diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,37637 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 53680,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007451564828614009,
+      "grad_norm": 1.7114099264144897,
+      "learning_rate": 0.0002,
+      "loss": 1.8316,
+      "step": 10
+    },
+    {
+      "epoch": 0.0014903129657228018,
+      "grad_norm": 2.16106915473938,
+      "learning_rate": 0.0002,
+      "loss": 1.9128,
+      "step": 20
+    },
+    {
+      "epoch": 0.0022354694485842027,
+      "grad_norm": 1.9972808361053467,
+      "learning_rate": 0.0002,
+      "loss": 1.8523,
+      "step": 30
+    },
+    {
+      "epoch": 0.0029806259314456036,
+      "grad_norm": 2.058236598968506,
+      "learning_rate": 0.0002,
+      "loss": 1.7295,
+      "step": 40
+    },
+    {
+      "epoch": 0.0037257824143070045,
+      "grad_norm": 2.164706230163574,
+      "learning_rate": 0.0002,
+      "loss": 1.6538,
+      "step": 50
+    },
+    {
+      "epoch": 0.004470938897168405,
+      "grad_norm": 1.7097787857055664,
+      "learning_rate": 0.0002,
+      "loss": 1.7809,
+      "step": 60
+    },
+    {
+      "epoch": 0.005216095380029807,
+      "grad_norm": 1.7462966442108154,
+      "learning_rate": 0.0002,
+      "loss": 1.7701,
+      "step": 70
+    },
+    {
+      "epoch": 0.005961251862891207,
+      "grad_norm": 2.259033679962158,
+      "learning_rate": 0.0002,
+      "loss": 1.7231,
+      "step": 80
+    },
+    {
+      "epoch": 0.0067064083457526085,
+      "grad_norm": 2.3188297748565674,
+      "learning_rate": 0.0002,
+      "loss": 1.9569,
+      "step": 90
+    },
+    {
+      "epoch": 0.007451564828614009,
+      "grad_norm": 1.7893937826156616,
+      "learning_rate": 0.0002,
+      "loss": 2.5206,
+      "step": 100
+    },
+    {
+      "epoch": 0.00819672131147541,
+      "grad_norm": 1.4487690925598145,
+      "learning_rate": 0.0002,
+      "loss": 2.5691,
+      "step": 110
+    },
+    {
+      "epoch": 0.00894187779433681,
+      "grad_norm": 1.6098469495773315,
+      "learning_rate": 0.0002,
+      "loss": 2.5209,
+      "step": 120
+    },
+    {
+      "epoch": 0.009687034277198211,
+      "grad_norm": 1.4259092807769775,
+      "learning_rate": 0.0002,
+      "loss": 2.5239,
+      "step": 130
+    },
+    {
+      "epoch": 0.010432190760059613,
+      "grad_norm": 1.4447660446166992,
+      "learning_rate": 0.0002,
+      "loss": 2.6533,
+      "step": 140
+    },
+    {
+      "epoch": 0.011177347242921014,
+      "grad_norm": 1.341070532798767,
+      "learning_rate": 0.0002,
+      "loss": 2.6231,
+      "step": 150
+    },
+    {
+      "epoch": 0.011922503725782414,
+      "grad_norm": 1.2440979480743408,
+      "learning_rate": 0.0002,
+      "loss": 2.4615,
+      "step": 160
+    },
+    {
+      "epoch": 0.012667660208643815,
+      "grad_norm": 1.4133532047271729,
+      "learning_rate": 0.0002,
+      "loss": 2.5564,
+      "step": 170
+    },
+    {
+      "epoch": 0.013412816691505217,
+      "grad_norm": 1.3202648162841797,
+      "learning_rate": 0.0002,
+      "loss": 2.4586,
+      "step": 180
+    },
+    {
+      "epoch": 0.014157973174366617,
+      "grad_norm": 1.2608141899108887,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 190
+    },
+    {
+      "epoch": 0.014903129657228018,
+      "grad_norm": 1.3800815343856812,
+      "learning_rate": 0.0002,
+      "loss": 2.4191,
+      "step": 200
+    },
+    {
+      "epoch": 0.01564828614008942,
+      "grad_norm": 1.4190888404846191,
+      "learning_rate": 0.0002,
+      "loss": 2.576,
+      "step": 210
+    },
+    {
+      "epoch": 0.01639344262295082,
+      "grad_norm": 1.36716628074646,
+      "learning_rate": 0.0002,
+      "loss": 2.4893,
+      "step": 220
+    },
+    {
+      "epoch": 0.01713859910581222,
+      "grad_norm": 1.4020277261734009,
+      "learning_rate": 0.0002,
+      "loss": 2.4958,
+      "step": 230
+    },
+    {
+      "epoch": 0.01788375558867362,
+      "grad_norm": 1.4159200191497803,
+      "learning_rate": 0.0002,
+      "loss": 2.2937,
+      "step": 240
+    },
+    {
+      "epoch": 0.018628912071535022,
+      "grad_norm": 1.6275138854980469,
+      "learning_rate": 0.0002,
+      "loss": 2.3386,
+      "step": 250
+    },
+    {
+      "epoch": 0.019374068554396422,
+      "grad_norm": 1.3647948503494263,
+      "learning_rate": 0.0002,
+      "loss": 2.5081,
+      "step": 260
+    },
+    {
+      "epoch": 0.020119225037257823,
+      "grad_norm": 1.237382173538208,
+      "learning_rate": 0.0002,
+      "loss": 2.4055,
+      "step": 270
+    },
+    {
+      "epoch": 0.020864381520119227,
+      "grad_norm": 1.3185752630233765,
+      "learning_rate": 0.0002,
+      "loss": 2.5449,
+      "step": 280
+    },
+    {
+      "epoch": 0.021609538002980627,
+      "grad_norm": 1.454095721244812,
+      "learning_rate": 0.0002,
+      "loss": 2.5499,
+      "step": 290
+    },
+    {
+      "epoch": 0.022354694485842028,
+      "grad_norm": 1.4591819047927856,
+      "learning_rate": 0.0002,
+      "loss": 2.4029,
+      "step": 300
+    },
+    {
+      "epoch": 0.023099850968703428,
+      "grad_norm": 1.3012772798538208,
+      "learning_rate": 0.0002,
+      "loss": 2.3855,
+      "step": 310
+    },
+    {
+      "epoch": 0.02384500745156483,
+      "grad_norm": 1.326201319694519,
+      "learning_rate": 0.0002,
+      "loss": 2.6307,
+      "step": 320
+    },
+    {
+      "epoch": 0.02459016393442623,
+      "grad_norm": 1.3953168392181396,
+      "learning_rate": 0.0002,
+      "loss": 2.6101,
+      "step": 330
+    },
+    {
+      "epoch": 0.02533532041728763,
+      "grad_norm": 1.5380605459213257,
+      "learning_rate": 0.0002,
+      "loss": 2.4536,
+      "step": 340
+    },
+    {
+      "epoch": 0.02608047690014903,
+      "grad_norm": 1.5608468055725098,
+      "learning_rate": 0.0002,
+      "loss": 2.1291,
+      "step": 350
+    },
+    {
+      "epoch": 0.026825633383010434,
+      "grad_norm": 1.316112756729126,
+      "learning_rate": 0.0002,
+      "loss": 2.5011,
+      "step": 360
+    },
+    {
+      "epoch": 0.027570789865871834,
+      "grad_norm": 1.4802881479263306,
+      "learning_rate": 0.0002,
+      "loss": 2.4637,
+      "step": 370
+    },
+    {
+      "epoch": 0.028315946348733235,
+      "grad_norm": 1.213468074798584,
+      "learning_rate": 0.0002,
+      "loss": 2.5339,
+      "step": 380
+    },
+    {
+      "epoch": 0.029061102831594635,
+      "grad_norm": 1.3106322288513184,
+      "learning_rate": 0.0002,
+      "loss": 2.5736,
+      "step": 390
+    },
+    {
+      "epoch": 0.029806259314456036,
+      "grad_norm": 1.7102941274642944,
+      "learning_rate": 0.0002,
+      "loss": 2.4255,
+      "step": 400
+    },
+    {
+      "epoch": 0.030551415797317436,
+      "grad_norm": 2.0210812091827393,
+      "learning_rate": 0.0002,
+      "loss": 2.4833,
+      "step": 410
+    },
+    {
+      "epoch": 0.03129657228017884,
+      "grad_norm": 1.38539457321167,
+      "learning_rate": 0.0002,
+      "loss": 2.3001,
+      "step": 420
+    },
+    {
+      "epoch": 0.03204172876304024,
+      "grad_norm": 1.439780592918396,
+      "learning_rate": 0.0002,
+      "loss": 2.5553,
+      "step": 430
+    },
+    {
+      "epoch": 0.03278688524590164,
+      "grad_norm": 1.488784670829773,
+      "learning_rate": 0.0002,
+      "loss": 2.3695,
+      "step": 440
+    },
+    {
+      "epoch": 0.03353204172876304,
+      "grad_norm": 1.4806923866271973,
+      "learning_rate": 0.0002,
+      "loss": 2.4231,
+      "step": 450
+    },
+    {
+      "epoch": 0.03427719821162444,
+      "grad_norm": 1.3404014110565186,
+      "learning_rate": 0.0002,
+      "loss": 2.5017,
+      "step": 460
+    },
+    {
+      "epoch": 0.03502235469448584,
+      "grad_norm": 1.6606372594833374,
+      "learning_rate": 0.0002,
+      "loss": 2.4502,
+      "step": 470
+    },
+    {
+      "epoch": 0.03576751117734724,
+      "grad_norm": 1.6447430849075317,
+      "learning_rate": 0.0002,
+      "loss": 2.2351,
+      "step": 480
+    },
+    {
+      "epoch": 0.03651266766020864,
+      "grad_norm": 1.2761911153793335,
+      "learning_rate": 0.0002,
+      "loss": 2.3787,
+      "step": 490
+    },
+    {
+      "epoch": 0.037257824143070044,
+      "grad_norm": 1.4461835622787476,
+      "learning_rate": 0.0002,
+      "loss": 2.5047,
+      "step": 500
+    },
+    {
+      "epoch": 0.038002980625931444,
+      "grad_norm": 1.9162489175796509,
+      "learning_rate": 0.0002,
+      "loss": 2.3294,
+      "step": 510
+    },
+    {
+      "epoch": 0.038748137108792845,
+      "grad_norm": 1.6676952838897705,
+      "learning_rate": 0.0002,
+      "loss": 2.4176,
+      "step": 520
+    },
+    {
+      "epoch": 0.039493293591654245,
+      "grad_norm": 1.7947252988815308,
+      "learning_rate": 0.0002,
+      "loss": 2.3651,
+      "step": 530
+    },
+    {
+      "epoch": 0.040238450074515646,
+      "grad_norm": 1.4677958488464355,
+      "learning_rate": 0.0002,
+      "loss": 2.6263,
+      "step": 540
+    },
+    {
+      "epoch": 0.040983606557377046,
+      "grad_norm": 1.4735350608825684,
+      "learning_rate": 0.0002,
+      "loss": 2.4261,
+      "step": 550
+    },
+    {
+      "epoch": 0.041728763040238454,
+      "grad_norm": 1.4515550136566162,
+      "learning_rate": 0.0002,
+      "loss": 2.4795,
+      "step": 560
+    },
+    {
+      "epoch": 0.042473919523099854,
+      "grad_norm": 1.6137213706970215,
+      "learning_rate": 0.0002,
+      "loss": 2.5102,
+      "step": 570
+    },
+    {
+      "epoch": 0.043219076005961254,
+      "grad_norm": 1.6257590055465698,
+      "learning_rate": 0.0002,
+      "loss": 2.4218,
+      "step": 580
+    },
+    {
+      "epoch": 0.043964232488822655,
+      "grad_norm": 1.515711784362793,
+      "learning_rate": 0.0002,
+      "loss": 2.3959,
+      "step": 590
+    },
+    {
+      "epoch": 0.044709388971684055,
+      "grad_norm": 1.4846817255020142,
+      "learning_rate": 0.0002,
+      "loss": 2.4579,
+      "step": 600
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 1.7104601860046387,
+      "learning_rate": 0.0002,
+      "loss": 2.4271,
+      "step": 610
+    },
+    {
+      "epoch": 0.046199701937406856,
+      "grad_norm": 1.4951586723327637,
+      "learning_rate": 0.0002,
+      "loss": 2.4518,
+      "step": 620
+    },
+    {
+      "epoch": 0.04694485842026826,
+      "grad_norm": 1.4925544261932373,
+      "learning_rate": 0.0002,
+      "loss": 2.4987,
+      "step": 630
+    },
+    {
+      "epoch": 0.04769001490312966,
+      "grad_norm": 1.4516690969467163,
+      "learning_rate": 0.0002,
+      "loss": 2.4968,
+      "step": 640
+    },
+    {
+      "epoch": 0.04843517138599106,
+      "grad_norm": 1.6420326232910156,
+      "learning_rate": 0.0002,
+      "loss": 2.4214,
+      "step": 650
+    },
+    {
+      "epoch": 0.04918032786885246,
+      "grad_norm": 1.6954593658447266,
+      "learning_rate": 0.0002,
+      "loss": 2.4363,
+      "step": 660
+    },
+    {
+      "epoch": 0.04992548435171386,
+      "grad_norm": 1.6805527210235596,
+      "learning_rate": 0.0002,
+      "loss": 2.2638,
+      "step": 670
+    },
+    {
+      "epoch": 0.05067064083457526,
+      "grad_norm": 1.5352181196212769,
+      "learning_rate": 0.0002,
+      "loss": 2.3839,
+      "step": 680
+    },
+    {
+      "epoch": 0.05141579731743666,
+      "grad_norm": 1.696765661239624,
+      "learning_rate": 0.0002,
+      "loss": 2.4535,
+      "step": 690
+    },
+    {
+      "epoch": 0.05216095380029806,
+      "grad_norm": 1.4545871019363403,
+      "learning_rate": 0.0002,
+      "loss": 2.3242,
+      "step": 700
+    },
+    {
+      "epoch": 0.05290611028315946,
+      "grad_norm": 1.4923096895217896,
+      "learning_rate": 0.0002,
+      "loss": 2.4713,
+      "step": 710
+    },
+    {
+      "epoch": 0.05365126676602087,
+      "grad_norm": 1.483655571937561,
+      "learning_rate": 0.0002,
+      "loss": 2.4,
+      "step": 720
+    },
+    {
+      "epoch": 0.05439642324888227,
+      "grad_norm": 1.5965553522109985,
+      "learning_rate": 0.0002,
+      "loss": 2.5278,
+      "step": 730
+    },
+    {
+      "epoch": 0.05514157973174367,
+      "grad_norm": 1.788888931274414,
+      "learning_rate": 0.0002,
+      "loss": 2.5566,
+      "step": 740
+    },
+    {
+      "epoch": 0.05588673621460507,
+      "grad_norm": 1.612257957458496,
+      "learning_rate": 0.0002,
+      "loss": 2.4796,
+      "step": 750
+    },
+    {
+      "epoch": 0.05663189269746647,
+      "grad_norm": 1.4303230047225952,
+      "learning_rate": 0.0002,
+      "loss": 2.4086,
+      "step": 760
+    },
+    {
+      "epoch": 0.05737704918032787,
+      "grad_norm": 1.534578800201416,
+      "learning_rate": 0.0002,
+      "loss": 2.6292,
+      "step": 770
+    },
+    {
+      "epoch": 0.05812220566318927,
+      "grad_norm": 1.4243452548980713,
+      "learning_rate": 0.0002,
+      "loss": 2.5726,
+      "step": 780
+    },
+    {
+      "epoch": 0.05886736214605067,
+      "grad_norm": 1.3264392614364624,
+      "learning_rate": 0.0002,
+      "loss": 2.552,
+      "step": 790
+    },
+    {
+      "epoch": 0.05961251862891207,
+      "grad_norm": 1.251664400100708,
+      "learning_rate": 0.0002,
+      "loss": 2.2833,
+      "step": 800
+    },
+    {
+      "epoch": 0.06035767511177347,
+      "grad_norm": 1.5511835813522339,
+      "learning_rate": 0.0002,
+      "loss": 2.5349,
+      "step": 810
+    },
+    {
+      "epoch": 0.06110283159463487,
+      "grad_norm": 1.3495992422103882,
+      "learning_rate": 0.0002,
+      "loss": 2.2091,
+      "step": 820
+    },
+    {
+      "epoch": 0.06184798807749627,
+      "grad_norm": 1.343342661857605,
+      "learning_rate": 0.0002,
+      "loss": 2.6122,
+      "step": 830
+    },
+    {
+      "epoch": 0.06259314456035768,
+      "grad_norm": 1.568200707435608,
+      "learning_rate": 0.0002,
+      "loss": 2.3481,
+      "step": 840
+    },
+    {
+      "epoch": 0.06333830104321908,
+      "grad_norm": 1.6847620010375977,
+      "learning_rate": 0.0002,
+      "loss": 2.4552,
+      "step": 850
+    },
+    {
+      "epoch": 0.06408345752608048,
+      "grad_norm": 2.0226731300354004,
+      "learning_rate": 0.0002,
+      "loss": 2.4509,
+      "step": 860
+    },
+    {
+      "epoch": 0.06482861400894188,
+      "grad_norm": 1.8059664964675903,
+      "learning_rate": 0.0002,
+      "loss": 2.4808,
+      "step": 870
+    },
+    {
+      "epoch": 0.06557377049180328,
+      "grad_norm": 1.5953114032745361,
+      "learning_rate": 0.0002,
+      "loss": 2.5304,
+      "step": 880
+    },
+    {
+      "epoch": 0.06631892697466468,
+      "grad_norm": 1.7731475830078125,
+      "learning_rate": 0.0002,
+      "loss": 2.258,
+      "step": 890
+    },
+    {
+      "epoch": 0.06706408345752608,
+      "grad_norm": 1.8624380826950073,
+      "learning_rate": 0.0002,
+      "loss": 2.3362,
+      "step": 900
+    },
+    {
+      "epoch": 0.06780923994038748,
+      "grad_norm": 1.6222556829452515,
+      "learning_rate": 0.0002,
+      "loss": 2.4066,
+      "step": 910
+    },
+    {
+      "epoch": 0.06855439642324888,
+      "grad_norm": 1.7085922956466675,
+      "learning_rate": 0.0002,
+      "loss": 2.5117,
+      "step": 920
+    },
+    {
+      "epoch": 0.06929955290611028,
+      "grad_norm": 1.6273415088653564,
+      "learning_rate": 0.0002,
+      "loss": 2.5034,
+      "step": 930
+    },
+    {
+      "epoch": 0.07004470938897168,
+      "grad_norm": 1.5758957862854004,
+      "learning_rate": 0.0002,
+      "loss": 2.4744,
+      "step": 940
+    },
+    {
+      "epoch": 0.07078986587183309,
+      "grad_norm": 1.7586950063705444,
+      "learning_rate": 0.0002,
+      "loss": 2.6029,
+      "step": 950
+    },
+    {
+      "epoch": 0.07153502235469449,
+      "grad_norm": 1.7354320287704468,
+      "learning_rate": 0.0002,
+      "loss": 2.4572,
+      "step": 960
+    },
+    {
+      "epoch": 0.07228017883755589,
+      "grad_norm": 1.440529704093933,
+      "learning_rate": 0.0002,
+      "loss": 2.4439,
+      "step": 970
+    },
+    {
+      "epoch": 0.07302533532041729,
+      "grad_norm": 1.4725509881973267,
+      "learning_rate": 0.0002,
+      "loss": 2.4279,
+      "step": 980
+    },
+    {
+      "epoch": 0.07377049180327869,
+      "grad_norm": 1.6604877710342407,
+      "learning_rate": 0.0002,
+      "loss": 2.4856,
+      "step": 990
+    },
+    {
+      "epoch": 0.07451564828614009,
+      "grad_norm": 1.5507324934005737,
+      "learning_rate": 0.0002,
+      "loss": 2.5235,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07526080476900149,
+      "grad_norm": 1.6070083379745483,
+      "learning_rate": 0.0002,
+      "loss": 2.3157,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07600596125186289,
+      "grad_norm": 1.5939691066741943,
+      "learning_rate": 0.0002,
+      "loss": 2.5106,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07675111773472429,
+      "grad_norm": 1.68025803565979,
+      "learning_rate": 0.0002,
+      "loss": 2.4918,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07749627421758569,
+      "grad_norm": 1.6039890050888062,
+      "learning_rate": 0.0002,
+      "loss": 2.5583,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07824143070044709,
+      "grad_norm": 1.8732694387435913,
+      "learning_rate": 0.0002,
+      "loss": 2.2141,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07898658718330849,
+      "grad_norm": 1.4604535102844238,
+      "learning_rate": 0.0002,
+      "loss": 2.5011,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07973174366616989,
+      "grad_norm": 1.5717531442642212,
+      "learning_rate": 0.0002,
+      "loss": 2.6481,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08047690014903129,
+      "grad_norm": 1.6097593307495117,
+      "learning_rate": 0.0002,
+      "loss": 2.6329,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08122205663189269,
+      "grad_norm": 1.6716941595077515,
+      "learning_rate": 0.0002,
+      "loss": 2.2852,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08196721311475409,
+      "grad_norm": 1.4320539236068726,
+      "learning_rate": 0.0002,
+      "loss": 2.5,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08271236959761549,
+      "grad_norm": 1.76832914352417,
+      "learning_rate": 0.0002,
+      "loss": 2.4243,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08345752608047691,
+      "grad_norm": 1.5858458280563354,
+      "learning_rate": 0.0002,
+      "loss": 2.4635,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08420268256333831,
+      "grad_norm": 1.627699851989746,
+      "learning_rate": 0.0002,
+      "loss": 2.429,
+      "step": 1130
+    },
+    {
+      "epoch": 0.08494783904619971,
+      "grad_norm": 1.723833441734314,
+      "learning_rate": 0.0002,
+      "loss": 2.5082,
+      "step": 1140
+    },
+    {
+      "epoch": 0.08569299552906111,
+      "grad_norm": 1.6531169414520264,
+      "learning_rate": 0.0002,
+      "loss": 2.3722,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08643815201192251,
+      "grad_norm": 1.6412060260772705,
+      "learning_rate": 0.0002,
+      "loss": 2.4493,
+      "step": 1160
+    },
+    {
+      "epoch": 0.08718330849478391,
+      "grad_norm": 1.6279152631759644,
+      "learning_rate": 0.0002,
+      "loss": 2.6463,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08792846497764531,
+      "grad_norm": 1.9763994216918945,
+      "learning_rate": 0.0002,
+      "loss": 2.5753,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08867362146050671,
+      "grad_norm": 1.5985665321350098,
+      "learning_rate": 0.0002,
+      "loss": 2.6349,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08941877794336811,
+      "grad_norm": 1.683590054512024,
+      "learning_rate": 0.0002,
+      "loss": 2.4348,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09016393442622951,
+      "grad_norm": 1.55161452293396,
+      "learning_rate": 0.0002,
+      "loss": 2.2263,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 1.7723956108093262,
+      "learning_rate": 0.0002,
+      "loss": 2.5187,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09165424739195231,
+      "grad_norm": 2.007422924041748,
+      "learning_rate": 0.0002,
+      "loss": 2.2847,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09239940387481371,
+      "grad_norm": 1.6839237213134766,
+      "learning_rate": 0.0002,
+      "loss": 2.395,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09314456035767511,
+      "grad_norm": 1.8041167259216309,
+      "learning_rate": 0.0002,
+      "loss": 2.4219,
+      "step": 1250
+    },
+    {
+      "epoch": 0.09388971684053651,
+      "grad_norm": 1.3533605337142944,
+      "learning_rate": 0.0002,
+      "loss": 2.1161,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09463487332339791,
+      "grad_norm": 1.5705360174179077,
+      "learning_rate": 0.0002,
+      "loss": 2.4224,
+      "step": 1270
+    },
+    {
+      "epoch": 0.09538002980625931,
+      "grad_norm": 1.3010971546173096,
+      "learning_rate": 0.0002,
+      "loss": 2.7023,
+      "step": 1280
+    },
+    {
+      "epoch": 0.09612518628912071,
+      "grad_norm": 1.61898934841156,
+      "learning_rate": 0.0002,
+      "loss": 2.4894,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09687034277198212,
+      "grad_norm": 1.570351243019104,
+      "learning_rate": 0.0002,
+      "loss": 2.6117,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09761549925484352,
+      "grad_norm": 1.5680445432662964,
+      "learning_rate": 0.0002,
+      "loss": 2.4442,
+      "step": 1310
+    },
+    {
+      "epoch": 0.09836065573770492,
+      "grad_norm": 1.6972705125808716,
+      "learning_rate": 0.0002,
+      "loss": 2.4776,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09910581222056632,
+      "grad_norm": 1.4837958812713623,
+      "learning_rate": 0.0002,
+      "loss": 2.4469,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09985096870342772,
+      "grad_norm": 1.241639494895935,
+      "learning_rate": 0.0002,
+      "loss": 2.2837,
+      "step": 1340
+    },
+    {
+      "epoch": 0.10059612518628912,
+      "grad_norm": 1.6141420602798462,
+      "learning_rate": 0.0002,
+      "loss": 2.4861,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10134128166915052,
+      "grad_norm": 1.3432039022445679,
+      "learning_rate": 0.0002,
+      "loss": 2.2677,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10208643815201192,
+      "grad_norm": 1.4515589475631714,
+      "learning_rate": 0.0002,
+      "loss": 2.4508,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10283159463487332,
+      "grad_norm": 1.6723533868789673,
+      "learning_rate": 0.0002,
+      "loss": 2.4985,
+      "step": 1380
+    },
+    {
+      "epoch": 0.10357675111773472,
+      "grad_norm": 1.6158736944198608,
+      "learning_rate": 0.0002,
+      "loss": 2.5645,
+      "step": 1390
+    },
+    {
+      "epoch": 0.10432190760059612,
+      "grad_norm": 1.7271533012390137,
+      "learning_rate": 0.0002,
+      "loss": 2.367,
+      "step": 1400
+    },
+    {
+      "epoch": 0.10506706408345752,
+      "grad_norm": 1.1721924543380737,
+      "learning_rate": 0.0002,
+      "loss": 2.3531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.10581222056631892,
+      "grad_norm": 1.509311318397522,
+      "learning_rate": 0.0002,
+      "loss": 2.3562,
+      "step": 1420
+    },
+    {
+      "epoch": 0.10655737704918032,
+      "grad_norm": 1.8995124101638794,
+      "learning_rate": 0.0002,
+      "loss": 2.325,
+      "step": 1430
+    },
+    {
+      "epoch": 0.10730253353204174,
+      "grad_norm": 1.5647929906845093,
+      "learning_rate": 0.0002,
+      "loss": 2.306,
+      "step": 1440
+    },
+    {
+      "epoch": 0.10804769001490314,
+      "grad_norm": 1.8889915943145752,
+      "learning_rate": 0.0002,
+      "loss": 2.3319,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10879284649776454,
+      "grad_norm": 1.7834174633026123,
+      "learning_rate": 0.0002,
+      "loss": 2.4158,
+      "step": 1460
+    },
+    {
+      "epoch": 0.10953800298062594,
+      "grad_norm": 1.7446699142456055,
+      "learning_rate": 0.0002,
+      "loss": 2.5858,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11028315946348734,
+      "grad_norm": 1.6821390390396118,
+      "learning_rate": 0.0002,
+      "loss": 2.6529,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11102831594634874,
+      "grad_norm": 1.7135263681411743,
+      "learning_rate": 0.0002,
+      "loss": 2.4174,
+      "step": 1490
+    },
+    {
+      "epoch": 0.11177347242921014,
+      "grad_norm": 1.6189689636230469,
+      "learning_rate": 0.0002,
+      "loss": 2.3491,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11251862891207154,
+      "grad_norm": 1.9968479871749878,
+      "learning_rate": 0.0002,
+      "loss": 2.3949,
+      "step": 1510
+    },
+    {
+      "epoch": 0.11326378539493294,
+      "grad_norm": 1.7658995389938354,
+      "learning_rate": 0.0002,
+      "loss": 2.3559,
+      "step": 1520
+    },
+    {
+      "epoch": 0.11400894187779434,
+      "grad_norm": 1.5398634672164917,
+      "learning_rate": 0.0002,
+      "loss": 2.3183,
+      "step": 1530
+    },
+    {
+      "epoch": 0.11475409836065574,
+      "grad_norm": 1.558796763420105,
+      "learning_rate": 0.0002,
+      "loss": 2.5318,
+      "step": 1540
+    },
+    {
+      "epoch": 0.11549925484351714,
+      "grad_norm": 1.7153369188308716,
+      "learning_rate": 0.0002,
+      "loss": 2.42,
+      "step": 1550
+    },
+    {
+      "epoch": 0.11624441132637854,
+      "grad_norm": 1.8950843811035156,
+      "learning_rate": 0.0002,
+      "loss": 2.4338,
+      "step": 1560
+    },
+    {
+      "epoch": 0.11698956780923994,
+      "grad_norm": 1.9553101062774658,
+      "learning_rate": 0.0002,
+      "loss": 2.3172,
+      "step": 1570
+    },
+    {
+      "epoch": 0.11773472429210134,
+      "grad_norm": 2.0377368927001953,
+      "learning_rate": 0.0002,
+      "loss": 2.5465,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11847988077496274,
+      "grad_norm": 1.596413254737854,
+      "learning_rate": 0.0002,
+      "loss": 2.3895,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11922503725782414,
+      "grad_norm": 1.9717952013015747,
+      "learning_rate": 0.0002,
+      "loss": 2.33,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11997019374068554,
+      "grad_norm": 1.597959041595459,
+      "learning_rate": 0.0002,
+      "loss": 2.4298,
+      "step": 1610
+    },
+    {
+      "epoch": 0.12071535022354694,
+      "grad_norm": 1.7834532260894775,
+      "learning_rate": 0.0002,
+      "loss": 2.4906,
+      "step": 1620
+    },
+    {
+      "epoch": 0.12146050670640834,
+      "grad_norm": 1.7709592580795288,
+      "learning_rate": 0.0002,
+      "loss": 2.4961,
+      "step": 1630
+    },
+    {
+      "epoch": 0.12220566318926974,
+      "grad_norm": 1.448915719985962,
+      "learning_rate": 0.0002,
+      "loss": 2.2579,
+      "step": 1640
+    },
+    {
+      "epoch": 0.12295081967213115,
+      "grad_norm": 1.9186158180236816,
+      "learning_rate": 0.0002,
+      "loss": 2.6403,
+      "step": 1650
+    },
+    {
+      "epoch": 0.12369597615499255,
+      "grad_norm": 1.7312026023864746,
+      "learning_rate": 0.0002,
+      "loss": 2.3614,
+      "step": 1660
+    },
+    {
+      "epoch": 0.12444113263785395,
+      "grad_norm": 1.6252959966659546,
+      "learning_rate": 0.0002,
+      "loss": 2.1939,
+      "step": 1670
+    },
+    {
+      "epoch": 0.12518628912071536,
+      "grad_norm": 1.485950231552124,
+      "learning_rate": 0.0002,
+      "loss": 2.3911,
+      "step": 1680
+    },
+    {
+      "epoch": 0.12593144560357675,
+      "grad_norm": 2.1096255779266357,
+      "learning_rate": 0.0002,
+      "loss": 2.3905,
+      "step": 1690
+    },
+    {
+      "epoch": 0.12667660208643816,
+      "grad_norm": 1.9784533977508545,
+      "learning_rate": 0.0002,
+      "loss": 2.5378,
+      "step": 1700
+    },
+    {
+      "epoch": 0.12742175856929955,
+      "grad_norm": 1.6286430358886719,
+      "learning_rate": 0.0002,
+      "loss": 2.3893,
+      "step": 1710
+    },
+    {
+      "epoch": 0.12816691505216096,
+      "grad_norm": 1.6326884031295776,
+      "learning_rate": 0.0002,
+      "loss": 2.2981,
+      "step": 1720
+    },
+    {
+      "epoch": 0.12891207153502235,
+      "grad_norm": 1.8827307224273682,
+      "learning_rate": 0.0002,
+      "loss": 2.386,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12965722801788376,
+      "grad_norm": 1.5798028707504272,
+      "learning_rate": 0.0002,
+      "loss": 2.4409,
+      "step": 1740
+    },
+    {
+      "epoch": 0.13040238450074515,
+      "grad_norm": 1.701941728591919,
+      "learning_rate": 0.0002,
+      "loss": 2.621,
+      "step": 1750
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 2.052530527114868,
+      "learning_rate": 0.0002,
+      "loss": 2.5072,
+      "step": 1760
+    },
+    {
+      "epoch": 0.13189269746646795,
+      "grad_norm": 1.7951122522354126,
+      "learning_rate": 0.0002,
+      "loss": 2.6495,
+      "step": 1770
+    },
+    {
+      "epoch": 0.13263785394932937,
+      "grad_norm": 1.713895320892334,
+      "learning_rate": 0.0002,
+      "loss": 2.4414,
+      "step": 1780
+    },
+    {
+      "epoch": 0.13338301043219075,
+      "grad_norm": 1.7314261198043823,
+      "learning_rate": 0.0002,
+      "loss": 2.5448,
+      "step": 1790
+    },
+    {
+      "epoch": 0.13412816691505217,
+      "grad_norm": 2.169917583465576,
+      "learning_rate": 0.0002,
+      "loss": 2.3403,
+      "step": 1800
+    },
+    {
+      "epoch": 0.13487332339791355,
+      "grad_norm": 2.260791063308716,
+      "learning_rate": 0.0002,
+      "loss": 2.3851,
+      "step": 1810
+    },
+    {
+      "epoch": 0.13561847988077497,
+      "grad_norm": 1.5603179931640625,
+      "learning_rate": 0.0002,
+      "loss": 2.385,
+      "step": 1820
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 1.854830026626587,
+      "learning_rate": 0.0002,
+      "loss": 2.5961,
+      "step": 1830
+    },
+    {
+      "epoch": 0.13710879284649777,
+      "grad_norm": 2.0177578926086426,
+      "learning_rate": 0.0002,
+      "loss": 2.5359,
+      "step": 1840
+    },
+    {
+      "epoch": 0.13785394932935915,
+      "grad_norm": 1.6412891149520874,
+      "learning_rate": 0.0002,
+      "loss": 2.5614,
+      "step": 1850
+    },
+    {
+      "epoch": 0.13859910581222057,
+      "grad_norm": 1.498461127281189,
+      "learning_rate": 0.0002,
+      "loss": 2.4013,
+      "step": 1860
+    },
+    {
+      "epoch": 0.13934426229508196,
+      "grad_norm": 1.7153528928756714,
+      "learning_rate": 0.0002,
+      "loss": 2.4077,
+      "step": 1870
+    },
+    {
+      "epoch": 0.14008941877794337,
+      "grad_norm": 1.8465684652328491,
+      "learning_rate": 0.0002,
+      "loss": 2.4389,
+      "step": 1880
+    },
+    {
+      "epoch": 0.14083457526080476,
+      "grad_norm": 1.687004566192627,
+      "learning_rate": 0.0002,
+      "loss": 2.629,
+      "step": 1890
+    },
+    {
+      "epoch": 0.14157973174366617,
+      "grad_norm": 2.4556238651275635,
+      "learning_rate": 0.0002,
+      "loss": 2.4681,
+      "step": 1900
+    },
+    {
+      "epoch": 0.14232488822652756,
+      "grad_norm": 1.6830357313156128,
+      "learning_rate": 0.0002,
+      "loss": 2.5962,
+      "step": 1910
+    },
+    {
+      "epoch": 0.14307004470938897,
+      "grad_norm": 1.5959599018096924,
+      "learning_rate": 0.0002,
+      "loss": 2.3358,
+      "step": 1920
+    },
+    {
+      "epoch": 0.14381520119225039,
+      "grad_norm": 1.9852540493011475,
+      "learning_rate": 0.0002,
+      "loss": 2.4023,
+      "step": 1930
+    },
+    {
+      "epoch": 0.14456035767511177,
+      "grad_norm": 1.334253191947937,
+      "learning_rate": 0.0002,
+      "loss": 2.3208,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1453055141579732,
+      "grad_norm": 1.7445403337478638,
+      "learning_rate": 0.0002,
+      "loss": 2.7472,
+      "step": 1950
+    },
+    {
+      "epoch": 0.14605067064083457,
+      "grad_norm": 2.1274497509002686,
+      "learning_rate": 0.0002,
+      "loss": 2.4091,
+      "step": 1960
+    },
+    {
+      "epoch": 0.146795827123696,
+      "grad_norm": 1.9186792373657227,
+      "learning_rate": 0.0002,
+      "loss": 2.5525,
+      "step": 1970
+    },
+    {
+      "epoch": 0.14754098360655737,
+      "grad_norm": 1.7423287630081177,
+      "learning_rate": 0.0002,
+      "loss": 2.4312,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1482861400894188,
+      "grad_norm": 1.6370166540145874,
+      "learning_rate": 0.0002,
+      "loss": 2.6371,
+      "step": 1990
+    },
+    {
+      "epoch": 0.14903129657228018,
+      "grad_norm": 1.812752366065979,
+      "learning_rate": 0.0002,
+      "loss": 2.4224,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1497764530551416,
+      "grad_norm": 1.6510322093963623,
+      "learning_rate": 0.0002,
+      "loss": 2.4596,
+      "step": 2010
+    },
+    {
+      "epoch": 0.15052160953800298,
+      "grad_norm": 1.7658458948135376,
+      "learning_rate": 0.0002,
+      "loss": 2.4772,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1512667660208644,
+      "grad_norm": 1.8121706247329712,
+      "learning_rate": 0.0002,
+      "loss": 2.3523,
+      "step": 2030
+    },
+    {
+      "epoch": 0.15201192250372578,
+      "grad_norm": 1.7872707843780518,
+      "learning_rate": 0.0002,
+      "loss": 2.4009,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1527570789865872,
+      "grad_norm": 1.8161529302597046,
+      "learning_rate": 0.0002,
+      "loss": 2.2658,
+      "step": 2050
+    },
+    {
+      "epoch": 0.15350223546944858,
+      "grad_norm": 3.1155641078948975,
+      "learning_rate": 0.0002,
+      "loss": 2.3572,
+      "step": 2060
+    },
+    {
+      "epoch": 0.15424739195231,
+      "grad_norm": 3.371666193008423,
+      "learning_rate": 0.0002,
+      "loss": 2.5711,
+      "step": 2070
+    },
+    {
+      "epoch": 0.15499254843517138,
+      "grad_norm": 2.1471800804138184,
+      "learning_rate": 0.0002,
+      "loss": 2.4525,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1557377049180328,
+      "grad_norm": 1.849825382232666,
+      "learning_rate": 0.0002,
+      "loss": 2.472,
+      "step": 2090
+    },
+    {
+      "epoch": 0.15648286140089418,
+      "grad_norm": 1.9633269309997559,
+      "learning_rate": 0.0002,
+      "loss": 2.5328,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1572280178837556,
+      "grad_norm": 2.1593234539031982,
+      "learning_rate": 0.0002,
+      "loss": 2.5724,
+      "step": 2110
+    },
+    {
+      "epoch": 0.15797317436661698,
+      "grad_norm": 1.8166416883468628,
+      "learning_rate": 0.0002,
+      "loss": 2.5599,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1587183308494784,
+      "grad_norm": 2.0000736713409424,
+      "learning_rate": 0.0002,
+      "loss": 2.5174,
+      "step": 2130
+    },
+    {
+      "epoch": 0.15946348733233978,
+      "grad_norm": 2.016064167022705,
+      "learning_rate": 0.0002,
+      "loss": 2.4799,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1602086438152012,
+      "grad_norm": 2.289914846420288,
+      "learning_rate": 0.0002,
+      "loss": 2.4579,
+      "step": 2150
+    },
+    {
+      "epoch": 0.16095380029806258,
+      "grad_norm": 2.0967648029327393,
+      "learning_rate": 0.0002,
+      "loss": 2.3974,
+      "step": 2160
+    },
+    {
+      "epoch": 0.161698956780924,
+      "grad_norm": 1.9680815935134888,
+      "learning_rate": 0.0002,
+      "loss": 2.5335,
+      "step": 2170
+    },
+    {
+      "epoch": 0.16244411326378538,
+      "grad_norm": 1.659155011177063,
+      "learning_rate": 0.0002,
+      "loss": 2.2785,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1631892697466468,
+      "grad_norm": 1.719580054283142,
+      "learning_rate": 0.0002,
+      "loss": 2.5193,
+      "step": 2190
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 1.4163504838943481,
+      "learning_rate": 0.0002,
+      "loss": 2.4477,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1646795827123696,
+      "grad_norm": 1.4476326704025269,
+      "learning_rate": 0.0002,
+      "loss": 2.3973,
+      "step": 2210
+    },
+    {
+      "epoch": 0.16542473919523099,
+      "grad_norm": 1.7087070941925049,
+      "learning_rate": 0.0002,
+      "loss": 2.5413,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1661698956780924,
+      "grad_norm": 1.813745379447937,
+      "learning_rate": 0.0002,
+      "loss": 2.3435,
+      "step": 2230
+    },
+    {
+      "epoch": 0.16691505216095381,
+      "grad_norm": 2.2703053951263428,
+      "learning_rate": 0.0002,
+      "loss": 2.5846,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1676602086438152,
+      "grad_norm": 1.743831753730774,
+      "learning_rate": 0.0002,
+      "loss": 2.2399,
+      "step": 2250
+    },
+    {
+      "epoch": 0.16840536512667661,
+      "grad_norm": 1.7628560066223145,
+      "learning_rate": 0.0002,
+      "loss": 2.5024,
+      "step": 2260
+    },
+    {
+      "epoch": 0.169150521609538,
+      "grad_norm": 2.0905685424804688,
+      "learning_rate": 0.0002,
+      "loss": 2.3357,
+      "step": 2270
+    },
+    {
+      "epoch": 0.16989567809239942,
+      "grad_norm": 1.7532408237457275,
+      "learning_rate": 0.0002,
+      "loss": 2.4824,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1706408345752608,
+      "grad_norm": 1.749121069908142,
+      "learning_rate": 0.0002,
+      "loss": 2.3893,
+      "step": 2290
+    },
+    {
+      "epoch": 0.17138599105812222,
+      "grad_norm": 1.529807448387146,
+      "learning_rate": 0.0002,
+      "loss": 2.3581,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1721311475409836,
+      "grad_norm": 1.8684520721435547,
+      "learning_rate": 0.0002,
+      "loss": 2.3585,
+      "step": 2310
+    },
+    {
+      "epoch": 0.17287630402384502,
+      "grad_norm": 1.7635807991027832,
+      "learning_rate": 0.0002,
+      "loss": 2.3862,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1736214605067064,
+      "grad_norm": 1.6727739572525024,
+      "learning_rate": 0.0002,
+      "loss": 2.5992,
+      "step": 2330
+    },
+    {
+      "epoch": 0.17436661698956782,
+      "grad_norm": 1.7773075103759766,
+      "learning_rate": 0.0002,
+      "loss": 2.4468,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1751117734724292,
+      "grad_norm": 2.184798002243042,
+      "learning_rate": 0.0002,
+      "loss": 2.4697,
+      "step": 2350
+    },
+    {
+      "epoch": 0.17585692995529062,
+      "grad_norm": 1.7383967638015747,
+      "learning_rate": 0.0002,
+      "loss": 2.4937,
+      "step": 2360
+    },
+    {
+      "epoch": 0.176602086438152,
+      "grad_norm": 2.211831569671631,
+      "learning_rate": 0.0002,
+      "loss": 2.471,
+      "step": 2370
+    },
+    {
+      "epoch": 0.17734724292101342,
+      "grad_norm": 1.5768284797668457,
+      "learning_rate": 0.0002,
+      "loss": 2.4485,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1780923994038748,
+      "grad_norm": 1.7147942781448364,
+      "learning_rate": 0.0002,
+      "loss": 2.3187,
+      "step": 2390
+    },
+    {
+      "epoch": 0.17883755588673622,
+      "grad_norm": 1.5341167449951172,
+      "learning_rate": 0.0002,
+      "loss": 2.3842,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1795827123695976,
+      "grad_norm": 1.8212217092514038,
+      "learning_rate": 0.0002,
+      "loss": 2.2845,
+      "step": 2410
+    },
+    {
+      "epoch": 0.18032786885245902,
+      "grad_norm": 1.5502907037734985,
+      "learning_rate": 0.0002,
+      "loss": 2.4956,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1810730253353204,
+      "grad_norm": 1.8575736284255981,
+      "learning_rate": 0.0002,
+      "loss": 2.4574,
+      "step": 2430
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 1.594504475593567,
+      "learning_rate": 0.0002,
+      "loss": 2.4808,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1825633383010432,
+      "grad_norm": 1.864490032196045,
+      "learning_rate": 0.0002,
