diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12634 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.963998270928964,
+  "eval_steps": 500,
+  "global_step": 18000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016467343200016467,
+      "grad_norm": 30.985063552856445,
+      "learning_rate": 3.2906764168190127e-07,
+      "loss": 1.3972,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032934686400032933,
+      "grad_norm": 31.839323043823242,
+      "learning_rate": 6.946983546617917e-07,
+      "loss": 1.3292,
+      "step": 20
+    },
+    {
+      "epoch": 0.0049402029600049404,
+      "grad_norm": 34.46781921386719,
+      "learning_rate": 1.060329067641682e-06,
+      "loss": 1.2296,
+      "step": 30
+    },
+    {
+      "epoch": 0.006586937280006587,
+      "grad_norm": 34.32807540893555,
+      "learning_rate": 1.4259597806215722e-06,
+      "loss": 1.2755,
+      "step": 40
+    },
+    {
+      "epoch": 0.008233671600008234,
+      "grad_norm": 24.143030166625977,
+      "learning_rate": 1.7915904936014627e-06,
+      "loss": 1.2333,
+      "step": 50
+    },
+    {
+      "epoch": 0.009880405920009881,
+      "grad_norm": 20.771278381347656,
+      "learning_rate": 2.157221206581353e-06,
+      "loss": 1.206,
+      "step": 60
+    },
+    {
+      "epoch": 0.011527140240011528,
+      "grad_norm": 25.031925201416016,
+      "learning_rate": 2.5228519195612434e-06,
+      "loss": 1.2008,
+      "step": 70
+    },
+    {
+      "epoch": 0.013173874560013173,
+      "grad_norm": 24.008222579956055,
+      "learning_rate": 2.8884826325411334e-06,
+      "loss": 1.1446,
+      "step": 80
+    },
+    {
+      "epoch": 0.01482060888001482,
+      "grad_norm": 23.959667205810547,
+      "learning_rate": 3.254113345521024e-06,
+      "loss": 1.2072,
+      "step": 90
+    },
+    {
+      "epoch": 0.016467343200016468,
+      "grad_norm": 16.00537109375,
+      "learning_rate": 3.6197440585009143e-06,
+      "loss": 1.1743,
+      "step": 100
+    },
+    {
+      "epoch": 0.018114077520018115,
+      "grad_norm": 13.254168510437012,
+      "learning_rate": 3.985374771480805e-06,
+      "loss": 1.1374,
+      "step": 110
+    },
+    {
+      "epoch": 0.019760811840019762,
+      "grad_norm": 14.698610305786133,
+      "learning_rate": 4.351005484460696e-06,
+      "loss": 1.1192,
+      "step": 120
+    },
+    {
+      "epoch": 0.02140754616002141,
+      "grad_norm": 12.052921295166016,
+      "learning_rate": 4.716636197440586e-06,
+      "loss": 1.089,
+      "step": 130
+    },
+    {
+      "epoch": 0.023054280480023056,
+      "grad_norm": 9.61983871459961,
+      "learning_rate": 5.082266910420476e-06,
+      "loss": 1.0972,
+      "step": 140
+    },
+    {
+      "epoch": 0.0247010148000247,
+      "grad_norm": 10.221634864807129,
+      "learning_rate": 5.447897623400366e-06,
+      "loss": 1.1082,
+      "step": 150
+    },
+    {
+      "epoch": 0.026347749120026347,
+      "grad_norm": 14.476886749267578,
+      "learning_rate": 5.813528336380257e-06,
+      "loss": 1.1348,
+      "step": 160
+    },
+    {
+      "epoch": 0.027994483440027994,
+      "grad_norm": 23.15399169921875,
+      "learning_rate": 6.1791590493601475e-06,
+      "loss": 1.0889,
+      "step": 170
+    },
+    {
+      "epoch": 0.02964121776002964,
+      "grad_norm": 16.883651733398438,
+      "learning_rate": 6.544789762340037e-06,
+      "loss": 1.1055,
+      "step": 180
+    },
+    {
+      "epoch": 0.03128795208003129,
+      "grad_norm": 12.747838020324707,
+      "learning_rate": 6.9104204753199275e-06,
+      "loss": 1.0862,
+      "step": 190
+    },
+    {
+      "epoch": 0.032934686400032935,
+      "grad_norm": 11.923359870910645,
+      "learning_rate": 7.2760511882998175e-06,
+      "loss": 1.0944,
+      "step": 200
+    },
+    {
+      "epoch": 0.03458142072003458,
+      "grad_norm": 10.496978759765625,
+      "learning_rate": 7.641681901279708e-06,
+      "loss": 1.0962,
+      "step": 210
+    },
+    {
+      "epoch": 0.03622815504003623,
+      "grad_norm": 8.666916847229004,
+      "learning_rate": 8.007312614259598e-06,
+      "loss": 1.0826,
+      "step": 220
+    },
+    {
+      "epoch": 0.037874889360037876,
+      "grad_norm": 7.300290584564209,
+      "learning_rate": 8.372943327239488e-06,
+      "loss": 1.0932,
+      "step": 230
+    },
+    {
+      "epoch": 0.039521623680039523,
+      "grad_norm": 10.057311058044434,
+      "learning_rate": 8.73857404021938e-06,
+      "loss": 1.0466,
+      "step": 240
+    },
+    {
+      "epoch": 0.04116835800004117,
+      "grad_norm": 11.838115692138672,
+      "learning_rate": 9.10420475319927e-06,
+      "loss": 1.0881,
+      "step": 250
+    },
+    {
+      "epoch": 0.04281509232004282,
+      "grad_norm": 6.902042388916016,
+      "learning_rate": 9.469835466179161e-06,
+      "loss": 1.0618,
+      "step": 260
+    },
+    {
+      "epoch": 0.044461826640044465,
+      "grad_norm": 5.424289703369141,
+      "learning_rate": 9.83546617915905e-06,
+      "loss": 1.0146,
+      "step": 270
+    },
+    {
+      "epoch": 0.04610856096004611,
+      "grad_norm": 5.259349346160889,
+      "learning_rate": 1.020109689213894e-05,
+      "loss": 0.9994,
+      "step": 280
+    },
+    {
+      "epoch": 0.04775529528004775,
+      "grad_norm": 6.474951267242432,
+      "learning_rate": 1.0566727605118832e-05,
+      "loss": 1.0003,
+      "step": 290
+    },
+    {
+      "epoch": 0.0494020296000494,
+      "grad_norm": 7.7709455490112305,
+      "learning_rate": 1.0932358318098721e-05,
+      "loss": 1.0263,
+      "step": 300
+    },
+    {
+      "epoch": 0.051048763920051046,
+      "grad_norm": 8.521011352539062,
+      "learning_rate": 1.129798903107861e-05,
+      "loss": 1.0383,
+      "step": 310
+    },
+    {
+      "epoch": 0.05269549824005269,
+      "grad_norm": 6.253118515014648,
+      "learning_rate": 1.1663619744058501e-05,
+      "loss": 1.0609,
+      "step": 320
+    },
+    {
+      "epoch": 0.05434223256005434,
+      "grad_norm": 6.294135093688965,
+      "learning_rate": 1.2029250457038392e-05,
+      "loss": 1.0212,
+      "step": 330
+    },
+    {
+      "epoch": 0.05598896688005599,
+      "grad_norm": 8.303707122802734,
+      "learning_rate": 1.2394881170018283e-05,
+      "loss": 0.9964,
+      "step": 340
+    },
+    {
+      "epoch": 0.057635701200057635,
+      "grad_norm": 4.852534294128418,
+      "learning_rate": 1.2760511882998172e-05,
+      "loss": 0.9875,
+      "step": 350
+    },
+    {
+      "epoch": 0.05928243552005928,
+      "grad_norm": 6.304441928863525,
+      "learning_rate": 1.3126142595978065e-05,
+      "loss": 1.0425,
+      "step": 360
+    },
+    {
+      "epoch": 0.06092916984006093,
+      "grad_norm": 5.18842887878418,
+      "learning_rate": 1.3491773308957954e-05,
+      "loss": 1.0101,
+      "step": 370
+    },
+    {
+      "epoch": 0.06257590416006258,
+      "grad_norm": 5.784980773925781,
+      "learning_rate": 1.3857404021937843e-05,
+      "loss": 1.0046,
+      "step": 380
+    },
+    {
+      "epoch": 0.06422263848006422,
+      "grad_norm": 8.19119930267334,
+      "learning_rate": 1.4223034734917734e-05,
+      "loss": 0.9889,
+      "step": 390
+    },
+    {
+      "epoch": 0.06586937280006587,
+      "grad_norm": 6.851486682891846,
+      "learning_rate": 1.4588665447897625e-05,
+      "loss": 1.0111,
+      "step": 400
+    },
+    {
+      "epoch": 0.06751610712006752,
+      "grad_norm": 4.777026176452637,
+      "learning_rate": 1.4954296160877516e-05,
+      "loss": 0.9922,
+      "step": 410
+    },
+    {
+      "epoch": 0.06916284144006916,
+      "grad_norm": 5.17830753326416,
+      "learning_rate": 1.5319926873857403e-05,
+      "loss": 0.9556,
+      "step": 420
+    },
+    {
+      "epoch": 0.07080957576007081,
+      "grad_norm": 5.6952080726623535,
+      "learning_rate": 1.5685557586837297e-05,
+      "loss": 0.9696,
+      "step": 430
+    },
+    {
+      "epoch": 0.07245631008007246,
+      "grad_norm": 4.926089286804199,
+      "learning_rate": 1.6051188299817185e-05,
+      "loss": 0.9668,
+      "step": 440
+    },
+    {
+      "epoch": 0.0741030444000741,
+      "grad_norm": 4.33101749420166,
+      "learning_rate": 1.6416819012797076e-05,
+      "loss": 0.9884,
+      "step": 450
+    },
+    {
+      "epoch": 0.07574977872007575,
+      "grad_norm": 3.4968671798706055,
+      "learning_rate": 1.6782449725776967e-05,
+      "loss": 0.9852,
+      "step": 460
+    },
+    {
+      "epoch": 0.0773965130400774,
+      "grad_norm": 3.997958183288574,
+      "learning_rate": 1.7148080438756858e-05,
+      "loss": 0.9675,
+      "step": 470
+    },
+    {
+      "epoch": 0.07904324736007905,
+      "grad_norm": 4.513462066650391,
+      "learning_rate": 1.751371115173675e-05,
+      "loss": 0.9886,
+      "step": 480
+    },
+    {
+      "epoch": 0.0806899816800807,
+      "grad_norm": 4.011235237121582,
+      "learning_rate": 1.7879341864716636e-05,
+      "loss": 0.9802,
+      "step": 490
+    },
+    {
+      "epoch": 0.08233671600008234,
+      "grad_norm": 4.776538848876953,
+      "learning_rate": 1.8244972577696527e-05,
+      "loss": 0.928,
+      "step": 500
+    },
+    {
+      "epoch": 0.08398345032008399,
+      "grad_norm": 4.51215934753418,
+      "learning_rate": 1.8610603290676418e-05,
+      "loss": 0.9286,
+      "step": 510
+    },
+    {
+      "epoch": 0.08563018464008564,
+      "grad_norm": 3.450950860977173,
+      "learning_rate": 1.897623400365631e-05,
+      "loss": 0.9485,
+      "step": 520
+    },
+    {
+      "epoch": 0.08727691896008728,
+      "grad_norm": 3.313753604888916,
+      "learning_rate": 1.93418647166362e-05,
+      "loss": 0.9803,
+      "step": 530
+    },
+    {
+      "epoch": 0.08892365328008893,
+      "grad_norm": 4.609573841094971,
+      "learning_rate": 1.970749542961609e-05,
+      "loss": 0.9782,
+      "step": 540
+    },
+    {
+      "epoch": 0.09057038760009058,
+      "grad_norm": 3.477215528488159,
+      "learning_rate": 1.9999999367939968e-05,
+      "loss": 0.9189,
+      "step": 550
+    },
+    {
+      "epoch": 0.09221712192009222,
+      "grad_norm": 5.594038963317871,
+      "learning_rate": 1.9999977245847137e-05,
+      "loss": 0.9146,
+      "step": 560
+    },
+    {
+      "epoch": 0.09386385624009386,
+      "grad_norm": 2.6363534927368164,
+      "learning_rate": 1.9999923520832466e-05,
+      "loss": 0.895,
+      "step": 570
+    },
+    {
+      "epoch": 0.0955105905600955,
+      "grad_norm": 3.1434268951416016,
+      "learning_rate": 1.999983819306574e-05,
+      "loss": 0.9041,
+      "step": 580
+    },
+    {
+      "epoch": 0.09715732488009715,
+      "grad_norm": 3.609675168991089,
+      "learning_rate": 1.999972126281662e-05,
+      "loss": 0.9154,
+      "step": 590
+    },
+    {
+      "epoch": 0.0988040592000988,
+      "grad_norm": 4.128945350646973,
+      "learning_rate": 1.9999572730454638e-05,
+      "loss": 0.9061,
+      "step": 600
+    },
+    {
+      "epoch": 0.10045079352010045,
+      "grad_norm": 3.6238315105438232,
+      "learning_rate": 1.999939259644921e-05,
+      "loss": 0.9326,
+      "step": 610
+    },
+    {
+      "epoch": 0.10209752784010209,
+      "grad_norm": 3.4944262504577637,
+      "learning_rate": 1.9999180861369605e-05,
+      "loss": 0.894,
+      "step": 620
+    },
+    {
+      "epoch": 0.10374426216010374,
+      "grad_norm": 2.8273308277130127,
+      "learning_rate": 1.999893752588497e-05,
+      "loss": 0.9058,
+      "step": 630
+    },
+    {
+      "epoch": 0.10539099648010539,
+      "grad_norm": 3.1057746410369873,
+      "learning_rate": 1.999866259076432e-05,
+      "loss": 0.8943,
+      "step": 640
+    },
+    {
+      "epoch": 0.10703773080010703,
+      "grad_norm": 4.640377998352051,
+      "learning_rate": 1.9998356056876532e-05,
+      "loss": 0.9156,
+      "step": 650
+    },
+    {
+      "epoch": 0.10868446512010868,
+      "grad_norm": 4.991531848907471,
+      "learning_rate": 1.9998017925190345e-05,
+      "loss": 0.9163,
+      "step": 660
+    },
+    {
+      "epoch": 0.11033119944011033,
+      "grad_norm": 7.1393327713012695,
+      "learning_rate": 1.9997648196774354e-05,
+      "loss": 0.8828,
+      "step": 670
+    },
+    {
+      "epoch": 0.11197793376011198,
+      "grad_norm": 3.893332004547119,
+      "learning_rate": 1.9997246872797018e-05,
+      "loss": 0.9226,
+      "step": 680
+    },
+    {
+      "epoch": 0.11362466808011362,
+      "grad_norm": 3.794612169265747,
+      "learning_rate": 1.999681395452663e-05,
+      "loss": 0.8738,
+      "step": 690
+    },
+    {
+      "epoch": 0.11527140240011527,
+      "grad_norm": 3.3877007961273193,
+      "learning_rate": 1.9996349443331354e-05,
+      "loss": 0.8742,
+      "step": 700
+    },
+    {
+      "epoch": 0.11691813672011692,
+      "grad_norm": 5.569369316101074,
+      "learning_rate": 1.9995853340679173e-05,
+      "loss": 0.8674,
+      "step": 710
+    },
+    {
+      "epoch": 0.11856487104011856,
+      "grad_norm": 3.396794080734253,
+      "learning_rate": 1.999532564813793e-05,
+      "loss": 0.914,
+      "step": 720
+    },
+    {
+      "epoch": 0.12021160536012021,
+      "grad_norm": 5.97723388671875,
+      "learning_rate": 1.9994766367375283e-05,
+      "loss": 0.8892,
+      "step": 730
+    },
+    {
+      "epoch": 0.12185833968012186,
+      "grad_norm": 3.1893229484558105,
+      "learning_rate": 1.999417550015873e-05,
+      "loss": 0.8554,
+      "step": 740
+    },
+    {
+      "epoch": 0.1235050740001235,
+      "grad_norm": 3.7248144149780273,
+      "learning_rate": 1.999355304835559e-05,
+      "loss": 0.857,
+      "step": 750
+    },
+    {
+      "epoch": 0.12515180832012515,
+      "grad_norm": 3.797175645828247,
+      "learning_rate": 1.9992899013932994e-05,
+      "loss": 0.8646,
+      "step": 760
+    },
+    {
+      "epoch": 0.1267985426401268,
+      "grad_norm": 2.7383298873901367,
+      "learning_rate": 1.999221339895789e-05,
+      "loss": 0.8749,
+      "step": 770
+    },
+    {
+      "epoch": 0.12844527696012845,
+      "grad_norm": 2.588918447494507,
+      "learning_rate": 1.9991496205597023e-05,
+      "loss": 0.8897,
+      "step": 780
+    },
+    {
+      "epoch": 0.1300920112801301,
+      "grad_norm": 2.533170223236084,
+      "learning_rate": 1.999074743611694e-05,
+      "loss": 0.8611,
+      "step": 790
+    },
+    {
+      "epoch": 0.13173874560013174,
+      "grad_norm": 2.219616174697876,
+      "learning_rate": 1.998996709288398e-05,
+      "loss": 0.8416,
+      "step": 800
+    },
+    {
+      "epoch": 0.1333854799201334,
+      "grad_norm": 2.3094987869262695,
+      "learning_rate": 1.9989155178364253e-05,
+      "loss": 0.8401,
+      "step": 810
+    },
+    {
+      "epoch": 0.13503221424013503,
+      "grad_norm": 1.8022428750991821,
+      "learning_rate": 1.998831169512366e-05,
+      "loss": 0.8049,
+      "step": 820
+    },
+    {
+      "epoch": 0.13667894856013668,
+      "grad_norm": 1.8802989721298218,
+      "learning_rate": 1.998743664582786e-05,
+      "loss": 0.829,
+      "step": 830
+    },
+    {
+      "epoch": 0.13832568288013833,
+      "grad_norm": 2.1697161197662354,
+      "learning_rate": 1.9986530033242263e-05,
+      "loss": 0.8503,
+      "step": 840
+    },
+    {
+      "epoch": 0.13997241720013998,
+      "grad_norm": 4.128193378448486,
+      "learning_rate": 1.9985591860232047e-05,
+      "loss": 0.8472,
+      "step": 850
+    },
+    {
+      "epoch": 0.14161915152014162,
+      "grad_norm": 3.771169900894165,
+      "learning_rate": 1.9984622129762116e-05,
+      "loss": 0.8543,
+      "step": 860
+    },
+    {
+      "epoch": 0.14326588584014327,
+      "grad_norm": 3.2667131423950195,
+      "learning_rate": 1.99836208448971e-05,
+      "loss": 0.8512,
+      "step": 870
+    },
+    {
+      "epoch": 0.14491262016014492,
+      "grad_norm": 2.6572048664093018,
+      "learning_rate": 1.9982588008801368e-05,
+      "loss": 0.8274,
+      "step": 880
+    },
+    {
+      "epoch": 0.14655935448014656,
+      "grad_norm": 2.6121675968170166,
+      "learning_rate": 1.998152362473899e-05,
+      "loss": 0.8373,
+      "step": 890
+    },
+    {
+      "epoch": 0.1482060888001482,
+      "grad_norm": 3.6851730346679688,
+      "learning_rate": 1.998042769607374e-05,
+      "loss": 0.8262,
+      "step": 900
+    },
+    {
+      "epoch": 0.14985282312014986,
+      "grad_norm": 3.212480306625366,
+      "learning_rate": 1.9979300226269077e-05,
+      "loss": 0.8379,
+      "step": 910
+    },
+    {
+      "epoch": 0.1514995574401515,
+      "grad_norm": 2.4245805740356445,
+      "learning_rate": 1.9978141218888143e-05,
+      "loss": 0.8513,
+      "step": 920
+    },
+    {
+      "epoch": 0.15314629176015315,
+      "grad_norm": 3.501965284347534,
+      "learning_rate": 1.997695067759375e-05,
+      "loss": 0.8045,
+      "step": 930
+    },
+    {
+      "epoch": 0.1547930260801548,
+      "grad_norm": 2.5098695755004883,
+      "learning_rate": 1.997572860614836e-05,
+      "loss": 0.8123,
+      "step": 940
+    },
+    {
+      "epoch": 0.15643976040015645,
+      "grad_norm": 2.632624387741089,
+      "learning_rate": 1.9974475008414095e-05,
+      "loss": 0.8261,
+      "step": 950
+    },
+    {
+      "epoch": 0.1580864947201581,
+      "grad_norm": 2.489697217941284,
+      "learning_rate": 1.997318988835269e-05,
+      "loss": 0.8348,
+      "step": 960
+    },
+    {
+      "epoch": 0.15973322904015974,
+      "grad_norm": 4.488799571990967,
+      "learning_rate": 1.9971873250025512e-05,
+      "loss": 0.8684,
+      "step": 970
+    },
+    {
+      "epoch": 0.1613799633601614,
+      "grad_norm": 2.742992877960205,
+      "learning_rate": 1.9970525097593537e-05,
+      "loss": 0.7831,
+      "step": 980
+    },
+    {
+      "epoch": 0.16302669768016304,
+      "grad_norm": 2.506899118423462,
+      "learning_rate": 1.996914543531732e-05,
+      "loss": 0.8285,
+      "step": 990
+    },
+    {
+      "epoch": 0.16467343200016468,
+      "grad_norm": 3.0873513221740723,
+      "learning_rate": 1.996773426755702e-05,
+      "loss": 0.8215,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16632016632016633,
+      "grad_norm": 1.7364745140075684,
+      "learning_rate": 1.9966291598772335e-05,
+      "loss": 0.8385,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16796690064016798,
+      "grad_norm": 2.113279104232788,
+      "learning_rate": 1.9964817433522537e-05,
+      "loss": 0.8246,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16961363496016962,
+      "grad_norm": 2.741102457046509,
+      "learning_rate": 1.9963311776466435e-05,
+      "loss": 0.8454,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17126036928017127,
+      "grad_norm": 2.6016159057617188,
+      "learning_rate": 1.996177463236235e-05,
+      "loss": 0.8078,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17290710360017292,
+      "grad_norm": 2.431082010269165,
+      "learning_rate": 1.9960206006068116e-05,
+      "loss": 0.8548,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17455383792017456,
+      "grad_norm": 2.077765703201294,
+      "learning_rate": 1.9958605902541065e-05,
+      "loss": 0.8492,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1762005722401762,
+      "grad_norm": 3.3199009895324707,
+      "learning_rate": 1.9956974326838004e-05,
+      "loss": 0.7955,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17784730656017786,
+      "grad_norm": 2.1479249000549316,
+      "learning_rate": 1.9955311284115198e-05,
+      "loss": 0.7802,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1794940408801795,
+      "grad_norm": 1.943623661994934,
+      "learning_rate": 1.9953616779628364e-05,
+      "loss": 0.8206,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18114077520018115,
+      "grad_norm": 2.654792070388794,
+      "learning_rate": 1.995189081873264e-05,
+      "loss": 0.8148,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1827875095201828,
+      "grad_norm": 1.9163497686386108,
+      "learning_rate": 1.9950133406882577e-05,
+      "loss": 0.7941,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18443424384018445,
+      "grad_norm": 2.9743542671203613,
+      "learning_rate": 1.9948344549632124e-05,
+      "loss": 0.8019,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1860809781601861,
+      "grad_norm": 2.794917106628418,
+      "learning_rate": 1.9946524252634612e-05,
+      "loss": 0.8506,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18772771248018771,
+      "grad_norm": 2.309389352798462,
+      "learning_rate": 1.9944672521642715e-05,
+      "loss": 0.7851,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18937444680018936,
+      "grad_norm": 1.9234291315078735,
+      "learning_rate": 1.9942789362508463e-05,
+      "loss": 0.7941,
+      "step": 1150
+    },
+    {
+      "epoch": 0.191021181120191,
+      "grad_norm": 2.477848768234253,
+      "learning_rate": 1.9940874781183203e-05,
+      "loss": 0.7892,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19266791544019266,
+      "grad_norm": 2.7016117572784424,
+      "learning_rate": 1.993892878371758e-05,
+      "loss": 0.8266,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1943146497601943,
+      "grad_norm": 1.9896224737167358,
+      "learning_rate": 1.9936951376261534e-05,
+      "loss": 0.8227,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19596138408019595,
+      "grad_norm": 2.1344499588012695,
+      "learning_rate": 1.993494256506426e-05,
+      "loss": 0.8019,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1976081184001976,
+      "grad_norm": 2.2229461669921875,
+      "learning_rate": 1.9932902356474208e-05,
+      "loss": 0.7962,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19925485272019924,
+      "grad_norm": 2.2535150051116943,
+      "learning_rate": 1.993083075693904e-05,
+      "loss": 0.8405,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2009015870402009,
+      "grad_norm": 2.778207778930664,
+      "learning_rate": 1.9928727773005644e-05,
+      "loss": 0.8082,
+      "step": 1220
+    },
+    {
+      "epoch": 0.20254832136020254,
+      "grad_norm": 3.0831124782562256,
+      "learning_rate": 1.9926593411320064e-05,
+      "loss": 0.796,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20419505568020418,
+      "grad_norm": 2.5188772678375244,
+      "learning_rate": 1.9924427678627533e-05,
+      "loss": 0.8401,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20584179000020583,
+      "grad_norm": 2.3578946590423584,
+      "learning_rate": 1.9922230581772405e-05,
+      "loss": 0.7995,
+      "step": 1250
+    },
+    {
+      "epoch": 0.20748852432020748,
+      "grad_norm": 2.7173349857330322,
+      "learning_rate": 1.992000212769817e-05,
+      "loss": 0.8271,
+      "step": 1260
+    },
+    {
+      "epoch": 0.20913525864020913,
+      "grad_norm": 2.0116770267486572,
+      "learning_rate": 1.9917742323447414e-05,
+      "loss": 0.7978,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21078199296021077,
+      "grad_norm": 1.9019718170166016,
+      "learning_rate": 1.9915451176161788e-05,
+      "loss": 0.8141,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21242872728021242,
+      "grad_norm": 2.0212275981903076,
+      "learning_rate": 1.9913128693082e-05,
+      "loss": 0.773,
+      "step": 1290
+    },
+    {
+      "epoch": 0.21407546160021407,
+      "grad_norm": 3.2075860500335693,
+      "learning_rate": 1.9910774881547803e-05,
+      "loss": 0.7901,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21572219592021571,
+      "grad_norm": 2.278242588043213,
+      "learning_rate": 1.9908389748997937e-05,
+      "loss": 0.7616,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21736893024021736,
+      "grad_norm": 1.8260267972946167,
+      "learning_rate": 1.990597330297014e-05,
+      "loss": 0.7901,
+      "step": 1320
+    },
+    {
+      "epoch": 0.219015664560219,
+      "grad_norm": 2.169857978820801,
+      "learning_rate": 1.9903525551101105e-05,
+      "loss": 0.7491,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22066239888022066,
+      "grad_norm": 1.7144447565078735,
+      "learning_rate": 1.9901046501126454e-05,
+      "loss": 0.7546,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2223091332002223,
+      "grad_norm": 1.9034463167190552,
+      "learning_rate": 1.9898536160880736e-05,
+      "loss": 0.7679,
+      "step": 1350
+    },
+    {
+      "epoch": 0.22395586752022395,
+      "grad_norm": 1.7277398109436035,
+      "learning_rate": 1.989599453829737e-05,
+      "loss": 0.7254,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2256026018402256,
+      "grad_norm": 1.7360395193099976,
+      "learning_rate": 1.989342164140865e-05,
+      "loss": 0.7861,
+      "step": 1370
+    },
+    {
+      "epoch": 0.22724933616022724,
+      "grad_norm": 1.7167317867279053,
+      "learning_rate": 1.98908174783457e-05,
+      "loss": 0.7765,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2288960704802289,
+      "grad_norm": 2.2761967182159424,
+      "learning_rate": 1.988818205733845e-05,
+      "loss": 0.8211,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23054280480023054,
+      "grad_norm": 2.0242958068847656,
+      "learning_rate": 1.9885515386715625e-05,
+      "loss": 0.7516,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23218953912023219,
+      "grad_norm": 1.9908565282821655,
+      "learning_rate": 1.9882817474904697e-05,
+      "loss": 0.7685,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23383627344023383,
+      "grad_norm": 2.352423667907715,
+      "learning_rate": 1.9880088330431883e-05,
+      "loss": 0.7235,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23548300776023548,
+      "grad_norm": 1.5180648565292358,
+      "learning_rate": 1.9877327961922085e-05,
+      "loss": 0.7557,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23712974208023713,
+      "grad_norm": 2.784830331802368,
+      "learning_rate": 1.9874536378098905e-05,
+      "loss": 0.7864,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23877647640023877,
+      "grad_norm": 1.8382682800292969,
+      "learning_rate": 1.987171358778458e-05,
+      "loss": 0.7465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24042321072024042,
+      "grad_norm": 1.4438285827636719,
+      "learning_rate": 1.986885959989997e-05,
+      "loss": 0.7892,
+      "step": 1460
+    },
+    {
+      "epoch": 0.24206994504024207,
+      "grad_norm": 1.4113389253616333,
+      "learning_rate": 1.986597442346453e-05,
+      "loss": 0.7586,
+      "step": 1470
+    },
+    {
+      "epoch": 0.24371667936024372,
+      "grad_norm": 3.4214653968811035,
+      "learning_rate": 1.9863058067596287e-05,
+      "loss": 0.7564,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24536341368024536,
+      "grad_norm": 1.6341257095336914,
+      "learning_rate": 1.9860110541511792e-05,
+      "loss": 0.7493,
+      "step": 1490
+    },
+    {
+      "epoch": 0.247010148000247,
+      "grad_norm": 4.065557956695557,
+      "learning_rate": 1.9857131854526117e-05,
+      "loss": 0.768,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24865688232024866,
+      "grad_norm": 1.5984668731689453,
+      "learning_rate": 1.9854122016052803e-05,
+      "loss": 0.7631,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2503036166402503,
+      "grad_norm": 1.571628212928772,
+      "learning_rate": 1.9851081035603836e-05,
+      "loss": 0.7575,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2519503509602519,
+      "grad_norm": 2.2074406147003174,
+      "learning_rate": 1.9848008922789625e-05,
+      "loss": 0.7501,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2535970852802536,
+      "grad_norm": 1.4130685329437256,
+      "learning_rate": 1.984490568731897e-05,
+      "loss": 0.7653,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2552438196002552,
+      "grad_norm": 1.4988036155700684,
+      "learning_rate": 1.9841771338999022e-05,
+      "loss": 0.7544,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2568905539202569,
+      "grad_norm": 1.799906611442566,
+      "learning_rate": 1.9838605887735266e-05,
+      "loss": 0.7264,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2585372882402585,
+      "grad_norm": 2.0572450160980225,
+      "learning_rate": 1.9835409343531465e-05,
+      "loss": 0.7212,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2601840225602602,
+      "grad_norm": 1.9132790565490723,
+      "learning_rate": 1.9832181716489664e-05,
+      "loss": 0.7176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2618307568802618,
+      "grad_norm": 1.7707911729812622,
+      "learning_rate": 1.9828923016810123e-05,
+      "loss": 0.7647,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2634774912002635,
+      "grad_norm": 1.9048503637313843,
+      "learning_rate": 1.9825633254791318e-05,
+      "loss": 0.7202,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2651242255202651,
+      "grad_norm": 2.583421230316162,
+      "learning_rate": 1.9822312440829876e-05,
+      "loss": 0.73,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2667709598402668,
+      "grad_norm": 1.651800513267517,
+      "learning_rate": 1.9818960585420562e-05,
+      "loss": 0.706,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2684176941602684,
+      "grad_norm": 2.220543622970581,
+      "learning_rate": 1.981557769915625e-05,
+      "loss": 0.7246,
+      "step": 1630
+    },
+    {
+      "epoch": 0.27006442848027007,
+      "grad_norm": 1.6591778993606567,
+      "learning_rate": 1.9812163792727864e-05,
+      "loss": 0.7453,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2717111628002717,
+      "grad_norm": 1.8537503480911255,
+      "learning_rate": 1.9808718876924376e-05,
+      "loss": 0.7483,
+      "step": 1650
+    },
+    {
+      "epoch": 0.27335789712027336,
+      "grad_norm": 2.1067299842834473,
+      "learning_rate": 1.9805242962632747e-05,
+      "loss": 0.7047,
+      "step": 1660
+    },
+    {
+      "epoch": 0.275004631440275,
+      "grad_norm": 1.5523687601089478,
+      "learning_rate": 1.9801736060837913e-05,
+      "loss": 0.7427,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27665136576027666,
+      "grad_norm": 1.404558777809143,
+      "learning_rate": 1.9798198182622734e-05,
+      "loss": 0.7976,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2782981000802783,
+      "grad_norm": 2.101656675338745,
+      "learning_rate": 1.979462933916795e-05,
+      "loss": 0.7621,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27994483440027995,
+      "grad_norm": 2.0541417598724365,
+      "learning_rate": 1.9791029541752197e-05,
+      "loss": 0.7805,
+      "step": 1700
+    },
+    {
+      "epoch": 0.28159156872028157,
+      "grad_norm": 1.5349130630493164,
+      "learning_rate": 1.9787398801751895e-05,
+      "loss": 0.7399,
+      "step": 1710
+    },
+    {
+      "epoch": 0.28323830304028325,
+      "grad_norm": 2.281158447265625,
+      "learning_rate": 1.9783737130641272e-05,
+      "loss": 0.7575,
+      "step": 1720
+    },
+    {
+      "epoch": 0.28488503736028487,
+      "grad_norm": 2.4197065830230713,
+      "learning_rate": 1.978004453999231e-05,
+      "loss": 0.7578,
+      "step": 1730
+    },
+    {
+      "epoch": 0.28653177168028654,
+      "grad_norm": 2.0881662368774414,
+      "learning_rate": 1.97763210414747e-05,
+      "loss": 0.7549,
+      "step": 1740
+    },
+    {
+      "epoch": 0.28817850600028816,
+      "grad_norm": 2.4676826000213623,
+      "learning_rate": 1.9772566646855814e-05,
+      "loss": 0.741,
+      "step": 1750
+    },
+    {
+      "epoch": 0.28982524032028983,
+      "grad_norm": 1.893269419670105,
+      "learning_rate": 1.9768781368000658e-05,
+      "loss": 0.7416,
+      "step": 1760
+    },
+    {
+      "epoch": 0.29147197464029145,
+      "grad_norm": 1.5994690656661987,
+      "learning_rate": 1.9764965216871848e-05,
+      "loss": 0.7181,
+      "step": 1770
+    },
+    {
+      "epoch": 0.29311870896029313,
+      "grad_norm": 1.4995242357254028,
+      "learning_rate": 1.9761118205529565e-05,
+      "loss": 0.7389,
+      "step": 1780
+    },
+    {
+      "epoch": 0.29476544328029475,
+      "grad_norm": 1.3291387557983398,
+      "learning_rate": 1.9757240346131517e-05,
+      "loss": 0.7356,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2964121776002964,
+      "grad_norm": 1.4342174530029297,
+      "learning_rate": 1.9753331650932898e-05,
+      "loss": 0.722,
+      "step": 1800
+    },
+    {
+      "epoch": 0.29805891192029804,
+      "grad_norm": 1.4941635131835938,
+      "learning_rate": 1.9749392132286356e-05,
+      "loss": 0.7491,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2997056462402997,
+      "grad_norm": 1.4663833379745483,
+      "learning_rate": 1.974542180264195e-05,
+      "loss": 0.7724,
+      "step": 1820
+    },
+    {
+      "epoch": 0.30135238056030134,
+      "grad_norm": 1.420600414276123,
+      "learning_rate": 1.974142067454711e-05,
+      "loss": 0.7844,
+      "step": 1830
+    },
+    {
+      "epoch": 0.302999114880303,
+      "grad_norm": 1.7661523818969727,
+      "learning_rate": 1.97373887606466e-05,
+      "loss": 0.7277,
+      "step": 1840
+    },
+    {
+      "epoch": 0.30464584920030463,
+      "grad_norm": 2.7334187030792236,
+      "learning_rate": 1.9733326073682475e-05,
+      "loss": 0.7209,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3062925835203063,
+      "grad_norm": 1.8105331659317017,
+      "learning_rate": 1.972923262649404e-05,
+      "loss": 0.7499,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3079393178403079,
+      "grad_norm": 1.575403094291687,
+      "learning_rate": 1.9725108432017812e-05,
+      "loss": 0.7097,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3095860521603096,
+      "grad_norm": 1.4154794216156006,
+      "learning_rate": 1.9720953503287487e-05,
+      "loss": 0.7141,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3112327864803112,
+      "grad_norm": 1.4072984457015991,
+      "learning_rate": 1.9716767853433877e-05,
+      "loss": 0.7378,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3128795208003129,
+      "grad_norm": 1.3542795181274414,
+      "learning_rate": 1.971255149568489e-05,
+      "loss": 0.7385,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3145262551203145,
+      "grad_norm": 1.7643039226531982,
+      "learning_rate": 1.970830444336548e-05,
+      "loss": 0.7148,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3161729894403162,
+      "grad_norm": 1.285137414932251,
+      "learning_rate": 1.9704026709897606e-05,
+      "loss": 0.6827,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3178197237603178,
+      "grad_norm": 1.2715585231781006,
+      "learning_rate": 1.9699718308800182e-05,
+      "loss": 0.7133,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3194664580803195,
+      "grad_norm": 1.5506412982940674,
+      "learning_rate": 1.9695379253689048e-05,
+      "loss": 0.7238,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3211131924003211,
+      "grad_norm": 1.3223036527633667,
+      "learning_rate": 1.9691009558276915e-05,
+      "loss": 0.7131,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3227599267203228,
+      "grad_norm": 1.505676031112671,
+      "learning_rate": 1.9686609236373333e-05,
+      "loss": 0.7216,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3244066610403244,
+      "grad_norm": 1.330424427986145,
+      "learning_rate": 1.9682178301884632e-05,
+      "loss": 0.7296,
+      "step": 1970
+    },
+    {
+      "epoch": 0.32605339536032607,
+      "grad_norm": 1.6218839883804321,
+      "learning_rate": 1.9677716768813893e-05,
+      "loss": 0.6836,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3277001296803277,
+      "grad_norm": 1.3749337196350098,
+      "learning_rate": 1.9673224651260894e-05,
+      "loss": 0.7039,
+      "step": 1990
+    },
+    {
+      "epoch": 0.32934686400032936,
+      "grad_norm": 1.3419939279556274,
+      "learning_rate": 1.9668701963422077e-05,
+      "loss": 0.7181,
+      "step": 2000
+    },
+    {
+      "epoch": 0.330993598320331,
+      "grad_norm": 1.531684398651123,
+      "learning_rate": 1.9664148719590486e-05,
+      "loss": 0.7263,
+      "step": 2010
+    },
+    {
+      "epoch": 0.33264033264033266,
+      "grad_norm": 1.6776705980300903,
+      "learning_rate": 1.9659564934155733e-05,
+      "loss": 0.7279,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3342870669603343,
+      "grad_norm": 1.4876450300216675,
+      "learning_rate": 1.9654950621603955e-05,
+      "loss": 0.7441,
+      "step": 2030
+    },
+    {
+      "epoch": 0.33593380128033595,
+      "grad_norm": 1.3358787298202515,
+      "learning_rate": 1.965030579651776e-05,
+      "loss": 0.6973,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3375805356003376,
+      "grad_norm": 1.2709858417510986,
+      "learning_rate": 1.9645630473576184e-05,
+      "loss": 0.6895,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33922726992033925,
+      "grad_norm": 1.523409366607666,
+      "learning_rate": 1.9640924667554654e-05,
+      "loss": 0.7273,
+      "step": 2060
+    },
+    {
+      "epoch": 0.34087400424034087,
+      "grad_norm": 1.1860836744308472,
+      "learning_rate": 1.9636188393324917e-05,
+      "loss": 0.6914,
+      "step": 2070
+    },
+    {
+      "epoch": 0.34252073856034254,
+      "grad_norm": 1.5158321857452393,
+      "learning_rate": 1.9631421665855023e-05,
+      "loss": 0.6884,
+      "step": 2080
+    },
+    {
+      "epoch": 0.34416747288034416,
+      "grad_norm": 1.5086334943771362,
+      "learning_rate": 1.9626624500209254e-05,
+      "loss": 0.7352,
+      "step": 2090
+    },
+    {
+      "epoch": 0.34581420720034584,
+      "grad_norm": 1.5807071924209595,
+      "learning_rate": 1.9621796911548097e-05,
+      "loss": 0.6873,
+      "step": 2100
+    },
+    {
+      "epoch": 0.34746094152034745,
+      "grad_norm": 1.472090244293213,
+      "learning_rate": 1.961693891512817e-05,
+      "loss": 0.6839,
+      "step": 2110
+    },
+    {
+      "epoch": 0.34910767584034913,
+      "grad_norm": 1.2841448783874512,
+      "learning_rate": 1.9612050526302195e-05,
+      "loss": 0.7038,
+      "step": 2120
+    },
+    {
+      "epoch": 0.35075441016035075,
+      "grad_norm": 1.4147202968597412,
+      "learning_rate": 1.9607131760518952e-05,
+      "loss": 0.7,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3524011444803524,
+      "grad_norm": 1.4587483406066895,
+      "learning_rate": 1.9602182633323205e-05,
+      "loss": 0.6995,
+      "step": 2140
+    },
+    {
+      "epoch": 0.35404787880035404,
+      "grad_norm": 1.179916501045227,
+      "learning_rate": 1.9597203160355684e-05,
+      "loss": 0.6975,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3556946131203557,
+      "grad_norm": 1.621735692024231,
+      "learning_rate": 1.9592193357353012e-05,
+      "loss": 0.6808,
+      "step": 2160
+    },
+    {
+      "epoch": 0.35734134744035734,
+      "grad_norm": 1.5990266799926758,
+      "learning_rate": 1.9587153240147663e-05,
+      "loss": 0.7072,
+      "step": 2170
+    },
+    {
+      "epoch": 0.358988081760359,
+      "grad_norm": 1.6757946014404297,
+      "learning_rate": 1.9582082824667924e-05,
+      "loss": 0.7017,
+      "step": 2180
+    },
+    {
+      "epoch": 0.36063481608036063,
+      "grad_norm": 1.4000555276870728,
+      "learning_rate": 1.957698212693782e-05,
+      "loss": 0.704,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3622815504003623,
+      "grad_norm": 1.4697622060775757,
+      "learning_rate": 1.9571851163077082e-05,
+      "loss": 0.6987,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3639282847203639,
+      "grad_norm": 1.608054518699646,
+      "learning_rate": 1.9566689949301097e-05,
+      "loss": 0.6947,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3655750190403656,
+      "grad_norm": 1.957412600517273,
+      "learning_rate": 1.956149850192084e-05,
+      "loss": 0.6938,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3672217533603672,
+      "grad_norm": 1.398135781288147,
+      "learning_rate": 1.955627683734284e-05,
+      "loss": 0.6606,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3688684876803689,
+      "grad_norm": 1.6263389587402344,
+      "learning_rate": 1.9551024972069127e-05,
