
+      "loss": 1.5051,
+      "step": 46030
+    },
+    {
+      "epoch": 9.699267920155895,
+      "grad_norm": 1.2750422954559326,
+      "learning_rate": 4.414394959640578e-07,
+      "loss": 1.5713,
+      "step": 46040
+    },
+    {
+      "epoch": 9.701374624743245,
+      "grad_norm": 1.2818180322647095,
+      "learning_rate": 4.352484290262249e-07,
+      "loss": 1.5517,
+      "step": 46050
+    },
+    {
+      "epoch": 9.703481329330595,
+      "grad_norm": 1.320322871208191,
+      "learning_rate": 4.291009885385333e-07,
+      "loss": 1.5886,
+      "step": 46060
+    },
+    {
+      "epoch": 9.705588033917945,
+      "grad_norm": 1.1885708570480347,
+      "learning_rate": 4.229971771946284e-07,
+      "loss": 1.5852,
+      "step": 46070
+    },
+    {
+      "epoch": 9.707694738505293,
+      "grad_norm": 1.2882665395736694,
+      "learning_rate": 4.1693699766902626e-07,
+      "loss": 1.5852,
+      "step": 46080
+    },
+    {
+      "epoch": 9.709801443092642,
+      "grad_norm": 1.1464838981628418,
+      "learning_rate": 4.10920452617114e-07,
+      "loss": 1.555,
+      "step": 46090
+    },
+    {
+      "epoch": 9.711908147679992,
+      "grad_norm": 1.2324957847595215,
+      "learning_rate": 4.049475446751827e-07,
+      "loss": 1.5373,
+      "step": 46100
+    },
+    {
+      "epoch": 9.71401485226734,
+      "grad_norm": 1.1413804292678833,
+      "learning_rate": 3.9901827646039446e-07,
+      "loss": 1.5398,
+      "step": 46110
+    },
+    {
+      "epoch": 9.71612155685469,
+      "grad_norm": 1.084005355834961,
+      "learning_rate": 3.931326505707822e-07,
+      "loss": 1.5267,
+      "step": 46120
+    },
+    {
+      "epoch": 9.71822826144204,
+      "grad_norm": 1.366945505142212,
+      "learning_rate": 3.872906695852607e-07,
+      "loss": 1.6069,
+      "step": 46130
+    },
+    {
+      "epoch": 9.72033496602939,
+      "grad_norm": 1.3505078554153442,
+      "learning_rate": 3.814923360636158e-07,
+      "loss": 1.6329,
+      "step": 46140
+    },
+    {
+      "epoch": 9.722441670616737,
+      "grad_norm": 1.1962380409240723,
+      "learning_rate": 3.7573765254651504e-07,
+      "loss": 1.5906,
+      "step": 46150
+    },
+    {
+      "epoch": 9.724548375204087,
+      "grad_norm": 1.1975687742233276,
+      "learning_rate": 3.700266215554971e-07,
+      "loss": 1.5692,
+      "step": 46160
+    },
+    {
+      "epoch": 9.726655079791437,
+      "grad_norm": 1.2534469366073608,
+      "learning_rate": 3.643592455929712e-07,
+      "loss": 1.5563,
+      "step": 46170
+    },
+    {
+      "epoch": 9.728761784378786,
+      "grad_norm": 1.1433435678482056,
+      "learning_rate": 3.587355271422288e-07,
+      "loss": 1.5346,
+      "step": 46180
+    },
+    {
+      "epoch": 9.730868488966134,
+      "grad_norm": 1.272630214691162,
+      "learning_rate": 3.53155468667421e-07,
+      "loss": 1.5332,
+      "step": 46190
+    },
+    {
+      "epoch": 9.732975193553484,
+      "grad_norm": 1.3106838464736938,
+      "learning_rate": 3.4761907261356976e-07,
+      "loss": 1.6156,
+      "step": 46200
+    },
+    {
+      "epoch": 9.735081898140834,
+      "grad_norm": 1.2662346363067627,
+      "learning_rate": 3.4212634140656784e-07,
+      "loss": 1.5496,
+      "step": 46210