+      "loss": 2.6226,
+      "step": 2450
+    },
+    {
+      "epoch": 0.18330849478390462,
+      "grad_norm": 1.9114779233932495,
+      "learning_rate": 0.0002,
+      "loss": 2.5032,
+      "step": 2460
+    },
+    {
+      "epoch": 0.184053651266766,
+      "grad_norm": 2.119196653366089,
+      "learning_rate": 0.0002,
+      "loss": 2.4699,
+      "step": 2470
+    },
+    {
+      "epoch": 0.18479880774962743,
+      "grad_norm": 1.8797602653503418,
+      "learning_rate": 0.0002,
+      "loss": 2.4705,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1855439642324888,
+      "grad_norm": 1.985633134841919,
+      "learning_rate": 0.0002,
+      "loss": 2.4216,
+      "step": 2490
+    },
+    {
+      "epoch": 0.18628912071535023,
+      "grad_norm": 1.831678032875061,
+      "learning_rate": 0.0002,
+      "loss": 2.538,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1870342771982116,
+      "grad_norm": 1.7474476099014282,
+      "learning_rate": 0.0002,
+      "loss": 2.6348,
+      "step": 2510
+    },
+    {
+      "epoch": 0.18777943368107303,
+      "grad_norm": 1.7333717346191406,
+      "learning_rate": 0.0002,
+      "loss": 2.4576,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1885245901639344,
+      "grad_norm": 2.0141854286193848,
+      "learning_rate": 0.0002,
+      "loss": 2.4277,
+      "step": 2530
+    },
+    {
+      "epoch": 0.18926974664679583,
+      "grad_norm": 1.7154362201690674,
+      "learning_rate": 0.0002,
+      "loss": 2.428,
+      "step": 2540
+    },
+    {
+      "epoch": 0.19001490312965721,
+      "grad_norm": 2.11297607421875,
+      "learning_rate": 0.0002,
+      "loss": 2.4868,
+      "step": 2550
+    },
+    {
+      "epoch": 0.19076005961251863,
+      "grad_norm": 1.8067042827606201,
+      "learning_rate": 0.0002,
+      "loss": 2.4782,
+      "step": 2560
+    },
+    {
+      "epoch": 0.19150521609538004,
+      "grad_norm": 2.3354551792144775,
+      "learning_rate": 0.0002,
+      "loss": 2.5139,
+      "step": 2570
+    },
+    {
+      "epoch": 0.19225037257824143,
+      "grad_norm": 1.7424496412277222,
+      "learning_rate": 0.0002,
+      "loss": 2.4437,
+      "step": 2580
+    },
+    {
+      "epoch": 0.19299552906110284,
+      "grad_norm": 2.1761672496795654,
+      "learning_rate": 0.0002,
+      "loss": 2.4239,
+      "step": 2590
+    },
+    {
+      "epoch": 0.19374068554396423,
+      "grad_norm": 2.0857791900634766,
+      "learning_rate": 0.0002,
+      "loss": 2.2968,
+      "step": 2600
+    },
+    {
+      "epoch": 0.19448584202682564,
+      "grad_norm": 2.1806142330169678,
+      "learning_rate": 0.0002,
+      "loss": 2.571,
+      "step": 2610
+    },
+    {
+      "epoch": 0.19523099850968703,
+      "grad_norm": 1.7046921253204346,
+      "learning_rate": 0.0002,
+      "loss": 2.5633,
+      "step": 2620
+    },
+    {
+      "epoch": 0.19597615499254845,
+      "grad_norm": 1.876073956489563,
+      "learning_rate": 0.0002,
+      "loss": 2.4484,
+      "step": 2630
+    },
+    {
+      "epoch": 0.19672131147540983,
+      "grad_norm": 1.948696255683899,
+      "learning_rate": 0.0002,
+      "loss": 2.4158,
+      "step": 2640
+    },
+    {
+      "epoch": 0.19746646795827125,
+      "grad_norm": 2.022000551223755,
+      "learning_rate": 0.0002,
+      "loss": 2.4718,
+      "step": 2650
+    },
+    {
+      "epoch": 0.19821162444113263,
+      "grad_norm": 1.776353120803833,
+      "learning_rate": 0.0002,
+      "loss": 2.5832,
+      "step": 2660
+    },
+    {
+      "epoch": 0.19895678092399405,
+      "grad_norm": 1.9837231636047363,
+      "learning_rate": 0.0002,
+      "loss": 2.5152,
+      "step": 2670
+    },
+    {
+      "epoch": 0.19970193740685543,
+      "grad_norm": 1.8511583805084229,
+      "learning_rate": 0.0002,
+      "loss": 2.3081,
+      "step": 2680
+    },
+    {
+      "epoch": 0.20044709388971685,
+      "grad_norm": 1.7415555715560913,
+      "learning_rate": 0.0002,
+      "loss": 2.3807,
+      "step": 2690
+    },
+    {
+      "epoch": 0.20119225037257824,
+      "grad_norm": 1.9808768033981323,
+      "learning_rate": 0.0002,
+      "loss": 2.4238,
+      "step": 2700
+    },
+    {
+      "epoch": 0.20193740685543965,
+      "grad_norm": 1.6676552295684814,
+      "learning_rate": 0.0002,
+      "loss": 2.4074,
+      "step": 2710
+    },
+    {
+      "epoch": 0.20268256333830104,
+      "grad_norm": 2.046172857284546,
+      "learning_rate": 0.0002,
+      "loss": 2.5146,
+      "step": 2720
+    },
+    {
+      "epoch": 0.20342771982116245,
+      "grad_norm": 2.0430774688720703,
+      "learning_rate": 0.0002,
+      "loss": 2.348,
+      "step": 2730
+    },
+    {
+      "epoch": 0.20417287630402384,
+      "grad_norm": 1.978427767753601,
+      "learning_rate": 0.0002,
+      "loss": 2.4703,
+      "step": 2740
+    },
+    {
+      "epoch": 0.20491803278688525,
+      "grad_norm": 2.2341721057891846,
+      "learning_rate": 0.0002,
+      "loss": 2.424,
+      "step": 2750
+    },
+    {
+      "epoch": 0.20566318926974664,
+      "grad_norm": 1.874298095703125,
+      "learning_rate": 0.0002,
+      "loss": 2.5449,
+      "step": 2760
+    },
+    {
+      "epoch": 0.20640834575260805,
+      "grad_norm": 2.0746755599975586,
+      "learning_rate": 0.0002,
+      "loss": 2.4735,
+      "step": 2770
+    },
+    {
+      "epoch": 0.20715350223546944,
+      "grad_norm": 1.671237587928772,
+      "learning_rate": 0.0002,
+      "loss": 2.4185,
+      "step": 2780
+    },
+    {
+      "epoch": 0.20789865871833085,
+      "grad_norm": 1.7436130046844482,
+      "learning_rate": 0.0002,
+      "loss": 2.5611,
+      "step": 2790
+    },
+    {
+      "epoch": 0.20864381520119224,
+      "grad_norm": 1.991050124168396,
+      "learning_rate": 0.0002,
+      "loss": 2.5108,
+      "step": 2800
+    },
+    {
+      "epoch": 0.20938897168405365,
+      "grad_norm": 1.893971562385559,
+      "learning_rate": 0.0002,
+      "loss": 2.4371,
+      "step": 2810
+    },
+    {
+      "epoch": 0.21013412816691504,
+      "grad_norm": 2.048959970474243,
+      "learning_rate": 0.0002,
+      "loss": 2.4887,
+      "step": 2820
+    },
+    {
+      "epoch": 0.21087928464977646,
+      "grad_norm": 1.4141104221343994,
+      "learning_rate": 0.0002,
+      "loss": 2.3666,
+      "step": 2830
+    },
+    {
+      "epoch": 0.21162444113263784,
+      "grad_norm": 1.9136624336242676,
+      "learning_rate": 0.0002,
+      "loss": 2.5108,
+      "step": 2840
+    },
+    {
+      "epoch": 0.21236959761549926,
+      "grad_norm": 1.8413383960723877,
+      "learning_rate": 0.0002,
+      "loss": 2.5956,
+      "step": 2850
+    },
+    {
+      "epoch": 0.21311475409836064,
+      "grad_norm": 1.8322744369506836,
+      "learning_rate": 0.0002,
+      "loss": 2.2766,
+      "step": 2860
+    },
+    {
+      "epoch": 0.21385991058122206,
+      "grad_norm": 2.0261011123657227,
+      "learning_rate": 0.0002,
+      "loss": 2.5494,
+      "step": 2870
+    },
+    {
+      "epoch": 0.21460506706408347,
+      "grad_norm": 1.7044886350631714,
+      "learning_rate": 0.0002,
+      "loss": 2.453,
+      "step": 2880
+    },
+    {
+      "epoch": 0.21535022354694486,
+      "grad_norm": 1.8733025789260864,
+      "learning_rate": 0.0002,
+      "loss": 2.4455,
+      "step": 2890
+    },
+    {
+      "epoch": 0.21609538002980627,
+      "grad_norm": 1.7016903162002563,
+      "learning_rate": 0.0002,
+      "loss": 2.542,
+      "step": 2900
+    },
+    {
+      "epoch": 0.21684053651266766,
+      "grad_norm": 1.8775333166122437,
+      "learning_rate": 0.0002,
+      "loss": 2.5536,
+      "step": 2910
+    },
+    {
+      "epoch": 0.21758569299552907,
+      "grad_norm": 1.9875683784484863,
+      "learning_rate": 0.0002,
+      "loss": 2.365,
+      "step": 2920
+    },
+    {
+      "epoch": 0.21833084947839046,
+      "grad_norm": 1.9816802740097046,
+      "learning_rate": 0.0002,
+      "loss": 2.5781,
+      "step": 2930
+    },
+    {
+      "epoch": 0.21907600596125187,
+      "grad_norm": 2.2699484825134277,
+      "learning_rate": 0.0002,
+      "loss": 2.369,
+      "step": 2940
+    },
+    {
+      "epoch": 0.21982116244411326,
+      "grad_norm": 1.795271635055542,
+      "learning_rate": 0.0002,
+      "loss": 2.5805,
+      "step": 2950
+    },
+    {
+      "epoch": 0.22056631892697467,
+      "grad_norm": 2.1191015243530273,
+      "learning_rate": 0.0002,
+      "loss": 2.522,
+      "step": 2960
+    },
+    {
+      "epoch": 0.22131147540983606,
+      "grad_norm": 1.8573635816574097,
+      "learning_rate": 0.0002,
+      "loss": 2.5885,
+      "step": 2970
+    },
+    {
+      "epoch": 0.22205663189269748,
+      "grad_norm": 1.8743098974227905,
+      "learning_rate": 0.0002,
+      "loss": 2.4619,
+      "step": 2980
+    },
+    {
+      "epoch": 0.22280178837555886,
+      "grad_norm": 2.030251979827881,
+      "learning_rate": 0.0002,
+      "loss": 2.4544,
+      "step": 2990
+    },
+    {
+      "epoch": 0.22354694485842028,
+      "grad_norm": 1.7911770343780518,
+      "learning_rate": 0.0002,
+      "loss": 2.3729,
+      "step": 3000
+    },
+    {
+      "epoch": 0.22429210134128166,
+      "grad_norm": 1.7852712869644165,
+      "learning_rate": 0.0002,
+      "loss": 2.411,
+      "step": 3010
+    },
+    {
+      "epoch": 0.22503725782414308,
+      "grad_norm": 2.3478708267211914,
+      "learning_rate": 0.0002,
+      "loss": 2.4913,
+      "step": 3020
+    },
+    {
+      "epoch": 0.22578241430700446,
+      "grad_norm": 1.9944161176681519,
+      "learning_rate": 0.0002,
+      "loss": 2.5399,
+      "step": 3030
+    },
+    {
+      "epoch": 0.22652757078986588,
+      "grad_norm": 2.0674550533294678,
+      "learning_rate": 0.0002,
+      "loss": 2.4142,
+      "step": 3040
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 1.7529555559158325,
+      "learning_rate": 0.0002,
+      "loss": 2.4124,
+      "step": 3050
+    },
+    {
+      "epoch": 0.22801788375558868,
+      "grad_norm": 1.950682520866394,
+      "learning_rate": 0.0002,
+      "loss": 2.5125,
+      "step": 3060
+    },
+    {
+      "epoch": 0.22876304023845007,
+      "grad_norm": 1.9079375267028809,
+      "learning_rate": 0.0002,
+      "loss": 2.2691,
+      "step": 3070
+    },
+    {
+      "epoch": 0.22950819672131148,
+      "grad_norm": 2.3478922843933105,
+      "learning_rate": 0.0002,
+      "loss": 2.6288,
+      "step": 3080
+    },
+    {
+      "epoch": 0.23025335320417287,
+      "grad_norm": 1.7229981422424316,
+      "learning_rate": 0.0002,
+      "loss": 2.4475,
+      "step": 3090
+    },
+    {
+      "epoch": 0.23099850968703428,
+      "grad_norm": 1.970567226409912,
+      "learning_rate": 0.0002,
+      "loss": 2.6176,
+      "step": 3100
+    },
+    {
+      "epoch": 0.23174366616989567,
+      "grad_norm": 2.1065728664398193,
+      "learning_rate": 0.0002,
+      "loss": 2.5292,
+      "step": 3110
+    },
+    {
+      "epoch": 0.23248882265275708,
+      "grad_norm": 2.137432336807251,
+      "learning_rate": 0.0002,
+      "loss": 2.5971,
+      "step": 3120
+    },
+    {
+      "epoch": 0.23323397913561847,
+      "grad_norm": 1.9328407049179077,
+      "learning_rate": 0.0002,
+      "loss": 2.4158,
+      "step": 3130
+    },
+    {
+      "epoch": 0.23397913561847988,
+      "grad_norm": 1.981392741203308,
+      "learning_rate": 0.0002,
+      "loss": 2.524,
+      "step": 3140
+    },
+    {
+      "epoch": 0.23472429210134127,
+      "grad_norm": 2.098461866378784,
+      "learning_rate": 0.0002,
+      "loss": 2.4319,
+      "step": 3150
+    },
+    {
+      "epoch": 0.23546944858420268,
+      "grad_norm": 2.2548506259918213,
+      "learning_rate": 0.0002,
+      "loss": 2.608,
+      "step": 3160
+    },
+    {
+      "epoch": 0.23621460506706407,
+      "grad_norm": 2.1737749576568604,
+      "learning_rate": 0.0002,
+      "loss": 2.4759,
+      "step": 3170
+    },
+    {
+      "epoch": 0.23695976154992549,
+      "grad_norm": 2.0309979915618896,
+      "learning_rate": 0.0002,
+      "loss": 2.3443,
+      "step": 3180
+    },
+    {
+      "epoch": 0.23770491803278687,
+      "grad_norm": 2.115638494491577,
+      "learning_rate": 0.0002,
+      "loss": 2.5721,
+      "step": 3190
+    },
+    {
+      "epoch": 0.23845007451564829,
+      "grad_norm": 2.2513480186462402,
+      "learning_rate": 0.0002,
+      "loss": 2.4294,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2391952309985097,
+      "grad_norm": 1.984811544418335,
+      "learning_rate": 0.0002,
+      "loss": 2.5133,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2399403874813711,
+      "grad_norm": 1.9788024425506592,
+      "learning_rate": 0.0002,
+      "loss": 2.3574,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2406855439642325,
+      "grad_norm": 1.8824642896652222,
+      "learning_rate": 0.0002,
+      "loss": 2.3478,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2414307004470939,
+      "grad_norm": 1.9927945137023926,
+      "learning_rate": 0.0002,
+      "loss": 2.6732,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2421758569299553,
+      "grad_norm": 1.7224589586257935,
+      "learning_rate": 0.0002,
+      "loss": 2.5057,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2429210134128167,
+      "grad_norm": 1.787179708480835,
+      "learning_rate": 0.0002,
+      "loss": 2.4139,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2436661698956781,
+      "grad_norm": 1.9321664571762085,
+      "learning_rate": 0.0002,
+      "loss": 2.5715,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2444113263785395,
+      "grad_norm": 1.9817686080932617,
+      "learning_rate": 0.0002,
+      "loss": 2.5466,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2451564828614009,
+      "grad_norm": 1.8614225387573242,
+      "learning_rate": 0.0002,
+      "loss": 2.4936,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2459016393442623,
+      "grad_norm": 1.6355311870574951,
+      "learning_rate": 0.0002,
+      "loss": 2.3833,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2466467958271237,
+      "grad_norm": 1.9222790002822876,
+      "learning_rate": 0.0002,
+      "loss": 2.3884,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2473919523099851,
+      "grad_norm": 1.8250503540039062,
+      "learning_rate": 0.0002,
+      "loss": 2.4804,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2481371087928465,
+      "grad_norm": 1.9070334434509277,
+      "learning_rate": 0.0002,
+      "loss": 2.4982,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2488822652757079,
+      "grad_norm": 2.0816781520843506,
+      "learning_rate": 0.0002,
+      "loss": 2.188,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2496274217585693,
+      "grad_norm": 2.8250019550323486,
+      "learning_rate": 0.0002,
+      "loss": 2.1956,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2503725782414307,
+      "grad_norm": 2.0118353366851807,
+      "learning_rate": 0.0002,
+      "loss": 2.3953,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2511177347242921,
+      "grad_norm": 1.9895654916763306,
+      "learning_rate": 0.0002,
+      "loss": 2.4261,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2518628912071535,
+      "grad_norm": 1.6906330585479736,
+      "learning_rate": 0.0002,
+      "loss": 2.3543,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2526080476900149,
+      "grad_norm": 1.8363487720489502,
+      "learning_rate": 0.0002,
+      "loss": 2.6099,
+      "step": 3390
+    },
+    {
+      "epoch": 0.2533532041728763,
+      "grad_norm": 1.8634984493255615,
+      "learning_rate": 0.0002,
+      "loss": 2.4341,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2540983606557377,
+      "grad_norm": 1.7584588527679443,
+      "learning_rate": 0.0002,
+      "loss": 2.3242,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2548435171385991,
+      "grad_norm": 1.542602300643921,
+      "learning_rate": 0.0002,
+      "loss": 2.5926,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2555886736214605,
+      "grad_norm": 1.7898099422454834,
+      "learning_rate": 0.0002,
+      "loss": 2.5387,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2563338301043219,
+      "grad_norm": 1.9139480590820312,
+      "learning_rate": 0.0002,
+      "loss": 2.1946,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2570789865871833,
+      "grad_norm": 1.7210464477539062,
+      "learning_rate": 0.0002,
+      "loss": 2.6303,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2578241430700447,
+      "grad_norm": 2.010288953781128,
+      "learning_rate": 0.0002,
+      "loss": 2.6645,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2585692995529061,
+      "grad_norm": 2.2922821044921875,
+      "learning_rate": 0.0002,
+      "loss": 2.353,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2593144560357675,
+      "grad_norm": 1.7492296695709229,
+      "learning_rate": 0.0002,
+      "loss": 2.5581,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2600596125186289,
+      "grad_norm": 1.845924735069275,
+      "learning_rate": 0.0002,
+      "loss": 2.4135,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2608047690014903,
+      "grad_norm": 1.926946759223938,
+      "learning_rate": 0.0002,
+      "loss": 2.5024,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2615499254843517,
+      "grad_norm": 2.1195075511932373,
+      "learning_rate": 0.0002,
+      "loss": 2.663,
+      "step": 3510
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 1.7661279439926147,
+      "learning_rate": 0.0002,
+      "loss": 2.5196,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2630402384500745,
+      "grad_norm": 2.357307195663452,
+      "learning_rate": 0.0002,
+      "loss": 2.3874,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2637853949329359,
+      "grad_norm": 1.9379281997680664,
+      "learning_rate": 0.0002,
+      "loss": 2.5646,
+      "step": 3540
+    },
+    {
+      "epoch": 0.26453055141579734,
+      "grad_norm": 1.89481782913208,
+      "learning_rate": 0.0002,
+      "loss": 2.5963,
+      "step": 3550
+    },
+    {
+      "epoch": 0.26527570789865873,
+      "grad_norm": 1.9536447525024414,
+      "learning_rate": 0.0002,
+      "loss": 2.2617,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2660208643815201,
+      "grad_norm": 1.9169467687606812,
+      "learning_rate": 0.0002,
+      "loss": 2.4662,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2667660208643815,
+      "grad_norm": 2.1301636695861816,
+      "learning_rate": 0.0002,
+      "loss": 2.4761,
+      "step": 3580
+    },
+    {
+      "epoch": 0.26751117734724295,
+      "grad_norm": 1.9134154319763184,
+      "learning_rate": 0.0002,
+      "loss": 2.5022,
+      "step": 3590
+    },
+    {
+      "epoch": 0.26825633383010433,
+      "grad_norm": 2.04421067237854,
+      "learning_rate": 0.0002,
+      "loss": 2.3228,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2690014903129657,
+      "grad_norm": 1.7705532312393188,
+      "learning_rate": 0.0002,
+      "loss": 2.4066,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2697466467958271,
+      "grad_norm": 2.080298900604248,
+      "learning_rate": 0.0002,
+      "loss": 2.6129,
+      "step": 3620
+    },
+    {
+      "epoch": 0.27049180327868855,
+      "grad_norm": 1.6139470338821411,
+      "learning_rate": 0.0002,
+      "loss": 2.5801,
+      "step": 3630
+    },
+    {
+      "epoch": 0.27123695976154993,
+      "grad_norm": 2.054302453994751,
+      "learning_rate": 0.0002,
+      "loss": 2.3711,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2719821162444113,
+      "grad_norm": 2.338289260864258,
+      "learning_rate": 0.0002,
+      "loss": 2.5838,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 1.7468085289001465,
+      "learning_rate": 0.0002,
+      "loss": 2.1876,
+      "step": 3660
+    },
+    {
+      "epoch": 0.27347242921013415,
+      "grad_norm": 2.027275562286377,
+      "learning_rate": 0.0002,
+      "loss": 2.5105,
+      "step": 3670
+    },
+    {
+      "epoch": 0.27421758569299554,
+      "grad_norm": 2.229505777359009,
+      "learning_rate": 0.0002,
+      "loss": 2.6577,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2749627421758569,
+      "grad_norm": 1.9320789575576782,
+      "learning_rate": 0.0002,
+      "loss": 2.2976,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2757078986587183,
+      "grad_norm": 2.186555862426758,
+      "learning_rate": 0.0002,
+      "loss": 2.3886,
+      "step": 3700
+    },
+    {
+      "epoch": 0.27645305514157975,
+      "grad_norm": 1.918982982635498,
+      "learning_rate": 0.0002,
+      "loss": 2.5599,
+      "step": 3710
+    },
+    {
+      "epoch": 0.27719821162444114,
+      "grad_norm": 1.8183304071426392,
+      "learning_rate": 0.0002,
+      "loss": 2.5157,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2779433681073025,
+      "grad_norm": 1.994059443473816,
+      "learning_rate": 0.0002,
+      "loss": 2.5395,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2786885245901639,
+      "grad_norm": 1.7807224988937378,
+      "learning_rate": 0.0002,
+      "loss": 2.3498,
+      "step": 3740
+    },
+    {
+      "epoch": 0.27943368107302535,
+      "grad_norm": 2.1187775135040283,
+      "learning_rate": 0.0002,
+      "loss": 2.5581,
+      "step": 3750
+    },
+    {
+      "epoch": 0.28017883755588674,
+      "grad_norm": 1.8966116905212402,
+      "learning_rate": 0.0002,
+      "loss": 2.5547,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2809239940387481,
+      "grad_norm": 2.044442653656006,
+      "learning_rate": 0.0002,
+      "loss": 2.5357,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2816691505216095,
+      "grad_norm": 1.8322268724441528,
+      "learning_rate": 0.0002,
+      "loss": 2.5751,
+      "step": 3780
+    },
+    {
+      "epoch": 0.28241430700447095,
+      "grad_norm": 2.1471657752990723,
+      "learning_rate": 0.0002,
+      "loss": 2.5739,
+      "step": 3790
+    },
+    {
+      "epoch": 0.28315946348733234,
+      "grad_norm": 1.54653000831604,
+      "learning_rate": 0.0002,
+      "loss": 2.4101,
+      "step": 3800
+    },
+    {
+      "epoch": 0.28390461997019373,
+      "grad_norm": 2.0441126823425293,
+      "learning_rate": 0.0002,
+      "loss": 2.4344,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2846497764530551,
+      "grad_norm": 2.10286021232605,
+      "learning_rate": 0.0002,
+      "loss": 2.4168,
+      "step": 3820
+    },
+    {
+      "epoch": 0.28539493293591656,
+      "grad_norm": 2.1255736351013184,
+      "learning_rate": 0.0002,
+      "loss": 2.4448,
+      "step": 3830
+    },
+    {
+      "epoch": 0.28614008941877794,
+      "grad_norm": 1.830354928970337,
+      "learning_rate": 0.0002,
+      "loss": 2.3284,
+      "step": 3840
+    },
+    {
+      "epoch": 0.28688524590163933,
+      "grad_norm": 1.9017317295074463,
+      "learning_rate": 0.0002,
+      "loss": 2.51,
+      "step": 3850
+    },
+    {
+      "epoch": 0.28763040238450077,
+      "grad_norm": 2.0456435680389404,
+      "learning_rate": 0.0002,
+      "loss": 2.5099,
+      "step": 3860
+    },
+    {
+      "epoch": 0.28837555886736216,
+      "grad_norm": 2.036862373352051,
+      "learning_rate": 0.0002,
+      "loss": 2.418,
+      "step": 3870
+    },
+    {
+      "epoch": 0.28912071535022354,
+      "grad_norm": 1.9456264972686768,
+      "learning_rate": 0.0002,
+      "loss": 2.371,
+      "step": 3880
+    },
+    {
+      "epoch": 0.28986587183308493,
+      "grad_norm": 2.3842172622680664,
+      "learning_rate": 0.0002,
+      "loss": 2.5213,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2906110283159464,
+      "grad_norm": 1.9152741432189941,
+      "learning_rate": 0.0002,
+      "loss": 2.5191,
+      "step": 3900
+    },
+    {
+      "epoch": 0.29135618479880776,
+      "grad_norm": 1.9884922504425049,
+      "learning_rate": 0.0002,
+      "loss": 2.4272,
+      "step": 3910
+    },
+    {
+      "epoch": 0.29210134128166915,
+      "grad_norm": 1.7578502893447876,
+      "learning_rate": 0.0002,
+      "loss": 2.4975,
+      "step": 3920
+    },
+    {
+      "epoch": 0.29284649776453053,
+      "grad_norm": 2.5302071571350098,
+      "learning_rate": 0.0002,
+      "loss": 2.3564,
+      "step": 3930
+    },
+    {
+      "epoch": 0.293591654247392,
+      "grad_norm": 2.2181880474090576,
+      "learning_rate": 0.0002,
+      "loss": 2.6365,
+      "step": 3940
+    },
+    {
+      "epoch": 0.29433681073025336,
+      "grad_norm": 1.8359179496765137,
+      "learning_rate": 0.0002,
+      "loss": 2.4825,
+      "step": 3950
+    },
+    {
+      "epoch": 0.29508196721311475,
+      "grad_norm": 1.8330084085464478,
+      "learning_rate": 0.0002,
+      "loss": 2.3746,
+      "step": 3960
+    },
+    {
+      "epoch": 0.29582712369597614,
+      "grad_norm": 2.0700743198394775,
+      "learning_rate": 0.0002,
+      "loss": 2.4456,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2965722801788376,
+      "grad_norm": 2.0610342025756836,
+      "learning_rate": 0.0002,
+      "loss": 2.5419,
+      "step": 3980
+    },
+    {
+      "epoch": 0.29731743666169896,
+      "grad_norm": 2.0633559226989746,
+      "learning_rate": 0.0002,
+      "loss": 2.4911,
+      "step": 3990
+    },
+    {
+      "epoch": 0.29806259314456035,
+      "grad_norm": 1.894938588142395,
+      "learning_rate": 0.0002,
+      "loss": 2.5311,
+      "step": 4000
+    },
+    {
+      "epoch": 0.29880774962742174,
+      "grad_norm": 2.153447389602661,
+      "learning_rate": 0.0002,
+      "loss": 2.5135,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2995529061102832,
+      "grad_norm": 1.8810901641845703,
+      "learning_rate": 0.0002,
+      "loss": 2.5123,
+      "step": 4020
+    },
+    {
+      "epoch": 0.30029806259314457,
+      "grad_norm": 2.828382730484009,
+      "learning_rate": 0.0002,
+      "loss": 2.4624,
+      "step": 4030
+    },
+    {
+      "epoch": 0.30104321907600595,
+      "grad_norm": 1.5955597162246704,
+      "learning_rate": 0.0002,
+      "loss": 2.4029,
+      "step": 4040
+    },
+    {
+      "epoch": 0.30178837555886734,
+      "grad_norm": 2.158923864364624,
+      "learning_rate": 0.0002,
+      "loss": 2.3133,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3025335320417288,
+      "grad_norm": 2.044170379638672,
+      "learning_rate": 0.0002,
+      "loss": 2.5106,
+      "step": 4060
+    },
+    {
+      "epoch": 0.30327868852459017,
+      "grad_norm": 2.1641745567321777,
+      "learning_rate": 0.0002,
+      "loss": 2.3511,
+      "step": 4070
+    },
+    {
+      "epoch": 0.30402384500745155,
+      "grad_norm": 1.7815321683883667,
+      "learning_rate": 0.0002,
+      "loss": 2.47,
+      "step": 4080
+    },
+    {
+      "epoch": 0.30476900149031294,
+      "grad_norm": 1.7963491678237915,
+      "learning_rate": 0.0002,
+      "loss": 2.431,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3055141579731744,
+      "grad_norm": 1.8735893964767456,
+      "learning_rate": 0.0002,
+      "loss": 2.4197,
+      "step": 4100
+    },
+    {
+      "epoch": 0.30625931445603577,
+      "grad_norm": 1.8785910606384277,
+      "learning_rate": 0.0002,
+      "loss": 2.4681,
+      "step": 4110
+    },
+    {
+      "epoch": 0.30700447093889716,
+      "grad_norm": 2.711383104324341,
+      "learning_rate": 0.0002,
+      "loss": 2.5373,
+      "step": 4120
+    },
+    {
+      "epoch": 0.30774962742175854,
+      "grad_norm": 1.837888479232788,
+      "learning_rate": 0.0002,
+      "loss": 2.4393,
+      "step": 4130
+    },
+    {
+      "epoch": 0.30849478390462,
+      "grad_norm": 2.044309616088867,
+      "learning_rate": 0.0002,
+      "loss": 2.5764,
+      "step": 4140
+    },
+    {
+      "epoch": 0.30923994038748137,
+      "grad_norm": 2.052886486053467,
+      "learning_rate": 0.0002,
+      "loss": 2.5038,
+      "step": 4150
+    },
+    {
+      "epoch": 0.30998509687034276,
+      "grad_norm": 2.0241613388061523,
+      "learning_rate": 0.0002,
+      "loss": 2.6383,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3107302533532042,
+      "grad_norm": 1.9976118803024292,
+      "learning_rate": 0.0002,
+      "loss": 2.5346,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3114754098360656,
+      "grad_norm": 1.7011710405349731,
+      "learning_rate": 0.0002,
+      "loss": 2.5725,
+      "step": 4180
+    },
+    {
+      "epoch": 0.312220566318927,
+      "grad_norm": 2.266010284423828,
+      "learning_rate": 0.0002,
+      "loss": 2.473,
+      "step": 4190
+    },
+    {
+      "epoch": 0.31296572280178836,
+      "grad_norm": 1.793825626373291,
+      "learning_rate": 0.0002,
+      "loss": 2.3846,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3137108792846498,
+      "grad_norm": 2.258878469467163,
+      "learning_rate": 0.0002,
+      "loss": 2.6184,
+      "step": 4210
+    },
+    {
+      "epoch": 0.3144560357675112,
+      "grad_norm": 2.1145365238189697,
+      "learning_rate": 0.0002,
+      "loss": 2.5497,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3152011922503726,
+      "grad_norm": 2.1772029399871826,
+      "learning_rate": 0.0002,
+      "loss": 2.4528,
+      "step": 4230
+    },
+    {
+      "epoch": 0.31594634873323396,
+      "grad_norm": 2.5457372665405273,
+      "learning_rate": 0.0002,
+      "loss": 2.6091,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3166915052160954,
+      "grad_norm": 1.8056083917617798,
+      "learning_rate": 0.0002,
+      "loss": 2.48,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3174366616989568,
+      "grad_norm": 2.0121543407440186,
+      "learning_rate": 0.0002,
+      "loss": 2.5622,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3181818181818182,
+      "grad_norm": 2.6106455326080322,
+      "learning_rate": 0.0002,
+      "loss": 2.5243,
+      "step": 4270
+    },
+    {
+      "epoch": 0.31892697466467956,
+      "grad_norm": 1.9244214296340942,
+      "learning_rate": 0.0002,
+      "loss": 2.4204,
+      "step": 4280
+    },
+    {
+      "epoch": 0.319672131147541,
+      "grad_norm": 2.1080305576324463,
+      "learning_rate": 0.0002,
+      "loss": 2.4296,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3204172876304024,
+      "grad_norm": 2.309900999069214,
+      "learning_rate": 0.0002,
+      "loss": 2.4689,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3211624441132638,
+      "grad_norm": 2.2716972827911377,
+      "learning_rate": 0.0002,
+      "loss": 2.3885,
+      "step": 4310
+    },
+    {
+      "epoch": 0.32190760059612517,
+      "grad_norm": 2.2576944828033447,
+      "learning_rate": 0.0002,
+      "loss": 2.3281,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3226527570789866,
+      "grad_norm": 1.943267583847046,
+      "learning_rate": 0.0002,
+      "loss": 2.4674,
+      "step": 4330
+    },
+    {
+      "epoch": 0.323397913561848,
+      "grad_norm": 1.886509895324707,
+      "learning_rate": 0.0002,
+      "loss": 2.491,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3241430700447094,
+      "grad_norm": 1.9226173162460327,
+      "learning_rate": 0.0002,
+      "loss": 2.4899,
+      "step": 4350
+    },
+    {
+      "epoch": 0.32488822652757077,
+      "grad_norm": 2.016324043273926,
+      "learning_rate": 0.0002,
+      "loss": 2.4347,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3256333830104322,
+      "grad_norm": 1.8798201084136963,
+      "learning_rate": 0.0002,
+      "loss": 2.7045,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3263785394932936,
+      "grad_norm": 2.0897727012634277,
+      "learning_rate": 0.0002,
+      "loss": 2.4779,
+      "step": 4380
+    },
+    {
+      "epoch": 0.327123695976155,
+      "grad_norm": 2.4956274032592773,
+      "learning_rate": 0.0002,
+      "loss": 2.5188,
+      "step": 4390
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 2.08056378364563,
+      "learning_rate": 0.0002,
+      "loss": 2.4447,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3286140089418778,
+      "grad_norm": 2.2066187858581543,
+      "learning_rate": 0.0002,
+      "loss": 2.5474,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3293591654247392,
+      "grad_norm": 1.8168262243270874,
+      "learning_rate": 0.0002,
+      "loss": 2.5207,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3301043219076006,
+      "grad_norm": 2.649477481842041,
+      "learning_rate": 0.0002,
+      "loss": 2.4138,
+      "step": 4430
+    },
+    {
+      "epoch": 0.33084947839046197,
+      "grad_norm": 2.3449325561523438,
+      "learning_rate": 0.0002,
+      "loss": 2.5474,
+      "step": 4440
+    },
+    {
+      "epoch": 0.3315946348733234,
+      "grad_norm": 1.7262647151947021,
+      "learning_rate": 0.0002,
+      "loss": 2.5911,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3323397913561848,
+      "grad_norm": 2.0296733379364014,
+      "learning_rate": 0.0002,
+      "loss": 2.5417,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3330849478390462,
+      "grad_norm": 2.036099672317505,
+      "learning_rate": 0.0002,
+      "loss": 2.3912,
+      "step": 4470
+    },
+    {
+      "epoch": 0.33383010432190763,
+      "grad_norm": 1.9934238195419312,
+      "learning_rate": 0.0002,
+      "loss": 2.5022,
+      "step": 4480
+    },
+    {
+      "epoch": 0.334575260804769,
+      "grad_norm": 2.1589412689208984,
+      "learning_rate": 0.0002,
+      "loss": 2.5989,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3353204172876304,
+      "grad_norm": 1.8449981212615967,
+      "learning_rate": 0.0002,
+      "loss": 2.233,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3360655737704918,
+      "grad_norm": 1.7662941217422485,
+      "learning_rate": 0.0002,
+      "loss": 2.5521,
+      "step": 4510
+    },
+    {
+      "epoch": 0.33681073025335323,
+      "grad_norm": 2.3483681678771973,
+      "learning_rate": 0.0002,
+      "loss": 2.4731,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3375558867362146,
+      "grad_norm": 1.8929402828216553,
+      "learning_rate": 0.0002,
+      "loss": 2.4925,
+      "step": 4530
+    },
+    {
+      "epoch": 0.338301043219076,
+      "grad_norm": 1.988408088684082,
+      "learning_rate": 0.0002,
+      "loss": 2.5581,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3390461997019374,
+      "grad_norm": 2.3717474937438965,
+      "learning_rate": 0.0002,
+      "loss": 2.5353,
+      "step": 4550
+    },
+    {
+      "epoch": 0.33979135618479883,
+      "grad_norm": 2.058992862701416,
+      "learning_rate": 0.0002,
+      "loss": 2.4383,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3405365126676602,
+      "grad_norm": 2.3037335872650146,
+      "learning_rate": 0.0002,
+      "loss": 2.6187,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3412816691505216,
+      "grad_norm": 2.0833449363708496,
+      "learning_rate": 0.0002,
+      "loss": 2.6367,
+      "step": 4580
+    },
+    {
+      "epoch": 0.342026825633383,
+      "grad_norm": 1.6986418962478638,
+      "learning_rate": 0.0002,
+      "loss": 2.343,
+      "step": 4590
+    },
+    {
+      "epoch": 0.34277198211624443,
+      "grad_norm": 4.1914472579956055,
+      "learning_rate": 0.0002,
+      "loss": 2.6611,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3435171385991058,
+      "grad_norm": 2.1022298336029053,
+      "learning_rate": 0.0002,
+      "loss": 2.5226,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3442622950819672,
+      "grad_norm": 2.1484179496765137,
+      "learning_rate": 0.0002,
+      "loss": 2.3475,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3450074515648286,
+      "grad_norm": 2.1597182750701904,
+      "learning_rate": 0.0002,
+      "loss": 2.5566,
+      "step": 4630
+    },
+    {
+      "epoch": 0.34575260804769004,
+      "grad_norm": 2.0153634548187256,
+      "learning_rate": 0.0002,
+      "loss": 2.399,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3464977645305514,
+      "grad_norm": 4.205637454986572,
+      "learning_rate": 0.0002,
+      "loss": 2.4738,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3472429210134128,
+      "grad_norm": 1.965390682220459,
+      "learning_rate": 0.0002,
+      "loss": 2.5928,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3479880774962742,
+      "grad_norm": 2.318188428878784,
+      "learning_rate": 0.0002,
+      "loss": 2.3906,
+      "step": 4670
+    },
+    {
+      "epoch": 0.34873323397913564,
+      "grad_norm": 2.16817045211792,
+      "learning_rate": 0.0002,
+      "loss": 2.5063,
+      "step": 4680
+    },
+    {
+      "epoch": 0.349478390461997,
+      "grad_norm": 1.8313651084899902,
+      "learning_rate": 0.0002,
+      "loss": 2.3634,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3502235469448584,
+      "grad_norm": 2.083974599838257,
+      "learning_rate": 0.0002,
+      "loss": 2.4984,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3509687034277198,
+      "grad_norm": 2.033154249191284,
+      "learning_rate": 0.0002,
+      "loss": 2.5167,
+      "step": 4710
+    },
+    {
+      "epoch": 0.35171385991058124,
+      "grad_norm": 1.7030832767486572,
+      "learning_rate": 0.0002,
+      "loss": 2.3332,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3524590163934426,
+      "grad_norm": 2.236445188522339,
+      "learning_rate": 0.0002,
+      "loss": 2.5903,
+      "step": 4730
+    },
+    {
+      "epoch": 0.353204172876304,
+      "grad_norm": 2.0333669185638428,
+      "learning_rate": 0.0002,
+      "loss": 2.6599,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3539493293591654,
+      "grad_norm": 2.043572425842285,
+      "learning_rate": 0.0002,
+      "loss": 2.5685,
+      "step": 4750
+    },
+    {
+      "epoch": 0.35469448584202684,
+      "grad_norm": 1.7955186367034912,
+      "learning_rate": 0.0002,
+      "loss": 2.4531,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3554396423248882,
+      "grad_norm": 2.064957857131958,
+      "learning_rate": 0.0002,
+      "loss": 2.5,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3561847988077496,
+      "grad_norm": 2.2814271450042725,
+      "learning_rate": 0.0002,
+      "loss": 2.3527,
+      "step": 4780
+    },
+    {
+      "epoch": 0.356929955290611,
+      "grad_norm": 2.179020881652832,
+      "learning_rate": 0.0002,
+      "loss": 2.4888,
+      "step": 4790
+    },
+    {
+      "epoch": 0.35767511177347244,
+      "grad_norm": 1.7410861253738403,
+      "learning_rate": 0.0002,
+      "loss": 2.3833,
+      "step": 4800
+    },
+    {
+      "epoch": 0.35842026825633383,
+      "grad_norm": 2.1326522827148438,
+      "learning_rate": 0.0002,
+      "loss": 2.5156,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3591654247391952,
+      "grad_norm": 2.075561761856079,
+      "learning_rate": 0.0002,
+      "loss": 2.4855,
+      "step": 4820
+    },
+    {
+      "epoch": 0.35991058122205666,
+      "grad_norm": 2.168584108352661,
+      "learning_rate": 0.0002,
+      "loss": 2.5904,
+      "step": 4830
+    },
+    {
+      "epoch": 0.36065573770491804,
+      "grad_norm": 2.0109267234802246,
+      "learning_rate": 0.0002,
+      "loss": 2.502,
+      "step": 4840
+    },
+    {
+      "epoch": 0.36140089418777943,
+      "grad_norm": 1.8693374395370483,
+      "learning_rate": 0.0002,
+      "loss": 2.531,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3621460506706408,
+      "grad_norm": 2.561384439468384,
+      "learning_rate": 0.0002,
+      "loss": 2.2466,
+      "step": 4860
+    },
+    {
+      "epoch": 0.36289120715350226,
+      "grad_norm": 2.319011688232422,
+      "learning_rate": 0.0002,
+      "loss": 2.4845,
+      "step": 4870
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 2.0164289474487305,
+      "learning_rate": 0.0002,
+      "loss": 2.6448,
+      "step": 4880
+    },
+    {
+      "epoch": 0.36438152011922503,
+      "grad_norm": 1.8674942255020142,
+      "learning_rate": 0.0002,
+      "loss": 2.4344,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3651266766020864,
+      "grad_norm": 1.8700525760650635,
+      "learning_rate": 0.0002,
+      "loss": 2.412,
+      "step": 4900
+    },
+    {
+      "epoch": 0.36587183308494786,
+      "grad_norm": 1.9874043464660645,
+      "learning_rate": 0.0002,