+      "loss": 0.7298,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3705152220003705,
+      "grad_norm": 1.461220145225525,
+      "learning_rate": 1.9545742922697157e-05,
+      "loss": 0.704,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3721619563203722,
+      "grad_norm": 1.1990222930908203,
+      "learning_rate": 1.9540430705919798e-05,
+      "loss": 0.6566,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3738086906403738,
+      "grad_norm": 1.6441187858581543,
+      "learning_rate": 1.9535088338525238e-05,
+      "loss": 0.7184,
+      "step": 2270
+    },
+    {
+      "epoch": 0.37545542496037543,
+      "grad_norm": 1.2525330781936646,
+      "learning_rate": 1.9529715837396956e-05,
+      "loss": 0.7407,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3771021592803771,
+      "grad_norm": 1.493485927581787,
+      "learning_rate": 1.952431321951367e-05,
+      "loss": 0.6515,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3787488936003787,
+      "grad_norm": 1.203551173210144,
+      "learning_rate": 1.9518880501949267e-05,
+      "loss": 0.7195,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3803956279203804,
+      "grad_norm": 1.6315183639526367,
+      "learning_rate": 1.9513417701872766e-05,
+      "loss": 0.6887,
+      "step": 2310
+    },
+    {
+      "epoch": 0.382042362240382,
+      "grad_norm": 1.2268880605697632,
+      "learning_rate": 1.9507924836548244e-05,
+      "loss": 0.7259,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3836890965603837,
+      "grad_norm": 1.1860028505325317,
+      "learning_rate": 1.9502401923334798e-05,
+      "loss": 0.671,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3853358308803853,
+      "grad_norm": 1.2253532409667969,
+      "learning_rate": 1.9496848979686493e-05,
+      "loss": 0.6954,
+      "step": 2340
+    },
+    {
+      "epoch": 0.386982565200387,
+      "grad_norm": 1.3530570268630981,
+      "learning_rate": 1.949126602315229e-05,
+      "loss": 0.7222,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3886292995203886,
+      "grad_norm": 1.353309988975525,
+      "learning_rate": 1.9485653071376004e-05,
+      "loss": 0.7268,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3902760338403903,
+      "grad_norm": 1.2534314393997192,
+      "learning_rate": 1.9480010142096245e-05,
+      "loss": 0.6925,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3919227681603919,
+      "grad_norm": 1.8099732398986816,
+      "learning_rate": 1.947433725314636e-05,
+      "loss": 0.659,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3935695024803936,
+      "grad_norm": 2.0360233783721924,
+      "learning_rate": 1.946863442245437e-05,
+      "loss": 0.6859,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3952162368003952,
+      "grad_norm": 1.5365337133407593,
+      "learning_rate": 1.946290166804293e-05,
+      "loss": 0.6973,
+      "step": 2400
+    },
+    {
+      "epoch": 0.39686297112039687,
+      "grad_norm": 1.4845314025878906,
+      "learning_rate": 1.9457139008029263e-05,
+      "loss": 0.7292,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3985097054403985,
+      "grad_norm": 1.3728680610656738,
+      "learning_rate": 1.94513464606251e-05,
+      "loss": 0.6981,
+      "step": 2420
+    },
+    {
+      "epoch": 0.40015643976040016,
+      "grad_norm": 1.3709412813186646,
+      "learning_rate": 1.9445524044136618e-05,
+      "loss": 0.7105,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4018031740804018,
+      "grad_norm": 1.321450114250183,
+      "learning_rate": 1.94396717769644e-05,
+      "loss": 0.6632,
+      "step": 2440
+    },
+    {
+      "epoch": 0.40344990840040346,
+      "grad_norm": 1.2133983373641968,
+      "learning_rate": 1.943378967760337e-05,
+      "loss": 0.6788,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4050966427204051,
+      "grad_norm": 1.2709228992462158,
+      "learning_rate": 1.9427877764642714e-05,
+      "loss": 0.683,
+      "step": 2460
+    },
+    {
+      "epoch": 0.40674337704040675,
+      "grad_norm": 1.167478084564209,
+      "learning_rate": 1.9421936056765847e-05,
+      "loss": 0.6599,
+      "step": 2470
+    },
+    {
+      "epoch": 0.40839011136040837,
+      "grad_norm": 1.3339264392852783,
+      "learning_rate": 1.9415964572750347e-05,
+      "loss": 0.6807,
+      "step": 2480
+    },
+    {
+      "epoch": 0.41003684568041004,
+      "grad_norm": 1.1604259014129639,
+      "learning_rate": 1.9409963331467893e-05,
+      "loss": 0.6634,
+      "step": 2490
+    },
+    {
+      "epoch": 0.41168358000041166,
+      "grad_norm": 1.3648931980133057,
+      "learning_rate": 1.94039323518842e-05,
+      "loss": 0.6572,
+      "step": 2500
+    },
+    {
+      "epoch": 0.41333031432041334,
+      "grad_norm": 1.3427473306655884,
+      "learning_rate": 1.9397871653058974e-05,
+      "loss": 0.6773,
+      "step": 2510
+    },
+    {
+      "epoch": 0.41497704864041496,
+      "grad_norm": 1.2733012437820435,
+      "learning_rate": 1.9391781254145833e-05,
+      "loss": 0.6822,
+      "step": 2520
+    },
+    {
+      "epoch": 0.41662378296041663,
+      "grad_norm": 1.1846665143966675,
+      "learning_rate": 1.9385661174392262e-05,
+      "loss": 0.6781,
+      "step": 2530
+    },
+    {
+      "epoch": 0.41827051728041825,
+      "grad_norm": 1.2302825450897217,
+      "learning_rate": 1.9379511433139547e-05,
+      "loss": 0.6608,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4199172516004199,
+      "grad_norm": 1.2838640213012695,
+      "learning_rate": 1.937333204982271e-05,
+      "loss": 0.6924,
+      "step": 2550
+    },
+    {
+      "epoch": 0.42156398592042155,
+      "grad_norm": 1.528942346572876,
+      "learning_rate": 1.9367123043970452e-05,
+      "loss": 0.6789,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4232107202404232,
+      "grad_norm": 1.4241206645965576,
+      "learning_rate": 1.936088443520509e-05,
+      "loss": 0.675,
+      "step": 2570
+    },
+    {
+      "epoch": 0.42485745456042484,
+      "grad_norm": 1.1903736591339111,
+      "learning_rate": 1.93546162432425e-05,
+      "loss": 0.6795,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4265041888804265,
+      "grad_norm": 1.1137593984603882,
+      "learning_rate": 1.9348318487892036e-05,
+      "loss": 0.6693,
+      "step": 2590
+    },
+    {
+      "epoch": 0.42815092320042814,
+      "grad_norm": 1.1012071371078491,
+      "learning_rate": 1.9341991189056498e-05,
+      "loss": 0.6982,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4297976575204298,
+      "grad_norm": 1.1505986452102661,
+      "learning_rate": 1.9335634366732044e-05,
+      "loss": 0.664,
+      "step": 2610
+    },
+    {
+      "epoch": 0.43144439184043143,
+      "grad_norm": 1.229068398475647,
+      "learning_rate": 1.9329248041008134e-05,
+      "loss": 0.6789,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4330911261604331,
+      "grad_norm": 1.1394855976104736,
+      "learning_rate": 1.9322832232067466e-05,
+      "loss": 0.6599,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4347378604804347,
+      "grad_norm": 1.1221281290054321,
+      "learning_rate": 1.9316386960185922e-05,
+      "loss": 0.6592,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4363845948004364,
+      "grad_norm": 1.1771143674850464,
+      "learning_rate": 1.930991224573249e-05,
+      "loss": 0.6922,
+      "step": 2650
+    },
+    {
+      "epoch": 0.438031329120438,
+      "grad_norm": 1.2491823434829712,
+      "learning_rate": 1.9303408109169205e-05,
+      "loss": 0.6544,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4396780634404397,
+      "grad_norm": 1.059578537940979,
+      "learning_rate": 1.9296874571051084e-05,
+      "loss": 0.6561,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4413247977604413,
+      "grad_norm": 1.4251331090927124,
+      "learning_rate": 1.9290311652026065e-05,
+      "loss": 0.6322,
+      "step": 2680
+    },
+    {
+      "epoch": 0.442971532080443,
+      "grad_norm": 1.9921714067459106,
+      "learning_rate": 1.9283719372834933e-05,
+      "loss": 0.6859,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4446182664004446,
+      "grad_norm": 1.2513914108276367,
+      "learning_rate": 1.9277097754311277e-05,
+      "loss": 0.6463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4462650007204463,
+      "grad_norm": 1.6695165634155273,
+      "learning_rate": 1.9270446817381377e-05,
+      "loss": 0.6827,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4479117350404479,
+      "grad_norm": 1.291625738143921,
+      "learning_rate": 1.9263766583064193e-05,
+      "loss": 0.6992,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4495584693604496,
+      "grad_norm": 1.7672340869903564,
+      "learning_rate": 1.925705707247127e-05,
+      "loss": 0.6754,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4512052036804512,
+      "grad_norm": 3.7783656120300293,
+      "learning_rate": 1.925031830680666e-05,
+      "loss": 0.671,
+      "step": 2740
+    },
+    {
+      "epoch": 0.45285193800045287,
+      "grad_norm": 1.448926568031311,
+      "learning_rate": 1.9243550307366884e-05,
+      "loss": 0.6677,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4544986723204545,
+      "grad_norm": 1.472121238708496,
+      "learning_rate": 1.923675309554085e-05,
+      "loss": 0.6609,
+      "step": 2760
+    },
+    {
+      "epoch": 0.45614540664045616,
+      "grad_norm": 1.7885098457336426,
+      "learning_rate": 1.9229926692809777e-05,
+      "loss": 0.6847,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4577921409604578,
+      "grad_norm": 1.6598893404006958,
+      "learning_rate": 1.9223071120747145e-05,
+      "loss": 0.6782,
+      "step": 2780
+    },
+    {
+      "epoch": 0.45943887528045946,
+      "grad_norm": 1.3006958961486816,
+      "learning_rate": 1.9216186401018614e-05,
+      "loss": 0.656,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4610856096004611,
+      "grad_norm": 1.0863409042358398,
+      "learning_rate": 1.920927255538196e-05,
+      "loss": 0.6482,
+      "step": 2800
+    },
+    {
+      "epoch": 0.46273234392046275,
+      "grad_norm": 1.5282680988311768,
+      "learning_rate": 1.9202329605687e-05,
+      "loss": 0.6434,
+      "step": 2810
+    },
+    {
+      "epoch": 0.46437907824046437,
+      "grad_norm": 1.2323793172836304,
+      "learning_rate": 1.9195357573875537e-05,
+      "loss": 0.6177,
+      "step": 2820
+    },
+    {
+      "epoch": 0.46602581256046605,
+      "grad_norm": 1.1313997507095337,
+      "learning_rate": 1.918835648198128e-05,
+      "loss": 0.674,
+      "step": 2830
+    },
+    {
+      "epoch": 0.46767254688046767,
+      "grad_norm": 1.3586218357086182,
+      "learning_rate": 1.9181326352129773e-05,
+      "loss": 0.6722,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46931928120046934,
+      "grad_norm": 1.0108249187469482,
+      "learning_rate": 1.9174267206538332e-05,
+      "loss": 0.6609,
+      "step": 2850
+    },
+    {
+      "epoch": 0.47096601552047096,
+      "grad_norm": 1.2038466930389404,
+      "learning_rate": 1.916717906751597e-05,
+      "loss": 0.6407,
+      "step": 2860
+    },
+    {
+      "epoch": 0.47261274984047263,
+      "grad_norm": 1.1672009229660034,
+      "learning_rate": 1.916006195746333e-05,
+      "loss": 0.6932,
+      "step": 2870
+    },
+    {
+      "epoch": 0.47425948416047425,
+      "grad_norm": 1.5974833965301514,
+      "learning_rate": 1.915291589887261e-05,
+      "loss": 0.6373,
+      "step": 2880
+    },
+    {
+      "epoch": 0.47590621848047593,
+      "grad_norm": 1.06233811378479,
+      "learning_rate": 1.914574091432749e-05,
+      "loss": 0.6429,
+      "step": 2890
+    },
+    {
+      "epoch": 0.47755295280047755,
+      "grad_norm": 1.8846768140792847,
+      "learning_rate": 1.9138537026503076e-05,
+      "loss": 0.6373,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4791996871204792,
+      "grad_norm": 1.164318323135376,
+      "learning_rate": 1.913130425816581e-05,
+      "loss": 0.6639,
+      "step": 2910
+    },
+    {
+      "epoch": 0.48084642144048084,
+      "grad_norm": 1.1987966299057007,
+      "learning_rate": 1.9124042632173398e-05,
+      "loss": 0.6699,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4824931557604825,
+      "grad_norm": 1.5233405828475952,
+      "learning_rate": 1.9116752171474754e-05,
+      "loss": 0.6355,
+      "step": 2930
+    },
+    {
+      "epoch": 0.48413989008048414,
+      "grad_norm": 1.2364847660064697,
+      "learning_rate": 1.9109432899109923e-05,
+      "loss": 0.6616,
+      "step": 2940
+    },
+    {
+      "epoch": 0.4857866244004858,
+      "grad_norm": 1.3125419616699219,
+      "learning_rate": 1.9102084838209992e-05,
+      "loss": 0.6585,
+      "step": 2950
+    },
+    {
+      "epoch": 0.48743335872048743,
+      "grad_norm": 1.315054178237915,
+      "learning_rate": 1.9094708011997033e-05,
+      "loss": 0.6443,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4890800930404891,
+      "grad_norm": 0.934471070766449,
+      "learning_rate": 1.908730244378403e-05,
+      "loss": 0.6209,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4907268273604907,
+      "grad_norm": 1.1312390565872192,
+      "learning_rate": 1.9079868156974788e-05,
+      "loss": 0.6699,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4923735616804924,
+      "grad_norm": 1.168817162513733,
+      "learning_rate": 1.9072405175063883e-05,
+      "loss": 0.6222,
+      "step": 2990
+    },
+    {
+      "epoch": 0.494020296000494,
+      "grad_norm": 1.1964330673217773,
+      "learning_rate": 1.9064913521636574e-05,
+      "loss": 0.6303,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4956670303204957,
+      "grad_norm": 1.1183303594589233,
+      "learning_rate": 1.9057393220368722e-05,
+      "loss": 0.6285,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4973137646404973,
+      "grad_norm": 1.0846532583236694,
+      "learning_rate": 1.9049844295026738e-05,
+      "loss": 0.6463,
+      "step": 3020
+    },
+    {
+      "epoch": 0.498960498960499,
+      "grad_norm": 1.4997867345809937,
+      "learning_rate": 1.904226676946748e-05,
+      "loss": 0.6729,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5006072332805006,
+      "grad_norm": 1.2239123582839966,
+      "learning_rate": 1.9034660667638206e-05,
+      "loss": 0.664,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5022539676005022,
+      "grad_norm": 1.2806061506271362,
+      "learning_rate": 1.9027026013576465e-05,
+      "loss": 0.6568,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5039007019205038,
+      "grad_norm": 1.2059847116470337,
+      "learning_rate": 1.9019362831410057e-05,
+      "loss": 0.6771,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5055474362405056,
+      "grad_norm": 1.2228035926818848,
+      "learning_rate": 1.9011671145356926e-05,
+      "loss": 0.6438,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5071941705605072,
+      "grad_norm": 1.16281259059906,
+      "learning_rate": 1.9003950979725103e-05,
+      "loss": 0.6518,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5088409048805088,
+      "grad_norm": 1.3429057598114014,
+      "learning_rate": 1.899620235891263e-05,
+      "loss": 0.6167,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5104876392005104,
+      "grad_norm": 1.0990668535232544,
+      "learning_rate": 1.8988425307407458e-05,
+      "loss": 0.624,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5121343735205122,
+      "grad_norm": 1.1892952919006348,
+      "learning_rate": 1.89806198497874e-05,
+      "loss": 0.6193,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5137811078405138,
+      "grad_norm": 1.194745421409607,
+      "learning_rate": 1.8972786010720046e-05,
+      "loss": 0.6825,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5154278421605154,
+      "grad_norm": 1.267279028892517,
+      "learning_rate": 1.8964923814962672e-05,
+      "loss": 0.6361,
+      "step": 3130
+    },
+    {
+      "epoch": 0.517074576480517,
+      "grad_norm": 1.3956196308135986,
+      "learning_rate": 1.8957033287362167e-05,
+      "loss": 0.6075,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5187213108005188,
+      "grad_norm": 1.4350258111953735,
+      "learning_rate": 1.8949114452854957e-05,
+      "loss": 0.6427,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5203680451205204,
+      "grad_norm": 1.3245211839675903,
+      "learning_rate": 1.8941167336466932e-05,
+      "loss": 0.6825,
+      "step": 3160
+    },
+    {
+      "epoch": 0.522014779440522,
+      "grad_norm": 1.3547745943069458,
+      "learning_rate": 1.893319196331336e-05,
+      "loss": 0.6353,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5236615137605236,
+      "grad_norm": 1.0884572267532349,
+      "learning_rate": 1.8925188358598815e-05,
+      "loss": 0.6354,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5253082480805253,
+      "grad_norm": 1.193581223487854,
+      "learning_rate": 1.8917156547617072e-05,
+      "loss": 0.651,
+      "step": 3190
+    },
+    {
+      "epoch": 0.526954982400527,
+      "grad_norm": 1.1126220226287842,
+      "learning_rate": 1.890909655575106e-05,
+      "loss": 0.6516,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5286017167205286,
+      "grad_norm": 1.4914835691452026,
+      "learning_rate": 1.8901008408472775e-05,
+      "loss": 0.6346,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5302484510405302,
+      "grad_norm": 1.3618090152740479,
+      "learning_rate": 1.8892892131343177e-05,
+      "loss": 0.6237,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5318951853605319,
+      "grad_norm": 1.3347357511520386,
+      "learning_rate": 1.888474775001213e-05,
+      "loss": 0.6727,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5335419196805335,
+      "grad_norm": 1.1067010164260864,
+      "learning_rate": 1.8876575290218323e-05,
+      "loss": 0.6533,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5351886540005352,
+      "grad_norm": 1.7199362516403198,
+      "learning_rate": 1.8868374777789172e-05,
+      "loss": 0.6546,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5368353883205368,
+      "grad_norm": 1.990593671798706,
+      "learning_rate": 1.886014623864075e-05,
+      "loss": 0.6129,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5384821226405385,
+      "grad_norm": 1.2225085496902466,
+      "learning_rate": 1.8851889698777707e-05,
+      "loss": 0.655,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5401288569605401,
+      "grad_norm": 1.4789817333221436,
+      "learning_rate": 1.8843605184293177e-05,
+      "loss": 0.6274,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5417755912805418,
+      "grad_norm": 1.4760066270828247,
+      "learning_rate": 1.8835292721368715e-05,
+      "loss": 0.649,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5434223256005434,
+      "grad_norm": 2.949131488800049,
+      "learning_rate": 1.8826952336274184e-05,
+      "loss": 0.6358,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5450690599205451,
+      "grad_norm": 1.5721391439437866,
+      "learning_rate": 1.88185840553677e-05,
+      "loss": 0.6461,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5467157942405467,
+      "grad_norm": 1.620802402496338,
+      "learning_rate": 1.881018790509553e-05,
+      "loss": 0.5999,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5483625285605483,
+      "grad_norm": 1.6917542219161987,
+      "learning_rate": 1.8801763911992035e-05,
+      "loss": 0.6141,
+      "step": 3330
+    },
+    {
+      "epoch": 0.55000926288055,
+      "grad_norm": 1.6579176187515259,
+      "learning_rate": 1.8793312102679548e-05,
+      "loss": 0.6865,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5516559972005517,
+      "grad_norm": 1.5372205972671509,
+      "learning_rate": 1.8784832503868314e-05,
+      "loss": 0.607,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5533027315205533,
+      "grad_norm": 2.1712288856506348,
+      "learning_rate": 1.8776325142356406e-05,
+      "loss": 0.6754,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5549494658405549,
+      "grad_norm": 1.6536438465118408,
+      "learning_rate": 1.876779004502964e-05,
+      "loss": 0.6484,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5565962001605566,
+      "grad_norm": 2.0324020385742188,
+      "learning_rate": 1.8759227238861467e-05,
+      "loss": 0.6278,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5582429344805583,
+      "grad_norm": 2.410161018371582,
+      "learning_rate": 1.8750636750912927e-05,
+      "loss": 0.611,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5598896688005599,
+      "grad_norm": 1.6419649124145508,
+      "learning_rate": 1.874201860833253e-05,
+      "loss": 0.643,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5615364031205615,
+      "grad_norm": 2.4913241863250732,
+      "learning_rate": 1.873337283835619e-05,
+      "loss": 0.6535,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5631831374405631,
+      "grad_norm": 1.8292442560195923,
+      "learning_rate": 1.8724699468307123e-05,
+      "loss": 0.643,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5648298717605649,
+      "grad_norm": 1.3961424827575684,
+      "learning_rate": 1.8715998525595775e-05,
+      "loss": 0.6279,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5664766060805665,
+      "grad_norm": 1.811924695968628,
+      "learning_rate": 1.8707270037719737e-05,
+      "loss": 0.6424,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5681233404005681,
+      "grad_norm": 1.5042741298675537,
+      "learning_rate": 1.8698514032263636e-05,
+      "loss": 0.6027,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5697700747205697,
+      "grad_norm": 1.6137949228286743,
+      "learning_rate": 1.8689730536899078e-05,
+      "loss": 0.6294,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5714168090405715,
+      "grad_norm": 1.3153676986694336,
+      "learning_rate": 1.868091957938453e-05,
+      "loss": 0.6277,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5730635433605731,
+      "grad_norm": 1.300405740737915,
+      "learning_rate": 1.867208118756526e-05,
+      "loss": 0.6461,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5747102776805747,
+      "grad_norm": 1.4557414054870605,
+      "learning_rate": 1.866321538937323e-05,
+      "loss": 0.6486,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5763570120005763,
+      "grad_norm": 1.8575087785720825,
+      "learning_rate": 1.8654322212827022e-05,
+      "loss": 0.6732,
+      "step": 3500
+    },
+    {
+      "epoch": 0.578003746320578,
+      "grad_norm": 1.3033815622329712,
+      "learning_rate": 1.864540168603173e-05,
+      "loss": 0.6361,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5796504806405797,
+      "grad_norm": 2.0723509788513184,
+      "learning_rate": 1.8636453837178893e-05,
+      "loss": 0.6148,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5812972149605813,
+      "grad_norm": 2.9872148036956787,
+      "learning_rate": 1.8627478694546387e-05,
+      "loss": 0.6768,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5829439492805829,
+      "grad_norm": 1.6916303634643555,
+      "learning_rate": 1.8618476286498356e-05,
+      "loss": 0.6318,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5845906836005846,
+      "grad_norm": 2.017186164855957,
+      "learning_rate": 1.86094466414851e-05,
+      "loss": 0.6583,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5862374179205863,
+      "grad_norm": 1.7270339727401733,
+      "learning_rate": 1.8600389788043003e-05,
+      "loss": 0.6026,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5878841522405879,
+      "grad_norm": 1.8550065755844116,
+      "learning_rate": 1.8591305754794434e-05,
+      "loss": 0.631,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5895308865605895,
+      "grad_norm": 2.131333112716675,
+      "learning_rate": 1.8582194570447654e-05,
+      "loss": 0.6287,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5911776208805912,
+      "grad_norm": 2.0691232681274414,
+      "learning_rate": 1.8573056263796732e-05,
+      "loss": 0.6032,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5928243552005928,
+      "grad_norm": 2.4947502613067627,
+      "learning_rate": 1.856389086372146e-05,
+      "loss": 0.6571,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5944710895205945,
+      "grad_norm": 1.4081095457077026,
+      "learning_rate": 1.8554698399187246e-05,
+      "loss": 0.649,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5961178238405961,
+      "grad_norm": 1.7815622091293335,
+      "learning_rate": 1.854547889924502e-05,
+      "loss": 0.6437,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5977645581605978,
+      "grad_norm": 1.900574803352356,
+      "learning_rate": 1.8536232393031173e-05,
+      "loss": 0.636,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5994112924805994,
+      "grad_norm": 1.3070859909057617,
+      "learning_rate": 1.8526958909767425e-05,
+      "loss": 0.6325,
+      "step": 3640
+    },
+    {
+      "epoch": 0.601058026800601,
+      "grad_norm": 1.3566079139709473,
+      "learning_rate": 1.851765847876076e-05,
+      "loss": 0.6235,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6027047611206027,
+      "grad_norm": 1.459912896156311,
+      "learning_rate": 1.8508331129403333e-05,
+      "loss": 0.6738,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6043514954406044,
+      "grad_norm": 1.7209458351135254,
+      "learning_rate": 1.849897689117235e-05,
+      "loss": 0.6478,
+      "step": 3670
+    },
+    {
+      "epoch": 0.605998229760606,
+      "grad_norm": 1.884052038192749,
+      "learning_rate": 1.848959579363001e-05,
+      "loss": 0.6695,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6076449640806076,
+      "grad_norm": 1.8747471570968628,
+      "learning_rate": 1.8480187866423386e-05,
+      "loss": 0.6453,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6092916984006093,
+      "grad_norm": 1.548047423362732,
+      "learning_rate": 1.8470753139284344e-05,
+      "loss": 0.6253,
+      "step": 3700
+    },
+    {
+      "epoch": 0.610938432720611,
+      "grad_norm": 3.6169090270996094,
+      "learning_rate": 1.8461291642029454e-05,
+      "loss": 0.6002,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6125851670406126,
+      "grad_norm": 1.727982997894287,
+      "learning_rate": 1.8451803404559873e-05,
+      "loss": 0.6077,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6142319013606142,
+      "grad_norm": 1.832689881324768,
+      "learning_rate": 1.844228845686127e-05,
+      "loss": 0.621,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6158786356806158,
+      "grad_norm": 1.83828866481781,
+      "learning_rate": 1.8432746829003732e-05,
+      "loss": 0.6464,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6175253700006176,
+      "grad_norm": 2.2948646545410156,
+      "learning_rate": 1.8423178551141662e-05,
+      "loss": 0.6234,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6191721043206192,
+      "grad_norm": 2.7820045948028564,
+      "learning_rate": 1.841358365351368e-05,
+      "loss": 0.6334,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6208188386406208,
+      "grad_norm": 1.8063865900039673,
+      "learning_rate": 1.8403962166442535e-05,
+      "loss": 0.6515,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6224655729606224,
+      "grad_norm": 3.7020092010498047,
+      "learning_rate": 1.8394314120335002e-05,
+      "loss": 0.6761,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6241123072806241,
+      "grad_norm": 3.1100969314575195,
+      "learning_rate": 1.8384639545681803e-05,
+      "loss": 0.5896,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6257590416006258,
+      "grad_norm": 3.011991262435913,
+      "learning_rate": 1.8374938473057486e-05,
+      "loss": 0.627,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6274057759206274,
+      "grad_norm": 1.5754694938659668,
+      "learning_rate": 1.8365210933120347e-05,
+      "loss": 0.6245,
+      "step": 3810
+    },
+    {
+      "epoch": 0.629052510240629,
+      "grad_norm": 1.8275468349456787,
+      "learning_rate": 1.835545695661232e-05,
+      "loss": 0.6179,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6306992445606306,
+      "grad_norm": 2.1361637115478516,
+      "learning_rate": 1.8345676574358897e-05,
+      "loss": 0.638,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6323459788806324,
+      "grad_norm": 1.829188585281372,
+      "learning_rate": 1.8335869817269006e-05,
+      "loss": 0.6304,
+      "step": 3840
+    },
+    {
+      "epoch": 0.633992713200634,
+      "grad_norm": 1.8671929836273193,
+      "learning_rate": 1.8326036716334942e-05,
+      "loss": 0.6322,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6356394475206356,
+      "grad_norm": 1.404921531677246,
+      "learning_rate": 1.8316177302632248e-05,
+      "loss": 0.6213,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6372861818406372,
+      "grad_norm": 1.64621901512146,
+      "learning_rate": 1.8306291607319618e-05,
+      "loss": 0.6668,
+      "step": 3870
+    },
+    {
+      "epoch": 0.638932916160639,
+      "grad_norm": 2.2080841064453125,
+      "learning_rate": 1.829637966163881e-05,
+      "loss": 0.6089,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6405796504806406,
+      "grad_norm": 1.4423542022705078,
+      "learning_rate": 1.8286441496914545e-05,
+      "loss": 0.6406,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6422263848006422,
+      "grad_norm": 1.7559462785720825,
+      "learning_rate": 1.8276477144554393e-05,
+      "loss": 0.6371,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6438731191206438,
+      "grad_norm": 1.5417958498001099,
+      "learning_rate": 1.826648663604869e-05,
+      "loss": 0.6262,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6455198534406456,
+      "grad_norm": 1.899674654006958,
+      "learning_rate": 1.8256470002970438e-05,
+      "loss": 0.6192,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6471665877606472,
+      "grad_norm": 1.715377688407898,
+      "learning_rate": 1.8246427276975196e-05,
+      "loss": 0.6189,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6488133220806488,
+      "grad_norm": 1.4496129751205444,
+      "learning_rate": 1.823635848980098e-05,
+      "loss": 0.6147,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6504600564006504,
+      "grad_norm": 1.5673155784606934,
+      "learning_rate": 1.822626367326818e-05,
+      "loss": 0.6159,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6521067907206521,
+      "grad_norm": 1.1710649728775024,
+      "learning_rate": 1.8216142859279432e-05,
+      "loss": 0.5911,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6537535250406538,
+      "grad_norm": 1.358017921447754,
+      "learning_rate": 1.8205996079819534e-05,
+      "loss": 0.6107,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6554002593606554,
+      "grad_norm": 1.4123727083206177,
+      "learning_rate": 1.8195823366955356e-05,
+      "loss": 0.629,
+      "step": 3980
+    },
+    {
+      "epoch": 0.657046993680657,
+      "grad_norm": 1.9289451837539673,
+      "learning_rate": 1.8185624752835714e-05,
+      "loss": 0.6609,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6586937280006587,
+      "grad_norm": 1.6234312057495117,
+      "learning_rate": 1.8175400269691278e-05,
+      "loss": 0.6191,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6603404623206603,
+      "grad_norm": 1.8857241868972778,
+      "learning_rate": 1.8165149949834474e-05,
+      "loss": 0.6466,
+      "step": 4010
+    },
+    {
+      "epoch": 0.661987196640662,
+      "grad_norm": 2.1216554641723633,
+      "learning_rate": 1.8154873825659393e-05,
+      "loss": 0.6257,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6636339309606636,
+      "grad_norm": 1.483231544494629,
+      "learning_rate": 1.814457192964165e-05,
+      "loss": 0.5745,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6652806652806653,
+      "grad_norm": 1.1727079153060913,
+      "learning_rate": 1.813424429433833e-05,
+      "loss": 0.6168,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6669273996006669,
+      "grad_norm": 1.454728364944458,
+      "learning_rate": 1.8123890952387848e-05,
+      "loss": 0.6006,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6685741339206686,
+      "grad_norm": 1.6542500257492065,
+      "learning_rate": 1.8113511936509864e-05,
+      "loss": 0.6296,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6702208682406702,
+      "grad_norm": 1.3419349193572998,
+      "learning_rate": 1.8103107279505177e-05,
+      "loss": 0.6597,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6718676025606719,
+      "grad_norm": 1.247767686843872,
+      "learning_rate": 1.809267701425562e-05,
+      "loss": 0.6231,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6735143368806735,
+      "grad_norm": 2.062878370285034,
+      "learning_rate": 1.808222117372395e-05,
+      "loss": 0.6577,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6751610712006751,
+      "grad_norm": 1.2169808149337769,
+      "learning_rate": 1.8071739790953754e-05,
+      "loss": 0.6052,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6768078055206768,
+      "grad_norm": 1.3929376602172852,
+      "learning_rate": 1.806123289906934e-05,
+      "loss": 0.5777,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6784545398406785,
+      "grad_norm": 1.6458179950714111,
+      "learning_rate": 1.8050700531275632e-05,
+      "loss": 0.6085,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6801012741606801,
+      "grad_norm": 1.402419090270996,
+      "learning_rate": 1.8040142720858064e-05,
+      "loss": 0.6112,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6817480084806817,
+      "grad_norm": 1.2374345064163208,
+      "learning_rate": 1.8029559501182482e-05,
+      "loss": 0.5976,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6833947428006834,
+      "grad_norm": 1.6818028688430786,
+      "learning_rate": 1.8018950905695022e-05,
+      "loss": 0.6461,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6850414771206851,
+      "grad_norm": 1.5200414657592773,
+      "learning_rate": 1.8008316967922027e-05,
+      "loss": 0.6412,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6866882114406867,
+      "grad_norm": 1.8898037672042847,
+      "learning_rate": 1.799765772146992e-05,
+      "loss": 0.6135,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6883349457606883,
+      "grad_norm": 2.4506630897521973,
+      "learning_rate": 1.7986973200025115e-05,
+      "loss": 0.6181,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6899816800806899,
+      "grad_norm": 1.4668997526168823,
+      "learning_rate": 1.7976263437353897e-05,
+      "loss": 0.5981,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6916284144006917,
+      "grad_norm": 1.4976595640182495,
+      "learning_rate": 1.796552846730232e-05,
+      "loss": 0.5983,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6932751487206933,
+      "grad_norm": 1.2524445056915283,
+      "learning_rate": 1.7954768323796107e-05,
+      "loss": 0.6191,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6949218830406949,
+      "grad_norm": 1.3345338106155396,
+      "learning_rate": 1.7943983040840527e-05,
+      "loss": 0.6019,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6965686173606965,
+      "grad_norm": 1.9218558073043823,
+      "learning_rate": 1.7933172652520308e-05,
+      "loss": 0.6084,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6982153516806983,
+      "grad_norm": 1.2218800783157349,
+      "learning_rate": 1.7922337192999514e-05,
+      "loss": 0.6521,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6998620860006999,
+      "grad_norm": 1.9215497970581055,
+      "learning_rate": 1.7911476696521437e-05,
+      "loss": 0.6229,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7015088203207015,
+      "grad_norm": 1.0203500986099243,
+      "learning_rate": 1.79005911974085e-05,
+      "loss": 0.5724,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7031555546407031,
+      "grad_norm": 1.290940761566162,
+      "learning_rate": 1.7889680730062137e-05,
+      "loss": 0.5783,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7048022889607048,
+      "grad_norm": 1.156741976737976,
+      "learning_rate": 1.7878745328962696e-05,
+      "loss": 0.6373,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7064490232807065,
+      "grad_norm": 1.3677122592926025,
+      "learning_rate": 1.7867785028669308e-05,
+      "loss": 0.6251,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7080957576007081,
+      "grad_norm": 1.15171217918396,
+      "learning_rate": 1.7856799863819814e-05,
+      "loss": 0.6389,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7097424919207097,
+      "grad_norm": 1.243003249168396,
+      "learning_rate": 1.784578986913062e-05,
+      "loss": 0.6155,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7113892262407114,
+      "grad_norm": 1.0966947078704834,
+      "learning_rate": 1.7834755079396604e-05,
+      "loss": 0.5783,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7130359605607131,
+      "grad_norm": 1.2809836864471436,
+      "learning_rate": 1.782369552949101e-05,
+      "loss": 0.6119,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7146826948807147,
+      "grad_norm": 1.148653507232666,
+      "learning_rate": 1.781261125436532e-05,
+      "loss": 0.6417,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7163294292007163,
+      "grad_norm": 1.11336088180542,
+      "learning_rate": 1.780150228904916e-05,
+      "loss": 0.6154,
+      "step": 4350
+    },
+    {
+      "epoch": 0.717976163520718,
+      "grad_norm": 1.1516305208206177,
+      "learning_rate": 1.779036866865019e-05,
+      "loss": 0.601,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7196228978407196,
+      "grad_norm": 1.2385212182998657,
+      "learning_rate": 1.777921042835399e-05,
+      "loss": 0.6114,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7212696321607213,
+      "grad_norm": 1.259101152420044,
+      "learning_rate": 1.776802760342393e-05,
+      "loss": 0.5704,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7229163664807229,
+      "grad_norm": 1.3446874618530273,
+      "learning_rate": 1.7756820229201092e-05,
+      "loss": 0.5561,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7245631008007246,
+      "grad_norm": 1.4307595491409302,
+      "learning_rate": 1.7745588341104127e-05,
+      "loss": 0.5923,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7262098351207262,
+      "grad_norm": 1.4261305332183838,
+      "learning_rate": 1.7734331974629166e-05,
+      "loss": 0.6218,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7278565694407279,
+      "grad_norm": 1.342528223991394,
+      "learning_rate": 1.77230511653497e-05,
+      "loss": 0.6146,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7295033037607295,
+      "grad_norm": 1.4735767841339111,
+      "learning_rate": 1.7711745948916464e-05,
+      "loss": 0.6167,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7311500380807312,
+      "grad_norm": 1.2360775470733643,
+      "learning_rate": 1.7700416361057322e-05,
+      "loss": 0.58,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7327967724007328,
+      "grad_norm": 1.2893751859664917,
+      "learning_rate": 1.7689062437577165e-05,
+      "loss": 0.5934,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7344435067207344,
+      "grad_norm": 1.6743543148040771,
+      "learning_rate": 1.7677684214357793e-05,
+      "loss": 0.5796,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7360902410407361,
+      "grad_norm": 1.218066692352295,
+      "learning_rate": 1.7666281727357792e-05,
+      "loss": 0.5908,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7377369753607378,
+      "grad_norm": 1.4119192361831665,
+      "learning_rate": 1.7654855012612442e-05,
+      "loss": 0.6163,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7393837096807394,
+      "grad_norm": 1.4504175186157227,
+      "learning_rate": 1.7643404106233573e-05,
+      "loss": 0.6091,
+      "step": 4490
+    },
+    {
+      "epoch": 0.741030444000741,
+      "grad_norm": 1.2811555862426758,