+    },
+    {
+      "epoch": 9.737188602728182,
+      "grad_norm": 1.1517857313156128,
+      "learning_rate": 3.366772774531679e-07,
+      "loss": 1.5525,
+      "step": 46220
+    },
+    {
+      "epoch": 9.739295307315532,
+      "grad_norm": 1.2349926233291626,
+      "learning_rate": 3.3127188314100444e-07,
+      "loss": 1.4994,
+      "step": 46230
+    },
+    {
+      "epoch": 9.741402011902881,
+      "grad_norm": 1.3378150463104248,
+      "learning_rate": 3.2591016083856064e-07,
+      "loss": 1.5132,
+      "step": 46240
+    },
+    {
+      "epoch": 9.743508716490231,
+      "grad_norm": 1.350303053855896,
+      "learning_rate": 3.205921128952016e-07,
+      "loss": 1.5775,
+      "step": 46250
+    },
+    {
+      "epoch": 9.745615421077579,
+      "grad_norm": 1.3889814615249634,
+      "learning_rate": 3.1531774164111903e-07,
+      "loss": 1.6269,
+      "step": 46260
+    },
+    {
+      "epoch": 9.747722125664929,
+      "grad_norm": 1.2698912620544434,
+      "learning_rate": 3.1008704938743084e-07,
+      "loss": 1.5344,
+      "step": 46270
+    },
+    {
+      "epoch": 9.749828830252278,
+      "grad_norm": 1.3080488443374634,
+      "learning_rate": 3.049000384260592e-07,
+      "loss": 1.5701,
+      "step": 46280
+    },
+    {
+      "epoch": 9.751935534839626,
+      "grad_norm": 1.3104041814804077,
+      "learning_rate": 2.997567110297861e-07,
+      "loss": 1.6486,
+      "step": 46290
+    },
+    {
+      "epoch": 9.754042239426976,
+      "grad_norm": 1.277556300163269,
+      "learning_rate": 2.9465706945230874e-07,
+      "loss": 1.5088,
+      "step": 46300
+    },
+    {
+      "epoch": 9.756148944014326,
+      "grad_norm": 1.2666455507278442,
+      "learning_rate": 2.896011159281176e-07,
+      "loss": 1.5295,
+      "step": 46310
+    },
+    {
+      "epoch": 9.758255648601676,
+      "grad_norm": 1.277604103088379,
+      "learning_rate": 2.8458885267260705e-07,
+      "loss": 1.5221,
+      "step": 46320
+    },
+    {
+      "epoch": 9.760362353189024,
+      "grad_norm": 1.2649208307266235,
+      "learning_rate": 2.7962028188198706e-07,
+      "loss": 1.555,
+      "step": 46330
+    },
+    {
+      "epoch": 9.762469057776373,
+      "grad_norm": 1.2684274911880493,
+      "learning_rate": 2.746954057333606e-07,
+      "loss": 1.5789,
+      "step": 46340
+    },
+    {
+      "epoch": 9.764575762363723,
+      "grad_norm": 1.2506211996078491,
+      "learning_rate": 2.6981422638466814e-07,
+      "loss": 1.5723,
+      "step": 46350
+    },
+    {
+      "epoch": 9.766682466951071,
+      "grad_norm": 1.1741187572479248,
+      "learning_rate": 2.649767459746988e-07,
+      "loss": 1.5262,
+      "step": 46360
+    },
+    {
+      "epoch": 9.76878917153842,
+      "grad_norm": 1.1646640300750732,
+      "learning_rate": 2.601829666231015e-07,
+      "loss": 1.4826,
+      "step": 46370
+    },
+    {
+      "epoch": 9.77089587612577,
+      "grad_norm": 1.2875044345855713,
+      "learning_rate": 2.554328904303738e-07,
+      "loss": 1.574,
+      "step": 46380
+    },
+    {
+      "epoch": 9.77300258071312,
+      "grad_norm": 1.1216455698013306,
+      "learning_rate": 2.5072651947786184e-07,
+      "loss": 1.534,
+      "step": 46390
+    },
+    {