+      "loss": 2.5567,
+      "step": 4910
+    },
+    {
+      "epoch": 0.36661698956780925,
+      "grad_norm": 1.914652943611145,
+      "learning_rate": 0.0002,
+      "loss": 2.5984,
+      "step": 4920
+    },
+    {
+      "epoch": 0.36736214605067063,
+      "grad_norm": 2.1003236770629883,
+      "learning_rate": 0.0002,
+      "loss": 2.6851,
+      "step": 4930
+    },
+    {
+      "epoch": 0.368107302533532,
+      "grad_norm": 1.9648061990737915,
+      "learning_rate": 0.0002,
+      "loss": 2.4207,
+      "step": 4940
+    },
+    {
+      "epoch": 0.36885245901639346,
+      "grad_norm": 2.0232656002044678,
+      "learning_rate": 0.0002,
+      "loss": 2.4922,
+      "step": 4950
+    },
+    {
+      "epoch": 0.36959761549925485,
+      "grad_norm": 2.0865535736083984,
+      "learning_rate": 0.0002,
+      "loss": 2.6409,
+      "step": 4960
+    },
+    {
+      "epoch": 0.37034277198211624,
+      "grad_norm": 1.8709211349487305,
+      "learning_rate": 0.0002,
+      "loss": 2.3267,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3710879284649776,
+      "grad_norm": 2.093190908432007,
+      "learning_rate": 0.0002,
+      "loss": 2.3293,
+      "step": 4980
+    },
+    {
+      "epoch": 0.37183308494783907,
+      "grad_norm": 2.2843077182769775,
+      "learning_rate": 0.0002,
+      "loss": 2.5344,
+      "step": 4990
+    },
+    {
+      "epoch": 0.37257824143070045,
+      "grad_norm": 1.9902077913284302,
+      "learning_rate": 0.0002,
+      "loss": 2.4221,
+      "step": 5000
+    },
+    {
+      "epoch": 0.37332339791356184,
+      "grad_norm": 1.827880859375,
+      "learning_rate": 0.0002,
+      "loss": 2.3791,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3740685543964232,
+      "grad_norm": 2.0986554622650146,
+      "learning_rate": 0.0002,
+      "loss": 2.5067,
+      "step": 5020
+    },
+    {
+      "epoch": 0.37481371087928467,
+      "grad_norm": 2.260951519012451,
+      "learning_rate": 0.0002,
+      "loss": 2.5156,
+      "step": 5030
+    },
+    {
+      "epoch": 0.37555886736214605,
+      "grad_norm": 2.1205878257751465,
+      "learning_rate": 0.0002,
+      "loss": 2.524,
+      "step": 5040
+    },
+    {
+      "epoch": 0.37630402384500744,
+      "grad_norm": 1.7332857847213745,
+      "learning_rate": 0.0002,
+      "loss": 2.4875,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3770491803278688,
+      "grad_norm": 2.1427862644195557,
+      "learning_rate": 0.0002,
+      "loss": 2.3529,
+      "step": 5060
+    },
+    {
+      "epoch": 0.37779433681073027,
+      "grad_norm": 2.6252365112304688,
+      "learning_rate": 0.0002,
+      "loss": 2.6104,
+      "step": 5070
+    },
+    {
+      "epoch": 0.37853949329359166,
+      "grad_norm": 2.0142056941986084,
+      "learning_rate": 0.0002,
+      "loss": 2.5434,
+      "step": 5080
+    },
+    {
+      "epoch": 0.37928464977645304,
+      "grad_norm": 2.4451770782470703,
+      "learning_rate": 0.0002,
+      "loss": 2.4673,
+      "step": 5090
+    },
+    {
+      "epoch": 0.38002980625931443,
+      "grad_norm": 2.5081820487976074,
+      "learning_rate": 0.0002,
+      "loss": 2.4091,
+      "step": 5100
+    },
+    {
+      "epoch": 0.38077496274217587,
+      "grad_norm": 2.149099588394165,
+      "learning_rate": 0.0002,
+      "loss": 2.4014,
+      "step": 5110
+    },
+    {
+      "epoch": 0.38152011922503726,
+      "grad_norm": 2.3913164138793945,
+      "learning_rate": 0.0002,
+      "loss": 2.437,
+      "step": 5120
+    },
+    {
+      "epoch": 0.38226527570789864,
+      "grad_norm": 2.1281864643096924,
+      "learning_rate": 0.0002,
+      "loss": 2.5205,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3830104321907601,
+      "grad_norm": 1.710132122039795,
+      "learning_rate": 0.0002,
+      "loss": 2.3987,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3837555886736215,
+      "grad_norm": 2.173602342605591,
+      "learning_rate": 0.0002,
+      "loss": 2.5606,
+      "step": 5150
+    },
+    {
+      "epoch": 0.38450074515648286,
+      "grad_norm": 2.032154083251953,
+      "learning_rate": 0.0002,
+      "loss": 2.5223,
+      "step": 5160
+    },
+    {
+      "epoch": 0.38524590163934425,
+      "grad_norm": 2.1516082286834717,
+      "learning_rate": 0.0002,
+      "loss": 2.4851,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3859910581222057,
+      "grad_norm": 1.9636366367340088,
+      "learning_rate": 0.0002,
+      "loss": 2.3907,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3867362146050671,
+      "grad_norm": 2.3789303302764893,
+      "learning_rate": 0.0002,
+      "loss": 2.3962,
+      "step": 5190
+    },
+    {
+      "epoch": 0.38748137108792846,
+      "grad_norm": 2.102897882461548,
+      "learning_rate": 0.0002,
+      "loss": 2.6373,
+      "step": 5200
+    },
+    {
+      "epoch": 0.38822652757078985,
+      "grad_norm": 2.0494508743286133,
+      "learning_rate": 0.0002,
+      "loss": 2.4939,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3889716840536513,
+      "grad_norm": 1.95903480052948,
+      "learning_rate": 0.0002,
+      "loss": 2.4768,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3897168405365127,
+      "grad_norm": 2.15665864944458,
+      "learning_rate": 0.0002,
+      "loss": 2.5624,
+      "step": 5230
+    },
+    {
+      "epoch": 0.39046199701937406,
+      "grad_norm": 2.1582460403442383,
+      "learning_rate": 0.0002,
+      "loss": 2.4983,
+      "step": 5240
+    },
+    {
+      "epoch": 0.39120715350223545,
+      "grad_norm": 1.939427375793457,
+      "learning_rate": 0.0002,
+      "loss": 2.4378,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3919523099850969,
+      "grad_norm": 1.7251907587051392,
+      "learning_rate": 0.0002,
+      "loss": 2.3679,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3926974664679583,
+      "grad_norm": 1.9326486587524414,
+      "learning_rate": 0.0002,
+      "loss": 2.5102,
+      "step": 5270
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 2.2457611560821533,
+      "learning_rate": 0.0002,
+      "loss": 2.6181,
+      "step": 5280
+    },
+    {
+      "epoch": 0.39418777943368105,
+      "grad_norm": 2.027223825454712,
+      "learning_rate": 0.0002,
+      "loss": 2.3882,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3949329359165425,
+      "grad_norm": 2.135723829269409,
+      "learning_rate": 0.0002,
+      "loss": 2.3614,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3956780923994039,
+      "grad_norm": 2.2400708198547363,
+      "learning_rate": 0.0002,
+      "loss": 2.4794,
+      "step": 5310
+    },
+    {
+      "epoch": 0.39642324888226527,
+      "grad_norm": 2.0185799598693848,
+      "learning_rate": 0.0002,
+      "loss": 2.5809,
+      "step": 5320
+    },
+    {
+      "epoch": 0.39716840536512665,
+      "grad_norm": 2.2141106128692627,
+      "learning_rate": 0.0002,
+      "loss": 2.5126,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3979135618479881,
+      "grad_norm": 2.2849326133728027,
+      "learning_rate": 0.0002,
+      "loss": 2.3855,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3986587183308495,
+      "grad_norm": 1.7335988283157349,
+      "learning_rate": 0.0002,
+      "loss": 2.3463,
+      "step": 5350
+    },
+    {
+      "epoch": 0.39940387481371087,
+      "grad_norm": 1.9355789422988892,
+      "learning_rate": 0.0002,
+      "loss": 2.5056,
+      "step": 5360
+    },
+    {
+      "epoch": 0.40014903129657226,
+      "grad_norm": 1.7740095853805542,
+      "learning_rate": 0.0002,
+      "loss": 2.5791,
+      "step": 5370
+    },
+    {
+      "epoch": 0.4008941877794337,
+      "grad_norm": 1.7497676610946655,
+      "learning_rate": 0.0002,
+      "loss": 2.4308,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4016393442622951,
+      "grad_norm": 2.0560314655303955,
+      "learning_rate": 0.0002,
+      "loss": 2.4581,
+      "step": 5390
+    },
+    {
+      "epoch": 0.40238450074515647,
+      "grad_norm": 2.065265655517578,
+      "learning_rate": 0.0002,
+      "loss": 2.5172,
+      "step": 5400
+    },
+    {
+      "epoch": 0.40312965722801786,
+      "grad_norm": 2.282453775405884,
+      "learning_rate": 0.0002,
+      "loss": 2.3409,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4038748137108793,
+      "grad_norm": 2.3466989994049072,
+      "learning_rate": 0.0002,
+      "loss": 2.5305,
+      "step": 5420
+    },
+    {
+      "epoch": 0.4046199701937407,
+      "grad_norm": 2.2113606929779053,
+      "learning_rate": 0.0002,
+      "loss": 2.6027,
+      "step": 5430
+    },
+    {
+      "epoch": 0.40536512667660207,
+      "grad_norm": 2.149338483810425,
+      "learning_rate": 0.0002,
+      "loss": 2.3704,
+      "step": 5440
+    },
+    {
+      "epoch": 0.4061102831594635,
+      "grad_norm": 2.129307270050049,
+      "learning_rate": 0.0002,
+      "loss": 2.3367,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4068554396423249,
+      "grad_norm": 2.2620790004730225,
+      "learning_rate": 0.0002,
+      "loss": 2.5256,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4076005961251863,
+      "grad_norm": 2.623889684677124,
+      "learning_rate": 0.0002,
+      "loss": 2.5516,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4083457526080477,
+      "grad_norm": 1.7352521419525146,
+      "learning_rate": 0.0002,
+      "loss": 2.2739,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 2.0829408168792725,
+      "learning_rate": 0.0002,
+      "loss": 2.5512,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4098360655737705,
+      "grad_norm": 2.000159978866577,
+      "learning_rate": 0.0002,
+      "loss": 2.3084,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4105812220566319,
+      "grad_norm": 2.665837526321411,
+      "learning_rate": 0.0002,
+      "loss": 2.3492,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4113263785394933,
+      "grad_norm": 2.220322608947754,
+      "learning_rate": 0.0002,
+      "loss": 2.5528,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4120715350223547,
+      "grad_norm": 2.322227954864502,
+      "learning_rate": 0.0002,
+      "loss": 2.4756,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4128166915052161,
+      "grad_norm": 2.0118496417999268,
+      "learning_rate": 0.0002,
+      "loss": 2.564,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4135618479880775,
+      "grad_norm": 2.0772318840026855,
+      "learning_rate": 0.0002,
+      "loss": 2.583,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4143070044709389,
+      "grad_norm": 2.002246141433716,
+      "learning_rate": 0.0002,
+      "loss": 2.6718,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4150521609538003,
+      "grad_norm": 2.208174228668213,
+      "learning_rate": 0.0002,
+      "loss": 2.4075,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4157973174366617,
+      "grad_norm": 2.2165260314941406,
+      "learning_rate": 0.0002,
+      "loss": 2.6347,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4165424739195231,
+      "grad_norm": 2.365262508392334,
+      "learning_rate": 0.0002,
+      "loss": 2.5209,
+      "step": 5590
+    },
+    {
+      "epoch": 0.4172876304023845,
+      "grad_norm": 2.264592170715332,
+      "learning_rate": 0.0002,
+      "loss": 2.4414,
+      "step": 5600
+    },
+    {
+      "epoch": 0.4180327868852459,
+      "grad_norm": 1.9426195621490479,
+      "learning_rate": 0.0002,
+      "loss": 2.313,
+      "step": 5610
+    },
+    {
+      "epoch": 0.4187779433681073,
+      "grad_norm": 2.155670642852783,
+      "learning_rate": 0.0002,
+      "loss": 2.5946,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4195230998509687,
+      "grad_norm": 1.9506397247314453,
+      "learning_rate": 0.0002,
+      "loss": 2.4709,
+      "step": 5630
+    },
+    {
+      "epoch": 0.4202682563338301,
+      "grad_norm": 2.3226125240325928,
+      "learning_rate": 0.0002,
+      "loss": 2.3527,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4210134128166915,
+      "grad_norm": 1.6601407527923584,
+      "learning_rate": 0.0002,
+      "loss": 2.4622,
+      "step": 5650
+    },
+    {
+      "epoch": 0.4217585692995529,
+      "grad_norm": 2.126014471054077,
+      "learning_rate": 0.0002,
+      "loss": 2.4602,
+      "step": 5660
+    },
+    {
+      "epoch": 0.4225037257824143,
+      "grad_norm": 2.292633295059204,
+      "learning_rate": 0.0002,
+      "loss": 2.7211,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4232488822652757,
+      "grad_norm": 2.1793317794799805,
+      "learning_rate": 0.0002,
+      "loss": 2.4626,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4239940387481371,
+      "grad_norm": 2.196563720703125,
+      "learning_rate": 0.0002,
+      "loss": 2.5533,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4247391952309985,
+      "grad_norm": 2.425448417663574,
+      "learning_rate": 0.0002,
+      "loss": 2.4851,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4254843517138599,
+      "grad_norm": 2.2193682193756104,
+      "learning_rate": 0.0002,
+      "loss": 2.3698,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4262295081967213,
+      "grad_norm": 2.054805040359497,
+      "learning_rate": 0.0002,
+      "loss": 2.5059,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4269746646795827,
+      "grad_norm": 1.8185921907424927,
+      "learning_rate": 0.0002,
+      "loss": 2.5176,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4277198211624441,
+      "grad_norm": 1.9967896938323975,
+      "learning_rate": 0.0002,
+      "loss": 2.4504,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4284649776453055,
+      "grad_norm": 2.662285327911377,
+      "learning_rate": 0.0002,
+      "loss": 2.4244,
+      "step": 5750
+    },
+    {
+      "epoch": 0.42921013412816694,
+      "grad_norm": 2.16428804397583,
+      "learning_rate": 0.0002,
+      "loss": 2.2946,
+      "step": 5760
+    },
+    {
+      "epoch": 0.42995529061102833,
+      "grad_norm": 2.0115864276885986,
+      "learning_rate": 0.0002,
+      "loss": 2.4657,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4307004470938897,
+      "grad_norm": 2.388542413711548,
+      "learning_rate": 0.0002,
+      "loss": 2.5151,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4314456035767511,
+      "grad_norm": 2.0310490131378174,
+      "learning_rate": 0.0002,
+      "loss": 2.3604,
+      "step": 5790
+    },
+    {
+      "epoch": 0.43219076005961254,
+      "grad_norm": 1.8433557748794556,
+      "learning_rate": 0.0002,
+      "loss": 2.5056,
+      "step": 5800
+    },
+    {
+      "epoch": 0.43293591654247393,
+      "grad_norm": 2.093193769454956,
+      "learning_rate": 0.0002,
+      "loss": 2.5444,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4336810730253353,
+      "grad_norm": 2.3394739627838135,
+      "learning_rate": 0.0002,
+      "loss": 2.4693,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4344262295081967,
+      "grad_norm": 2.294088363647461,
+      "learning_rate": 0.0002,
+      "loss": 2.2499,
+      "step": 5830
+    },
+    {
+      "epoch": 0.43517138599105815,
+      "grad_norm": 2.3843374252319336,
+      "learning_rate": 0.0002,
+      "loss": 2.5797,
+      "step": 5840
+    },
+    {
+      "epoch": 0.43591654247391953,
+      "grad_norm": 2.0269100666046143,
+      "learning_rate": 0.0002,
+      "loss": 2.5623,
+      "step": 5850
+    },
+    {
+      "epoch": 0.4366616989567809,
+      "grad_norm": 2.5061495304107666,
+      "learning_rate": 0.0002,
+      "loss": 2.3113,
+      "step": 5860
+    },
+    {
+      "epoch": 0.4374068554396423,
+      "grad_norm": 2.5287938117980957,
+      "learning_rate": 0.0002,
+      "loss": 2.3915,
+      "step": 5870
+    },
+    {
+      "epoch": 0.43815201192250375,
+      "grad_norm": 2.0537567138671875,
+      "learning_rate": 0.0002,
+      "loss": 2.6012,
+      "step": 5880
+    },
+    {
+      "epoch": 0.43889716840536513,
+      "grad_norm": 2.5927767753601074,
+      "learning_rate": 0.0002,
+      "loss": 2.4097,
+      "step": 5890
+    },
+    {
+      "epoch": 0.4396423248882265,
+      "grad_norm": 2.193775177001953,
+      "learning_rate": 0.0002,
+      "loss": 2.4681,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4403874813710879,
+      "grad_norm": 1.99056077003479,
+      "learning_rate": 0.0002,
+      "loss": 2.4156,
+      "step": 5910
+    },
+    {
+      "epoch": 0.44113263785394935,
+      "grad_norm": 2.4100918769836426,
+      "learning_rate": 0.0002,
+      "loss": 2.5194,
+      "step": 5920
+    },
+    {
+      "epoch": 0.44187779433681074,
+      "grad_norm": 2.178215980529785,
+      "learning_rate": 0.0002,
+      "loss": 2.4394,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4426229508196721,
+      "grad_norm": 2.3645856380462646,
+      "learning_rate": 0.0002,
+      "loss": 2.4488,
+      "step": 5940
+    },
+    {
+      "epoch": 0.4433681073025335,
+      "grad_norm": 1.8661149740219116,
+      "learning_rate": 0.0002,
+      "loss": 2.4434,
+      "step": 5950
+    },
+    {
+      "epoch": 0.44411326378539495,
+      "grad_norm": 1.9467486143112183,
+      "learning_rate": 0.0002,
+      "loss": 2.462,
+      "step": 5960
+    },
+    {
+      "epoch": 0.44485842026825634,
+      "grad_norm": 2.4336559772491455,
+      "learning_rate": 0.0002,
+      "loss": 2.6196,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4456035767511177,
+      "grad_norm": 2.2935919761657715,
+      "learning_rate": 0.0002,
+      "loss": 2.6004,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4463487332339791,
+      "grad_norm": 2.2421000003814697,
+      "learning_rate": 0.0002,
+      "loss": 2.4005,
+      "step": 5990
+    },
+    {
+      "epoch": 0.44709388971684055,
+      "grad_norm": 1.7802613973617554,
+      "learning_rate": 0.0002,
+      "loss": 2.4354,
+      "step": 6000
+    },
+    {
+      "epoch": 0.44783904619970194,
+      "grad_norm": 2.0072882175445557,
+      "learning_rate": 0.0002,
+      "loss": 2.4345,
+      "step": 6010
+    },
+    {
+      "epoch": 0.4485842026825633,
+      "grad_norm": 1.9635995626449585,
+      "learning_rate": 0.0002,
+      "loss": 2.5976,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4493293591654247,
+      "grad_norm": 2.6438255310058594,
+      "learning_rate": 0.0002,
+      "loss": 2.4887,
+      "step": 6030
+    },
+    {
+      "epoch": 0.45007451564828616,
+      "grad_norm": 2.0173120498657227,
+      "learning_rate": 0.0002,
+      "loss": 2.6285,
+      "step": 6040
+    },
+    {
+      "epoch": 0.45081967213114754,
+      "grad_norm": 2.0639147758483887,
+      "learning_rate": 0.0002,
+      "loss": 2.6229,
+      "step": 6050
+    },
+    {
+      "epoch": 0.45156482861400893,
+      "grad_norm": 2.2229113578796387,
+      "learning_rate": 0.0002,
+      "loss": 2.3337,
+      "step": 6060
+    },
+    {
+      "epoch": 0.4523099850968703,
+      "grad_norm": 2.0675735473632812,
+      "learning_rate": 0.0002,
+      "loss": 2.3311,
+      "step": 6070
+    },
+    {
+      "epoch": 0.45305514157973176,
+      "grad_norm": 2.0282604694366455,
+      "learning_rate": 0.0002,
+      "loss": 2.5878,
+      "step": 6080
+    },
+    {
+      "epoch": 0.45380029806259314,
+      "grad_norm": 2.4298789501190186,
+      "learning_rate": 0.0002,
+      "loss": 2.5402,
+      "step": 6090
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 2.1782422065734863,
+      "learning_rate": 0.0002,
+      "loss": 2.3497,
+      "step": 6100
+    },
+    {
+      "epoch": 0.455290611028316,
+      "grad_norm": 2.3076884746551514,
+      "learning_rate": 0.0002,
+      "loss": 2.4455,
+      "step": 6110
+    },
+    {
+      "epoch": 0.45603576751117736,
+      "grad_norm": 2.051884651184082,
+      "learning_rate": 0.0002,
+      "loss": 2.4875,
+      "step": 6120
+    },
+    {
+      "epoch": 0.45678092399403875,
+      "grad_norm": 2.0429365634918213,
+      "learning_rate": 0.0002,
+      "loss": 2.5699,
+      "step": 6130
+    },
+    {
+      "epoch": 0.45752608047690013,
+      "grad_norm": 2.0632383823394775,
+      "learning_rate": 0.0002,
+      "loss": 2.5193,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4582712369597616,
+      "grad_norm": 2.0191586017608643,
+      "learning_rate": 0.0002,
+      "loss": 2.4983,
+      "step": 6150
+    },
+    {
+      "epoch": 0.45901639344262296,
+      "grad_norm": 2.1893763542175293,
+      "learning_rate": 0.0002,
+      "loss": 2.4648,
+      "step": 6160
+    },
+    {
+      "epoch": 0.45976154992548435,
+      "grad_norm": 2.150398015975952,
+      "learning_rate": 0.0002,
+      "loss": 2.6308,
+      "step": 6170
+    },
+    {
+      "epoch": 0.46050670640834573,
+      "grad_norm": 2.084906578063965,
+      "learning_rate": 0.0002,
+      "loss": 2.5127,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4612518628912072,
+      "grad_norm": 2.3456621170043945,
+      "learning_rate": 0.0002,
+      "loss": 2.5092,
+      "step": 6190
+    },
+    {
+      "epoch": 0.46199701937406856,
+      "grad_norm": 2.259631633758545,
+      "learning_rate": 0.0002,
+      "loss": 2.4077,
+      "step": 6200
+    },
+    {
+      "epoch": 0.46274217585692995,
+      "grad_norm": 1.9756247997283936,
+      "learning_rate": 0.0002,
+      "loss": 2.3623,
+      "step": 6210
+    },
+    {
+      "epoch": 0.46348733233979134,
+      "grad_norm": 1.8990384340286255,
+      "learning_rate": 0.0002,
+      "loss": 2.5315,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4642324888226528,
+      "grad_norm": 2.1848907470703125,
+      "learning_rate": 0.0002,
+      "loss": 2.5418,
+      "step": 6230
+    },
+    {
+      "epoch": 0.46497764530551416,
+      "grad_norm": 2.267343521118164,
+      "learning_rate": 0.0002,
+      "loss": 2.4859,
+      "step": 6240
+    },
+    {
+      "epoch": 0.46572280178837555,
+      "grad_norm": 2.3617684841156006,
+      "learning_rate": 0.0002,
+      "loss": 2.3138,
+      "step": 6250
+    },
+    {
+      "epoch": 0.46646795827123694,
+      "grad_norm": 1.972421646118164,
+      "learning_rate": 0.0002,
+      "loss": 2.5332,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4672131147540984,
+      "grad_norm": 2.0907230377197266,
+      "learning_rate": 0.0002,
+      "loss": 2.7235,
+      "step": 6270
+    },
+    {
+      "epoch": 0.46795827123695977,
+      "grad_norm": 2.452946901321411,
+      "learning_rate": 0.0002,
+      "loss": 2.4452,
+      "step": 6280
+    },
+    {
+      "epoch": 0.46870342771982115,
+      "grad_norm": 2.432163953781128,
+      "learning_rate": 0.0002,
+      "loss": 2.4138,
+      "step": 6290
+    },
+    {
+      "epoch": 0.46944858420268254,
+      "grad_norm": 2.2494499683380127,
+      "learning_rate": 0.0002,
+      "loss": 2.4877,
+      "step": 6300
+    },
+    {
+      "epoch": 0.470193740685544,
+      "grad_norm": 1.6413583755493164,
+      "learning_rate": 0.0002,
+      "loss": 2.5623,
+      "step": 6310
+    },
+    {
+      "epoch": 0.47093889716840537,
+      "grad_norm": 2.4607837200164795,
+      "learning_rate": 0.0002,
+      "loss": 2.4455,
+      "step": 6320
+    },
+    {
+      "epoch": 0.47168405365126675,
+      "grad_norm": 2.0634145736694336,
+      "learning_rate": 0.0002,
+      "loss": 2.5643,
+      "step": 6330
+    },
+    {
+      "epoch": 0.47242921013412814,
+      "grad_norm": 2.0293703079223633,
+      "learning_rate": 0.0002,
+      "loss": 2.4346,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4731743666169896,
+      "grad_norm": 2.567640781402588,
+      "learning_rate": 0.0002,
+      "loss": 2.2669,
+      "step": 6350
+    },
+    {
+      "epoch": 0.47391952309985097,
+      "grad_norm": 2.118058204650879,
+      "learning_rate": 0.0002,
+      "loss": 2.5486,
+      "step": 6360
+    },
+    {
+      "epoch": 0.47466467958271236,
+      "grad_norm": 2.1259288787841797,
+      "learning_rate": 0.0002,
+      "loss": 2.5674,
+      "step": 6370
+    },
+    {
+      "epoch": 0.47540983606557374,
+      "grad_norm": 2.1126372814178467,
+      "learning_rate": 0.0002,
+      "loss": 2.6126,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4761549925484352,
+      "grad_norm": 1.626694917678833,
+      "learning_rate": 0.0002,
+      "loss": 2.5019,
+      "step": 6390
+    },
+    {
+      "epoch": 0.47690014903129657,
+      "grad_norm": 2.0831708908081055,
+      "learning_rate": 0.0002,
+      "loss": 2.4327,
+      "step": 6400
+    },
+    {
+      "epoch": 0.47764530551415796,
+      "grad_norm": 1.9350471496582031,
+      "learning_rate": 0.0002,
+      "loss": 2.4867,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4783904619970194,
+      "grad_norm": 2.1217947006225586,
+      "learning_rate": 0.0002,
+      "loss": 2.5254,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4791356184798808,
+      "grad_norm": 2.3201823234558105,
+      "learning_rate": 0.0002,
+      "loss": 2.4434,
+      "step": 6430
+    },
+    {
+      "epoch": 0.4798807749627422,
+      "grad_norm": 2.2600150108337402,
+      "learning_rate": 0.0002,
+      "loss": 2.4616,
+      "step": 6440
+    },
+    {
+      "epoch": 0.48062593144560356,
+      "grad_norm": 2.360180377960205,
+      "learning_rate": 0.0002,
+      "loss": 2.4934,
+      "step": 6450
+    },
+    {
+      "epoch": 0.481371087928465,
+      "grad_norm": 2.4700534343719482,
+      "learning_rate": 0.0002,
+      "loss": 2.3958,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4821162444113264,
+      "grad_norm": 2.0691604614257812,
+      "learning_rate": 0.0002,
+      "loss": 2.4818,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4828614008941878,
+      "grad_norm": 1.950579285621643,
+      "learning_rate": 0.0002,
+      "loss": 2.4003,
+      "step": 6480
+    },
+    {
+      "epoch": 0.48360655737704916,
+      "grad_norm": 2.0040478706359863,
+      "learning_rate": 0.0002,
+      "loss": 2.5706,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4843517138599106,
+      "grad_norm": 2.4494431018829346,
+      "learning_rate": 0.0002,
+      "loss": 2.4676,
+      "step": 6500
+    },
+    {
+      "epoch": 0.485096870342772,
+      "grad_norm": 2.25048565864563,
+      "learning_rate": 0.0002,
+      "loss": 2.6437,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4858420268256334,
+      "grad_norm": 2.4388575553894043,
+      "learning_rate": 0.0002,
+      "loss": 2.4013,
+      "step": 6520
+    },
+    {
+      "epoch": 0.48658718330849476,
+      "grad_norm": 2.7190568447113037,
+      "learning_rate": 0.0002,
+      "loss": 2.1267,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4873323397913562,
+      "grad_norm": 2.00464129447937,
+      "learning_rate": 0.0002,
+      "loss": 2.246,
+      "step": 6540
+    },
+    {
+      "epoch": 0.4880774962742176,
+      "grad_norm": 2.230637550354004,
+      "learning_rate": 0.0002,
+      "loss": 2.3657,
+      "step": 6550
+    },
+    {
+      "epoch": 0.488822652757079,
+      "grad_norm": 2.030550479888916,
+      "learning_rate": 0.0002,
+      "loss": 2.5825,
+      "step": 6560
+    },
+    {
+      "epoch": 0.48956780923994037,
+      "grad_norm": 1.7973986864089966,
+      "learning_rate": 0.0002,
+      "loss": 2.2891,
+      "step": 6570
+    },
+    {
+      "epoch": 0.4903129657228018,
+      "grad_norm": 2.1454880237579346,
+      "learning_rate": 0.0002,
+      "loss": 2.3413,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4910581222056632,
+      "grad_norm": 2.1549787521362305,
+      "learning_rate": 0.0002,
+      "loss": 2.3444,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 2.1284945011138916,
+      "learning_rate": 0.0002,
+      "loss": 2.5555,
+      "step": 6600
+    },
+    {
+      "epoch": 0.49254843517138597,
+      "grad_norm": 2.23075008392334,
+      "learning_rate": 0.0002,
+      "loss": 2.3384,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4932935916542474,
+      "grad_norm": 2.489455223083496,
+      "learning_rate": 0.0002,
+      "loss": 2.6383,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4940387481371088,
+      "grad_norm": 2.2579898834228516,
+      "learning_rate": 0.0002,
+      "loss": 2.6226,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4947839046199702,
+      "grad_norm": 2.3494224548339844,
+      "learning_rate": 0.0002,
+      "loss": 2.502,
+      "step": 6640
+    },
+    {
+      "epoch": 0.49552906110283157,
+      "grad_norm": 2.082937240600586,
+      "learning_rate": 0.0002,
+      "loss": 2.5592,
+      "step": 6650
+    },
+    {
+      "epoch": 0.496274217585693,
+      "grad_norm": 2.0874314308166504,
+      "learning_rate": 0.0002,
+      "loss": 2.5922,
+      "step": 6660
+    },
+    {
+      "epoch": 0.4970193740685544,
+      "grad_norm": 2.1997947692871094,
+      "learning_rate": 0.0002,
+      "loss": 2.6147,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4977645305514158,
+      "grad_norm": 2.215691089630127,
+      "learning_rate": 0.0002,
+      "loss": 2.4961,
+      "step": 6680
+    },
+    {
+      "epoch": 0.49850968703427717,
+      "grad_norm": 2.699936866760254,
+      "learning_rate": 0.0002,
+      "loss": 2.6774,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4992548435171386,
+      "grad_norm": 2.2943921089172363,
+      "learning_rate": 0.0002,
+      "loss": 2.4801,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3712542057037354,
+      "learning_rate": 0.0002,
+      "loss": 2.4751,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5007451564828614,
+      "grad_norm": 2.5637362003326416,
+      "learning_rate": 0.0002,
+      "loss": 2.2542,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5014903129657228,
+      "grad_norm": 2.330156087875366,
+      "learning_rate": 0.0002,
+      "loss": 2.4645,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5022354694485842,
+      "grad_norm": 2.053035020828247,
+      "learning_rate": 0.0002,
+      "loss": 2.5752,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5029806259314457,
+      "grad_norm": 2.305776834487915,
+      "learning_rate": 0.0002,
+      "loss": 2.5567,
+      "step": 6750
+    },
+    {
+      "epoch": 0.503725782414307,
+      "grad_norm": 2.023801326751709,
+      "learning_rate": 0.0002,
+      "loss": 2.3164,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5044709388971684,
+      "grad_norm": 2.477642059326172,
+      "learning_rate": 0.0002,
+      "loss": 2.5777,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5052160953800298,
+      "grad_norm": 2.298116683959961,
+      "learning_rate": 0.0002,
+      "loss": 2.4506,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5059612518628912,
+      "grad_norm": 2.2904672622680664,
+      "learning_rate": 0.0002,
+      "loss": 2.3897,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5067064083457526,
+      "grad_norm": 2.0368447303771973,
+      "learning_rate": 0.0002,
+      "loss": 2.42,
+      "step": 6800
+    },
+    {
+      "epoch": 0.507451564828614,
+      "grad_norm": 2.016451835632324,
+      "learning_rate": 0.0002,
+      "loss": 2.3806,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5081967213114754,
+      "grad_norm": 1.8767539262771606,
+      "learning_rate": 0.0002,
+      "loss": 2.3789,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5089418777943369,
+      "grad_norm": 2.399251937866211,
+      "learning_rate": 0.0002,
+      "loss": 2.5804,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5096870342771982,
+      "grad_norm": 2.187103509902954,
+      "learning_rate": 0.0002,
+      "loss": 2.5263,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5104321907600596,
+      "grad_norm": 1.9529699087142944,
+      "learning_rate": 0.0002,
+      "loss": 2.423,
+      "step": 6850
+    },
+    {
+      "epoch": 0.511177347242921,
+      "grad_norm": 4.577221393585205,
+      "learning_rate": 0.0002,
+      "loss": 2.5348,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5119225037257824,
+      "grad_norm": 2.331979513168335,
+      "learning_rate": 0.0002,
+      "loss": 2.6584,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5126676602086438,
+      "grad_norm": 2.2853405475616455,
+      "learning_rate": 0.0002,
+      "loss": 2.4096,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5134128166915052,
+      "grad_norm": 2.516995906829834,
+      "learning_rate": 0.0002,
+      "loss": 2.4762,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5141579731743666,
+      "grad_norm": 2.183138370513916,
+      "learning_rate": 0.0002,
+      "loss": 2.4239,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5149031296572281,
+      "grad_norm": 2.2524988651275635,
+      "learning_rate": 0.0002,
+      "loss": 2.4089,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5156482861400894,
+      "grad_norm": 2.2979516983032227,
+      "learning_rate": 0.0002,
+      "loss": 2.443,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5163934426229508,
+      "grad_norm": 2.099586009979248,
+      "learning_rate": 0.0002,
+      "loss": 2.4679,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5171385991058122,
+      "grad_norm": 2.2441070079803467,
+      "learning_rate": 0.0002,
+      "loss": 2.5278,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5178837555886736,
+      "grad_norm": 2.376936197280884,
+      "learning_rate": 0.0002,
+      "loss": 2.5222,
+      "step": 6950
+    },
+    {
+      "epoch": 0.518628912071535,
+      "grad_norm": 2.196542739868164,
+      "learning_rate": 0.0002,
+      "loss": 2.4336,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5193740685543964,
+      "grad_norm": 2.265864610671997,
+      "learning_rate": 0.0002,
+      "loss": 2.4876,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5201192250372578,
+      "grad_norm": 2.2958121299743652,
+      "learning_rate": 0.0002,
+      "loss": 2.6756,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5208643815201193,
+      "grad_norm": 2.2851719856262207,
+      "learning_rate": 0.0002,
+      "loss": 2.4873,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5216095380029806,
+      "grad_norm": 2.3405418395996094,
+      "learning_rate": 0.0002,
+      "loss": 2.3907,
+      "step": 7000
+    },
+    {
+      "epoch": 0.522354694485842,
+      "grad_norm": 2.4564900398254395,
+      "learning_rate": 0.0002,
+      "loss": 2.5203,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5230998509687034,
+      "grad_norm": 2.123331308364868,
+      "learning_rate": 0.0002,
+      "loss": 2.3261,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5238450074515648,
+      "grad_norm": 2.0890355110168457,
+      "learning_rate": 0.0002,
+      "loss": 2.5784,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 2.4165306091308594,
+      "learning_rate": 0.0002,
+      "loss": 2.4716,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5253353204172876,
+      "grad_norm": 2.1103525161743164,
+      "learning_rate": 0.0002,
+      "loss": 2.5825,
+      "step": 7050
+    },
+    {
+      "epoch": 0.526080476900149,
+      "grad_norm": 2.33457612991333,
+      "learning_rate": 0.0002,
+      "loss": 2.3823,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5268256333830105,
+      "grad_norm": 2.0597524642944336,
+      "learning_rate": 0.0002,
+      "loss": 2.4671,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5275707898658718,
+      "grad_norm": 3.1226096153259277,
+      "learning_rate": 0.0002,
+      "loss": 2.4017,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5283159463487332,
+      "grad_norm": 2.0051512718200684,
+      "learning_rate": 0.0002,
+      "loss": 2.359,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5290611028315947,
+      "grad_norm": 2.500908136367798,
+      "learning_rate": 0.0002,
+      "loss": 2.6868,
+      "step": 7100
+    },
+    {
+      "epoch": 0.529806259314456,
+      "grad_norm": 2.04001784324646,
+      "learning_rate": 0.0002,
+      "loss": 2.717,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5305514157973175,
+      "grad_norm": 2.715292453765869,
+      "learning_rate": 0.0002,
+      "loss": 2.5725,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5312965722801788,
+      "grad_norm": 2.3398818969726562,
+      "learning_rate": 0.0002,
+      "loss": 2.4834,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5320417287630402,
+      "grad_norm": 2.456146240234375,
+      "learning_rate": 0.0002,
+      "loss": 2.4792,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5327868852459017,
+      "grad_norm": 2.2321231365203857,
+      "learning_rate": 0.0002,
+      "loss": 2.6432,
+      "step": 7150
+    },
+    {
+      "epoch": 0.533532041728763,
+      "grad_norm": 2.06449294090271,
+      "learning_rate": 0.0002,
+      "loss": 2.3722,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5342771982116244,
+      "grad_norm": 1.98611581325531,
+      "learning_rate": 0.0002,
+      "loss": 2.2874,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5350223546944859,
+      "grad_norm": 2.2005727291107178,
+      "learning_rate": 0.0002,
+      "loss": 2.3682,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5357675111773472,
+      "grad_norm": 2.3024485111236572,
+      "learning_rate": 0.0002,
+      "loss": 2.5261,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5365126676602087,
+      "grad_norm": 2.2706873416900635,
+      "learning_rate": 0.0002,
+      "loss": 2.5962,
+      "step": 7200
+    },
+    {
+      "epoch": 0.53725782414307,
+      "grad_norm": 1.8197662830352783,
+      "learning_rate": 0.0002,
+      "loss": 2.5705,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5380029806259314,
+      "grad_norm": 1.9607528448104858,
+      "learning_rate": 0.0002,
+      "loss": 2.3861,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5387481371087929,
+      "grad_norm": 2.5178678035736084,
+      "learning_rate": 0.0002,
+      "loss": 2.5357,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5394932935916542,
+      "grad_norm": 2.2647557258605957,
+      "learning_rate": 0.0002,
+      "loss": 2.3796,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5402384500745157,
+      "grad_norm": 2.090864419937134,
+      "learning_rate": 0.0002,
+      "loss": 2.5204,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5409836065573771,
+      "grad_norm": 2.0715291500091553,
+      "learning_rate": 0.0002,
+      "loss": 2.5327,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5417287630402384,
+      "grad_norm": 1.7275205850601196,
+      "learning_rate": 0.0002,
+      "loss": 2.2495,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5424739195230999,
+      "grad_norm": 2.1511425971984863,
+      "learning_rate": 0.0002,
+      "loss": 2.4449,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5432190760059612,
+      "grad_norm": 2.3975725173950195,
+      "learning_rate": 0.0002,
+      "loss": 2.6171,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5439642324888226,
+      "grad_norm": 2.278902769088745,
+      "learning_rate": 0.0002,
+      "loss": 2.4238,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5447093889716841,
+      "grad_norm": 2.1751863956451416,
+      "learning_rate": 0.0002,
+      "loss": 2.3223,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 2.1303343772888184,
+      "learning_rate": 0.0002,
+      "loss": 2.292,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5461997019374069,
+      "grad_norm": 2.0914053916931152,
+      "learning_rate": 0.0002,
+      "loss": 2.6599,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5469448584202683,
+      "grad_norm": 2.3275091648101807,
+      "learning_rate": 0.0002,
+      "loss": 2.4827,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5476900149031296,
+      "grad_norm": 2.047351598739624,
+      "learning_rate": 0.0002,
+      "loss": 2.4824,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5484351713859911,
+      "grad_norm": 2.2209582328796387,
+      "learning_rate": 0.0002,
+      "loss": 2.5068,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5491803278688525,
+      "grad_norm": 2.029001235961914,
+      "learning_rate": 0.0002,
+      "loss": 2.3278,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5499254843517138,
+      "grad_norm": 2.0651822090148926,
+      "learning_rate": 0.0002,