+      "learning_rate": 1.763192904440949e-05,
+      "loss": 0.5846,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7426771783207426,
+      "grad_norm": 1.9486907720565796,
+      "learning_rate": 1.762042986340481e-05,
+      "loss": 0.6064,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7443239126407444,
+      "grad_norm": 1.731048822402954,
+      "learning_rate": 1.76089065995604e-05,
+      "loss": 0.6002,
+      "step": 4520
+    },
+    {
+      "epoch": 0.745970646960746,
+      "grad_norm": 1.3845677375793457,
+      "learning_rate": 1.7597359289293213e-05,
+      "loss": 0.5834,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7476173812807476,
+      "grad_norm": 2.7521491050720215,
+      "learning_rate": 1.758578796909621e-05,
+      "loss": 0.6001,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7492641156007492,
+      "grad_norm": 2.141174554824829,
+      "learning_rate": 1.7574192675538232e-05,
+      "loss": 0.5961,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7509108499207509,
+      "grad_norm": 2.0954525470733643,
+      "learning_rate": 1.756257344526387e-05,
+      "loss": 0.5862,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7525575842407526,
+      "grad_norm": 1.395780324935913,
+      "learning_rate": 1.755093031499338e-05,
+      "loss": 0.5753,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7542043185607542,
+      "grad_norm": 1.6621346473693848,
+      "learning_rate": 1.7539263321522528e-05,
+      "loss": 0.608,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7558510528807558,
+      "grad_norm": 1.2174135446548462,
+      "learning_rate": 1.7527572501722516e-05,
+      "loss": 0.5927,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7574977872007574,
+      "grad_norm": 2.0868325233459473,
+      "learning_rate": 1.7515857892539828e-05,
+      "loss": 0.5783,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7591445215207592,
+      "grad_norm": 1.0850328207015991,
+      "learning_rate": 1.7504119530996138e-05,
+      "loss": 0.6089,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7607912558407608,
+      "grad_norm": 0.910921573638916,
+      "learning_rate": 1.749235745418818e-05,
+      "loss": 0.5967,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7624379901607624,
+      "grad_norm": 1.7855496406555176,
+      "learning_rate": 1.7480571699287647e-05,
+      "loss": 0.5669,
+      "step": 4630
+    },
+    {
+      "epoch": 0.764084724480764,
+      "grad_norm": 1.3511872291564941,
+      "learning_rate": 1.7468762303541044e-05,
+      "loss": 0.6084,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7657314588007658,
+      "grad_norm": 1.2458902597427368,
+      "learning_rate": 1.7456929304269598e-05,
+      "loss": 0.5668,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7673781931207674,
+      "grad_norm": 1.1821246147155762,
+      "learning_rate": 1.7445072738869134e-05,
+      "loss": 0.6231,
+      "step": 4660
+    },
+    {
+      "epoch": 0.769024927440769,
+      "grad_norm": 1.6277104616165161,
+      "learning_rate": 1.7433192644809942e-05,
+      "loss": 0.6051,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7706716617607706,
+      "grad_norm": 1.184881329536438,
+      "learning_rate": 1.742128905963668e-05,
+      "loss": 0.6102,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7723183960807724,
+      "grad_norm": 1.4022691249847412,
+      "learning_rate": 1.7409362020968242e-05,
+      "loss": 0.6079,
+      "step": 4690
+    },
+    {
+      "epoch": 0.773965130400774,
+      "grad_norm": 1.271551251411438,
+      "learning_rate": 1.7397411566497638e-05,
+      "loss": 0.6041,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7756118647207756,
+      "grad_norm": 1.211241364479065,
+      "learning_rate": 1.738543773399188e-05,
+      "loss": 0.5973,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7772585990407772,
+      "grad_norm": 1.336120367050171,
+      "learning_rate": 1.737344056129187e-05,
+      "loss": 0.6225,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7789053333607789,
+      "grad_norm": 1.2912676334381104,
+      "learning_rate": 1.7361420086312255e-05,
+      "loss": 0.5732,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7805520676807806,
+      "grad_norm": 1.1799460649490356,
+      "learning_rate": 1.7349376347041346e-05,
+      "loss": 0.5675,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7821988020007822,
+      "grad_norm": 1.8032108545303345,
+      "learning_rate": 1.7337309381540955e-05,
+      "loss": 0.5871,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7838455363207838,
+      "grad_norm": 1.2031786441802979,
+      "learning_rate": 1.7325219227946314e-05,
+      "loss": 0.5934,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7854922706407855,
+      "grad_norm": 1.412174105644226,
+      "learning_rate": 1.7313105924465923e-05,
+      "loss": 0.5923,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7871390049607871,
+      "grad_norm": 1.4549241065979004,
+      "learning_rate": 1.7300969509381448e-05,
+      "loss": 0.5747,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7887857392807888,
+      "grad_norm": 1.831076979637146,
+      "learning_rate": 1.7288810021047597e-05,
+      "loss": 0.6102,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7904324736007904,
+      "grad_norm": 1.530872106552124,
+      "learning_rate": 1.7276627497891984e-05,
+      "loss": 0.5789,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7920792079207921,
+      "grad_norm": 1.1748154163360596,
+      "learning_rate": 1.726442197841504e-05,
+      "loss": 0.5854,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7937259422407937,
+      "grad_norm": 1.1575807332992554,
+      "learning_rate": 1.7252193501189857e-05,
+      "loss": 0.5712,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7953726765607954,
+      "grad_norm": 1.2026344537734985,
+      "learning_rate": 1.723994210486208e-05,
+      "loss": 0.5471,
+      "step": 4830
+    },
+    {
+      "epoch": 0.797019410880797,
+      "grad_norm": 1.4718669652938843,
+      "learning_rate": 1.7227667828149795e-05,
+      "loss": 0.6096,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7986661452007987,
+      "grad_norm": 1.6093403100967407,
+      "learning_rate": 1.721537070984339e-05,
+      "loss": 0.5773,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8003128795208003,
+      "grad_norm": 1.2670097351074219,
+      "learning_rate": 1.720305078880544e-05,
+      "loss": 0.5763,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8019596138408019,
+      "grad_norm": 1.2672990560531616,
+      "learning_rate": 1.719070810397058e-05,
+      "loss": 0.5859,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8036063481608036,
+      "grad_norm": 1.4210219383239746,
+      "learning_rate": 1.7178342694345395e-05,
+      "loss": 0.5781,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8052530824808053,
+      "grad_norm": 1.1239454746246338,
+      "learning_rate": 1.7165954599008275e-05,
+      "loss": 0.5778,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8068998168008069,
+      "grad_norm": 1.0624126195907593,
+      "learning_rate": 1.7153543857109314e-05,
+      "loss": 0.5804,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8085465511208085,
+      "grad_norm": 1.0993456840515137,
+      "learning_rate": 1.7141110507870172e-05,
+      "loss": 0.5778,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8101932854408102,
+      "grad_norm": 1.2814329862594604,
+      "learning_rate": 1.7128654590583953e-05,
+      "loss": 0.5687,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8118400197608119,
+      "grad_norm": 1.2554537057876587,
+      "learning_rate": 1.7116176144615085e-05,
+      "loss": 0.5967,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8134867540808135,
+      "grad_norm": 1.1652063131332397,
+      "learning_rate": 1.7103675209399194e-05,
+      "loss": 0.5907,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8151334884008151,
+      "grad_norm": 1.1235425472259521,
+      "learning_rate": 1.7091151824442978e-05,
+      "loss": 0.6217,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8167802227208167,
+      "grad_norm": 1.1538571119308472,
+      "learning_rate": 1.707860602932408e-05,
+      "loss": 0.5807,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8184269570408185,
+      "grad_norm": 1.3833861351013184,
+      "learning_rate": 1.7066037863690975e-05,
+      "loss": 0.5916,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8200736913608201,
+      "grad_norm": 1.847043752670288,
+      "learning_rate": 1.7053447367262817e-05,
+      "loss": 0.5662,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8217204256808217,
+      "grad_norm": 1.4586071968078613,
+      "learning_rate": 1.7040834579829358e-05,
+      "loss": 0.5624,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8233671600008233,
+      "grad_norm": 1.2225836515426636,
+      "learning_rate": 1.702819954125077e-05,
+      "loss": 0.5864,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8250138943208251,
+      "grad_norm": 1.430363655090332,
+      "learning_rate": 1.7015542291457567e-05,
+      "loss": 0.5816,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8266606286408267,
+      "grad_norm": 1.2088559865951538,
+      "learning_rate": 1.7002862870450446e-05,
+      "loss": 0.5924,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8283073629608283,
+      "grad_norm": 1.0974377393722534,
+      "learning_rate": 1.6990161318300167e-05,
+      "loss": 0.556,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8299540972808299,
+      "grad_norm": 1.475848913192749,
+      "learning_rate": 1.697743767514744e-05,
+      "loss": 0.5711,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8316008316008316,
+      "grad_norm": 1.3842908143997192,
+      "learning_rate": 1.696469198120279e-05,
+      "loss": 0.5931,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8332475659208333,
+      "grad_norm": 1.4212431907653809,
+      "learning_rate": 1.6951924276746425e-05,
+      "loss": 0.5905,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8348943002408349,
+      "grad_norm": 1.4288952350616455,
+      "learning_rate": 1.693913460212811e-05,
+      "loss": 0.5891,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8365410345608365,
+      "grad_norm": 1.2996931076049805,
+      "learning_rate": 1.6926322997767045e-05,
+      "loss": 0.5876,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8381877688808382,
+      "grad_norm": 1.6933460235595703,
+      "learning_rate": 1.6913489504151743e-05,
+      "loss": 0.5642,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8398345032008399,
+      "grad_norm": 1.210008144378662,
+      "learning_rate": 1.690063416183988e-05,
+      "loss": 0.5543,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8414812375208415,
+      "grad_norm": 1.4546929597854614,
+      "learning_rate": 1.6887757011458184e-05,
+      "loss": 0.5843,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8431279718408431,
+      "grad_norm": 1.6936150789260864,
+      "learning_rate": 1.687485809370231e-05,
+      "loss": 0.5882,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8447747061608448,
+      "grad_norm": 1.3216242790222168,
+      "learning_rate": 1.6861937449336697e-05,
+      "loss": 0.5567,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8464214404808464,
+      "grad_norm": 1.691382884979248,
+      "learning_rate": 1.6848995119194453e-05,
+      "loss": 0.559,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8480681748008481,
+      "grad_norm": 1.0977565050125122,
+      "learning_rate": 1.683603114417721e-05,
+      "loss": 0.5877,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8497149091208497,
+      "grad_norm": 1.1555675268173218,
+      "learning_rate": 1.6823045565255018e-05,
+      "loss": 0.5665,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8513616434408514,
+      "grad_norm": 1.320023775100708,
+      "learning_rate": 1.6810038423466184e-05,
+      "loss": 0.5939,
+      "step": 5170
+    },
+    {
+      "epoch": 0.853008377760853,
+      "grad_norm": 1.6562938690185547,
+      "learning_rate": 1.6797009759917177e-05,
+      "loss": 0.5441,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8546551120808547,
+      "grad_norm": 1.2160990238189697,
+      "learning_rate": 1.678395961578247e-05,
+      "loss": 0.5647,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8563018464008563,
+      "grad_norm": 1.3416587114334106,
+      "learning_rate": 1.6770888032304437e-05,
+      "loss": 0.5696,
+      "step": 5200
+    },
+    {
+      "epoch": 0.857948580720858,
+      "grad_norm": 1.0583994388580322,
+      "learning_rate": 1.6757795050793175e-05,
+      "loss": 0.5509,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8595953150408596,
+      "grad_norm": 1.2085150480270386,
+      "learning_rate": 1.674468071262644e-05,
+      "loss": 0.5405,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8612420493608612,
+      "grad_norm": 1.6761043071746826,
+      "learning_rate": 1.6731545059249467e-05,
+      "loss": 0.5666,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8628887836808629,
+      "grad_norm": 1.3844279050827026,
+      "learning_rate": 1.671838813217485e-05,
+      "loss": 0.5458,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8645355180008646,
+      "grad_norm": 2.4434731006622314,
+      "learning_rate": 1.670520997298241e-05,
+      "loss": 0.5734,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8661822523208662,
+      "grad_norm": 1.2524642944335938,
+      "learning_rate": 1.6692010623319087e-05,
+      "loss": 0.5698,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8678289866408678,
+      "grad_norm": 1.8466917276382446,
+      "learning_rate": 1.667879012489877e-05,
+      "loss": 0.5698,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8694757209608694,
+      "grad_norm": 1.3722879886627197,
+      "learning_rate": 1.6665548519502196e-05,
+      "loss": 0.5402,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8711224552808712,
+      "grad_norm": 1.4703186750411987,
+      "learning_rate": 1.66522858489768e-05,
+      "loss": 0.5886,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8727691896008728,
+      "grad_norm": 1.5705962181091309,
+      "learning_rate": 1.663900215523659e-05,
+      "loss": 0.6006,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8744159239208744,
+      "grad_norm": 1.7487270832061768,
+      "learning_rate": 1.662569748026202e-05,
+      "loss": 0.5421,
+      "step": 5310
+    },
+    {
+      "epoch": 0.876062658240876,
+      "grad_norm": 1.986843228340149,
+      "learning_rate": 1.661237186609984e-05,
+      "loss": 0.5275,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8777093925608777,
+      "grad_norm": 1.661085844039917,
+      "learning_rate": 1.6599025354862983e-05,
+      "loss": 0.5491,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8793561268808794,
+      "grad_norm": 1.975029706954956,
+      "learning_rate": 1.6585657988730424e-05,
+      "loss": 0.5743,
+      "step": 5340
+    },
+    {
+      "epoch": 0.881002861200881,
+      "grad_norm": 1.8658064603805542,
+      "learning_rate": 1.6572269809947035e-05,
+      "loss": 0.5627,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8826495955208826,
+      "grad_norm": 1.540292739868164,
+      "learning_rate": 1.6558860860823473e-05,
+      "loss": 0.5227,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8842963298408842,
+      "grad_norm": 1.792221188545227,
+      "learning_rate": 1.654543118373603e-05,
+      "loss": 0.5766,
+      "step": 5370
+    },
+    {
+      "epoch": 0.885943064160886,
+      "grad_norm": 1.3535420894622803,
+      "learning_rate": 1.6531980821126508e-05,
+      "loss": 0.5803,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8875897984808876,
+      "grad_norm": 1.7461328506469727,
+      "learning_rate": 1.651850981550208e-05,
+      "loss": 0.5511,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8892365328008892,
+      "grad_norm": 1.6010706424713135,
+      "learning_rate": 1.6505018209435152e-05,
+      "loss": 0.5656,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8908832671208908,
+      "grad_norm": 1.3865113258361816,
+      "learning_rate": 1.649150604556324e-05,
+      "loss": 0.57,
+      "step": 5410
+    },
+    {
+      "epoch": 0.8925300014408926,
+      "grad_norm": 1.609237790107727,
+      "learning_rate": 1.6477973366588833e-05,
+      "loss": 0.5566,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8941767357608942,
+      "grad_norm": 2.0316720008850098,
+      "learning_rate": 1.6464420215279237e-05,
+      "loss": 0.5601,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8958234700808958,
+      "grad_norm": 1.759619951248169,
+      "learning_rate": 1.6450846634466476e-05,
+      "loss": 0.5665,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8974702044008974,
+      "grad_norm": 1.5108795166015625,
+      "learning_rate": 1.643725266704713e-05,
+      "loss": 0.5432,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8991169387208992,
+      "grad_norm": 1.3208014965057373,
+      "learning_rate": 1.6423638355982202e-05,
+      "loss": 0.5867,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9007636730409008,
+      "grad_norm": 1.530247688293457,
+      "learning_rate": 1.6410003744296984e-05,
+      "loss": 0.5721,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9024104073609024,
+      "grad_norm": 2.346730947494507,
+      "learning_rate": 1.6396348875080945e-05,
+      "loss": 0.5569,
+      "step": 5480
+    },
+    {
+      "epoch": 0.904057141680904,
+      "grad_norm": 1.4136897325515747,
+      "learning_rate": 1.638267379148755e-05,
+      "loss": 0.5876,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9057038760009057,
+      "grad_norm": 1.5225974321365356,
+      "learning_rate": 1.6368978536734162e-05,
+      "loss": 0.57,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9073506103209074,
+      "grad_norm": 1.5766575336456299,
+      "learning_rate": 1.6355263154101884e-05,
+      "loss": 0.5282,
+      "step": 5510
+    },
+    {
+      "epoch": 0.908997344640909,
+      "grad_norm": 1.784691333770752,
+      "learning_rate": 1.634152768693543e-05,
+      "loss": 0.5805,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9106440789609106,
+      "grad_norm": 1.479247808456421,
+      "learning_rate": 1.6327772178642986e-05,
+      "loss": 0.5525,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9122908132809123,
+      "grad_norm": 1.4190905094146729,
+      "learning_rate": 1.6313996672696083e-05,
+      "loss": 0.5598,
+      "step": 5540
+    },
+    {
+      "epoch": 0.913937547600914,
+      "grad_norm": 1.5564013719558716,
+      "learning_rate": 1.6300201212629437e-05,
+      "loss": 0.5521,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9155842819209156,
+      "grad_norm": 1.3574228286743164,
+      "learning_rate": 1.6286385842040843e-05,
+      "loss": 0.55,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9172310162409172,
+      "grad_norm": 1.4063721895217896,
+      "learning_rate": 1.6272550604590993e-05,
+      "loss": 0.5731,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9188777505609189,
+      "grad_norm": 1.4861043691635132,
+      "learning_rate": 1.625869554400339e-05,
+      "loss": 0.5248,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9205244848809205,
+      "grad_norm": 1.3798218965530396,
+      "learning_rate": 1.624482070406417e-05,
+      "loss": 0.5626,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9221712192009222,
+      "grad_norm": 1.3654139041900635,
+      "learning_rate": 1.623092612862198e-05,
+      "loss": 0.5273,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9238179535209238,
+      "grad_norm": 1.593929648399353,
+      "learning_rate": 1.6217011861587834e-05,
+      "loss": 0.5609,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9254646878409255,
+      "grad_norm": 1.439780592918396,
+      "learning_rate": 1.6203077946934992e-05,
+      "loss": 0.5699,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9271114221609271,
+      "grad_norm": 2.25289249420166,
+      "learning_rate": 1.6189124428698787e-05,
+      "loss": 0.5472,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9287581564809287,
+      "grad_norm": 1.3383618593215942,
+      "learning_rate": 1.6175151350976518e-05,
+      "loss": 0.5551,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9304048908009304,
+      "grad_norm": 1.414027452468872,
+      "learning_rate": 1.6161158757927292e-05,
+      "loss": 0.5793,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9320516251209321,
+      "grad_norm": 1.126561164855957,
+      "learning_rate": 1.6147146693771896e-05,
+      "loss": 0.5704,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9336983594409337,
+      "grad_norm": 1.2248793840408325,
+      "learning_rate": 1.6133115202792645e-05,
+      "loss": 0.5614,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9353450937609353,
+      "grad_norm": 1.4462449550628662,
+      "learning_rate": 1.6119064329333248e-05,
+      "loss": 0.5544,
+      "step": 5680
+    },
+    {
+      "epoch": 0.936991828080937,
+      "grad_norm": 1.1664077043533325,
+      "learning_rate": 1.6104994117798674e-05,
+      "loss": 0.5596,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9386385624009387,
+      "grad_norm": 1.239491581916809,
+      "learning_rate": 1.6090904612655007e-05,
+      "loss": 0.5636,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9402852967209403,
+      "grad_norm": 1.3634086847305298,
+      "learning_rate": 1.6076795858429296e-05,
+      "loss": 0.558,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9419320310409419,
+      "grad_norm": 1.319319725036621,
+      "learning_rate": 1.606266789970943e-05,
+      "loss": 0.5563,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9435787653609435,
+      "grad_norm": 1.2931233644485474,
+      "learning_rate": 1.6048520781143988e-05,
+      "loss": 0.5646,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9452254996809453,
+      "grad_norm": 1.3727686405181885,
+      "learning_rate": 1.6034354547442104e-05,
+      "loss": 0.5288,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9468722340009469,
+      "grad_norm": 1.4657319784164429,
+      "learning_rate": 1.6020169243373313e-05,
+      "loss": 0.5605,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9485189683209485,
+      "grad_norm": 1.3323101997375488,
+      "learning_rate": 1.600596491376742e-05,
+      "loss": 0.5575,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9501657026409501,
+      "grad_norm": 1.4169237613677979,
+      "learning_rate": 1.5991741603514367e-05,
+      "loss": 0.5448,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9518124369609519,
+      "grad_norm": 2.1234960556030273,
+      "learning_rate": 1.5977499357564067e-05,
+      "loss": 0.5453,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9534591712809535,
+      "grad_norm": 1.3298665285110474,
+      "learning_rate": 1.596323822092628e-05,
+      "loss": 0.5205,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9551059056009551,
+      "grad_norm": 1.7144644260406494,
+      "learning_rate": 1.594895823867047e-05,
+      "loss": 0.5333,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9567526399209567,
+      "grad_norm": 1.2539159059524536,
+      "learning_rate": 1.5934659455925658e-05,
+      "loss": 0.552,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9583993742409584,
+      "grad_norm": 1.7403203248977661,
+      "learning_rate": 1.5920341917880277e-05,
+      "loss": 0.5166,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9600461085609601,
+      "grad_norm": 1.802240252494812,
+      "learning_rate": 1.5906005669782027e-05,
+      "loss": 0.5392,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9616928428809617,
+      "grad_norm": 1.3411508798599243,
+      "learning_rate": 1.5891650756937755e-05,
+      "loss": 0.5621,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9633395772009633,
+      "grad_norm": 1.388187289237976,
+      "learning_rate": 1.587727722471327e-05,
+      "loss": 0.5331,
+      "step": 5850
+    },
+    {
+      "epoch": 0.964986311520965,
+      "grad_norm": 1.2933506965637207,
+      "learning_rate": 1.5862885118533244e-05,
+      "loss": 0.5449,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9666330458409667,
+      "grad_norm": 1.637378454208374,
+      "learning_rate": 1.5848474483881044e-05,
+      "loss": 0.5586,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9682797801609683,
+      "grad_norm": 1.2650068998336792,
+      "learning_rate": 1.5834045366298593e-05,
+      "loss": 0.5504,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9699265144809699,
+      "grad_norm": 1.3587515354156494,
+      "learning_rate": 1.5819597811386208e-05,
+      "loss": 0.5494,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9715732488009716,
+      "grad_norm": 1.1852201223373413,
+      "learning_rate": 1.5805131864802496e-05,
+      "loss": 0.5639,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9732199831209732,
+      "grad_norm": 1.3955892324447632,
+      "learning_rate": 1.579064757226418e-05,
+      "loss": 0.5651,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9748667174409749,
+      "grad_norm": 1.284073829650879,
+      "learning_rate": 1.5776144979545963e-05,
+      "loss": 0.5304,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9765134517609765,
+      "grad_norm": 1.7228704690933228,
+      "learning_rate": 1.5761624132480372e-05,
+      "loss": 0.5762,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9781601860809782,
+      "grad_norm": 1.9546711444854736,
+      "learning_rate": 1.5747085076957635e-05,
+      "loss": 0.5715,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9798069204009798,
+      "grad_norm": 1.2742135524749756,
+      "learning_rate": 1.5732527858925523e-05,
+      "loss": 0.5501,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9814536547209814,
+      "grad_norm": 1.293632984161377,
+      "learning_rate": 1.57179525243892e-05,
+      "loss": 0.5593,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9831003890409831,
+      "grad_norm": 1.244485855102539,
+      "learning_rate": 1.5703359119411087e-05,
+      "loss": 0.5229,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9847471233609848,
+      "grad_norm": 1.1649023294448853,
+      "learning_rate": 1.5688747690110708e-05,
+      "loss": 0.5411,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9863938576809864,
+      "grad_norm": 1.1061643362045288,
+      "learning_rate": 1.5674118282664563e-05,
+      "loss": 0.5528,
+      "step": 5990
+    },
+    {
+      "epoch": 0.988040592000988,
+      "grad_norm": 1.6961778402328491,
+      "learning_rate": 1.5659470943305956e-05,
+      "loss": 0.5484,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9896873263209897,
+      "grad_norm": 1.4628689289093018,
+      "learning_rate": 1.5644805718324854e-05,
+      "loss": 0.5446,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9913340606409914,
+      "grad_norm": 1.3264354467391968,
+      "learning_rate": 1.5630122654067778e-05,
+      "loss": 0.5684,
+      "step": 6020
+    },
+    {
+      "epoch": 0.992980794960993,
+      "grad_norm": 1.5602631568908691,
+      "learning_rate": 1.5615421796937593e-05,
+      "loss": 0.5514,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9946275292809946,
+      "grad_norm": 2.169344663619995,
+      "learning_rate": 1.560070319339341e-05,
+      "loss": 0.5681,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9962742636009962,
+      "grad_norm": 1.190785527229309,
+      "learning_rate": 1.5585966889950423e-05,
+      "loss": 0.5266,
+      "step": 6050
+    },
+    {
+      "epoch": 0.997920997920998,
+      "grad_norm": 1.3038309812545776,
+      "learning_rate": 1.5571212933179766e-05,
+      "loss": 0.5326,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9995677322409996,
+      "grad_norm": 1.3445013761520386,
+      "learning_rate": 1.5556441369708358e-05,
+      "loss": 0.5473,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0013173874560013,
+      "grad_norm": 1.4349257946014404,
+      "learning_rate": 1.554165224621876e-05,
+      "loss": 0.5755,
+      "step": 6080
+    },
+    {
+      "epoch": 1.002964121776003,
+      "grad_norm": 1.7247047424316406,
+      "learning_rate": 1.552684560944903e-05,
+      "loss": 0.5148,
+      "step": 6090
+    },
+    {
+      "epoch": 1.0046108560960045,
+      "grad_norm": 1.518079161643982,
+      "learning_rate": 1.5512021506192575e-05,
+      "loss": 0.5156,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0062575904160063,
+      "grad_norm": 1.1310617923736572,
+      "learning_rate": 1.5497179983297998e-05,
+      "loss": 0.4851,
+      "step": 6110
+    },
+    {
+      "epoch": 1.007904324736008,
+      "grad_norm": 1.521835207939148,
+      "learning_rate": 1.5482321087668954e-05,
+      "loss": 0.507,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0095510590560095,
+      "grad_norm": 1.224595308303833,
+      "learning_rate": 1.5467444866264e-05,
+      "loss": 0.5086,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0111977933760112,
+      "grad_norm": 1.4510592222213745,
+      "learning_rate": 1.5452551366096457e-05,
+      "loss": 0.4839,
+      "step": 6140
+    },
+    {
+      "epoch": 1.0128445276960127,
+      "grad_norm": 1.4431626796722412,
+      "learning_rate": 1.5437640634234234e-05,
+      "loss": 0.4992,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0144912620160145,
+      "grad_norm": 1.131432056427002,
+      "learning_rate": 1.5422712717799714e-05,
+      "loss": 0.5107,
+      "step": 6160
+    },
+    {
+      "epoch": 1.0161379963360162,
+      "grad_norm": 1.4015928506851196,
+      "learning_rate": 1.5407767663969576e-05,
+      "loss": 0.5125,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0177847306560177,
+      "grad_norm": 1.4604285955429077,
+      "learning_rate": 1.5392805519974678e-05,
+      "loss": 0.5076,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0194314649760194,
+      "grad_norm": 1.2894060611724854,
+      "learning_rate": 1.5377826333099855e-05,
+      "loss": 0.51,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0210781992960212,
+      "grad_norm": 1.5654339790344238,
+      "learning_rate": 1.5362830150683838e-05,
+      "loss": 0.4946,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0227249336160227,
+      "grad_norm": 1.7387830018997192,
+      "learning_rate": 1.5347817020119044e-05,
+      "loss": 0.5302,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0243716679360244,
+      "grad_norm": 1.534555196762085,
+      "learning_rate": 1.5332786988851462e-05,
+      "loss": 0.5191,
+      "step": 6220
+    },
+    {
+      "epoch": 1.026018402256026,
+      "grad_norm": 1.6992589235305786,
+      "learning_rate": 1.531774010438049e-05,
+      "loss": 0.5076,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0276651365760276,
+      "grad_norm": 1.5963928699493408,
+      "learning_rate": 1.530267641425879e-05,
+      "loss": 0.4864,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0293118708960294,
+      "grad_norm": 1.390317678451538,
+      "learning_rate": 1.528759596609213e-05,
+      "loss": 0.5072,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0309586052160309,
+      "grad_norm": 1.8305940628051758,
+      "learning_rate": 1.5272498807539248e-05,
+      "loss": 0.4892,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0326053395360326,
+      "grad_norm": 4.596288204193115,
+      "learning_rate": 1.5257384986311671e-05,
+      "loss": 0.513,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0342520738560343,
+      "grad_norm": 1.9707483053207397,
+      "learning_rate": 1.5242254550173612e-05,
+      "loss": 0.4815,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0358988081760359,
+      "grad_norm": 1.657809853553772,
+      "learning_rate": 1.5227107546941772e-05,
+      "loss": 0.5244,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0375455424960376,
+      "grad_norm": 1.530784010887146,
+      "learning_rate": 1.5211944024485216e-05,
+      "loss": 0.5041,
+      "step": 6300
+    },
+    {
+      "epoch": 1.039192276816039,
+      "grad_norm": 1.7985795736312866,
+      "learning_rate": 1.5196764030725217e-05,
+      "loss": 0.4922,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0408390111360408,
+      "grad_norm": 1.5775318145751953,
+      "learning_rate": 1.5181567613635099e-05,
+      "loss": 0.5004,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0424857454560426,
+      "grad_norm": 1.5690876245498657,
+      "learning_rate": 1.5166354821240093e-05,
+      "loss": 0.5047,
+      "step": 6330
+    },
+    {
+      "epoch": 1.044132479776044,
+      "grad_norm": 1.4452539682388306,
+      "learning_rate": 1.5151125701617168e-05,
+      "loss": 0.4845,
+      "step": 6340
+    },
+    {
+      "epoch": 1.0457792140960458,
+      "grad_norm": 2.1906886100769043,
+      "learning_rate": 1.513588030289491e-05,
+      "loss": 0.4827,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0474259484160475,
+      "grad_norm": 2.8921637535095215,
+      "learning_rate": 1.5120618673253335e-05,
+      "loss": 0.5144,
+      "step": 6360
+    },
+    {
+      "epoch": 1.049072682736049,
+      "grad_norm": 2.0884976387023926,
+      "learning_rate": 1.510534086092377e-05,
+      "loss": 0.5055,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0507194170560508,
+      "grad_norm": 2.065589189529419,
+      "learning_rate": 1.5090046914188672e-05,
+      "loss": 0.5281,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0523661513760523,
+      "grad_norm": 2.384566068649292,
+      "learning_rate": 1.507473688138149e-05,
+      "loss": 0.5073,
+      "step": 6390
+    },
+    {
+      "epoch": 1.054012885696054,
+      "grad_norm": 2.007929801940918,
+      "learning_rate": 1.5059410810886515e-05,
+      "loss": 0.4977,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0556596200160557,
+      "grad_norm": 1.7083964347839355,
+      "learning_rate": 1.5044068751138716e-05,
+      "loss": 0.5095,
+      "step": 6410
+    },
+    {
+      "epoch": 1.0573063543360572,
+      "grad_norm": 1.9435292482376099,
+      "learning_rate": 1.5028710750623592e-05,
+      "loss": 0.5173,
+      "step": 6420
+    },
+    {
+      "epoch": 1.058953088656059,
+      "grad_norm": 3.824756145477295,
+      "learning_rate": 1.5013336857877025e-05,
+      "loss": 0.482,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0605998229760607,
+      "grad_norm": 2.91257381439209,
+      "learning_rate": 1.4997947121485119e-05,
+      "loss": 0.5235,
+      "step": 6440
+    },
+    {
+      "epoch": 1.0622465572960622,
+      "grad_norm": 1.5288364887237549,
+      "learning_rate": 1.4982541590084047e-05,
+      "loss": 0.4804,
+      "step": 6450
+    },
+    {
+      "epoch": 1.063893291616064,
+      "grad_norm": 1.5502305030822754,
+      "learning_rate": 1.4967120312359902e-05,
+      "loss": 0.516,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0655400259360654,
+      "grad_norm": 1.62122642993927,
+      "learning_rate": 1.4951683337048536e-05,
+      "loss": 0.4812,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0671867602560672,
+      "grad_norm": 1.508252739906311,
+      "learning_rate": 1.4936230712935416e-05,
+      "loss": 0.4897,
+      "step": 6480
+    },
+    {
+      "epoch": 1.068833494576069,
+      "grad_norm": 1.665816068649292,
+      "learning_rate": 1.4920762488855457e-05,
+      "loss": 0.4891,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0704802288960704,
+      "grad_norm": 1.297685146331787,
+      "learning_rate": 1.490527871369288e-05,
+      "loss": 0.4643,
+      "step": 6500
+    },
+    {
+      "epoch": 1.071962289784072,
+      "grad_norm": 1.9492204189300537,
+      "learning_rate": 1.4889779436381046e-05,
+      "loss": 0.5146,
+      "step": 6510
+    },
+    {
+      "epoch": 1.0736090241040737,
+      "grad_norm": 1.655253291130066,
+      "learning_rate": 1.4874264705902319e-05,
+      "loss": 0.5261,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0752557584240752,
+      "grad_norm": 1.1686067581176758,
+      "learning_rate": 1.4858734571287885e-05,
+      "loss": 0.5401,
+      "step": 6530
+    },
+    {
+      "epoch": 1.076902492744077,
+      "grad_norm": 1.5212386846542358,
+      "learning_rate": 1.4843189081617622e-05,
+      "loss": 0.5403,
+      "step": 6540
+    },
+    {
+      "epoch": 1.0785492270640786,
+      "grad_norm": 1.1677196025848389,
+      "learning_rate": 1.4827628286019928e-05,
+      "loss": 0.5261,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0801959613840801,
+      "grad_norm": 1.2745987176895142,
+      "learning_rate": 1.4812052233671581e-05,
+      "loss": 0.5309,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0818426957040819,
+      "grad_norm": 1.4600034952163696,
+      "learning_rate": 1.4796460973797566e-05,
+      "loss": 0.5168,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0834894300240836,
+      "grad_norm": 1.8619006872177124,
+      "learning_rate": 1.4780854555670932e-05,
+      "loss": 0.5423,
+      "step": 6580
+    },
+    {
+      "epoch": 1.085136164344085,
+      "grad_norm": 1.418540120124817,
+      "learning_rate": 1.4765233028612633e-05,
+      "loss": 0.501,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0867828986640868,
+      "grad_norm": 1.2168614864349365,
+      "learning_rate": 1.474959644199137e-05,
+      "loss": 0.5219,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0884296329840883,
+      "grad_norm": 1.4315396547317505,
+      "learning_rate": 1.4733944845223441e-05,
+      "loss": 0.5274,
+      "step": 6610
+    },
+    {
+      "epoch": 1.09007636730409,
+      "grad_norm": 1.5139554738998413,
+      "learning_rate": 1.4718278287772574e-05,
+      "loss": 0.5181,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0917231016240918,
+      "grad_norm": 1.6579830646514893,
+      "learning_rate": 1.4702596819149784e-05,
+      "loss": 0.5303,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0933698359440933,
+      "grad_norm": 1.6588560342788696,
+      "learning_rate": 1.46869004889132e-05,
+      "loss": 0.4853,
+      "step": 6640
+    },
+    {
+      "epoch": 1.095016570264095,
+      "grad_norm": 1.925136923789978,
+      "learning_rate": 1.4671189346667933e-05,
+      "loss": 0.5222,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0966633045840966,
+      "grad_norm": 1.1029921770095825,
+      "learning_rate": 1.4655463442065893e-05,
+      "loss": 0.4927,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0983100389040983,
+      "grad_norm": 1.2987926006317139,
+      "learning_rate": 1.4639722824805644e-05,
+      "loss": 0.5328,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0999567732241,
+      "grad_norm": 1.4294719696044922,
+      "learning_rate": 1.4623967544632252e-05,
+      "loss": 0.5172,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1016035075441015,
+      "grad_norm": 1.1788091659545898,
+      "learning_rate": 1.4608197651337117e-05,
+      "loss": 0.5056,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1032502418641033,
+      "grad_norm": 1.2116508483886719,
+      "learning_rate": 1.4592413194757826e-05,
+      "loss": 0.5182,
+      "step": 6700
+    },
+    {
+      "epoch": 1.104896976184105,
+      "grad_norm": 1.1964901685714722,
+      "learning_rate": 1.4576614224777982e-05,
+      "loss": 0.4942,
+      "step": 6710
+    },
+    {
+      "epoch": 1.1065437105041065,
+      "grad_norm": 1.5274699926376343,
+      "learning_rate": 1.4560800791327063e-05,
+      "loss": 0.5232,
+      "step": 6720
+    },
+    {
+      "epoch": 1.1081904448241082,
+      "grad_norm": 1.4706414937973022,
+      "learning_rate": 1.4544972944380256e-05,
+      "loss": 0.5432,
+      "step": 6730
+    },
+    {
+      "epoch": 1.10983717914411,
+      "grad_norm": 1.6734999418258667,
+      "learning_rate": 1.4529130733958292e-05,
+      "loss": 0.5161,
+      "step": 6740
+    },
+    {
+      "epoch": 1.1114839134641115,
+      "grad_norm": 1.2489820718765259,
+      "learning_rate": 1.4513274210127298e-05,
+      "loss": 0.5232,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1131306477841132,