+      "epoch": 9.775109285300468,
+      "grad_norm": 1.1633546352386475,
+      "learning_rate": 2.460638558277606e-07,
+      "loss": 1.5665,
+      "step": 46400
+    },
+    {
+      "epoch": 9.777215989887818,
+      "grad_norm": 1.1941152811050415,
+      "learning_rate": 2.414449015231357e-07,
+      "loss": 1.5668,
+      "step": 46410
+    },
+    {
+      "epoch": 9.779322694475168,
+      "grad_norm": 1.2184324264526367,
+      "learning_rate": 2.3686965858786824e-07,
+      "loss": 1.5645,
+      "step": 46420
+    },
+    {
+      "epoch": 9.781429399062516,
+      "grad_norm": 1.2699726819992065,
+      "learning_rate": 2.3233812902669905e-07,
+      "loss": 1.5608,
+      "step": 46430
+    },
+    {
+      "epoch": 9.783536103649865,
+      "grad_norm": 1.1520774364471436,
+      "learning_rate": 2.2785031482521758e-07,
+      "loss": 1.5574,
+      "step": 46440
+    },
+    {
+      "epoch": 9.785642808237215,
+      "grad_norm": 1.4768619537353516,
+      "learning_rate": 2.2340621794985084e-07,
+      "loss": 1.5991,
+      "step": 46450
+    },
+    {
+      "epoch": 9.787749512824565,
+      "grad_norm": 1.1728389263153076,
+      "learning_rate": 2.1900584034788561e-07,
+      "loss": 1.505,
+      "step": 46460
+    },
+    {
+      "epoch": 9.789856217411913,
+      "grad_norm": 1.2773675918579102,
+      "learning_rate": 2.1464918394743516e-07,
+      "loss": 1.5297,
+      "step": 46470
+    },
+    {
+      "epoch": 9.791962921999263,
+      "grad_norm": 1.2537978887557983,
+      "learning_rate": 2.1033625065747242e-07,
+      "loss": 1.5844,
+      "step": 46480
+    },
+    {
+      "epoch": 9.794069626586612,
+      "grad_norm": 1.206842303276062,
+      "learning_rate": 2.0606704236779683e-07,
+      "loss": 1.5266,
+      "step": 46490
+    },
+    {
+      "epoch": 9.79617633117396,
+      "grad_norm": 1.3502146005630493,
+      "learning_rate": 2.0184156094905648e-07,
+      "loss": 1.5746,
+      "step": 46500
+    },
+    {
+      "epoch": 9.79828303576131,
+      "grad_norm": 1.3354579210281372,
+      "learning_rate": 1.976598082527259e-07,
+      "loss": 1.5081,
+      "step": 46510
+    },
+    {
+      "epoch": 9.80038974034866,
+      "grad_norm": 1.1795066595077515,
+      "learning_rate": 1.9352178611115046e-07,
+      "loss": 1.5523,
+      "step": 46520
+    },
+    {
+      "epoch": 9.80249644493601,
+      "grad_norm": 1.2156345844268799,
+      "learning_rate": 1.894274963374798e-07,
+      "loss": 1.491,
+      "step": 46530
+    },
+    {
+      "epoch": 9.804603149523357,
+      "grad_norm": 1.2723069190979004,
+      "learning_rate": 1.853769407257122e-07,
+      "loss": 1.5152,
+      "step": 46540
+    },
+    {
+      "epoch": 9.806709854110707,
+      "grad_norm": 1.303544521331787,
+      "learning_rate": 1.813701210506946e-07,
+      "loss": 1.5611,
+      "step": 46550
+    },
+    {
+      "epoch": 9.808816558698057,
+      "grad_norm": 1.1918131113052368,
+      "learning_rate": 1.7740703906810042e-07,
+      "loss": 1.544,
+      "step": 46560
+    },
+    {
+      "epoch": 9.810923263285407,
+      "grad_norm": 1.2438985109329224,
+      "learning_rate": 1.7348769651445164e-07,
+      "loss": 1.554,
+      "step": 46570
+    },
+    {
+      "epoch": 9.813029967872755,