+      "loss": 2.4738,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5506706408345753,
+      "grad_norm": 1.8926769495010376,
+      "learning_rate": 0.0002,
+      "loss": 2.312,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5514157973174366,
+      "grad_norm": 2.1285948753356934,
+      "learning_rate": 0.0002,
+      "loss": 2.515,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5521609538002981,
+      "grad_norm": 2.026381254196167,
+      "learning_rate": 0.0002,
+      "loss": 2.6837,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5529061102831595,
+      "grad_norm": 2.052429437637329,
+      "learning_rate": 0.0002,
+      "loss": 2.3549,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5536512667660208,
+      "grad_norm": 2.634350538253784,
+      "learning_rate": 0.0002,
+      "loss": 2.4303,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5543964232488823,
+      "grad_norm": 2.1491518020629883,
+      "learning_rate": 0.0002,
+      "loss": 2.5716,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5551415797317437,
+      "grad_norm": 1.9551408290863037,
+      "learning_rate": 0.0002,
+      "loss": 2.4071,
+      "step": 7450
+    },
+    {
+      "epoch": 0.555886736214605,
+      "grad_norm": 2.2387406826019287,
+      "learning_rate": 0.0002,
+      "loss": 2.383,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5566318926974665,
+      "grad_norm": 2.1040196418762207,
+      "learning_rate": 0.0002,
+      "loss": 2.3814,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5573770491803278,
+      "grad_norm": 2.3352060317993164,
+      "learning_rate": 0.0002,
+      "loss": 2.5804,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5581222056631893,
+      "grad_norm": 2.1420586109161377,
+      "learning_rate": 0.0002,
+      "loss": 2.3543,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5588673621460507,
+      "grad_norm": 2.3646399974823,
+      "learning_rate": 0.0002,
+      "loss": 2.3857,
+      "step": 7500
+    },
+    {
+      "epoch": 0.559612518628912,
+      "grad_norm": 1.9728518724441528,
+      "learning_rate": 0.0002,
+      "loss": 2.5749,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5603576751117735,
+      "grad_norm": 2.2200262546539307,
+      "learning_rate": 0.0002,
+      "loss": 2.5357,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5611028315946349,
+      "grad_norm": 2.305957078933716,
+      "learning_rate": 0.0002,
+      "loss": 2.4489,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5618479880774963,
+      "grad_norm": 2.1752254962921143,
+      "learning_rate": 0.0002,
+      "loss": 2.2947,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5625931445603577,
+      "grad_norm": 2.1745660305023193,
+      "learning_rate": 0.0002,
+      "loss": 2.5143,
+      "step": 7550
+    },
+    {
+      "epoch": 0.563338301043219,
+      "grad_norm": 1.8101173639297485,
+      "learning_rate": 0.0002,
+      "loss": 2.5449,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5640834575260805,
+      "grad_norm": 2.5516979694366455,
+      "learning_rate": 0.0002,
+      "loss": 2.419,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5648286140089419,
+      "grad_norm": 2.087670087814331,
+      "learning_rate": 0.0002,
+      "loss": 2.5587,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5655737704918032,
+      "grad_norm": 2.1860992908477783,
+      "learning_rate": 0.0002,
+      "loss": 2.6475,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5663189269746647,
+      "grad_norm": 2.522256851196289,
+      "learning_rate": 0.0002,
+      "loss": 2.7054,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5670640834575261,
+      "grad_norm": 2.6697170734405518,
+      "learning_rate": 0.0002,
+      "loss": 2.5817,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5678092399403875,
+      "grad_norm": 2.129748821258545,
+      "learning_rate": 0.0002,
+      "loss": 2.4949,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5685543964232489,
+      "grad_norm": 1.946333646774292,
+      "learning_rate": 0.0002,
+      "loss": 2.4987,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5692995529061102,
+      "grad_norm": 2.6684484481811523,
+      "learning_rate": 0.0002,
+      "loss": 2.612,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5700447093889717,
+      "grad_norm": 2.1237940788269043,
+      "learning_rate": 0.0002,
+      "loss": 2.5082,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5707898658718331,
+      "grad_norm": 2.68740177154541,
+      "learning_rate": 0.0002,
+      "loss": 2.5652,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5715350223546944,
+      "grad_norm": 2.360792875289917,
+      "learning_rate": 0.0002,
+      "loss": 2.4631,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5722801788375559,
+      "grad_norm": 2.3592369556427,
+      "learning_rate": 0.0002,
+      "loss": 2.4923,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5730253353204173,
+      "grad_norm": 2.328521251678467,
+      "learning_rate": 0.0002,
+      "loss": 2.4656,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5737704918032787,
+      "grad_norm": 2.1476731300354004,
+      "learning_rate": 0.0002,
+      "loss": 2.3461,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5745156482861401,
+      "grad_norm": 2.1318559646606445,
+      "learning_rate": 0.0002,
+      "loss": 2.4251,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5752608047690015,
+      "grad_norm": 1.997536301612854,
+      "learning_rate": 0.0002,
+      "loss": 2.5326,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5760059612518629,
+      "grad_norm": 2.2597386837005615,
+      "learning_rate": 0.0002,
+      "loss": 2.6773,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5767511177347243,
+      "grad_norm": 2.8514564037323,
+      "learning_rate": 0.0002,
+      "loss": 2.5119,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5774962742175856,
+      "grad_norm": 2.486799716949463,
+      "learning_rate": 0.0002,
+      "loss": 2.6904,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5782414307004471,
+      "grad_norm": 2.237799882888794,
+      "learning_rate": 0.0002,
+      "loss": 2.5005,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5789865871833085,
+      "grad_norm": 2.4719021320343018,
+      "learning_rate": 0.0002,
+      "loss": 2.5874,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5797317436661699,
+      "grad_norm": 2.2470688819885254,
+      "learning_rate": 0.0002,
+      "loss": 2.5151,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5804769001490313,
+      "grad_norm": 2.4005558490753174,
+      "learning_rate": 0.0002,
+      "loss": 2.6334,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5812220566318927,
+      "grad_norm": 2.0954015254974365,
+      "learning_rate": 0.0002,
+      "loss": 2.3567,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5819672131147541,
+      "grad_norm": 2.228788375854492,
+      "learning_rate": 0.0002,
+      "loss": 2.6451,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5827123695976155,
+      "grad_norm": 1.78871488571167,
+      "learning_rate": 0.0002,
+      "loss": 2.6242,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5834575260804769,
+      "grad_norm": 1.7899997234344482,
+      "learning_rate": 0.0002,
+      "loss": 2.4975,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5842026825633383,
+      "grad_norm": 2.1144442558288574,
+      "learning_rate": 0.0002,
+      "loss": 2.6146,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5849478390461997,
+      "grad_norm": 1.8392325639724731,
+      "learning_rate": 0.0002,
+      "loss": 2.4227,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5856929955290611,
+      "grad_norm": 1.8613855838775635,
+      "learning_rate": 0.0002,
+      "loss": 2.5583,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5864381520119225,
+      "grad_norm": 2.1754300594329834,
+      "learning_rate": 0.0002,
+      "loss": 2.6082,
+      "step": 7870
+    },
+    {
+      "epoch": 0.587183308494784,
+      "grad_norm": 2.4607224464416504,
+      "learning_rate": 0.0002,
+      "loss": 2.3939,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5879284649776453,
+      "grad_norm": 2.129397392272949,
+      "learning_rate": 0.0002,
+      "loss": 2.5056,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5886736214605067,
+      "grad_norm": 2.302616596221924,
+      "learning_rate": 0.0002,
+      "loss": 2.4976,
+      "step": 7900
+    },
+    {
+      "epoch": 0.589418777943368,
+      "grad_norm": 3.0152175426483154,
+      "learning_rate": 0.0002,
+      "loss": 2.3935,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5901639344262295,
+      "grad_norm": 2.051461696624756,
+      "learning_rate": 0.0002,
+      "loss": 2.4928,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5909090909090909,
+      "grad_norm": 2.1009974479675293,
+      "learning_rate": 0.0002,
+      "loss": 2.6526,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5916542473919523,
+      "grad_norm": 2.290898561477661,
+      "learning_rate": 0.0002,
+      "loss": 2.5205,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5923994038748137,
+      "grad_norm": 2.202995777130127,
+      "learning_rate": 0.0002,
+      "loss": 2.5271,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5931445603576752,
+      "grad_norm": 2.211921453475952,
+      "learning_rate": 0.0002,
+      "loss": 2.6762,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5938897168405365,
+      "grad_norm": 2.294769763946533,
+      "learning_rate": 0.0002,
+      "loss": 2.5565,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5946348733233979,
+      "grad_norm": 2.0982816219329834,
+      "learning_rate": 0.0002,
+      "loss": 2.4512,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5953800298062594,
+      "grad_norm": 2.14776611328125,
+      "learning_rate": 0.0002,
+      "loss": 2.5092,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5961251862891207,
+      "grad_norm": 2.2703404426574707,
+      "learning_rate": 0.0002,
+      "loss": 2.3723,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5968703427719821,
+      "grad_norm": 2.265050172805786,
+      "learning_rate": 0.0002,
+      "loss": 2.4642,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5976154992548435,
+      "grad_norm": 2.3132333755493164,
+      "learning_rate": 0.0002,
+      "loss": 2.5886,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5983606557377049,
+      "grad_norm": 2.4410815238952637,
+      "learning_rate": 0.0002,
+      "loss": 2.3898,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5991058122205664,
+      "grad_norm": 2.282869577407837,
+      "learning_rate": 0.0002,
+      "loss": 2.3925,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5998509687034277,
+      "grad_norm": 2.3430824279785156,
+      "learning_rate": 0.0002,
+      "loss": 2.5852,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6005961251862891,
+      "grad_norm": 2.7821292877197266,
+      "learning_rate": 0.0002,
+      "loss": 2.1127,
+      "step": 8060
+    },
+    {
+      "epoch": 0.6013412816691506,
+      "grad_norm": 2.4642081260681152,
+      "learning_rate": 0.0002,
+      "loss": 2.5766,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6020864381520119,
+      "grad_norm": 2.013272285461426,
+      "learning_rate": 0.0002,
+      "loss": 2.4526,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6028315946348733,
+      "grad_norm": 2.0950276851654053,
+      "learning_rate": 0.0002,
+      "loss": 2.672,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6035767511177347,
+      "grad_norm": 2.2408697605133057,
+      "learning_rate": 0.0002,
+      "loss": 2.4886,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6043219076005961,
+      "grad_norm": 2.4338343143463135,
+      "learning_rate": 0.0002,
+      "loss": 2.4129,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6050670640834576,
+      "grad_norm": 2.3819990158081055,
+      "learning_rate": 0.0002,
+      "loss": 2.7111,
+      "step": 8120
+    },
+    {
+      "epoch": 0.6058122205663189,
+      "grad_norm": 2.3578953742980957,
+      "learning_rate": 0.0002,
+      "loss": 2.5222,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6065573770491803,
+      "grad_norm": 2.0468990802764893,
+      "learning_rate": 0.0002,
+      "loss": 2.4965,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6073025335320418,
+      "grad_norm": 2.439807415008545,
+      "learning_rate": 0.0002,
+      "loss": 2.6172,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6080476900149031,
+      "grad_norm": 2.1083173751831055,
+      "learning_rate": 0.0002,
+      "loss": 2.5669,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6087928464977646,
+      "grad_norm": 2.1767308712005615,
+      "learning_rate": 0.0002,
+      "loss": 2.5498,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6095380029806259,
+      "grad_norm": 2.1427078247070312,
+      "learning_rate": 0.0002,
+      "loss": 2.4109,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6102831594634873,
+      "grad_norm": 2.282959222793579,
+      "learning_rate": 0.0002,
+      "loss": 2.5091,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6110283159463488,
+      "grad_norm": 2.2628536224365234,
+      "learning_rate": 0.0002,
+      "loss": 2.5386,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6117734724292101,
+      "grad_norm": 2.706434488296509,
+      "learning_rate": 0.0002,
+      "loss": 2.5989,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6125186289120715,
+      "grad_norm": 2.3741445541381836,
+      "learning_rate": 0.0002,
+      "loss": 2.2712,
+      "step": 8220
+    },
+    {
+      "epoch": 0.613263785394933,
+      "grad_norm": 2.2221875190734863,
+      "learning_rate": 0.0002,
+      "loss": 2.5309,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6140089418777943,
+      "grad_norm": 1.9854212999343872,
+      "learning_rate": 0.0002,
+      "loss": 2.2749,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6147540983606558,
+      "grad_norm": 2.229374885559082,
+      "learning_rate": 0.0002,
+      "loss": 2.5129,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6154992548435171,
+      "grad_norm": 2.4126970767974854,
+      "learning_rate": 0.0002,
+      "loss": 2.325,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6162444113263785,
+      "grad_norm": 2.2903852462768555,
+      "learning_rate": 0.0002,
+      "loss": 2.5048,
+      "step": 8270
+    },
+    {
+      "epoch": 0.61698956780924,
+      "grad_norm": 2.691183090209961,
+      "learning_rate": 0.0002,
+      "loss": 2.4766,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6177347242921013,
+      "grad_norm": 2.386356830596924,
+      "learning_rate": 0.0002,
+      "loss": 2.5979,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6184798807749627,
+      "grad_norm": 2.695887804031372,
+      "learning_rate": 0.0002,
+      "loss": 2.642,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6192250372578242,
+      "grad_norm": 2.5322632789611816,
+      "learning_rate": 0.0002,
+      "loss": 2.5457,
+      "step": 8310
+    },
+    {
+      "epoch": 0.6199701937406855,
+      "grad_norm": 2.29015851020813,
+      "learning_rate": 0.0002,
+      "loss": 2.5388,
+      "step": 8320
+    },
+    {
+      "epoch": 0.620715350223547,
+      "grad_norm": 2.326113700866699,
+      "learning_rate": 0.0002,
+      "loss": 2.6757,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6214605067064084,
+      "grad_norm": 2.18438458442688,
+      "learning_rate": 0.0002,
+      "loss": 2.3377,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6222056631892697,
+      "grad_norm": 2.195730686187744,
+      "learning_rate": 0.0002,
+      "loss": 2.334,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6229508196721312,
+      "grad_norm": 2.1896743774414062,
+      "learning_rate": 0.0002,
+      "loss": 2.4039,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6236959761549925,
+      "grad_norm": 2.47771954536438,
+      "learning_rate": 0.0002,
+      "loss": 2.625,
+      "step": 8370
+    },
+    {
+      "epoch": 0.624441132637854,
+      "grad_norm": 2.5502564907073975,
+      "learning_rate": 0.0002,
+      "loss": 2.5164,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6251862891207154,
+      "grad_norm": 1.9382567405700684,
+      "learning_rate": 0.0002,
+      "loss": 2.5153,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6259314456035767,
+      "grad_norm": 2.078873872756958,
+      "learning_rate": 0.0002,
+      "loss": 2.4817,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6266766020864382,
+      "grad_norm": 2.1636760234832764,
+      "learning_rate": 0.0002,
+      "loss": 2.7134,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6274217585692996,
+      "grad_norm": 2.2006876468658447,
+      "learning_rate": 0.0002,
+      "loss": 2.6016,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6281669150521609,
+      "grad_norm": 2.364816665649414,
+      "learning_rate": 0.0002,
+      "loss": 2.5201,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6289120715350224,
+      "grad_norm": 2.450207471847534,
+      "learning_rate": 0.0002,
+      "loss": 2.6406,
+      "step": 8440
+    },
+    {
+      "epoch": 0.6296572280178837,
+      "grad_norm": 2.3795676231384277,
+      "learning_rate": 0.0002,
+      "loss": 2.2957,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6304023845007451,
+      "grad_norm": 2.1926169395446777,
+      "learning_rate": 0.0002,
+      "loss": 2.2288,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6311475409836066,
+      "grad_norm": 1.9396635293960571,
+      "learning_rate": 0.0002,
+      "loss": 2.3839,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6318926974664679,
+      "grad_norm": 2.6635711193084717,
+      "learning_rate": 0.0002,
+      "loss": 2.4389,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6326378539493294,
+      "grad_norm": 2.657240390777588,
+      "learning_rate": 0.0002,
+      "loss": 2.3886,
+      "step": 8490
+    },
+    {
+      "epoch": 0.6333830104321908,
+      "grad_norm": 2.050353765487671,
+      "learning_rate": 0.0002,
+      "loss": 2.3714,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6341281669150521,
+      "grad_norm": 2.3058016300201416,
+      "learning_rate": 0.0002,
+      "loss": 2.4301,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6348733233979136,
+      "grad_norm": 2.3272721767425537,
+      "learning_rate": 0.0002,
+      "loss": 2.442,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6356184798807749,
+      "grad_norm": 2.105719566345215,
+      "learning_rate": 0.0002,
+      "loss": 2.3532,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 2.2481689453125,
+      "learning_rate": 0.0002,
+      "loss": 2.7017,
+      "step": 8540
+    },
+    {
+      "epoch": 0.6371087928464978,
+      "grad_norm": 2.0684092044830322,
+      "learning_rate": 0.0002,
+      "loss": 2.5492,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6378539493293591,
+      "grad_norm": 2.2087674140930176,
+      "learning_rate": 0.0002,
+      "loss": 2.4946,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6385991058122206,
+      "grad_norm": 2.0686557292938232,
+      "learning_rate": 0.0002,
+      "loss": 2.603,
+      "step": 8570
+    },
+    {
+      "epoch": 0.639344262295082,
+      "grad_norm": 2.223733901977539,
+      "learning_rate": 0.0002,
+      "loss": 2.5515,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6400894187779433,
+      "grad_norm": 2.0543527603149414,
+      "learning_rate": 0.0002,
+      "loss": 2.2859,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6408345752608048,
+      "grad_norm": 2.119685411453247,
+      "learning_rate": 0.0002,
+      "loss": 2.4915,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6415797317436661,
+      "grad_norm": 2.1664891242980957,
+      "learning_rate": 0.0002,
+      "loss": 2.4224,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6423248882265276,
+      "grad_norm": 2.2479021549224854,
+      "learning_rate": 0.0002,
+      "loss": 2.5316,
+      "step": 8620
+    },
+    {
+      "epoch": 0.643070044709389,
+      "grad_norm": 2.2841110229492188,
+      "learning_rate": 0.0002,
+      "loss": 2.6606,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6438152011922503,
+      "grad_norm": 2.4399871826171875,
+      "learning_rate": 0.0002,
+      "loss": 2.5727,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6445603576751118,
+      "grad_norm": 1.9307136535644531,
+      "learning_rate": 0.0002,
+      "loss": 2.3302,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6453055141579732,
+      "grad_norm": 2.2575156688690186,
+      "learning_rate": 0.0002,
+      "loss": 2.5079,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6460506706408345,
+      "grad_norm": 2.704486131668091,
+      "learning_rate": 0.0002,
+      "loss": 2.64,
+      "step": 8670
+    },
+    {
+      "epoch": 0.646795827123696,
+      "grad_norm": 2.2060296535491943,
+      "learning_rate": 0.0002,
+      "loss": 2.5639,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6475409836065574,
+      "grad_norm": 2.2497682571411133,
+      "learning_rate": 0.0002,
+      "loss": 2.5753,
+      "step": 8690
+    },
+    {
+      "epoch": 0.6482861400894188,
+      "grad_norm": 2.1391713619232178,
+      "learning_rate": 0.0002,
+      "loss": 2.5179,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6490312965722802,
+      "grad_norm": 2.4493465423583984,
+      "learning_rate": 0.0002,
+      "loss": 2.5404,
+      "step": 8710
+    },
+    {
+      "epoch": 0.6497764530551415,
+      "grad_norm": 2.2963478565216064,
+      "learning_rate": 0.0002,
+      "loss": 2.395,
+      "step": 8720
+    },
+    {
+      "epoch": 0.650521609538003,
+      "grad_norm": 2.3371636867523193,
+      "learning_rate": 0.0002,
+      "loss": 2.649,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6512667660208644,
+      "grad_norm": 2.1336076259613037,
+      "learning_rate": 0.0002,
+      "loss": 2.5782,
+      "step": 8740
+    },
+    {
+      "epoch": 0.6520119225037257,
+      "grad_norm": 1.9927014112472534,
+      "learning_rate": 0.0002,
+      "loss": 2.4791,
+      "step": 8750
+    },
+    {
+      "epoch": 0.6527570789865872,
+      "grad_norm": 2.5760622024536133,
+      "learning_rate": 0.0002,
+      "loss": 2.5103,
+      "step": 8760
+    },
+    {
+      "epoch": 0.6535022354694486,
+      "grad_norm": 2.3019092082977295,
+      "learning_rate": 0.0002,
+      "loss": 2.5403,
+      "step": 8770
+    },
+    {
+      "epoch": 0.65424739195231,
+      "grad_norm": 2.1122303009033203,
+      "learning_rate": 0.0002,
+      "loss": 2.4911,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6549925484351714,
+      "grad_norm": 2.4388267993927,
+      "learning_rate": 0.0002,
+      "loss": 2.5928,
+      "step": 8790
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 2.3956820964813232,
+      "learning_rate": 0.0002,
+      "loss": 2.4162,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6564828614008942,
+      "grad_norm": 2.251885175704956,
+      "learning_rate": 0.0002,
+      "loss": 2.3572,
+      "step": 8810
+    },
+    {
+      "epoch": 0.6572280178837556,
+      "grad_norm": 2.1196508407592773,
+      "learning_rate": 0.0002,
+      "loss": 2.271,
+      "step": 8820
+    },
+    {
+      "epoch": 0.657973174366617,
+      "grad_norm": 2.6327478885650635,
+      "learning_rate": 0.0002,
+      "loss": 2.5528,
+      "step": 8830
+    },
+    {
+      "epoch": 0.6587183308494784,
+      "grad_norm": 3.1525380611419678,
+      "learning_rate": 0.0002,
+      "loss": 2.671,
+      "step": 8840
+    },
+    {
+      "epoch": 0.6594634873323398,
+      "grad_norm": 2.371023178100586,
+      "learning_rate": 0.0002,
+      "loss": 2.3042,
+      "step": 8850
+    },
+    {
+      "epoch": 0.6602086438152012,
+      "grad_norm": 2.1151058673858643,
+      "learning_rate": 0.0002,
+      "loss": 2.3586,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6609538002980626,
+      "grad_norm": 2.2981162071228027,
+      "learning_rate": 0.0002,
+      "loss": 2.3033,
+      "step": 8870
+    },
+    {
+      "epoch": 0.6616989567809239,
+      "grad_norm": 2.3385653495788574,
+      "learning_rate": 0.0002,
+      "loss": 2.4125,
+      "step": 8880
+    },
+    {
+      "epoch": 0.6624441132637854,
+      "grad_norm": 2.282998561859131,
+      "learning_rate": 0.0002,
+      "loss": 2.5846,
+      "step": 8890
+    },
+    {
+      "epoch": 0.6631892697466468,
+      "grad_norm": 2.276402473449707,
+      "learning_rate": 0.0002,
+      "loss": 2.5881,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6639344262295082,
+      "grad_norm": 2.753835439682007,
+      "learning_rate": 0.0002,
+      "loss": 2.5402,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6646795827123696,
+      "grad_norm": 2.155869960784912,
+      "learning_rate": 0.0002,
+      "loss": 2.3382,
+      "step": 8920
+    },
+    {
+      "epoch": 0.665424739195231,
+      "grad_norm": 2.25738263130188,
+      "learning_rate": 0.0002,
+      "loss": 2.5028,
+      "step": 8930
+    },
+    {
+      "epoch": 0.6661698956780924,
+      "grad_norm": 2.3716089725494385,
+      "learning_rate": 0.0002,
+      "loss": 2.4135,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6669150521609538,
+      "grad_norm": 2.5012192726135254,
+      "learning_rate": 0.0002,
+      "loss": 2.4297,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6676602086438153,
+      "grad_norm": 2.177103281021118,
+      "learning_rate": 0.0002,
+      "loss": 2.401,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6684053651266766,
+      "grad_norm": 2.500803232192993,
+      "learning_rate": 0.0002,
+      "loss": 2.6323,
+      "step": 8970
+    },
+    {
+      "epoch": 0.669150521609538,
+      "grad_norm": 2.7237913608551025,
+      "learning_rate": 0.0002,
+      "loss": 2.4543,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6698956780923994,
+      "grad_norm": 2.0506207942962646,
+      "learning_rate": 0.0002,
+      "loss": 2.3305,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6706408345752608,
+      "grad_norm": 2.182495355606079,
+      "learning_rate": 0.0002,
+      "loss": 2.3134,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6713859910581222,
+      "grad_norm": 2.3970160484313965,
+      "learning_rate": 0.0002,
+      "loss": 2.6011,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6721311475409836,
+      "grad_norm": 2.2599663734436035,
+      "learning_rate": 0.0002,
+      "loss": 2.2633,
+      "step": 9020
+    },
+    {
+      "epoch": 0.672876304023845,
+      "grad_norm": 2.202136754989624,
+      "learning_rate": 0.0002,
+      "loss": 2.5181,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6736214605067065,
+      "grad_norm": 2.4681708812713623,
+      "learning_rate": 0.0002,
+      "loss": 2.4997,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6743666169895678,
+      "grad_norm": 2.455841302871704,
+      "learning_rate": 0.0002,
+      "loss": 2.67,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6751117734724292,
+      "grad_norm": 2.679401397705078,
+      "learning_rate": 0.0002,
+      "loss": 2.4532,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6758569299552906,
+      "grad_norm": 2.003723621368408,
+      "learning_rate": 0.0002,
+      "loss": 2.6612,
+      "step": 9070
+    },
+    {
+      "epoch": 0.676602086438152,
+      "grad_norm": 2.256204128265381,
+      "learning_rate": 0.0002,
+      "loss": 2.6165,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6773472429210134,
+      "grad_norm": 2.3091988563537598,
+      "learning_rate": 0.0002,
+      "loss": 2.5914,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6780923994038748,
+      "grad_norm": 2.3021037578582764,
+      "learning_rate": 0.0002,
+      "loss": 2.5028,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6788375558867362,
+      "grad_norm": 2.1524159908294678,
+      "learning_rate": 0.0002,
+      "loss": 2.496,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6795827123695977,
+      "grad_norm": 2.2679061889648438,
+      "learning_rate": 0.0002,
+      "loss": 2.5092,
+      "step": 9120
+    },
+    {
+      "epoch": 0.680327868852459,
+      "grad_norm": 2.5078539848327637,
+      "learning_rate": 0.0002,
+      "loss": 2.2919,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6810730253353204,
+      "grad_norm": 2.5029258728027344,
+      "learning_rate": 0.0002,
+      "loss": 2.4382,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6818181818181818,
+      "grad_norm": 2.5736641883850098,
+      "learning_rate": 0.0002,
+      "loss": 2.4778,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6825633383010432,
+      "grad_norm": 2.327671527862549,
+      "learning_rate": 0.0002,
+      "loss": 2.5559,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6833084947839047,
+      "grad_norm": 2.310634136199951,
+      "learning_rate": 0.0002,
+      "loss": 2.4736,
+      "step": 9170
+    },
+    {
+      "epoch": 0.684053651266766,
+      "grad_norm": 2.2063729763031006,
+      "learning_rate": 0.0002,
+      "loss": 2.5829,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6847988077496274,
+      "grad_norm": 2.5868451595306396,
+      "learning_rate": 0.0002,
+      "loss": 2.4465,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6855439642324889,
+      "grad_norm": 1.984372854232788,
+      "learning_rate": 0.0002,
+      "loss": 2.4654,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6862891207153502,
+      "grad_norm": 2.2977192401885986,
+      "learning_rate": 0.0002,
+      "loss": 2.5643,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6870342771982116,
+      "grad_norm": 2.3139028549194336,
+      "learning_rate": 0.0002,
+      "loss": 2.7359,
+      "step": 9220
+    },
+    {
+      "epoch": 0.687779433681073,
+      "grad_norm": 2.124882698059082,
+      "learning_rate": 0.0002,
+      "loss": 2.4335,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6885245901639344,
+      "grad_norm": 2.434063673019409,
+      "learning_rate": 0.0002,
+      "loss": 2.3576,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6892697466467959,
+      "grad_norm": 2.290684700012207,
+      "learning_rate": 0.0002,
+      "loss": 2.7062,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6900149031296572,
+      "grad_norm": 2.5476014614105225,
+      "learning_rate": 0.0002,
+      "loss": 2.6915,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6907600596125186,
+      "grad_norm": 2.1429226398468018,
+      "learning_rate": 0.0002,
+      "loss": 2.5736,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6915052160953801,
+      "grad_norm": 2.399526834487915,
+      "learning_rate": 0.0002,
+      "loss": 2.3798,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6922503725782414,
+      "grad_norm": 2.5570576190948486,
+      "learning_rate": 0.0002,
+      "loss": 2.5188,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6929955290611028,
+      "grad_norm": 2.248030185699463,
+      "learning_rate": 0.0002,
+      "loss": 2.4761,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6937406855439643,
+      "grad_norm": 2.649503231048584,
+      "learning_rate": 0.0002,
+      "loss": 2.6177,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6944858420268256,
+      "grad_norm": 2.1536803245544434,
+      "learning_rate": 0.0002,
+      "loss": 2.5347,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6952309985096871,
+      "grad_norm": 2.3103137016296387,
+      "learning_rate": 0.0002,
+      "loss": 2.4589,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6959761549925484,
+      "grad_norm": 2.5560615062713623,
+      "learning_rate": 0.0002,
+      "loss": 2.6268,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6967213114754098,
+      "grad_norm": 2.149562120437622,
+      "learning_rate": 0.0002,
+      "loss": 2.7589,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6974664679582713,
+      "grad_norm": 2.180457592010498,
+      "learning_rate": 0.0002,
+      "loss": 2.3707,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6982116244411326,
+      "grad_norm": 2.1361920833587646,
+      "learning_rate": 0.0002,
+      "loss": 2.5647,
+      "step": 9370
+    },
+    {
+      "epoch": 0.698956780923994,
+      "grad_norm": 2.5958340167999268,
+      "learning_rate": 0.0002,
+      "loss": 2.6107,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6997019374068555,
+      "grad_norm": 2.298337936401367,
+      "learning_rate": 0.0002,
+      "loss": 2.5033,
+      "step": 9390
+    },
+    {
+      "epoch": 0.7004470938897168,
+      "grad_norm": 2.5497617721557617,
+      "learning_rate": 0.0002,
+      "loss": 2.6554,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7011922503725783,
+      "grad_norm": 2.4965898990631104,
+      "learning_rate": 0.0002,
+      "loss": 2.2896,
+      "step": 9410
+    },
+    {
+      "epoch": 0.7019374068554396,
+      "grad_norm": 1.6895242929458618,
+      "learning_rate": 0.0002,
+      "loss": 2.5128,
+      "step": 9420
+    },
+    {
+      "epoch": 0.702682563338301,
+      "grad_norm": 2.5977084636688232,
+      "learning_rate": 0.0002,
+      "loss": 2.6261,
+      "step": 9430
+    },
+    {
+      "epoch": 0.7034277198211625,
+      "grad_norm": 2.2413127422332764,
+      "learning_rate": 0.0002,
+      "loss": 2.5724,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7041728763040238,
+      "grad_norm": 1.9375770092010498,
+      "learning_rate": 0.0002,
+      "loss": 2.3573,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7049180327868853,
+      "grad_norm": 2.1468889713287354,
+      "learning_rate": 0.0002,
+      "loss": 2.4854,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7056631892697467,
+      "grad_norm": 2.3307502269744873,
+      "learning_rate": 0.0002,
+      "loss": 2.6102,
+      "step": 9470
+    },
+    {
+      "epoch": 0.706408345752608,
+      "grad_norm": 2.025935411453247,
+      "learning_rate": 0.0002,
+      "loss": 2.5878,
+      "step": 9480
+    },
+    {
+      "epoch": 0.7071535022354695,
+      "grad_norm": 2.0282442569732666,
+      "learning_rate": 0.0002,
+      "loss": 2.634,
+      "step": 9490
+    },
+    {
+      "epoch": 0.7078986587183308,
+      "grad_norm": 2.3716142177581787,
+      "learning_rate": 0.0002,
+      "loss": 2.523,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7086438152011922,
+      "grad_norm": 2.1333203315734863,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7093889716840537,
+      "grad_norm": 2.215022563934326,
+      "learning_rate": 0.0002,
+      "loss": 2.4586,
+      "step": 9520
+    },
+    {
+      "epoch": 0.710134128166915,
+      "grad_norm": 2.392059087753296,
+      "learning_rate": 0.0002,
+      "loss": 2.7049,
+      "step": 9530
+    },
+    {
+      "epoch": 0.7108792846497765,
+      "grad_norm": 2.0697712898254395,
+      "learning_rate": 0.0002,
+      "loss": 2.4652,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7116244411326379,
+      "grad_norm": 2.451186418533325,
+      "learning_rate": 0.0002,
+      "loss": 2.6232,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7123695976154992,
+      "grad_norm": 2.7246387004852295,
+      "learning_rate": 0.0002,
+      "loss": 2.6331,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7131147540983607,
+      "grad_norm": 2.2628626823425293,
+      "learning_rate": 0.0002,
+      "loss": 2.4974,
+      "step": 9570
+    },
+    {
+      "epoch": 0.713859910581222,
+      "grad_norm": 2.2943952083587646,
+      "learning_rate": 0.0002,
+      "loss": 2.5514,
+      "step": 9580
+    },
+    {
+      "epoch": 0.7146050670640834,
+      "grad_norm": 2.394134044647217,
+      "learning_rate": 0.0002,
+      "loss": 2.4207,
+      "step": 9590
+    },
+    {
+      "epoch": 0.7153502235469449,
+      "grad_norm": 2.2121686935424805,
+      "learning_rate": 0.0002,
+      "loss": 2.4458,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7160953800298062,
+      "grad_norm": 2.43963885307312,
+      "learning_rate": 0.0002,
+      "loss": 2.5707,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7168405365126677,
+      "grad_norm": 2.44991135597229,
+      "learning_rate": 0.0002,
+      "loss": 2.3977,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7175856929955291,
+      "grad_norm": 2.600816488265991,
+      "learning_rate": 0.0002,
+      "loss": 2.6366,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7183308494783904,
+      "grad_norm": 2.5606367588043213,
+      "learning_rate": 0.0002,
+      "loss": 2.3104,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7190760059612519,
+      "grad_norm": 2.0649945735931396,
+      "learning_rate": 0.0002,
+      "loss": 2.4726,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7198211624441133,
+      "grad_norm": 2.463927984237671,
+      "learning_rate": 0.0002,
+      "loss": 2.437,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7205663189269746,
+      "grad_norm": 2.189600706100464,
+      "learning_rate": 0.0002,
+      "loss": 2.6419,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7213114754098361,
+      "grad_norm": 1.8361003398895264,
+      "learning_rate": 0.0002,
+      "loss": 2.3392,
+      "step": 9680
+    },
+    {
+      "epoch": 0.7220566318926974,
+      "grad_norm": 2.4830501079559326,
+      "learning_rate": 0.0002,
+      "loss": 2.6257,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7228017883755589,
+      "grad_norm": 2.311711072921753,
+      "learning_rate": 0.0002,
+      "loss": 2.4023,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7235469448584203,
+      "grad_norm": 1.9280378818511963,
+      "learning_rate": 0.0002,
+      "loss": 2.6589,
+      "step": 9710
+    },
+    {
+      "epoch": 0.7242921013412816,
+      "grad_norm": 2.1411705017089844,
+      "learning_rate": 0.0002,
+      "loss": 2.539,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7250372578241431,
+      "grad_norm": 2.262427568435669,
+      "learning_rate": 0.0002,
+      "loss": 2.3129,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7257824143070045,
+      "grad_norm": 2.005398988723755,
+      "learning_rate": 0.0002,
+      "loss": 2.4774,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7265275707898659,
+      "grad_norm": 2.4369115829467773,
+      "learning_rate": 0.0002,
+      "loss": 2.5077,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 2.5426080226898193,
+      "learning_rate": 0.0002,
+      "loss": 2.4542,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7280178837555886,
+      "grad_norm": 2.222259044647217,
+      "learning_rate": 0.0002,
+      "loss": 2.5733,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7287630402384501,
+      "grad_norm": 3.0009191036224365,
+      "learning_rate": 0.0002,
+      "loss": 2.5769,
+      "step": 9780
+    },
+    {
+      "epoch": 0.7295081967213115,
+      "grad_norm": 2.354903221130371,
+      "learning_rate": 0.0002,
+      "loss": 2.4887,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7302533532041728,
+      "grad_norm": 2.4170987606048584,
+      "learning_rate": 0.0002,
+      "loss": 2.3048,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7309985096870343,
+      "grad_norm": 2.6301980018615723,
+      "learning_rate": 0.0002,
+      "loss": 2.2994,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7317436661698957,
+      "grad_norm": 1.6262503862380981,
+      "learning_rate": 0.0002,
+      "loss": 2.0802,
+      "step": 9820
+    },
+    {
+      "epoch": 0.732488822652757,
+      "grad_norm": 2.165588855743408,
+      "learning_rate": 0.0002,
+      "loss": 2.5009,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7332339791356185,
+      "grad_norm": 2.3280584812164307,
+      "learning_rate": 0.0002,
+      "loss": 2.5506,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7339791356184798,
+      "grad_norm": 2.505038261413574,
+      "learning_rate": 0.0002,
+      "loss": 2.4491,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7347242921013413,
+      "grad_norm": 2.5628268718719482,