+      "grad_norm": 1.5743283033370972,
+      "learning_rate": 1.4497403422998634e-05,
+      "loss": 0.5028,
+      "step": 6760
+    },
+    {
+      "epoch": 1.1147773821041147,
+      "grad_norm": 1.1314998865127563,
+      "learning_rate": 1.448151842272875e-05,
+      "loss": 0.5087,
+      "step": 6770
+    },
+    {
+      "epoch": 1.1164241164241164,
+      "grad_norm": 1.446236491203308,
+      "learning_rate": 1.446561925951899e-05,
+      "loss": 0.5107,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1180708507441182,
+      "grad_norm": 1.3917357921600342,
+      "learning_rate": 1.4449705983615474e-05,
+      "loss": 0.5119,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1197175850641197,
+      "grad_norm": 1.290035605430603,
+      "learning_rate": 1.443377864530892e-05,
+      "loss": 0.522,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1213643193841214,
+      "grad_norm": 1.2674860954284668,
+      "learning_rate": 1.4417837294934484e-05,
+      "loss": 0.5249,
+      "step": 6810
+    },
+    {
+      "epoch": 1.123011053704123,
+      "grad_norm": 1.2753984928131104,
+      "learning_rate": 1.4401881982871604e-05,
+      "loss": 0.4972,
+      "step": 6820
+    },
+    {
+      "epoch": 1.1246577880241246,
+      "grad_norm": 1.801187515258789,
+      "learning_rate": 1.4385912759543843e-05,
+      "loss": 0.5043,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1263045223441264,
+      "grad_norm": 1.2739251852035522,
+      "learning_rate": 1.436992967541873e-05,
+      "loss": 0.509,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1279512566641279,
+      "grad_norm": 1.319750189781189,
+      "learning_rate": 1.4353932781007594e-05,
+      "loss": 0.529,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1295979909841296,
+      "grad_norm": 1.7615410089492798,
+      "learning_rate": 1.433792212686541e-05,
+      "loss": 0.5045,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1312447253041313,
+      "grad_norm": 1.2809779644012451,
+      "learning_rate": 1.432189776359064e-05,
+      "loss": 0.4994,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1328914596241328,
+      "grad_norm": 1.5949429273605347,
+      "learning_rate": 1.4305859741825068e-05,
+      "loss": 0.502,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1345381939441346,
+      "grad_norm": 1.4932994842529297,
+      "learning_rate": 1.4289808112253643e-05,
+      "loss": 0.5026,
+      "step": 6890
+    },
+    {
+      "epoch": 1.1361849282641363,
+      "grad_norm": 1.708443284034729,
+      "learning_rate": 1.4273742925604322e-05,
+      "loss": 0.4747,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1378316625841378,
+      "grad_norm": 1.2732783555984497,
+      "learning_rate": 1.4257664232647903e-05,
+      "loss": 0.5154,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1394783969041395,
+      "grad_norm": 1.4677982330322266,
+      "learning_rate": 1.424157208419787e-05,
+      "loss": 0.4879,
+      "step": 6920
+    },
+    {
+      "epoch": 1.141125131224141,
+      "grad_norm": 1.3541361093521118,
+      "learning_rate": 1.422546653111023e-05,
+      "loss": 0.5129,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1427718655441428,
+      "grad_norm": 1.359041690826416,
+      "learning_rate": 1.4209347624283352e-05,
+      "loss": 0.4885,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1444185998641445,
+      "grad_norm": 1.3765989542007446,
+      "learning_rate": 1.4193215414657808e-05,
+      "loss": 0.543,
+      "step": 6950
+    },
+    {
+      "epoch": 1.146065334184146,
+      "grad_norm": 1.3113765716552734,
+      "learning_rate": 1.4177069953216207e-05,
+      "loss": 0.4778,
+      "step": 6960
+    },
+    {
+      "epoch": 1.1477120685041478,
+      "grad_norm": 1.1242080926895142,
+      "learning_rate": 1.4160911290983041e-05,
+      "loss": 0.4809,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1493588028241493,
+      "grad_norm": 1.536487340927124,
+      "learning_rate": 1.4144739479024527e-05,
+      "loss": 0.4818,
+      "step": 6980
+    },
+    {
+      "epoch": 1.151005537144151,
+      "grad_norm": 1.2508544921875,
+      "learning_rate": 1.4128554568448425e-05,
+      "loss": 0.514,
+      "step": 6990
+    },
+    {
+      "epoch": 1.1526522714641527,
+      "grad_norm": 1.192643404006958,
+      "learning_rate": 1.4112356610403897e-05,
+      "loss": 0.4733,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1542990057841542,
+      "grad_norm": 1.2769783735275269,
+      "learning_rate": 1.4096145656081347e-05,
+      "loss": 0.4891,
+      "step": 7010
+    },
+    {
+      "epoch": 1.155945740104156,
+      "grad_norm": 1.1201382875442505,
+      "learning_rate": 1.4079921756712238e-05,
+      "loss": 0.4824,
+      "step": 7020
+    },
+    {
+      "epoch": 1.1575924744241575,
+      "grad_norm": 1.3404284715652466,
+      "learning_rate": 1.4063684963568948e-05,
+      "loss": 0.5242,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1592392087441592,
+      "grad_norm": 1.3030028343200684,
+      "learning_rate": 1.4047435327964609e-05,
+      "loss": 0.5023,
+      "step": 7040
+    },
+    {
+      "epoch": 1.160885943064161,
+      "grad_norm": 1.7895567417144775,
+      "learning_rate": 1.4031172901252931e-05,
+      "loss": 0.4866,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1625326773841624,
+      "grad_norm": 1.2120519876480103,
+      "learning_rate": 1.4014897734828055e-05,
+      "loss": 0.4819,
+      "step": 7060
+    },
+    {
+      "epoch": 1.1641794117041642,
+      "grad_norm": 1.3895535469055176,
+      "learning_rate": 1.3998609880124373e-05,
+      "loss": 0.4955,
+      "step": 7070
+    },
+    {
+      "epoch": 1.165826146024166,
+      "grad_norm": 1.2869995832443237,
+      "learning_rate": 1.3982309388616385e-05,
+      "loss": 0.4857,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1674728803441674,
+      "grad_norm": 1.2224863767623901,
+      "learning_rate": 1.3965996311818526e-05,
+      "loss": 0.5069,
+      "step": 7090
+    },
+    {
+      "epoch": 1.1691196146641691,
+      "grad_norm": 1.1428492069244385,
+      "learning_rate": 1.3949670701285e-05,
+      "loss": 0.4908,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1707663489841709,
+      "grad_norm": 1.8123971223831177,
+      "learning_rate": 1.3933332608609624e-05,
+      "loss": 0.504,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1724130833041724,
+      "grad_norm": 1.2169560194015503,
+      "learning_rate": 1.3916982085425663e-05,
+      "loss": 0.5015,
+      "step": 7120
+    },
+    {
+      "epoch": 1.174059817624174,
+      "grad_norm": 1.34042489528656,
+      "learning_rate": 1.3900619183405666e-05,
+      "loss": 0.5077,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1757065519441756,
+      "grad_norm": 1.6218396425247192,
+      "learning_rate": 1.38842439542613e-05,
+      "loss": 0.5148,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1773532862641773,
+      "grad_norm": 1.162052869796753,
+      "learning_rate": 1.3867856449743191e-05,
+      "loss": 0.4832,
+      "step": 7150
+    },
+    {
+      "epoch": 1.179000020584179,
+      "grad_norm": 1.2878142595291138,
+      "learning_rate": 1.3851456721640762e-05,
+      "loss": 0.5124,
+      "step": 7160
+    },
+    {
+      "epoch": 1.1806467549041806,
+      "grad_norm": 1.1885998249053955,
+      "learning_rate": 1.3835044821782063e-05,
+      "loss": 0.4829,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1822934892241823,
+      "grad_norm": 1.176771879196167,
+      "learning_rate": 1.3818620802033609e-05,
+      "loss": 0.4881,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1839402235441838,
+      "grad_norm": 1.699350357055664,
+      "learning_rate": 1.3802184714300219e-05,
+      "loss": 0.4995,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1855869578641856,
+      "grad_norm": 1.660423994064331,
+      "learning_rate": 1.3785736610524854e-05,
+      "loss": 0.5151,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1872336921841873,
+      "grad_norm": 1.5366398096084595,
+      "learning_rate": 1.3769276542688444e-05,
+      "loss": 0.5002,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1888804265041888,
+      "grad_norm": 1.597117304801941,
+      "learning_rate": 1.3752804562809731e-05,
+      "loss": 0.4737,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1905271608241905,
+      "grad_norm": 1.2548353672027588,
+      "learning_rate": 1.3736320722945103e-05,
+      "loss": 0.4971,
+      "step": 7230
+    },
+    {
+      "epoch": 1.1921738951441923,
+      "grad_norm": 1.5998353958129883,
+      "learning_rate": 1.3719825075188427e-05,
+      "loss": 0.4988,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1938206294641938,
+      "grad_norm": 1.853366732597351,
+      "learning_rate": 1.370331767167089e-05,
+      "loss": 0.5122,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1954673637841955,
+      "grad_norm": 1.6614432334899902,
+      "learning_rate": 1.3686798564560831e-05,
+      "loss": 0.5014,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1971140981041972,
+      "grad_norm": 1.8630009889602661,
+      "learning_rate": 1.3670267806063573e-05,
+      "loss": 0.4648,
+      "step": 7270
+    },
+    {
+      "epoch": 1.1987608324241987,
+      "grad_norm": 1.5616339445114136,
+      "learning_rate": 1.3653725448421258e-05,
+      "loss": 0.5241,
+      "step": 7280
+    },
+    {
+      "epoch": 1.2004075667442005,
+      "grad_norm": 1.2994410991668701,
+      "learning_rate": 1.3637171543912693e-05,
+      "loss": 0.4812,
+      "step": 7290
+    },
+    {
+      "epoch": 1.202054301064202,
+      "grad_norm": 1.523964524269104,
+      "learning_rate": 1.362060614485317e-05,
+      "loss": 0.4898,
+      "step": 7300
+    },
+    {
+      "epoch": 1.2037010353842037,
+      "grad_norm": 2.751901865005493,
+      "learning_rate": 1.3604029303594311e-05,
+      "loss": 0.5339,
+      "step": 7310
+    },
+    {
+      "epoch": 1.2053477697042054,
+      "grad_norm": 1.832046389579773,
+      "learning_rate": 1.3587441072523893e-05,
+      "loss": 0.4728,
+      "step": 7320
+    },
+    {
+      "epoch": 1.206994504024207,
+      "grad_norm": 1.397538423538208,
+      "learning_rate": 1.3570841504065695e-05,
+      "loss": 0.4737,
+      "step": 7330
+    },
+    {
+      "epoch": 1.2086412383442087,
+      "grad_norm": 2.1878819465637207,
+      "learning_rate": 1.3554230650679326e-05,
+      "loss": 0.4977,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2102879726642102,
+      "grad_norm": 1.9552628993988037,
+      "learning_rate": 1.3537608564860053e-05,
+      "loss": 0.4879,
+      "step": 7350
+    },
+    {
+      "epoch": 1.211934706984212,
+      "grad_norm": 1.4052588939666748,
+      "learning_rate": 1.3520975299138637e-05,
+      "loss": 0.4753,
+      "step": 7360
+    },
+    {
+      "epoch": 1.2135814413042136,
+      "grad_norm": 1.8420672416687012,
+      "learning_rate": 1.3504330906081187e-05,
+      "loss": 0.4877,
+      "step": 7370
+    },
+    {
+      "epoch": 1.2152281756242151,
+      "grad_norm": 1.29952073097229,
+      "learning_rate": 1.3487675438288965e-05,
+      "loss": 0.4822,
+      "step": 7380
+    },
+    {
+      "epoch": 1.2168749099442169,
+      "grad_norm": 1.3250713348388672,
+      "learning_rate": 1.3471008948398233e-05,
+      "loss": 0.465,
+      "step": 7390
+    },
+    {
+      "epoch": 1.2185216442642186,
+      "grad_norm": 1.2507812976837158,
+      "learning_rate": 1.345433148908009e-05,
+      "loss": 0.4588,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2201683785842201,
+      "grad_norm": 1.415877103805542,
+      "learning_rate": 1.3437643113040302e-05,
+      "loss": 0.4691,
+      "step": 7410
+    },
+    {
+      "epoch": 1.2218151129042218,
+      "grad_norm": 1.3288524150848389,
+      "learning_rate": 1.3420943873019128e-05,
+      "loss": 0.4664,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2234618472242236,
+      "grad_norm": 1.4450100660324097,
+      "learning_rate": 1.3404233821791172e-05,
+      "loss": 0.4613,
+      "step": 7430
+    },
+    {
+      "epoch": 1.225108581544225,
+      "grad_norm": 1.3197182416915894,
+      "learning_rate": 1.3387513012165196e-05,
+      "loss": 0.485,
+      "step": 7440
+    },
+    {
+      "epoch": 1.2267553158642268,
+      "grad_norm": 1.192579746246338,
+      "learning_rate": 1.3370781496983965e-05,
+      "loss": 0.4967,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2284020501842283,
+      "grad_norm": 1.670664668083191,
+      "learning_rate": 1.3354039329124074e-05,
+      "loss": 0.4973,
+      "step": 7460
+    },
+    {
+      "epoch": 1.23004878450423,
+      "grad_norm": 1.251954197883606,
+      "learning_rate": 1.3337286561495788e-05,
+      "loss": 0.473,
+      "step": 7470
+    },
+    {
+      "epoch": 1.2316955188242318,
+      "grad_norm": 1.7240633964538574,
+      "learning_rate": 1.3320523247042867e-05,
+      "loss": 0.4754,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2333422531442333,
+      "grad_norm": 1.3650866746902466,
+      "learning_rate": 1.33037494387424e-05,
+      "loss": 0.4382,
+      "step": 7490
+    },
+    {
+      "epoch": 1.234988987464235,
+      "grad_norm": 1.5670567750930786,
+      "learning_rate": 1.3286965189604648e-05,
+      "loss": 0.4662,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2366357217842365,
+      "grad_norm": 1.2606995105743408,
+      "learning_rate": 1.3270170552672864e-05,
+      "loss": 0.4672,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2382824561042383,
+      "grad_norm": 1.3360075950622559,
+      "learning_rate": 1.3253365581023123e-05,
+      "loss": 0.4842,
+      "step": 7520
+    },
+    {
+      "epoch": 1.23992919042424,
+      "grad_norm": 1.5268915891647339,
+      "learning_rate": 1.3236550327764169e-05,
+      "loss": 0.4793,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2415759247442415,
+      "grad_norm": 1.3975640535354614,
+      "learning_rate": 1.3219724846037237e-05,
+      "loss": 0.4645,
+      "step": 7540
+    },
+    {
+      "epoch": 1.2432226590642432,
+      "grad_norm": 1.4461370706558228,
+      "learning_rate": 1.3202889189015883e-05,
+      "loss": 0.4699,
+      "step": 7550
+    },
+    {
+      "epoch": 1.244869393384245,
+      "grad_norm": 1.379309892654419,
+      "learning_rate": 1.3186043409905831e-05,
+      "loss": 0.4732,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2465161277042465,
+      "grad_norm": 1.8799365758895874,
+      "learning_rate": 1.3169187561944776e-05,
+      "loss": 0.4716,
+      "step": 7570
+    },
+    {
+      "epoch": 1.2481628620242482,
+      "grad_norm": 1.2510124444961548,
+      "learning_rate": 1.3152321698402253e-05,
+      "loss": 0.4666,
+      "step": 7580
+    },
+    {
+      "epoch": 1.24980959634425,
+      "grad_norm": 1.4777454137802124,
+      "learning_rate": 1.3135445872579435e-05,
+      "loss": 0.4837,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2514563306642514,
+      "grad_norm": 1.3962723016738892,
+      "learning_rate": 1.3118560137808985e-05,
+      "loss": 0.4803,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2531030649842532,
+      "grad_norm": 1.7384527921676636,
+      "learning_rate": 1.310166454745488e-05,
+      "loss": 0.4683,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2547497993042547,
+      "grad_norm": 1.8363560438156128,
+      "learning_rate": 1.3084759154912243e-05,
+      "loss": 0.4645,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2563965336242564,
+      "grad_norm": 1.5209704637527466,
+      "learning_rate": 1.3067844013607179e-05,
+      "loss": 0.4482,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2580432679442581,
+      "grad_norm": 3.5697286128997803,
+      "learning_rate": 1.3050919176996592e-05,
+      "loss": 0.4452,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2596900022642596,
+      "grad_norm": 1.9864082336425781,
+      "learning_rate": 1.3033984698568038e-05,
+      "loss": 0.4683,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2613367365842614,
+      "grad_norm": 2.205308437347412,
+      "learning_rate": 1.3017040631839536e-05,
+      "loss": 0.4605,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2629834709042629,
+      "grad_norm": 1.6735590696334839,
+      "learning_rate": 1.3000087030359411e-05,
+      "loss": 0.4727,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2646302052242646,
+      "grad_norm": 1.703411340713501,
+      "learning_rate": 1.298312394770612e-05,
+      "loss": 0.4644,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2662769395442663,
+      "grad_norm": 1.6092456579208374,
+      "learning_rate": 1.2966151437488078e-05,
+      "loss": 0.4496,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2679236738642679,
+      "grad_norm": 2.1162524223327637,
+      "learning_rate": 1.2949169553343504e-05,
+      "loss": 0.4658,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2695704081842696,
+      "grad_norm": 1.9072030782699585,
+      "learning_rate": 1.2932178348940231e-05,
+      "loss": 0.4503,
+      "step": 7710
+    },
+    {
+      "epoch": 1.271217142504271,
+      "grad_norm": 1.529270887374878,
+      "learning_rate": 1.2915177877975556e-05,
+      "loss": 0.4608,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2728638768242728,
+      "grad_norm": 1.5956228971481323,
+      "learning_rate": 1.2898168194176056e-05,
+      "loss": 0.4385,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2745106111442746,
+      "grad_norm": 1.6679530143737793,
+      "learning_rate": 1.288114935129742e-05,
+      "loss": 0.4593,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2761573454642763,
+      "grad_norm": 1.8094053268432617,
+      "learning_rate": 1.2864121403124288e-05,
+      "loss": 0.4918,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2778040797842778,
+      "grad_norm": 2.4337899684906006,
+      "learning_rate": 1.2847084403470076e-05,
+      "loss": 0.4849,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2794508141042795,
+      "grad_norm": 1.9572261571884155,
+      "learning_rate": 1.2830038406176804e-05,
+      "loss": 0.4923,
+      "step": 7770
+    },
+    {
+      "epoch": 1.281097548424281,
+      "grad_norm": 1.9552801847457886,
+      "learning_rate": 1.2812983465114925e-05,
+      "loss": 0.4959,
+      "step": 7780
+    },
+    {
+      "epoch": 1.2827442827442828,
+      "grad_norm": 1.6549162864685059,
+      "learning_rate": 1.2795919634183159e-05,
+      "loss": 0.4571,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2843910170642845,
+      "grad_norm": 1.427476167678833,
+      "learning_rate": 1.277884696730832e-05,
+      "loss": 0.4831,
+      "step": 7800
+    },
+    {
+      "epoch": 1.286037751384286,
+      "grad_norm": 2.2045860290527344,
+      "learning_rate": 1.2761765518445146e-05,
+      "loss": 0.4709,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2876844857042877,
+      "grad_norm": 1.556210994720459,
+      "learning_rate": 1.274467534157613e-05,
+      "loss": 0.4733,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2893312200242892,
+      "grad_norm": 1.9513988494873047,
+      "learning_rate": 1.2727576490711344e-05,
+      "loss": 0.4755,
+      "step": 7830
+    },
+    {
+      "epoch": 1.290977954344291,
+      "grad_norm": 1.7663830518722534,
+      "learning_rate": 1.2710469019888283e-05,
+      "loss": 0.4407,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2926246886642927,
+      "grad_norm": 1.8734620809555054,
+      "learning_rate": 1.2693352983171669e-05,
+      "loss": 0.4683,
+      "step": 7850
+    },
+    {
+      "epoch": 1.2942714229842942,
+      "grad_norm": 2.0744926929473877,
+      "learning_rate": 1.2676228434653307e-05,
+      "loss": 0.468,
+      "step": 7860
+    },
+    {
+      "epoch": 1.295918157304296,
+      "grad_norm": 1.9999136924743652,
+      "learning_rate": 1.2659095428451898e-05,
+      "loss": 0.4709,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2975648916242974,
+      "grad_norm": 1.605631709098816,
+      "learning_rate": 1.2641954018712863e-05,
+      "loss": 0.439,
+      "step": 7880
+    },
+    {
+      "epoch": 1.2992116259442992,
+      "grad_norm": 2.568913221359253,
+      "learning_rate": 1.2624804259608194e-05,
+      "loss": 0.4881,
+      "step": 7890
+    },
+    {
+      "epoch": 1.300858360264301,
+      "grad_norm": 2.5562288761138916,
+      "learning_rate": 1.2607646205336264e-05,
+      "loss": 0.507,
+      "step": 7900
+    },
+    {
+      "epoch": 1.3025050945843026,
+      "grad_norm": 2.0905675888061523,
+      "learning_rate": 1.2590479910121662e-05,
+      "loss": 0.467,
+      "step": 7910
+    },
+    {
+      "epoch": 1.3041518289043041,
+      "grad_norm": 2.1957993507385254,
+      "learning_rate": 1.2573305428215017e-05,
+      "loss": 0.4565,
+      "step": 7920
+    },
+    {
+      "epoch": 1.3057985632243059,
+      "grad_norm": 1.6840404272079468,
+      "learning_rate": 1.2556122813892834e-05,
+      "loss": 0.4654,
+      "step": 7930
+    },
+    {
+      "epoch": 1.3074452975443074,
+      "grad_norm": 1.8692106008529663,
+      "learning_rate": 1.2538932121457322e-05,
+      "loss": 0.452,
+      "step": 7940
+    },
+    {
+      "epoch": 1.3090920318643091,
+      "grad_norm": 1.645704746246338,
+      "learning_rate": 1.2521733405236214e-05,
+      "loss": 0.451,
+      "step": 7950
+    },
+    {
+      "epoch": 1.3107387661843108,
+      "grad_norm": 1.3794227838516235,
+      "learning_rate": 1.2504526719582596e-05,
+      "loss": 0.479,
+      "step": 7960
+    },
+    {
+      "epoch": 1.3123855005043124,
+      "grad_norm": 1.594277262687683,
+      "learning_rate": 1.2487312118874755e-05,
+      "loss": 0.4686,
+      "step": 7970
+    },
+    {
+      "epoch": 1.314032234824314,
+      "grad_norm": 2.5194473266601562,
+      "learning_rate": 1.247008965751598e-05,
+      "loss": 0.4672,
+      "step": 7980
+    },
+    {
+      "epoch": 1.3156789691443156,
+      "grad_norm": 1.4308621883392334,
+      "learning_rate": 1.2452859389934399e-05,
+      "loss": 0.4429,
+      "step": 7990
+    },
+    {
+      "epoch": 1.3173257034643173,
+      "grad_norm": 1.6084959506988525,
+      "learning_rate": 1.2435621370582824e-05,
+      "loss": 0.4557,
+      "step": 8000
+    },
+    {
+      "epoch": 1.318972437784319,
+      "grad_norm": 2.098825693130493,
+      "learning_rate": 1.241837565393855e-05,
+      "loss": 0.4538,
+      "step": 8010
+    },
+    {
+      "epoch": 1.3206191721043206,
+      "grad_norm": 1.6159589290618896,
+      "learning_rate": 1.2401122294503212e-05,
+      "loss": 0.4345,
+      "step": 8020
+    },
+    {
+      "epoch": 1.3222659064243223,
+      "grad_norm": 1.534826636314392,
+      "learning_rate": 1.2383861346802585e-05,
+      "loss": 0.444,
+      "step": 8030
+    },
+    {
+      "epoch": 1.3239126407443238,
+      "grad_norm": 1.4128071069717407,
+      "learning_rate": 1.2366592865386433e-05,
+      "loss": 0.4529,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3255593750643255,
+      "grad_norm": 1.3641729354858398,
+      "learning_rate": 1.2349316904828327e-05,
+      "loss": 0.4563,
+      "step": 8050
+    },
+    {
+      "epoch": 1.3272061093843273,
+      "grad_norm": 1.5807952880859375,
+      "learning_rate": 1.2332033519725474e-05,
+      "loss": 0.4433,
+      "step": 8060
+    },
+    {
+      "epoch": 1.328852843704329,
+      "grad_norm": 1.464307188987732,
+      "learning_rate": 1.231474276469855e-05,
+      "loss": 0.4405,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3304995780243305,
+      "grad_norm": 1.6941524744033813,
+      "learning_rate": 1.2297444694391509e-05,
+      "loss": 0.4552,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3321463123443322,
+      "grad_norm": 1.869858741760254,
+      "learning_rate": 1.2280139363471437e-05,
+      "loss": 0.4543,
+      "step": 8090
+    },
+    {
+      "epoch": 1.3337930466643337,
+      "grad_norm": 2.1110851764678955,
+      "learning_rate": 1.2262826826628357e-05,
+      "loss": 0.4671,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3354397809843355,
+      "grad_norm": 1.4754462242126465,
+      "learning_rate": 1.2245507138575072e-05,
+      "loss": 0.4646,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3370865153043372,
+      "grad_norm": 1.5958442687988281,
+      "learning_rate": 1.2228180354046983e-05,
+      "loss": 0.4384,
+      "step": 8120
+    },
+    {
+      "epoch": 1.3387332496243387,
+      "grad_norm": 1.969876766204834,
+      "learning_rate": 1.2210846527801912e-05,
+      "loss": 0.4393,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3403799839443404,
+      "grad_norm": 1.1938165426254272,
+      "learning_rate": 1.2193505714619937e-05,
+      "loss": 0.4304,
+      "step": 8140
+    },
+    {
+      "epoch": 1.342026718264342,
+      "grad_norm": 1.4246577024459839,
+      "learning_rate": 1.2176157969303227e-05,
+      "loss": 0.4483,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3436734525843437,
+      "grad_norm": 1.4756056070327759,
+      "learning_rate": 1.2158803346675845e-05,
+      "loss": 0.4525,
+      "step": 8160
+    },
+    {
+      "epoch": 1.3453201869043454,
+      "grad_norm": 1.381611704826355,
+      "learning_rate": 1.2141441901583593e-05,
+      "loss": 0.4407,
+      "step": 8170
+    },
+    {
+      "epoch": 1.346966921224347,
+      "grad_norm": 1.3701461553573608,
+      "learning_rate": 1.2124073688893838e-05,
+      "loss": 0.439,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3486136555443486,
+      "grad_norm": 2.179842710494995,
+      "learning_rate": 1.2106698763495333e-05,
+      "loss": 0.4494,
+      "step": 8190
+    },
+    {
+      "epoch": 1.3502603898643502,
+      "grad_norm": 1.3610347509384155,
+      "learning_rate": 1.2089317180298043e-05,
+      "loss": 0.4301,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3519071241843519,
+      "grad_norm": 1.461213231086731,
+      "learning_rate": 1.207192899423297e-05,
+      "loss": 0.4565,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3535538585043536,
+      "grad_norm": 1.9947609901428223,
+      "learning_rate": 1.2054534260251995e-05,
+      "loss": 0.4513,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3552005928243551,
+      "grad_norm": 1.3805066347122192,
+      "learning_rate": 1.2037133033327679e-05,
+      "loss": 0.4198,
+      "step": 8230
+    },
+    {
+      "epoch": 1.3568473271443569,
+      "grad_norm": 1.6667447090148926,
+      "learning_rate": 1.2019725368453111e-05,
+      "loss": 0.4525,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3584940614643584,
+      "grad_norm": 1.3600236177444458,
+      "learning_rate": 1.2002311320641722e-05,
+      "loss": 0.4311,
+      "step": 8250
+    },
+    {
+      "epoch": 1.36014079578436,
+      "grad_norm": 1.2351371049880981,
+      "learning_rate": 1.1984890944927119e-05,
+      "loss": 0.4625,
+      "step": 8260
+    },
+    {
+      "epoch": 1.3617875301043618,
+      "grad_norm": 1.8119263648986816,
+      "learning_rate": 1.1967464296362903e-05,
+      "loss": 0.4306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3634342644243636,
+      "grad_norm": 1.2814642190933228,
+      "learning_rate": 1.1950031430022499e-05,
+      "loss": 0.4399,
+      "step": 8280
+    },
+    {
+      "epoch": 1.365080998744365,
+      "grad_norm": 1.4676977396011353,
+      "learning_rate": 1.1932592400998985e-05,
+      "loss": 0.4423,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3667277330643668,
+      "grad_norm": 1.8431611061096191,
+      "learning_rate": 1.1915147264404916e-05,
+      "loss": 0.4298,
+      "step": 8300
+    },
+    {
+      "epoch": 1.3683744673843683,
+      "grad_norm": 1.3189785480499268,
+      "learning_rate": 1.1897696075372143e-05,
+      "loss": 0.4415,
+      "step": 8310
+    },
+    {
+      "epoch": 1.37002120170437,
+      "grad_norm": 1.5338736772537231,
+      "learning_rate": 1.1880238889051647e-05,
+      "loss": 0.4405,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3716679360243718,
+      "grad_norm": 2.1897244453430176,
+      "learning_rate": 1.1862775760613365e-05,
+      "loss": 0.4169,
+      "step": 8330
+    },
+    {
+      "epoch": 1.3733146703443733,
+      "grad_norm": 1.3329521417617798,
+      "learning_rate": 1.1845306745246012e-05,
+      "loss": 0.4433,
+      "step": 8340
+    },
+    {
+      "epoch": 1.374961404664375,
+      "grad_norm": 1.6167490482330322,
+      "learning_rate": 1.1827831898156905e-05,
+      "loss": 0.4594,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3766081389843765,
+      "grad_norm": 1.357681393623352,
+      "learning_rate": 1.1810351274571792e-05,
+      "loss": 0.4196,
+      "step": 8360
+    },
+    {
+      "epoch": 1.3782548733043782,
+      "grad_norm": 1.3341563940048218,
+      "learning_rate": 1.1792864929734678e-05,
+      "loss": 0.4466,
+      "step": 8370
+    },
+    {
+      "epoch": 1.37990160762438,
+      "grad_norm": 1.418643832206726,
+      "learning_rate": 1.1775372918907646e-05,
+      "loss": 0.4178,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3815483419443815,
+      "grad_norm": 1.5278998613357544,
+      "learning_rate": 1.1757875297370687e-05,
+      "loss": 0.4381,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3831950762643832,
+      "grad_norm": 1.9926173686981201,
+      "learning_rate": 1.174037212042152e-05,
+      "loss": 0.4454,
+      "step": 8400
+    },
+    {
+      "epoch": 1.3848418105843847,
+      "grad_norm": 3.1398065090179443,
+      "learning_rate": 1.1722863443375437e-05,
+      "loss": 0.4214,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3864885449043864,
+      "grad_norm": 1.9283721446990967,
+      "learning_rate": 1.1705349321565085e-05,
+      "loss": 0.4275,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3881352792243882,
+      "grad_norm": 1.2598986625671387,
+      "learning_rate": 1.1687829810340338e-05,
+      "loss": 0.4551,
+      "step": 8430
+    },
+    {
+      "epoch": 1.38978201354439,
+      "grad_norm": 1.5143520832061768,
+      "learning_rate": 1.1670304965068098e-05,
+      "loss": 0.4438,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3914287478643914,
+      "grad_norm": 2.051283597946167,
+      "learning_rate": 1.165277484113212e-05,
+      "loss": 0.4186,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3930754821843931,
+      "grad_norm": 1.9647774696350098,
+      "learning_rate": 1.1635239493932842e-05,
+      "loss": 0.4315,
+      "step": 8460
+    },
+    {
+      "epoch": 1.3947222165043947,
+      "grad_norm": 3.153352737426758,
+      "learning_rate": 1.1617698978887214e-05,
+      "loss": 0.4627,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3963689508243964,
+      "grad_norm": 2.1228764057159424,
+      "learning_rate": 1.1600153351428516e-05,
+      "loss": 0.4405,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3980156851443981,
+      "grad_norm": 1.7035056352615356,
+      "learning_rate": 1.158260266700618e-05,
+      "loss": 0.4512,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3996624194643996,
+      "grad_norm": 2.2831528186798096,
+      "learning_rate": 1.1565046981085621e-05,
+      "loss": 0.4442,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4013091537844014,
+      "grad_norm": 1.483190894126892,
+      "learning_rate": 1.154748634914807e-05,
+      "loss": 0.4119,
+      "step": 8510
+    },
+    {
+      "epoch": 1.4029558881044029,
+      "grad_norm": 1.4663301706314087,
+      "learning_rate": 1.1529920826690375e-05,
+      "loss": 0.4234,
+      "step": 8520
+    },
+    {
+      "epoch": 1.4046026224244046,
+      "grad_norm": 2.566152572631836,
+      "learning_rate": 1.1512350469224846e-05,
+      "loss": 0.423,
+      "step": 8530
+    },
+    {
+      "epoch": 1.4062493567444063,
+      "grad_norm": 1.1131696701049805,
+      "learning_rate": 1.1494775332279076e-05,
+      "loss": 0.4087,
+      "step": 8540
+    },
+    {
+      "epoch": 1.4078960910644078,
+      "grad_norm": 1.7250810861587524,
+      "learning_rate": 1.1477195471395759e-05,
+      "loss": 0.4192,
+      "step": 8550
+    },
+    {
+      "epoch": 1.4095428253844096,
+      "grad_norm": 2.6664037704467773,
+      "learning_rate": 1.1459610942132513e-05,
+      "loss": 0.4299,
+      "step": 8560
+    },
+    {
+      "epoch": 1.411189559704411,
+      "grad_norm": 1.2339338064193726,
+      "learning_rate": 1.144202180006172e-05,
+      "loss": 0.3896,
+      "step": 8570
+    },
+    {
+      "epoch": 1.4128362940244128,
+      "grad_norm": 1.5988686084747314,
+      "learning_rate": 1.1424428100770333e-05,
+      "loss": 0.438,
+      "step": 8580
+    },
+    {
+      "epoch": 1.4144830283444145,
+      "grad_norm": 1.4023103713989258,
+      "learning_rate": 1.140682989985971e-05,
+      "loss": 0.4207,
+      "step": 8590
+    },
+    {
+      "epoch": 1.4161297626644163,
+      "grad_norm": 1.4158636331558228,
+      "learning_rate": 1.1389227252945434e-05,
+      "loss": 0.4342,
+      "step": 8600
+    },
+    {
+      "epoch": 1.4177764969844178,
+      "grad_norm": 1.6897794008255005,
+      "learning_rate": 1.1371620215657136e-05,
+      "loss": 0.4199,
+      "step": 8610
+    },
+    {
+      "epoch": 1.4194232313044195,
+      "grad_norm": 2.0550646781921387,
+      "learning_rate": 1.1354008843638331e-05,
+      "loss": 0.411,
+      "step": 8620
+    },
+    {
+      "epoch": 1.421069965624421,
+      "grad_norm": 1.6463772058486938,
+      "learning_rate": 1.133639319254622e-05,
+      "loss": 0.4246,
+      "step": 8630
+    },
+    {
+      "epoch": 1.4227166999444227,
+      "grad_norm": 1.7673988342285156,
+      "learning_rate": 1.131877331805154e-05,
+      "loss": 0.4296,
+      "step": 8640
+    },
+    {
+      "epoch": 1.4243634342644245,
+      "grad_norm": 1.807904601097107,
+      "learning_rate": 1.1301149275838363e-05,
+      "loss": 0.4131,
+      "step": 8650
+    },
+    {
+      "epoch": 1.426010168584426,
+      "grad_norm": 1.274460792541504,
+      "learning_rate": 1.1283521121603948e-05,
+      "loss": 0.4114,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4276569029044277,
+      "grad_norm": 1.1608306169509888,
+      "learning_rate": 1.1265888911058532e-05,
+      "loss": 0.4429,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4293036372244292,
+      "grad_norm": 1.3403079509735107,
+      "learning_rate": 1.124825269992518e-05,
+      "loss": 0.4456,
+      "step": 8680
+    },
+    {
+      "epoch": 1.430950371544431,
+      "grad_norm": 1.4677191972732544,
+      "learning_rate": 1.1230612543939603e-05,
+      "loss": 0.4251,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4325971058644327,
+      "grad_norm": 1.2995331287384033,
+      "learning_rate": 1.1212968498849973e-05,
+      "loss": 0.4305,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4342438401844342,
+      "grad_norm": 1.7009153366088867,
+      "learning_rate": 1.119532062041675e-05,
+      "loss": 0.4071,
+      "step": 8710
+    },
+    {
+      "epoch": 1.435890574504436,
+      "grad_norm": 1.6432769298553467,
+      "learning_rate": 1.1177668964412519e-05,
+      "loss": 0.4388,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4375373088244374,
+      "grad_norm": 1.4236555099487305,
+      "learning_rate": 1.1160013586621796e-05,
+      "loss": 0.4028,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4391840431444392,
+      "grad_norm": 1.7424023151397705,
+      "learning_rate": 1.1142354542840859e-05,
+      "loss": 0.4037,
+      "step": 8740
+    },
+    {
+      "epoch": 1.4408307774644409,
+      "grad_norm": 1.2468111515045166,
+      "learning_rate": 1.1124691888877575e-05,
+      "loss": 0.4084,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4424775117844426,
+      "grad_norm": 1.6972237825393677,
+      "learning_rate": 1.1107025680551216e-05,
+      "loss": 0.4287,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4441242461044441,
+      "grad_norm": 1.772537112236023,
+      "learning_rate": 1.1089355973692292e-05,
+      "loss": 0.4122,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4457709804244459,
+      "grad_norm": 2.2410008907318115,
+      "learning_rate": 1.1071682824142365e-05,
+      "loss": 0.4277,
+      "step": 8780
+    },
+    {
+      "epoch": 1.4474177147444474,
+      "grad_norm": 1.619391918182373,
+      "learning_rate": 1.1054006287753876e-05,
+      "loss": 0.44,
+      "step": 8790
+    },
+    {
+      "epoch": 1.449064449064449,
+      "grad_norm": 1.8790594339370728,
+      "learning_rate": 1.1036326420389978e-05,
+      "loss": 0.4215,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4507111833844508,
+      "grad_norm": 2.067535877227783,
+      "learning_rate": 1.1018643277924338e-05,
+      "loss": 0.4136,
+      "step": 8810
+    },
+    {
+      "epoch": 1.4523579177044523,
+      "grad_norm": 1.8926339149475098,
+      "learning_rate": 1.1000956916240985e-05,
+      "loss": 0.4328,
+      "step": 8820
+    },
+    {
+      "epoch": 1.454004652024454,
+      "grad_norm": 1.6244444847106934,
+      "learning_rate": 1.0983267391234113e-05,
+      "loss": 0.4103,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4556513863444556,
+      "grad_norm": 1.3383644819259644,
+      "learning_rate": 1.0965574758807924e-05,
+      "loss": 0.4274,
+      "step": 8840
+    },
+    {
+      "epoch": 1.4572981206644573,
+      "grad_norm": 1.4654048681259155,
+      "learning_rate": 1.0947879074876425e-05,
+      "loss": 0.4314,
+      "step": 8850
+    },
+    {
+      "epoch": 1.458944854984459,
+      "grad_norm": 1.5433770418167114,
+      "learning_rate": 1.0930180395363275e-05,
+      "loss": 0.439,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4605915893044605,
+      "grad_norm": 1.5175681114196777,
+      "learning_rate": 1.0912478776201605e-05,
+      "loss": 0.4175,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4622383236244623,
+      "grad_norm": 1.402180552482605,
+      "learning_rate": 1.089477427333383e-05,
+      "loss": 0.389,
+      "step": 8880
+    },
+    {
+      "epoch": 1.4638850579444638,
+      "grad_norm": 1.3952786922454834,
+      "learning_rate": 1.0877066942711476e-05,
+      "loss": 0.4107,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4655317922644655,
+      "grad_norm": 1.4274441003799438,
+      "learning_rate": 1.0859356840295013e-05,
+      "loss": 0.4036,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4671785265844672,
+      "grad_norm": 1.6509827375411987,
+      "learning_rate": 1.0841644022053662e-05,
+      "loss": 0.4023,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4688252609044687,
+      "grad_norm": 1.4822527170181274,
+      "learning_rate": 1.0823928543965236e-05,
+      "loss": 0.417,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4704719952244705,
+      "grad_norm": 1.6752021312713623,
+      "learning_rate": 1.0806210462015946e-05,
+      "loss": 0.3962,
+      "step": 8930
+    },
+    {
+      "epoch": 1.472118729544472,
+      "grad_norm": 2.425339460372925,
+      "learning_rate": 1.0788489832200237e-05,
+      "loss": 0.4551,
+      "step": 8940
+    },
+    {
+      "epoch": 1.4737654638644737,
+      "grad_norm": 1.5440939664840698,
+      "learning_rate": 1.0770766710520607e-05,
+      "loss": 0.4006,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4754121981844754,
+      "grad_norm": 1.8079001903533936,
+      "learning_rate": 1.075304115298742e-05,
+      "loss": 0.397,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4770589325044772,
+      "grad_norm": 1.4747633934020996,
+      "learning_rate": 1.0735313215618748e-05,
+      "loss": 0.4002,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4787056668244787,
+      "grad_norm": 1.5912610292434692,
+      "learning_rate": 1.071758295444018e-05,
+      "loss": 0.4153,
+      "step": 8980
+    },
+    {
+      "epoch": 1.4803524011444804,
+      "grad_norm": 1.4173518419265747,
+      "learning_rate": 1.069985042548465e-05,
+      "loss": 0.4142,
+      "step": 8990
+    },
+    {
+      "epoch": 1.481999135464482,
+      "grad_norm": 1.8204532861709595,
+      "learning_rate": 1.0682115684792256e-05,
+      "loss": 0.4103,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4836458697844837,
+      "grad_norm": 1.718644380569458,