+      "grad_norm": 1.3150713443756104,
+      "learning_rate": 1.6961209510707454e-07,
+      "loss": 1.4749,
+      "step": 46580
+    },
+    {
+      "epoch": 9.815136672460104,
+      "grad_norm": 1.1011872291564941,
+      "learning_rate": 1.657802365441441e-07,
+      "loss": 1.568,
+      "step": 46590
+    },
+    {
+      "epoch": 9.817243377047454,
+      "grad_norm": 1.401932716369629,
+      "learning_rate": 1.6199212250469498e-07,
+      "loss": 1.5435,
+      "step": 46600
+    },
+    {
+      "epoch": 9.819350081634802,
+      "grad_norm": 1.2226777076721191,
+      "learning_rate": 1.58247754648555e-07,
+      "loss": 1.536,
+      "step": 46610
+    },
+    {
+      "epoch": 9.821456786222152,
+      "grad_norm": 1.3291361331939697,
+      "learning_rate": 1.545471346164007e-07,
+      "loss": 1.537,
+      "step": 46620
+    },
+    {
+      "epoch": 9.823563490809502,
+      "grad_norm": 1.2543821334838867,
+      "learning_rate": 1.5089026402973493e-07,
+      "loss": 1.5187,
+      "step": 46630
+    },
+    {
+      "epoch": 9.825670195396851,
+      "grad_norm": 1.2340812683105469,
+      "learning_rate": 1.4727714449090935e-07,
+      "loss": 1.5296,
+      "step": 46640
+    },
+    {
+      "epoch": 9.8277768999842,
+      "grad_norm": 1.2020738124847412,
+      "learning_rate": 1.4370777758307974e-07,
+      "loss": 1.5796,
+      "step": 46650
+    },
+    {
+      "epoch": 9.829883604571549,
+      "grad_norm": 1.1650514602661133,
+      "learning_rate": 1.401821648702506e-07,
+      "loss": 1.5099,
+      "step": 46660
+    },
+    {
+      "epoch": 9.831990309158899,
+      "grad_norm": 1.2301157712936401,
+      "learning_rate": 1.3670030789723066e-07,
+      "loss": 1.5486,
+      "step": 46670
+    },
+    {
+      "epoch": 9.834097013746247,
+      "grad_norm": 1.166978359222412,
+      "learning_rate": 1.3326220818968838e-07,
+      "loss": 1.5177,
+      "step": 46680
+    },
+    {
+      "epoch": 9.836203718333596,
+      "grad_norm": 1.241199016571045,
+      "learning_rate": 1.298678672540854e-07,
+      "loss": 1.6009,
+      "step": 46690
+    },
+    {
+      "epoch": 9.838310422920946,
+      "grad_norm": 1.2882767915725708,
+      "learning_rate": 1.2651728657773198e-07,
+      "loss": 1.579,
+      "step": 46700
+    },
+    {
+      "epoch": 9.840417127508296,
+      "grad_norm": 1.233310580253601,
+      "learning_rate": 1.2321046762876487e-07,
+      "loss": 1.6353,
+      "step": 46710
+    },
+    {
+      "epoch": 9.842523832095644,
+      "grad_norm": 1.211338758468628,
+      "learning_rate": 1.1994741185612502e-07,
+      "loss": 1.5947,
+      "step": 46720
+    },
+    {
+      "epoch": 9.844630536682994,
+      "grad_norm": 1.278219223022461,
+      "learning_rate": 1.1672812068960204e-07,
+      "loss": 1.5459,
+      "step": 46730
+    },
+    {
+      "epoch": 9.846737241270343,
+      "grad_norm": 1.3067710399627686,
+      "learning_rate": 1.1355259553978981e-07,
+      "loss": 1.5399,
+      "step": 46740
+    },
+    {
+      "epoch": 9.848843945857691,
+      "grad_norm": 1.2565813064575195,
+      "learning_rate": 1.1042083779811973e-07,
+      "loss": 1.5436,
+      "step": 46750
+    },
+    {
+      "epoch": 9.850950650445041,
+      "grad_norm": 1.2496920824050903,