+      "learning_rate": 0.0002,
+      "loss": 2.5678,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7354694485842027,
+      "grad_norm": 2.371814489364624,
+      "learning_rate": 0.0002,
+      "loss": 2.4223,
+      "step": 9870
+    },
+    {
+      "epoch": 0.736214605067064,
+      "grad_norm": 1.9160370826721191,
+      "learning_rate": 0.0002,
+      "loss": 2.5315,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7369597615499255,
+      "grad_norm": 2.015497922897339,
+      "learning_rate": 0.0002,
+      "loss": 2.6611,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7377049180327869,
+      "grad_norm": 2.402764081954956,
+      "learning_rate": 0.0002,
+      "loss": 2.5224,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7384500745156483,
+      "grad_norm": 2.2813656330108643,
+      "learning_rate": 0.0002,
+      "loss": 2.6005,
+      "step": 9910
+    },
+    {
+      "epoch": 0.7391952309985097,
+      "grad_norm": 2.1747665405273438,
+      "learning_rate": 0.0002,
+      "loss": 2.6124,
+      "step": 9920
+    },
+    {
+      "epoch": 0.7399403874813711,
+      "grad_norm": 2.8765082359313965,
+      "learning_rate": 0.0002,
+      "loss": 2.4271,
+      "step": 9930
+    },
+    {
+      "epoch": 0.7406855439642325,
+      "grad_norm": 2.9332666397094727,
+      "learning_rate": 0.0002,
+      "loss": 2.5666,
+      "step": 9940
+    },
+    {
+      "epoch": 0.7414307004470939,
+      "grad_norm": 2.209160566329956,
+      "learning_rate": 0.0002,
+      "loss": 2.5214,
+      "step": 9950
+    },
+    {
+      "epoch": 0.7421758569299552,
+      "grad_norm": 2.141798734664917,
+      "learning_rate": 0.0002,
+      "loss": 2.5629,
+      "step": 9960
+    },
+    {
+      "epoch": 0.7429210134128167,
+      "grad_norm": 2.4280612468719482,
+      "learning_rate": 0.0002,
+      "loss": 2.6251,
+      "step": 9970
+    },
+    {
+      "epoch": 0.7436661698956781,
+      "grad_norm": 2.588738441467285,
+      "learning_rate": 0.0002,
+      "loss": 2.5636,
+      "step": 9980
+    },
+    {
+      "epoch": 0.7444113263785395,
+      "grad_norm": 2.423440456390381,
+      "learning_rate": 0.0002,
+      "loss": 2.4463,
+      "step": 9990
+    },
+    {
+      "epoch": 0.7451564828614009,
+      "grad_norm": 2.3735451698303223,
+      "learning_rate": 0.0002,
+      "loss": 2.2597,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7459016393442623,
+      "grad_norm": 2.585657835006714,
+      "learning_rate": 0.0002,
+      "loss": 2.5384,
+      "step": 10010
+    },
+    {
+      "epoch": 0.7466467958271237,
+      "grad_norm": 2.6739962100982666,
+      "learning_rate": 0.0002,
+      "loss": 2.413,
+      "step": 10020
+    },
+    {
+      "epoch": 0.7473919523099851,
+      "grad_norm": 2.564932346343994,
+      "learning_rate": 0.0002,
+      "loss": 2.4084,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7481371087928465,
+      "grad_norm": 2.0360469818115234,
+      "learning_rate": 0.0002,
+      "loss": 2.5756,
+      "step": 10040
+    },
+    {
+      "epoch": 0.7488822652757079,
+      "grad_norm": 2.26521897315979,
+      "learning_rate": 0.0002,
+      "loss": 2.3981,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7496274217585693,
+      "grad_norm": 2.3003921508789062,
+      "learning_rate": 0.0002,
+      "loss": 2.599,
+      "step": 10060
+    },
+    {
+      "epoch": 0.7503725782414307,
+      "grad_norm": 2.3010787963867188,
+      "learning_rate": 0.0002,
+      "loss": 2.6884,
+      "step": 10070
+    },
+    {
+      "epoch": 0.7511177347242921,
+      "grad_norm": 2.4945406913757324,
+      "learning_rate": 0.0002,
+      "loss": 2.6397,
+      "step": 10080
+    },
+    {
+      "epoch": 0.7518628912071535,
+      "grad_norm": 2.391580104827881,
+      "learning_rate": 0.0002,
+      "loss": 2.4918,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7526080476900149,
+      "grad_norm": 2.094149589538574,
+      "learning_rate": 0.0002,
+      "loss": 2.5465,
+      "step": 10100
+    },
+    {
+      "epoch": 0.7533532041728763,
+      "grad_norm": 2.440086603164673,
+      "learning_rate": 0.0002,
+      "loss": 2.665,
+      "step": 10110
+    },
+    {
+      "epoch": 0.7540983606557377,
+      "grad_norm": 2.482935905456543,
+      "learning_rate": 0.0002,
+      "loss": 2.5492,
+      "step": 10120
+    },
+    {
+      "epoch": 0.7548435171385991,
+      "grad_norm": 2.4048640727996826,
+      "learning_rate": 0.0002,
+      "loss": 2.2824,
+      "step": 10130
+    },
+    {
+      "epoch": 0.7555886736214605,
+      "grad_norm": 2.5780625343322754,
+      "learning_rate": 0.0002,
+      "loss": 2.5935,
+      "step": 10140
+    },
+    {
+      "epoch": 0.7563338301043219,
+      "grad_norm": 1.9736360311508179,
+      "learning_rate": 0.0002,
+      "loss": 2.6113,
+      "step": 10150
+    },
+    {
+      "epoch": 0.7570789865871833,
+      "grad_norm": 2.432325839996338,
+      "learning_rate": 0.0002,
+      "loss": 2.5351,
+      "step": 10160
+    },
+    {
+      "epoch": 0.7578241430700448,
+      "grad_norm": 1.9205713272094727,
+      "learning_rate": 0.0002,
+      "loss": 2.4855,
+      "step": 10170
+    },
+    {
+      "epoch": 0.7585692995529061,
+      "grad_norm": 2.043088436126709,
+      "learning_rate": 0.0002,
+      "loss": 2.5828,
+      "step": 10180
+    },
+    {
+      "epoch": 0.7593144560357675,
+      "grad_norm": 3.1941723823547363,
+      "learning_rate": 0.0002,
+      "loss": 2.6751,
+      "step": 10190
+    },
+    {
+      "epoch": 0.7600596125186289,
+      "grad_norm": 2.210202693939209,
+      "learning_rate": 0.0002,
+      "loss": 2.4979,
+      "step": 10200
+    },
+    {
+      "epoch": 0.7608047690014903,
+      "grad_norm": 2.0645289421081543,
+      "learning_rate": 0.0002,
+      "loss": 2.5,
+      "step": 10210
+    },
+    {
+      "epoch": 0.7615499254843517,
+      "grad_norm": 2.4983456134796143,
+      "learning_rate": 0.0002,
+      "loss": 2.501,
+      "step": 10220
+    },
+    {
+      "epoch": 0.7622950819672131,
+      "grad_norm": 2.4388515949249268,
+      "learning_rate": 0.0002,
+      "loss": 2.5407,
+      "step": 10230
+    },
+    {
+      "epoch": 0.7630402384500745,
+      "grad_norm": 1.960217833518982,
+      "learning_rate": 0.0002,
+      "loss": 2.3579,
+      "step": 10240
+    },
+    {
+      "epoch": 0.763785394932936,
+      "grad_norm": 2.206451177597046,
+      "learning_rate": 0.0002,
+      "loss": 2.5991,
+      "step": 10250
+    },
+    {
+      "epoch": 0.7645305514157973,
+      "grad_norm": 3.0600955486297607,
+      "learning_rate": 0.0002,
+      "loss": 2.534,
+      "step": 10260
+    },
+    {
+      "epoch": 0.7652757078986587,
+      "grad_norm": 2.1760904788970947,
+      "learning_rate": 0.0002,
+      "loss": 2.5936,
+      "step": 10270
+    },
+    {
+      "epoch": 0.7660208643815202,
+      "grad_norm": 2.1374666690826416,
+      "learning_rate": 0.0002,
+      "loss": 2.3698,
+      "step": 10280
+    },
+    {
+      "epoch": 0.7667660208643815,
+      "grad_norm": 1.8994402885437012,
+      "learning_rate": 0.0002,
+      "loss": 2.5065,
+      "step": 10290
+    },
+    {
+      "epoch": 0.767511177347243,
+      "grad_norm": 1.7293874025344849,
+      "learning_rate": 0.0002,
+      "loss": 2.5013,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7682563338301043,
+      "grad_norm": 2.463646650314331,
+      "learning_rate": 0.0002,
+      "loss": 2.606,
+      "step": 10310
+    },
+    {
+      "epoch": 0.7690014903129657,
+      "grad_norm": 2.521929979324341,
+      "learning_rate": 0.0002,
+      "loss": 2.3497,
+      "step": 10320
+    },
+    {
+      "epoch": 0.7697466467958272,
+      "grad_norm": 2.0710620880126953,
+      "learning_rate": 0.0002,
+      "loss": 2.3254,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7704918032786885,
+      "grad_norm": 2.2939655780792236,
+      "learning_rate": 0.0002,
+      "loss": 2.5636,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7712369597615499,
+      "grad_norm": 2.3778505325317383,
+      "learning_rate": 0.0002,
+      "loss": 2.426,
+      "step": 10350
+    },
+    {
+      "epoch": 0.7719821162444114,
+      "grad_norm": 2.2639496326446533,
+      "learning_rate": 0.0002,
+      "loss": 2.5813,
+      "step": 10360
+    },
+    {
+      "epoch": 0.7727272727272727,
+      "grad_norm": 1.964016318321228,
+      "learning_rate": 0.0002,
+      "loss": 2.554,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7734724292101341,
+      "grad_norm": 2.1532862186431885,
+      "learning_rate": 0.0002,
+      "loss": 2.4438,
+      "step": 10380
+    },
+    {
+      "epoch": 0.7742175856929955,
+      "grad_norm": 2.354395627975464,
+      "learning_rate": 0.0002,
+      "loss": 2.6894,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7749627421758569,
+      "grad_norm": 2.0978426933288574,
+      "learning_rate": 0.0002,
+      "loss": 2.5408,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7757078986587184,
+      "grad_norm": 2.2250595092773438,
+      "learning_rate": 0.0002,
+      "loss": 2.4457,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7764530551415797,
+      "grad_norm": 2.65061354637146,
+      "learning_rate": 0.0002,
+      "loss": 2.3338,
+      "step": 10420
+    },
+    {
+      "epoch": 0.7771982116244411,
+      "grad_norm": 2.519925117492676,
+      "learning_rate": 0.0002,
+      "loss": 2.6637,
+      "step": 10430
+    },
+    {
+      "epoch": 0.7779433681073026,
+      "grad_norm": 1.8276631832122803,
+      "learning_rate": 0.0002,
+      "loss": 2.4743,
+      "step": 10440
+    },
+    {
+      "epoch": 0.7786885245901639,
+      "grad_norm": 2.745326042175293,
+      "learning_rate": 0.0002,
+      "loss": 2.4987,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7794336810730254,
+      "grad_norm": 2.3664751052856445,
+      "learning_rate": 0.0002,
+      "loss": 2.6172,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7801788375558867,
+      "grad_norm": 2.136486768722534,
+      "learning_rate": 0.0002,
+      "loss": 2.581,
+      "step": 10470
+    },
+    {
+      "epoch": 0.7809239940387481,
+      "grad_norm": 2.13596773147583,
+      "learning_rate": 0.0002,
+      "loss": 2.4819,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7816691505216096,
+      "grad_norm": 2.0259013175964355,
+      "learning_rate": 0.0002,
+      "loss": 2.4299,
+      "step": 10490
+    },
+    {
+      "epoch": 0.7824143070044709,
+      "grad_norm": 2.3508143424987793,
+      "learning_rate": 0.0002,
+      "loss": 2.6114,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7831594634873323,
+      "grad_norm": 2.127795457839966,
+      "learning_rate": 0.0002,
+      "loss": 2.5393,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7839046199701938,
+      "grad_norm": 2.1637115478515625,
+      "learning_rate": 0.0002,
+      "loss": 2.6723,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7846497764530551,
+      "grad_norm": 2.2965123653411865,
+      "learning_rate": 0.0002,
+      "loss": 2.5549,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7853949329359166,
+      "grad_norm": 2.3020741939544678,
+      "learning_rate": 0.0002,
+      "loss": 2.5971,
+      "step": 10540
+    },
+    {
+      "epoch": 0.786140089418778,
+      "grad_norm": 2.3106722831726074,
+      "learning_rate": 0.0002,
+      "loss": 2.5852,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 2.332108736038208,
+      "learning_rate": 0.0002,
+      "loss": 2.3608,
+      "step": 10560
+    },
+    {
+      "epoch": 0.7876304023845008,
+      "grad_norm": 2.337618589401245,
+      "learning_rate": 0.0002,
+      "loss": 2.3785,
+      "step": 10570
+    },
+    {
+      "epoch": 0.7883755588673621,
+      "grad_norm": 2.413151502609253,
+      "learning_rate": 0.0002,
+      "loss": 2.6424,
+      "step": 10580
+    },
+    {
+      "epoch": 0.7891207153502235,
+      "grad_norm": 1.9372957944869995,
+      "learning_rate": 0.0002,
+      "loss": 2.5991,
+      "step": 10590
+    },
+    {
+      "epoch": 0.789865871833085,
+      "grad_norm": 2.435525417327881,
+      "learning_rate": 0.0002,
+      "loss": 2.5649,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7906110283159463,
+      "grad_norm": 2.385913372039795,
+      "learning_rate": 0.0002,
+      "loss": 2.386,
+      "step": 10610
+    },
+    {
+      "epoch": 0.7913561847988078,
+      "grad_norm": 2.0109829902648926,
+      "learning_rate": 0.0002,
+      "loss": 2.4575,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7921013412816692,
+      "grad_norm": 2.28608775138855,
+      "learning_rate": 0.0002,
+      "loss": 2.5546,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7928464977645305,
+      "grad_norm": 2.1904425621032715,
+      "learning_rate": 0.0002,
+      "loss": 2.4409,
+      "step": 10640
+    },
+    {
+      "epoch": 0.793591654247392,
+      "grad_norm": 2.351679801940918,
+      "learning_rate": 0.0002,
+      "loss": 2.6908,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7943368107302533,
+      "grad_norm": 2.2788383960723877,
+      "learning_rate": 0.0002,
+      "loss": 2.4305,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7950819672131147,
+      "grad_norm": 2.0748603343963623,
+      "learning_rate": 0.0002,
+      "loss": 2.2788,
+      "step": 10670
+    },
+    {
+      "epoch": 0.7958271236959762,
+      "grad_norm": 2.1148083209991455,
+      "learning_rate": 0.0002,
+      "loss": 2.5214,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7965722801788375,
+      "grad_norm": 2.439086675643921,
+      "learning_rate": 0.0002,
+      "loss": 2.572,
+      "step": 10690
+    },
+    {
+      "epoch": 0.797317436661699,
+      "grad_norm": 2.302316188812256,
+      "learning_rate": 0.0002,
+      "loss": 2.6505,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7980625931445604,
+      "grad_norm": 1.977591633796692,
+      "learning_rate": 0.0002,
+      "loss": 2.6918,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7988077496274217,
+      "grad_norm": 2.8533225059509277,
+      "learning_rate": 0.0002,
+      "loss": 2.6504,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7995529061102832,
+      "grad_norm": 2.3402719497680664,
+      "learning_rate": 0.0002,
+      "loss": 2.5394,
+      "step": 10730
+    },
+    {
+      "epoch": 0.8002980625931445,
+      "grad_norm": 1.8375385999679565,
+      "learning_rate": 0.0002,
+      "loss": 2.4388,
+      "step": 10740
+    },
+    {
+      "epoch": 0.801043219076006,
+      "grad_norm": 2.375514030456543,
+      "learning_rate": 0.0002,
+      "loss": 2.2368,
+      "step": 10750
+    },
+    {
+      "epoch": 0.8017883755588674,
+      "grad_norm": 2.218656063079834,
+      "learning_rate": 0.0002,
+      "loss": 2.6589,
+      "step": 10760
+    },
+    {
+      "epoch": 0.8025335320417287,
+      "grad_norm": 2.560631513595581,
+      "learning_rate": 0.0002,
+      "loss": 2.3786,
+      "step": 10770
+    },
+    {
+      "epoch": 0.8032786885245902,
+      "grad_norm": 2.4114229679107666,
+      "learning_rate": 0.0002,
+      "loss": 2.5688,
+      "step": 10780
+    },
+    {
+      "epoch": 0.8040238450074516,
+      "grad_norm": 2.948805570602417,
+      "learning_rate": 0.0002,
+      "loss": 2.5296,
+      "step": 10790
+    },
+    {
+      "epoch": 0.8047690014903129,
+      "grad_norm": 2.384042501449585,
+      "learning_rate": 0.0002,
+      "loss": 2.6218,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8055141579731744,
+      "grad_norm": 2.3185818195343018,
+      "learning_rate": 0.0002,
+      "loss": 2.5289,
+      "step": 10810
+    },
+    {
+      "epoch": 0.8062593144560357,
+      "grad_norm": 2.2557380199432373,
+      "learning_rate": 0.0002,
+      "loss": 2.4439,
+      "step": 10820
+    },
+    {
+      "epoch": 0.8070044709388972,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 0.0002,
+      "loss": 2.4567,
+      "step": 10830
+    },
+    {
+      "epoch": 0.8077496274217586,
+      "grad_norm": 2.480273962020874,
+      "learning_rate": 0.0002,
+      "loss": 2.4788,
+      "step": 10840
+    },
+    {
+      "epoch": 0.8084947839046199,
+      "grad_norm": 2.4017157554626465,
+      "learning_rate": 0.0002,
+      "loss": 2.4639,
+      "step": 10850
+    },
+    {
+      "epoch": 0.8092399403874814,
+      "grad_norm": 2.3682639598846436,
+      "learning_rate": 0.0002,
+      "loss": 2.5533,
+      "step": 10860
+    },
+    {
+      "epoch": 0.8099850968703428,
+      "grad_norm": 2.693796157836914,
+      "learning_rate": 0.0002,
+      "loss": 2.4756,
+      "step": 10870
+    },
+    {
+      "epoch": 0.8107302533532041,
+      "grad_norm": 2.371288776397705,
+      "learning_rate": 0.0002,
+      "loss": 2.4941,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8114754098360656,
+      "grad_norm": 2.4420065879821777,
+      "learning_rate": 0.0002,
+      "loss": 2.5901,
+      "step": 10890
+    },
+    {
+      "epoch": 0.812220566318927,
+      "grad_norm": 2.154177665710449,
+      "learning_rate": 0.0002,
+      "loss": 2.4709,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8129657228017884,
+      "grad_norm": 2.3035285472869873,
+      "learning_rate": 0.0002,
+      "loss": 2.2832,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8137108792846498,
+      "grad_norm": 2.2672736644744873,
+      "learning_rate": 0.0002,
+      "loss": 2.8001,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8144560357675111,
+      "grad_norm": 2.507875442504883,
+      "learning_rate": 0.0002,
+      "loss": 2.5625,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8152011922503726,
+      "grad_norm": 2.542093276977539,
+      "learning_rate": 0.0002,
+      "loss": 2.5738,
+      "step": 10940
+    },
+    {
+      "epoch": 0.815946348733234,
+      "grad_norm": 2.764739990234375,
+      "learning_rate": 0.0002,
+      "loss": 2.6649,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8166915052160953,
+      "grad_norm": 2.2417197227478027,
+      "learning_rate": 0.0002,
+      "loss": 2.5701,
+      "step": 10960
+    },
+    {
+      "epoch": 0.8174366616989568,
+      "grad_norm": 2.3203530311584473,
+      "learning_rate": 0.0002,
+      "loss": 2.57,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 2.8699557781219482,
+      "learning_rate": 0.0002,
+      "loss": 2.5023,
+      "step": 10980
+    },
+    {
+      "epoch": 0.8189269746646796,
+      "grad_norm": 2.4314422607421875,
+      "learning_rate": 0.0002,
+      "loss": 2.5016,
+      "step": 10990
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 2.330038547515869,
+      "learning_rate": 0.0002,
+      "loss": 2.3436,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8204172876304023,
+      "grad_norm": 2.177577257156372,
+      "learning_rate": 0.0002,
+      "loss": 2.5652,
+      "step": 11010
+    },
+    {
+      "epoch": 0.8211624441132638,
+      "grad_norm": 1.8706187009811401,
+      "learning_rate": 0.0002,
+      "loss": 2.7162,
+      "step": 11020
+    },
+    {
+      "epoch": 0.8219076005961252,
+      "grad_norm": 2.2145588397979736,
+      "learning_rate": 0.0002,
+      "loss": 2.5284,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8226527570789866,
+      "grad_norm": 2.344332456588745,
+      "learning_rate": 0.0002,
+      "loss": 2.6446,
+      "step": 11040
+    },
+    {
+      "epoch": 0.823397913561848,
+      "grad_norm": 2.391944646835327,
+      "learning_rate": 0.0002,
+      "loss": 2.5797,
+      "step": 11050
+    },
+    {
+      "epoch": 0.8241430700447094,
+      "grad_norm": 2.2976560592651367,
+      "learning_rate": 0.0002,
+      "loss": 2.3885,
+      "step": 11060
+    },
+    {
+      "epoch": 0.8248882265275708,
+      "grad_norm": 2.0775372982025146,
+      "learning_rate": 0.0002,
+      "loss": 2.5791,
+      "step": 11070
+    },
+    {
+      "epoch": 0.8256333830104322,
+      "grad_norm": 2.264828681945801,
+      "learning_rate": 0.0002,
+      "loss": 2.6107,
+      "step": 11080
+    },
+    {
+      "epoch": 0.8263785394932935,
+      "grad_norm": 2.1409361362457275,
+      "learning_rate": 0.0002,
+      "loss": 2.5045,
+      "step": 11090
+    },
+    {
+      "epoch": 0.827123695976155,
+      "grad_norm": 2.565410614013672,
+      "learning_rate": 0.0002,
+      "loss": 2.5521,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8278688524590164,
+      "grad_norm": 2.375131607055664,
+      "learning_rate": 0.0002,
+      "loss": 2.3359,
+      "step": 11110
+    },
+    {
+      "epoch": 0.8286140089418778,
+      "grad_norm": 2.2293524742126465,
+      "learning_rate": 0.0002,
+      "loss": 2.3675,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8293591654247392,
+      "grad_norm": 2.4063050746917725,
+      "learning_rate": 0.0002,
+      "loss": 2.4713,
+      "step": 11130
+    },
+    {
+      "epoch": 0.8301043219076006,
+      "grad_norm": 2.408867597579956,
+      "learning_rate": 0.0002,
+      "loss": 2.5096,
+      "step": 11140
+    },
+    {
+      "epoch": 0.830849478390462,
+      "grad_norm": 2.5441880226135254,
+      "learning_rate": 0.0002,
+      "loss": 2.5106,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8315946348733234,
+      "grad_norm": 2.539547920227051,
+      "learning_rate": 0.0002,
+      "loss": 2.4346,
+      "step": 11160
+    },
+    {
+      "epoch": 0.8323397913561847,
+      "grad_norm": 2.414623260498047,
+      "learning_rate": 0.0002,
+      "loss": 2.5962,
+      "step": 11170
+    },
+    {
+      "epoch": 0.8330849478390462,
+      "grad_norm": 2.219356060028076,
+      "learning_rate": 0.0002,
+      "loss": 2.5249,
+      "step": 11180
+    },
+    {
+      "epoch": 0.8338301043219076,
+      "grad_norm": 2.43898606300354,
+      "learning_rate": 0.0002,
+      "loss": 2.6075,
+      "step": 11190
+    },
+    {
+      "epoch": 0.834575260804769,
+      "grad_norm": 2.644263744354248,
+      "learning_rate": 0.0002,
+      "loss": 2.5456,
+      "step": 11200
+    },
+    {
+      "epoch": 0.8353204172876304,
+      "grad_norm": 2.2671680450439453,
+      "learning_rate": 0.0002,
+      "loss": 2.4596,
+      "step": 11210
+    },
+    {
+      "epoch": 0.8360655737704918,
+      "grad_norm": 2.3897671699523926,
+      "learning_rate": 0.0002,
+      "loss": 2.5104,
+      "step": 11220
+    },
+    {
+      "epoch": 0.8368107302533532,
+      "grad_norm": 2.2913897037506104,
+      "learning_rate": 0.0002,
+      "loss": 2.6487,
+      "step": 11230
+    },
+    {
+      "epoch": 0.8375558867362146,
+      "grad_norm": 2.287472724914551,
+      "learning_rate": 0.0002,
+      "loss": 2.6213,
+      "step": 11240
+    },
+    {
+      "epoch": 0.8383010432190761,
+      "grad_norm": 2.184960126876831,
+      "learning_rate": 0.0002,
+      "loss": 2.559,
+      "step": 11250
+    },
+    {
+      "epoch": 0.8390461997019374,
+      "grad_norm": 2.3212270736694336,
+      "learning_rate": 0.0002,
+      "loss": 2.4804,
+      "step": 11260
+    },
+    {
+      "epoch": 0.8397913561847988,
+      "grad_norm": 2.361088991165161,
+      "learning_rate": 0.0002,
+      "loss": 2.7022,
+      "step": 11270
+    },
+    {
+      "epoch": 0.8405365126676602,
+      "grad_norm": 2.6766278743743896,
+      "learning_rate": 0.0002,
+      "loss": 2.5002,
+      "step": 11280
+    },
+    {
+      "epoch": 0.8412816691505216,
+      "grad_norm": 2.4925765991210938,
+      "learning_rate": 0.0002,
+      "loss": 2.5971,
+      "step": 11290
+    },
+    {
+      "epoch": 0.842026825633383,
+      "grad_norm": 2.4175381660461426,
+      "learning_rate": 0.0002,
+      "loss": 2.5388,
+      "step": 11300
+    },
+    {
+      "epoch": 0.8427719821162444,
+      "grad_norm": 2.3096563816070557,
+      "learning_rate": 0.0002,
+      "loss": 2.6924,
+      "step": 11310
+    },
+    {
+      "epoch": 0.8435171385991058,
+      "grad_norm": 2.5850837230682373,
+      "learning_rate": 0.0002,
+      "loss": 2.5236,
+      "step": 11320
+    },
+    {
+      "epoch": 0.8442622950819673,
+      "grad_norm": 2.491636276245117,
+      "learning_rate": 0.0002,
+      "loss": 2.3833,
+      "step": 11330
+    },
+    {
+      "epoch": 0.8450074515648286,
+      "grad_norm": 2.627934455871582,
+      "learning_rate": 0.0002,
+      "loss": 2.4309,
+      "step": 11340
+    },
+    {
+      "epoch": 0.84575260804769,
+      "grad_norm": 2.562291383743286,
+      "learning_rate": 0.0002,
+      "loss": 2.6193,
+      "step": 11350
+    },
+    {
+      "epoch": 0.8464977645305514,
+      "grad_norm": 2.1559548377990723,
+      "learning_rate": 0.0002,
+      "loss": 2.3262,
+      "step": 11360
+    },
+    {
+      "epoch": 0.8472429210134128,
+      "grad_norm": 2.06011962890625,
+      "learning_rate": 0.0002,
+      "loss": 2.4812,
+      "step": 11370
+    },
+    {
+      "epoch": 0.8479880774962743,
+      "grad_norm": 1.6976170539855957,
+      "learning_rate": 0.0002,
+      "loss": 2.482,
+      "step": 11380
+    },
+    {
+      "epoch": 0.8487332339791356,
+      "grad_norm": 2.2001445293426514,
+      "learning_rate": 0.0002,
+      "loss": 2.6044,
+      "step": 11390
+    },
+    {
+      "epoch": 0.849478390461997,
+      "grad_norm": 2.1249303817749023,
+      "learning_rate": 0.0002,
+      "loss": 2.395,
+      "step": 11400
+    },
+    {
+      "epoch": 0.8502235469448585,
+      "grad_norm": 2.477747678756714,
+      "learning_rate": 0.0002,
+      "loss": 2.3953,
+      "step": 11410
+    },
+    {
+      "epoch": 0.8509687034277198,
+      "grad_norm": 2.436199426651001,
+      "learning_rate": 0.0002,
+      "loss": 2.535,
+      "step": 11420
+    },
+    {
+      "epoch": 0.8517138599105812,
+      "grad_norm": 2.53760027885437,
+      "learning_rate": 0.0002,
+      "loss": 2.4221,
+      "step": 11430
+    },
+    {
+      "epoch": 0.8524590163934426,
+      "grad_norm": 2.1996359825134277,
+      "learning_rate": 0.0002,
+      "loss": 2.3221,
+      "step": 11440
+    },
+    {
+      "epoch": 0.853204172876304,
+      "grad_norm": 2.1827547550201416,
+      "learning_rate": 0.0002,
+      "loss": 2.4165,
+      "step": 11450
+    },
+    {
+      "epoch": 0.8539493293591655,
+      "grad_norm": 2.6674787998199463,
+      "learning_rate": 0.0002,
+      "loss": 2.524,
+      "step": 11460
+    },
+    {
+      "epoch": 0.8546944858420268,
+      "grad_norm": 2.6001546382904053,
+      "learning_rate": 0.0002,
+      "loss": 2.6605,
+      "step": 11470
+    },
+    {
+      "epoch": 0.8554396423248882,
+      "grad_norm": 2.287266254425049,
+      "learning_rate": 0.0002,
+      "loss": 2.6006,
+      "step": 11480
+    },
+    {
+      "epoch": 0.8561847988077497,
+      "grad_norm": 2.252857208251953,
+      "learning_rate": 0.0002,
+      "loss": 2.544,
+      "step": 11490
+    },
+    {
+      "epoch": 0.856929955290611,
+      "grad_norm": 2.273512601852417,
+      "learning_rate": 0.0002,
+      "loss": 2.5486,
+      "step": 11500
+    },
+    {
+      "epoch": 0.8576751117734724,
+      "grad_norm": 2.7023732662200928,
+      "learning_rate": 0.0002,
+      "loss": 2.592,
+      "step": 11510
+    },
+    {
+      "epoch": 0.8584202682563339,
+      "grad_norm": 2.303313970565796,
+      "learning_rate": 0.0002,
+      "loss": 2.5199,
+      "step": 11520
+    },
+    {
+      "epoch": 0.8591654247391952,
+      "grad_norm": 2.4638290405273438,
+      "learning_rate": 0.0002,
+      "loss": 2.5568,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8599105812220567,
+      "grad_norm": 2.451723337173462,
+      "learning_rate": 0.0002,
+      "loss": 2.6141,
+      "step": 11540
+    },
+    {
+      "epoch": 0.860655737704918,
+      "grad_norm": 2.3285202980041504,
+      "learning_rate": 0.0002,
+      "loss": 2.5687,
+      "step": 11550
+    },
+    {
+      "epoch": 0.8614008941877794,
+      "grad_norm": 2.085536241531372,
+      "learning_rate": 0.0002,
+      "loss": 2.6193,
+      "step": 11560
+    },
+    {
+      "epoch": 0.8621460506706409,
+      "grad_norm": 2.5319440364837646,
+      "learning_rate": 0.0002,
+      "loss": 2.5323,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8628912071535022,
+      "grad_norm": 2.246924638748169,
+      "learning_rate": 0.0002,
+      "loss": 2.4782,
+      "step": 11580
+    },
+    {
+      "epoch": 0.8636363636363636,
+      "grad_norm": 2.1741697788238525,
+      "learning_rate": 0.0002,
+      "loss": 2.4644,
+      "step": 11590
+    },
+    {
+      "epoch": 0.8643815201192251,
+      "grad_norm": 2.434746265411377,
+      "learning_rate": 0.0002,
+      "loss": 2.3757,
+      "step": 11600
+    },
+    {
+      "epoch": 0.8651266766020864,
+      "grad_norm": 2.406317710876465,
+      "learning_rate": 0.0002,
+      "loss": 2.5596,
+      "step": 11610
+    },
+    {
+      "epoch": 0.8658718330849479,
+      "grad_norm": 2.0323896408081055,
+      "learning_rate": 0.0002,
+      "loss": 2.491,
+      "step": 11620
+    },
+    {
+      "epoch": 0.8666169895678092,
+      "grad_norm": 1.8637727499008179,
+      "learning_rate": 0.0002,
+      "loss": 2.5684,
+      "step": 11630
+    },
+    {
+      "epoch": 0.8673621460506706,
+      "grad_norm": 2.324962854385376,
+      "learning_rate": 0.0002,
+      "loss": 2.5031,
+      "step": 11640
+    },
+    {
+      "epoch": 0.8681073025335321,
+      "grad_norm": 2.261016607284546,
+      "learning_rate": 0.0002,
+      "loss": 2.4632,
+      "step": 11650
+    },
+    {
+      "epoch": 0.8688524590163934,
+      "grad_norm": 2.4027981758117676,
+      "learning_rate": 0.0002,
+      "loss": 2.5733,
+      "step": 11660
+    },
+    {
+      "epoch": 0.8695976154992549,
+      "grad_norm": 2.2393276691436768,
+      "learning_rate": 0.0002,
+      "loss": 2.4513,
+      "step": 11670
+    },
+    {
+      "epoch": 0.8703427719821163,
+      "grad_norm": 2.2299957275390625,
+      "learning_rate": 0.0002,
+      "loss": 2.4286,
+      "step": 11680
+    },
+    {
+      "epoch": 0.8710879284649776,
+      "grad_norm": 3.5715925693511963,
+      "learning_rate": 0.0002,
+      "loss": 2.6196,
+      "step": 11690
+    },
+    {
+      "epoch": 0.8718330849478391,
+      "grad_norm": 2.3358867168426514,
+      "learning_rate": 0.0002,
+      "loss": 2.6385,
+      "step": 11700
+    },
+    {
+      "epoch": 0.8725782414307004,
+      "grad_norm": 2.237659454345703,
+      "learning_rate": 0.0002,
+      "loss": 2.5753,
+      "step": 11710
+    },
+    {
+      "epoch": 0.8733233979135618,
+      "grad_norm": 2.2841837406158447,
+      "learning_rate": 0.0002,
+      "loss": 2.6415,
+      "step": 11720
+    },
+    {
+      "epoch": 0.8740685543964233,
+      "grad_norm": 2.559757709503174,
+      "learning_rate": 0.0002,
+      "loss": 2.5872,
+      "step": 11730
+    },
+    {
+      "epoch": 0.8748137108792846,
+      "grad_norm": 2.5326175689697266,
+      "learning_rate": 0.0002,
+      "loss": 2.4485,
+      "step": 11740
+    },
+    {
+      "epoch": 0.875558867362146,
+      "grad_norm": 2.234788656234741,
+      "learning_rate": 0.0002,
+      "loss": 2.2943,
+      "step": 11750
+    },
+    {
+      "epoch": 0.8763040238450075,
+      "grad_norm": 2.4031131267547607,
+      "learning_rate": 0.0002,
+      "loss": 2.5176,
+      "step": 11760
+    },
+    {
+      "epoch": 0.8770491803278688,
+      "grad_norm": 2.832005023956299,
+      "learning_rate": 0.0002,
+      "loss": 2.5063,
+      "step": 11770
+    },
+    {
+      "epoch": 0.8777943368107303,
+      "grad_norm": 2.0008394718170166,
+      "learning_rate": 0.0002,
+      "loss": 2.5267,
+      "step": 11780
+    },
+    {
+      "epoch": 0.8785394932935916,
+      "grad_norm": 2.0909340381622314,
+      "learning_rate": 0.0002,
+      "loss": 2.5573,
+      "step": 11790
+    },
+    {
+      "epoch": 0.879284649776453,
+      "grad_norm": 2.437964677810669,
+      "learning_rate": 0.0002,
+      "loss": 2.6744,
+      "step": 11800
+    },
+    {
+      "epoch": 0.8800298062593145,
+      "grad_norm": 2.1106114387512207,
+      "learning_rate": 0.0002,
+      "loss": 2.5049,
+      "step": 11810
+    },
+    {
+      "epoch": 0.8807749627421758,
+      "grad_norm": 2.672287940979004,
+      "learning_rate": 0.0002,
+      "loss": 2.7222,
+      "step": 11820
+    },
+    {
+      "epoch": 0.8815201192250373,
+      "grad_norm": 2.584033727645874,
+      "learning_rate": 0.0002,
+      "loss": 2.5348,
+      "step": 11830
+    },
+    {
+      "epoch": 0.8822652757078987,
+      "grad_norm": 2.440488576889038,
+      "learning_rate": 0.0002,
+      "loss": 2.4024,
+      "step": 11840
+    },
+    {
+      "epoch": 0.88301043219076,
+      "grad_norm": 2.302765130996704,
+      "learning_rate": 0.0002,
+      "loss": 2.441,
+      "step": 11850
+    },
+    {
+      "epoch": 0.8837555886736215,
+      "grad_norm": 3.5020196437835693,
+      "learning_rate": 0.0002,
+      "loss": 2.5192,
+      "step": 11860
+    },
+    {
+      "epoch": 0.8845007451564829,
+      "grad_norm": 1.8765350580215454,
+      "learning_rate": 0.0002,
+      "loss": 2.5156,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8852459016393442,
+      "grad_norm": 2.4863150119781494,
+      "learning_rate": 0.0002,
+      "loss": 2.5239,
+      "step": 11880
+    },
+    {
+      "epoch": 0.8859910581222057,
+      "grad_norm": 2.492176055908203,
+      "learning_rate": 0.0002,
+      "loss": 2.5697,
+      "step": 11890
+    },
+    {
+      "epoch": 0.886736214605067,
+      "grad_norm": 3.2195818424224854,
+      "learning_rate": 0.0002,
+      "loss": 2.6402,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8874813710879285,
+      "grad_norm": 2.2776970863342285,
+      "learning_rate": 0.0002,
+      "loss": 2.4884,
+      "step": 11910
+    },
+    {
+      "epoch": 0.8882265275707899,
+      "grad_norm": 2.553642749786377,
+      "learning_rate": 0.0002,
+      "loss": 2.6425,
+      "step": 11920
+    },
+    {
+      "epoch": 0.8889716840536512,
+      "grad_norm": 2.269604206085205,
+      "learning_rate": 0.0002,
+      "loss": 2.5291,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8897168405365127,
+      "grad_norm": 2.3157992362976074,
+      "learning_rate": 0.0002,
+      "loss": 2.7461,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8904619970193741,
+      "grad_norm": 2.3829448223114014,
+      "learning_rate": 0.0002,
+      "loss": 2.6959,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8912071535022354,
+      "grad_norm": 2.368562698364258,
+      "learning_rate": 0.0002,
+      "loss": 2.3614,
+      "step": 11960
+    },
+    {
+      "epoch": 0.8919523099850969,
+      "grad_norm": 2.5107600688934326,
+      "learning_rate": 0.0002,
+      "loss": 2.5187,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8926974664679582,
+      "grad_norm": 2.376817226409912,
+      "learning_rate": 0.0002,
+      "loss": 2.5952,
+      "step": 11980
+    },
+    {
+      "epoch": 0.8934426229508197,
+      "grad_norm": 2.224245071411133,
+      "learning_rate": 0.0002,
+      "loss": 2.4105,
+      "step": 11990
+    },
+    {
+      "epoch": 0.8941877794336811,
+      "grad_norm": 2.118123769760132,
+      "learning_rate": 0.0002,
+      "loss": 2.442,
+      "step": 12000
+    },
+    {
+      "epoch": 0.8949329359165424,
+      "grad_norm": 2.6127233505249023,
+      "learning_rate": 0.0002,
+      "loss": 2.3396,
+      "step": 12010
+    },
+    {
+      "epoch": 0.8956780923994039,
+      "grad_norm": 2.319699287414551,
+      "learning_rate": 0.0002,
+      "loss": 2.5345,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8964232488822653,
+      "grad_norm": 2.394341468811035,
+      "learning_rate": 0.0002,
+      "loss": 2.6386,
+      "step": 12030
+    },
+    {
+      "epoch": 0.8971684053651267,
+      "grad_norm": 1.9906131029129028,
+      "learning_rate": 0.0002,
+      "loss": 2.3783,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8979135618479881,
+      "grad_norm": 2.038069009780884,
+      "learning_rate": 0.0002,
+      "loss": 2.4852,
+      "step": 12050
+    },
+    {
+      "epoch": 0.8986587183308494,
+      "grad_norm": 2.18753981590271,
+      "learning_rate": 0.0002,
+      "loss": 2.3601,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8994038748137109,
+      "grad_norm": 2.302685499191284,
+      "learning_rate": 0.0002,
+      "loss": 2.4762,
+      "step": 12070
+    },
+    {
+      "epoch": 0.9001490312965723,
+      "grad_norm": 2.177448272705078,
+      "learning_rate": 0.0002,
+      "loss": 2.3813,
+      "step": 12080
+    },
+    {
+      "epoch": 0.9008941877794336,
+      "grad_norm": 2.6771061420440674,
+      "learning_rate": 0.0002,
+      "loss": 2.5623,
+      "step": 12090
+    },
+    {
+      "epoch": 0.9016393442622951,
+      "grad_norm": 2.364640712738037,
+      "learning_rate": 0.0002,
+      "loss": 2.4862,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9023845007451565,
+      "grad_norm": 2.2396554946899414,
+      "learning_rate": 0.0002,
+      "loss": 2.3349,
+      "step": 12110
+    },
+    {
+      "epoch": 0.9031296572280179,
+      "grad_norm": 2.5424320697784424,
+      "learning_rate": 0.0002,
+      "loss": 2.5552,
+      "step": 12120
+    },
+    {
+      "epoch": 0.9038748137108793,
+      "grad_norm": 2.2599573135375977,
+      "learning_rate": 0.0002,
+      "loss": 2.4783,
+      "step": 12130
+    },
+    {
+      "epoch": 0.9046199701937406,
+      "grad_norm": 2.3719987869262695,
+      "learning_rate": 0.0002,
+      "loss": 2.3134,
+      "step": 12140
+    },
+    {
+      "epoch": 0.9053651266766021,
+      "grad_norm": 2.3063857555389404,
+      "learning_rate": 0.0002,
+      "loss": 2.375,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9061102831594635,
+      "grad_norm": 2.3336477279663086,
+      "learning_rate": 0.0002,
+      "loss": 2.5508,
+      "step": 12160
+    },
+    {
+      "epoch": 0.9068554396423248,
+      "grad_norm": 2.325166940689087,
+      "learning_rate": 0.0002,
+      "loss": 2.5011,
+      "step": 12170
+    },
+    {
+      "epoch": 0.9076005961251863,
+      "grad_norm": 2.2718868255615234,
+      "learning_rate": 0.0002,
+      "loss": 2.6404,
+      "step": 12180
+    },
+    {
+      "epoch": 0.9083457526080477,
+      "grad_norm": 2.607816696166992,
+      "learning_rate": 0.0002,
+      "loss": 2.5262,
+      "step": 12190
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 2.1004855632781982,
+      "learning_rate": 0.0002,
+      "loss": 2.5023,
+      "step": 12200
+    },
+    {
+      "epoch": 0.9098360655737705,
+      "grad_norm": 2.336148977279663,
+      "learning_rate": 0.0002,
+      "loss": 2.6659,
+      "step": 12210
+    },
+    {
+      "epoch": 0.910581222056632,
+      "grad_norm": 2.531151294708252,
+      "learning_rate": 0.0002,
+      "loss": 2.7188,
+      "step": 12220
+    },
+    {
+      "epoch": 0.9113263785394933,
+      "grad_norm": 2.559283494949341,
+      "learning_rate": 0.0002,
+      "loss": 2.6394,
+      "step": 12230
+    },
+    {
+      "epoch": 0.9120715350223547,
+      "grad_norm": 2.272649049758911,
+      "learning_rate": 0.0002,
+      "loss": 2.6711,
+      "step": 12240
+    },
+    {
+      "epoch": 0.912816691505216,
+      "grad_norm": 2.356740713119507,
+      "learning_rate": 0.0002,
+      "loss": 2.6009,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9135618479880775,
+      "grad_norm": 2.4409749507904053,
+      "learning_rate": 0.0002,
+      "loss": 2.6274,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9143070044709389,
+      "grad_norm": 2.7427172660827637,
+      "learning_rate": 0.0002,
+      "loss": 2.7499,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9150521609538003,
+      "grad_norm": 2.309023380279541,
+      "learning_rate": 0.0002,
+      "loss": 2.3997,
+      "step": 12280
+    },
+    {
+      "epoch": 0.9157973174366617,
+      "grad_norm": 2.1871066093444824,
+      "learning_rate": 0.0002,
+      "loss": 2.6141,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9165424739195231,
+      "grad_norm": 2.556929588317871,
+      "learning_rate": 0.0002,
+      "loss": 2.4421,
+      "step": 12300
+    },
+    {
+      "epoch": 0.9172876304023845,
+      "grad_norm": 2.5296781063079834,
+      "learning_rate": 0.0002,
+      "loss": 2.5385,
+      "step": 12310
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 2.4364492893218994,