+      "learning_rate": 1.0664378788410092e-05,
+      "loss": 0.4099,
+      "step": 9010
+    },
+    {
+      "epoch": 1.4852926041044854,
+      "grad_norm": 1.6308566331863403,
+      "learning_rate": 1.0646639792392057e-05,
+      "loss": 0.4108,
+      "step": 9020
+    },
+    {
+      "epoch": 1.486939338424487,
+      "grad_norm": 1.60697340965271,
+      "learning_rate": 1.062889875279869e-05,
+      "loss": 0.4109,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4885860727444886,
+      "grad_norm": 1.5269246101379395,
+      "learning_rate": 1.0611155725696988e-05,
+      "loss": 0.3809,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4902328070644901,
+      "grad_norm": 1.802048921585083,
+      "learning_rate": 1.0593410767160229e-05,
+      "loss": 0.4015,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4918795413844919,
+      "grad_norm": 1.5887212753295898,
+      "learning_rate": 1.0575663933267793e-05,
+      "loss": 0.3746,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4935262757044936,
+      "grad_norm": 5.430899620056152,
+      "learning_rate": 1.0557915280104987e-05,
+      "loss": 0.3979,
+      "step": 9070
+    },
+    {
+      "epoch": 1.495173010024495,
+      "grad_norm": 1.8319224119186401,
+      "learning_rate": 1.0540164863762867e-05,
+      "loss": 0.3903,
+      "step": 9080
+    },
+    {
+      "epoch": 1.4968197443444968,
+      "grad_norm": 1.445753574371338,
+      "learning_rate": 1.0522412740338072e-05,
+      "loss": 0.3993,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4984664786644983,
+      "grad_norm": 2.215062141418457,
+      "learning_rate": 1.0504658965932617e-05,
+      "loss": 0.4079,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5001132129845,
+      "grad_norm": 1.9936473369598389,
+      "learning_rate": 1.0486903596653746e-05,
+      "loss": 0.4127,
+      "step": 9110
+    },
+    {
+      "epoch": 1.5017599473045018,
+      "grad_norm": 2.1176092624664307,
+      "learning_rate": 1.0469146688613744e-05,
+      "loss": 0.3993,
+      "step": 9120
+    },
+    {
+      "epoch": 1.5034066816245035,
+      "grad_norm": 2.069631576538086,
+      "learning_rate": 1.0451388297929757e-05,
+      "loss": 0.4327,
+      "step": 9130
+    },
+    {
+      "epoch": 1.505053415944505,
+      "grad_norm": 1.5769414901733398,
+      "learning_rate": 1.043362848072361e-05,
+      "loss": 0.3866,
+      "step": 9140
+    },
+    {
+      "epoch": 1.5067001502645065,
+      "grad_norm": 1.590387225151062,
+      "learning_rate": 1.041586729312165e-05,
+      "loss": 0.3923,
+      "step": 9150
+    },
+    {
+      "epoch": 1.5083468845845083,
+      "grad_norm": 1.8121472597122192,
+      "learning_rate": 1.0398104791254542e-05,
+      "loss": 0.3963,
+      "step": 9160
+    },
+    {
+      "epoch": 1.50999361890451,
+      "grad_norm": 1.914831280708313,
+      "learning_rate": 1.038034103125711e-05,
+      "loss": 0.3835,
+      "step": 9170
+    },
+    {
+      "epoch": 1.5116403532245117,
+      "grad_norm": 1.556657314300537,
+      "learning_rate": 1.0362576069268156e-05,
+      "loss": 0.396,
+      "step": 9180
+    },
+    {
+      "epoch": 1.5132870875445132,
+      "grad_norm": 1.9325206279754639,
+      "learning_rate": 1.0344809961430277e-05,
+      "loss": 0.4204,
+      "step": 9190
+    },
+    {
+      "epoch": 1.514933821864515,
+      "grad_norm": 2.9902329444885254,
+      "learning_rate": 1.0327042763889692e-05,
+      "loss": 0.4115,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5165805561845165,
+      "grad_norm": 1.8379610776901245,
+      "learning_rate": 1.0309274532796064e-05,
+      "loss": 0.3794,
+      "step": 9210
+    },
+    {
+      "epoch": 1.5182272905045182,
+      "grad_norm": 1.8868550062179565,
+      "learning_rate": 1.0291505324302322e-05,
+      "loss": 0.4127,
+      "step": 9220
+    },
+    {
+      "epoch": 1.51987402482452,
+      "grad_norm": 2.223417282104492,
+      "learning_rate": 1.027373519456449e-05,
+      "loss": 0.4269,
+      "step": 9230
+    },
+    {
+      "epoch": 1.5215207591445217,
+      "grad_norm": 2.04238224029541,
+      "learning_rate": 1.0255964199741488e-05,
+      "loss": 0.3928,
+      "step": 9240
+    },
+    {
+      "epoch": 1.5231674934645232,
+      "grad_norm": 2.3193910121917725,
+      "learning_rate": 1.0238192395994989e-05,
+      "loss": 0.4056,
+      "step": 9250
+    },
+    {
+      "epoch": 1.5248142277845247,
+      "grad_norm": 2.5879781246185303,
+      "learning_rate": 1.022041983948921e-05,
+      "loss": 0.3993,
+      "step": 9260
+    },
+    {
+      "epoch": 1.5264609621045264,
+      "grad_norm": 1.9214088916778564,
+      "learning_rate": 1.020264658639075e-05,
+      "loss": 0.3947,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5281076964245282,
+      "grad_norm": 1.8478078842163086,
+      "learning_rate": 1.0184872692868409e-05,
+      "loss": 0.4056,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5297544307445299,
+      "grad_norm": 3.103062629699707,
+      "learning_rate": 1.016709821509301e-05,
+      "loss": 0.3801,
+      "step": 9290
+    },
+    {
+      "epoch": 1.5314011650645314,
+      "grad_norm": 2.530630111694336,
+      "learning_rate": 1.014932320923723e-05,
+      "loss": 0.4195,
+      "step": 9300
+    },
+    {
+      "epoch": 1.533047899384533,
+      "grad_norm": 1.8765909671783447,
+      "learning_rate": 1.0131547731475401e-05,
+      "loss": 0.4163,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5346946337045346,
+      "grad_norm": 1.9875380992889404,
+      "learning_rate": 1.0113771837983361e-05,
+      "loss": 0.4135,
+      "step": 9320
+    },
+    {
+      "epoch": 1.5363413680245364,
+      "grad_norm": 2.6800906658172607,
+      "learning_rate": 1.0095995584938252e-05,
+      "loss": 0.3949,
+      "step": 9330
+    },
+    {
+      "epoch": 1.537988102344538,
+      "grad_norm": 2.1471517086029053,
+      "learning_rate": 1.0078219028518359e-05,
+      "loss": 0.4129,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5396348366645396,
+      "grad_norm": 2.109405517578125,
+      "learning_rate": 1.0060442224902915e-05,
+      "loss": 0.404,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5412815709845413,
+      "grad_norm": 1.336411714553833,
+      "learning_rate": 1.0042665230271947e-05,
+      "loss": 0.423,
+      "step": 9360
+    },
+    {
+      "epoch": 1.5429283053045428,
+      "grad_norm": 1.543381690979004,
+      "learning_rate": 1.0024888100806079e-05,
+      "loss": 0.4204,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5445750396245446,
+      "grad_norm": 1.7605801820755005,
+      "learning_rate": 1.000711089268636e-05,
+      "loss": 0.3922,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5462217739445463,
+      "grad_norm": 1.8124173879623413,
+      "learning_rate": 9.989333662094092e-06,
+      "loss": 0.4036,
+      "step": 9390
+    },
+    {
+      "epoch": 1.547868508264548,
+      "grad_norm": 2.485732316970825,
+      "learning_rate": 9.971556465210643e-06,
+      "loss": 0.3985,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5495152425845495,
+      "grad_norm": 2.339867353439331,
+      "learning_rate": 9.953779358217281e-06,
+      "loss": 0.4196,
+      "step": 9410
+    },
+    {
+      "epoch": 1.551161976904551,
+      "grad_norm": 2.12572979927063,
+      "learning_rate": 9.93600239729499e-06,
+      "loss": 0.3888,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5528087112245528,
+      "grad_norm": 2.382894277572632,
+      "learning_rate": 9.918225638624276e-06,
+      "loss": 0.4123,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5544554455445545,
+      "grad_norm": 2.5262231826782227,
+      "learning_rate": 9.900449138385026e-06,
+      "loss": 0.425,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5561021798645562,
+      "grad_norm": 2.0982043743133545,
+      "learning_rate": 9.882672952756301e-06,
+      "loss": 0.398,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5577489141845577,
+      "grad_norm": 1.5875078439712524,
+      "learning_rate": 9.86489713791617e-06,
+      "loss": 0.3729,
+      "step": 9460
+    },
+    {
+      "epoch": 1.5593956485045593,
+      "grad_norm": 1.499527096748352,
+      "learning_rate": 9.847121750041532e-06,
+      "loss": 0.4208,
+      "step": 9470
+    },
+    {
+      "epoch": 1.561042382824561,
+      "grad_norm": 1.7554186582565308,
+      "learning_rate": 9.829346845307929e-06,
+      "loss": 0.4171,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5626891171445627,
+      "grad_norm": 2.1074600219726562,
+      "learning_rate": 9.811572479889387e-06,
+      "loss": 0.4018,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5643358514645644,
+      "grad_norm": 3.0306382179260254,
+      "learning_rate": 9.793798709958221e-06,
+      "loss": 0.4319,
+      "step": 9500
+    },
+    {
+      "epoch": 1.565982585784566,
+      "grad_norm": 1.5060926675796509,
+      "learning_rate": 9.77602559168486e-06,
+      "loss": 0.3968,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5676293201045677,
+      "grad_norm": 1.7786906957626343,
+      "learning_rate": 9.75825318123768e-06,
+      "loss": 0.3867,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5692760544245692,
+      "grad_norm": 1.4499342441558838,
+      "learning_rate": 9.740481534782822e-06,
+      "loss": 0.3877,
+      "step": 9530
+    },
+    {
+      "epoch": 1.570922788744571,
+      "grad_norm": 1.7993638515472412,
+      "learning_rate": 9.722710708484009e-06,
+      "loss": 0.3935,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5725695230645726,
+      "grad_norm": 1.487587571144104,
+      "learning_rate": 9.704940758502367e-06,
+      "loss": 0.4333,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5742162573845744,
+      "grad_norm": 1.5965005159378052,
+      "learning_rate": 9.687171740996262e-06,
+      "loss": 0.3824,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5758629917045759,
+      "grad_norm": 1.9359855651855469,
+      "learning_rate": 9.669403712121116e-06,
+      "loss": 0.4349,
+      "step": 9570
+    },
+    {
+      "epoch": 1.5775097260245774,
+      "grad_norm": 1.7077593803405762,
+      "learning_rate": 9.651636728029205e-06,
+      "loss": 0.4088,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5791564603445791,
+      "grad_norm": 1.3524852991104126,
+      "learning_rate": 9.633870844869526e-06,
+      "loss": 0.3674,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5808031946645809,
+      "grad_norm": 1.4619336128234863,
+      "learning_rate": 9.616106118787586e-06,
+      "loss": 0.4186,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5824499289845826,
+      "grad_norm": 1.5433001518249512,
+      "learning_rate": 9.59834260592524e-06,
+      "loss": 0.3836,
+      "step": 9610
+    },
+    {
+      "epoch": 1.584096663304584,
+      "grad_norm": 1.562225580215454,
+      "learning_rate": 9.580580362420505e-06,
+      "loss": 0.3979,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5857433976245856,
+      "grad_norm": 1.6424192190170288,
+      "learning_rate": 9.562819444407389e-06,
+      "loss": 0.3839,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5873901319445873,
+      "grad_norm": 1.7441072463989258,
+      "learning_rate": 9.545059908015713e-06,
+      "loss": 0.3896,
+      "step": 9640
+    },
+    {
+      "epoch": 1.589036866264589,
+      "grad_norm": 1.3649961948394775,
+      "learning_rate": 9.527301809370922e-06,
+      "loss": 0.3891,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5906836005845908,
+      "grad_norm": 2.137296676635742,
+      "learning_rate": 9.509545204593928e-06,
+      "loss": 0.3771,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5923303349045923,
+      "grad_norm": 3.6009469032287598,
+      "learning_rate": 9.491790149800916e-06,
+      "loss": 0.3759,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5939770692245938,
+      "grad_norm": 2.4487457275390625,
+      "learning_rate": 9.474036701103178e-06,
+      "loss": 0.4081,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5956238035445955,
+      "grad_norm": 2.1855509281158447,
+      "learning_rate": 9.456284914606924e-06,
+      "loss": 0.394,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5972705378645973,
+      "grad_norm": 1.6992751359939575,
+      "learning_rate": 9.438534846413115e-06,
+      "loss": 0.3946,
+      "step": 9700
+    },
+    {
+      "epoch": 1.598917272184599,
+      "grad_norm": 1.9765753746032715,
+      "learning_rate": 9.420786552617281e-06,
+      "loss": 0.3921,
+      "step": 9710
+    },
+    {
+      "epoch": 1.6005640065046005,
+      "grad_norm": 1.6092065572738647,
+      "learning_rate": 9.40304008930934e-06,
+      "loss": 0.3737,
+      "step": 9720
+    },
+    {
+      "epoch": 1.6022107408246022,
+      "grad_norm": 1.9483388662338257,
+      "learning_rate": 9.385295512573436e-06,
+      "loss": 0.4191,
+      "step": 9730
+    },
+    {
+      "epoch": 1.6038574751446038,
+      "grad_norm": 1.6057438850402832,
+      "learning_rate": 9.367552878487736e-06,
+      "loss": 0.4186,
+      "step": 9740
+    },
+    {
+      "epoch": 1.6055042094646055,
+      "grad_norm": 2.68620228767395,
+      "learning_rate": 9.34981224312428e-06,
+      "loss": 0.3937,
+      "step": 9750
+    },
+    {
+      "epoch": 1.6071509437846072,
+      "grad_norm": 2.7004518508911133,
+      "learning_rate": 9.332073662548785e-06,
+      "loss": 0.4057,
+      "step": 9760
+    },
+    {
+      "epoch": 1.608797678104609,
+      "grad_norm": 2.4203567504882812,
+      "learning_rate": 9.314337192820477e-06,
+      "loss": 0.3898,
+      "step": 9770
+    },
+    {
+      "epoch": 1.6104444124246104,
+      "grad_norm": 1.4761004447937012,
+      "learning_rate": 9.296602889991914e-06,
+      "loss": 0.3759,
+      "step": 9780
+    },
+    {
+      "epoch": 1.612091146744612,
+      "grad_norm": 1.881105899810791,
+      "learning_rate": 9.278870810108794e-06,
+      "loss": 0.3927,
+      "step": 9790
+    },
+    {
+      "epoch": 1.6137378810646137,
+      "grad_norm": 1.8509386777877808,
+      "learning_rate": 9.261141009209803e-06,
+      "loss": 0.4013,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 1.8641047477722168,
+      "learning_rate": 9.24341354332642e-06,
+      "loss": 0.3911,
+      "step": 9810
+    },
+    {
+      "epoch": 1.6170313497046171,
+      "grad_norm": 1.479648470878601,
+      "learning_rate": 9.225688468482743e-06,
+      "loss": 0.384,
+      "step": 9820
+    },
+    {
+      "epoch": 1.6186780840246187,
+      "grad_norm": 1.395717740058899,
+      "learning_rate": 9.207965840695314e-06,
+      "loss": 0.3862,
+      "step": 9830
+    },
+    {
+      "epoch": 1.6203248183446202,
+      "grad_norm": 1.989012598991394,
+      "learning_rate": 9.190245715972946e-06,
+      "loss": 0.405,
+      "step": 9840
+    },
+    {
+      "epoch": 1.621971552664622,
+      "grad_norm": 2.0621602535247803,
+      "learning_rate": 9.172528150316536e-06,
+      "loss": 0.4259,
+      "step": 9850
+    },
+    {
+      "epoch": 1.6236182869846236,
+      "grad_norm": 2.660885810852051,
+      "learning_rate": 9.154813199718893e-06,
+      "loss": 0.3823,
+      "step": 9860
+    },
+    {
+      "epoch": 1.6252650213046254,
+      "grad_norm": 2.435926914215088,
+      "learning_rate": 9.137100920164567e-06,
+      "loss": 0.3809,
+      "step": 9870
+    },
+    {
+      "epoch": 1.6269117556246269,
+      "grad_norm": 1.566701054573059,
+      "learning_rate": 9.119391367629665e-06,
+      "loss": 0.3987,
+      "step": 9880
+    },
+    {
+      "epoch": 1.6285584899446286,
+      "grad_norm": 1.947234034538269,
+      "learning_rate": 9.101684598081672e-06,
+      "loss": 0.3806,
+      "step": 9890
+    },
+    {
+      "epoch": 1.63020522426463,
+      "grad_norm": 1.8058415651321411,
+      "learning_rate": 9.083980667479286e-06,
+      "loss": 0.412,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6318519585846318,
+      "grad_norm": 1.4982244968414307,
+      "learning_rate": 9.066279631772222e-06,
+      "loss": 0.3814,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6334986929046336,
+      "grad_norm": 1.4314496517181396,
+      "learning_rate": 9.048581546901056e-06,
+      "loss": 0.3928,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6351454272246353,
+      "grad_norm": 1.6061030626296997,
+      "learning_rate": 9.03088646879703e-06,
+      "loss": 0.3695,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6367921615446368,
+      "grad_norm": 1.627259373664856,
+      "learning_rate": 9.013194453381892e-06,
+      "loss": 0.4073,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6384388958646383,
+      "grad_norm": 1.8757750988006592,
+      "learning_rate": 8.995505556567707e-06,
+      "loss": 0.3773,
+      "step": 9950
+    },
+    {
+      "epoch": 1.64008563018464,
+      "grad_norm": 2.141075372695923,
+      "learning_rate": 8.977819834256683e-06,
+      "loss": 0.3893,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6417323645046418,
+      "grad_norm": 1.8441202640533447,
+      "learning_rate": 8.960137342340997e-06,
+      "loss": 0.3921,
+      "step": 9970
+    },
+    {
+      "epoch": 1.6433790988246435,
+      "grad_norm": 1.523070216178894,
+      "learning_rate": 8.94245813670262e-06,
+      "loss": 0.3899,
+      "step": 9980
+    },
+    {
+      "epoch": 1.645025833144645,
+      "grad_norm": 1.7577736377716064,
+      "learning_rate": 8.924782273213137e-06,
+      "loss": 0.3742,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6466725674646465,
+      "grad_norm": 1.6085615158081055,
+      "learning_rate": 8.907109807733559e-06,
+      "loss": 0.3694,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6483193017846482,
+      "grad_norm": 2.142134666442871,
+      "learning_rate": 8.889440796114174e-06,
+      "loss": 0.3847,
+      "step": 10010
+    },
+    {
+      "epoch": 1.64996603610465,
+      "grad_norm": 1.5217559337615967,
+      "learning_rate": 8.871775294194346e-06,
+      "loss": 0.3493,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6516127704246517,
+      "grad_norm": 2.990302085876465,
+      "learning_rate": 8.854113357802353e-06,
+      "loss": 0.367,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6532595047446532,
+      "grad_norm": 1.62135648727417,
+      "learning_rate": 8.836455042755197e-06,
+      "loss": 0.3797,
+      "step": 10040
+    },
+    {
+      "epoch": 1.654906239064655,
+      "grad_norm": 1.5902869701385498,
+      "learning_rate": 8.818800404858441e-06,
+      "loss": 0.3803,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6565529733846565,
+      "grad_norm": 1.5055309534072876,
+      "learning_rate": 8.801149499906032e-06,
+      "loss": 0.379,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6581997077046582,
+      "grad_norm": 1.5412495136260986,
+      "learning_rate": 8.78350238368011e-06,
+      "loss": 0.3718,
+      "step": 10070
+    },
+    {
+      "epoch": 1.65984644202466,
+      "grad_norm": 1.5508426427841187,
+      "learning_rate": 8.765859111950842e-06,
+      "loss": 0.3957,
+      "step": 10080
+    },
+    {
+      "epoch": 1.6614931763446616,
+      "grad_norm": 1.5197268724441528,
+      "learning_rate": 8.74821974047625e-06,
+      "loss": 0.3655,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6631399106646632,
+      "grad_norm": 2.0279932022094727,
+      "learning_rate": 8.730584325002031e-06,
+      "loss": 0.3608,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6647866449846647,
+      "grad_norm": 1.2136750221252441,
+      "learning_rate": 8.712952921261377e-06,
+      "loss": 0.3698,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6664333793046664,
+      "grad_norm": 1.5392953157424927,
+      "learning_rate": 8.695325584974802e-06,
+      "loss": 0.385,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6680801136246681,
+      "grad_norm": 1.2899067401885986,
+      "learning_rate": 8.677702371849965e-06,
+      "loss": 0.3637,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6697268479446699,
+      "grad_norm": 1.5220967531204224,
+      "learning_rate": 8.6600833375815e-06,
+      "loss": 0.3788,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6713735822646714,
+      "grad_norm": 2.6489806175231934,
+      "learning_rate": 8.642468537850822e-06,
+      "loss": 0.3889,
+      "step": 10150
+    },
+    {
+      "epoch": 1.6730203165846729,
+      "grad_norm": 2.6164515018463135,
+      "learning_rate": 8.624858028325976e-06,
+      "loss": 0.384,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6746670509046746,
+      "grad_norm": 2.1148691177368164,
+      "learning_rate": 8.607251864661443e-06,
+      "loss": 0.3684,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6763137852246763,
+      "grad_norm": 1.1956931352615356,
+      "learning_rate": 8.589650102497973e-06,
+      "loss": 0.3512,
+      "step": 10180
+    },
+    {
+      "epoch": 1.677960519544678,
+      "grad_norm": 1.5633714199066162,
+      "learning_rate": 8.572052797462403e-06,
+      "loss": 0.3658,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6796072538646796,
+      "grad_norm": 1.9797472953796387,
+      "learning_rate": 8.554460005167483e-06,
+      "loss": 0.3643,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6812539881846813,
+      "grad_norm": 1.8225812911987305,
+      "learning_rate": 8.536871781211711e-06,
+      "loss": 0.3615,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6829007225046828,
+      "grad_norm": 2.6055593490600586,
+      "learning_rate": 8.51928818117913e-06,
+      "loss": 0.3932,
+      "step": 10220
+    },
+    {
+      "epoch": 1.6845474568246845,
+      "grad_norm": 2.2658164501190186,
+      "learning_rate": 8.501709260639187e-06,
+      "loss": 0.3912,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6861941911446863,
+      "grad_norm": 2.0631959438323975,
+      "learning_rate": 8.484135075146528e-06,
+      "loss": 0.3777,
+      "step": 10240
+    },
+    {
+      "epoch": 1.687840925464688,
+      "grad_norm": 2.3120977878570557,
+      "learning_rate": 8.466565680240847e-06,
+      "loss": 0.3918,
+      "step": 10250
+    },
+    {
+      "epoch": 1.6894876597846895,
+      "grad_norm": 1.802088737487793,
+      "learning_rate": 8.449001131446687e-06,
+      "loss": 0.3785,
+      "step": 10260
+    },
+    {
+      "epoch": 1.691134394104691,
+      "grad_norm": 2.1044161319732666,
+      "learning_rate": 8.431441484273282e-06,
+      "loss": 0.3679,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6927811284246927,
+      "grad_norm": 2.071812868118286,
+      "learning_rate": 8.413886794214379e-06,
+      "loss": 0.3771,
+      "step": 10280
+    },
+    {
+      "epoch": 1.6944278627446945,
+      "grad_norm": 2.1053099632263184,
+      "learning_rate": 8.396337116748046e-06,
+      "loss": 0.3496,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6960745970646962,
+      "grad_norm": 1.8317806720733643,
+      "learning_rate": 8.37879250733652e-06,
+      "loss": 0.3549,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6977213313846977,
+      "grad_norm": 2.297271728515625,
+      "learning_rate": 8.361253021426019e-06,
+      "loss": 0.3801,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6993680657046992,
+      "grad_norm": 2.0474696159362793,
+      "learning_rate": 8.343718714446572e-06,
+      "loss": 0.3814,
+      "step": 10320
+    },
+    {
+      "epoch": 1.701014800024701,
+      "grad_norm": 1.3918821811676025,
+      "learning_rate": 8.326189641811835e-06,
+      "loss": 0.3513,
+      "step": 10330
+    },
+    {
+      "epoch": 1.7026615343447027,
+      "grad_norm": 1.5274895429611206,
+      "learning_rate": 8.308665858918928e-06,
+      "loss": 0.3358,
+      "step": 10340
+    },
+    {
+      "epoch": 1.7043082686647044,
+      "grad_norm": 2.130195140838623,
+      "learning_rate": 8.291147421148255e-06,
+      "loss": 0.3682,
+      "step": 10350
+    },
+    {
+      "epoch": 1.705955002984706,
+      "grad_norm": 1.9873136281967163,
+      "learning_rate": 8.273634383863315e-06,
+      "loss": 0.367,
+      "step": 10360
+    },
+    {
+      "epoch": 1.7076017373047077,
+      "grad_norm": 1.433440923690796,
+      "learning_rate": 8.256126802410554e-06,
+      "loss": 0.3572,
+      "step": 10370
+    },
+    {
+      "epoch": 1.7092484716247092,
+      "grad_norm": 1.550261378288269,
+      "learning_rate": 8.238624732119169e-06,
+      "loss": 0.383,
+      "step": 10380
+    },
+    {
+      "epoch": 1.710895205944711,
+      "grad_norm": 1.8397653102874756,
+      "learning_rate": 8.221128228300941e-06,
+      "loss": 0.3497,
+      "step": 10390
+    },
+    {
+      "epoch": 1.7125419402647126,
+      "grad_norm": 1.6404129266738892,
+      "learning_rate": 8.203637346250062e-06,
+      "loss": 0.3514,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7141886745847141,
+      "grad_norm": 2.078935146331787,
+      "learning_rate": 8.186152141242957e-06,
+      "loss": 0.3694,
+      "step": 10410
+    },
+    {
+      "epoch": 1.7158354089047159,
+      "grad_norm": 1.4735045433044434,
+      "learning_rate": 8.16867266853811e-06,
+      "loss": 0.3602,
+      "step": 10420
+    },
+    {
+      "epoch": 1.7174821432247174,
+      "grad_norm": 1.6575242280960083,
+      "learning_rate": 8.15119898337588e-06,
+      "loss": 0.3569,
+      "step": 10430
+    },
+    {
+      "epoch": 1.719128877544719,
+      "grad_norm": 1.4052168130874634,
+      "learning_rate": 8.133731140978347e-06,
+      "loss": 0.3703,
+      "step": 10440
+    },
+    {
+      "epoch": 1.7207756118647208,
+      "grad_norm": 2.0821497440338135,
+      "learning_rate": 8.116269196549124e-06,
+      "loss": 0.3442,
+      "step": 10450
+    },
+    {
+      "epoch": 1.7224223461847226,
+      "grad_norm": 1.4449158906936646,
+      "learning_rate": 8.098813205273183e-06,
+      "loss": 0.3179,
+      "step": 10460
+    },
+    {
+      "epoch": 1.724069080504724,
+      "grad_norm": 1.4337159395217896,
+      "learning_rate": 8.081363222316681e-06,
+      "loss": 0.3608,
+      "step": 10470
+    },
+    {
+      "epoch": 1.7257158148247256,
+      "grad_norm": 1.587299108505249,
+      "learning_rate": 8.063919302826787e-06,
+      "loss": 0.366,
+      "step": 10480
+    },
+    {
+      "epoch": 1.7273625491447273,
+      "grad_norm": 1.4321982860565186,
+      "learning_rate": 8.046481501931516e-06,
+      "loss": 0.36,
+      "step": 10490
+    },
+    {
+      "epoch": 1.729009283464729,
+      "grad_norm": 1.3218423128128052,
+      "learning_rate": 8.02904987473953e-06,
+      "loss": 0.3678,
+      "step": 10500
+    },
+    {
+      "epoch": 1.7306560177847308,
+      "grad_norm": 1.4841070175170898,
+      "learning_rate": 8.011624476339993e-06,
+      "loss": 0.3565,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7323027521047323,
+      "grad_norm": 1.6882789134979248,
+      "learning_rate": 7.99420536180238e-06,
+      "loss": 0.3598,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7339494864247338,
+      "grad_norm": 1.6460531949996948,
+      "learning_rate": 7.976792586176311e-06,
+      "loss": 0.3601,
+      "step": 10530
+    },
+    {
+      "epoch": 1.7355962207447355,
+      "grad_norm": 2.165046215057373,
+      "learning_rate": 7.959386204491365e-06,
+      "loss": 0.3529,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7372429550647372,
+      "grad_norm": 2.083591938018799,
+      "learning_rate": 7.941986271756926e-06,
+      "loss": 0.3664,
+      "step": 10550
+    },
+    {
+      "epoch": 1.738889689384739,
+      "grad_norm": 1.4335622787475586,
+      "learning_rate": 7.924592842961985e-06,
+      "loss": 0.3742,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7405364237047405,
+      "grad_norm": 1.9920486211776733,
+      "learning_rate": 7.907205973074987e-06,
+      "loss": 0.3652,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7421831580247422,
+      "grad_norm": 1.6265625953674316,
+      "learning_rate": 7.889825717043643e-06,
+      "loss": 0.3714,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7438298923447437,
+      "grad_norm": 1.6098805665969849,
+      "learning_rate": 7.872452129794765e-06,
+      "loss": 0.3434,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7454766266647455,
+      "grad_norm": 1.5913910865783691,
+      "learning_rate": 7.855085266234093e-06,
+      "loss": 0.3601,
+      "step": 10600
+    },
+    {
+      "epoch": 1.7471233609847472,
+      "grad_norm": 1.3813248872756958,
+      "learning_rate": 7.837725181246116e-06,
+      "loss": 0.3658,
+      "step": 10610
+    },
+    {
+      "epoch": 1.748770095304749,
+      "grad_norm": 1.6209338903427124,
+      "learning_rate": 7.820371929693894e-06,
+      "loss": 0.3515,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7504168296247504,
+      "grad_norm": 1.6621341705322266,
+      "learning_rate": 7.803025566418904e-06,
+      "loss": 0.364,
+      "step": 10630
+    },
+    {
+      "epoch": 1.752063563944752,
+      "grad_norm": 1.360546588897705,
+      "learning_rate": 7.785686146240844e-06,
+      "loss": 0.3582,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7537102982647537,
+      "grad_norm": 2.2365732192993164,
+      "learning_rate": 7.76835372395747e-06,
+      "loss": 0.3538,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7553570325847554,
+      "grad_norm": 1.673392415046692,
+      "learning_rate": 7.751028354344432e-06,
+      "loss": 0.3661,
+      "step": 10660
+    },
+    {
+      "epoch": 1.7570037669047571,
+      "grad_norm": 1.4482479095458984,
+      "learning_rate": 7.733710092155076e-06,
+      "loss": 0.3468,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7586505012247586,
+      "grad_norm": 1.5075653791427612,
+      "learning_rate": 7.716398992120302e-06,
+      "loss": 0.3588,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7602972355447601,
+      "grad_norm": 1.881640911102295,
+      "learning_rate": 7.699095108948365e-06,
+      "loss": 0.3538,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7619439698647619,
+      "grad_norm": 1.933388590812683,
+      "learning_rate": 7.681798497324717e-06,
+      "loss": 0.3527,
+      "step": 10700
+    },
+    {
+      "epoch": 1.7635907041847636,
+      "grad_norm": 1.514220118522644,
+      "learning_rate": 7.664509211911833e-06,
+      "loss": 0.3679,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7652374385047653,
+      "grad_norm": 1.9936243295669556,
+      "learning_rate": 7.647227307349024e-06,
+      "loss": 0.3808,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7668841728247668,
+      "grad_norm": 1.567679762840271,
+      "learning_rate": 7.629952838252287e-06,
+      "loss": 0.3515,
+      "step": 10730
+    },
+    {
+      "epoch": 1.7685309071447686,
+      "grad_norm": 1.4825366735458374,
+      "learning_rate": 7.612685859214113e-06,
+      "loss": 0.3677,
+      "step": 10740
+    },
+    {
+      "epoch": 1.77017764146477,
+      "grad_norm": 1.8481963872909546,
+      "learning_rate": 7.59542642480333e-06,
+      "loss": 0.3642,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7718243757847718,
+      "grad_norm": 1.9915103912353516,
+      "learning_rate": 7.578174589564911e-06,
+      "loss": 0.3512,
+      "step": 10760
+    },
+    {
+      "epoch": 1.7734711101047735,
+      "grad_norm": 2.2940824031829834,
+      "learning_rate": 7.560930408019823e-06,
+      "loss": 0.3713,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7751178444247753,
+      "grad_norm": 1.6668184995651245,
+      "learning_rate": 7.543693934664846e-06,
+      "loss": 0.3782,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7767645787447768,
+      "grad_norm": 1.4784590005874634,
+      "learning_rate": 7.52646522397239e-06,
+      "loss": 0.3632,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7784113130647783,
+      "grad_norm": 1.6654629707336426,
+      "learning_rate": 7.5092443303903404e-06,
+      "loss": 0.351,
+      "step": 10800
+    },
+    {
+      "epoch": 1.78005804738478,
+      "grad_norm": 1.913283109664917,
+      "learning_rate": 7.492031308341879e-06,
+      "loss": 0.3557,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7817047817047817,
+      "grad_norm": 1.9951850175857544,
+      "learning_rate": 7.474826212225305e-06,
+      "loss": 0.3358,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7833515160247835,
+      "grad_norm": 1.5256726741790771,
+      "learning_rate": 7.457629096413874e-06,
+      "loss": 0.3418,
+      "step": 10830
+    },
+    {
+      "epoch": 1.784998250344785,
+      "grad_norm": 1.617591142654419,
+      "learning_rate": 7.440440015255625e-06,
+      "loss": 0.3448,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7866449846647865,
+      "grad_norm": 1.3453478813171387,
+      "learning_rate": 7.423259023073197e-06,
+      "loss": 0.3479,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7882917189847882,
+      "grad_norm": 1.643965721130371,
+      "learning_rate": 7.406086174163665e-06,
+      "loss": 0.3634,
+      "step": 10860
+    },
+    {
+      "epoch": 1.78993845330479,
+      "grad_norm": 1.6207489967346191,
+      "learning_rate": 7.388921522798376e-06,
+      "loss": 0.3419,
+      "step": 10870
+    },
+    {
+      "epoch": 1.7915851876247917,
+      "grad_norm": 2.043084144592285,
+      "learning_rate": 7.371765123222767e-06,
+      "loss": 0.348,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7932319219447932,
+      "grad_norm": 1.35907781124115,
+      "learning_rate": 7.354617029656198e-06,
+      "loss": 0.3344,
+      "step": 10890
+    },
+    {
+      "epoch": 1.794878656264795,
+      "grad_norm": 1.5821566581726074,
+      "learning_rate": 7.337477296291778e-06,
+      "loss": 0.3273,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7965253905847964,
+      "grad_norm": 1.6671669483184814,
+      "learning_rate": 7.3203459772961924e-06,
+      "loss": 0.3508,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7981721249047982,
+      "grad_norm": 1.5016417503356934,
+      "learning_rate": 7.303223126809546e-06,
+      "loss": 0.3348,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7998188592248,
+      "grad_norm": 1.4615623950958252,
+      "learning_rate": 7.286108798945162e-06,
+      "loss": 0.3501,
+      "step": 10930
+    },
+    {
+      "epoch": 1.8014655935448016,
+      "grad_norm": 1.4260656833648682,
+      "learning_rate": 7.269003047789446e-06,
+      "loss": 0.338,
+      "step": 10940
+    },
+    {
+      "epoch": 1.8031123278648031,
+      "grad_norm": 2.318674087524414,
+      "learning_rate": 7.251905927401691e-06,
+      "loss": 0.3422,
+      "step": 10950
+    },
+    {
+      "epoch": 1.8047590621848046,
+      "grad_norm": 1.3945016860961914,
+      "learning_rate": 7.234817491813917e-06,
+      "loss": 0.3291,
+      "step": 10960
+    },
+    {
+      "epoch": 1.8064057965048064,
+      "grad_norm": 2.036652088165283,
+      "learning_rate": 7.217737795030695e-06,
+      "loss": 0.3412,
+      "step": 10970
+    },
+    {
+      "epoch": 1.808052530824808,
+      "grad_norm": 1.9263806343078613,
+      "learning_rate": 7.200666891028983e-06,
+      "loss": 0.3341,
+      "step": 10980
+    },
+    {
+      "epoch": 1.8096992651448098,
+      "grad_norm": 1.7770569324493408,
+      "learning_rate": 7.183604833757949e-06,
+      "loss": 0.3375,
+      "step": 10990
+    },
+    {
+      "epoch": 1.8113459994648113,
+      "grad_norm": 1.8444048166275024,
+      "learning_rate": 7.166551677138794e-06,
+      "loss": 0.3298,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8129927337848128,
+      "grad_norm": 1.5634081363677979,
+      "learning_rate": 7.149507475064606e-06,
+      "loss": 0.337,
+      "step": 11010
+    },
+    {
+      "epoch": 1.8146394681048146,
+      "grad_norm": 2.272538900375366,
+      "learning_rate": 7.13247228140016e-06,
+      "loss": 0.3514,
+      "step": 11020
+    },
+    {
+      "epoch": 1.8162862024248163,
+      "grad_norm": 1.4847793579101562,
+      "learning_rate": 7.115446149981774e-06,
+      "loss": 0.3442,
+      "step": 11030
+    },
+    {
+      "epoch": 1.817932936744818,
+      "grad_norm": 1.7815018892288208,
+      "learning_rate": 7.098429134617117e-06,
+      "loss": 0.3527,
+      "step": 11040
+    },
+    {
+      "epoch": 1.8195796710648195,
+      "grad_norm": 1.8198860883712769,
+      "learning_rate": 7.081421289085053e-06,
+      "loss": 0.3289,
+      "step": 11050
+    },
+    {
+      "epoch": 1.8212264053848213,
+      "grad_norm": 1.5678714513778687,
+      "learning_rate": 7.06442266713547e-06,
+      "loss": 0.3212,
+      "step": 11060
+    },
+    {
+      "epoch": 1.8228731397048228,
+      "grad_norm": 1.6661272048950195,
+      "learning_rate": 7.047433322489094e-06,
+      "loss": 0.3511,
+      "step": 11070
+    },
+    {
+      "epoch": 1.8245198740248245,
+      "grad_norm": 1.59157395362854,
+      "learning_rate": 7.030453308837344e-06,
+      "loss": 0.3392,
+      "step": 11080
+    },
+    {
+      "epoch": 1.8261666083448262,
+      "grad_norm": 1.9493590593338013,
+      "learning_rate": 7.013482679842145e-06,
+      "loss": 0.3496,
+      "step": 11090
+    },
+    {
+      "epoch": 1.827813342664828,
+      "grad_norm": 2.397372245788574,
+      "learning_rate": 6.996521489135768e-06,
+      "loss": 0.3428,
+      "step": 11100
+    },
+    {
+      "epoch": 1.8294600769848295,
+      "grad_norm": 2.058790445327759,
+      "learning_rate": 6.979569790320653e-06,
+      "loss": 0.3099,
+      "step": 11110
+    },
+    {
+      "epoch": 1.831106811304831,
+      "grad_norm": 1.6316676139831543,
+      "learning_rate": 6.962627636969241e-06,
+      "loss": 0.3536,
+      "step": 11120
+    },
+    {
+      "epoch": 1.8327535456248327,
+      "grad_norm": 1.8087751865386963,
+      "learning_rate": 6.945695082623816e-06,
+      "loss": 0.3548,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8344002799448345,
+      "grad_norm": 1.6913039684295654,
+      "learning_rate": 6.928772180796308e-06,
+      "loss": 0.3325,
+      "step": 11140
+    },
+    {
+      "epoch": 1.8360470142648362,
+      "grad_norm": 1.5240042209625244,
+      "learning_rate": 6.911858984968158e-06,
+      "loss": 0.3381,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8376937485848377,
+      "grad_norm": 1.340308427810669,
+      "learning_rate": 6.894955548590128e-06,
+      "loss": 0.3293,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8393404829048392,
+      "grad_norm": 1.640592336654663,
+      "learning_rate": 6.878061925082138e-06,
+      "loss": 0.3394,
+      "step": 11170
+    },
+    {
+      "epoch": 1.840987217224841,
+      "grad_norm": 1.7119625806808472,
+      "learning_rate": 6.861178167833096e-06,
+      "loss": 0.3495,
+      "step": 11180
+    },
+    {
+      "epoch": 1.8426339515448427,
+      "grad_norm": 1.365484356880188,
+      "learning_rate": 6.844304330200728e-06,
+      "loss": 0.3353,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8442806858648444,
+      "grad_norm": 2.2884132862091064,
+      "learning_rate": 6.827440465511414e-06,
+      "loss": 0.331,
+      "step": 11200
+    },
+    {
+      "epoch": 1.845927420184846,
+      "grad_norm": 1.5407003164291382,
+      "learning_rate": 6.810586627060019e-06,
+      "loss": 0.3182,
+      "step": 11210
+    },
+    {
+      "epoch": 1.8475741545048474,
+      "grad_norm": 1.7855136394500732,
+      "learning_rate": 6.793742868109709e-06,
+      "loss": 0.3248,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8492208888248491,
+      "grad_norm": 1.5685489177703857,
+      "learning_rate": 6.776909241891809e-06,
+      "loss": 0.3332,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8508676231448509,
+      "grad_norm": 2.068676233291626,
+      "learning_rate": 6.76008580160562e-06,
+      "loss": 0.3421,
+      "step": 11240
+    },
+    {
+      "epoch": 1.8525143574648526,
+      "grad_norm": 1.9478691816329956,
+      "learning_rate": 6.743272600418246e-06,
+      "loss": 0.3194,
+      "step": 11250
+    },
+    {
+      "epoch": 1.854161091784854,
+      "grad_norm": 1.7436619997024536,
+      "learning_rate": 6.726469691464439e-06,
+      "loss": 0.3261,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8558078261048558,
+      "grad_norm": 1.7236602306365967,