+      "learning_rate": 1.0733284883682749e-07,
+      "loss": 1.5691,
+      "step": 46760
+    },
+    {
+      "epoch": 9.85305735503239,
+      "grad_norm": 1.1770131587982178,
+      "learning_rate": 1.0428863000899735e-07,
+      "loss": 1.6555,
+      "step": 46770
+    },
+    {
+      "epoch": 9.85516405961974,
+      "grad_norm": 1.2371290922164917,
+      "learning_rate": 1.0128818264851791e-07,
+      "loss": 1.5574,
+      "step": 46780
+    },
+    {
+      "epoch": 9.857270764207088,
+      "grad_norm": 1.1748226881027222,
+      "learning_rate": 9.833150807009306e-08,
+      "loss": 1.5877,
+      "step": 46790
+    },
+    {
+      "epoch": 9.859377468794438,
+      "grad_norm": 1.3029236793518066,
+      "learning_rate": 9.541860756925314e-08,
+      "loss": 1.5687,
+      "step": 46800
+    },
+    {
+      "epoch": 9.861484173381788,
+      "grad_norm": 1.1565899848937988,
+      "learning_rate": 9.254948242235495e-08,
+      "loss": 1.5244,
+      "step": 46810
+    },
+    {
+      "epoch": 9.863590877969138,
+      "grad_norm": 1.278841495513916,
+      "learning_rate": 8.972413388657064e-08,
+      "loss": 1.5342,
+      "step": 46820
+    },
+    {
+      "epoch": 9.865697582556486,
+      "grad_norm": 1.2975276708602905,
+      "learning_rate": 8.694256319987659e-08,
+      "loss": 1.5355,
+      "step": 46830
+    },
+    {
+      "epoch": 9.867804287143835,
+      "grad_norm": 1.1724909543991089,
+      "learning_rate": 8.420477158107565e-08,
+      "loss": 1.5269,
+      "step": 46840
+    },
+    {
+      "epoch": 9.869910991731185,
+      "grad_norm": 1.3155508041381836,
+      "learning_rate": 8.151076022980819e-08,
+      "loss": 1.5505,
+      "step": 46850
+    },
+    {
+      "epoch": 9.872017696318533,
+      "grad_norm": 1.251495599746704,
+      "learning_rate": 7.886053032649665e-08,
+      "loss": 1.5621,
+      "step": 46860
+    },
+    {
+      "epoch": 9.874124400905883,
+      "grad_norm": 1.429298996925354,
+      "learning_rate": 7.62540830324121e-08,
+      "loss": 1.5938,
+      "step": 46870
+    },
+    {
+      "epoch": 9.876231105493233,
+      "grad_norm": 1.2667317390441895,
+      "learning_rate": 7.369141948960767e-08,
+      "loss": 1.544,
+      "step": 46880
+    },
+    {
+      "epoch": 9.87833781008058,
+      "grad_norm": 1.188432216644287,
+      "learning_rate": 7.117254082098512e-08,
+      "loss": 1.5984,
+      "step": 46890
+    },
+    {
+      "epoch": 9.88044451466793,
+      "grad_norm": 1.162582516670227,
+      "learning_rate": 6.869744813023937e-08,
+      "loss": 1.5626,
+      "step": 46900
+    },
+    {
+      "epoch": 9.88255121925528,
+      "grad_norm": 1.2468620538711548,
+      "learning_rate": 6.626614250188068e-08,
+      "loss": 1.6147,
+      "step": 46910
+    },
+    {
+      "epoch": 9.88465792384263,
+      "grad_norm": 1.2421079874038696,
+      "learning_rate": 6.387862500125685e-08,
+      "loss": 1.5524,
+      "step": 46920
+    },
+    {
+      "epoch": 9.886764628429978,
+      "grad_norm": 1.3621149063110352,
+      "learning_rate": 6.153489667448664e-08,
+      "loss": 1.5235,
+      "step": 46930
+    },
+    {
+      "epoch": 9.888871333017327,
+      "grad_norm": 1.3506252765655518,
+      "learning_rate": 5.9234958548537445e-08,