+      "learning_rate": 0.0002,
+      "loss": 2.5862,
+      "step": 12320
+    },
+    {
+      "epoch": 0.9187779433681073,
+      "grad_norm": 2.1894443035125732,
+      "learning_rate": 0.0002,
+      "loss": 2.5224,
+      "step": 12330
+    },
+    {
+      "epoch": 0.9195230998509687,
+      "grad_norm": 2.3544352054595947,
+      "learning_rate": 0.0002,
+      "loss": 2.4424,
+      "step": 12340
+    },
+    {
+      "epoch": 0.9202682563338301,
+      "grad_norm": 2.2780649662017822,
+      "learning_rate": 0.0002,
+      "loss": 2.4522,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9210134128166915,
+      "grad_norm": 2.3933236598968506,
+      "learning_rate": 0.0002,
+      "loss": 2.4272,
+      "step": 12360
+    },
+    {
+      "epoch": 0.9217585692995529,
+      "grad_norm": 2.260446071624756,
+      "learning_rate": 0.0002,
+      "loss": 2.4688,
+      "step": 12370
+    },
+    {
+      "epoch": 0.9225037257824144,
+      "grad_norm": 2.4620795249938965,
+      "learning_rate": 0.0002,
+      "loss": 2.1951,
+      "step": 12380
+    },
+    {
+      "epoch": 0.9232488822652757,
+      "grad_norm": 2.4303131103515625,
+      "learning_rate": 0.0002,
+      "loss": 2.2946,
+      "step": 12390
+    },
+    {
+      "epoch": 0.9239940387481371,
+      "grad_norm": 2.8419077396392822,
+      "learning_rate": 0.0002,
+      "loss": 2.4692,
+      "step": 12400
+    },
+    {
+      "epoch": 0.9247391952309985,
+      "grad_norm": 2.9645135402679443,
+      "learning_rate": 0.0002,
+      "loss": 2.6461,
+      "step": 12410
+    },
+    {
+      "epoch": 0.9254843517138599,
+      "grad_norm": 2.391739845275879,
+      "learning_rate": 0.0002,
+      "loss": 2.5335,
+      "step": 12420
+    },
+    {
+      "epoch": 0.9262295081967213,
+      "grad_norm": 2.686976194381714,
+      "learning_rate": 0.0002,
+      "loss": 2.6333,
+      "step": 12430
+    },
+    {
+      "epoch": 0.9269746646795827,
+      "grad_norm": 2.723388195037842,
+      "learning_rate": 0.0002,
+      "loss": 2.354,
+      "step": 12440
+    },
+    {
+      "epoch": 0.9277198211624441,
+      "grad_norm": 2.4658613204956055,
+      "learning_rate": 0.0002,
+      "loss": 2.5475,
+      "step": 12450
+    },
+    {
+      "epoch": 0.9284649776453056,
+      "grad_norm": 2.6882481575012207,
+      "learning_rate": 0.0002,
+      "loss": 2.3591,
+      "step": 12460
+    },
+    {
+      "epoch": 0.9292101341281669,
+      "grad_norm": 2.646805763244629,
+      "learning_rate": 0.0002,
+      "loss": 2.4765,
+      "step": 12470
+    },
+    {
+      "epoch": 0.9299552906110283,
+      "grad_norm": 2.4550538063049316,
+      "learning_rate": 0.0002,
+      "loss": 2.6768,
+      "step": 12480
+    },
+    {
+      "epoch": 0.9307004470938898,
+      "grad_norm": 2.296886920928955,
+      "learning_rate": 0.0002,
+      "loss": 2.6753,
+      "step": 12490
+    },
+    {
+      "epoch": 0.9314456035767511,
+      "grad_norm": 2.1057779788970947,
+      "learning_rate": 0.0002,
+      "loss": 2.5129,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9321907600596125,
+      "grad_norm": 2.1545591354370117,
+      "learning_rate": 0.0002,
+      "loss": 2.4823,
+      "step": 12510
+    },
+    {
+      "epoch": 0.9329359165424739,
+      "grad_norm": 2.354118824005127,
+      "learning_rate": 0.0002,
+      "loss": 2.6566,
+      "step": 12520
+    },
+    {
+      "epoch": 0.9336810730253353,
+      "grad_norm": 2.294067859649658,
+      "learning_rate": 0.0002,
+      "loss": 2.5172,
+      "step": 12530
+    },
+    {
+      "epoch": 0.9344262295081968,
+      "grad_norm": 2.2037270069122314,
+      "learning_rate": 0.0002,
+      "loss": 2.5489,
+      "step": 12540
+    },
+    {
+      "epoch": 0.9351713859910581,
+      "grad_norm": 2.276435136795044,
+      "learning_rate": 0.0002,
+      "loss": 2.3769,
+      "step": 12550
+    },
+    {
+      "epoch": 0.9359165424739195,
+      "grad_norm": 2.4285993576049805,
+      "learning_rate": 0.0002,
+      "loss": 2.7328,
+      "step": 12560
+    },
+    {
+      "epoch": 0.936661698956781,
+      "grad_norm": 2.661801338195801,
+      "learning_rate": 0.0002,
+      "loss": 2.5124,
+      "step": 12570
+    },
+    {
+      "epoch": 0.9374068554396423,
+      "grad_norm": 1.9602450132369995,
+      "learning_rate": 0.0002,
+      "loss": 2.4483,
+      "step": 12580
+    },
+    {
+      "epoch": 0.9381520119225037,
+      "grad_norm": 2.1745917797088623,
+      "learning_rate": 0.0002,
+      "loss": 2.4838,
+      "step": 12590
+    },
+    {
+      "epoch": 0.9388971684053651,
+      "grad_norm": 2.373079776763916,
+      "learning_rate": 0.0002,
+      "loss": 2.6714,
+      "step": 12600
+    },
+    {
+      "epoch": 0.9396423248882265,
+      "grad_norm": 2.567545175552368,
+      "learning_rate": 0.0002,
+      "loss": 2.6517,
+      "step": 12610
+    },
+    {
+      "epoch": 0.940387481371088,
+      "grad_norm": 2.426687240600586,
+      "learning_rate": 0.0002,
+      "loss": 2.698,
+      "step": 12620
+    },
+    {
+      "epoch": 0.9411326378539493,
+      "grad_norm": 2.2519803047180176,
+      "learning_rate": 0.0002,
+      "loss": 2.3313,
+      "step": 12630
+    },
+    {
+      "epoch": 0.9418777943368107,
+      "grad_norm": 2.221134662628174,
+      "learning_rate": 0.0002,
+      "loss": 2.4458,
+      "step": 12640
+    },
+    {
+      "epoch": 0.9426229508196722,
+      "grad_norm": 2.3901071548461914,
+      "learning_rate": 0.0002,
+      "loss": 2.6864,
+      "step": 12650
+    },
+    {
+      "epoch": 0.9433681073025335,
+      "grad_norm": 2.3478126525878906,
+      "learning_rate": 0.0002,
+      "loss": 2.5446,
+      "step": 12660
+    },
+    {
+      "epoch": 0.944113263785395,
+      "grad_norm": 2.271385908126831,
+      "learning_rate": 0.0002,
+      "loss": 2.6798,
+      "step": 12670
+    },
+    {
+      "epoch": 0.9448584202682563,
+      "grad_norm": 2.620520830154419,
+      "learning_rate": 0.0002,
+      "loss": 2.5766,
+      "step": 12680
+    },
+    {
+      "epoch": 0.9456035767511177,
+      "grad_norm": 2.460322380065918,
+      "learning_rate": 0.0002,
+      "loss": 2.3643,
+      "step": 12690
+    },
+    {
+      "epoch": 0.9463487332339792,
+      "grad_norm": 2.526590585708618,
+      "learning_rate": 0.0002,
+      "loss": 2.4779,
+      "step": 12700
+    },
+    {
+      "epoch": 0.9470938897168405,
+      "grad_norm": 2.0752880573272705,
+      "learning_rate": 0.0002,
+      "loss": 2.4393,
+      "step": 12710
+    },
+    {
+      "epoch": 0.9478390461997019,
+      "grad_norm": 2.8343307971954346,
+      "learning_rate": 0.0002,
+      "loss": 2.5893,
+      "step": 12720
+    },
+    {
+      "epoch": 0.9485842026825634,
+      "grad_norm": 2.369196653366089,
+      "learning_rate": 0.0002,
+      "loss": 2.5725,
+      "step": 12730
+    },
+    {
+      "epoch": 0.9493293591654247,
+      "grad_norm": 2.3680763244628906,
+      "learning_rate": 0.0002,
+      "loss": 2.3934,
+      "step": 12740
+    },
+    {
+      "epoch": 0.9500745156482862,
+      "grad_norm": 2.43011736869812,
+      "learning_rate": 0.0002,
+      "loss": 2.1303,
+      "step": 12750
+    },
+    {
+      "epoch": 0.9508196721311475,
+      "grad_norm": 2.4753246307373047,
+      "learning_rate": 0.0002,
+      "loss": 2.6349,
+      "step": 12760
+    },
+    {
+      "epoch": 0.9515648286140089,
+      "grad_norm": 2.558833122253418,
+      "learning_rate": 0.0002,
+      "loss": 2.523,
+      "step": 12770
+    },
+    {
+      "epoch": 0.9523099850968704,
+      "grad_norm": 2.847932815551758,
+      "learning_rate": 0.0002,
+      "loss": 2.4538,
+      "step": 12780
+    },
+    {
+      "epoch": 0.9530551415797317,
+      "grad_norm": 2.3892436027526855,
+      "learning_rate": 0.0002,
+      "loss": 2.3756,
+      "step": 12790
+    },
+    {
+      "epoch": 0.9538002980625931,
+      "grad_norm": 2.214956760406494,
+      "learning_rate": 0.0002,
+      "loss": 2.4268,
+      "step": 12800
+    },
+    {
+      "epoch": 0.9545454545454546,
+      "grad_norm": 2.3451969623565674,
+      "learning_rate": 0.0002,
+      "loss": 2.423,
+      "step": 12810
+    },
+    {
+      "epoch": 0.9552906110283159,
+      "grad_norm": 2.4705963134765625,
+      "learning_rate": 0.0002,
+      "loss": 2.6488,
+      "step": 12820
+    },
+    {
+      "epoch": 0.9560357675111774,
+      "grad_norm": 2.7425971031188965,
+      "learning_rate": 0.0002,
+      "loss": 2.5578,
+      "step": 12830
+    },
+    {
+      "epoch": 0.9567809239940388,
+      "grad_norm": 2.3511667251586914,
+      "learning_rate": 0.0002,
+      "loss": 2.3301,
+      "step": 12840
+    },
+    {
+      "epoch": 0.9575260804769001,
+      "grad_norm": 2.307358980178833,
+      "learning_rate": 0.0002,
+      "loss": 2.4876,
+      "step": 12850
+    },
+    {
+      "epoch": 0.9582712369597616,
+      "grad_norm": 2.233081102371216,
+      "learning_rate": 0.0002,
+      "loss": 2.2724,
+      "step": 12860
+    },
+    {
+      "epoch": 0.9590163934426229,
+      "grad_norm": 2.7037274837493896,
+      "learning_rate": 0.0002,
+      "loss": 2.3392,
+      "step": 12870
+    },
+    {
+      "epoch": 0.9597615499254843,
+      "grad_norm": 2.435098886489868,
+      "learning_rate": 0.0002,
+      "loss": 2.5374,
+      "step": 12880
+    },
+    {
+      "epoch": 0.9605067064083458,
+      "grad_norm": 2.4666409492492676,
+      "learning_rate": 0.0002,
+      "loss": 2.5573,
+      "step": 12890
+    },
+    {
+      "epoch": 0.9612518628912071,
+      "grad_norm": 2.260859251022339,
+      "learning_rate": 0.0002,
+      "loss": 2.4869,
+      "step": 12900
+    },
+    {
+      "epoch": 0.9619970193740686,
+      "grad_norm": 2.661461353302002,
+      "learning_rate": 0.0002,
+      "loss": 2.5166,
+      "step": 12910
+    },
+    {
+      "epoch": 0.96274217585693,
+      "grad_norm": 2.374238967895508,
+      "learning_rate": 0.0002,
+      "loss": 2.5812,
+      "step": 12920
+    },
+    {
+      "epoch": 0.9634873323397913,
+      "grad_norm": 2.1044247150421143,
+      "learning_rate": 0.0002,
+      "loss": 2.5499,
+      "step": 12930
+    },
+    {
+      "epoch": 0.9642324888226528,
+      "grad_norm": 2.3307137489318848,
+      "learning_rate": 0.0002,
+      "loss": 2.6675,
+      "step": 12940
+    },
+    {
+      "epoch": 0.9649776453055141,
+      "grad_norm": 2.323369264602661,
+      "learning_rate": 0.0002,
+      "loss": 2.6393,
+      "step": 12950
+    },
+    {
+      "epoch": 0.9657228017883756,
+      "grad_norm": 2.198737621307373,
+      "learning_rate": 0.0002,
+      "loss": 2.659,
+      "step": 12960
+    },
+    {
+      "epoch": 0.966467958271237,
+      "grad_norm": 2.636990547180176,
+      "learning_rate": 0.0002,
+      "loss": 2.3037,
+      "step": 12970
+    },
+    {
+      "epoch": 0.9672131147540983,
+      "grad_norm": 2.4657247066497803,
+      "learning_rate": 0.0002,
+      "loss": 2.4913,
+      "step": 12980
+    },
+    {
+      "epoch": 0.9679582712369598,
+      "grad_norm": 2.854710340499878,
+      "learning_rate": 0.0002,
+      "loss": 2.6958,
+      "step": 12990
+    },
+    {
+      "epoch": 0.9687034277198212,
+      "grad_norm": 1.9861502647399902,
+      "learning_rate": 0.0002,
+      "loss": 2.4177,
+      "step": 13000
+    },
+    {
+      "epoch": 0.9694485842026825,
+      "grad_norm": 2.2618319988250732,
+      "learning_rate": 0.0002,
+      "loss": 2.6304,
+      "step": 13010
+    },
+    {
+      "epoch": 0.970193740685544,
+      "grad_norm": 2.253206491470337,
+      "learning_rate": 0.0002,
+      "loss": 2.539,
+      "step": 13020
+    },
+    {
+      "epoch": 0.9709388971684053,
+      "grad_norm": 2.4439284801483154,
+      "learning_rate": 0.0002,
+      "loss": 2.6837,
+      "step": 13030
+    },
+    {
+      "epoch": 0.9716840536512668,
+      "grad_norm": 2.5577943325042725,
+      "learning_rate": 0.0002,
+      "loss": 2.6352,
+      "step": 13040
+    },
+    {
+      "epoch": 0.9724292101341282,
+      "grad_norm": 2.345137596130371,
+      "learning_rate": 0.0002,
+      "loss": 2.5917,
+      "step": 13050
+    },
+    {
+      "epoch": 0.9731743666169895,
+      "grad_norm": 2.3082261085510254,
+      "learning_rate": 0.0002,
+      "loss": 2.484,
+      "step": 13060
+    },
+    {
+      "epoch": 0.973919523099851,
+      "grad_norm": 2.542104721069336,
+      "learning_rate": 0.0002,
+      "loss": 2.5961,
+      "step": 13070
+    },
+    {
+      "epoch": 0.9746646795827124,
+      "grad_norm": 2.420619487762451,
+      "learning_rate": 0.0002,
+      "loss": 2.6367,
+      "step": 13080
+    },
+    {
+      "epoch": 0.9754098360655737,
+      "grad_norm": 2.230170488357544,
+      "learning_rate": 0.0002,
+      "loss": 2.4443,
+      "step": 13090
+    },
+    {
+      "epoch": 0.9761549925484352,
+      "grad_norm": 2.4202942848205566,
+      "learning_rate": 0.0002,
+      "loss": 2.4689,
+      "step": 13100
+    },
+    {
+      "epoch": 0.9769001490312966,
+      "grad_norm": 2.6881895065307617,
+      "learning_rate": 0.0002,
+      "loss": 2.4001,
+      "step": 13110
+    },
+    {
+      "epoch": 0.977645305514158,
+      "grad_norm": 1.8320391178131104,
+      "learning_rate": 0.0002,
+      "loss": 2.545,
+      "step": 13120
+    },
+    {
+      "epoch": 0.9783904619970194,
+      "grad_norm": 2.4107582569122314,
+      "learning_rate": 0.0002,
+      "loss": 2.5337,
+      "step": 13130
+    },
+    {
+      "epoch": 0.9791356184798807,
+      "grad_norm": 2.323265790939331,
+      "learning_rate": 0.0002,
+      "loss": 2.57,
+      "step": 13140
+    },
+    {
+      "epoch": 0.9798807749627422,
+      "grad_norm": 2.5477750301361084,
+      "learning_rate": 0.0002,
+      "loss": 2.5769,
+      "step": 13150
+    },
+    {
+      "epoch": 0.9806259314456036,
+      "grad_norm": 3.1621835231781006,
+      "learning_rate": 0.0002,
+      "loss": 2.4171,
+      "step": 13160
+    },
+    {
+      "epoch": 0.981371087928465,
+      "grad_norm": 2.4050159454345703,
+      "learning_rate": 0.0002,
+      "loss": 2.5969,
+      "step": 13170
+    },
+    {
+      "epoch": 0.9821162444113264,
+      "grad_norm": 2.261742353439331,
+      "learning_rate": 0.0002,
+      "loss": 2.5855,
+      "step": 13180
+    },
+    {
+      "epoch": 0.9828614008941878,
+      "grad_norm": 2.472062349319458,
+      "learning_rate": 0.0002,
+      "loss": 2.6003,
+      "step": 13190
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 2.328455924987793,
+      "learning_rate": 0.0002,
+      "loss": 2.6398,
+      "step": 13200
+    },
+    {
+      "epoch": 0.9843517138599106,
+      "grad_norm": 2.5007076263427734,
+      "learning_rate": 0.0002,
+      "loss": 2.6027,
+      "step": 13210
+    },
+    {
+      "epoch": 0.9850968703427719,
+      "grad_norm": 2.1729538440704346,
+      "learning_rate": 0.0002,
+      "loss": 2.5315,
+      "step": 13220
+    },
+    {
+      "epoch": 0.9858420268256334,
+      "grad_norm": 2.4826149940490723,
+      "learning_rate": 0.0002,
+      "loss": 2.5213,
+      "step": 13230
+    },
+    {
+      "epoch": 0.9865871833084948,
+      "grad_norm": 2.1284587383270264,
+      "learning_rate": 0.0002,
+      "loss": 2.3732,
+      "step": 13240
+    },
+    {
+      "epoch": 0.9873323397913562,
+      "grad_norm": 2.260220766067505,
+      "learning_rate": 0.0002,
+      "loss": 2.5361,
+      "step": 13250
+    },
+    {
+      "epoch": 0.9880774962742176,
+      "grad_norm": 2.427016019821167,
+      "learning_rate": 0.0002,
+      "loss": 2.7074,
+      "step": 13260
+    },
+    {
+      "epoch": 0.988822652757079,
+      "grad_norm": 2.6063551902770996,
+      "learning_rate": 0.0002,
+      "loss": 2.7052,
+      "step": 13270
+    },
+    {
+      "epoch": 0.9895678092399404,
+      "grad_norm": 2.5883805751800537,
+      "learning_rate": 0.0002,
+      "loss": 2.6739,
+      "step": 13280
+    },
+    {
+      "epoch": 0.9903129657228018,
+      "grad_norm": 2.7127444744110107,
+      "learning_rate": 0.0002,
+      "loss": 2.5796,
+      "step": 13290
+    },
+    {
+      "epoch": 0.9910581222056631,
+      "grad_norm": 2.2932395935058594,
+      "learning_rate": 0.0002,
+      "loss": 2.6266,
+      "step": 13300
+    },
+    {
+      "epoch": 0.9918032786885246,
+      "grad_norm": 2.1839466094970703,
+      "learning_rate": 0.0002,
+      "loss": 2.4277,
+      "step": 13310
+    },
+    {
+      "epoch": 0.992548435171386,
+      "grad_norm": 2.1350860595703125,
+      "learning_rate": 0.0002,
+      "loss": 2.4021,
+      "step": 13320
+    },
+    {
+      "epoch": 0.9932935916542474,
+      "grad_norm": 2.821213960647583,
+      "learning_rate": 0.0002,
+      "loss": 2.4104,
+      "step": 13330
+    },
+    {
+      "epoch": 0.9940387481371088,
+      "grad_norm": 2.3386242389678955,
+      "learning_rate": 0.0002,
+      "loss": 2.6611,
+      "step": 13340
+    },
+    {
+      "epoch": 0.9947839046199702,
+      "grad_norm": 2.477780342102051,
+      "learning_rate": 0.0002,
+      "loss": 2.5641,
+      "step": 13350
+    },
+    {
+      "epoch": 0.9955290611028316,
+      "grad_norm": 2.1101603507995605,
+      "learning_rate": 0.0002,
+      "loss": 2.3935,
+      "step": 13360
+    },
+    {
+      "epoch": 0.996274217585693,
+      "grad_norm": 2.371098041534424,
+      "learning_rate": 0.0002,
+      "loss": 2.4277,
+      "step": 13370
+    },
+    {
+      "epoch": 0.9970193740685543,
+      "grad_norm": 2.4047231674194336,
+      "learning_rate": 0.0002,
+      "loss": 2.5452,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9977645305514158,
+      "grad_norm": 2.2400450706481934,
+      "learning_rate": 0.0002,
+      "loss": 2.5556,
+      "step": 13390
+    },
+    {
+      "epoch": 0.9985096870342772,
+      "grad_norm": 2.187760829925537,
+      "learning_rate": 0.0002,
+      "loss": 2.645,
+      "step": 13400
+    },
+    {
+      "epoch": 0.9992548435171386,
+      "grad_norm": 2.3010387420654297,
+      "learning_rate": 0.0002,
+      "loss": 2.4113,
+      "step": 13410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8511722087860107,
+      "learning_rate": 0.0002,
+      "loss": 2.5681,
+      "step": 13420
+    },
+    {
+      "epoch": 1.0,
+      "eval_runtime": 2766.9434,
+      "eval_samples_per_second": 4.85,
+      "eval_steps_per_second": 0.606,
+      "step": 13420
+    },
+    {
+      "epoch": 1.0007451564828613,
+      "grad_norm": 2.541956901550293,
+      "learning_rate": 0.0002,
+      "loss": 2.4417,
+      "step": 13430
+    },
+    {
+      "epoch": 1.0014903129657229,
+      "grad_norm": 2.2113637924194336,
+      "learning_rate": 0.0002,
+      "loss": 2.3728,
+      "step": 13440
+    },
+    {
+      "epoch": 1.0022354694485842,
+      "grad_norm": 2.6709325313568115,
+      "learning_rate": 0.0002,
+      "loss": 2.3259,
+      "step": 13450
+    },
+    {
+      "epoch": 1.0029806259314455,
+      "grad_norm": 2.1708545684814453,
+      "learning_rate": 0.0002,
+      "loss": 2.164,
+      "step": 13460
+    },
+    {
+      "epoch": 1.003725782414307,
+      "grad_norm": 2.2694172859191895,
+      "learning_rate": 0.0002,
+      "loss": 2.37,
+      "step": 13470
+    },
+    {
+      "epoch": 1.0044709388971684,
+      "grad_norm": 2.839197874069214,
+      "learning_rate": 0.0002,
+      "loss": 2.383,
+      "step": 13480
+    },
+    {
+      "epoch": 1.0052160953800298,
+      "grad_norm": 2.7437849044799805,
+      "learning_rate": 0.0002,
+      "loss": 2.3811,
+      "step": 13490
+    },
+    {
+      "epoch": 1.0059612518628913,
+      "grad_norm": 2.8862855434417725,
+      "learning_rate": 0.0002,
+      "loss": 2.2016,
+      "step": 13500
+    },
+    {
+      "epoch": 1.0067064083457526,
+      "grad_norm": 2.6776318550109863,
+      "learning_rate": 0.0002,
+      "loss": 2.2084,
+      "step": 13510
+    },
+    {
+      "epoch": 1.007451564828614,
+      "grad_norm": 2.473053455352783,
+      "learning_rate": 0.0002,
+      "loss": 2.4357,
+      "step": 13520
+    },
+    {
+      "epoch": 1.0081967213114753,
+      "grad_norm": 2.5545501708984375,
+      "learning_rate": 0.0002,
+      "loss": 2.4439,
+      "step": 13530
+    },
+    {
+      "epoch": 1.0089418777943369,
+      "grad_norm": 2.6826674938201904,
+      "learning_rate": 0.0002,
+      "loss": 2.5071,
+      "step": 13540
+    },
+    {
+      "epoch": 1.0096870342771982,
+      "grad_norm": 2.596944570541382,
+      "learning_rate": 0.0002,
+      "loss": 2.5319,
+      "step": 13550
+    },
+    {
+      "epoch": 1.0104321907600595,
+      "grad_norm": 2.0925228595733643,
+      "learning_rate": 0.0002,
+      "loss": 2.1662,
+      "step": 13560
+    },
+    {
+      "epoch": 1.011177347242921,
+      "grad_norm": 2.8665518760681152,
+      "learning_rate": 0.0002,
+      "loss": 2.5069,
+      "step": 13570
+    },
+    {
+      "epoch": 1.0119225037257824,
+      "grad_norm": 2.4106383323669434,
+      "learning_rate": 0.0002,
+      "loss": 2.3902,
+      "step": 13580
+    },
+    {
+      "epoch": 1.0126676602086437,
+      "grad_norm": 2.5520753860473633,
+      "learning_rate": 0.0002,
+      "loss": 2.4638,
+      "step": 13590
+    },
+    {
+      "epoch": 1.0134128166915053,
+      "grad_norm": 2.505903482437134,
+      "learning_rate": 0.0002,
+      "loss": 2.5054,
+      "step": 13600
+    },
+    {
+      "epoch": 1.0141579731743666,
+      "grad_norm": 2.5196523666381836,
+      "learning_rate": 0.0002,
+      "loss": 2.1964,
+      "step": 13610
+    },
+    {
+      "epoch": 1.014903129657228,
+      "grad_norm": 2.414044141769409,
+      "learning_rate": 0.0002,
+      "loss": 2.2939,
+      "step": 13620
+    },
+    {
+      "epoch": 1.0156482861400895,
+      "grad_norm": 2.586638927459717,
+      "learning_rate": 0.0002,
+      "loss": 2.403,
+      "step": 13630
+    },
+    {
+      "epoch": 1.0163934426229508,
+      "grad_norm": 2.509230852127075,
+      "learning_rate": 0.0002,
+      "loss": 2.3131,
+      "step": 13640
+    },
+    {
+      "epoch": 1.0171385991058122,
+      "grad_norm": 2.8576319217681885,
+      "learning_rate": 0.0002,
+      "loss": 2.3398,
+      "step": 13650
+    },
+    {
+      "epoch": 1.0178837555886737,
+      "grad_norm": 3.0436184406280518,
+      "learning_rate": 0.0002,
+      "loss": 2.4026,
+      "step": 13660
+    },
+    {
+      "epoch": 1.018628912071535,
+      "grad_norm": 2.6709108352661133,
+      "learning_rate": 0.0002,
+      "loss": 2.3594,
+      "step": 13670
+    },
+    {
+      "epoch": 1.0193740685543964,
+      "grad_norm": 2.6364598274230957,
+      "learning_rate": 0.0002,
+      "loss": 2.3834,
+      "step": 13680
+    },
+    {
+      "epoch": 1.0201192250372577,
+      "grad_norm": 2.666531562805176,
+      "learning_rate": 0.0002,
+      "loss": 2.4294,
+      "step": 13690
+    },
+    {
+      "epoch": 1.0208643815201193,
+      "grad_norm": 2.945071220397949,
+      "learning_rate": 0.0002,
+      "loss": 2.4322,
+      "step": 13700
+    },
+    {
+      "epoch": 1.0216095380029806,
+      "grad_norm": 2.6446971893310547,
+      "learning_rate": 0.0002,
+      "loss": 2.3047,
+      "step": 13710
+    },
+    {
+      "epoch": 1.022354694485842,
+      "grad_norm": 2.769773483276367,
+      "learning_rate": 0.0002,
+      "loss": 2.328,
+      "step": 13720
+    },
+    {
+      "epoch": 1.0230998509687035,
+      "grad_norm": 2.7484028339385986,
+      "learning_rate": 0.0002,
+      "loss": 2.4968,
+      "step": 13730
+    },
+    {
+      "epoch": 1.0238450074515648,
+      "grad_norm": 2.563288927078247,
+      "learning_rate": 0.0002,
+      "loss": 2.3943,
+      "step": 13740
+    },
+    {
+      "epoch": 1.0245901639344261,
+      "grad_norm": 2.358903408050537,
+      "learning_rate": 0.0002,
+      "loss": 2.3394,
+      "step": 13750
+    },
+    {
+      "epoch": 1.0253353204172877,
+      "grad_norm": 3.086336851119995,
+      "learning_rate": 0.0002,
+      "loss": 2.4493,
+      "step": 13760
+    },
+    {
+      "epoch": 1.026080476900149,
+      "grad_norm": 2.4959349632263184,
+      "learning_rate": 0.0002,
+      "loss": 2.3651,
+      "step": 13770
+    },
+    {
+      "epoch": 1.0268256333830104,
+      "grad_norm": 2.5488016605377197,
+      "learning_rate": 0.0002,
+      "loss": 2.3654,
+      "step": 13780
+    },
+    {
+      "epoch": 1.027570789865872,
+      "grad_norm": 2.6956136226654053,
+      "learning_rate": 0.0002,
+      "loss": 2.5436,
+      "step": 13790
+    },
+    {
+      "epoch": 1.0283159463487332,
+      "grad_norm": 2.507394790649414,
+      "learning_rate": 0.0002,
+      "loss": 2.393,
+      "step": 13800
+    },
+    {
+      "epoch": 1.0290611028315946,
+      "grad_norm": 2.583897113800049,
+      "learning_rate": 0.0002,
+      "loss": 2.369,
+      "step": 13810
+    },
+    {
+      "epoch": 1.0298062593144561,
+      "grad_norm": 2.804624080657959,
+      "learning_rate": 0.0002,
+      "loss": 2.4578,
+      "step": 13820
+    },
+    {
+      "epoch": 1.0305514157973175,
+      "grad_norm": 2.8353066444396973,
+      "learning_rate": 0.0002,
+      "loss": 2.415,
+      "step": 13830
+    },
+    {
+      "epoch": 1.0312965722801788,
+      "grad_norm": 2.8079729080200195,
+      "learning_rate": 0.0002,
+      "loss": 2.3498,
+      "step": 13840
+    },
+    {
+      "epoch": 1.0320417287630403,
+      "grad_norm": 3.153747320175171,
+      "learning_rate": 0.0002,
+      "loss": 2.4533,
+      "step": 13850
+    },
+    {
+      "epoch": 1.0327868852459017,
+      "grad_norm": 2.170792579650879,
+      "learning_rate": 0.0002,
+      "loss": 2.2922,
+      "step": 13860
+    },
+    {
+      "epoch": 1.033532041728763,
+      "grad_norm": 2.5001676082611084,
+      "learning_rate": 0.0002,
+      "loss": 2.4288,
+      "step": 13870
+    },
+    {
+      "epoch": 1.0342771982116243,
+      "grad_norm": 2.776643753051758,
+      "learning_rate": 0.0002,
+      "loss": 2.3554,
+      "step": 13880
+    },
+    {
+      "epoch": 1.035022354694486,
+      "grad_norm": 2.7371437549591064,
+      "learning_rate": 0.0002,
+      "loss": 2.3946,
+      "step": 13890
+    },
+    {
+      "epoch": 1.0357675111773472,
+      "grad_norm": 2.1921558380126953,
+      "learning_rate": 0.0002,
+      "loss": 2.344,
+      "step": 13900
+    },
+    {
+      "epoch": 1.0365126676602086,
+      "grad_norm": 2.4711101055145264,
+      "learning_rate": 0.0002,
+      "loss": 2.4239,
+      "step": 13910
+    },
+    {
+      "epoch": 1.03725782414307,
+      "grad_norm": 1.9869732856750488,
+      "learning_rate": 0.0002,
+      "loss": 2.3064,
+      "step": 13920
+    },
+    {
+      "epoch": 1.0380029806259314,
+      "grad_norm": 2.50216007232666,
+      "learning_rate": 0.0002,
+      "loss": 2.5143,
+      "step": 13930
+    },
+    {
+      "epoch": 1.0387481371087928,
+      "grad_norm": 2.3530120849609375,
+      "learning_rate": 0.0002,
+      "loss": 2.46,
+      "step": 13940
+    },
+    {
+      "epoch": 1.0394932935916543,
+      "grad_norm": 2.8673856258392334,
+      "learning_rate": 0.0002,
+      "loss": 2.3562,
+      "step": 13950
+    },
+    {
+      "epoch": 1.0402384500745157,
+      "grad_norm": 2.380708932876587,
+      "learning_rate": 0.0002,
+      "loss": 2.4367,
+      "step": 13960
+    },
+    {
+      "epoch": 1.040983606557377,
+      "grad_norm": 2.6121153831481934,
+      "learning_rate": 0.0002,
+      "loss": 2.4539,
+      "step": 13970
+    },
+    {
+      "epoch": 1.0417287630402385,
+      "grad_norm": 2.1682288646698,
+      "learning_rate": 0.0002,
+      "loss": 2.3299,
+      "step": 13980
+    },
+    {
+      "epoch": 1.0424739195230999,
+      "grad_norm": 2.493143081665039,
+      "learning_rate": 0.0002,
+      "loss": 2.2522,
+      "step": 13990
+    },
+    {
+      "epoch": 1.0432190760059612,
+      "grad_norm": 2.3164801597595215,
+      "learning_rate": 0.0002,
+      "loss": 2.3501,
+      "step": 14000
+    },
+    {
+      "epoch": 1.0439642324888228,
+      "grad_norm": 2.372344493865967,
+      "learning_rate": 0.0002,
+      "loss": 2.4671,
+      "step": 14010
+    },
+    {
+      "epoch": 1.044709388971684,
+      "grad_norm": 2.2820746898651123,
+      "learning_rate": 0.0002,
+      "loss": 2.3155,
+      "step": 14020
+    },
+    {
+      "epoch": 1.0454545454545454,
+      "grad_norm": 2.676579475402832,
+      "learning_rate": 0.0002,
+      "loss": 2.3697,
+      "step": 14030
+    },
+    {
+      "epoch": 1.046199701937407,
+      "grad_norm": 2.5534327030181885,
+      "learning_rate": 0.0002,
+      "loss": 2.1582,
+      "step": 14040
+    },
+    {
+      "epoch": 1.0469448584202683,
+      "grad_norm": 2.6932196617126465,
+      "learning_rate": 0.0002,
+      "loss": 2.3688,
+      "step": 14050
+    },
+    {
+      "epoch": 1.0476900149031296,
+      "grad_norm": 2.8461594581604004,
+      "learning_rate": 0.0002,
+      "loss": 2.6371,
+      "step": 14060
+    },
+    {
+      "epoch": 1.048435171385991,
+      "grad_norm": 2.3841376304626465,
+      "learning_rate": 0.0002,
+      "loss": 2.2052,
+      "step": 14070
+    },
+    {
+      "epoch": 1.0491803278688525,
+      "grad_norm": 2.358163356781006,
+      "learning_rate": 0.0002,
+      "loss": 2.2105,
+      "step": 14080
+    },
+    {
+      "epoch": 1.0499254843517138,
+      "grad_norm": 2.9143970012664795,
+      "learning_rate": 0.0002,
+      "loss": 2.3372,
+      "step": 14090
+    },
+    {
+      "epoch": 1.0506706408345752,
+      "grad_norm": 2.2948029041290283,
+      "learning_rate": 0.0002,
+      "loss": 2.4952,
+      "step": 14100
+    },
+    {
+      "epoch": 1.0514157973174367,
+      "grad_norm": 2.725156307220459,
+      "learning_rate": 0.0002,
+      "loss": 2.4442,
+      "step": 14110
+    },
+    {
+      "epoch": 1.052160953800298,
+      "grad_norm": 2.590603828430176,
+      "learning_rate": 0.0002,
+      "loss": 2.4739,
+      "step": 14120
+    },
+    {
+      "epoch": 1.0529061102831594,
+      "grad_norm": 2.1326684951782227,
+      "learning_rate": 0.0002,
+      "loss": 2.3584,
+      "step": 14130
+    },
+    {
+      "epoch": 1.053651266766021,
+      "grad_norm": 2.2197265625,
+      "learning_rate": 0.0002,
+      "loss": 2.4694,
+      "step": 14140
+    },
+    {
+      "epoch": 1.0543964232488823,
+      "grad_norm": 2.6531333923339844,
+      "learning_rate": 0.0002,
+      "loss": 2.3527,
+      "step": 14150
+    },
+    {
+      "epoch": 1.0551415797317436,
+      "grad_norm": 2.60204815864563,
+      "learning_rate": 0.0002,
+      "loss": 2.4601,
+      "step": 14160
+    },
+    {
+      "epoch": 1.0558867362146052,
+      "grad_norm": 1.9536486864089966,
+      "learning_rate": 0.0002,
+      "loss": 2.3322,
+      "step": 14170
+    },
+    {
+      "epoch": 1.0566318926974665,
+      "grad_norm": 2.206132650375366,
+      "learning_rate": 0.0002,
+      "loss": 2.388,
+      "step": 14180
+    },
+    {
+      "epoch": 1.0573770491803278,
+      "grad_norm": 2.372546911239624,
+      "learning_rate": 0.0002,
+      "loss": 2.4207,
+      "step": 14190
+    },
+    {
+      "epoch": 1.0581222056631894,
+      "grad_norm": 2.5340518951416016,
+      "learning_rate": 0.0002,
+      "loss": 2.4014,
+      "step": 14200
+    },
+    {
+      "epoch": 1.0588673621460507,
+      "grad_norm": 2.513394594192505,
+      "learning_rate": 0.0002,
+      "loss": 2.5686,
+      "step": 14210
+    },
+    {
+      "epoch": 1.059612518628912,
+      "grad_norm": 2.649597644805908,
+      "learning_rate": 0.0002,
+      "loss": 2.4436,
+      "step": 14220
+    },
+    {
+      "epoch": 1.0603576751117734,
+      "grad_norm": 2.531172513961792,
+      "learning_rate": 0.0002,
+      "loss": 2.2963,
+      "step": 14230
+    },
+    {
+      "epoch": 1.061102831594635,
+      "grad_norm": 2.393568515777588,
+      "learning_rate": 0.0002,
+      "loss": 2.2868,
+      "step": 14240
+    },
+    {
+      "epoch": 1.0618479880774963,
+      "grad_norm": 2.474477767944336,
+      "learning_rate": 0.0002,
+      "loss": 2.2407,
+      "step": 14250
+    },
+    {
+      "epoch": 1.0625931445603576,
+      "grad_norm": 2.3346738815307617,
+      "learning_rate": 0.0002,
+      "loss": 2.3307,
+      "step": 14260
+    },
+    {
+      "epoch": 1.0633383010432191,
+      "grad_norm": 2.627002239227295,
+      "learning_rate": 0.0002,
+      "loss": 2.5363,
+      "step": 14270
+    },
+    {
+      "epoch": 1.0640834575260805,
+      "grad_norm": 2.602421522140503,
+      "learning_rate": 0.0002,
+      "loss": 2.5165,
+      "step": 14280
+    },
+    {
+      "epoch": 1.0648286140089418,
+      "grad_norm": 2.7509329319000244,
+      "learning_rate": 0.0002,
+      "loss": 2.3219,
+      "step": 14290
+    },
+    {
+      "epoch": 1.0655737704918034,
+      "grad_norm": 2.4855947494506836,
+      "learning_rate": 0.0002,
+      "loss": 2.5004,
+      "step": 14300
+    },
+    {
+      "epoch": 1.0663189269746647,
+      "grad_norm": 2.7259209156036377,
+      "learning_rate": 0.0002,
+      "loss": 2.3724,
+      "step": 14310
+    },
+    {
+      "epoch": 1.067064083457526,
+      "grad_norm": 2.578622579574585,
+      "learning_rate": 0.0002,
+      "loss": 2.3865,
+      "step": 14320
+    },
+    {
+      "epoch": 1.0678092399403876,
+      "grad_norm": 2.640737771987915,
+      "learning_rate": 0.0002,
+      "loss": 2.2266,
+      "step": 14330
+    },
+    {
+      "epoch": 1.068554396423249,
+      "grad_norm": 2.536515712738037,
+      "learning_rate": 0.0002,
+      "loss": 2.4457,
+      "step": 14340
+    },
+    {
+      "epoch": 1.0692995529061102,
+      "grad_norm": 2.3752591609954834,
+      "learning_rate": 0.0002,
+      "loss": 2.3925,
+      "step": 14350
+    },
+    {
+      "epoch": 1.0700447093889718,
+      "grad_norm": 2.668381929397583,
+      "learning_rate": 0.0002,
+      "loss": 2.3275,
+      "step": 14360
+    },
+    {
+      "epoch": 1.0707898658718331,
+      "grad_norm": 2.636784315109253,
+      "learning_rate": 0.0002,
+      "loss": 2.3801,
+      "step": 14370
+    },
+    {
+      "epoch": 1.0715350223546944,
+      "grad_norm": 2.4896252155303955,
+      "learning_rate": 0.0002,
+      "loss": 2.3056,
+      "step": 14380
+    },
+    {
+      "epoch": 1.072280178837556,
+      "grad_norm": 2.3532116413116455,
+      "learning_rate": 0.0002,
+      "loss": 2.2644,
+      "step": 14390
+    },
+    {
+      "epoch": 1.0730253353204173,
+      "grad_norm": 2.7929883003234863,
+      "learning_rate": 0.0002,
+      "loss": 2.326,
+      "step": 14400
+    },
+    {
+      "epoch": 1.0737704918032787,
+      "grad_norm": 2.66312837600708,
+      "learning_rate": 0.0002,
+      "loss": 2.4889,
+      "step": 14410
+    },
+    {
+      "epoch": 1.07451564828614,
+      "grad_norm": 2.480541706085205,
+      "learning_rate": 0.0002,
+      "loss": 2.4544,
+      "step": 14420
+    },
+    {
+      "epoch": 1.0752608047690015,
+      "grad_norm": 2.473315954208374,
+      "learning_rate": 0.0002,
+      "loss": 2.4154,
+      "step": 14430
+    },
+    {
+      "epoch": 1.0760059612518629,
+      "grad_norm": 2.4049172401428223,
+      "learning_rate": 0.0002,
+      "loss": 2.3524,
+      "step": 14440
+    },
+    {
+      "epoch": 1.0767511177347242,
+      "grad_norm": 2.4751241207122803,
+      "learning_rate": 0.0002,
+      "loss": 2.1488,
+      "step": 14450
+    },
+    {
+      "epoch": 1.0774962742175858,
+      "grad_norm": 2.5170867443084717,
+      "learning_rate": 0.0002,
+      "loss": 2.5725,
+      "step": 14460
+    },
+    {
+      "epoch": 1.078241430700447,
+      "grad_norm": 2.9997737407684326,
+      "learning_rate": 0.0002,
+      "loss": 2.4776,
+      "step": 14470
+    },
+    {
+      "epoch": 1.0789865871833084,
+      "grad_norm": 2.396097421646118,
+      "learning_rate": 0.0002,
+      "loss": 2.249,
+      "step": 14480
+    },
+    {
+      "epoch": 1.07973174366617,
+      "grad_norm": 3.120967149734497,
+      "learning_rate": 0.0002,
+      "loss": 2.4061,
+      "step": 14490
+    },
+    {
+      "epoch": 1.0804769001490313,
+      "grad_norm": 3.100985527038574,
+      "learning_rate": 0.0002,
+      "loss": 2.4715,
+      "step": 14500
+    },
+    {
+      "epoch": 1.0812220566318926,
+      "grad_norm": 2.4423913955688477,
+      "learning_rate": 0.0002,
+      "loss": 2.296,
+      "step": 14510
+    },
+    {
+      "epoch": 1.0819672131147542,
+      "grad_norm": 2.439854860305786,
+      "learning_rate": 0.0002,
+      "loss": 2.4691,
+      "step": 14520
+    },
+    {
+      "epoch": 1.0827123695976155,
+      "grad_norm": 2.5412724018096924,
+      "learning_rate": 0.0002,
+      "loss": 2.3789,
+      "step": 14530
+    },
+    {
+      "epoch": 1.0834575260804769,
+      "grad_norm": 1.9563956260681152,
+      "learning_rate": 0.0002,
+      "loss": 2.571,
+      "step": 14540
+    },
+    {
+      "epoch": 1.0842026825633384,
+      "grad_norm": 2.4468328952789307,
+      "learning_rate": 0.0002,
+      "loss": 2.3139,
+      "step": 14550
+    },
+    {
+      "epoch": 1.0849478390461997,
+      "grad_norm": 2.4930431842803955,
+      "learning_rate": 0.0002,
+      "loss": 2.313,
+      "step": 14560
+    },
+    {
+      "epoch": 1.085692995529061,
+      "grad_norm": 2.567399501800537,
+      "learning_rate": 0.0002,
+      "loss": 2.4385,
+      "step": 14570
+    },
+    {
+      "epoch": 1.0864381520119224,
+      "grad_norm": 2.871107339859009,
+      "learning_rate": 0.0002,
+      "loss": 2.5201,
+      "step": 14580
+    },
+    {
+      "epoch": 1.087183308494784,
+      "grad_norm": 2.9992733001708984,
+      "learning_rate": 0.0002,
+      "loss": 2.2764,
+      "step": 14590
+    },
+    {
+      "epoch": 1.0879284649776453,
+      "grad_norm": 2.3706576824188232,
+      "learning_rate": 0.0002,
+      "loss": 2.604,
+      "step": 14600
+    },
+    {
+      "epoch": 1.0886736214605066,
+      "grad_norm": 2.642545461654663,
+      "learning_rate": 0.0002,
+      "loss": 2.3878,
+      "step": 14610
+    },
+    {
+      "epoch": 1.0894187779433682,
+      "grad_norm": 2.364602565765381,
+      "learning_rate": 0.0002,
+      "loss": 2.338,
+      "step": 14620
+    },
+    {
+      "epoch": 1.0901639344262295,
+      "grad_norm": 2.9395387172698975,
+      "learning_rate": 0.0002,
+      "loss": 2.3848,
+      "step": 14630
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 3.6950175762176514,
+      "learning_rate": 0.0002,
+      "loss": 2.4532,
+      "step": 14640
+    },
+    {
+      "epoch": 1.0916542473919524,
+      "grad_norm": 2.566967487335205,
+      "learning_rate": 0.0002,
+      "loss": 2.5999,
+      "step": 14650
+    },
+    {
+      "epoch": 1.0923994038748137,
+      "grad_norm": 2.3279924392700195,
+      "learning_rate": 0.0002,
+      "loss": 2.5143,
+      "step": 14660
+    },
+    {
+      "epoch": 1.093144560357675,
+      "grad_norm": 3.0130693912506104,
+      "learning_rate": 0.0002,
+      "loss": 2.4298,
+      "step": 14670
+    },
+    {
+      "epoch": 1.0938897168405366,
+      "grad_norm": 2.0942749977111816,
+      "learning_rate": 0.0002,
+      "loss": 2.394,
+      "step": 14680
+    },
+    {
+      "epoch": 1.094634873323398,
+      "grad_norm": 2.4784209728240967,
+      "learning_rate": 0.0002,
+      "loss": 2.474,
+      "step": 14690
+    },
+    {
+      "epoch": 1.0953800298062593,
+      "grad_norm": 2.7817776203155518,
+      "learning_rate": 0.0002,
+      "loss": 2.3312,
+      "step": 14700
+    },
+    {
+      "epoch": 1.0961251862891208,
+      "grad_norm": 2.4290931224823,
+      "learning_rate": 0.0002,
+      "loss": 2.3258,
+      "step": 14710
+    },
+    {
+      "epoch": 1.0968703427719821,
+      "grad_norm": 2.543133020401001,
+      "learning_rate": 0.0002,
+      "loss": 2.4291,
+      "step": 14720
+    },
+    {
+      "epoch": 1.0976154992548435,
+      "grad_norm": 2.4686405658721924,
+      "learning_rate": 0.0002,
+      "loss": 2.2782,
+      "step": 14730
+    },
+    {
+      "epoch": 1.098360655737705,
+      "grad_norm": 2.5764386653900146,
+      "learning_rate": 0.0002,
+      "loss": 2.3708,
+      "step": 14740
+    },
+    {
+      "epoch": 1.0991058122205664,
+      "grad_norm": 2.827773094177246,
+      "learning_rate": 0.0002,
+      "loss": 2.3506,
+      "step": 14750
+    },
+    {
+      "epoch": 1.0998509687034277,