+      "learning_rate": 6.70967712784642e-06,
+      "loss": 0.3115,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8574545604248573,
+      "grad_norm": 1.4117182493209839,
+      "learning_rate": 6.692894962633722e-06,
+      "loss": 0.3181,
+      "step": 11280
+    },
+    {
+      "epoch": 1.859101294744859,
+      "grad_norm": 1.5423839092254639,
+      "learning_rate": 6.6761232488630046e-06,
+      "loss": 0.3137,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8607480290648608,
+      "grad_norm": 1.5166431665420532,
+      "learning_rate": 6.659362039537907e-06,
+      "loss": 0.3257,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8623947633848625,
+      "grad_norm": 1.261831521987915,
+      "learning_rate": 6.6426113876288665e-06,
+      "loss": 0.3134,
+      "step": 11310
+    },
+    {
+      "epoch": 1.864041497704864,
+      "grad_norm": 1.5698233842849731,
+      "learning_rate": 6.6258713460729604e-06,
+      "loss": 0.3269,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8656882320248656,
+      "grad_norm": 1.4800500869750977,
+      "learning_rate": 6.609141967773733e-06,
+      "loss": 0.3393,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8673349663448673,
+      "grad_norm": 1.6218483448028564,
+      "learning_rate": 6.592423305601025e-06,
+      "loss": 0.3542,
+      "step": 11340
+    },
+    {
+      "epoch": 1.868981700664869,
+      "grad_norm": 1.5717345476150513,
+      "learning_rate": 6.5757154123908185e-06,
+      "loss": 0.3347,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8706284349848707,
+      "grad_norm": 2.0220766067504883,
+      "learning_rate": 6.559018340945051e-06,
+      "loss": 0.3264,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8722751693048723,
+      "grad_norm": 1.756557822227478,
+      "learning_rate": 6.542332144031471e-06,
+      "loss": 0.3721,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8739219036248738,
+      "grad_norm": 1.273116946220398,
+      "learning_rate": 6.525656874383456e-06,
+      "loss": 0.3272,
+      "step": 11380
+    },
+    {
+      "epoch": 1.8755686379448755,
+      "grad_norm": 1.276261568069458,
+      "learning_rate": 6.508992584699849e-06,
+      "loss": 0.3009,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8772153722648772,
+      "grad_norm": 1.6424288749694824,
+      "learning_rate": 6.492339327644797e-06,
+      "loss": 0.3341,
+      "step": 11400
+    },
+    {
+      "epoch": 1.878862106584879,
+      "grad_norm": 1.6955491304397583,
+      "learning_rate": 6.4756971558475755e-06,
+      "loss": 0.3335,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8805088409048805,
+      "grad_norm": 1.661267638206482,
+      "learning_rate": 6.459066121902433e-06,
+      "loss": 0.3396,
+      "step": 11420
+    },
+    {
+      "epoch": 1.8821555752248822,
+      "grad_norm": 1.469783067703247,
+      "learning_rate": 6.442446278368411e-06,
+      "loss": 0.2996,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8838023095448837,
+      "grad_norm": 1.665208101272583,
+      "learning_rate": 6.425837677769191e-06,
+      "loss": 0.3379,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8854490438648854,
+      "grad_norm": 1.6089602708816528,
+      "learning_rate": 6.409240372592926e-06,
+      "loss": 0.3355,
+      "step": 11450
+    },
+    {
+      "epoch": 1.8870957781848872,
+      "grad_norm": 1.9932987689971924,
+      "learning_rate": 6.392654415292068e-06,
+      "loss": 0.3335,
+      "step": 11460
+    },
+    {
+      "epoch": 1.888742512504889,
+      "grad_norm": 1.4461263418197632,
+      "learning_rate": 6.3760798582832065e-06,
+      "loss": 0.327,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8903892468248904,
+      "grad_norm": 1.5308934450149536,
+      "learning_rate": 6.359516753946905e-06,
+      "loss": 0.3283,
+      "step": 11480
+    },
+    {
+      "epoch": 1.892035981144892,
+      "grad_norm": 1.4926317930221558,
+      "learning_rate": 6.342965154627534e-06,
+      "loss": 0.3308,
+      "step": 11490
+    },
+    {
+      "epoch": 1.8936827154648936,
+      "grad_norm": 1.491185188293457,
+      "learning_rate": 6.326425112633097e-06,
+      "loss": 0.3179,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8953294497848954,
+      "grad_norm": 1.4022414684295654,
+      "learning_rate": 6.309896680235082e-06,
+      "loss": 0.331,
+      "step": 11510
+    },
+    {
+      "epoch": 1.896976184104897,
+      "grad_norm": 1.5799373388290405,
+      "learning_rate": 6.293379909668282e-06,
+      "loss": 0.3099,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8986229184248986,
+      "grad_norm": 1.7698333263397217,
+      "learning_rate": 6.276874853130639e-06,
+      "loss": 0.3459,
+      "step": 11530
+    },
+    {
+      "epoch": 1.9002696527449001,
+      "grad_norm": 1.4904412031173706,
+      "learning_rate": 6.2603815627830685e-06,
+      "loss": 0.3258,
+      "step": 11540
+    },
+    {
+      "epoch": 1.9019163870649018,
+      "grad_norm": 1.5331462621688843,
+      "learning_rate": 6.2439000907493105e-06,
+      "loss": 0.3347,
+      "step": 11550
+    },
+    {
+      "epoch": 1.9035631213849036,
+      "grad_norm": 1.8723499774932861,
+      "learning_rate": 6.227430489115751e-06,
+      "loss": 0.3369,
+      "step": 11560
+    },
+    {
+      "epoch": 1.9052098557049053,
+      "grad_norm": 1.607606291770935,
+      "learning_rate": 6.210972809931257e-06,
+      "loss": 0.3183,
+      "step": 11570
+    },
+    {
+      "epoch": 1.9068565900249068,
+      "grad_norm": 1.2479761838912964,
+      "learning_rate": 6.194527105207024e-06,
+      "loss": 0.3357,
+      "step": 11580
+    },
+    {
+      "epoch": 1.9085033243449085,
+      "grad_norm": 1.7267532348632812,
+      "learning_rate": 6.178093426916403e-06,
+      "loss": 0.3089,
+      "step": 11590
+    },
+    {
+      "epoch": 1.91015005866491,
+      "grad_norm": 1.616453766822815,
+      "learning_rate": 6.161671826994739e-06,
+      "loss": 0.3416,
+      "step": 11600
+    },
+    {
+      "epoch": 1.9117967929849118,
+      "grad_norm": 1.7664293050765991,
+      "learning_rate": 6.1452623573392e-06,
+      "loss": 0.3323,
+      "step": 11610
+    },
+    {
+      "epoch": 1.9134435273049135,
+      "grad_norm": 1.3341567516326904,
+      "learning_rate": 6.128865069808625e-06,
+      "loss": 0.3194,
+      "step": 11620
+    },
+    {
+      "epoch": 1.9150902616249152,
+      "grad_norm": 1.6267433166503906,
+      "learning_rate": 6.112480016223352e-06,
+      "loss": 0.321,
+      "step": 11630
+    },
+    {
+      "epoch": 1.9167369959449168,
+      "grad_norm": 1.7511682510375977,
+      "learning_rate": 6.0961072483650526e-06,
+      "loss": 0.3216,
+      "step": 11640
+    },
+    {
+      "epoch": 1.9183837302649183,
+      "grad_norm": 1.8487275838851929,
+      "learning_rate": 6.0797468179765785e-06,
+      "loss": 0.3226,
+      "step": 11650
+    },
+    {
+      "epoch": 1.92003046458492,
+      "grad_norm": 1.737802267074585,
+      "learning_rate": 6.063398776761785e-06,
+      "loss": 0.3123,
+      "step": 11660
+    },
+    {
+      "epoch": 1.9216771989049217,
+      "grad_norm": 1.4471075534820557,
+      "learning_rate": 6.047063176385378e-06,
+      "loss": 0.2977,
+      "step": 11670
+    },
+    {
+      "epoch": 1.9233239332249235,
+      "grad_norm": 1.8854256868362427,
+      "learning_rate": 6.030740068472745e-06,
+      "loss": 0.3323,
+      "step": 11680
+    },
+    {
+      "epoch": 1.924970667544925,
+      "grad_norm": 1.4351447820663452,
+      "learning_rate": 6.014429504609796e-06,
+      "loss": 0.3164,
+      "step": 11690
+    },
+    {
+      "epoch": 1.9266174018649265,
+      "grad_norm": 1.6448378562927246,
+      "learning_rate": 5.998131536342792e-06,
+      "loss": 0.3268,
+      "step": 11700
+    },
+    {
+      "epoch": 1.9282641361849282,
+      "grad_norm": 1.5195516347885132,
+      "learning_rate": 5.981846215178191e-06,
+      "loss": 0.3031,
+      "step": 11710
+    },
+    {
+      "epoch": 1.92991087050493,
+      "grad_norm": 1.6448338031768799,
+      "learning_rate": 5.965573592582488e-06,
+      "loss": 0.3234,
+      "step": 11720
+    },
+    {
+      "epoch": 1.9315576048249317,
+      "grad_norm": 1.4545010328292847,
+      "learning_rate": 5.9493137199820376e-06,
+      "loss": 0.3354,
+      "step": 11730
+    },
+    {
+      "epoch": 1.9332043391449332,
+      "grad_norm": 1.2834254503250122,
+      "learning_rate": 5.933066648762907e-06,
+      "loss": 0.3229,
+      "step": 11740
+    },
+    {
+      "epoch": 1.934851073464935,
+      "grad_norm": 1.7601953744888306,
+      "learning_rate": 5.916832430270705e-06,
+      "loss": 0.3112,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9364978077849364,
+      "grad_norm": 1.315958857536316,
+      "learning_rate": 5.900611115810423e-06,
+      "loss": 0.3119,
+      "step": 11760
+    },
+    {
+      "epoch": 1.9381445421049381,
+      "grad_norm": 2.057389259338379,
+      "learning_rate": 5.884402756646273e-06,
+      "loss": 0.3206,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9397912764249399,
+      "grad_norm": 2.368434190750122,
+      "learning_rate": 5.868207404001518e-06,
+      "loss": 0.3211,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9414380107449416,
+      "grad_norm": 1.6587117910385132,
+      "learning_rate": 5.852025109058321e-06,
+      "loss": 0.3078,
+      "step": 11790
+    },
+    {
+      "epoch": 1.943084745064943,
+      "grad_norm": 2.1465649604797363,
+      "learning_rate": 5.835855922957583e-06,
+      "loss": 0.3168,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9447314793849446,
+      "grad_norm": 2.077082395553589,
+      "learning_rate": 5.819699896798765e-06,
+      "loss": 0.3214,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9463782137049463,
+      "grad_norm": 1.8317384719848633,
+      "learning_rate": 5.803557081639757e-06,
+      "loss": 0.3085,
+      "step": 11820
+    },
+    {
+      "epoch": 1.948024948024948,
+      "grad_norm": 1.5140388011932373,
+      "learning_rate": 5.787427528496676e-06,
+      "loss": 0.3213,
+      "step": 11830
+    },
+    {
+      "epoch": 1.9496716823449498,
+      "grad_norm": 1.7246615886688232,
+      "learning_rate": 5.771311288343748e-06,
+      "loss": 0.3098,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9513184166649513,
+      "grad_norm": 2.0293779373168945,
+      "learning_rate": 5.755208412113116e-06,
+      "loss": 0.3206,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9529651509849528,
+      "grad_norm": 1.6586174964904785,
+      "learning_rate": 5.739118950694684e-06,
+      "loss": 0.3122,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9546118853049546,
+      "grad_norm": 1.7157447338104248,
+      "learning_rate": 5.723042954935968e-06,
+      "loss": 0.2913,
+      "step": 11870
+    },
+    {
+      "epoch": 1.9562586196249563,
+      "grad_norm": 1.383748173713684,
+      "learning_rate": 5.7069804756419326e-06,
+      "loss": 0.3248,
+      "step": 11880
+    },
+    {
+      "epoch": 1.957905353944958,
+      "grad_norm": 2.3896443843841553,
+      "learning_rate": 5.690931563574813e-06,
+      "loss": 0.3112,
+      "step": 11890
+    },
+    {
+      "epoch": 1.9595520882649595,
+      "grad_norm": 1.9655721187591553,
+      "learning_rate": 5.6748962694539855e-06,
+      "loss": 0.3181,
+      "step": 11900
+    },
+    {
+      "epoch": 1.961198822584961,
+      "grad_norm": 1.6184086799621582,
+      "learning_rate": 5.6588746439557706e-06,
+      "loss": 0.3272,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9628455569049628,
+      "grad_norm": 1.5995012521743774,
+      "learning_rate": 5.642866737713311e-06,
+      "loss": 0.3039,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9644922912249645,
+      "grad_norm": 1.8000292778015137,
+      "learning_rate": 5.6268726013163764e-06,
+      "loss": 0.3061,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9661390255449662,
+      "grad_norm": 1.651097059249878,
+      "learning_rate": 5.610892285311229e-06,
+      "loss": 0.3,
+      "step": 11940
+    },
+    {
+      "epoch": 1.9677857598649677,
+      "grad_norm": 1.8784924745559692,
+      "learning_rate": 5.5949258402004446e-06,
+      "loss": 0.3261,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9694324941849695,
+      "grad_norm": 1.6601523160934448,
+      "learning_rate": 5.578973316442779e-06,
+      "loss": 0.3081,
+      "step": 11960
+    },
+    {
+      "epoch": 1.971079228504971,
+      "grad_norm": 1.904227614402771,
+      "learning_rate": 5.563034764452976e-06,
+      "loss": 0.3211,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9727259628249727,
+      "grad_norm": 1.6444785594940186,
+      "learning_rate": 5.5471102346016385e-06,
+      "loss": 0.3315,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9743726971449744,
+      "grad_norm": 1.8297940492630005,
+      "learning_rate": 5.531199777215044e-06,
+      "loss": 0.2992,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9760194314649762,
+      "grad_norm": 1.717757225036621,
+      "learning_rate": 5.515303442574997e-06,
+      "loss": 0.3323,
+      "step": 12000
+    },
+    {
+      "epoch": 1.9776661657849777,
+      "grad_norm": 1.6197963953018188,
+      "learning_rate": 5.499421280918682e-06,
+      "loss": 0.3401,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9793129001049792,
+      "grad_norm": 1.5700541734695435,
+      "learning_rate": 5.4835533424384825e-06,
+      "loss": 0.3204,
+      "step": 12020
+    },
+    {
+      "epoch": 1.980959634424981,
+      "grad_norm": 1.5739840269088745,
+      "learning_rate": 5.467699677281828e-06,
+      "loss": 0.3115,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9826063687449826,
+      "grad_norm": 1.8662614822387695,
+      "learning_rate": 5.451860335551056e-06,
+      "loss": 0.3207,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9842531030649844,
+      "grad_norm": 1.725319504737854,
+      "learning_rate": 5.4360353673032185e-06,
+      "loss": 0.3038,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9858998373849859,
+      "grad_norm": 1.7426807880401611,
+      "learning_rate": 5.420224822549963e-06,
+      "loss": 0.3012,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9875465717049874,
+      "grad_norm": 1.5318362712860107,
+      "learning_rate": 5.404428751257339e-06,
+      "loss": 0.3037,
+      "step": 12070
+    },
+    {
+      "epoch": 1.9891933060249891,
+      "grad_norm": 1.514984369277954,
+      "learning_rate": 5.388647203345659e-06,
+      "loss": 0.3329,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9908400403449908,
+      "grad_norm": 1.3293901681900024,
+      "learning_rate": 5.372880228689341e-06,
+      "loss": 0.3109,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9924867746649926,
+      "grad_norm": 1.5703567266464233,
+      "learning_rate": 5.357127877116743e-06,
+      "loss": 0.3262,
+      "step": 12100
+    },
+    {
+      "epoch": 1.994133508984994,
+      "grad_norm": 1.395535945892334,
+      "learning_rate": 5.3413901984100195e-06,
+      "loss": 0.3314,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9957802433049958,
+      "grad_norm": 1.7006887197494507,
+      "learning_rate": 5.3256672423049396e-06,
+      "loss": 0.3132,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9974269776249973,
+      "grad_norm": 1.6310688257217407,
+      "learning_rate": 5.309959058490754e-06,
+      "loss": 0.2896,
+      "step": 12130
+    },
+    {
+      "epoch": 1.999073711944999,
+      "grad_norm": 1.698585033416748,
+      "learning_rate": 5.294265696610022e-06,
+      "loss": 0.3006,
+      "step": 12140
+    },
+    {
+      "epoch": 2.0008233671600006,
+      "grad_norm": 1.5965461730957031,
+      "learning_rate": 5.2785872062584705e-06,
+      "loss": 0.3434,
+      "step": 12150
+    },
+    {
+      "epoch": 2.0024701014800024,
+      "grad_norm": 1.6904765367507935,
+      "learning_rate": 5.262923636984818e-06,
+      "loss": 0.2823,
+      "step": 12160
+    },
+    {
+      "epoch": 2.004116835800004,
+      "grad_norm": 1.2032884359359741,
+      "learning_rate": 5.24727503829064e-06,
+      "loss": 0.2712,
+      "step": 12170
+    },
+    {
+      "epoch": 2.005763570120006,
+      "grad_norm": 1.4353348016738892,
+      "learning_rate": 5.2316414596301855e-06,
+      "loss": 0.285,
+      "step": 12180
+    },
+    {
+      "epoch": 2.0074103044400076,
+      "grad_norm": 1.5126605033874512,
+      "learning_rate": 5.216022950410251e-06,
+      "loss": 0.2763,
+      "step": 12190
+    },
+    {
+      "epoch": 2.009057038760009,
+      "grad_norm": 2.8418431282043457,
+      "learning_rate": 5.2004195599899966e-06,
+      "loss": 0.2863,
+      "step": 12200
+    },
+    {
+      "epoch": 2.0107037730800106,
+      "grad_norm": 2.0845654010772705,
+      "learning_rate": 5.1848313376808065e-06,
+      "loss": 0.29,
+      "step": 12210
+    },
+    {
+      "epoch": 2.0123505074000123,
+      "grad_norm": 1.393932580947876,
+      "learning_rate": 5.16925833274613e-06,
+      "loss": 0.2904,
+      "step": 12220
+    },
+    {
+      "epoch": 2.013997241720014,
+      "grad_norm": 1.4617998600006104,
+      "learning_rate": 5.153700594401328e-06,
+      "loss": 0.2663,
+      "step": 12230
+    },
+    {
+      "epoch": 2.0156439760400158,
+      "grad_norm": 1.865277886390686,
+      "learning_rate": 5.138158171813507e-06,
+      "loss": 0.2876,
+      "step": 12240
+    },
+    {
+      "epoch": 2.0172907103600175,
+      "grad_norm": 1.4628126621246338,
+      "learning_rate": 5.12263111410138e-06,
+      "loss": 0.2608,
+      "step": 12250
+    },
+    {
+      "epoch": 2.018937444680019,
+      "grad_norm": 1.8890234231948853,
+      "learning_rate": 5.107119470335093e-06,
+      "loss": 0.2869,
+      "step": 12260
+    },
+    {
+      "epoch": 2.0205841790000205,
+      "grad_norm": 2.153687000274658,
+      "learning_rate": 5.091623289536095e-06,
+      "loss": 0.2778,
+      "step": 12270
+    },
+    {
+      "epoch": 2.0222309133200222,
+      "grad_norm": 1.9289218187332153,
+      "learning_rate": 5.076142620676941e-06,
+      "loss": 0.3025,
+      "step": 12280
+    },
+    {
+      "epoch": 2.023877647640024,
+      "grad_norm": 1.6347593069076538,
+      "learning_rate": 5.060677512681187e-06,
+      "loss": 0.2692,
+      "step": 12290
+    },
+    {
+      "epoch": 2.0255243819600257,
+      "grad_norm": 2.1839990615844727,
+      "learning_rate": 5.045228014423203e-06,
+      "loss": 0.2843,
+      "step": 12300
+    },
+    {
+      "epoch": 2.027171116280027,
+      "grad_norm": 2.4028539657592773,
+      "learning_rate": 5.029794174728031e-06,
+      "loss": 0.2798,
+      "step": 12310
+    },
+    {
+      "epoch": 2.0288178506000287,
+      "grad_norm": 2.70993709564209,
+      "learning_rate": 5.014376042371221e-06,
+      "loss": 0.2773,
+      "step": 12320
+    },
+    {
+      "epoch": 2.0304645849200305,
+      "grad_norm": 1.8589779138565063,
+      "learning_rate": 4.998973666078692e-06,
+      "loss": 0.2775,
+      "step": 12330
+    },
+    {
+      "epoch": 2.032111319240032,
+      "grad_norm": 1.971356749534607,
+      "learning_rate": 4.983587094526556e-06,
+      "loss": 0.2776,
+      "step": 12340
+    },
+    {
+      "epoch": 2.033758053560034,
+      "grad_norm": 1.5491329431533813,
+      "learning_rate": 4.9682163763410005e-06,
+      "loss": 0.2729,
+      "step": 12350
+    },
+    {
+      "epoch": 2.035404787880035,
+      "grad_norm": 1.7983886003494263,
+      "learning_rate": 4.952861560098079e-06,
+      "loss": 0.2835,
+      "step": 12360
+    },
+    {
+      "epoch": 2.037051522200037,
+      "grad_norm": 2.259243965148926,
+      "learning_rate": 4.937522694323618e-06,
+      "loss": 0.3009,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0386982565200387,
+      "grad_norm": 2.2802488803863525,
+      "learning_rate": 4.922199827493022e-06,
+      "loss": 0.2921,
+      "step": 12380
+    },
+    {
+      "epoch": 2.0403449908400404,
+      "grad_norm": 1.9333369731903076,
+      "learning_rate": 4.906893008031141e-06,
+      "loss": 0.2746,
+      "step": 12390
+    },
+    {
+      "epoch": 2.041991725160042,
+      "grad_norm": 1.7847884893417358,
+      "learning_rate": 4.8916022843121e-06,
+      "loss": 0.2816,
+      "step": 12400
+    },
+    {
+      "epoch": 2.043638459480044,
+      "grad_norm": 1.5014792680740356,
+      "learning_rate": 4.876327704659172e-06,
+      "loss": 0.2897,
+      "step": 12410
+    },
+    {
+      "epoch": 2.045285193800045,
+      "grad_norm": 2.1002280712127686,
+      "learning_rate": 4.861069317344598e-06,
+      "loss": 0.2744,
+      "step": 12420
+    },
+    {
+      "epoch": 2.046931928120047,
+      "grad_norm": 1.5581461191177368,
+      "learning_rate": 4.845827170589449e-06,
+      "loss": 0.2886,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0485786624400486,
+      "grad_norm": 1.6326346397399902,
+      "learning_rate": 4.830601312563469e-06,
+      "loss": 0.2859,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0502253967600503,
+      "grad_norm": 1.5935137271881104,
+      "learning_rate": 4.815391791384933e-06,
+      "loss": 0.2932,
+      "step": 12450
+    },
+    {
+      "epoch": 2.051872131080052,
+      "grad_norm": 1.5060147047042847,
+      "learning_rate": 4.800198655120478e-06,
+      "loss": 0.2899,
+      "step": 12460
+    },
+    {
+      "epoch": 2.0535188654000534,
+      "grad_norm": 1.7694473266601562,
+      "learning_rate": 4.785021951784967e-06,
+      "loss": 0.2852,
+      "step": 12470
+    },
+    {
+      "epoch": 2.055165599720055,
+      "grad_norm": 1.579512596130371,
+      "learning_rate": 4.76986172934132e-06,
+      "loss": 0.2837,
+      "step": 12480
+    },
+    {
+      "epoch": 2.056812334040057,
+      "grad_norm": 1.4137424230575562,
+      "learning_rate": 4.7547180357003885e-06,
+      "loss": 0.2793,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0584590683600585,
+      "grad_norm": 1.5642904043197632,
+      "learning_rate": 4.739590918720765e-06,
+      "loss": 0.2766,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0601058026800603,
+      "grad_norm": 2.033942937850952,
+      "learning_rate": 4.724480426208678e-06,
+      "loss": 0.3038,
+      "step": 12510
+    },
+    {
+      "epoch": 2.0617525370000616,
+      "grad_norm": 1.408429503440857,
+      "learning_rate": 4.709386605917798e-06,
+      "loss": 0.2747,
+      "step": 12520
+    },
+    {
+      "epoch": 2.0633992713200633,
+      "grad_norm": 1.467748761177063,
+      "learning_rate": 4.694309505549128e-06,
+      "loss": 0.2833,
+      "step": 12530
+    },
+    {
+      "epoch": 2.065046005640065,
+      "grad_norm": 2.374821662902832,
+      "learning_rate": 4.6792491727508076e-06,
+      "loss": 0.2828,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0666927399600667,
+      "grad_norm": 2.1544690132141113,
+      "learning_rate": 4.664205655118006e-06,
+      "loss": 0.2872,
+      "step": 12550
+    },
+    {
+      "epoch": 2.0683394742800685,
+      "grad_norm": 1.5849260091781616,
+      "learning_rate": 4.6491790001927385e-06,
+      "loss": 0.2748,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0699862086000698,
+      "grad_norm": 1.403977394104004,
+      "learning_rate": 4.634169255463734e-06,
+      "loss": 0.2899,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0716329429200715,
+      "grad_norm": 1.8115674257278442,
+      "learning_rate": 4.619176468366274e-06,
+      "loss": 0.2723,
+      "step": 12580
+    },
+    {
+      "epoch": 2.0732796772400732,
+      "grad_norm": 1.3875244855880737,
+      "learning_rate": 4.604200686282063e-06,
+      "loss": 0.2661,
+      "step": 12590
+    },
+    {
+      "epoch": 2.074926411560075,
+      "grad_norm": 2.000905990600586,
+      "learning_rate": 4.5892419565390486e-06,
+      "loss": 0.2808,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0765731458800767,
+      "grad_norm": 2.43119740486145,
+      "learning_rate": 4.5743003264113015e-06,
+      "loss": 0.2826,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0782198802000784,
+      "grad_norm": 3.4667115211486816,
+      "learning_rate": 4.559375843118839e-06,
+      "loss": 0.2803,
+      "step": 12620
+    },
+    {
+      "epoch": 2.0798666145200797,
+      "grad_norm": 1.7227445840835571,
+      "learning_rate": 4.544468553827508e-06,
+      "loss": 0.2836,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0815133488400814,
+      "grad_norm": 1.7388190031051636,
+      "learning_rate": 4.529578505648789e-06,
+      "loss": 0.288,
+      "step": 12640
+    },
+    {
+      "epoch": 2.083160083160083,
+      "grad_norm": 1.63683021068573,
+      "learning_rate": 4.514705745639706e-06,
+      "loss": 0.2896,
+      "step": 12650
+    },
+    {
+      "epoch": 2.084806817480085,
+      "grad_norm": 2.0901145935058594,
+      "learning_rate": 4.499850320802623e-06,
+      "loss": 0.2793,
+      "step": 12660
+    },
+    {
+      "epoch": 2.0864535518000866,
+      "grad_norm": 2.0766639709472656,
+      "learning_rate": 4.485012278085139e-06,
+      "loss": 0.2731,
+      "step": 12670
+    },
+    {
+      "epoch": 2.088100286120088,
+      "grad_norm": 1.5726747512817383,
+      "learning_rate": 4.470191664379903e-06,
+      "loss": 0.2751,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0897470204400896,
+      "grad_norm": 2.0181379318237305,
+      "learning_rate": 4.455388526524498e-06,
+      "loss": 0.2992,
+      "step": 12690
+    },
+    {
+      "epoch": 2.0913937547600914,
+      "grad_norm": 2.143312692642212,
+      "learning_rate": 4.440602911301267e-06,
+      "loss": 0.2939,
+      "step": 12700
+    },
+    {
+      "epoch": 2.093040489080093,
+      "grad_norm": 1.3811988830566406,
+      "learning_rate": 4.425834865437184e-06,
+      "loss": 0.2795,
+      "step": 12710
+    },
+    {
+      "epoch": 2.094687223400095,
+      "grad_norm": 1.4296287298202515,
+      "learning_rate": 4.411084435603688e-06,
+      "loss": 0.2831,
+      "step": 12720
+    },
+    {
+      "epoch": 2.096333957720096,
+      "grad_norm": 1.4301252365112305,
+      "learning_rate": 4.396351668416562e-06,
+      "loss": 0.2963,
+      "step": 12730
+    },
+    {
+      "epoch": 2.097980692040098,
+      "grad_norm": 2.8089404106140137,
+      "learning_rate": 4.3816366104357545e-06,
+      "loss": 0.2935,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0996274263600996,
+      "grad_norm": 1.936610460281372,
+      "learning_rate": 4.366939308165259e-06,
+      "loss": 0.3008,
+      "step": 12750
+    },
+    {
+      "epoch": 2.1012741606801013,
+      "grad_norm": 1.634929895401001,
+      "learning_rate": 4.352259808052944e-06,
+      "loss": 0.2853,
+      "step": 12760
+    },
+    {
+      "epoch": 2.102920895000103,
+      "grad_norm": 1.791841745376587,
+      "learning_rate": 4.337598156490435e-06,
+      "loss": 0.2583,
+      "step": 12770
+    },
+    {
+      "epoch": 2.1045676293201048,
+      "grad_norm": 1.6003674268722534,
+      "learning_rate": 4.322954399812932e-06,
+      "loss": 0.2787,
+      "step": 12780
+    },
+    {
+      "epoch": 2.106214363640106,
+      "grad_norm": 1.8749767541885376,
+      "learning_rate": 4.308328584299092e-06,
+      "loss": 0.2829,
+      "step": 12790
+    },
+    {
+      "epoch": 2.107861097960108,
+      "grad_norm": 1.6271084547042847,
+      "learning_rate": 4.293720756170868e-06,
+      "loss": 0.2713,
+      "step": 12800
+    },
+    {
+      "epoch": 2.1095078322801095,
+      "grad_norm": 1.9377316236495972,
+      "learning_rate": 4.2791309615933764e-06,
+      "loss": 0.2913,
+      "step": 12810
+    },
+    {
+      "epoch": 2.1111545666001112,
+      "grad_norm": 1.353582501411438,
+      "learning_rate": 4.264559246674727e-06,
+      "loss": 0.2805,
+      "step": 12820
+    },
+    {
+      "epoch": 2.112801300920113,
+      "grad_norm": 1.5382156372070312,
+      "learning_rate": 4.2500056574659135e-06,
+      "loss": 0.2853,
+      "step": 12830
+    },
+    {
+      "epoch": 2.1144480352401143,
+      "grad_norm": 1.6927893161773682,
+      "learning_rate": 4.235470239960623e-06,
+      "loss": 0.2789,
+      "step": 12840
+    },
+    {
+      "epoch": 2.116094769560116,
+      "grad_norm": 1.7223583459854126,
+      "learning_rate": 4.2209530400951335e-06,
+      "loss": 0.2786,
+      "step": 12850
+    },
+    {
+      "epoch": 2.1177415038801177,
+      "grad_norm": 1.528069257736206,
+      "learning_rate": 4.206454103748142e-06,
+      "loss": 0.2854,
+      "step": 12860
+    },
+    {
+      "epoch": 2.1193882382001195,
+      "grad_norm": 1.9811831712722778,
+      "learning_rate": 4.191973476740628e-06,
+      "loss": 0.2687,
+      "step": 12870
+    },
+    {
+      "epoch": 2.121034972520121,
+      "grad_norm": 1.6877635717391968,
+      "learning_rate": 4.1775112048357e-06,
+      "loss": 0.2725,
+      "step": 12880
+    },
+    {
+      "epoch": 2.1226817068401225,
+      "grad_norm": 1.6502715349197388,
+      "learning_rate": 4.163067333738479e-06,
+      "loss": 0.2712,
+      "step": 12890
+    },
+    {
+      "epoch": 2.124328441160124,
+      "grad_norm": 1.460304617881775,
+      "learning_rate": 4.148641909095911e-06,
+      "loss": 0.2653,
+      "step": 12900
+    },
+    {
+      "epoch": 2.125975175480126,
+      "grad_norm": 1.7706825733184814,
+      "learning_rate": 4.134234976496666e-06,
+      "loss": 0.2853,
+      "step": 12910
+    },
+    {
+      "epoch": 2.1276219098001277,
+      "grad_norm": 1.4802557229995728,
+      "learning_rate": 4.11984658147096e-06,
+      "loss": 0.2824,
+      "step": 12920
+    },
+    {
+      "epoch": 2.1292686441201294,
+      "grad_norm": 1.5747597217559814,
+      "learning_rate": 4.105476769490424e-06,
+      "loss": 0.268,
+      "step": 12930
+    },
+    {
+      "epoch": 2.1309153784401307,
+      "grad_norm": 2.0587172508239746,
+      "learning_rate": 4.091125585967975e-06,
+      "loss": 0.2962,
+      "step": 12940
+    },
+    {
+      "epoch": 2.1325621127601324,
+      "grad_norm": 1.5220030546188354,
+      "learning_rate": 4.0767930762576415e-06,
+      "loss": 0.2717,
+      "step": 12950
+    },
+    {
+      "epoch": 2.134208847080134,
+      "grad_norm": 1.6755307912826538,
+      "learning_rate": 4.0624792856544505e-06,
+      "loss": 0.2795,
+      "step": 12960
+    },
+    {
+      "epoch": 2.135855581400136,
+      "grad_norm": 1.359605073928833,
+      "learning_rate": 4.0481842593942636e-06,
+      "loss": 0.2689,
+      "step": 12970
+    },
+    {
+      "epoch": 2.1375023157201376,
+      "grad_norm": 1.5714139938354492,
+      "learning_rate": 4.033908042653639e-06,
+      "loss": 0.2772,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1391490500401393,
+      "grad_norm": 1.4556095600128174,
+      "learning_rate": 4.019650680549704e-06,
+      "loss": 0.276,
+      "step": 12990
+    },
+    {
+      "epoch": 2.1407957843601406,
+      "grad_norm": 1.8429503440856934,
+      "learning_rate": 4.005412218139986e-06,
+      "loss": 0.2829,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1424425186801423,
+      "grad_norm": 1.6822142601013184,
+      "learning_rate": 3.991192700422286e-06,
+      "loss": 0.2829,
+      "step": 13010
+    },
+    {
+      "epoch": 2.144089253000144,
+      "grad_norm": 1.859412670135498,
+      "learning_rate": 3.976992172334544e-06,
+      "loss": 0.2764,
+      "step": 13020
+    },
+    {
+      "epoch": 2.145735987320146,
+      "grad_norm": 1.6306754350662231,
+      "learning_rate": 3.962810678754674e-06,
+      "loss": 0.2731,
+      "step": 13030
+    },
+    {
+      "epoch": 2.1473827216401475,
+      "grad_norm": 1.5259572267532349,
+      "learning_rate": 3.948648264500445e-06,
+      "loss": 0.2608,
+      "step": 13040
+    },
+    {
+      "epoch": 2.149029455960149,
+      "grad_norm": 1.6222518682479858,
+      "learning_rate": 3.934504974329326e-06,
+      "loss": 0.299,
+      "step": 13050
+    },
+    {
+      "epoch": 2.1506761902801506,
+      "grad_norm": 1.880744457244873,
+      "learning_rate": 3.920380852938348e-06,
+      "loss": 0.2744,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1523229246001523,
+      "grad_norm": 1.6208007335662842,
+      "learning_rate": 3.906275944963957e-06,
+      "loss": 0.2817,
+      "step": 13070
+    },
+    {
+      "epoch": 2.153969658920154,
+      "grad_norm": 1.4789073467254639,
+      "learning_rate": 3.892190294981893e-06,
+      "loss": 0.2813,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1556163932401557,
+      "grad_norm": 1.7686960697174072,
+      "learning_rate": 3.8781239475070194e-06,
+      "loss": 0.2651,
+      "step": 13090
+    },
+    {
+      "epoch": 2.1572631275601575,
+      "grad_norm": 1.6891413927078247,
+      "learning_rate": 3.864076946993215e-06,
+      "loss": 0.2503,
+      "step": 13100
+    },
+    {
+      "epoch": 2.1589098618801588,
+      "grad_norm": 1.6192152500152588,
+      "learning_rate": 3.850049337833196e-06,
+      "loss": 0.2693,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1605565962001605,
+      "grad_norm": 1.6435240507125854,
+      "learning_rate": 3.836041164358416e-06,
+      "loss": 0.2769,
+      "step": 13120
+    },
+    {
+      "epoch": 2.1622033305201622,
+      "grad_norm": 1.7345331907272339,
+      "learning_rate": 3.822052470838893e-06,
+      "loss": 0.2804,
+      "step": 13130
+    },
+    {
+      "epoch": 2.163850064840164,
+      "grad_norm": 1.42905855178833,
+      "learning_rate": 3.8080833014830865e-06,
+      "loss": 0.2749,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1654967991601657,
+      "grad_norm": 1.6597148180007935,
+      "learning_rate": 3.7941337004377497e-06,
+      "loss": 0.2748,
+      "step": 13150
+    },
+    {
+      "epoch": 2.167143533480167,
+      "grad_norm": 1.6788873672485352,
+      "learning_rate": 3.7802037117878053e-06,
+      "loss": 0.2832,
+      "step": 13160
+    },
+    {
+      "epoch": 2.1687902678001687,
+      "grad_norm": 1.761786699295044,
+      "learning_rate": 3.7662933795561805e-06,
+      "loss": 0.2943,
+      "step": 13170
+    },
+    {
+      "epoch": 2.1704370021201704,
+      "grad_norm": 1.6826972961425781,
+      "learning_rate": 3.7524027477036974e-06,
+      "loss": 0.2763,
+      "step": 13180
+    },
+    {
+      "epoch": 2.172083736440172,
+      "grad_norm": 1.4096064567565918,
+      "learning_rate": 3.7385318601289034e-06,
+      "loss": 0.2904,
+      "step": 13190
+    },
+    {
+      "epoch": 2.173730470760174,
+      "grad_norm": 2.226126194000244,
+      "learning_rate": 3.724680760667967e-06,
+      "loss": 0.2656,
+      "step": 13200
+    },
+    {
+      "epoch": 2.175377205080175,
+      "grad_norm": 1.3945817947387695,
+      "learning_rate": 3.7108494930944937e-06,
+      "loss": 0.2762,
+      "step": 13210
+    },
+    {
+      "epoch": 2.177023939400177,
+      "grad_norm": 1.428282618522644,
+      "learning_rate": 3.697038101119442e-06,
+      "loss": 0.2728,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1786706737201786,
+      "grad_norm": 1.9224436283111572,
+      "learning_rate": 3.6832466283909386e-06,
+      "loss": 0.2577,
+      "step": 13230
+    },
+    {
+      "epoch": 2.1803174080401804,
+      "grad_norm": 2.1055612564086914,
+      "learning_rate": 3.6694751184941712e-06,
+      "loss": 0.2695,
+      "step": 13240
+    },
+    {
+      "epoch": 2.181964142360182,
+      "grad_norm": 2.1279563903808594,
+      "learning_rate": 3.6557236149512276e-06,
+      "loss": 0.2871,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1836108766801834,
+      "grad_norm": 2.5193889141082764,
+      "learning_rate": 3.641992161220983e-06,
+      "loss": 0.2814,
+      "step": 13260
+    },
+    {
+      "epoch": 2.185257611000185,
+      "grad_norm": 1.3304851055145264,
+      "learning_rate": 3.628280800698939e-06,
+      "loss": 0.2704,
+      "step": 13270
+    },
+    {
+      "epoch": 2.186904345320187,
+      "grad_norm": 1.5844635963439941,
+      "learning_rate": 3.6145895767171e-06,
+      "loss": 0.2627,
+      "step": 13280
+    },
+    {
+      "epoch": 2.1885510796401886,
+      "grad_norm": 1.840064287185669,
+      "learning_rate": 3.6009185325438278e-06,
+      "loss": 0.2649,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1901978139601903,
+      "grad_norm": 1.7065536975860596,
+      "learning_rate": 3.5872677113837227e-06,
+      "loss": 0.2494,
+      "step": 13300
+    },
+    {
+      "epoch": 2.191844548280192,
+      "grad_norm": 1.525583267211914,
+      "learning_rate": 3.5736371563774587e-06,
+      "loss": 0.2716,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1934912826001933,
+      "grad_norm": 1.5239418745040894,
+      "learning_rate": 3.560026910601678e-06,
+      "loss": 0.272,
+      "step": 13320
+    },
+    {
+      "epoch": 2.195138016920195,
+      "grad_norm": 1.7140849828720093,
+      "learning_rate": 3.5464370170688244e-06,
+      "loss": 0.2802,
+      "step": 13330
+    },
+    {
+      "epoch": 2.196784751240197,
+      "grad_norm": 1.8460406064987183,
+      "learning_rate": 3.53286751872704e-06,
+      "loss": 0.2674,
+      "step": 13340
+    },
+    {
+      "epoch": 2.1984314855601985,
+      "grad_norm": 1.5761222839355469,
+      "learning_rate": 3.519318458459988e-06,
+      "loss": 0.2673,
+      "step": 13350
+    },
+    {
+      "epoch": 2.2000782198802002,
+      "grad_norm": 1.5991119146347046,
+      "learning_rate": 3.5057898790867673e-06,
+      "loss": 0.2904,
+      "step": 13360
+    },
+    {
+      "epoch": 2.2017249542002015,
+      "grad_norm": 1.726907730102539,
+      "learning_rate": 3.4922818233617295e-06,
+      "loss": 0.2729,
+      "step": 13370
+    },
+    {
+      "epoch": 2.2033716885202033,
+      "grad_norm": 1.3864545822143555,
+      "learning_rate": 3.4787943339743836e-06,
+      "loss": 0.2753,
+      "step": 13380
+    },
+    {
+      "epoch": 2.205018422840205,
+      "grad_norm": 2.468479871749878,
+      "learning_rate": 3.4653274535492255e-06,
+      "loss": 0.2777,
+      "step": 13390
+    },
+    {
+      "epoch": 2.2066651571602067,
+      "grad_norm": 1.603942632675171,
+      "learning_rate": 3.451881224645637e-06,
+      "loss": 0.2721,
+      "step": 13400
+    },
+    {
+      "epoch": 2.2083118914802085,
+      "grad_norm": 1.7777491807937622,
+      "learning_rate": 3.4384556897577183e-06,
+      "loss": 0.2949,
+      "step": 13410
+    },
+    {
+      "epoch": 2.20995862580021,
+      "grad_norm": 1.8029536008834839,
+      "learning_rate": 3.425050891314191e-06,
+      "loss": 0.272,
+      "step": 13420
+    },
+    {
+      "epoch": 2.2116053601202115,
+      "grad_norm": 1.74064302444458,
+      "learning_rate": 3.4116668716782164e-06,
+      "loss": 0.267,
+      "step": 13430
+    },
+    {
+      "epoch": 2.213252094440213,
+      "grad_norm": 1.70004403591156,
+      "learning_rate": 3.398303673147314e-06,
+      "loss": 0.2528,
+      "step": 13440
+    },
+    {
+      "epoch": 2.214898828760215,
+      "grad_norm": 1.407127022743225,
+      "learning_rate": 3.3849613379531865e-06,
+      "loss": 0.27,
+      "step": 13450
+    },
+    {
+      "epoch": 2.2165455630802167,
+      "grad_norm": 2.015048027038574,
+      "learning_rate": 3.371639908261611e-06,
+      "loss": 0.2576,
+      "step": 13460
+    },
+    {
+      "epoch": 2.2181922974002184,
+      "grad_norm": 2.1229333877563477,
+      "learning_rate": 3.35833942617229e-06,
+      "loss": 0.2779,
+      "step": 13470
+    },
+    {
+      "epoch": 2.2198390317202197,
+      "grad_norm": 1.634856939315796,
+      "learning_rate": 3.3450599337187326e-06,
+      "loss": 0.2564,
+      "step": 13480
+    },
+    {
+      "epoch": 2.2214857660402214,
+      "grad_norm": 1.7512359619140625,
+      "learning_rate": 3.3318014728681104e-06,
+      "loss": 0.2773,
+      "step": 13490
+    },
+    {
+      "epoch": 2.223132500360223,
+      "grad_norm": 1.8369240760803223,
+      "learning_rate": 3.3185640855211255e-06,
+      "loss": 0.2501,
+      "step": 13500
+    },
+    {
+      "epoch": 2.224779234680225,
+      "grad_norm": 2.120023012161255,
+      "learning_rate": 3.3053478135118845e-06,
+      "loss": 0.2836,
+      "step": 13510
+    },
+    {
+      "epoch": 2.2264259690002266,
+      "grad_norm": 1.684515118598938,
+      "learning_rate": 3.292152698607768e-06,