+      "loss": 1.5795,
+      "step": 46940
+    },
+    {
+      "epoch": 9.890978037604677,
+      "grad_norm": 1.1938551664352417,
+      "learning_rate": 5.697881163118091e-08,
+      "loss": 1.513,
+      "step": 46950
+    },
+    {
+      "epoch": 9.893084742192027,
+      "grad_norm": 1.2343593835830688,
+      "learning_rate": 5.476645691098181e-08,
+      "loss": 1.5676,
+      "step": 46960
+    },
+    {
+      "epoch": 9.895191446779375,
+      "grad_norm": 1.3996899127960205,
+      "learning_rate": 5.25978953573536e-08,
+      "loss": 1.5635,
+      "step": 46970
+    },
+    {
+      "epoch": 9.897298151366725,
+      "grad_norm": 1.1408532857894897,
+      "learning_rate": 5.047312792046954e-08,
+      "loss": 1.5245,
+      "step": 46980
+    },
+    {
+      "epoch": 9.899404855954074,
+      "grad_norm": 1.3579094409942627,
+      "learning_rate": 4.839215553137377e-08,
+      "loss": 1.503,
+      "step": 46990
+    },
+    {
+      "epoch": 9.901511560541422,
+      "grad_norm": 1.0767526626586914,
+      "learning_rate": 4.6354979101870254e-08,
+      "loss": 1.5036,
+      "step": 47000
+    },
+    {
+      "epoch": 9.903618265128772,
+      "grad_norm": 1.2614816427230835,
+      "learning_rate": 4.4361599524589406e-08,
+      "loss": 1.6215,
+      "step": 47010
+    },
+    {
+      "epoch": 9.905724969716122,
+      "grad_norm": 1.347352147102356,
+      "learning_rate": 4.241201767298808e-08,
+      "loss": 1.5626,
+      "step": 47020
+    },
+    {
+      "epoch": 9.907831674303472,
+      "grad_norm": 1.3003053665161133,
+      "learning_rate": 4.0506234401305186e-08,
+      "loss": 1.6067,
+      "step": 47030
+    },
+    {
+      "epoch": 9.90993837889082,
+      "grad_norm": 1.2332658767700195,
+      "learning_rate": 3.8644250544594975e-08,
+      "loss": 1.5574,
+      "step": 47040
+    },
+    {
+      "epoch": 9.91204508347817,
+      "grad_norm": 1.1698170900344849,
+      "learning_rate": 3.682606691874924e-08,
+      "loss": 1.5384,
+      "step": 47050
+    },
+    {
+      "epoch": 9.914151788065519,
+      "grad_norm": 1.2463845014572144,
+      "learning_rate": 3.505168432043071e-08,
+      "loss": 1.5131,
+      "step": 47060
+    },
+    {
+      "epoch": 9.916258492652867,
+      "grad_norm": 1.2173086404800415,
+      "learning_rate": 3.332110352712858e-08,
+      "loss": 1.5339,
+      "step": 47070
+    },
+    {
+      "epoch": 9.918365197240217,
+      "grad_norm": 1.2471425533294678,
+      "learning_rate": 3.1634325297114076e-08,
+      "loss": 1.589,
+      "step": 47080
+    },
+    {
+      "epoch": 9.920471901827566,
+      "grad_norm": 1.269729733467102,
+      "learning_rate": 2.999135036951817e-08,
+      "loss": 1.5994,
+      "step": 47090
+    },
+    {
+      "epoch": 9.922578606414916,
+      "grad_norm": 1.228976845741272,
+      "learning_rate": 2.839217946422057e-08,
+      "loss": 1.5785,
+      "step": 47100
+    },
+    {
+      "epoch": 9.924685311002264,
+      "grad_norm": 1.2782270908355713,
+      "learning_rate": 2.6836813281938543e-08,
+      "loss": 1.4929,
+      "step": 47110
+    },
+    {
+      "epoch": 9.926792015589614,
+      "grad_norm": 1.2272552251815796,
+      "learning_rate": 2.53252525042047e-08,
+      "loss": 1.5998,
+      "step": 47120