+      "grad_norm": 2.2853167057037354,
+      "learning_rate": 0.0002,
+      "loss": 2.4782,
+      "step": 14760
+    },
+    {
+      "epoch": 1.100596125186289,
+      "grad_norm": 2.094871997833252,
+      "learning_rate": 0.0002,
+      "loss": 2.3033,
+      "step": 14770
+    },
+    {
+      "epoch": 1.1013412816691506,
+      "grad_norm": 2.252192735671997,
+      "learning_rate": 0.0002,
+      "loss": 2.5064,
+      "step": 14780
+    },
+    {
+      "epoch": 1.102086438152012,
+      "grad_norm": 2.530860424041748,
+      "learning_rate": 0.0002,
+      "loss": 2.3885,
+      "step": 14790
+    },
+    {
+      "epoch": 1.1028315946348732,
+      "grad_norm": 2.822106122970581,
+      "learning_rate": 0.0002,
+      "loss": 2.4288,
+      "step": 14800
+    },
+    {
+      "epoch": 1.1035767511177348,
+      "grad_norm": 2.6030118465423584,
+      "learning_rate": 0.0002,
+      "loss": 2.3705,
+      "step": 14810
+    },
+    {
+      "epoch": 1.1043219076005961,
+      "grad_norm": 3.0381968021392822,
+      "learning_rate": 0.0002,
+      "loss": 2.3168,
+      "step": 14820
+    },
+    {
+      "epoch": 1.1050670640834575,
+      "grad_norm": 2.1125636100769043,
+      "learning_rate": 0.0002,
+      "loss": 2.2316,
+      "step": 14830
+    },
+    {
+      "epoch": 1.105812220566319,
+      "grad_norm": 2.649933338165283,
+      "learning_rate": 0.0002,
+      "loss": 2.3767,
+      "step": 14840
+    },
+    {
+      "epoch": 1.1065573770491803,
+      "grad_norm": 2.4779107570648193,
+      "learning_rate": 0.0002,
+      "loss": 2.4179,
+      "step": 14850
+    },
+    {
+      "epoch": 1.1073025335320417,
+      "grad_norm": 2.370229721069336,
+      "learning_rate": 0.0002,
+      "loss": 2.3823,
+      "step": 14860
+    },
+    {
+      "epoch": 1.1080476900149032,
+      "grad_norm": 2.200868844985962,
+      "learning_rate": 0.0002,
+      "loss": 2.4024,
+      "step": 14870
+    },
+    {
+      "epoch": 1.1087928464977646,
+      "grad_norm": 2.385791063308716,
+      "learning_rate": 0.0002,
+      "loss": 2.5435,
+      "step": 14880
+    },
+    {
+      "epoch": 1.1095380029806259,
+      "grad_norm": 2.731098175048828,
+      "learning_rate": 0.0002,
+      "loss": 2.4073,
+      "step": 14890
+    },
+    {
+      "epoch": 1.1102831594634874,
+      "grad_norm": 2.1808290481567383,
+      "learning_rate": 0.0002,
+      "loss": 2.4306,
+      "step": 14900
+    },
+    {
+      "epoch": 1.1110283159463488,
+      "grad_norm": 2.3113632202148438,
+      "learning_rate": 0.0002,
+      "loss": 2.4484,
+      "step": 14910
+    },
+    {
+      "epoch": 1.11177347242921,
+      "grad_norm": 2.473583459854126,
+      "learning_rate": 0.0002,
+      "loss": 2.3573,
+      "step": 14920
+    },
+    {
+      "epoch": 1.1125186289120714,
+      "grad_norm": 2.4973528385162354,
+      "learning_rate": 0.0002,
+      "loss": 2.4679,
+      "step": 14930
+    },
+    {
+      "epoch": 1.113263785394933,
+      "grad_norm": 2.698992967605591,
+      "learning_rate": 0.0002,
+      "loss": 2.573,
+      "step": 14940
+    },
+    {
+      "epoch": 1.1140089418777943,
+      "grad_norm": 2.4769937992095947,
+      "learning_rate": 0.0002,
+      "loss": 2.5083,
+      "step": 14950
+    },
+    {
+      "epoch": 1.1147540983606556,
+      "grad_norm": 2.54304838180542,
+      "learning_rate": 0.0002,
+      "loss": 2.3611,
+      "step": 14960
+    },
+    {
+      "epoch": 1.1154992548435172,
+      "grad_norm": 2.456918478012085,
+      "learning_rate": 0.0002,
+      "loss": 2.512,
+      "step": 14970
+    },
+    {
+      "epoch": 1.1162444113263785,
+      "grad_norm": 2.8047049045562744,
+      "learning_rate": 0.0002,
+      "loss": 2.4403,
+      "step": 14980
+    },
+    {
+      "epoch": 1.1169895678092399,
+      "grad_norm": 2.716897964477539,
+      "learning_rate": 0.0002,
+      "loss": 2.4872,
+      "step": 14990
+    },
+    {
+      "epoch": 1.1177347242921014,
+      "grad_norm": 2.525214910507202,
+      "learning_rate": 0.0002,
+      "loss": 2.3246,
+      "step": 15000
+    },
+    {
+      "epoch": 1.1184798807749627,
+      "grad_norm": 2.955278158187866,
+      "learning_rate": 0.0002,
+      "loss": 2.4764,
+      "step": 15010
+    },
+    {
+      "epoch": 1.119225037257824,
+      "grad_norm": 2.454103708267212,
+      "learning_rate": 0.0002,
+      "loss": 2.4668,
+      "step": 15020
+    },
+    {
+      "epoch": 1.1199701937406856,
+      "grad_norm": 2.495340585708618,
+      "learning_rate": 0.0002,
+      "loss": 2.3346,
+      "step": 15030
+    },
+    {
+      "epoch": 1.120715350223547,
+      "grad_norm": 2.2860970497131348,
+      "learning_rate": 0.0002,
+      "loss": 2.5325,
+      "step": 15040
+    },
+    {
+      "epoch": 1.1214605067064083,
+      "grad_norm": 2.4267046451568604,
+      "learning_rate": 0.0002,
+      "loss": 2.3052,
+      "step": 15050
+    },
+    {
+      "epoch": 1.1222056631892698,
+      "grad_norm": 2.8328089714050293,
+      "learning_rate": 0.0002,
+      "loss": 2.412,
+      "step": 15060
+    },
+    {
+      "epoch": 1.1229508196721312,
+      "grad_norm": 2.378267288208008,
+      "learning_rate": 0.0002,
+      "loss": 2.3667,
+      "step": 15070
+    },
+    {
+      "epoch": 1.1236959761549925,
+      "grad_norm": 2.625948667526245,
+      "learning_rate": 0.0002,
+      "loss": 2.3818,
+      "step": 15080
+    },
+    {
+      "epoch": 1.124441132637854,
+      "grad_norm": 2.269430160522461,
+      "learning_rate": 0.0002,
+      "loss": 2.5105,
+      "step": 15090
+    },
+    {
+      "epoch": 1.1251862891207154,
+      "grad_norm": 2.5549607276916504,
+      "learning_rate": 0.0002,
+      "loss": 2.4043,
+      "step": 15100
+    },
+    {
+      "epoch": 1.1259314456035767,
+      "grad_norm": 2.5551154613494873,
+      "learning_rate": 0.0002,
+      "loss": 2.4814,
+      "step": 15110
+    },
+    {
+      "epoch": 1.1266766020864383,
+      "grad_norm": 2.48346209526062,
+      "learning_rate": 0.0002,
+      "loss": 2.2917,
+      "step": 15120
+    },
+    {
+      "epoch": 1.1274217585692996,
+      "grad_norm": 2.214686632156372,
+      "learning_rate": 0.0002,
+      "loss": 2.4799,
+      "step": 15130
+    },
+    {
+      "epoch": 1.128166915052161,
+      "grad_norm": 2.6155035495758057,
+      "learning_rate": 0.0002,
+      "loss": 2.4024,
+      "step": 15140
+    },
+    {
+      "epoch": 1.1289120715350223,
+      "grad_norm": 2.485732316970825,
+      "learning_rate": 0.0002,
+      "loss": 2.4218,
+      "step": 15150
+    },
+    {
+      "epoch": 1.1296572280178838,
+      "grad_norm": 2.489976167678833,
+      "learning_rate": 0.0002,
+      "loss": 2.4819,
+      "step": 15160
+    },
+    {
+      "epoch": 1.1304023845007451,
+      "grad_norm": 2.557816743850708,
+      "learning_rate": 0.0002,
+      "loss": 2.404,
+      "step": 15170
+    },
+    {
+      "epoch": 1.1311475409836065,
+      "grad_norm": 2.362257242202759,
+      "learning_rate": 0.0002,
+      "loss": 2.413,
+      "step": 15180
+    },
+    {
+      "epoch": 1.131892697466468,
+      "grad_norm": 2.555720806121826,
+      "learning_rate": 0.0002,
+      "loss": 2.4822,
+      "step": 15190
+    },
+    {
+      "epoch": 1.1326378539493294,
+      "grad_norm": 2.301452398300171,
+      "learning_rate": 0.0002,
+      "loss": 2.4775,
+      "step": 15200
+    },
+    {
+      "epoch": 1.1333830104321907,
+      "grad_norm": 2.718770980834961,
+      "learning_rate": 0.0002,
+      "loss": 2.4039,
+      "step": 15210
+    },
+    {
+      "epoch": 1.1341281669150522,
+      "grad_norm": 2.561338424682617,
+      "learning_rate": 0.0002,
+      "loss": 2.4091,
+      "step": 15220
+    },
+    {
+      "epoch": 1.1348733233979136,
+      "grad_norm": 2.619065523147583,
+      "learning_rate": 0.0002,
+      "loss": 2.5057,
+      "step": 15230
+    },
+    {
+      "epoch": 1.135618479880775,
+      "grad_norm": 3.1791832447052,
+      "learning_rate": 0.0002,
+      "loss": 2.4179,
+      "step": 15240
+    },
+    {
+      "epoch": 1.1363636363636362,
+      "grad_norm": 2.456456422805786,
+      "learning_rate": 0.0002,
+      "loss": 2.2546,
+      "step": 15250
+    },
+    {
+      "epoch": 1.1371087928464978,
+      "grad_norm": 2.696702718734741,
+      "learning_rate": 0.0002,
+      "loss": 2.3953,
+      "step": 15260
+    },
+    {
+      "epoch": 1.1378539493293591,
+      "grad_norm": 2.065814733505249,
+      "learning_rate": 0.0002,
+      "loss": 2.1887,
+      "step": 15270
+    },
+    {
+      "epoch": 1.1385991058122205,
+      "grad_norm": 2.3281009197235107,
+      "learning_rate": 0.0002,
+      "loss": 2.4815,
+      "step": 15280
+    },
+    {
+      "epoch": 1.139344262295082,
+      "grad_norm": 2.4700629711151123,
+      "learning_rate": 0.0002,
+      "loss": 2.3037,
+      "step": 15290
+    },
+    {
+      "epoch": 1.1400894187779433,
+      "grad_norm": 2.3414204120635986,
+      "learning_rate": 0.0002,
+      "loss": 2.469,
+      "step": 15300
+    },
+    {
+      "epoch": 1.1408345752608047,
+      "grad_norm": 2.8033456802368164,
+      "learning_rate": 0.0002,
+      "loss": 2.2909,
+      "step": 15310
+    },
+    {
+      "epoch": 1.1415797317436662,
+      "grad_norm": 2.513645648956299,
+      "learning_rate": 0.0002,
+      "loss": 2.3957,
+      "step": 15320
+    },
+    {
+      "epoch": 1.1423248882265276,
+      "grad_norm": 3.029627799987793,
+      "learning_rate": 0.0002,
+      "loss": 2.4792,
+      "step": 15330
+    },
+    {
+      "epoch": 1.1430700447093889,
+      "grad_norm": 2.7976560592651367,
+      "learning_rate": 0.0002,
+      "loss": 2.4781,
+      "step": 15340
+    },
+    {
+      "epoch": 1.1438152011922504,
+      "grad_norm": 2.6071226596832275,
+      "learning_rate": 0.0002,
+      "loss": 2.4836,
+      "step": 15350
+    },
+    {
+      "epoch": 1.1445603576751118,
+      "grad_norm": 2.4131622314453125,
+      "learning_rate": 0.0002,
+      "loss": 2.4202,
+      "step": 15360
+    },
+    {
+      "epoch": 1.145305514157973,
+      "grad_norm": 2.6419029235839844,
+      "learning_rate": 0.0002,
+      "loss": 2.5272,
+      "step": 15370
+    },
+    {
+      "epoch": 1.1460506706408347,
+      "grad_norm": 2.893653154373169,
+      "learning_rate": 0.0002,
+      "loss": 2.5116,
+      "step": 15380
+    },
+    {
+      "epoch": 1.146795827123696,
+      "grad_norm": 2.739550828933716,
+      "learning_rate": 0.0002,
+      "loss": 2.3836,
+      "step": 15390
+    },
+    {
+      "epoch": 1.1475409836065573,
+      "grad_norm": 2.9094793796539307,
+      "learning_rate": 0.0002,
+      "loss": 2.3208,
+      "step": 15400
+    },
+    {
+      "epoch": 1.1482861400894189,
+      "grad_norm": 2.420971155166626,
+      "learning_rate": 0.0002,
+      "loss": 2.3705,
+      "step": 15410
+    },
+    {
+      "epoch": 1.1490312965722802,
+      "grad_norm": 2.5930230617523193,
+      "learning_rate": 0.0002,
+      "loss": 2.5412,
+      "step": 15420
+    },
+    {
+      "epoch": 1.1497764530551415,
+      "grad_norm": 2.6284139156341553,
+      "learning_rate": 0.0002,
+      "loss": 2.2475,
+      "step": 15430
+    },
+    {
+      "epoch": 1.150521609538003,
+      "grad_norm": 2.640470266342163,
+      "learning_rate": 0.0002,
+      "loss": 2.4292,
+      "step": 15440
+    },
+    {
+      "epoch": 1.1512667660208644,
+      "grad_norm": 2.7943484783172607,
+      "learning_rate": 0.0002,
+      "loss": 2.4816,
+      "step": 15450
+    },
+    {
+      "epoch": 1.1520119225037257,
+      "grad_norm": 2.281245470046997,
+      "learning_rate": 0.0002,
+      "loss": 2.3931,
+      "step": 15460
+    },
+    {
+      "epoch": 1.1527570789865873,
+      "grad_norm": 2.250092029571533,
+      "learning_rate": 0.0002,
+      "loss": 2.4882,
+      "step": 15470
+    },
+    {
+      "epoch": 1.1535022354694486,
+      "grad_norm": 2.4525022506713867,
+      "learning_rate": 0.0002,
+      "loss": 2.4132,
+      "step": 15480
+    },
+    {
+      "epoch": 1.15424739195231,
+      "grad_norm": 2.4779772758483887,
+      "learning_rate": 0.0002,
+      "loss": 2.3182,
+      "step": 15490
+    },
+    {
+      "epoch": 1.1549925484351713,
+      "grad_norm": 2.4261624813079834,
+      "learning_rate": 0.0002,
+      "loss": 2.5122,
+      "step": 15500
+    },
+    {
+      "epoch": 1.1557377049180328,
+      "grad_norm": 2.989243745803833,
+      "learning_rate": 0.0002,
+      "loss": 2.5352,
+      "step": 15510
+    },
+    {
+      "epoch": 1.1564828614008942,
+      "grad_norm": 3.4298269748687744,
+      "learning_rate": 0.0002,
+      "loss": 2.4098,
+      "step": 15520
+    },
+    {
+      "epoch": 1.1572280178837555,
+      "grad_norm": 2.6402604579925537,
+      "learning_rate": 0.0002,
+      "loss": 2.4133,
+      "step": 15530
+    },
+    {
+      "epoch": 1.157973174366617,
+      "grad_norm": 2.4061222076416016,
+      "learning_rate": 0.0002,
+      "loss": 2.3814,
+      "step": 15540
+    },
+    {
+      "epoch": 1.1587183308494784,
+      "grad_norm": 2.915400266647339,
+      "learning_rate": 0.0002,
+      "loss": 2.3256,
+      "step": 15550
+    },
+    {
+      "epoch": 1.1594634873323397,
+      "grad_norm": 2.7170588970184326,
+      "learning_rate": 0.0002,
+      "loss": 2.4919,
+      "step": 15560
+    },
+    {
+      "epoch": 1.1602086438152013,
+      "grad_norm": 2.5556771755218506,
+      "learning_rate": 0.0002,
+      "loss": 2.5706,
+      "step": 15570
+    },
+    {
+      "epoch": 1.1609538002980626,
+      "grad_norm": 2.762441873550415,
+      "learning_rate": 0.0002,
+      "loss": 2.3807,
+      "step": 15580
+    },
+    {
+      "epoch": 1.161698956780924,
+      "grad_norm": 2.6300220489501953,
+      "learning_rate": 0.0002,
+      "loss": 2.405,
+      "step": 15590
+    },
+    {
+      "epoch": 1.1624441132637853,
+      "grad_norm": 2.586217164993286,
+      "learning_rate": 0.0002,
+      "loss": 2.5505,
+      "step": 15600
+    },
+    {
+      "epoch": 1.1631892697466468,
+      "grad_norm": 2.6956429481506348,
+      "learning_rate": 0.0002,
+      "loss": 2.259,
+      "step": 15610
+    },
+    {
+      "epoch": 1.1639344262295082,
+      "grad_norm": 2.6788172721862793,
+      "learning_rate": 0.0002,
+      "loss": 2.281,
+      "step": 15620
+    },
+    {
+      "epoch": 1.1646795827123695,
+      "grad_norm": 2.6697628498077393,
+      "learning_rate": 0.0002,
+      "loss": 2.4952,
+      "step": 15630
+    },
+    {
+      "epoch": 1.165424739195231,
+      "grad_norm": 2.434091567993164,
+      "learning_rate": 0.0002,
+      "loss": 2.5588,
+      "step": 15640
+    },
+    {
+      "epoch": 1.1661698956780924,
+      "grad_norm": 2.590214252471924,
+      "learning_rate": 0.0002,
+      "loss": 2.4701,
+      "step": 15650
+    },
+    {
+      "epoch": 1.1669150521609537,
+      "grad_norm": 2.3940234184265137,
+      "learning_rate": 0.0002,
+      "loss": 2.4546,
+      "step": 15660
+    },
+    {
+      "epoch": 1.1676602086438153,
+      "grad_norm": 2.5530643463134766,
+      "learning_rate": 0.0002,
+      "loss": 2.3913,
+      "step": 15670
+    },
+    {
+      "epoch": 1.1684053651266766,
+      "grad_norm": 2.599776029586792,
+      "learning_rate": 0.0002,
+      "loss": 2.3379,
+      "step": 15680
+    },
+    {
+      "epoch": 1.169150521609538,
+      "grad_norm": 2.3956868648529053,
+      "learning_rate": 0.0002,
+      "loss": 2.4249,
+      "step": 15690
+    },
+    {
+      "epoch": 1.1698956780923995,
+      "grad_norm": 2.866769313812256,
+      "learning_rate": 0.0002,
+      "loss": 2.5113,
+      "step": 15700
+    },
+    {
+      "epoch": 1.1706408345752608,
+      "grad_norm": 2.3817973136901855,
+      "learning_rate": 0.0002,
+      "loss": 2.4549,
+      "step": 15710
+    },
+    {
+      "epoch": 1.1713859910581221,
+      "grad_norm": 2.8611204624176025,
+      "learning_rate": 0.0002,
+      "loss": 2.4499,
+      "step": 15720
+    },
+    {
+      "epoch": 1.1721311475409837,
+      "grad_norm": 2.5031049251556396,
+      "learning_rate": 0.0002,
+      "loss": 2.474,
+      "step": 15730
+    },
+    {
+      "epoch": 1.172876304023845,
+      "grad_norm": 2.7926108837127686,
+      "learning_rate": 0.0002,
+      "loss": 2.6565,
+      "step": 15740
+    },
+    {
+      "epoch": 1.1736214605067063,
+      "grad_norm": 2.480978488922119,
+      "learning_rate": 0.0002,
+      "loss": 2.5754,
+      "step": 15750
+    },
+    {
+      "epoch": 1.174366616989568,
+      "grad_norm": 2.831561803817749,
+      "learning_rate": 0.0002,
+      "loss": 2.2785,
+      "step": 15760
+    },
+    {
+      "epoch": 1.1751117734724292,
+      "grad_norm": 2.7527670860290527,
+      "learning_rate": 0.0002,
+      "loss": 2.4922,
+      "step": 15770
+    },
+    {
+      "epoch": 1.1758569299552906,
+      "grad_norm": 2.640598773956299,
+      "learning_rate": 0.0002,
+      "loss": 2.4216,
+      "step": 15780
+    },
+    {
+      "epoch": 1.1766020864381521,
+      "grad_norm": 2.8140854835510254,
+      "learning_rate": 0.0002,
+      "loss": 2.509,
+      "step": 15790
+    },
+    {
+      "epoch": 1.1773472429210134,
+      "grad_norm": 2.448840618133545,
+      "learning_rate": 0.0002,
+      "loss": 2.2896,
+      "step": 15800
+    },
+    {
+      "epoch": 1.1780923994038748,
+      "grad_norm": 2.5497422218322754,
+      "learning_rate": 0.0002,
+      "loss": 2.329,
+      "step": 15810
+    },
+    {
+      "epoch": 1.1788375558867363,
+      "grad_norm": 2.042785882949829,
+      "learning_rate": 0.0002,
+      "loss": 2.4259,
+      "step": 15820
+    },
+    {
+      "epoch": 1.1795827123695977,
+      "grad_norm": 2.9720048904418945,
+      "learning_rate": 0.0002,
+      "loss": 2.2438,
+      "step": 15830
+    },
+    {
+      "epoch": 1.180327868852459,
+      "grad_norm": 2.516237258911133,
+      "learning_rate": 0.0002,
+      "loss": 2.4921,
+      "step": 15840
+    },
+    {
+      "epoch": 1.1810730253353203,
+      "grad_norm": 2.7879035472869873,
+      "learning_rate": 0.0002,
+      "loss": 2.407,
+      "step": 15850
+    },
+    {
+      "epoch": 1.1818181818181819,
+      "grad_norm": 2.4089086055755615,
+      "learning_rate": 0.0002,
+      "loss": 2.4276,
+      "step": 15860
+    },
+    {
+      "epoch": 1.1825633383010432,
+      "grad_norm": 2.6390304565429688,
+      "learning_rate": 0.0002,
+      "loss": 2.428,
+      "step": 15870
+    },
+    {
+      "epoch": 1.1833084947839045,
+      "grad_norm": 3.222198963165283,
+      "learning_rate": 0.0002,
+      "loss": 2.3888,
+      "step": 15880
+    },
+    {
+      "epoch": 1.184053651266766,
+      "grad_norm": 2.4840657711029053,
+      "learning_rate": 0.0002,
+      "loss": 2.4878,
+      "step": 15890
+    },
+    {
+      "epoch": 1.1847988077496274,
+      "grad_norm": 2.1169028282165527,
+      "learning_rate": 0.0002,
+      "loss": 2.424,
+      "step": 15900
+    },
+    {
+      "epoch": 1.1855439642324888,
+      "grad_norm": 2.243370294570923,
+      "learning_rate": 0.0002,
+      "loss": 2.2954,
+      "step": 15910
+    },
+    {
+      "epoch": 1.1862891207153503,
+      "grad_norm": 2.8752856254577637,
+      "learning_rate": 0.0002,
+      "loss": 2.3783,
+      "step": 15920
+    },
+    {
+      "epoch": 1.1870342771982116,
+      "grad_norm": 2.4351089000701904,
+      "learning_rate": 0.0002,
+      "loss": 2.347,
+      "step": 15930
+    },
+    {
+      "epoch": 1.187779433681073,
+      "grad_norm": 2.618943214416504,
+      "learning_rate": 0.0002,
+      "loss": 2.4062,
+      "step": 15940
+    },
+    {
+      "epoch": 1.1885245901639343,
+      "grad_norm": 2.5448083877563477,
+      "learning_rate": 0.0002,
+      "loss": 2.4515,
+      "step": 15950
+    },
+    {
+      "epoch": 1.1892697466467959,
+      "grad_norm": 2.431405544281006,
+      "learning_rate": 0.0002,
+      "loss": 2.4382,
+      "step": 15960
+    },
+    {
+      "epoch": 1.1900149031296572,
+      "grad_norm": 2.914207696914673,
+      "learning_rate": 0.0002,
+      "loss": 2.5043,
+      "step": 15970
+    },
+    {
+      "epoch": 1.1907600596125185,
+      "grad_norm": 2.584994316101074,
+      "learning_rate": 0.0002,
+      "loss": 2.3106,
+      "step": 15980
+    },
+    {
+      "epoch": 1.19150521609538,
+      "grad_norm": 2.518873691558838,
+      "learning_rate": 0.0002,
+      "loss": 2.3248,
+      "step": 15990
+    },
+    {
+      "epoch": 1.1922503725782414,
+      "grad_norm": 2.7757396697998047,
+      "learning_rate": 0.0002,
+      "loss": 2.2855,
+      "step": 16000
+    },
+    {
+      "epoch": 1.1929955290611027,
+      "grad_norm": 2.546560764312744,
+      "learning_rate": 0.0002,
+      "loss": 2.2436,
+      "step": 16010
+    },
+    {
+      "epoch": 1.1937406855439643,
+      "grad_norm": 2.4890432357788086,
+      "learning_rate": 0.0002,
+      "loss": 2.5493,
+      "step": 16020
+    },
+    {
+      "epoch": 1.1944858420268256,
+      "grad_norm": 2.523747682571411,
+      "learning_rate": 0.0002,
+      "loss": 2.3246,
+      "step": 16030
+    },
+    {
+      "epoch": 1.195230998509687,
+      "grad_norm": 2.861879825592041,
+      "learning_rate": 0.0002,
+      "loss": 2.3921,
+      "step": 16040
+    },
+    {
+      "epoch": 1.1959761549925485,
+      "grad_norm": 2.518486261367798,
+      "learning_rate": 0.0002,
+      "loss": 2.4388,
+      "step": 16050
+    },
+    {
+      "epoch": 1.1967213114754098,
+      "grad_norm": 2.4700870513916016,
+      "learning_rate": 0.0002,
+      "loss": 2.5634,
+      "step": 16060
+    },
+    {
+      "epoch": 1.1974664679582712,
+      "grad_norm": 2.402773141860962,
+      "learning_rate": 0.0002,
+      "loss": 2.3364,
+      "step": 16070
+    },
+    {
+      "epoch": 1.1982116244411327,
+      "grad_norm": 2.673292636871338,
+      "learning_rate": 0.0002,
+      "loss": 2.5241,
+      "step": 16080
+    },
+    {
+      "epoch": 1.198956780923994,
+      "grad_norm": 2.3342599868774414,
+      "learning_rate": 0.0002,
+      "loss": 2.56,
+      "step": 16090
+    },
+    {
+      "epoch": 1.1997019374068554,
+      "grad_norm": 2.9310660362243652,
+      "learning_rate": 0.0002,
+      "loss": 2.492,
+      "step": 16100
+    },
+    {
+      "epoch": 1.200447093889717,
+      "grad_norm": 2.380824327468872,
+      "learning_rate": 0.0002,
+      "loss": 2.511,
+      "step": 16110
+    },
+    {
+      "epoch": 1.2011922503725783,
+      "grad_norm": 2.292423963546753,
+      "learning_rate": 0.0002,
+      "loss": 2.5952,
+      "step": 16120
+    },
+    {
+      "epoch": 1.2019374068554396,
+      "grad_norm": 2.2969772815704346,
+      "learning_rate": 0.0002,
+      "loss": 2.3879,
+      "step": 16130
+    },
+    {
+      "epoch": 1.2026825633383011,
+      "grad_norm": 2.4675862789154053,
+      "learning_rate": 0.0002,
+      "loss": 2.4499,
+      "step": 16140
+    },
+    {
+      "epoch": 1.2034277198211625,
+      "grad_norm": 2.35262393951416,
+      "learning_rate": 0.0002,
+      "loss": 2.4689,
+      "step": 16150
+    },
+    {
+      "epoch": 1.2041728763040238,
+      "grad_norm": 2.7040700912475586,
+      "learning_rate": 0.0002,
+      "loss": 2.547,
+      "step": 16160
+    },
+    {
+      "epoch": 1.2049180327868854,
+      "grad_norm": 2.248189926147461,
+      "learning_rate": 0.0002,
+      "loss": 2.5511,
+      "step": 16170
+    },
+    {
+      "epoch": 1.2056631892697467,
+      "grad_norm": 2.631309747695923,
+      "learning_rate": 0.0002,
+      "loss": 2.5509,
+      "step": 16180
+    },
+    {
+      "epoch": 1.206408345752608,
+      "grad_norm": 2.66676926612854,
+      "learning_rate": 0.0002,
+      "loss": 2.555,
+      "step": 16190
+    },
+    {
+      "epoch": 1.2071535022354694,
+      "grad_norm": 2.6497623920440674,
+      "learning_rate": 0.0002,
+      "loss": 2.445,
+      "step": 16200
+    },
+    {
+      "epoch": 1.207898658718331,
+      "grad_norm": 2.9009816646575928,
+      "learning_rate": 0.0002,
+      "loss": 2.1566,
+      "step": 16210
+    },
+    {
+      "epoch": 1.2086438152011922,
+      "grad_norm": 2.3509867191314697,
+      "learning_rate": 0.0002,
+      "loss": 2.5733,
+      "step": 16220
+    },
+    {
+      "epoch": 1.2093889716840536,
+      "grad_norm": 2.4485838413238525,
+      "learning_rate": 0.0002,
+      "loss": 2.4517,
+      "step": 16230
+    },
+    {
+      "epoch": 1.2101341281669151,
+      "grad_norm": 2.4066200256347656,
+      "learning_rate": 0.0002,
+      "loss": 2.4892,
+      "step": 16240
+    },
+    {
+      "epoch": 1.2108792846497765,
+      "grad_norm": 2.5041356086730957,
+      "learning_rate": 0.0002,
+      "loss": 2.4068,
+      "step": 16250
+    },
+    {
+      "epoch": 1.2116244411326378,
+      "grad_norm": 2.457516670227051,
+      "learning_rate": 0.0002,
+      "loss": 2.4337,
+      "step": 16260
+    },
+    {
+      "epoch": 1.2123695976154993,
+      "grad_norm": 2.524627923965454,
+      "learning_rate": 0.0002,
+      "loss": 2.6584,
+      "step": 16270
+    },
+    {
+      "epoch": 1.2131147540983607,
+      "grad_norm": 2.730018138885498,
+      "learning_rate": 0.0002,
+      "loss": 2.4592,
+      "step": 16280
+    },
+    {
+      "epoch": 1.213859910581222,
+      "grad_norm": 3.1194841861724854,
+      "learning_rate": 0.0002,
+      "loss": 2.5211,
+      "step": 16290
+    },
+    {
+      "epoch": 1.2146050670640836,
+      "grad_norm": 2.484910726547241,
+      "learning_rate": 0.0002,
+      "loss": 2.4436,
+      "step": 16300
+    },
+    {
+      "epoch": 1.2153502235469449,
+      "grad_norm": 2.5972893238067627,
+      "learning_rate": 0.0002,
+      "loss": 2.5412,
+      "step": 16310
+    },
+    {
+      "epoch": 1.2160953800298062,
+      "grad_norm": 2.599231719970703,
+      "learning_rate": 0.0002,
+      "loss": 2.5554,
+      "step": 16320
+    },
+    {
+      "epoch": 1.2168405365126675,
+      "grad_norm": 2.5555241107940674,
+      "learning_rate": 0.0002,
+      "loss": 2.5618,
+      "step": 16330
+    },
+    {
+      "epoch": 1.217585692995529,
+      "grad_norm": 2.6577017307281494,
+      "learning_rate": 0.0002,
+      "loss": 2.3419,
+      "step": 16340
+    },
+    {
+      "epoch": 1.2183308494783904,
+      "grad_norm": 2.239018440246582,
+      "learning_rate": 0.0002,
+      "loss": 2.4517,
+      "step": 16350
+    },
+    {
+      "epoch": 1.2190760059612518,
+      "grad_norm": 2.2550699710845947,
+      "learning_rate": 0.0002,
+      "loss": 2.5087,
+      "step": 16360
+    },
+    {
+      "epoch": 1.2198211624441133,
+      "grad_norm": 2.7445180416107178,
+      "learning_rate": 0.0002,
+      "loss": 2.5011,
+      "step": 16370
+    },
+    {
+      "epoch": 1.2205663189269746,
+      "grad_norm": 2.637608528137207,
+      "learning_rate": 0.0002,
+      "loss": 2.3431,
+      "step": 16380
+    },
+    {
+      "epoch": 1.221311475409836,
+      "grad_norm": 2.4551427364349365,
+      "learning_rate": 0.0002,
+      "loss": 2.4485,
+      "step": 16390
+    },
+    {
+      "epoch": 1.2220566318926975,
+      "grad_norm": 2.595010757446289,
+      "learning_rate": 0.0002,
+      "loss": 2.6419,
+      "step": 16400
+    },
+    {
+      "epoch": 1.2228017883755589,
+      "grad_norm": 3.2001779079437256,
+      "learning_rate": 0.0002,
+      "loss": 2.3009,
+      "step": 16410
+    },
+    {
+      "epoch": 1.2235469448584202,
+      "grad_norm": 2.3680529594421387,
+      "learning_rate": 0.0002,
+      "loss": 2.4545,
+      "step": 16420
+    },
+    {
+      "epoch": 1.2242921013412817,
+      "grad_norm": 2.284846782684326,
+      "learning_rate": 0.0002,
+      "loss": 2.4662,
+      "step": 16430
+    },
+    {
+      "epoch": 1.225037257824143,
+      "grad_norm": 2.7528367042541504,
+      "learning_rate": 0.0002,
+      "loss": 2.504,
+      "step": 16440
+    },
+    {
+      "epoch": 1.2257824143070044,
+      "grad_norm": 2.5540902614593506,
+      "learning_rate": 0.0002,
+      "loss": 2.5621,
+      "step": 16450
+    },
+    {
+      "epoch": 1.226527570789866,
+      "grad_norm": 2.6369526386260986,
+      "learning_rate": 0.0002,
+      "loss": 2.4563,
+      "step": 16460
+    },
+    {
+      "epoch": 1.2272727272727273,
+      "grad_norm": 2.560155153274536,
+      "learning_rate": 0.0002,
+      "loss": 2.3171,
+      "step": 16470
+    },
+    {
+      "epoch": 1.2280178837555886,
+      "grad_norm": 2.8841707706451416,
+      "learning_rate": 0.0002,
+      "loss": 2.5621,
+      "step": 16480
+    },
+    {
+      "epoch": 1.2287630402384502,
+      "grad_norm": 2.587996482849121,
+      "learning_rate": 0.0002,
+      "loss": 2.5119,
+      "step": 16490
+    },
+    {
+      "epoch": 1.2295081967213115,
+      "grad_norm": 2.455828905105591,
+      "learning_rate": 0.0002,
+      "loss": 2.3992,
+      "step": 16500
+    },
+    {
+      "epoch": 1.2302533532041728,
+      "grad_norm": 2.5524632930755615,
+      "learning_rate": 0.0002,
+      "loss": 2.5209,
+      "step": 16510
+    },
+    {
+      "epoch": 1.2309985096870344,
+      "grad_norm": 2.436572551727295,
+      "learning_rate": 0.0002,
+      "loss": 2.4511,
+      "step": 16520
+    },
+    {
+      "epoch": 1.2317436661698957,
+      "grad_norm": 2.5850322246551514,
+      "learning_rate": 0.0002,
+      "loss": 2.3393,
+      "step": 16530
+    },
+    {
+      "epoch": 1.232488822652757,
+      "grad_norm": 2.884225845336914,
+      "learning_rate": 0.0002,
+      "loss": 2.4517,
+      "step": 16540
+    },
+    {
+      "epoch": 1.2332339791356184,
+      "grad_norm": 2.3954968452453613,
+      "learning_rate": 0.0002,
+      "loss": 2.5374,
+      "step": 16550
+    },
+    {
+      "epoch": 1.23397913561848,
+      "grad_norm": 2.329937696456909,
+      "learning_rate": 0.0002,
+      "loss": 2.631,
+      "step": 16560
+    },
+    {
+      "epoch": 1.2347242921013413,
+      "grad_norm": 1.706188440322876,
+      "learning_rate": 0.0002,
+      "loss": 2.3386,
+      "step": 16570
+    },
+    {
+      "epoch": 1.2354694485842026,
+      "grad_norm": 2.635439395904541,
+      "learning_rate": 0.0002,
+      "loss": 2.408,
+      "step": 16580
+    },
+    {
+      "epoch": 1.2362146050670642,
+      "grad_norm": 3.037846326828003,
+      "learning_rate": 0.0002,
+      "loss": 2.3775,
+      "step": 16590
+    },
+    {
+      "epoch": 1.2369597615499255,
+      "grad_norm": 2.435914993286133,
+      "learning_rate": 0.0002,
+      "loss": 2.4828,
+      "step": 16600
+    },
+    {
+      "epoch": 1.2377049180327868,
+      "grad_norm": 2.3864994049072266,
+      "learning_rate": 0.0002,
+      "loss": 2.307,
+      "step": 16610
+    },
+    {
+      "epoch": 1.2384500745156484,
+      "grad_norm": 2.5248770713806152,
+      "learning_rate": 0.0002,
+      "loss": 2.4023,
+      "step": 16620
+    },
+    {
+      "epoch": 1.2391952309985097,
+      "grad_norm": 2.4763107299804688,
+      "learning_rate": 0.0002,
+      "loss": 2.5354,
+      "step": 16630
+    },
+    {
+      "epoch": 1.239940387481371,
+      "grad_norm": 2.6904945373535156,
+      "learning_rate": 0.0002,
+      "loss": 2.5673,
+      "step": 16640
+    },
+    {
+      "epoch": 1.2406855439642326,
+      "grad_norm": 2.5238559246063232,
+      "learning_rate": 0.0002,
+      "loss": 2.4272,
+      "step": 16650
+    },
+    {
+      "epoch": 1.241430700447094,
+      "grad_norm": 2.451791763305664,
+      "learning_rate": 0.0002,
+      "loss": 2.5297,
+      "step": 16660
+    },
+    {
+      "epoch": 1.2421758569299552,
+      "grad_norm": 2.5765328407287598,
+      "learning_rate": 0.0002,
+      "loss": 2.4879,
+      "step": 16670
+    },
+    {
+      "epoch": 1.2429210134128166,
+      "grad_norm": 2.518141269683838,
+      "learning_rate": 0.0002,
+      "loss": 2.3585,
+      "step": 16680
+    },
+    {
+      "epoch": 1.2436661698956781,
+      "grad_norm": 2.4746904373168945,
+      "learning_rate": 0.0002,
+      "loss": 2.5626,
+      "step": 16690
+    },
+    {
+      "epoch": 1.2444113263785395,
+      "grad_norm": 2.6540675163269043,
+      "learning_rate": 0.0002,
+      "loss": 2.4304,
+      "step": 16700
+    },
+    {
+      "epoch": 1.2451564828614008,
+      "grad_norm": 1.8600475788116455,
+      "learning_rate": 0.0002,
+      "loss": 2.394,
+      "step": 16710
+    },
+    {
+      "epoch": 1.2459016393442623,
+      "grad_norm": 2.5817408561706543,
+      "learning_rate": 0.0002,
+      "loss": 2.5504,
+      "step": 16720
+    },
+    {
+      "epoch": 1.2466467958271237,
+      "grad_norm": 2.5648725032806396,
+      "learning_rate": 0.0002,
+      "loss": 2.2296,
+      "step": 16730
+    },
+    {
+      "epoch": 1.247391952309985,
+      "grad_norm": 2.600857734680176,
+      "learning_rate": 0.0002,
+      "loss": 2.3259,
+      "step": 16740
+    },
+    {
+      "epoch": 1.2481371087928466,
+      "grad_norm": 2.6139464378356934,
+      "learning_rate": 0.0002,
+      "loss": 2.4264,
+      "step": 16750
+    },
+    {
+      "epoch": 1.248882265275708,
+      "grad_norm": 2.387685537338257,
+      "learning_rate": 0.0002,
+      "loss": 2.6434,
+      "step": 16760
+    },
+    {
+      "epoch": 1.2496274217585692,
+      "grad_norm": 2.712690830230713,
+      "learning_rate": 0.0002,
+      "loss": 2.461,
+      "step": 16770
+    },
+    {
+      "epoch": 1.2503725782414308,
+      "grad_norm": 2.4673256874084473,
+      "learning_rate": 0.0002,
+      "loss": 2.4162,
+      "step": 16780
+    },
+    {
+      "epoch": 1.251117734724292,
+      "grad_norm": 2.5331919193267822,
+      "learning_rate": 0.0002,
+      "loss": 2.591,
+      "step": 16790
+    },
+    {
+      "epoch": 1.2518628912071534,
+      "grad_norm": 2.48547625541687,
+      "learning_rate": 0.0002,
+      "loss": 2.2615,
+      "step": 16800
+    },
+    {
+      "epoch": 1.252608047690015,
+      "grad_norm": 2.41300368309021,
+      "learning_rate": 0.0002,
+      "loss": 2.3431,
+      "step": 16810
+    },
+    {
+      "epoch": 1.2533532041728763,
+      "grad_norm": 2.566871166229248,
+      "learning_rate": 0.0002,
+      "loss": 2.3982,
+      "step": 16820
+    },
+    {
+      "epoch": 1.2540983606557377,
+      "grad_norm": 2.4768545627593994,
+      "learning_rate": 0.0002,
+      "loss": 2.346,
+      "step": 16830
+    },
+    {
+      "epoch": 1.2548435171385992,
+      "grad_norm": 2.677468776702881,
+      "learning_rate": 0.0002,
+      "loss": 2.2478,
+      "step": 16840
+    },
+    {
+      "epoch": 1.2555886736214605,
+      "grad_norm": 2.2552690505981445,
+      "learning_rate": 0.0002,
+      "loss": 2.5072,
+      "step": 16850
+    },
+    {
+      "epoch": 1.2563338301043219,
+      "grad_norm": 2.868969202041626,
+      "learning_rate": 0.0002,
+      "loss": 2.2606,
+      "step": 16860
+    },
+    {
+      "epoch": 1.2570789865871834,
+      "grad_norm": 2.6906239986419678,
+      "learning_rate": 0.0002,
+      "loss": 2.5686,
+      "step": 16870
+    },
+    {
+      "epoch": 1.2578241430700448,
+      "grad_norm": 2.433225393295288,
+      "learning_rate": 0.0002,
+      "loss": 2.3974,
+      "step": 16880
+    },
+    {
+      "epoch": 1.258569299552906,
+      "grad_norm": 2.37595534324646,
+      "learning_rate": 0.0002,
+      "loss": 2.4342,
+      "step": 16890
+    },
+    {
+      "epoch": 1.2593144560357676,
+      "grad_norm": 2.5448384284973145,
+      "learning_rate": 0.0002,
+      "loss": 2.3194,
+      "step": 16900
+    },
+    {
+      "epoch": 1.260059612518629,
+      "grad_norm": 2.858198404312134,
+      "learning_rate": 0.0002,
+      "loss": 2.4199,
+      "step": 16910
+    },
+    {
+      "epoch": 1.2608047690014903,
+      "grad_norm": 2.546701431274414,
+      "learning_rate": 0.0002,
+      "loss": 2.4509,
+      "step": 16920
+    },
+    {
+      "epoch": 1.2615499254843516,
+      "grad_norm": 2.5379738807678223,
+      "learning_rate": 0.0002,
+      "loss": 2.4427,
+      "step": 16930
+    },
+    {
+      "epoch": 1.2622950819672132,
+      "grad_norm": 2.6814608573913574,
+      "learning_rate": 0.0002,
+      "loss": 2.4686,
+      "step": 16940
+    },
+    {
+      "epoch": 1.2630402384500745,
+      "grad_norm": 2.3676536083221436,
+      "learning_rate": 0.0002,
+      "loss": 2.4649,
+      "step": 16950
+    },
+    {
+      "epoch": 1.2637853949329358,
+      "grad_norm": 2.789276123046875,
+      "learning_rate": 0.0002,
+      "loss": 2.3954,
+      "step": 16960
+    },
+    {
+      "epoch": 1.2645305514157974,
+      "grad_norm": 2.770944595336914,
+      "learning_rate": 0.0002,
+      "loss": 2.5665,
+      "step": 16970
+    },
+    {
+      "epoch": 1.2652757078986587,
+      "grad_norm": 2.385180950164795,
+      "learning_rate": 0.0002,
+      "loss": 2.227,
+      "step": 16980
+    },
+    {
+      "epoch": 1.26602086438152,
+      "grad_norm": 2.8469879627227783,
+      "learning_rate": 0.0002,
+      "loss": 2.2351,
+      "step": 16990
+    },
+    {
+      "epoch": 1.2667660208643814,
+      "grad_norm": 2.438333034515381,
+      "learning_rate": 0.0002,
+      "loss": 2.3147,
+      "step": 17000
+    },
+    {
+      "epoch": 1.267511177347243,
+      "grad_norm": 2.537980794906616,
+      "learning_rate": 0.0002,
+      "loss": 2.4113,
+      "step": 17010
+    },
+    {
+      "epoch": 1.2682563338301043,
+      "grad_norm": 2.5566976070404053,
+      "learning_rate": 0.0002,
+      "loss": 2.3972,
+      "step": 17020
+    },
+    {
+      "epoch": 1.2690014903129656,
+      "grad_norm": 2.5756947994232178,
+      "learning_rate": 0.0002,
+      "loss": 2.3624,
+      "step": 17030
+    },
+    {
+      "epoch": 1.2697466467958272,
+      "grad_norm": 2.6613008975982666,
+      "learning_rate": 0.0002,
+      "loss": 2.4689,
+      "step": 17040
+    },
+    {
+      "epoch": 1.2704918032786885,
+      "grad_norm": 2.802849054336548,
+      "learning_rate": 0.0002,
+      "loss": 2.3772,
+      "step": 17050
+    },
+    {
+      "epoch": 1.2712369597615498,
+      "grad_norm": 2.221499443054199,
+      "learning_rate": 0.0002,
+      "loss": 2.5116,
+      "step": 17060
+    },
+    {
+      "epoch": 1.2719821162444114,
+      "grad_norm": 2.3637337684631348,
+      "learning_rate": 0.0002,
+      "loss": 2.4486,
+      "step": 17070
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 2.217311382293701,
+      "learning_rate": 0.0002,
+      "loss": 2.4252,
+      "step": 17080
+    },
+    {
+      "epoch": 1.273472429210134,
+      "grad_norm": 2.36021089553833,
+      "learning_rate": 0.0002,
+      "loss": 2.5555,
+      "step": 17090
+    },
+    {
+      "epoch": 1.2742175856929956,
+      "grad_norm": 2.5092270374298096,
+      "learning_rate": 0.0002,
+      "loss": 2.298,
+      "step": 17100
+    },
+    {
+      "epoch": 1.274962742175857,
+      "grad_norm": 2.396413803100586,
+      "learning_rate": 0.0002,
+      "loss": 2.113,
+      "step": 17110
+    },
+    {
+      "epoch": 1.2757078986587183,
+      "grad_norm": 2.5738630294799805,
+      "learning_rate": 0.0002,
+      "loss": 2.2539,
+      "step": 17120
+    },
+    {
+      "epoch": 1.2764530551415798,
+      "grad_norm": 2.5101852416992188,
+      "learning_rate": 0.0002,
+      "loss": 2.3469,
+      "step": 17130
+    },
+    {
+      "epoch": 1.2771982116244411,
+      "grad_norm": 2.4591407775878906,
+      "learning_rate": 0.0002,
+      "loss": 2.6166,
+      "step": 17140
+    },
+    {
+      "epoch": 1.2779433681073025,
+      "grad_norm": 2.526982069015503,
+      "learning_rate": 0.0002,
+      "loss": 2.4245,
+      "step": 17150
+    },
+    {
+      "epoch": 1.278688524590164,
+      "grad_norm": 2.8430566787719727,
+      "learning_rate": 0.0002,
+      "loss": 2.3319,
+      "step": 17160
+    },
+    {
+      "epoch": 1.2794336810730254,
+      "grad_norm": 3.2182416915893555,
+      "learning_rate": 0.0002,
+      "loss": 2.4737,
+      "step": 17170
+    },
+    {
+      "epoch": 1.2801788375558867,
+      "grad_norm": 1.936545491218567,
+      "learning_rate": 0.0002,
+      "loss": 2.3686,
+      "step": 17180
+    },
+    {
+      "epoch": 1.2809239940387482,
+      "grad_norm": 2.754791736602783,
+      "learning_rate": 0.0002,
+      "loss": 2.5935,
+      "step": 17190
+    },
+    {
+      "epoch": 1.2816691505216096,
+      "grad_norm": 2.615044116973877,
+      "learning_rate": 0.0002,
+      "loss": 2.1516,
+      "step": 17200