+      "loss": 0.2638,
+      "step": 13520
+    },
+    {
+      "epoch": 2.228072703320228,
+      "grad_norm": 1.8493366241455078,
+      "learning_rate": 3.2789787825092834e-06,
+      "loss": 0.2579,
+      "step": 13530
+    },
+    {
+      "epoch": 2.2297194376402296,
+      "grad_norm": 1.272315502166748,
+      "learning_rate": 3.2658261068499562e-06,
+      "loss": 0.2729,
+      "step": 13540
+    },
+    {
+      "epoch": 2.2313661719602313,
+      "grad_norm": 2.1729161739349365,
+      "learning_rate": 3.252694713196173e-06,
+      "loss": 0.275,
+      "step": 13550
+    },
+    {
+      "epoch": 2.233012906280233,
+      "grad_norm": 1.7917054891586304,
+      "learning_rate": 3.239584643047078e-06,
+      "loss": 0.2788,
+      "step": 13560
+    },
+    {
+      "epoch": 2.234659640600235,
+      "grad_norm": 1.5601428747177124,
+      "learning_rate": 3.2264959378344053e-06,
+      "loss": 0.2728,
+      "step": 13570
+    },
+    {
+      "epoch": 2.236306374920236,
+      "grad_norm": 1.8332723379135132,
+      "learning_rate": 3.2134286389223913e-06,
+      "loss": 0.2782,
+      "step": 13580
+    },
+    {
+      "epoch": 2.237953109240238,
+      "grad_norm": 1.5861481428146362,
+      "learning_rate": 3.2003827876076066e-06,
+      "loss": 0.2638,
+      "step": 13590
+    },
+    {
+      "epoch": 2.2395998435602396,
+      "grad_norm": 1.4836739301681519,
+      "learning_rate": 3.1873584251188527e-06,
+      "loss": 0.2454,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2412465778802413,
+      "grad_norm": 1.721747636795044,
+      "learning_rate": 3.174355592617008e-06,
+      "loss": 0.2766,
+      "step": 13610
+    },
+    {
+      "epoch": 2.242893312200243,
+      "grad_norm": 2.278952121734619,
+      "learning_rate": 3.161374331194922e-06,
+      "loss": 0.2613,
+      "step": 13620
+    },
+    {
+      "epoch": 2.2445400465202447,
+      "grad_norm": 2.0035574436187744,
+      "learning_rate": 3.1484146818772644e-06,
+      "loss": 0.2816,
+      "step": 13630
+    },
+    {
+      "epoch": 2.246186780840246,
+      "grad_norm": 1.6324348449707031,
+      "learning_rate": 3.1354766856204066e-06,
+      "loss": 0.2613,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2478335151602478,
+      "grad_norm": 2.2162437438964844,
+      "learning_rate": 3.1225603833122866e-06,
+      "loss": 0.2682,
+      "step": 13650
+    },
+    {
+      "epoch": 2.2494802494802495,
+      "grad_norm": 1.5400594472885132,
+      "learning_rate": 3.1096658157722936e-06,
+      "loss": 0.2776,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2511269838002512,
+      "grad_norm": 2.152050495147705,
+      "learning_rate": 3.0967930237511144e-06,
+      "loss": 0.2573,
+      "step": 13670
+    },
+    {
+      "epoch": 2.252773718120253,
+      "grad_norm": 1.3941951990127563,
+      "learning_rate": 3.0839420479306325e-06,
+      "loss": 0.2738,
+      "step": 13680
+    },
+    {
+      "epoch": 2.2544204524402542,
+      "grad_norm": 1.7376865148544312,
+      "learning_rate": 3.071112928923773e-06,
+      "loss": 0.2625,
+      "step": 13690
+    },
+    {
+      "epoch": 2.256067186760256,
+      "grad_norm": 2.0723204612731934,
+      "learning_rate": 3.058305707274396e-06,
+      "loss": 0.2632,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2577139210802577,
+      "grad_norm": 2.126526117324829,
+      "learning_rate": 3.0455204234571568e-06,
+      "loss": 0.2649,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2593606554002594,
+      "grad_norm": 2.2294485569000244,
+      "learning_rate": 3.0327571178773772e-06,
+      "loss": 0.2506,
+      "step": 13720
+    },
+    {
+      "epoch": 2.261007389720261,
+      "grad_norm": 1.6885284185409546,
+      "learning_rate": 3.0200158308709217e-06,
+      "loss": 0.2715,
+      "step": 13730
+    },
+    {
+      "epoch": 2.262654124040263,
+      "grad_norm": 1.9802756309509277,
+      "learning_rate": 3.0072966027040785e-06,
+      "loss": 0.2647,
+      "step": 13740
+    },
+    {
+      "epoch": 2.264300858360264,
+      "grad_norm": 1.758371353149414,
+      "learning_rate": 2.9945994735734085e-06,
+      "loss": 0.2741,
+      "step": 13750
+    },
+    {
+      "epoch": 2.265947592680266,
+      "grad_norm": 1.7555452585220337,
+      "learning_rate": 2.9819244836056482e-06,
+      "loss": 0.2659,
+      "step": 13760
+    },
+    {
+      "epoch": 2.2675943270002676,
+      "grad_norm": 1.7351974248886108,
+      "learning_rate": 2.9692716728575576e-06,
+      "loss": 0.2482,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2692410613202694,
+      "grad_norm": 1.67487633228302,
+      "learning_rate": 2.956641081315803e-06,
+      "loss": 0.2655,
+      "step": 13780
+    },
+    {
+      "epoch": 2.270887795640271,
+      "grad_norm": 1.7826485633850098,
+      "learning_rate": 2.944032748896842e-06,
+      "loss": 0.2706,
+      "step": 13790
+    },
+    {
+      "epoch": 2.2725345299602724,
+      "grad_norm": 1.6247366666793823,
+      "learning_rate": 2.9314467154467728e-06,
+      "loss": 0.2877,
+      "step": 13800
+    },
+    {
+      "epoch": 2.274181264280274,
+      "grad_norm": 1.7416200637817383,
+      "learning_rate": 2.9188830207412355e-06,
+      "loss": 0.2653,
+      "step": 13810
+    },
+    {
+      "epoch": 2.275827998600276,
+      "grad_norm": 1.554840087890625,
+      "learning_rate": 2.9063417044852627e-06,
+      "loss": 0.2528,
+      "step": 13820
+    },
+    {
+      "epoch": 2.2774747329202776,
+      "grad_norm": 1.978947639465332,
+      "learning_rate": 2.8938228063131655e-06,
+      "loss": 0.2884,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2791214672402793,
+      "grad_norm": 2.049222946166992,
+      "learning_rate": 2.881326365788417e-06,
+      "loss": 0.2699,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2807682015602806,
+      "grad_norm": 1.588185429573059,
+      "learning_rate": 2.8688524224035076e-06,
+      "loss": 0.2725,
+      "step": 13850
+    },
+    {
+      "epoch": 2.2824149358802823,
+      "grad_norm": 1.8063595294952393,
+      "learning_rate": 2.856401015579828e-06,
+      "loss": 0.2775,
+      "step": 13860
+    },
+    {
+      "epoch": 2.284061670200284,
+      "grad_norm": 1.70321786403656,
+      "learning_rate": 2.8439721846675595e-06,
+      "loss": 0.259,
+      "step": 13870
+    },
+    {
+      "epoch": 2.285708404520286,
+      "grad_norm": 1.4761788845062256,
+      "learning_rate": 2.83156596894552e-06,
+      "loss": 0.2497,
+      "step": 13880
+    },
+    {
+      "epoch": 2.2873551388402875,
+      "grad_norm": 2.2861974239349365,
+      "learning_rate": 2.819182407621074e-06,
+      "loss": 0.2603,
+      "step": 13890
+    },
+    {
+      "epoch": 2.289001873160289,
+      "grad_norm": 2.404402494430542,
+      "learning_rate": 2.806821539829978e-06,
+      "loss": 0.2708,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2906486074802905,
+      "grad_norm": 1.8514249324798584,
+      "learning_rate": 2.7944834046362755e-06,
+      "loss": 0.259,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2922953418002923,
+      "grad_norm": 2.039151191711426,
+      "learning_rate": 2.7821680410321638e-06,
+      "loss": 0.2385,
+      "step": 13920
+    },
+    {
+      "epoch": 2.293942076120294,
+      "grad_norm": 1.4477638006210327,
+      "learning_rate": 2.7698754879378853e-06,
+      "loss": 0.2805,
+      "step": 13930
+    },
+    {
+      "epoch": 2.2955888104402957,
+      "grad_norm": 1.8691905736923218,
+      "learning_rate": 2.7576057842015823e-06,
+      "loss": 0.2642,
+      "step": 13940
+    },
+    {
+      "epoch": 2.297235544760297,
+      "grad_norm": 2.3928122520446777,
+      "learning_rate": 2.7453589685991964e-06,
+      "loss": 0.2619,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2988822790802987,
+      "grad_norm": 1.9695171117782593,
+      "learning_rate": 2.733135079834327e-06,
+      "loss": 0.2569,
+      "step": 13960
+    },
+    {
+      "epoch": 2.3005290134003005,
+      "grad_norm": 1.6952944993972778,
+      "learning_rate": 2.7209341565381275e-06,
+      "loss": 0.272,
+      "step": 13970
+    },
+    {
+      "epoch": 2.302175747720302,
+      "grad_norm": 2.042433500289917,
+      "learning_rate": 2.7087562372691644e-06,
+      "loss": 0.2553,
+      "step": 13980
+    },
+    {
+      "epoch": 2.303822482040304,
+      "grad_norm": 1.8709336519241333,
+      "learning_rate": 2.696601360513309e-06,
+      "loss": 0.2592,
+      "step": 13990
+    },
+    {
+      "epoch": 2.3054692163603057,
+      "grad_norm": 1.952884316444397,
+      "learning_rate": 2.684469564683608e-06,
+      "loss": 0.269,
+      "step": 14000
+    },
+    {
+      "epoch": 2.307115950680307,
+      "grad_norm": 1.885693907737732,
+      "learning_rate": 2.6723608881201737e-06,
+      "loss": 0.2986,
+      "step": 14010
+    },
+    {
+      "epoch": 2.3087626850003087,
+      "grad_norm": 1.9475207328796387,
+      "learning_rate": 2.660275369090043e-06,
+      "loss": 0.2396,
+      "step": 14020
+    },
+    {
+      "epoch": 2.3104094193203104,
+      "grad_norm": 1.693192958831787,
+      "learning_rate": 2.6482130457870813e-06,
+      "loss": 0.2611,
+      "step": 14030
+    },
+    {
+      "epoch": 2.312056153640312,
+      "grad_norm": 1.7588555812835693,
+      "learning_rate": 2.6361739563318334e-06,
+      "loss": 0.2704,
+      "step": 14040
+    },
+    {
+      "epoch": 2.313702887960314,
+      "grad_norm": 2.1660373210906982,
+      "learning_rate": 2.6241581387714333e-06,
+      "loss": 0.2869,
+      "step": 14050
+    },
+    {
+      "epoch": 2.315349622280315,
+      "grad_norm": 1.8832919597625732,
+      "learning_rate": 2.61216563107946e-06,
+      "loss": 0.2388,
+      "step": 14060
+    },
+    {
+      "epoch": 2.316996356600317,
+      "grad_norm": 1.77772855758667,
+      "learning_rate": 2.6001964711558245e-06,
+      "loss": 0.2586,
+      "step": 14070
+    },
+    {
+      "epoch": 2.3186430909203186,
+      "grad_norm": 1.9591187238693237,
+      "learning_rate": 2.5882506968266564e-06,
+      "loss": 0.2469,
+      "step": 14080
+    },
+    {
+      "epoch": 2.3202898252403203,
+      "grad_norm": 1.8513628244400024,
+      "learning_rate": 2.5763283458441823e-06,
+      "loss": 0.2728,
+      "step": 14090
+    },
+    {
+      "epoch": 2.321936559560322,
+      "grad_norm": 1.9876492023468018,
+      "learning_rate": 2.5644294558865955e-06,
+      "loss": 0.254,
+      "step": 14100
+    },
+    {
+      "epoch": 2.323583293880324,
+      "grad_norm": 1.812187910079956,
+      "learning_rate": 2.5525540645579573e-06,
+      "loss": 0.2641,
+      "step": 14110
+    },
+    {
+      "epoch": 2.325230028200325,
+      "grad_norm": 1.6044301986694336,
+      "learning_rate": 2.540702209388052e-06,
+      "loss": 0.2712,
+      "step": 14120
+    },
+    {
+      "epoch": 2.326876762520327,
+      "grad_norm": 1.71054208278656,
+      "learning_rate": 2.5288739278322992e-06,
+      "loss": 0.2619,
+      "step": 14130
+    },
+    {
+      "epoch": 2.3285234968403286,
+      "grad_norm": 1.707785964012146,
+      "learning_rate": 2.5170692572715983e-06,
+      "loss": 0.2493,
+      "step": 14140
+    },
+    {
+      "epoch": 2.3301702311603303,
+      "grad_norm": 1.6551952362060547,
+      "learning_rate": 2.505288235012251e-06,
+      "loss": 0.2635,
+      "step": 14150
+    },
+    {
+      "epoch": 2.331816965480332,
+      "grad_norm": 1.5544058084487915,
+      "learning_rate": 2.4935308982858097e-06,
+      "loss": 0.2601,
+      "step": 14160
+    },
+    {
+      "epoch": 2.3334636998003333,
+      "grad_norm": 2.0415706634521484,
+      "learning_rate": 2.4817972842489824e-06,
+      "loss": 0.2691,
+      "step": 14170
+    },
+    {
+      "epoch": 2.335110434120335,
+      "grad_norm": 1.9987118244171143,
+      "learning_rate": 2.4700874299834975e-06,
+      "loss": 0.257,
+      "step": 14180
+    },
+    {
+      "epoch": 2.3367571684403368,
+      "grad_norm": 2.0765695571899414,
+      "learning_rate": 2.4584013724960063e-06,
+      "loss": 0.2277,
+      "step": 14190
+    },
+    {
+      "epoch": 2.3384039027603385,
+      "grad_norm": 2.2434241771698,
+      "learning_rate": 2.4467391487179446e-06,
+      "loss": 0.245,
+      "step": 14200
+    },
+    {
+      "epoch": 2.3400506370803402,
+      "grad_norm": 1.8256983757019043,
+      "learning_rate": 2.4351007955054316e-06,
+      "loss": 0.2637,
+      "step": 14210
+    },
+    {
+      "epoch": 2.3416973714003415,
+      "grad_norm": 1.6313284635543823,
+      "learning_rate": 2.4234863496391458e-06,
+      "loss": 0.2765,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3433441057203432,
+      "grad_norm": 1.5957192182540894,
+      "learning_rate": 2.411895847824218e-06,
+      "loss": 0.2675,
+      "step": 14230
+    },
+    {
+      "epoch": 2.344990840040345,
+      "grad_norm": 1.8532218933105469,
+      "learning_rate": 2.4003293266900985e-06,
+      "loss": 0.2673,
+      "step": 14240
+    },
+    {
+      "epoch": 2.3466375743603467,
+      "grad_norm": 2.489428997039795,
+      "learning_rate": 2.388786822790462e-06,
+      "loss": 0.2629,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3482843086803484,
+      "grad_norm": 1.6500635147094727,
+      "learning_rate": 2.377268372603071e-06,
+      "loss": 0.2726,
+      "step": 14260
+    },
+    {
+      "epoch": 2.3499310430003497,
+      "grad_norm": 2.1244235038757324,
+      "learning_rate": 2.3657740125296845e-06,
+      "loss": 0.2656,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3515777773203514,
+      "grad_norm": 1.7785730361938477,
+      "learning_rate": 2.354303778895911e-06,
+      "loss": 0.2606,
+      "step": 14280
+    },
+    {
+      "epoch": 2.353224511640353,
+      "grad_norm": 1.852630615234375,
+      "learning_rate": 2.3428577079511304e-06,
+      "loss": 0.2656,
+      "step": 14290
+    },
+    {
+      "epoch": 2.354871245960355,
+      "grad_norm": 1.7617170810699463,
+      "learning_rate": 2.331435835868349e-06,
+      "loss": 0.2685,
+      "step": 14300
+    },
+    {
+      "epoch": 2.3565179802803566,
+      "grad_norm": 1.348222017288208,
+      "learning_rate": 2.3200381987441067e-06,
+      "loss": 0.2524,
+      "step": 14310
+    },
+    {
+      "epoch": 2.358164714600358,
+      "grad_norm": 1.567896842956543,
+      "learning_rate": 2.308664832598343e-06,
+      "loss": 0.2574,
+      "step": 14320
+    },
+    {
+      "epoch": 2.3598114489203597,
+      "grad_norm": 2.113232374191284,
+      "learning_rate": 2.2973157733743055e-06,
+      "loss": 0.2589,
+      "step": 14330
+    },
+    {
+      "epoch": 2.3614581832403614,
+      "grad_norm": 2.2120754718780518,
+      "learning_rate": 2.285991056938418e-06,
+      "loss": 0.2562,
+      "step": 14340
+    },
+    {
+      "epoch": 2.363104917560363,
+      "grad_norm": 1.2926445007324219,
+      "learning_rate": 2.2746907190801724e-06,
+      "loss": 0.2383,
+      "step": 14350
+    },
+    {
+      "epoch": 2.364751651880365,
+      "grad_norm": 2.4468600749969482,
+      "learning_rate": 2.2634147955120176e-06,
+      "loss": 0.2621,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3663983862003666,
+      "grad_norm": 1.7285964488983154,
+      "learning_rate": 2.252163321869254e-06,
+      "loss": 0.2619,
+      "step": 14370
+    },
+    {
+      "epoch": 2.368045120520368,
+      "grad_norm": 1.6784744262695312,
+      "learning_rate": 2.2409363337099e-06,
+      "loss": 0.2551,
+      "step": 14380
+    },
+    {
+      "epoch": 2.3696918548403696,
+      "grad_norm": 1.5659254789352417,
+      "learning_rate": 2.229733866514605e-06,
+      "loss": 0.2449,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3713385891603713,
+      "grad_norm": 1.6228045225143433,
+      "learning_rate": 2.2185559556865145e-06,
+      "loss": 0.2598,
+      "step": 14400
+    },
+    {
+      "epoch": 2.372985323480373,
+      "grad_norm": 1.4725788831710815,
+      "learning_rate": 2.20740263655118e-06,
+      "loss": 0.2619,
+      "step": 14410
+    },
+    {
+      "epoch": 2.374632057800375,
+      "grad_norm": 1.6542408466339111,
+      "learning_rate": 2.1962739443564196e-06,
+      "loss": 0.2548,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3762787921203765,
+      "grad_norm": 1.6933832168579102,
+      "learning_rate": 2.1851699142722395e-06,
+      "loss": 0.2413,
+      "step": 14430
+    },
+    {
+      "epoch": 2.377925526440378,
+      "grad_norm": 1.9051034450531006,
+      "learning_rate": 2.1740905813906945e-06,
+      "loss": 0.2533,
+      "step": 14440
+    },
+    {
+      "epoch": 2.3795722607603795,
+      "grad_norm": 1.8932108879089355,
+      "learning_rate": 2.1630359807257962e-06,
+      "loss": 0.2603,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3812189950803813,
+      "grad_norm": 2.07220721244812,
+      "learning_rate": 2.1520061472133903e-06,
+      "loss": 0.2544,
+      "step": 14460
+    },
+    {
+      "epoch": 2.382865729400383,
+      "grad_norm": 1.947472333908081,
+      "learning_rate": 2.1410011157110556e-06,
+      "loss": 0.2462,
+      "step": 14470
+    },
+    {
+      "epoch": 2.3845124637203847,
+      "grad_norm": 2.6550047397613525,
+      "learning_rate": 2.130020920997985e-06,
+      "loss": 0.2725,
+      "step": 14480
+    },
+    {
+      "epoch": 2.386159198040386,
+      "grad_norm": 1.6511452198028564,
+      "learning_rate": 2.11906559777488e-06,
+      "loss": 0.26,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3878059323603877,
+      "grad_norm": 1.7327511310577393,
+      "learning_rate": 2.1081351806638395e-06,
+      "loss": 0.2479,
+      "step": 14500
+    },
+    {
+      "epoch": 2.3892879932483893,
+      "grad_norm": 1.907187819480896,
+      "learning_rate": 2.09722970420826e-06,
+      "loss": 0.2281,
+      "step": 14510
+    },
+    {
+      "epoch": 2.390934727568391,
+      "grad_norm": 1.4497332572937012,
+      "learning_rate": 2.086349202872705e-06,
+      "loss": 0.2465,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3925814618883927,
+      "grad_norm": 1.7492471933364868,
+      "learning_rate": 2.075493711042823e-06,
+      "loss": 0.2407,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3942281962083944,
+      "grad_norm": 2.5765883922576904,
+      "learning_rate": 2.0646632630252104e-06,
+      "loss": 0.2505,
+      "step": 14540
+    },
+    {
+      "epoch": 2.3958749305283957,
+      "grad_norm": 1.5902353525161743,
+      "learning_rate": 2.053857893047334e-06,
+      "loss": 0.2381,
+      "step": 14550
+    },
+    {
+      "epoch": 2.3975216648483975,
+      "grad_norm": 2.103832721710205,
+      "learning_rate": 2.0430776352573924e-06,
+      "loss": 0.2603,
+      "step": 14560
+    },
+    {
+      "epoch": 2.399168399168399,
+      "grad_norm": 2.955065965652466,
+      "learning_rate": 2.0323225237242285e-06,
+      "loss": 0.2414,
+      "step": 14570
+    },
+    {
+      "epoch": 2.400815133488401,
+      "grad_norm": 1.6870169639587402,
+      "learning_rate": 2.0215925924372126e-06,
+      "loss": 0.2196,
+      "step": 14580
+    },
+    {
+      "epoch": 2.4024618678084027,
+      "grad_norm": 1.891734004020691,
+      "learning_rate": 2.0108878753061434e-06,
+      "loss": 0.2278,
+      "step": 14590
+    },
+    {
+      "epoch": 2.404108602128404,
+      "grad_norm": 2.0214812755584717,
+      "learning_rate": 2.0002084061611282e-06,
+      "loss": 0.2463,
+      "step": 14600
+    },
+    {
+      "epoch": 2.4057553364484057,
+      "grad_norm": 1.8380804061889648,
+      "learning_rate": 1.9895542187524906e-06,
+      "loss": 0.2256,
+      "step": 14610
+    },
+    {
+      "epoch": 2.4074020707684074,
+      "grad_norm": 1.8675373792648315,
+      "learning_rate": 1.9789253467506474e-06,
+      "loss": 0.2442,
+      "step": 14620
+    },
+    {
+      "epoch": 2.409048805088409,
+      "grad_norm": 1.725472092628479,
+      "learning_rate": 1.9683218237460233e-06,
+      "loss": 0.2396,
+      "step": 14630
+    },
+    {
+      "epoch": 2.410695539408411,
+      "grad_norm": 1.398227334022522,
+      "learning_rate": 1.9577436832489206e-06,
+      "loss": 0.2201,
+      "step": 14640
+    },
+    {
+      "epoch": 2.412342273728412,
+      "grad_norm": 2.566589593887329,
+      "learning_rate": 1.947190958689428e-06,
+      "loss": 0.2475,
+      "step": 14650
+    },
+    {
+      "epoch": 2.413989008048414,
+      "grad_norm": 1.501044511795044,
+      "learning_rate": 1.9366636834173193e-06,
+      "loss": 0.2488,
+      "step": 14660
+    },
+    {
+      "epoch": 2.4156357423684156,
+      "grad_norm": 1.5589001178741455,
+      "learning_rate": 1.926161890701934e-06,
+      "loss": 0.2375,
+      "step": 14670
+    },
+    {
+      "epoch": 2.4172824766884173,
+      "grad_norm": 1.3738040924072266,
+      "learning_rate": 1.915685613732079e-06,
+      "loss": 0.2353,
+      "step": 14680
+    },
+    {
+      "epoch": 2.418929211008419,
+      "grad_norm": 1.7642990350723267,
+      "learning_rate": 1.9052348856159298e-06,
+      "loss": 0.2333,
+      "step": 14690
+    },
+    {
+      "epoch": 2.4205759453284204,
+      "grad_norm": 1.651908040046692,
+      "learning_rate": 1.8948097393809127e-06,
+      "loss": 0.2347,
+      "step": 14700
+    },
+    {
+      "epoch": 2.422222679648422,
+      "grad_norm": 1.8113752603530884,
+      "learning_rate": 1.8844102079736114e-06,
+      "loss": 0.2444,
+      "step": 14710
+    },
+    {
+      "epoch": 2.423869413968424,
+      "grad_norm": 1.66728675365448,
+      "learning_rate": 1.8740363242596605e-06,
+      "loss": 0.222,
+      "step": 14720
+    },
+    {
+      "epoch": 2.4255161482884255,
+      "grad_norm": 1.575468897819519,
+      "learning_rate": 1.8636881210236346e-06,
+      "loss": 0.2395,
+      "step": 14730
+    },
+    {
+      "epoch": 2.4271628826084273,
+      "grad_norm": 1.5622098445892334,
+      "learning_rate": 1.8533656309689584e-06,
+      "loss": 0.2589,
+      "step": 14740
+    },
+    {
+      "epoch": 2.4288096169284286,
+      "grad_norm": 1.7539955377578735,
+      "learning_rate": 1.8430688867177882e-06,
+      "loss": 0.2442,
+      "step": 14750
+    },
+    {
+      "epoch": 2.4304563512484303,
+      "grad_norm": 1.620202898979187,
+      "learning_rate": 1.832797920810917e-06,
+      "loss": 0.2434,
+      "step": 14760
+    },
+    {
+      "epoch": 2.432103085568432,
+      "grad_norm": 1.9027595520019531,
+      "learning_rate": 1.822552765707676e-06,
+      "loss": 0.2541,
+      "step": 14770
+    },
+    {
+      "epoch": 2.4337498198884338,
+      "grad_norm": 1.3224104642868042,
+      "learning_rate": 1.8123334537858195e-06,
+      "loss": 0.2264,
+      "step": 14780
+    },
+    {
+      "epoch": 2.4353965542084355,
+      "grad_norm": 2.0354669094085693,
+      "learning_rate": 1.8021400173414306e-06,
+      "loss": 0.237,
+      "step": 14790
+    },
+    {
+      "epoch": 2.437043288528437,
+      "grad_norm": 1.2932946681976318,
+      "learning_rate": 1.7919724885888256e-06,
+      "loss": 0.2355,
+      "step": 14800
+    },
+    {
+      "epoch": 2.4386900228484385,
+      "grad_norm": 1.3264442682266235,
+      "learning_rate": 1.781830899660434e-06,
+      "loss": 0.2237,
+      "step": 14810
+    },
+    {
+      "epoch": 2.4403367571684402,
+      "grad_norm": 1.640587329864502,
+      "learning_rate": 1.7717152826067175e-06,
+      "loss": 0.2311,
+      "step": 14820
+    },
+    {
+      "epoch": 2.441983491488442,
+      "grad_norm": 1.6640249490737915,
+      "learning_rate": 1.7616256693960532e-06,
+      "loss": 0.2193,
+      "step": 14830
+    },
+    {
+      "epoch": 2.4436302258084437,
+      "grad_norm": 1.4918125867843628,
+      "learning_rate": 1.751562091914637e-06,
+      "loss": 0.2356,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4452769601284454,
+      "grad_norm": 1.52755606174469,
+      "learning_rate": 1.7415245819663861e-06,
+      "loss": 0.2266,
+      "step": 14850
+    },
+    {
+      "epoch": 2.446923694448447,
+      "grad_norm": 2.0704216957092285,
+      "learning_rate": 1.7315131712728417e-06,
+      "loss": 0.2535,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4485704287684484,
+      "grad_norm": 1.83346426486969,
+      "learning_rate": 1.7215278914730527e-06,
+      "loss": 0.2363,
+      "step": 14870
+    },
+    {
+      "epoch": 2.45021716308845,
+      "grad_norm": 1.4527473449707031,
+      "learning_rate": 1.7115687741234987e-06,
+      "loss": 0.2296,
+      "step": 14880
+    },
+    {
+      "epoch": 2.451863897408452,
+      "grad_norm": 1.9535012245178223,
+      "learning_rate": 1.7016358506979657e-06,
+      "loss": 0.2336,
+      "step": 14890
+    },
+    {
+      "epoch": 2.4535106317284536,
+      "grad_norm": 2.087193250656128,
+      "learning_rate": 1.6917291525874723e-06,
+      "loss": 0.2527,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4551573660484554,
+      "grad_norm": 1.9018986225128174,
+      "learning_rate": 1.681848711100146e-06,
+      "loss": 0.2463,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4568041003684566,
+      "grad_norm": 1.6955307722091675,
+      "learning_rate": 1.6719945574611418e-06,
+      "loss": 0.2348,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4584508346884584,
+      "grad_norm": 1.538035273551941,
+      "learning_rate": 1.6621667228125305e-06,
+      "loss": 0.2511,
+      "step": 14930
+    },
+    {
+      "epoch": 2.46009756900846,
+      "grad_norm": 1.6792057752609253,
+      "learning_rate": 1.6523652382132183e-06,
+      "loss": 0.2592,
+      "step": 14940
+    },
+    {
+      "epoch": 2.461744303328462,
+      "grad_norm": 1.4391840696334839,
+      "learning_rate": 1.6425901346388263e-06,
+      "loss": 0.2233,
+      "step": 14950
+    },
+    {
+      "epoch": 2.4633910376484636,
+      "grad_norm": 1.333770990371704,
+      "learning_rate": 1.6328414429816109e-06,
+      "loss": 0.2303,
+      "step": 14960
+    },
+    {
+      "epoch": 2.465037771968465,
+      "grad_norm": 2.2043004035949707,
+      "learning_rate": 1.6231191940503543e-06,
+      "loss": 0.2294,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4666845062884666,
+      "grad_norm": 1.5073317289352417,
+      "learning_rate": 1.613423418570279e-06,
+      "loss": 0.2368,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4683312406084683,
+      "grad_norm": 1.6827707290649414,
+      "learning_rate": 1.6037541471829288e-06,
+      "loss": 0.238,
+      "step": 14990
+    },
+    {
+      "epoch": 2.46997797492847,
+      "grad_norm": 1.4171675443649292,
+      "learning_rate": 1.594111410446104e-06,
+      "loss": 0.2162,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4716247092484718,
+      "grad_norm": 1.938513994216919,
+      "learning_rate": 1.5844952388337332e-06,
+      "loss": 0.269,
+      "step": 15010
+    },
+    {
+      "epoch": 2.473271443568473,
+      "grad_norm": 1.7886428833007812,
+      "learning_rate": 1.574905662735805e-06,
+      "loss": 0.2393,
+      "step": 15020
+    },
+    {
+      "epoch": 2.474918177888475,
+      "grad_norm": 1.4953855276107788,
+      "learning_rate": 1.5653427124582431e-06,
+      "loss": 0.2265,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4765649122084765,
+      "grad_norm": 1.8918662071228027,
+      "learning_rate": 1.5558064182228393e-06,
+      "loss": 0.2295,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4782116465284783,
+      "grad_norm": 1.7208409309387207,
+      "learning_rate": 1.5462968101671361e-06,
+      "loss": 0.2429,
+      "step": 15050
+    },
+    {
+      "epoch": 2.47985838084848,
+      "grad_norm": 1.5308380126953125,
+      "learning_rate": 1.5368139183443421e-06,
+      "loss": 0.2396,
+      "step": 15060
+    },
+    {
+      "epoch": 2.4815051151684813,
+      "grad_norm": 1.626708984375,
+      "learning_rate": 1.5273577727232314e-06,
+      "loss": 0.2398,
+      "step": 15070
+    },
+    {
+      "epoch": 2.483151849488483,
+      "grad_norm": 1.9254333972930908,
+      "learning_rate": 1.5179284031880603e-06,
+      "loss": 0.2288,
+      "step": 15080
+    },
+    {
+      "epoch": 2.4847985838084847,
+      "grad_norm": 2.0169408321380615,
+      "learning_rate": 1.5085258395384538e-06,
+      "loss": 0.2416,
+      "step": 15090
+    },
+    {
+      "epoch": 2.4864453181284865,
+      "grad_norm": 1.554512858390808,
+      "learning_rate": 1.4991501114893336e-06,
+      "loss": 0.2522,
+      "step": 15100
+    },
+    {
+      "epoch": 2.488092052448488,
+      "grad_norm": 1.447223424911499,
+      "learning_rate": 1.4898012486708024e-06,
+      "loss": 0.2248,
+      "step": 15110
+    },
+    {
+      "epoch": 2.48973878676849,
+      "grad_norm": 1.233121395111084,
+      "learning_rate": 1.48047928062807e-06,
+      "loss": 0.2216,
+      "step": 15120
+    },
+    {
+      "epoch": 2.491385521088491,
+      "grad_norm": 1.4239838123321533,
+      "learning_rate": 1.4711842368213437e-06,
+      "loss": 0.2151,
+      "step": 15130
+    },
+    {
+      "epoch": 2.493032255408493,
+      "grad_norm": 1.5615428686141968,
+      "learning_rate": 1.4619161466257459e-06,
+      "loss": 0.2303,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4946789897284947,
+      "grad_norm": 1.5079021453857422,
+      "learning_rate": 1.4526750393312118e-06,
+      "loss": 0.2295,
+      "step": 15150
+    },
+    {
+      "epoch": 2.4963257240484964,
+      "grad_norm": 1.7538816928863525,
+      "learning_rate": 1.443460944142413e-06,
+      "loss": 0.2289,
+      "step": 15160
+    },
+    {
+      "epoch": 2.497972458368498,
+      "grad_norm": 1.7975422143936157,
+      "learning_rate": 1.4342738901786434e-06,
+      "loss": 0.234,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4996191926885,
+      "grad_norm": 1.8438105583190918,
+      "learning_rate": 1.4251139064737485e-06,
+      "loss": 0.239,
+      "step": 15180
+    },
+    {
+      "epoch": 2.501265927008501,
+      "grad_norm": 4.315160274505615,
+      "learning_rate": 1.4159810219760161e-06,
+      "loss": 0.2261,
+      "step": 15190
+    },
+    {
+      "epoch": 2.502912661328503,
+      "grad_norm": 1.4963452816009521,
+      "learning_rate": 1.406875265548101e-06,
+      "loss": 0.2228,
+      "step": 15200
+    },
+    {
+      "epoch": 2.5045593956485046,
+      "grad_norm": 1.717155933380127,
+      "learning_rate": 1.3977966659669096e-06,
+      "loss": 0.2385,
+      "step": 15210
+    },
+    {
+      "epoch": 2.5062061299685063,
+      "grad_norm": 1.8200072050094604,
+      "learning_rate": 1.3887452519235434e-06,
+      "loss": 0.2179,
+      "step": 15220
+    },
+    {
+      "epoch": 2.507852864288508,
+      "grad_norm": 1.7720727920532227,
+      "learning_rate": 1.379721052023174e-06,
+      "loss": 0.2252,
+      "step": 15230
+    },
+    {
+      "epoch": 2.5094995986085094,
+      "grad_norm": 1.5537362098693848,
+      "learning_rate": 1.3707240947849797e-06,
+      "loss": 0.2273,
+      "step": 15240
+    },
+    {
+      "epoch": 2.511146332928511,
+      "grad_norm": 1.4824028015136719,
+      "learning_rate": 1.3617544086420353e-06,
+      "loss": 0.241,
+      "step": 15250
+    },
+    {
+      "epoch": 2.512793067248513,
+      "grad_norm": 1.6623420715332031,
+      "learning_rate": 1.3528120219412377e-06,
+      "loss": 0.2494,
+      "step": 15260
+    },
+    {
+      "epoch": 2.5144398015685145,
+      "grad_norm": 1.8645778894424438,
+      "learning_rate": 1.3438969629432042e-06,
+      "loss": 0.2561,
+      "step": 15270
+    },
+    {
+      "epoch": 2.5160865358885163,
+      "grad_norm": 2.436868190765381,
+      "learning_rate": 1.335009259822191e-06,
+      "loss": 0.2311,
+      "step": 15280
+    },
+    {
+      "epoch": 2.5177332702085176,
+      "grad_norm": 1.7088630199432373,
+      "learning_rate": 1.3261489406659978e-06,
+      "loss": 0.2494,
+      "step": 15290
+    },
+    {
+      "epoch": 2.5193800045285193,
+      "grad_norm": 1.6420857906341553,
+      "learning_rate": 1.3173160334758895e-06,
+      "loss": 0.2465,
+      "step": 15300
+    },
+    {
+      "epoch": 2.521026738848521,
+      "grad_norm": 1.4090502262115479,
+      "learning_rate": 1.3085105661664933e-06,
+      "loss": 0.2263,
+      "step": 15310
+    },
+    {
+      "epoch": 2.5226734731685228,
+      "grad_norm": 1.924774169921875,
+      "learning_rate": 1.2997325665657257e-06,
+      "loss": 0.2348,
+      "step": 15320
+    },
+    {
+      "epoch": 2.5243202074885245,
+      "grad_norm": 1.5904576778411865,
+      "learning_rate": 1.2909820624146908e-06,
+      "loss": 0.2272,
+      "step": 15330
+    },
+    {
+      "epoch": 2.5259669418085258,
+      "grad_norm": 1.8772406578063965,
+      "learning_rate": 1.282259081367606e-06,
+      "loss": 0.2255,
+      "step": 15340
+    },
+    {
+      "epoch": 2.5276136761285275,
+      "grad_norm": 1.8594352006912231,
+      "learning_rate": 1.273563650991696e-06,
+      "loss": 0.2329,
+      "step": 15350
+    },
+    {
+      "epoch": 2.5292604104485292,
+      "grad_norm": 1.4280927181243896,
+      "learning_rate": 1.2648957987671295e-06,
+      "loss": 0.234,
+      "step": 15360
+    },
+    {
+      "epoch": 2.530907144768531,
+      "grad_norm": 1.7298115491867065,
+      "learning_rate": 1.256255552086909e-06,
+      "loss": 0.2416,
+      "step": 15370
+    },
+    {
+      "epoch": 2.5325538790885327,
+      "grad_norm": 1.3822236061096191,
+      "learning_rate": 1.2476429382568067e-06,
+      "loss": 0.2454,
+      "step": 15380
+    },
+    {
+      "epoch": 2.534200613408534,
+      "grad_norm": 1.9294795989990234,
+      "learning_rate": 1.2390579844952565e-06,
+      "loss": 0.2421,
+      "step": 15390
+    },
+    {
+      "epoch": 2.5358473477285357,
+      "grad_norm": 1.349674940109253,
+      "learning_rate": 1.2305007179332851e-06,
+      "loss": 0.2348,
+      "step": 15400
+    },
+    {
+      "epoch": 2.5374940820485374,
+      "grad_norm": 1.3522496223449707,
+      "learning_rate": 1.2219711656144161e-06,
+      "loss": 0.2404,
+      "step": 15410
+    },
+    {
+      "epoch": 2.539140816368539,
+      "grad_norm": 1.3507784605026245,
+      "learning_rate": 1.2134693544945875e-06,
+      "loss": 0.246,
+      "step": 15420
+    },
+    {
+      "epoch": 2.540787550688541,
+      "grad_norm": 1.6919246912002563,
+      "learning_rate": 1.2049953114420654e-06,
+      "loss": 0.2455,
+      "step": 15430
+    },
+    {
+      "epoch": 2.542434285008542,
+      "grad_norm": 1.62079918384552,
+      "learning_rate": 1.1965490632373677e-06,
+      "loss": 0.2317,
+      "step": 15440
+    },
+    {
+      "epoch": 2.5440810193285444,
+      "grad_norm": 1.7440752983093262,
+      "learning_rate": 1.1881306365731638e-06,
+      "loss": 0.2322,
+      "step": 15450
+    },
+    {
+      "epoch": 2.5457277536485456,
+      "grad_norm": 1.5291510820388794,
+      "learning_rate": 1.179740058054204e-06,
+      "loss": 0.2287,
+      "step": 15460
+    },
+    {
+      "epoch": 2.5473744879685474,
+      "grad_norm": 1.448723316192627,
+      "learning_rate": 1.1713773541972263e-06,
+      "loss": 0.2278,
+      "step": 15470
+    },
+    {
+      "epoch": 2.549021222288549,
+      "grad_norm": 1.5722907781600952,
+      "learning_rate": 1.1630425514308819e-06,
+      "loss": 0.2366,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5506679566085504,
+      "grad_norm": 1.579613447189331,
+      "learning_rate": 1.1547356760956397e-06,
+      "loss": 0.2361,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5523146909285526,
+      "grad_norm": 1.437633991241455,
+      "learning_rate": 1.1464567544437144e-06,
+      "loss": 0.2245,
+      "step": 15500
+    },
+    {
+      "epoch": 2.553961425248554,
+      "grad_norm": 1.626078724861145,
+      "learning_rate": 1.138205812638975e-06,
+      "loss": 0.2505,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5556081595685556,
+      "grad_norm": 1.7124260663986206,
+      "learning_rate": 1.12998287675687e-06,
+      "loss": 0.2494,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5572548938885573,
+      "grad_norm": 1.749879002571106,
+      "learning_rate": 1.1217879727843351e-06,
+      "loss": 0.2194,
+      "step": 15530
+    },
+    {
+      "epoch": 2.558901628208559,
+      "grad_norm": 2.109128713607788,
+      "learning_rate": 1.113621126619725e-06,
+      "loss": 0.2321,
+      "step": 15540
+    },
+    {
+      "epoch": 2.5605483625285608,
+      "grad_norm": 1.6905772686004639,
+      "learning_rate": 1.1054823640727163e-06,
+      "loss": 0.2457,
+      "step": 15550
+    },
+    {
+      "epoch": 2.562195096848562,
+      "grad_norm": 1.2987953424453735,
+      "learning_rate": 1.0973717108642323e-06,
+      "loss": 0.2325,
+      "step": 15560
+    },
+    {
+      "epoch": 2.563841831168564,
+      "grad_norm": 1.5485819578170776,
+      "learning_rate": 1.0892891926263703e-06,
+      "loss": 0.2462,
+      "step": 15570
+    },
+    {
+      "epoch": 2.5654885654885655,
+      "grad_norm": 1.8916813135147095,
+      "learning_rate": 1.0812348349023038e-06,
+      "loss": 0.2335,
+      "step": 15580
+    },
+    {
+      "epoch": 2.5671352998085673,
+      "grad_norm": 1.2640624046325684,
+      "learning_rate": 1.073208663146218e-06,
+      "loss": 0.2301,
+      "step": 15590
+    },
+    {
+      "epoch": 2.568782034128569,
+      "grad_norm": 1.434807300567627,
+      "learning_rate": 1.065210702723215e-06,
+      "loss": 0.2167,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5704287684485703,
+      "grad_norm": 1.6343811750411987,
+      "learning_rate": 1.0572409789092452e-06,
+      "loss": 0.2227,
+      "step": 15610
+    },
+    {
+      "epoch": 2.572075502768572,
+      "grad_norm": 1.6789041757583618,
+      "learning_rate": 1.0492995168910225e-06,
+      "loss": 0.2451,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5737222370885737,
+      "grad_norm": 1.42600417137146,
+      "learning_rate": 1.0413863417659454e-06,
+      "loss": 0.23,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5753689714085755,
+      "grad_norm": 1.9068646430969238,
+      "learning_rate": 1.0335014785420128e-06,
+      "loss": 0.2413,
+      "step": 15640
+    },
+    {
+      "epoch": 2.577015705728577,
+      "grad_norm": 1.5069235563278198,
+      "learning_rate": 1.0256449521377565e-06,
+      "loss": 0.2581,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5786624400485785,
+      "grad_norm": 1.8914705514907837,
+      "learning_rate": 1.0178167873821487e-06,
+      "loss": 0.2205,
+      "step": 15660
+    },
+    {
+      "epoch": 2.58030917436858,
+      "grad_norm": 1.843311071395874,
+      "learning_rate": 1.0100170090145379e-06,
+      "loss": 0.2357,
+      "step": 15670
+    },
+    {
+      "epoch": 2.581955908688582,
+      "grad_norm": 1.3863564729690552,
+      "learning_rate": 1.0022456416845561e-06,
+      "loss": 0.222,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5836026430085837,
+      "grad_norm": 1.5262445211410522,
+      "learning_rate": 9.945027099520489e-07,
+      "loss": 0.2182,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5852493773285854,
+      "grad_norm": 1.4200090169906616,
+      "learning_rate": 9.86788238287003e-07,
+      "loss": 0.2287,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5868961116485867,
+      "grad_norm": 1.338196039199829,
+      "learning_rate": 9.79102251069456e-07,
+      "loss": 0.2213,
+      "step": 15710
+    },
+    {
+      "epoch": 2.5885428459685884,
+      "grad_norm": 1.5565379858016968,
+      "learning_rate": 9.71444772589426e-07,
+      "loss": 0.2243,
+      "step": 15720
+    },
+    {
+      "epoch": 2.59018958028859,
+      "grad_norm": 1.8993768692016602,
+      "learning_rate": 9.638158270468423e-07,
+      "loss": 0.2194,
+      "step": 15730
+    },
+    {
+      "epoch": 2.591836314608592,
+      "grad_norm": 1.739005446434021,
+      "learning_rate": 9.56215438551452e-07,
+      "loss": 0.2241,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5934830489285936,
+      "grad_norm": 1.7477450370788574,
+      "learning_rate": 9.486436311227631e-07,
+      "loss": 0.2443,
+      "step": 15750
+    },
+    {
+      "epoch": 2.595129783248595,
+      "grad_norm": 1.8742027282714844,
+      "learning_rate": 9.411004286899495e-07,
+      "loss": 0.2386,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5967765175685966,