+    },
+    {
+      "epoch": 9.928898720176964,
+      "grad_norm": 1.2180017232894897,
+      "learning_rate": 2.385749779332258e-08,
+      "loss": 1.5487,
+      "step": 47130
+    },
+    {
+      "epoch": 9.931005424764312,
+      "grad_norm": 1.21531081199646,
+      "learning_rate": 2.243354979242218e-08,
+      "loss": 1.5415,
+      "step": 47140
+    },
+    {
+      "epoch": 9.933112129351661,
+      "grad_norm": 1.1451653242111206,
+      "learning_rate": 2.1053409125448843e-08,
+      "loss": 1.5198,
+      "step": 47150
+    },
+    {
+      "epoch": 9.935218833939011,
+      "grad_norm": 1.2328366041183472,
+      "learning_rate": 1.971707639712994e-08,
+      "loss": 1.5671,
+      "step": 47160
+    },
+    {
+      "epoch": 9.93732553852636,
+      "grad_norm": 1.4899163246154785,
+      "learning_rate": 1.8424552193019308e-08,
+      "loss": 1.604,
+      "step": 47170
+    },
+    {
+      "epoch": 9.939432243113709,
+      "grad_norm": 1.3283289670944214,
+      "learning_rate": 1.7175837079452804e-08,
+      "loss": 1.5649,
+      "step": 47180
+    },
+    {
+      "epoch": 9.941538947701059,
+      "grad_norm": 1.2755736112594604,
+      "learning_rate": 1.5970931603592752e-08,
+      "loss": 1.5319,
+      "step": 47190
+    },
+    {
+      "epoch": 9.943645652288408,
+      "grad_norm": 1.3269785642623901,
+      "learning_rate": 1.4809836293394608e-08,
+      "loss": 1.5431,
+      "step": 47200
+    },
+    {
+      "epoch": 9.945752356875758,
+      "grad_norm": 1.18012535572052,
+      "learning_rate": 1.3692551657595865e-08,
+      "loss": 1.5568,
+      "step": 47210
+    },
+    {
+      "epoch": 9.947859061463106,
+      "grad_norm": 1.5265027284622192,
+      "learning_rate": 1.2619078185793776e-08,
+      "loss": 1.5734,
+      "step": 47220
+    },
+    {
+      "epoch": 9.949965766050456,
+      "grad_norm": 1.2340668439865112,
+      "learning_rate": 1.1589416348323224e-08,
+      "loss": 1.6149,
+      "step": 47230
+    },
+    {
+      "epoch": 9.952072470637805,
+      "grad_norm": 1.254093885421753,
+      "learning_rate": 1.0603566596367742e-08,
+      "loss": 1.5872,
+      "step": 47240
+    },
+    {
+      "epoch": 9.954179175225153,
+      "grad_norm": 1.224780797958374,
+      "learning_rate": 9.661529361892907e-09,
+      "loss": 1.5545,
+      "step": 47250
+    },
+    {
+      "epoch": 9.956285879812503,
+      "grad_norm": 1.172776699066162,
+      "learning_rate": 8.76330505769074e-09,
+      "loss": 1.5728,
+      "step": 47260
+    },
+    {
+      "epoch": 9.958392584399853,
+      "grad_norm": 1.1891138553619385,
+      "learning_rate": 7.908894077301998e-09,
+      "loss": 1.522,
+      "step": 47270
+    },
+    {
+      "epoch": 9.9604992889872,
+      "grad_norm": 1.2386298179626465,
+      "learning_rate": 7.098296795138293e-09,
+      "loss": 1.5344,
+      "step": 47280
+    },
+    {
+      "epoch": 9.96260599357455,
+      "grad_norm": 1.303057074546814,
+      "learning_rate": 6.331513566371072e-09,
+      "loss": 1.5502,
+      "step": 47290
+    },
+    {
+      "epoch": 9.9647126981619,
+      "grad_norm": 1.3631552457809448,
+      "learning_rate": 5.608544726976029e-09,
+      "loss": 1.5891,
+      "step": 47300
+    },
+    {
+      "epoch": 9.96681940274925,