+      "grad_norm": 1.9991704225540161,
+      "learning_rate": 9.335858550917942e-07,
+      "loss": 0.2306,
+      "step": 15770
+    },
+    {
+      "epoch": 2.5984232518885984,
+      "grad_norm": 1.8734098672866821,
+      "learning_rate": 9.26099934076593e-07,
+      "loss": 0.2307,
+      "step": 15780
+    },
+    {
+      "epoch": 2.6000699862086,
+      "grad_norm": 1.9999672174453735,
+      "learning_rate": 9.186426893021016e-07,
+      "loss": 0.2242,
+      "step": 15790
+    },
+    {
+      "epoch": 2.601716720528602,
+      "grad_norm": 1.490793228149414,
+      "learning_rate": 9.112141443354439e-07,
+      "loss": 0.2493,
+      "step": 15800
+    },
+    {
+      "epoch": 2.603363454848603,
+      "grad_norm": 1.619504690170288,
+      "learning_rate": 9.038143226530482e-07,
+      "loss": 0.2458,
+      "step": 15810
+    },
+    {
+      "epoch": 2.6050101891686053,
+      "grad_norm": 1.3605459928512573,
+      "learning_rate": 8.964432476405638e-07,
+      "loss": 0.2279,
+      "step": 15820
+    },
+    {
+      "epoch": 2.6066569234886066,
+      "grad_norm": 1.5391523838043213,
+      "learning_rate": 8.891009425927977e-07,
+      "loss": 0.2332,
+      "step": 15830
+    },
+    {
+      "epoch": 2.6083036578086083,
+      "grad_norm": 1.6004542112350464,
+      "learning_rate": 8.817874307136298e-07,
+      "loss": 0.2404,
+      "step": 15840
+    },
+    {
+      "epoch": 2.60995039212861,
+      "grad_norm": 1.6447652578353882,
+      "learning_rate": 8.745027351159486e-07,
+      "loss": 0.2363,
+      "step": 15850
+    },
+    {
+      "epoch": 2.6115971264486117,
+      "grad_norm": 2.2156291007995605,
+      "learning_rate": 8.672468788215682e-07,
+      "loss": 0.2413,
+      "step": 15860
+    },
+    {
+      "epoch": 2.6132438607686135,
+      "grad_norm": 1.7267791032791138,
+      "learning_rate": 8.60019884761173e-07,
+      "loss": 0.2444,
+      "step": 15870
+    },
+    {
+      "epoch": 2.6148905950886148,
+      "grad_norm": 1.5263491868972778,
+      "learning_rate": 8.52821775774223e-07,
+      "loss": 0.227,
+      "step": 15880
+    },
+    {
+      "epoch": 2.6165373294086165,
+      "grad_norm": 1.5678350925445557,
+      "learning_rate": 8.456525746089017e-07,
+      "loss": 0.2185,
+      "step": 15890
+    },
+    {
+      "epoch": 2.6181840637286182,
+      "grad_norm": 2.008378744125366,
+      "learning_rate": 8.385123039220277e-07,
+      "loss": 0.237,
+      "step": 15900
+    },
+    {
+      "epoch": 2.61983079804862,
+      "grad_norm": 2.221608877182007,
+      "learning_rate": 8.314009862789984e-07,
+      "loss": 0.2433,
+      "step": 15910
+    },
+    {
+      "epoch": 2.6214775323686217,
+      "grad_norm": 2.2678582668304443,
+      "learning_rate": 8.243186441536999e-07,
+      "loss": 0.2457,
+      "step": 15920
+    },
+    {
+      "epoch": 2.623124266688623,
+      "grad_norm": 1.7668777704238892,
+      "learning_rate": 8.172652999284592e-07,
+      "loss": 0.2409,
+      "step": 15930
+    },
+    {
+      "epoch": 2.6247710010086247,
+      "grad_norm": 1.7608022689819336,
+      "learning_rate": 8.102409758939522e-07,
+      "loss": 0.243,
+      "step": 15940
+    },
+    {
+      "epoch": 2.6264177353286264,
+      "grad_norm": 2.060465097427368,
+      "learning_rate": 8.032456942491484e-07,
+      "loss": 0.2452,
+      "step": 15950
+    },
+    {
+      "epoch": 2.628064469648628,
+      "grad_norm": 1.3911464214324951,
+      "learning_rate": 7.962794771012284e-07,
+      "loss": 0.2331,
+      "step": 15960
+    },
+    {
+      "epoch": 2.62971120396863,
+      "grad_norm": 1.3499740362167358,
+      "learning_rate": 7.893423464655292e-07,
+      "loss": 0.2287,
+      "step": 15970
+    },
+    {
+      "epoch": 2.631357938288631,
+      "grad_norm": 1.5909963846206665,
+      "learning_rate": 7.824343242654564e-07,
+      "loss": 0.2338,
+      "step": 15980
+    },
+    {
+      "epoch": 2.633004672608633,
+      "grad_norm": 1.7845244407653809,
+      "learning_rate": 7.755554323324299e-07,
+      "loss": 0.233,
+      "step": 15990
+    },
+    {
+      "epoch": 2.6346514069286346,
+      "grad_norm": 1.5275846719741821,
+      "learning_rate": 7.687056924058056e-07,
+      "loss": 0.2209,
+      "step": 16000
+    },
+    {
+      "epoch": 2.6362981412486364,
+      "grad_norm": 1.519129753112793,
+      "learning_rate": 7.618851261328153e-07,
+      "loss": 0.227,
+      "step": 16010
+    },
+    {
+      "epoch": 2.637944875568638,
+      "grad_norm": 1.5210860967636108,
+      "learning_rate": 7.550937550684867e-07,
+      "loss": 0.2288,
+      "step": 16020
+    },
+    {
+      "epoch": 2.6395916098886394,
+      "grad_norm": 2.0805931091308594,
+      "learning_rate": 7.483316006755892e-07,
+      "loss": 0.234,
+      "step": 16030
+    },
+    {
+      "epoch": 2.641238344208641,
+      "grad_norm": 1.706473708152771,
+      "learning_rate": 7.415986843245515e-07,
+      "loss": 0.2407,
+      "step": 16040
+    },
+    {
+      "epoch": 2.642885078528643,
+      "grad_norm": 1.63625168800354,
+      "learning_rate": 7.348950272934107e-07,
+      "loss": 0.2213,
+      "step": 16050
+    },
+    {
+      "epoch": 2.6445318128486446,
+      "grad_norm": 1.5497370958328247,
+      "learning_rate": 7.282206507677225e-07,
+      "loss": 0.2214,
+      "step": 16060
+    },
+    {
+      "epoch": 2.6461785471686463,
+      "grad_norm": 1.570217490196228,
+      "learning_rate": 7.215755758405208e-07,
+      "loss": 0.2112,
+      "step": 16070
+    },
+    {
+      "epoch": 2.6478252814886476,
+      "grad_norm": 1.8267185688018799,
+      "learning_rate": 7.149598235122279e-07,
+      "loss": 0.224,
+      "step": 16080
+    },
+    {
+      "epoch": 2.6494720158086493,
+      "grad_norm": 1.6849004030227661,
+      "learning_rate": 7.08373414690604e-07,
+      "loss": 0.2273,
+      "step": 16090
+    },
+    {
+      "epoch": 2.651118750128651,
+      "grad_norm": 1.8440622091293335,
+      "learning_rate": 7.01816370190671e-07,
+      "loss": 0.2058,
+      "step": 16100
+    },
+    {
+      "epoch": 2.652765484448653,
+      "grad_norm": 1.6094481945037842,
+      "learning_rate": 6.952887107346551e-07,
+      "loss": 0.2358,
+      "step": 16110
+    },
+    {
+      "epoch": 2.6544122187686545,
+      "grad_norm": 1.7363183498382568,
+      "learning_rate": 6.887904569519133e-07,
+      "loss": 0.2254,
+      "step": 16120
+    },
+    {
+      "epoch": 2.656058953088656,
+      "grad_norm": 1.7017285823822021,
+      "learning_rate": 6.823216293788715e-07,
+      "loss": 0.2234,
+      "step": 16130
+    },
+    {
+      "epoch": 2.657705687408658,
+      "grad_norm": 1.396016240119934,
+      "learning_rate": 6.758822484589622e-07,
+      "loss": 0.232,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6593524217286593,
+      "grad_norm": 1.758091688156128,
+      "learning_rate": 6.69472334542558e-07,
+      "loss": 0.2362,
+      "step": 16150
+    },
+    {
+      "epoch": 2.660999156048661,
+      "grad_norm": 1.6662510633468628,
+      "learning_rate": 6.630919078869036e-07,
+      "loss": 0.2182,
+      "step": 16160
+    },
+    {
+      "epoch": 2.6626458903686627,
+      "grad_norm": 2.1852259635925293,
+      "learning_rate": 6.567409886560605e-07,
+      "loss": 0.2109,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6642926246886645,
+      "grad_norm": 1.657429575920105,
+      "learning_rate": 6.504195969208315e-07,
+      "loss": 0.2273,
+      "step": 16180
+    },
+    {
+      "epoch": 2.665939359008666,
+      "grad_norm": 2.1529386043548584,
+      "learning_rate": 6.44127752658712e-07,
+      "loss": 0.2448,
+      "step": 16190
+    },
+    {
+      "epoch": 2.6675860933286675,
+      "grad_norm": 1.56782066822052,
+      "learning_rate": 6.378654757538072e-07,
+      "loss": 0.2201,
+      "step": 16200
+    },
+    {
+      "epoch": 2.669232827648669,
+      "grad_norm": 2.715771198272705,
+      "learning_rate": 6.316327859967907e-07,
+      "loss": 0.2286,
+      "step": 16210
+    },
+    {
+      "epoch": 2.670879561968671,
+      "grad_norm": 1.7289623022079468,
+      "learning_rate": 6.254297030848255e-07,
+      "loss": 0.2243,
+      "step": 16220
+    },
+    {
+      "epoch": 2.6725262962886727,
+      "grad_norm": 1.6193941831588745,
+      "learning_rate": 6.192562466215135e-07,
+      "loss": 0.2266,
+      "step": 16230
+    },
+    {
+      "epoch": 2.6741730306086744,
+      "grad_norm": 1.712371826171875,
+      "learning_rate": 6.131124361168228e-07,
+      "loss": 0.221,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6758197649286757,
+      "grad_norm": 1.8381526470184326,
+      "learning_rate": 6.069982909870376e-07,
+      "loss": 0.2216,
+      "step": 16250
+    },
+    {
+      "epoch": 2.6774664992486774,
+      "grad_norm": 2.2945876121520996,
+      "learning_rate": 6.009138305546813e-07,
+      "loss": 0.2157,
+      "step": 16260
+    },
+    {
+      "epoch": 2.679113233568679,
+      "grad_norm": 1.9753563404083252,
+      "learning_rate": 5.948590740484783e-07,
+      "loss": 0.2298,
+      "step": 16270
+    },
+    {
+      "epoch": 2.680759967888681,
+      "grad_norm": 1.4538737535476685,
+      "learning_rate": 5.888340406032633e-07,
+      "loss": 0.2151,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6824067022086826,
+      "grad_norm": 1.7806299924850464,
+      "learning_rate": 5.828387492599507e-07,
+      "loss": 0.244,
+      "step": 16290
+    },
+    {
+      "epoch": 2.684053436528684,
+      "grad_norm": 1.6264704465866089,
+      "learning_rate": 5.768732189654535e-07,
+      "loss": 0.2418,
+      "step": 16300
+    },
+    {
+      "epoch": 2.6857001708486856,
+      "grad_norm": 1.9107524156570435,
+      "learning_rate": 5.709374685726365e-07,
+      "loss": 0.2453,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6873469051686873,
+      "grad_norm": 2.1297032833099365,
+      "learning_rate": 5.650315168402443e-07,
+      "loss": 0.2351,
+      "step": 16320
+    },
+    {
+      "epoch": 2.688993639488689,
+      "grad_norm": 5.344056606292725,
+      "learning_rate": 5.591553824328555e-07,
+      "loss": 0.2455,
+      "step": 16330
+    },
+    {
+      "epoch": 2.690640373808691,
+      "grad_norm": 1.7449675798416138,
+      "learning_rate": 5.533090839208133e-07,
+      "loss": 0.2377,
+      "step": 16340
+    },
+    {
+      "epoch": 2.692287108128692,
+      "grad_norm": 2.327143907546997,
+      "learning_rate": 5.474926397801705e-07,
+      "loss": 0.2251,
+      "step": 16350
+    },
+    {
+      "epoch": 2.693933842448694,
+      "grad_norm": 1.8635077476501465,
+      "learning_rate": 5.417060683926301e-07,
+      "loss": 0.2174,
+      "step": 16360
+    },
+    {
+      "epoch": 2.6955805767686956,
+      "grad_norm": 1.3951900005340576,
+      "learning_rate": 5.359493880454935e-07,
+      "loss": 0.2114,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6972273110886973,
+      "grad_norm": 2.1820077896118164,
+      "learning_rate": 5.302226169315927e-07,
+      "loss": 0.2318,
+      "step": 16380
+    },
+    {
+      "epoch": 2.698874045408699,
+      "grad_norm": 1.5952907800674438,
+      "learning_rate": 5.245257731492381e-07,
+      "loss": 0.2344,
+      "step": 16390
+    },
+    {
+      "epoch": 2.7005207797287003,
+      "grad_norm": 1.400417447090149,
+      "learning_rate": 5.188588747021628e-07,
+      "loss": 0.2343,
+      "step": 16400
+    },
+    {
+      "epoch": 2.702167514048702,
+      "grad_norm": 1.4244686365127563,
+      "learning_rate": 5.13221939499462e-07,
+      "loss": 0.2151,
+      "step": 16410
+    },
+    {
+      "epoch": 2.7038142483687038,
+      "grad_norm": 2.0728249549865723,
+      "learning_rate": 5.076149853555379e-07,
+      "loss": 0.2045,
+      "step": 16420
+    },
+    {
+      "epoch": 2.7054609826887055,
+      "grad_norm": 1.8236521482467651,
+      "learning_rate": 5.02038029990044e-07,
+      "loss": 0.2209,
+      "step": 16430
+    },
+    {
+      "epoch": 2.7071077170087072,
+      "grad_norm": 2.1654088497161865,
+      "learning_rate": 4.964910910278298e-07,
+      "loss": 0.2313,
+      "step": 16440
+    },
+    {
+      "epoch": 2.7087544513287085,
+      "grad_norm": 1.472296953201294,
+      "learning_rate": 4.909741859988837e-07,
+      "loss": 0.2372,
+      "step": 16450
+    },
+    {
+      "epoch": 2.7104011856487102,
+      "grad_norm": 1.6187572479248047,
+      "learning_rate": 4.854873323382747e-07,
+      "loss": 0.2295,
+      "step": 16460
+    },
+    {
+      "epoch": 2.712047919968712,
+      "grad_norm": 1.675091028213501,
+      "learning_rate": 4.800305473861056e-07,
+      "loss": 0.2211,
+      "step": 16470
+    },
+    {
+      "epoch": 2.7136946542887137,
+      "grad_norm": 1.6285738945007324,
+      "learning_rate": 4.7460384838744934e-07,
+      "loss": 0.2188,
+      "step": 16480
+    },
+    {
+      "epoch": 2.7153413886087154,
+      "grad_norm": 1.419481873512268,
+      "learning_rate": 4.692072524922975e-07,
+      "loss": 0.2209,
+      "step": 16490
+    },
+    {
+      "epoch": 2.7169881229287167,
+      "grad_norm": 1.381244421005249,
+      "learning_rate": 4.638407767555131e-07,
+      "loss": 0.2193,
+      "step": 16500
+    },
+    {
+      "epoch": 2.718634857248719,
+      "grad_norm": 1.5735642910003662,
+      "learning_rate": 4.585044381367609e-07,
+      "loss": 0.2397,
+      "step": 16510
+    },
+    {
+      "epoch": 2.72028159156872,
+      "grad_norm": 1.268129825592041,
+      "learning_rate": 4.531982535004731e-07,
+      "loss": 0.2109,
+      "step": 16520
+    },
+    {
+      "epoch": 2.721928325888722,
+      "grad_norm": 1.3103333711624146,
+      "learning_rate": 4.4792223961578006e-07,
+      "loss": 0.2041,
+      "step": 16530
+    },
+    {
+      "epoch": 2.7235750602087236,
+      "grad_norm": 1.9005833864212036,
+      "learning_rate": 4.4267641315646313e-07,
+      "loss": 0.2229,
+      "step": 16540
+    },
+    {
+      "epoch": 2.7252217945287254,
+      "grad_norm": 1.3628641366958618,
+      "learning_rate": 4.3746079070090765e-07,
+      "loss": 0.2132,
+      "step": 16550
+    },
+    {
+      "epoch": 2.726868528848727,
+      "grad_norm": 1.7968915700912476,
+      "learning_rate": 4.3227538873204076e-07,
+      "loss": 0.2321,
+      "step": 16560
+    },
+    {
+      "epoch": 2.7285152631687284,
+      "grad_norm": 1.7925100326538086,
+      "learning_rate": 4.271202236372829e-07,
+      "loss": 0.2242,
+      "step": 16570
+    },
+    {
+      "epoch": 2.73016199748873,
+      "grad_norm": 1.5291099548339844,
+      "learning_rate": 4.2199531170850296e-07,
+      "loss": 0.2206,
+      "step": 16580
+    },
+    {
+      "epoch": 2.731808731808732,
+      "grad_norm": 1.671055793762207,
+      "learning_rate": 4.1690066914195306e-07,
+      "loss": 0.2104,
+      "step": 16590
+    },
+    {
+      "epoch": 2.7334554661287336,
+      "grad_norm": 1.9849021434783936,
+      "learning_rate": 4.118363120382318e-07,
+      "loss": 0.2212,
+      "step": 16600
+    },
+    {
+      "epoch": 2.7351022004487353,
+      "grad_norm": 1.765097975730896,
+      "learning_rate": 4.0680225640222227e-07,
+      "loss": 0.2274,
+      "step": 16610
+    },
+    {
+      "epoch": 2.7367489347687366,
+      "grad_norm": 1.7051899433135986,
+      "learning_rate": 4.017985181430495e-07,
+      "loss": 0.2262,
+      "step": 16620
+    },
+    {
+      "epoch": 2.7383956690887383,
+      "grad_norm": 1.5623761415481567,
+      "learning_rate": 3.9682511307402083e-07,
+      "loss": 0.2442,
+      "step": 16630
+    },
+    {
+      "epoch": 2.74004240340874,
+      "grad_norm": 1.9426796436309814,
+      "learning_rate": 3.918820569125881e-07,
+      "loss": 0.2348,
+      "step": 16640
+    },
+    {
+      "epoch": 2.741689137728742,
+      "grad_norm": 1.656064510345459,
+      "learning_rate": 3.869693652802864e-07,
+      "loss": 0.232,
+      "step": 16650
+    },
+    {
+      "epoch": 2.7433358720487435,
+      "grad_norm": 1.2131900787353516,
+      "learning_rate": 3.820870537026944e-07,
+      "loss": 0.2123,
+      "step": 16660
+    },
+    {
+      "epoch": 2.744982606368745,
+      "grad_norm": 1.733933448791504,
+      "learning_rate": 3.7723513760937525e-07,
+      "loss": 0.2327,
+      "step": 16670
+    },
+    {
+      "epoch": 2.7466293406887465,
+      "grad_norm": 1.47609281539917,
+      "learning_rate": 3.7241363233384007e-07,
+      "loss": 0.2185,
+      "step": 16680
+    },
+    {
+      "epoch": 2.7482760750087483,
+      "grad_norm": 1.770021915435791,
+      "learning_rate": 3.6762255311348696e-07,
+      "loss": 0.2237,
+      "step": 16690
+    },
+    {
+      "epoch": 2.74992280932875,
+      "grad_norm": 1.9828838109970093,
+      "learning_rate": 3.628619150895607e-07,
+      "loss": 0.2295,
+      "step": 16700
+    },
+    {
+      "epoch": 2.7515695436487517,
+      "grad_norm": 2.307547092437744,
+      "learning_rate": 3.5813173330710215e-07,
+      "loss": 0.2399,
+      "step": 16710
+    },
+    {
+      "epoch": 2.753216277968753,
+      "grad_norm": 1.719611644744873,
+      "learning_rate": 3.534320227149035e-07,
+      "loss": 0.2246,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7548630122887547,
+      "grad_norm": 1.925981879234314,
+      "learning_rate": 3.48762798165454e-07,
+      "loss": 0.2115,
+      "step": 16730
+    },
+    {
+      "epoch": 2.7565097466087565,
+      "grad_norm": 1.7162208557128906,
+      "learning_rate": 3.441240744149055e-07,
+      "loss": 0.2397,
+      "step": 16740
+    },
+    {
+      "epoch": 2.758156480928758,
+      "grad_norm": 2.125133991241455,
+      "learning_rate": 3.3951586612300914e-07,
+      "loss": 0.2179,
+      "step": 16750
+    },
+    {
+      "epoch": 2.75980321524876,
+      "grad_norm": 1.792880654335022,
+      "learning_rate": 3.3493818785308886e-07,
+      "loss": 0.2181,
+      "step": 16760
+    },
+    {
+      "epoch": 2.761449949568761,
+      "grad_norm": 1.3869192600250244,
+      "learning_rate": 3.3039105407197127e-07,
+      "loss": 0.2357,
+      "step": 16770
+    },
+    {
+      "epoch": 2.763096683888763,
+      "grad_norm": 2.811720371246338,
+      "learning_rate": 3.2587447914996463e-07,
+      "loss": 0.2457,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7647434182087647,
+      "grad_norm": 1.611673355102539,
+      "learning_rate": 3.213884773607967e-07,
+      "loss": 0.2396,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7663901525287664,
+      "grad_norm": 2.732407808303833,
+      "learning_rate": 3.1693306288157697e-07,
+      "loss": 0.2192,
+      "step": 16800
+    },
+    {
+      "epoch": 2.768036886848768,
+      "grad_norm": 1.6391844749450684,
+      "learning_rate": 3.1250824979274675e-07,
+      "loss": 0.2464,
+      "step": 16810
+    },
+    {
+      "epoch": 2.7696836211687694,
+      "grad_norm": 1.7055604457855225,
+      "learning_rate": 3.0811405207804456e-07,
+      "loss": 0.2196,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7713303554887716,
+      "grad_norm": 1.6378053426742554,
+      "learning_rate": 3.0375048362444535e-07,
+      "loss": 0.2241,
+      "step": 16830
+    },
+    {
+      "epoch": 2.772977089808773,
+      "grad_norm": 1.4006156921386719,
+      "learning_rate": 2.9941755822213704e-07,
+      "loss": 0.2291,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7746238241287746,
+      "grad_norm": 1.677067756652832,
+      "learning_rate": 2.95115289564456e-07,
+      "loss": 0.2527,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7762705584487763,
+      "grad_norm": 1.372916340827942,
+      "learning_rate": 2.9084369124786293e-07,
+      "loss": 0.2291,
+      "step": 16860
+    },
+    {
+      "epoch": 2.777917292768778,
+      "grad_norm": 1.4055744409561157,
+      "learning_rate": 2.8660277677188487e-07,
+      "loss": 0.2199,
+      "step": 16870
+    },
+    {
+      "epoch": 2.77956402708878,
+      "grad_norm": 1.945945382118225,
+      "learning_rate": 2.8239255953908305e-07,
+      "loss": 0.2199,
+      "step": 16880
+    },
+    {
+      "epoch": 2.781210761408781,
+      "grad_norm": 1.2029650211334229,
+      "learning_rate": 2.782130528550031e-07,
+      "loss": 0.2292,
+      "step": 16890
+    },
+    {
+      "epoch": 2.782857495728783,
+      "grad_norm": 2.0258445739746094,
+      "learning_rate": 2.740642699281382e-07,
+      "loss": 0.2193,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7845042300487846,
+      "grad_norm": 1.2839769124984741,
+      "learning_rate": 2.699462238698847e-07,
+      "loss": 0.2206,
+      "step": 16910
+    },
+    {
+      "epoch": 2.7861509643687863,
+      "grad_norm": 1.4832942485809326,
+      "learning_rate": 2.6585892769450005e-07,
+      "loss": 0.2196,
+      "step": 16920
+    },
+    {
+      "epoch": 2.787797698688788,
+      "grad_norm": 1.6463637351989746,
+      "learning_rate": 2.6180239431906284e-07,
+      "loss": 0.2127,
+      "step": 16930
+    },
+    {
+      "epoch": 2.7894444330087893,
+      "grad_norm": 1.6157695055007935,
+      "learning_rate": 2.57776636563436e-07,
+      "loss": 0.2241,
+      "step": 16940
+    },
+    {
+      "epoch": 2.791091167328791,
+      "grad_norm": 1.478996753692627,
+      "learning_rate": 2.537816671502158e-07,
+      "loss": 0.2124,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7927379016487928,
+      "grad_norm": 1.402809739112854,
+      "learning_rate": 2.498174987047042e-07,
+      "loss": 0.2311,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7943846359687945,
+      "grad_norm": 1.756034016609192,
+      "learning_rate": 2.458841437548587e-07,
+      "loss": 0.208,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7960313702887962,
+      "grad_norm": 1.5966026782989502,
+      "learning_rate": 2.4198161473126147e-07,
+      "loss": 0.2216,
+      "step": 16980
+    },
+    {
+      "epoch": 2.7976781046087975,
+      "grad_norm": 1.536316990852356,
+      "learning_rate": 2.3810992396706812e-07,
+      "loss": 0.2261,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7993248389287992,
+      "grad_norm": 2.4802510738372803,
+      "learning_rate": 2.342690836979833e-07,
+      "loss": 0.2265,
+      "step": 17000
+    },
+    {
+      "epoch": 2.800971573248801,
+      "grad_norm": 1.9055966138839722,
+      "learning_rate": 2.3045910606221078e-07,
+      "loss": 0.2162,
+      "step": 17010
+    },
+    {
+      "epoch": 2.8026183075688027,
+      "grad_norm": 1.6244001388549805,
+      "learning_rate": 2.2668000310042237e-07,
+      "loss": 0.2292,
+      "step": 17020
+    },
+    {
+      "epoch": 2.8042650418888044,
+      "grad_norm": 1.7967705726623535,
+      "learning_rate": 2.2293178675571236e-07,
+      "loss": 0.2163,
+      "step": 17030
+    },
+    {
+      "epoch": 2.8059117762088057,
+      "grad_norm": 1.432999610900879,
+      "learning_rate": 2.1921446887356869e-07,
+      "loss": 0.2227,
+      "step": 17040
+    },
+    {
+      "epoch": 2.8075585105288074,
+      "grad_norm": 1.8171266317367554,
+      "learning_rate": 2.1552806120182734e-07,
+      "loss": 0.219,
+      "step": 17050
+    },
+    {
+      "epoch": 2.809205244848809,
+      "grad_norm": 1.6710107326507568,
+      "learning_rate": 2.1187257539064143e-07,
+      "loss": 0.2164,
+      "step": 17060
+    },
+    {
+      "epoch": 2.810851979168811,
+      "grad_norm": 2.254349708557129,
+      "learning_rate": 2.0824802299243775e-07,
+      "loss": 0.2159,
+      "step": 17070
+    },
+    {
+      "epoch": 2.8124987134888126,
+      "grad_norm": 1.3687024116516113,
+      "learning_rate": 2.0465441546189125e-07,
+      "loss": 0.2183,
+      "step": 17080
+    },
+    {
+      "epoch": 2.814145447808814,
+      "grad_norm": 1.6281770467758179,
+      "learning_rate": 2.0109176415587294e-07,
+      "loss": 0.2154,
+      "step": 17090
+    },
+    {
+      "epoch": 2.8157921821288157,
+      "grad_norm": 1.8856794834136963,
+      "learning_rate": 1.9756008033343211e-07,
+      "loss": 0.213,
+      "step": 17100
+    },
+    {
+      "epoch": 2.8174389164488174,
+      "grad_norm": 1.6289527416229248,
+      "learning_rate": 1.940593751557429e-07,
+      "loss": 0.2276,
+      "step": 17110
+    },
+    {
+      "epoch": 2.819085650768819,
+      "grad_norm": 1.4464855194091797,
+      "learning_rate": 1.9058965968608567e-07,
+      "loss": 0.2174,
+      "step": 17120
+    },
+    {
+      "epoch": 2.820732385088821,
+      "grad_norm": 1.4708118438720703,
+      "learning_rate": 1.8715094488979568e-07,
+      "loss": 0.2193,
+      "step": 17130
+    },
+    {
+      "epoch": 2.822379119408822,
+      "grad_norm": 1.6330119371414185,
+      "learning_rate": 1.837432416342444e-07,
+      "loss": 0.2153,
+      "step": 17140
+    },
+    {
+      "epoch": 2.824025853728824,
+      "grad_norm": 1.0777000188827515,
+      "learning_rate": 1.8036656068879166e-07,
+      "loss": 0.2284,
+      "step": 17150
+    },
+    {
+      "epoch": 2.8256725880488256,
+      "grad_norm": 1.7415492534637451,
+      "learning_rate": 1.770209127247635e-07,
+      "loss": 0.2217,
+      "step": 17160
+    },
+    {
+      "epoch": 2.8273193223688273,
+      "grad_norm": 1.8589420318603516,
+      "learning_rate": 1.7370630831540668e-07,
+      "loss": 0.2278,
+      "step": 17170
+    },
+    {
+      "epoch": 2.828966056688829,
+      "grad_norm": 1.300451636314392,
+      "learning_rate": 1.7042275793586416e-07,
+      "loss": 0.212,
+      "step": 17180
+    },
+    {
+      "epoch": 2.8306127910088303,
+      "grad_norm": 1.7537988424301147,
+      "learning_rate": 1.671702719631374e-07,
+      "loss": 0.222,
+      "step": 17190
+    },
+    {
+      "epoch": 2.8322595253288325,
+      "grad_norm": 1.643229365348816,
+      "learning_rate": 1.6394886067605752e-07,
+      "loss": 0.2198,
+      "step": 17200
+    },
+    {
+      "epoch": 2.833906259648834,
+      "grad_norm": 1.942588210105896,
+      "learning_rate": 1.6075853425524646e-07,
+      "loss": 0.2279,
+      "step": 17210
+    },
+    {
+      "epoch": 2.8355529939688355,
+      "grad_norm": 1.780501365661621,
+      "learning_rate": 1.5759930278309243e-07,
+      "loss": 0.2243,
+      "step": 17220
+    },
+    {
+      "epoch": 2.8371997282888373,
+      "grad_norm": 1.2578332424163818,
+      "learning_rate": 1.5447117624371122e-07,
+      "loss": 0.2193,
+      "step": 17230
+    },
+    {
+      "epoch": 2.838846462608839,
+      "grad_norm": 1.4017175436019897,
+      "learning_rate": 1.5137416452292164e-07,
+      "loss": 0.1921,
+      "step": 17240
+    },
+    {
+      "epoch": 2.8404931969288407,
+      "grad_norm": 1.6967748403549194,
+      "learning_rate": 1.4830827740820453e-07,
+      "loss": 0.233,
+      "step": 17250
+    },
+    {
+      "epoch": 2.842139931248842,
+      "grad_norm": 1.5690455436706543,
+      "learning_rate": 1.4527352458868494e-07,
+      "loss": 0.2263,
+      "step": 17260
+    },
+    {
+      "epoch": 2.8437866655688437,
+      "grad_norm": 2.2616426944732666,
+      "learning_rate": 1.4226991565508662e-07,
+      "loss": 0.2235,
+      "step": 17270
+    },
+    {
+      "epoch": 2.8454333998888455,
+      "grad_norm": 2.0758249759674072,
+      "learning_rate": 1.3929746009971434e-07,
+      "loss": 0.2137,
+      "step": 17280
+    },
+    {
+      "epoch": 2.847080134208847,
+      "grad_norm": 1.4178270101547241,
+      "learning_rate": 1.3635616731641933e-07,
+      "loss": 0.2044,
+      "step": 17290
+    },
+    {
+      "epoch": 2.848726868528849,
+      "grad_norm": 1.5788546800613403,
+      "learning_rate": 1.3344604660056494e-07,
+      "loss": 0.2198,
+      "step": 17300
+    },
+    {
+      "epoch": 2.85037360284885,
+      "grad_norm": 1.6083252429962158,
+      "learning_rate": 1.3056710714900334e-07,
+      "loss": 0.2158,
+      "step": 17310
+    },
+    {
+      "epoch": 2.852020337168852,
+      "grad_norm": 1.376654863357544,
+      "learning_rate": 1.2771935806004776e-07,
+      "loss": 0.2158,
+      "step": 17320
+    },
+    {
+      "epoch": 2.8536670714888537,
+      "grad_norm": 1.5676956176757812,
+      "learning_rate": 1.2490280833343694e-07,
+      "loss": 0.2127,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8553138058088554,
+      "grad_norm": 1.510061502456665,
+      "learning_rate": 1.2211746687030958e-07,
+      "loss": 0.2066,
+      "step": 17340
+    },
+    {
+      "epoch": 2.856960540128857,
+      "grad_norm": 1.0092027187347412,
+      "learning_rate": 1.1936334247318104e-07,
+      "loss": 0.2072,
+      "step": 17350
+    },
+    {
+      "epoch": 2.8586072744488584,
+      "grad_norm": 1.4783754348754883,
+      "learning_rate": 1.1664044384590679e-07,
+      "loss": 0.2055,
+      "step": 17360
+    },
+    {
+      "epoch": 2.86025400876886,
+      "grad_norm": 1.3680094480514526,
+      "learning_rate": 1.1394877959366223e-07,
+      "loss": 0.2041,
+      "step": 17370
+    },
+    {
+      "epoch": 2.861900743088862,
+      "grad_norm": 1.7390999794006348,
+      "learning_rate": 1.1128835822291406e-07,
+      "loss": 0.2109,
+      "step": 17380
+    },
+    {
+      "epoch": 2.8635474774088636,
+      "grad_norm": 1.5219812393188477,
+      "learning_rate": 1.0865918814138677e-07,
+      "loss": 0.2101,
+      "step": 17390
+    },
+    {
+      "epoch": 2.8651942117288653,
+      "grad_norm": 1.1699146032333374,
+      "learning_rate": 1.060612776580483e-07,
+      "loss": 0.227,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8668409460488666,
+      "grad_norm": 1.836059331893921,
+      "learning_rate": 1.0349463498307233e-07,
+      "loss": 0.2264,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8684876803688684,
+      "grad_norm": 1.6597439050674438,
+      "learning_rate": 1.009592682278171e-07,
+      "loss": 0.2205,
+      "step": 17420
+    },
+    {
+      "epoch": 2.87013441468887,
+      "grad_norm": 1.5050686597824097,
+      "learning_rate": 9.845518540480214e-08,
+      "loss": 0.2203,
+      "step": 17430
+    },
+    {
+      "epoch": 2.871781149008872,
+      "grad_norm": 1.6144781112670898,
+      "learning_rate": 9.598239442767721e-08,
+      "loss": 0.249,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8734278833288736,
+      "grad_norm": 1.542999029159546,
+      "learning_rate": 9.354090311120334e-08,
+      "loss": 0.2278,
+      "step": 17450
+    },
+    {
+      "epoch": 2.875074617648875,
+      "grad_norm": 1.3249629735946655,
+      "learning_rate": 9.113071917122407e-08,
+      "loss": 0.2044,
+      "step": 17460
+    },
+    {
+      "epoch": 2.8767213519688766,
+      "grad_norm": 1.4665296077728271,
+      "learning_rate": 8.875185022464094e-08,
+      "loss": 0.2165,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8783680862888783,
+      "grad_norm": 1.5788675546646118,
+      "learning_rate": 8.640430378939246e-08,
+      "loss": 0.2187,
+      "step": 17480
+    },
+    {
+      "epoch": 2.88001482060888,
+      "grad_norm": 1.677793025970459,
+      "learning_rate": 8.408808728442963e-08,
+      "loss": 0.2058,
+      "step": 17490
+    },
+    {
+      "epoch": 2.8816615549288818,
+      "grad_norm": 1.5491596460342407,
+      "learning_rate": 8.180320802968822e-08,
+      "loss": 0.2036,
+      "step": 17500
+    },
+    {
+      "epoch": 2.883308289248883,
+      "grad_norm": 1.7089343070983887,
+      "learning_rate": 7.95496732460721e-08,
+      "loss": 0.2199,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8849550235688852,
+      "grad_norm": 1.4340360164642334,
+      "learning_rate": 7.732749005542439e-08,
+      "loss": 0.2135,
+      "step": 17520
+    },
+    {
+      "epoch": 2.8866017578888865,
+      "grad_norm": 1.716632604598999,
+      "learning_rate": 7.51366654805108e-08,
+      "loss": 0.2367,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8882484922088882,
+      "grad_norm": 1.836976170539856,
+      "learning_rate": 7.297720644499073e-08,
+      "loss": 0.2203,
+      "step": 17540
+    },
+    {
+      "epoch": 2.88989522652889,
+      "grad_norm": 1.2886078357696533,
+      "learning_rate": 7.084911977340404e-08,
+      "loss": 0.2142,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8915419608488917,
+      "grad_norm": 1.4025774002075195,
+      "learning_rate": 6.875241219113982e-08,
+      "loss": 0.2067,
+      "step": 17560
+    },
+    {
+      "epoch": 2.8931886951688934,
+      "grad_norm": 1.7300608158111572,
+      "learning_rate": 6.66870903244221e-08,
+      "loss": 0.2204,
+      "step": 17570
+    },
+    {
+      "epoch": 2.8948354294888947,
+      "grad_norm": 1.5215460062026978,
+      "learning_rate": 6.465316070028538e-08,
+      "loss": 0.2343,
+      "step": 17580
+    },
+    {
+      "epoch": 2.8964821638088964,
+      "grad_norm": 1.8203786611557007,
+      "learning_rate": 6.265062974655789e-08,
+      "loss": 0.21,
+      "step": 17590
+    },
+    {
+      "epoch": 2.898128898128898,
+      "grad_norm": 1.4137710332870483,
+      "learning_rate": 6.067950379183619e-08,
+      "loss": 0.224,
+      "step": 17600
+    },
+    {
+      "epoch": 2.8997756324489,
+      "grad_norm": 1.8194389343261719,
+      "learning_rate": 5.87397890654684e-08,
+      "loss": 0.2254,
+      "step": 17610
+    },
+    {
+      "epoch": 2.9014223667689016,
+      "grad_norm": 2.260550022125244,
+      "learning_rate": 5.683149169753433e-08,
+      "loss": 0.2237,
+      "step": 17620
+    },
+    {
+      "epoch": 2.903069101088903,
+      "grad_norm": 1.4916115999221802,
+      "learning_rate": 5.4954617718823154e-08,
+      "loss": 0.2235,
+      "step": 17630
+    },
+    {
+      "epoch": 2.9047158354089047,
+      "grad_norm": 1.691039800643921,
+      "learning_rate": 5.3109173060820196e-08,
+      "loss": 0.2222,
+      "step": 17640
+    },
+    {
+      "epoch": 2.9063625697289064,
+      "grad_norm": 1.4430158138275146,
+      "learning_rate": 5.129516355568354e-08,
+      "loss": 0.2101,
+      "step": 17650
+    },
+    {
+      "epoch": 2.908009304048908,
+      "grad_norm": 1.406510829925537,
+      "learning_rate": 4.9512594936224065e-08,
+      "loss": 0.2136,
+      "step": 17660
+    },
+    {
+      "epoch": 2.90965603836891,
+      "grad_norm": 1.539867639541626,
+      "learning_rate": 4.776147283589438e-08,
+      "loss": 0.2305,
+      "step": 17670
+    },
+    {
+      "epoch": 2.911302772688911,
+      "grad_norm": 1.5464060306549072,
+      "learning_rate": 4.6041802788762136e-08,
+      "loss": 0.218,
+      "step": 17680
+    },
+    {
+      "epoch": 2.912949507008913,
+      "grad_norm": 1.3350013494491577,
+      "learning_rate": 4.435359022950336e-08,
+      "loss": 0.2234,
+      "step": 17690
+    },
+    {
+      "epoch": 2.9145962413289146,
+      "grad_norm": 1.1990008354187012,
+      "learning_rate": 4.269684049337142e-08,
+      "loss": 0.2208,
+      "step": 17700
+    },
+    {
+      "epoch": 2.9162429756489163,
+      "grad_norm": 1.4853509664535522,
+      "learning_rate": 4.1071558816193626e-08,
+      "loss": 0.219,
+      "step": 17710
+    },
+    {
+      "epoch": 2.917889709968918,
+      "grad_norm": 1.4338077306747437,
+      "learning_rate": 3.947775033434575e-08,
+      "loss": 0.2077,
+      "step": 17720
+    },
+    {
+      "epoch": 2.9195364442889193,
+      "grad_norm": 2.2821855545043945,
+      "learning_rate": 3.7915420084740915e-08,
+      "loss": 0.2177,
+      "step": 17730
+    },
+    {
+      "epoch": 2.921183178608921,
+      "grad_norm": 1.7811158895492554,
+      "learning_rate": 3.6384573004808465e-08,
+      "loss": 0.2172,
+      "step": 17740
+    },
+    {
+      "epoch": 2.922829912928923,
+      "grad_norm": 1.8270400762557983,
+      "learning_rate": 3.488521393248401e-08,
+      "loss": 0.2127,
+      "step": 17750
+    },
+    {
+      "epoch": 2.9244766472489245,
+      "grad_norm": 1.455257534980774,
+      "learning_rate": 3.341734760619275e-08,
+      "loss": 0.2223,
+      "step": 17760
+    },
+    {
+      "epoch": 2.9261233815689263,
+      "grad_norm": 1.5582276582717896,
+      "learning_rate": 3.198097866483063e-08,
+      "loss": 0.2194,
+      "step": 17770
+    },
+    {
+      "epoch": 2.9277701158889275,
+      "grad_norm": 1.5785218477249146,
+      "learning_rate": 3.0576111647752096e-08,
+      "loss": 0.2032,
+      "step": 17780
+    },
+    {
+      "epoch": 2.9294168502089293,
+      "grad_norm": 1.5375938415527344,
+      "learning_rate": 2.920275099476011e-08,
+      "loss": 0.2294,
+      "step": 17790
+    },
+    {
+      "epoch": 2.931063584528931,
+      "grad_norm": 2.171330451965332,
+      "learning_rate": 2.7860901046082856e-08,
+      "loss": 0.2358,
+      "step": 17800
+    },
+    {
+      "epoch": 2.9327103188489327,
+      "grad_norm": 1.675736665725708,
+      "learning_rate": 2.6550566042370386e-08,
+      "loss": 0.2215,
+      "step": 17810
+    },
+    {
+      "epoch": 2.9343570531689345,
+      "grad_norm": 2.0025336742401123,
+      "learning_rate": 2.5271750124672423e-08,
+      "loss": 0.2151,
+      "step": 17820
+    },
+    {
+      "epoch": 2.9360037874889358,
+      "grad_norm": 1.4711110591888428,
+      "learning_rate": 2.4024457334430595e-08,
+      "loss": 0.2173,
+      "step": 17830
+    },
+    {
+      "epoch": 2.9376505218089375,
+      "grad_norm": 1.8637895584106445,
+      "learning_rate": 2.2808691613461776e-08,
+      "loss": 0.2215,
+      "step": 17840
+    },
+    {
+      "epoch": 2.939297256128939,
+      "grad_norm": 2.4678914546966553,
+      "learning_rate": 2.162445680395142e-08,
+      "loss": 0.2373,
+      "step": 17850
+    },
+    {
+      "epoch": 2.940943990448941,
+      "grad_norm": 1.4459493160247803,
+      "learning_rate": 2.0471756648435814e-08,
+      "loss": 0.2033,
+      "step": 17860
+    },
+    {
+      "epoch": 2.9425907247689427,
+      "grad_norm": 1.8387855291366577,
+      "learning_rate": 1.9350594789792064e-08,
+      "loss": 0.2107,
+      "step": 17870
+    },
+    {
+      "epoch": 2.944237459088944,
+      "grad_norm": 1.8373688459396362,
+      "learning_rate": 1.8260974771227015e-08,
+      "loss": 0.2318,
+      "step": 17880
+    },
+    {
+      "epoch": 2.945884193408946,
+      "grad_norm": 1.6983997821807861,
+      "learning_rate": 1.7202900036268343e-08,
+      "loss": 0.2178,
+      "step": 17890
+    },
+    {
+      "epoch": 2.9475309277289474,
+      "grad_norm": 1.5288745164871216,
+      "learning_rate": 1.6176373928745715e-08,
+      "loss": 0.2205,
+      "step": 17900
+    },
+    {
+      "epoch": 2.949177662048949,
+      "grad_norm": 1.6072468757629395,
+      "learning_rate": 1.5181399692790756e-08,
+      "loss": 0.2076,
+      "step": 17910
+    },
+    {
+      "epoch": 2.950824396368951,
+      "grad_norm": 3.108124017715454,
+      "learning_rate": 1.4217980472819304e-08,
+      "loss": 0.233,
+      "step": 17920
+    },
+    {
+      "epoch": 2.9524711306889526,
+      "grad_norm": 1.8453097343444824,
+      "learning_rate": 1.3286119313525858e-08,
+      "loss": 0.2303,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9541178650089543,
+      "grad_norm": 2.0312986373901367,
+      "learning_rate": 1.2385819159869138e-08,
+      "loss": 0.2122,
+      "step": 17940
+    },
+    {
+      "epoch": 2.9557645993289556,
+      "grad_norm": 1.338973879814148,
+      "learning_rate": 1.1517082857067652e-08,
+      "loss": 0.2248,
+      "step": 17950
+    },
+    {
+      "epoch": 2.9574113336489574,
+      "grad_norm": 1.7432622909545898,
+      "learning_rate": 1.0679913150588584e-08,
+      "loss": 0.2213,
+      "step": 17960
+    },
+    {
+      "epoch": 2.959058067968959,
+      "grad_norm": 1.9854906797409058,
+      "learning_rate": 9.87431268613781e-09,
+      "loss": 0.2312,
+      "step": 17970
+    },
+    {
+      "epoch": 2.960704802288961,
+      "grad_norm": 1.817718267440796,
+      "learning_rate": 9.100284009655458e-09,
+      "loss": 0.2276,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9623515366089626,
+      "grad_norm": 1.655806064605713,
+      "learning_rate": 8.357829567302577e-09,
+      "loss": 0.2261,
+      "step": 17990
+    },
+    {
+      "epoch": 2.963998270928964,
+      "grad_norm": 1.608216643333435,
+      "learning_rate": 7.646951705457817e-09,
+      "loss": 0.2112,
+      "step": 18000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 18219,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2831001863587446e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}