+      "grad_norm": 1.2292252779006958,
+      "learning_rate": 4.929390593744199e-09,
+      "loss": 1.6031,
+      "step": 47310
+    },
+    {
+      "epoch": 9.968926107336598,
+      "grad_norm": 1.2337449789047241,
+      "learning_rate": 4.2940514642597626e-09,
+      "loss": 1.5657,
+      "step": 47320
+    },
+    {
+      "epoch": 9.971032811923948,
+      "grad_norm": 1.3655712604522705,
+      "learning_rate": 3.7025276169000424e-09,
+      "loss": 1.5833,
+      "step": 47330
+    },
+    {
+      "epoch": 9.973139516511297,
+      "grad_norm": 1.205053687095642,
+      "learning_rate": 3.154819310868806e-09,
+      "loss": 1.5473,
+      "step": 47340
+    },
+    {
+      "epoch": 9.975246221098647,
+      "grad_norm": 1.3029731512069702,
+      "learning_rate": 2.650926786151864e-09,
+      "loss": 1.538,
+      "step": 47350
+    },
+    {
+      "epoch": 9.977352925685995,
+      "grad_norm": 1.1505680084228516,
+      "learning_rate": 2.19085026355037e-09,
+      "loss": 1.5799,
+      "step": 47360
+    },
+    {
+      "epoch": 9.979459630273345,
+      "grad_norm": 1.2553126811981201,
+      "learning_rate": 1.7745899446364178e-09,
+      "loss": 1.5353,
+      "step": 47370
+    },
+    {
+      "epoch": 9.981566334860695,
+      "grad_norm": 1.342659831047058,
+      "learning_rate": 1.4021460118085472e-09,
+      "loss": 1.4656,
+      "step": 47380
+    },
+    {
+      "epoch": 9.983673039448043,
+      "grad_norm": 1.3252551555633545,
+      "learning_rate": 1.073518628269543e-09,
+      "loss": 1.5514,
+      "step": 47390
+    },
+    {
+      "epoch": 9.985779744035392,
+      "grad_norm": 1.1547834873199463,
+      "learning_rate": 7.887079380153317e-10,
+      "loss": 1.5357,
+      "step": 47400
+    },
+    {
+      "epoch": 9.987886448622742,
+      "grad_norm": 1.2872856855392456,
+      "learning_rate": 5.477140658349811e-10,
+      "loss": 1.4944,
+      "step": 47410
+    },
+    {
+      "epoch": 9.989993153210092,
+      "grad_norm": 1.1849569082260132,
+      "learning_rate": 3.5053711733290527e-10,
+      "loss": 1.6425,
+      "step": 47420
+    },
+    {
+      "epoch": 9.99209985779744,
+      "grad_norm": 1.1810338497161865,
+      "learning_rate": 1.9717717889555787e-10,
+      "loss": 1.5292,
+      "step": 47430
+    },
+    {
+      "epoch": 9.99420656238479,
+      "grad_norm": 1.142162799835205,
+      "learning_rate": 8.763431773584075e-11,
+      "loss": 1.5462,
+      "step": 47440
+    },
+    {
+      "epoch": 9.99631326697214,
+      "grad_norm": 1.3802740573883057,
+      "learning_rate": 2.1908581837593033e-11,
+      "loss": 1.5825,
+      "step": 47450
+    },
+    {
+      "epoch": 9.998419971559487,
+      "grad_norm": 1.2215389013290405,
+      "learning_rate": 0.0,
+      "loss": 1.6347,
+      "step": 47460
+    },
+    {
+      "epoch": 9.998419971559487,
+      "step": 47460,
+      "total_flos": 1.2118866375632486e+18,
+      "train_loss": 1.6495774427446894,
+      "train_runtime": 50831.8889,
+      "train_samples_per_second": 3.735,
+      "train_steps_per_second": 0.934
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 47460,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 1.2118866375632486e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}