+    },
+    {
+      "epoch": 1.282414307004471,
+      "grad_norm": 2.3363049030303955,
+      "learning_rate": 0.0002,
+      "loss": 2.4082,
+      "step": 17210
+    },
+    {
+      "epoch": 1.2831594634873325,
+      "grad_norm": 2.6004695892333984,
+      "learning_rate": 0.0002,
+      "loss": 2.2609,
+      "step": 17220
+    },
+    {
+      "epoch": 1.2839046199701938,
+      "grad_norm": 2.6206114292144775,
+      "learning_rate": 0.0002,
+      "loss": 2.6122,
+      "step": 17230
+    },
+    {
+      "epoch": 1.2846497764530551,
+      "grad_norm": 2.5685789585113525,
+      "learning_rate": 0.0002,
+      "loss": 2.4993,
+      "step": 17240
+    },
+    {
+      "epoch": 1.2853949329359167,
+      "grad_norm": 2.702047824859619,
+      "learning_rate": 0.0002,
+      "loss": 2.5556,
+      "step": 17250
+    },
+    {
+      "epoch": 1.286140089418778,
+      "grad_norm": 2.7429511547088623,
+      "learning_rate": 0.0002,
+      "loss": 2.3498,
+      "step": 17260
+    },
+    {
+      "epoch": 1.2868852459016393,
+      "grad_norm": 2.4050281047821045,
+      "learning_rate": 0.0002,
+      "loss": 2.4434,
+      "step": 17270
+    },
+    {
+      "epoch": 1.2876304023845009,
+      "grad_norm": 2.3521718978881836,
+      "learning_rate": 0.0002,
+      "loss": 2.2286,
+      "step": 17280
+    },
+    {
+      "epoch": 1.2883755588673622,
+      "grad_norm": 2.6510558128356934,
+      "learning_rate": 0.0002,
+      "loss": 2.4448,
+      "step": 17290
+    },
+    {
+      "epoch": 1.2891207153502235,
+      "grad_norm": 2.949458599090576,
+      "learning_rate": 0.0002,
+      "loss": 2.6088,
+      "step": 17300
+    },
+    {
+      "epoch": 1.2898658718330849,
+      "grad_norm": 2.5707809925079346,
+      "learning_rate": 0.0002,
+      "loss": 2.5102,
+      "step": 17310
+    },
+    {
+      "epoch": 1.2906110283159464,
+      "grad_norm": 2.6018505096435547,
+      "learning_rate": 0.0002,
+      "loss": 2.5562,
+      "step": 17320
+    },
+    {
+      "epoch": 1.2913561847988078,
+      "grad_norm": 2.5191781520843506,
+      "learning_rate": 0.0002,
+      "loss": 2.2755,
+      "step": 17330
+    },
+    {
+      "epoch": 1.292101341281669,
+      "grad_norm": 2.482866048812866,
+      "learning_rate": 0.0002,
+      "loss": 2.3706,
+      "step": 17340
+    },
+    {
+      "epoch": 1.2928464977645304,
+      "grad_norm": 2.840179681777954,
+      "learning_rate": 0.0002,
+      "loss": 2.6733,
+      "step": 17350
+    },
+    {
+      "epoch": 1.293591654247392,
+      "grad_norm": 2.3511884212493896,
+      "learning_rate": 0.0002,
+      "loss": 2.1673,
+      "step": 17360
+    },
+    {
+      "epoch": 1.2943368107302533,
+      "grad_norm": 2.9184696674346924,
+      "learning_rate": 0.0002,
+      "loss": 2.4761,
+      "step": 17370
+    },
+    {
+      "epoch": 1.2950819672131146,
+      "grad_norm": 2.500131130218506,
+      "learning_rate": 0.0002,
+      "loss": 2.5081,
+      "step": 17380
+    },
+    {
+      "epoch": 1.2958271236959762,
+      "grad_norm": 2.2473769187927246,
+      "learning_rate": 0.0002,
+      "loss": 2.421,
+      "step": 17390
+    },
+    {
+      "epoch": 1.2965722801788375,
+      "grad_norm": 2.166339159011841,
+      "learning_rate": 0.0002,
+      "loss": 2.4269,
+      "step": 17400
+    },
+    {
+      "epoch": 1.2973174366616989,
+      "grad_norm": 2.546684741973877,
+      "learning_rate": 0.0002,
+      "loss": 2.3341,
+      "step": 17410
+    },
+    {
+      "epoch": 1.2980625931445604,
+      "grad_norm": 2.3546717166900635,
+      "learning_rate": 0.0002,
+      "loss": 2.2886,
+      "step": 17420
+    },
+    {
+      "epoch": 1.2988077496274217,
+      "grad_norm": 3.0054266452789307,
+      "learning_rate": 0.0002,
+      "loss": 2.5172,
+      "step": 17430
+    },
+    {
+      "epoch": 1.299552906110283,
+      "grad_norm": 2.7535972595214844,
+      "learning_rate": 0.0002,
+      "loss": 2.5921,
+      "step": 17440
+    },
+    {
+      "epoch": 1.3002980625931446,
+      "grad_norm": 2.568647861480713,
+      "learning_rate": 0.0002,
+      "loss": 2.4328,
+      "step": 17450
+    },
+    {
+      "epoch": 1.301043219076006,
+      "grad_norm": 2.550978899002075,
+      "learning_rate": 0.0002,
+      "loss": 2.4866,
+      "step": 17460
+    },
+    {
+      "epoch": 1.3017883755588673,
+      "grad_norm": 2.5628457069396973,
+      "learning_rate": 0.0002,
+      "loss": 2.4284,
+      "step": 17470
+    },
+    {
+      "epoch": 1.3025335320417288,
+      "grad_norm": 2.3512864112854004,
+      "learning_rate": 0.0002,
+      "loss": 2.4614,
+      "step": 17480
+    },
+    {
+      "epoch": 1.3032786885245902,
+      "grad_norm": 2.6441545486450195,
+      "learning_rate": 0.0002,
+      "loss": 2.4846,
+      "step": 17490
+    },
+    {
+      "epoch": 1.3040238450074515,
+      "grad_norm": 2.4340755939483643,
+      "learning_rate": 0.0002,
+      "loss": 2.2236,
+      "step": 17500
+    },
+    {
+      "epoch": 1.304769001490313,
+      "grad_norm": 2.4928324222564697,
+      "learning_rate": 0.0002,
+      "loss": 2.4082,
+      "step": 17510
+    },
+    {
+      "epoch": 1.3055141579731744,
+      "grad_norm": 2.5223214626312256,
+      "learning_rate": 0.0002,
+      "loss": 2.5462,
+      "step": 17520
+    },
+    {
+      "epoch": 1.3062593144560357,
+      "grad_norm": 2.824338674545288,
+      "learning_rate": 0.0002,
+      "loss": 2.4898,
+      "step": 17530
+    },
+    {
+      "epoch": 1.3070044709388973,
+      "grad_norm": 2.299622058868408,
+      "learning_rate": 0.0002,
+      "loss": 2.4033,
+      "step": 17540
+    },
+    {
+      "epoch": 1.3077496274217586,
+      "grad_norm": 2.4792845249176025,
+      "learning_rate": 0.0002,
+      "loss": 2.5937,
+      "step": 17550
+    },
+    {
+      "epoch": 1.30849478390462,
+      "grad_norm": 2.402050495147705,
+      "learning_rate": 0.0002,
+      "loss": 2.5102,
+      "step": 17560
+    },
+    {
+      "epoch": 1.3092399403874815,
+      "grad_norm": 2.833794593811035,
+      "learning_rate": 0.0002,
+      "loss": 2.6331,
+      "step": 17570
+    },
+    {
+      "epoch": 1.3099850968703428,
+      "grad_norm": 2.4737417697906494,
+      "learning_rate": 0.0002,
+      "loss": 2.2122,
+      "step": 17580
+    },
+    {
+      "epoch": 1.3107302533532041,
+      "grad_norm": 2.520129680633545,
+      "learning_rate": 0.0002,
+      "loss": 2.4989,
+      "step": 17590
+    },
+    {
+      "epoch": 1.3114754098360657,
+      "grad_norm": 2.310004472732544,
+      "learning_rate": 0.0002,
+      "loss": 2.3101,
+      "step": 17600
+    },
+    {
+      "epoch": 1.312220566318927,
+      "grad_norm": 2.863279342651367,
+      "learning_rate": 0.0002,
+      "loss": 2.4879,
+      "step": 17610
+    },
+    {
+      "epoch": 1.3129657228017884,
+      "grad_norm": 2.590439558029175,
+      "learning_rate": 0.0002,
+      "loss": 2.6645,
+      "step": 17620
+    },
+    {
+      "epoch": 1.31371087928465,
+      "grad_norm": 2.840513229370117,
+      "learning_rate": 0.0002,
+      "loss": 2.4547,
+      "step": 17630
+    },
+    {
+      "epoch": 1.3144560357675112,
+      "grad_norm": 2.6092474460601807,
+      "learning_rate": 0.0002,
+      "loss": 2.5813,
+      "step": 17640
+    },
+    {
+      "epoch": 1.3152011922503726,
+      "grad_norm": 2.347975969314575,
+      "learning_rate": 0.0002,
+      "loss": 2.3886,
+      "step": 17650
+    },
+    {
+      "epoch": 1.315946348733234,
+      "grad_norm": 2.4244983196258545,
+      "learning_rate": 0.0002,
+      "loss": 2.4909,
+      "step": 17660
+    },
+    {
+      "epoch": 1.3166915052160955,
+      "grad_norm": 2.1615536212921143,
+      "learning_rate": 0.0002,
+      "loss": 2.4675,
+      "step": 17670
+    },
+    {
+      "epoch": 1.3174366616989568,
+      "grad_norm": 2.350623846054077,
+      "learning_rate": 0.0002,
+      "loss": 2.4184,
+      "step": 17680
+    },
+    {
+      "epoch": 1.3181818181818181,
+      "grad_norm": 2.4277079105377197,
+      "learning_rate": 0.0002,
+      "loss": 2.5482,
+      "step": 17690
+    },
+    {
+      "epoch": 1.3189269746646795,
+      "grad_norm": 2.431518793106079,
+      "learning_rate": 0.0002,
+      "loss": 2.5813,
+      "step": 17700
+    },
+    {
+      "epoch": 1.319672131147541,
+      "grad_norm": 2.4477312564849854,
+      "learning_rate": 0.0002,
+      "loss": 2.4473,
+      "step": 17710
+    },
+    {
+      "epoch": 1.3204172876304023,
+      "grad_norm": 2.8006045818328857,
+      "learning_rate": 0.0002,
+      "loss": 2.639,
+      "step": 17720
+    },
+    {
+      "epoch": 1.3211624441132637,
+      "grad_norm": 2.362448215484619,
+      "learning_rate": 0.0002,
+      "loss": 2.3965,
+      "step": 17730
+    },
+    {
+      "epoch": 1.3219076005961252,
+      "grad_norm": 2.489379405975342,
+      "learning_rate": 0.0002,
+      "loss": 2.4664,
+      "step": 17740
+    },
+    {
+      "epoch": 1.3226527570789866,
+      "grad_norm": 2.386716365814209,
+      "learning_rate": 0.0002,
+      "loss": 2.5855,
+      "step": 17750
+    },
+    {
+      "epoch": 1.3233979135618479,
+      "grad_norm": 2.5991947650909424,
+      "learning_rate": 0.0002,
+      "loss": 2.5633,
+      "step": 17760
+    },
+    {
+      "epoch": 1.3241430700447094,
+      "grad_norm": 2.499544858932495,
+      "learning_rate": 0.0002,
+      "loss": 2.4749,
+      "step": 17770
+    },
+    {
+      "epoch": 1.3248882265275708,
+      "grad_norm": 2.4425177574157715,
+      "learning_rate": 0.0002,
+      "loss": 2.5387,
+      "step": 17780
+    },
+    {
+      "epoch": 1.325633383010432,
+      "grad_norm": 2.6338937282562256,
+      "learning_rate": 0.0002,
+      "loss": 2.4913,
+      "step": 17790
+    },
+    {
+      "epoch": 1.3263785394932937,
+      "grad_norm": 2.280900716781616,
+      "learning_rate": 0.0002,
+      "loss": 2.548,
+      "step": 17800
+    },
+    {
+      "epoch": 1.327123695976155,
+      "grad_norm": 2.5411298274993896,
+      "learning_rate": 0.0002,
+      "loss": 2.5131,
+      "step": 17810
+    },
+    {
+      "epoch": 1.3278688524590163,
+      "grad_norm": 2.546914577484131,
+      "learning_rate": 0.0002,
+      "loss": 2.667,
+      "step": 17820
+    },
+    {
+      "epoch": 1.3286140089418779,
+      "grad_norm": 2.6048712730407715,
+      "learning_rate": 0.0002,
+      "loss": 2.3578,
+      "step": 17830
+    },
+    {
+      "epoch": 1.3293591654247392,
+      "grad_norm": 2.714755058288574,
+      "learning_rate": 0.0002,
+      "loss": 2.6083,
+      "step": 17840
+    },
+    {
+      "epoch": 1.3301043219076005,
+      "grad_norm": 2.559908151626587,
+      "learning_rate": 0.0002,
+      "loss": 2.4031,
+      "step": 17850
+    },
+    {
+      "epoch": 1.330849478390462,
+      "grad_norm": 2.308744192123413,
+      "learning_rate": 0.0002,
+      "loss": 2.3735,
+      "step": 17860
+    },
+    {
+      "epoch": 1.3315946348733234,
+      "grad_norm": 2.331125497817993,
+      "learning_rate": 0.0002,
+      "loss": 2.2689,
+      "step": 17870
+    },
+    {
+      "epoch": 1.3323397913561847,
+      "grad_norm": 2.7024810314178467,
+      "learning_rate": 0.0002,
+      "loss": 2.4517,
+      "step": 17880
+    },
+    {
+      "epoch": 1.3330849478390463,
+      "grad_norm": 2.6669280529022217,
+      "learning_rate": 0.0002,
+      "loss": 2.3356,
+      "step": 17890
+    },
+    {
+      "epoch": 1.3338301043219076,
+      "grad_norm": 2.739065408706665,
+      "learning_rate": 0.0002,
+      "loss": 2.3587,
+      "step": 17900
+    },
+    {
+      "epoch": 1.334575260804769,
+      "grad_norm": 2.8512372970581055,
+      "learning_rate": 0.0002,
+      "loss": 2.5018,
+      "step": 17910
+    },
+    {
+      "epoch": 1.3353204172876305,
+      "grad_norm": 2.336021900177002,
+      "learning_rate": 0.0002,
+      "loss": 2.3453,
+      "step": 17920
+    },
+    {
+      "epoch": 1.3360655737704918,
+      "grad_norm": 2.4425575733184814,
+      "learning_rate": 0.0002,
+      "loss": 2.5346,
+      "step": 17930
+    },
+    {
+      "epoch": 1.3368107302533532,
+      "grad_norm": 3.027134656906128,
+      "learning_rate": 0.0002,
+      "loss": 2.3945,
+      "step": 17940
+    },
+    {
+      "epoch": 1.3375558867362147,
+      "grad_norm": 2.451101541519165,
+      "learning_rate": 0.0002,
+      "loss": 2.4273,
+      "step": 17950
+    },
+    {
+      "epoch": 1.338301043219076,
+      "grad_norm": 2.5687952041625977,
+      "learning_rate": 0.0002,
+      "loss": 2.5681,
+      "step": 17960
+    },
+    {
+      "epoch": 1.3390461997019374,
+      "grad_norm": 2.5672757625579834,
+      "learning_rate": 0.0002,
+      "loss": 2.5521,
+      "step": 17970
+    },
+    {
+      "epoch": 1.339791356184799,
+      "grad_norm": 2.4332363605499268,
+      "learning_rate": 0.0002,
+      "loss": 2.4988,
+      "step": 17980
+    },
+    {
+      "epoch": 1.3405365126676603,
+      "grad_norm": 2.3175182342529297,
+      "learning_rate": 0.0002,
+      "loss": 2.1908,
+      "step": 17990
+    },
+    {
+      "epoch": 1.3412816691505216,
+      "grad_norm": 2.677385091781616,
+      "learning_rate": 0.0002,
+      "loss": 2.6282,
+      "step": 18000
+    },
+    {
+      "epoch": 1.342026825633383,
+      "grad_norm": 2.5982978343963623,
+      "learning_rate": 0.0002,
+      "loss": 2.3879,
+      "step": 18010
+    },
+    {
+      "epoch": 1.3427719821162445,
+      "grad_norm": 2.657008171081543,
+      "learning_rate": 0.0002,
+      "loss": 2.5917,
+      "step": 18020
+    },
+    {
+      "epoch": 1.3435171385991058,
+      "grad_norm": 2.4030990600585938,
+      "learning_rate": 0.0002,
+      "loss": 2.4942,
+      "step": 18030
+    },
+    {
+      "epoch": 1.3442622950819672,
+      "grad_norm": 2.760948657989502,
+      "learning_rate": 0.0002,
+      "loss": 2.5241,
+      "step": 18040
+    },
+    {
+      "epoch": 1.3450074515648285,
+      "grad_norm": 2.4696218967437744,
+      "learning_rate": 0.0002,
+      "loss": 2.5653,
+      "step": 18050
+    },
+    {
+      "epoch": 1.34575260804769,
+      "grad_norm": 2.8503217697143555,
+      "learning_rate": 0.0002,
+      "loss": 2.5897,
+      "step": 18060
+    },
+    {
+      "epoch": 1.3464977645305514,
+      "grad_norm": 2.43888258934021,
+      "learning_rate": 0.0002,
+      "loss": 2.5546,
+      "step": 18070
+    },
+    {
+      "epoch": 1.3472429210134127,
+      "grad_norm": 2.212247610092163,
+      "learning_rate": 0.0002,
+      "loss": 2.2402,
+      "step": 18080
+    },
+    {
+      "epoch": 1.3479880774962743,
+      "grad_norm": 2.612032413482666,
+      "learning_rate": 0.0002,
+      "loss": 2.4855,
+      "step": 18090
+    },
+    {
+      "epoch": 1.3487332339791356,
+      "grad_norm": 1.9392924308776855,
+      "learning_rate": 0.0002,
+      "loss": 2.485,
+      "step": 18100
+    },
+    {
+      "epoch": 1.349478390461997,
+      "grad_norm": 2.618718147277832,
+      "learning_rate": 0.0002,
+      "loss": 2.4073,
+      "step": 18110
+    },
+    {
+      "epoch": 1.3502235469448585,
+      "grad_norm": 2.3610072135925293,
+      "learning_rate": 0.0002,
+      "loss": 2.5229,
+      "step": 18120
+    },
+    {
+      "epoch": 1.3509687034277198,
+      "grad_norm": 2.595275640487671,
+      "learning_rate": 0.0002,
+      "loss": 2.4553,
+      "step": 18130
+    },
+    {
+      "epoch": 1.3517138599105811,
+      "grad_norm": 2.991643190383911,
+      "learning_rate": 0.0002,
+      "loss": 2.676,
+      "step": 18140
+    },
+    {
+      "epoch": 1.3524590163934427,
+      "grad_norm": 2.907435655593872,
+      "learning_rate": 0.0002,
+      "loss": 2.5444,
+      "step": 18150
+    },
+    {
+      "epoch": 1.353204172876304,
+      "grad_norm": 2.5445618629455566,
+      "learning_rate": 0.0002,
+      "loss": 2.2842,
+      "step": 18160
+    },
+    {
+      "epoch": 1.3539493293591653,
+      "grad_norm": 2.4656434059143066,
+      "learning_rate": 0.0002,
+      "loss": 2.4092,
+      "step": 18170
+    },
+    {
+      "epoch": 1.354694485842027,
+      "grad_norm": 2.802135467529297,
+      "learning_rate": 0.0002,
+      "loss": 2.546,
+      "step": 18180
+    },
+    {
+      "epoch": 1.3554396423248882,
+      "grad_norm": 2.503904342651367,
+      "learning_rate": 0.0002,
+      "loss": 2.1597,
+      "step": 18190
+    },
+    {
+      "epoch": 1.3561847988077496,
+      "grad_norm": 2.6705873012542725,
+      "learning_rate": 0.0002,
+      "loss": 2.5067,
+      "step": 18200
+    },
+    {
+      "epoch": 1.3569299552906111,
+      "grad_norm": 2.7576780319213867,
+      "learning_rate": 0.0002,
+      "loss": 2.5998,
+      "step": 18210
+    },
+    {
+      "epoch": 1.3576751117734724,
+      "grad_norm": 2.821742296218872,
+      "learning_rate": 0.0002,
+      "loss": 2.5604,
+      "step": 18220
+    },
+    {
+      "epoch": 1.3584202682563338,
+      "grad_norm": 2.8857474327087402,
+      "learning_rate": 0.0002,
+      "loss": 2.4744,
+      "step": 18230
+    },
+    {
+      "epoch": 1.3591654247391953,
+      "grad_norm": 2.5896153450012207,
+      "learning_rate": 0.0002,
+      "loss": 2.4459,
+      "step": 18240
+    },
+    {
+      "epoch": 1.3599105812220567,
+      "grad_norm": 2.5406441688537598,
+      "learning_rate": 0.0002,
+      "loss": 2.6658,
+      "step": 18250
+    },
+    {
+      "epoch": 1.360655737704918,
+      "grad_norm": 2.489504337310791,
+      "learning_rate": 0.0002,
+      "loss": 2.4687,
+      "step": 18260
+    },
+    {
+      "epoch": 1.3614008941877795,
+      "grad_norm": 2.5663485527038574,
+      "learning_rate": 0.0002,
+      "loss": 2.4253,
+      "step": 18270
+    },
+    {
+      "epoch": 1.3621460506706409,
+      "grad_norm": 2.5449881553649902,
+      "learning_rate": 0.0002,
+      "loss": 2.4748,
+      "step": 18280
+    },
+    {
+      "epoch": 1.3628912071535022,
+      "grad_norm": 2.5304982662200928,
+      "learning_rate": 0.0002,
+      "loss": 2.5618,
+      "step": 18290
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 2.8851704597473145,
+      "learning_rate": 0.0002,
+      "loss": 2.3942,
+      "step": 18300
+    },
+    {
+      "epoch": 1.364381520119225,
+      "grad_norm": 2.3670241832733154,
+      "learning_rate": 0.0002,
+      "loss": 2.5914,
+      "step": 18310
+    },
+    {
+      "epoch": 1.3651266766020864,
+      "grad_norm": 2.421011447906494,
+      "learning_rate": 0.0002,
+      "loss": 2.4694,
+      "step": 18320
+    },
+    {
+      "epoch": 1.365871833084948,
+      "grad_norm": 2.5096373558044434,
+      "learning_rate": 0.0002,
+      "loss": 2.5616,
+      "step": 18330
+    },
+    {
+      "epoch": 1.3666169895678093,
+      "grad_norm": 2.6197783946990967,
+      "learning_rate": 0.0002,
+      "loss": 2.351,
+      "step": 18340
+    },
+    {
+      "epoch": 1.3673621460506706,
+      "grad_norm": 2.3040313720703125,
+      "learning_rate": 0.0002,
+      "loss": 2.5583,
+      "step": 18350
+    },
+    {
+      "epoch": 1.368107302533532,
+      "grad_norm": 2.4461286067962646,
+      "learning_rate": 0.0002,
+      "loss": 2.5253,
+      "step": 18360
+    },
+    {
+      "epoch": 1.3688524590163935,
+      "grad_norm": 2.6195592880249023,
+      "learning_rate": 0.0002,
+      "loss": 2.4353,
+      "step": 18370
+    },
+    {
+      "epoch": 1.3695976154992549,
+      "grad_norm": 2.711920976638794,
+      "learning_rate": 0.0002,
+      "loss": 2.4055,
+      "step": 18380
+    },
+    {
+      "epoch": 1.3703427719821162,
+      "grad_norm": 2.2745280265808105,
+      "learning_rate": 0.0002,
+      "loss": 2.2759,
+      "step": 18390
+    },
+    {
+      "epoch": 1.3710879284649775,
+      "grad_norm": 2.446627140045166,
+      "learning_rate": 0.0002,
+      "loss": 2.3023,
+      "step": 18400
+    },
+    {
+      "epoch": 1.371833084947839,
+      "grad_norm": 2.3650870323181152,
+      "learning_rate": 0.0002,
+      "loss": 2.1179,
+      "step": 18410
+    },
+    {
+      "epoch": 1.3725782414307004,
+      "grad_norm": 2.3850514888763428,
+      "learning_rate": 0.0002,
+      "loss": 2.5995,
+      "step": 18420
+    },
+    {
+      "epoch": 1.3733233979135617,
+      "grad_norm": 2.4336416721343994,
+      "learning_rate": 0.0002,
+      "loss": 2.5989,
+      "step": 18430
+    },
+    {
+      "epoch": 1.3740685543964233,
+      "grad_norm": 2.4156320095062256,
+      "learning_rate": 0.0002,
+      "loss": 2.4384,
+      "step": 18440
+    },
+    {
+      "epoch": 1.3748137108792846,
+      "grad_norm": 2.216463327407837,
+      "learning_rate": 0.0002,
+      "loss": 2.288,
+      "step": 18450
+    },
+    {
+      "epoch": 1.375558867362146,
+      "grad_norm": 2.4132537841796875,
+      "learning_rate": 0.0002,
+      "loss": 2.5789,
+      "step": 18460
+    },
+    {
+      "epoch": 1.3763040238450075,
+      "grad_norm": 2.4926469326019287,
+      "learning_rate": 0.0002,
+      "loss": 2.2998,
+      "step": 18470
+    },
+    {
+      "epoch": 1.3770491803278688,
+      "grad_norm": 2.515826940536499,
+      "learning_rate": 0.0002,
+      "loss": 2.4162,
+      "step": 18480
+    },
+    {
+      "epoch": 1.3777943368107302,
+      "grad_norm": 2.26667857170105,
+      "learning_rate": 0.0002,
+      "loss": 2.5093,
+      "step": 18490
+    },
+    {
+      "epoch": 1.3785394932935917,
+      "grad_norm": 2.739278793334961,
+      "learning_rate": 0.0002,
+      "loss": 2.6155,
+      "step": 18500
+    },
+    {
+      "epoch": 1.379284649776453,
+      "grad_norm": 2.4237091541290283,
+      "learning_rate": 0.0002,
+      "loss": 2.5357,
+      "step": 18510
+    },
+    {
+      "epoch": 1.3800298062593144,
+      "grad_norm": 2.349325180053711,
+      "learning_rate": 0.0002,
+      "loss": 2.6382,
+      "step": 18520
+    },
+    {
+      "epoch": 1.380774962742176,
+      "grad_norm": 2.0295040607452393,
+      "learning_rate": 0.0002,
+      "loss": 2.2775,
+      "step": 18530
+    },
+    {
+      "epoch": 1.3815201192250373,
+      "grad_norm": 2.54569411277771,
+      "learning_rate": 0.0002,
+      "loss": 2.6035,
+      "step": 18540
+    },
+    {
+      "epoch": 1.3822652757078986,
+      "grad_norm": 2.625166416168213,
+      "learning_rate": 0.0002,
+      "loss": 2.3252,
+      "step": 18550
+    },
+    {
+      "epoch": 1.3830104321907601,
+      "grad_norm": 2.5229341983795166,
+      "learning_rate": 0.0002,
+      "loss": 2.5045,
+      "step": 18560
+    },
+    {
+      "epoch": 1.3837555886736215,
+      "grad_norm": 2.5632095336914062,
+      "learning_rate": 0.0002,
+      "loss": 2.4786,
+      "step": 18570
+    },
+    {
+      "epoch": 1.3845007451564828,
+      "grad_norm": 2.6461541652679443,
+      "learning_rate": 0.0002,
+      "loss": 2.3556,
+      "step": 18580
+    },
+    {
+      "epoch": 1.3852459016393444,
+      "grad_norm": 2.256908893585205,
+      "learning_rate": 0.0002,
+      "loss": 2.4763,
+      "step": 18590
+    },
+    {
+      "epoch": 1.3859910581222057,
+      "grad_norm": 2.930579662322998,
+      "learning_rate": 0.0002,
+      "loss": 2.3258,
+      "step": 18600
+    },
+    {
+      "epoch": 1.386736214605067,
+      "grad_norm": 2.400162696838379,
+      "learning_rate": 0.0002,
+      "loss": 2.5043,
+      "step": 18610
+    },
+    {
+      "epoch": 1.3874813710879286,
+      "grad_norm": 2.6638431549072266,
+      "learning_rate": 0.0002,
+      "loss": 2.3977,
+      "step": 18620
+    },
+    {
+      "epoch": 1.38822652757079,
+      "grad_norm": 2.60463285446167,
+      "learning_rate": 0.0002,
+      "loss": 2.4366,
+      "step": 18630
+    },
+    {
+      "epoch": 1.3889716840536512,
+      "grad_norm": 2.412717342376709,
+      "learning_rate": 0.0002,
+      "loss": 2.4246,
+      "step": 18640
+    },
+    {
+      "epoch": 1.3897168405365128,
+      "grad_norm": 3.0523712635040283,
+      "learning_rate": 0.0002,
+      "loss": 2.4677,
+      "step": 18650
+    },
+    {
+      "epoch": 1.3904619970193741,
+      "grad_norm": 2.9874138832092285,
+      "learning_rate": 0.0002,
+      "loss": 2.612,
+      "step": 18660
+    },
+    {
+      "epoch": 1.3912071535022354,
+      "grad_norm": 2.532456398010254,
+      "learning_rate": 0.0002,
+      "loss": 2.4691,
+      "step": 18670
+    },
+    {
+      "epoch": 1.391952309985097,
+      "grad_norm": 2.7136852741241455,
+      "learning_rate": 0.0002,
+      "loss": 2.4155,
+      "step": 18680
+    },
+    {
+      "epoch": 1.3926974664679583,
+      "grad_norm": 2.5134053230285645,
+      "learning_rate": 0.0002,
+      "loss": 2.5658,
+      "step": 18690
+    },
+    {
+      "epoch": 1.3934426229508197,
+      "grad_norm": 2.934316396713257,
+      "learning_rate": 0.0002,
+      "loss": 2.5088,
+      "step": 18700
+    },
+    {
+      "epoch": 1.394187779433681,
+      "grad_norm": 2.23718523979187,
+      "learning_rate": 0.0002,
+      "loss": 2.3933,
+      "step": 18710
+    },
+    {
+      "epoch": 1.3949329359165425,
+      "grad_norm": 2.1990694999694824,
+      "learning_rate": 0.0002,
+      "loss": 2.3508,
+      "step": 18720
+    },
+    {
+      "epoch": 1.3956780923994039,
+      "grad_norm": 2.4167299270629883,
+      "learning_rate": 0.0002,
+      "loss": 2.5543,
+      "step": 18730
+    },
+    {
+      "epoch": 1.3964232488822652,
+      "grad_norm": 2.468496561050415,
+      "learning_rate": 0.0002,
+      "loss": 2.4384,
+      "step": 18740
+    },
+    {
+      "epoch": 1.3971684053651265,
+      "grad_norm": 2.5203256607055664,
+      "learning_rate": 0.0002,
+      "loss": 2.4853,
+      "step": 18750
+    },
+    {
+      "epoch": 1.397913561847988,
+      "grad_norm": 2.669184446334839,
+      "learning_rate": 0.0002,
+      "loss": 2.4931,
+      "step": 18760
+    },
+    {
+      "epoch": 1.3986587183308494,
+      "grad_norm": 2.7965617179870605,
+      "learning_rate": 0.0002,
+      "loss": 2.6282,
+      "step": 18770
+    },
+    {
+      "epoch": 1.3994038748137108,
+      "grad_norm": 2.606688976287842,
+      "learning_rate": 0.0002,
+      "loss": 2.5169,
+      "step": 18780
+    },
+    {
+      "epoch": 1.4001490312965723,
+      "grad_norm": 3.333808422088623,
+      "learning_rate": 0.0002,
+      "loss": 2.4799,
+      "step": 18790
+    },
+    {
+      "epoch": 1.4008941877794336,
+      "grad_norm": 2.470353841781616,
+      "learning_rate": 0.0002,
+      "loss": 2.6148,
+      "step": 18800
+    },
+    {
+      "epoch": 1.401639344262295,
+      "grad_norm": 3.7731544971466064,
+      "learning_rate": 0.0002,
+      "loss": 2.5979,
+      "step": 18810
+    },
+    {
+      "epoch": 1.4023845007451565,
+      "grad_norm": 2.7229580879211426,
+      "learning_rate": 0.0002,
+      "loss": 2.5466,
+      "step": 18820
+    },
+    {
+      "epoch": 1.4031296572280179,
+      "grad_norm": 2.463998556137085,
+      "learning_rate": 0.0002,
+      "loss": 2.5742,
+      "step": 18830
+    },
+    {
+      "epoch": 1.4038748137108792,
+      "grad_norm": 2.2875285148620605,
+      "learning_rate": 0.0002,
+      "loss": 2.3219,
+      "step": 18840
+    },
+    {
+      "epoch": 1.4046199701937407,
+      "grad_norm": 2.479088068008423,
+      "learning_rate": 0.0002,
+      "loss": 2.2547,
+      "step": 18850
+    },
+    {
+      "epoch": 1.405365126676602,
+      "grad_norm": 2.7648444175720215,
+      "learning_rate": 0.0002,
+      "loss": 2.4328,
+      "step": 18860
+    },
+    {
+      "epoch": 1.4061102831594634,
+      "grad_norm": 2.649199962615967,
+      "learning_rate": 0.0002,
+      "loss": 2.5119,
+      "step": 18870
+    },
+    {
+      "epoch": 1.406855439642325,
+      "grad_norm": 2.6668455600738525,
+      "learning_rate": 0.0002,
+      "loss": 2.3982,
+      "step": 18880
+    },
+    {
+      "epoch": 1.4076005961251863,
+      "grad_norm": 2.207362651824951,
+      "learning_rate": 0.0002,
+      "loss": 2.5986,
+      "step": 18890
+    },
+    {
+      "epoch": 1.4083457526080476,
+      "grad_norm": 2.383084535598755,
+      "learning_rate": 0.0002,
+      "loss": 2.3486,
+      "step": 18900
+    },
+    {
+      "epoch": 1.4090909090909092,
+      "grad_norm": 2.490837574005127,
+      "learning_rate": 0.0002,
+      "loss": 2.4754,
+      "step": 18910
+    },
+    {
+      "epoch": 1.4098360655737705,
+      "grad_norm": 2.479707717895508,
+      "learning_rate": 0.0002,
+      "loss": 2.4408,
+      "step": 18920
+    },
+    {
+      "epoch": 1.4105812220566318,
+      "grad_norm": 2.9965310096740723,
+      "learning_rate": 0.0002,
+      "loss": 2.5574,
+      "step": 18930
+    },
+    {
+      "epoch": 1.4113263785394934,
+      "grad_norm": 2.707078456878662,
+      "learning_rate": 0.0002,
+      "loss": 2.4176,
+      "step": 18940
+    },
+    {
+      "epoch": 1.4120715350223547,
+      "grad_norm": 2.5538723468780518,
+      "learning_rate": 0.0002,
+      "loss": 2.4505,
+      "step": 18950
+    },
+    {
+      "epoch": 1.412816691505216,
+      "grad_norm": 2.399869441986084,
+      "learning_rate": 0.0002,
+      "loss": 2.391,
+      "step": 18960
+    },
+    {
+      "epoch": 1.4135618479880776,
+      "grad_norm": 2.150087594985962,
+      "learning_rate": 0.0002,
+      "loss": 2.5095,
+      "step": 18970
+    },
+    {
+      "epoch": 1.414307004470939,
+      "grad_norm": 2.7738189697265625,
+      "learning_rate": 0.0002,
+      "loss": 2.4873,
+      "step": 18980
+    },
+    {
+      "epoch": 1.4150521609538003,
+      "grad_norm": 2.8907742500305176,
+      "learning_rate": 0.0002,
+      "loss": 2.7251,
+      "step": 18990
+    },
+    {
+      "epoch": 1.4157973174366618,
+      "grad_norm": 2.500542640686035,
+      "learning_rate": 0.0002,
+      "loss": 2.6472,
+      "step": 19000
+    },
+    {
+      "epoch": 1.4165424739195231,
+      "grad_norm": 2.930190324783325,
+      "learning_rate": 0.0002,
+      "loss": 2.4754,
+      "step": 19010
+    },
+    {
+      "epoch": 1.4172876304023845,
+      "grad_norm": 2.717930793762207,
+      "learning_rate": 0.0002,
+      "loss": 2.5804,
+      "step": 19020
+    },
+    {
+      "epoch": 1.418032786885246,
+      "grad_norm": 2.6405129432678223,
+      "learning_rate": 0.0002,
+      "loss": 2.4247,
+      "step": 19030
+    },
+    {
+      "epoch": 1.4187779433681074,
+      "grad_norm": 2.3362009525299072,
+      "learning_rate": 0.0002,
+      "loss": 2.4234,
+      "step": 19040
+    },
+    {
+      "epoch": 1.4195230998509687,
+      "grad_norm": 2.1805837154388428,
+      "learning_rate": 0.0002,
+      "loss": 2.4447,
+      "step": 19050
+    },
+    {
+      "epoch": 1.42026825633383,
+      "grad_norm": 2.530463695526123,
+      "learning_rate": 0.0002,
+      "loss": 2.5781,
+      "step": 19060
+    },
+    {
+      "epoch": 1.4210134128166916,
+      "grad_norm": 2.360027313232422,
+      "learning_rate": 0.0002,
+      "loss": 2.4854,
+      "step": 19070
+    },
+    {
+      "epoch": 1.421758569299553,
+      "grad_norm": 2.552933931350708,
+      "learning_rate": 0.0002,
+      "loss": 2.4995,
+      "step": 19080
+    },
+    {
+      "epoch": 1.4225037257824142,
+      "grad_norm": 2.643507719039917,
+      "learning_rate": 0.0002,
+      "loss": 2.4029,
+      "step": 19090
+    },
+    {
+      "epoch": 1.4232488822652756,
+      "grad_norm": 2.3608195781707764,
+      "learning_rate": 0.0002,
+      "loss": 2.3169,
+      "step": 19100
+    },
+    {
+      "epoch": 1.4239940387481371,
+      "grad_norm": 2.282942295074463,
+      "learning_rate": 0.0002,
+      "loss": 2.632,
+      "step": 19110
+    },
+    {
+      "epoch": 1.4247391952309985,
+      "grad_norm": 2.8628063201904297,
+      "learning_rate": 0.0002,
+      "loss": 2.3114,
+      "step": 19120
+    },
+    {
+      "epoch": 1.4254843517138598,
+      "grad_norm": 2.357126235961914,
+      "learning_rate": 0.0002,
+      "loss": 2.483,
+      "step": 19130
+    },
+    {
+      "epoch": 1.4262295081967213,
+      "grad_norm": 2.1612229347229004,
+      "learning_rate": 0.0002,
+      "loss": 2.4844,
+      "step": 19140
+    },
+    {
+      "epoch": 1.4269746646795827,
+      "grad_norm": 2.248215436935425,
+      "learning_rate": 0.0002,
+      "loss": 2.3824,
+      "step": 19150
+    },
+    {
+      "epoch": 1.427719821162444,
+      "grad_norm": 1.9331451654434204,
+      "learning_rate": 0.0002,
+      "loss": 2.3143,
+      "step": 19160
+    },
+    {
+      "epoch": 1.4284649776453056,
+      "grad_norm": 2.4009180068969727,
+      "learning_rate": 0.0002,
+      "loss": 2.382,
+      "step": 19170
+    },
+    {
+      "epoch": 1.4292101341281669,
+      "grad_norm": 2.6418373584747314,
+      "learning_rate": 0.0002,
+      "loss": 2.4591,
+      "step": 19180
+    },
+    {
+      "epoch": 1.4299552906110282,
+      "grad_norm": 2.737656593322754,
+      "learning_rate": 0.0002,
+      "loss": 2.3134,
+      "step": 19190
+    },
+    {
+      "epoch": 1.4307004470938898,
+      "grad_norm": 2.2396743297576904,
+      "learning_rate": 0.0002,
+      "loss": 2.4965,
+      "step": 19200
+    },
+    {
+      "epoch": 1.431445603576751,
+      "grad_norm": 2.542550802230835,
+      "learning_rate": 0.0002,
+      "loss": 2.4396,
+      "step": 19210
+    },
+    {
+      "epoch": 1.4321907600596124,
+      "grad_norm": 2.6923890113830566,
+      "learning_rate": 0.0002,
+      "loss": 2.3213,
+      "step": 19220
+    },
+    {
+      "epoch": 1.432935916542474,
+      "grad_norm": 2.479887008666992,
+      "learning_rate": 0.0002,
+      "loss": 2.528,
+      "step": 19230
+    },
+    {
+      "epoch": 1.4336810730253353,
+      "grad_norm": 2.2909326553344727,
+      "learning_rate": 0.0002,
+      "loss": 2.4784,
+      "step": 19240
+    },
+    {
+      "epoch": 1.4344262295081966,
+      "grad_norm": 2.543410301208496,
+      "learning_rate": 0.0002,
+      "loss": 2.603,
+      "step": 19250
+    },
+    {
+      "epoch": 1.4351713859910582,
+      "grad_norm": 2.4764413833618164,
+      "learning_rate": 0.0002,
+      "loss": 2.3813,
+      "step": 19260
+    },
+    {
+      "epoch": 1.4359165424739195,
+      "grad_norm": 2.4031872749328613,
+      "learning_rate": 0.0002,
+      "loss": 2.4507,
+      "step": 19270
+    },
+    {
+      "epoch": 1.4366616989567809,
+      "grad_norm": 2.174621105194092,
+      "learning_rate": 0.0002,
+      "loss": 2.505,
+      "step": 19280
+    },
+    {
+      "epoch": 1.4374068554396424,
+      "grad_norm": 2.6859323978424072,
+      "learning_rate": 0.0002,
+      "loss": 2.5794,
+      "step": 19290
+    },
+    {
+      "epoch": 1.4381520119225037,
+      "grad_norm": 2.5702786445617676,
+      "learning_rate": 0.0002,
+      "loss": 2.429,
+      "step": 19300
+    },
+    {
+      "epoch": 1.438897168405365,
+      "grad_norm": 2.4313433170318604,
+      "learning_rate": 0.0002,
+      "loss": 2.4518,
+      "step": 19310
+    },
+    {
+      "epoch": 1.4396423248882266,
+      "grad_norm": 2.800537347793579,
+      "learning_rate": 0.0002,
+      "loss": 2.4637,
+      "step": 19320
+    },
+    {
+      "epoch": 1.440387481371088,
+      "grad_norm": 2.221001625061035,
+      "learning_rate": 0.0002,
+      "loss": 2.5614,
+      "step": 19330
+    },
+    {
+      "epoch": 1.4411326378539493,
+      "grad_norm": 2.630506753921509,
+      "learning_rate": 0.0002,
+      "loss": 2.5497,
+      "step": 19340
+    },
+    {
+      "epoch": 1.4418777943368108,
+      "grad_norm": 3.166796922683716,
+      "learning_rate": 0.0002,
+      "loss": 2.4768,
+      "step": 19350
+    },
+    {
+      "epoch": 1.4426229508196722,
+      "grad_norm": 2.3211870193481445,
+      "learning_rate": 0.0002,
+      "loss": 2.4818,
+      "step": 19360
+    },
+    {
+      "epoch": 1.4433681073025335,
+      "grad_norm": 2.428395986557007,
+      "learning_rate": 0.0002,
+      "loss": 2.2887,
+      "step": 19370
+    },
+    {
+      "epoch": 1.444113263785395,
+      "grad_norm": 2.2605369091033936,
+      "learning_rate": 0.0002,
+      "loss": 2.5042,
+      "step": 19380
+    },
+    {
+      "epoch": 1.4448584202682564,
+      "grad_norm": 2.404771089553833,
+      "learning_rate": 0.0002,
+      "loss": 2.5354,
+      "step": 19390
+    },
+    {
+      "epoch": 1.4456035767511177,
+      "grad_norm": 2.3696742057800293,
+      "learning_rate": 0.0002,
+      "loss": 2.5734,
+      "step": 19400
+    },
+    {
+      "epoch": 1.446348733233979,
+      "grad_norm": 2.0790860652923584,
+      "learning_rate": 0.0002,
+      "loss": 2.379,
+      "step": 19410
+    },
+    {
+      "epoch": 1.4470938897168406,
+      "grad_norm": 2.8864376544952393,
+      "learning_rate": 0.0002,
+      "loss": 2.5431,
+      "step": 19420
+    },
+    {
+      "epoch": 1.447839046199702,
+      "grad_norm": 2.4749741554260254,
+      "learning_rate": 0.0002,
+      "loss": 2.7492,
+      "step": 19430
+    },
+    {
+      "epoch": 1.4485842026825633,
+      "grad_norm": 2.4821410179138184,
+      "learning_rate": 0.0002,
+      "loss": 2.6195,
+      "step": 19440
+    },
+    {
+      "epoch": 1.4493293591654246,
+      "grad_norm": 2.2254769802093506,
+      "learning_rate": 0.0002,
+      "loss": 2.5504,
+      "step": 19450
+    },
+    {
+      "epoch": 1.4500745156482862,
+      "grad_norm": 2.85517954826355,
+      "learning_rate": 0.0002,
+      "loss": 2.5548,
+      "step": 19460
+    },
+    {
+      "epoch": 1.4508196721311475,
+      "grad_norm": 2.567664384841919,
+      "learning_rate": 0.0002,
+      "loss": 2.5434,
+      "step": 19470
+    },
+    {
+      "epoch": 1.4515648286140088,
+      "grad_norm": 2.842428684234619,
+      "learning_rate": 0.0002,
+      "loss": 2.6169,
+      "step": 19480
+    },
+    {
+      "epoch": 1.4523099850968704,
+      "grad_norm": 2.630206346511841,
+      "learning_rate": 0.0002,
+      "loss": 2.46,
+      "step": 19490
+    },
+    {
+      "epoch": 1.4530551415797317,
+      "grad_norm": 2.2468862533569336,
+      "learning_rate": 0.0002,
+      "loss": 2.5329,
+      "step": 19500
+    },
+    {
+      "epoch": 1.453800298062593,
+      "grad_norm": 2.688547134399414,
+      "learning_rate": 0.0002,
+      "loss": 2.4376,
+      "step": 19510
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 2.439633846282959,
+      "learning_rate": 0.0002,
+      "loss": 2.572,
+      "step": 19520
+    },
+    {
+      "epoch": 1.455290611028316,
+      "grad_norm": 2.374743700027466,
+      "learning_rate": 0.0002,
+      "loss": 2.579,
+      "step": 19530
+    },
+    {
+      "epoch": 1.4560357675111772,
+      "grad_norm": 2.5189223289489746,
+      "learning_rate": 0.0002,
+      "loss": 2.4794,
+      "step": 19540
+    },
+    {
+      "epoch": 1.4567809239940388,
+      "grad_norm": 2.6235687732696533,
+      "learning_rate": 0.0002,
+      "loss": 2.4074,
+      "step": 19550
+    },
+    {
+      "epoch": 1.4575260804769001,
+      "grad_norm": 2.6033565998077393,
+      "learning_rate": 0.0002,
+      "loss": 2.4671,
+      "step": 19560
+    },
+    {
+      "epoch": 1.4582712369597615,
+      "grad_norm": 2.5955042839050293,
+      "learning_rate": 0.0002,
+      "loss": 2.4777,
+      "step": 19570
+    },
+    {
+      "epoch": 1.459016393442623,
+      "grad_norm": 2.8207786083221436,
+      "learning_rate": 0.0002,
+      "loss": 2.5766,
+      "step": 19580
+    },
+    {
+      "epoch": 1.4597615499254843,
+      "grad_norm": 2.597846269607544,
+      "learning_rate": 0.0002,
+      "loss": 2.4987,
+      "step": 19590
+    },
+    {
+      "epoch": 1.4605067064083457,
+      "grad_norm": 2.6083250045776367,
+      "learning_rate": 0.0002,
+      "loss": 2.6282,
+      "step": 19600
+    },
+    {
+      "epoch": 1.4612518628912072,
+      "grad_norm": 2.8631696701049805,
+      "learning_rate": 0.0002,
+      "loss": 2.3029,
+      "step": 19610
+    },
+    {
+      "epoch": 1.4619970193740686,
+      "grad_norm": 2.701707124710083,
+      "learning_rate": 0.0002,
+      "loss": 2.4164,
+      "step": 19620
+    },
+    {
+      "epoch": 1.46274217585693,
+      "grad_norm": 2.4737026691436768,
+      "learning_rate": 0.0002,
+      "loss": 2.4605,
+      "step": 19630
+    },
+    {
+      "epoch": 1.4634873323397914,
+      "grad_norm": 2.5299856662750244,
+      "learning_rate": 0.0002,
+      "loss": 2.2568,
+      "step": 19640
+    },
+    {
+      "epoch": 1.4642324888226528,
+      "grad_norm": 2.5376803874969482,
+      "learning_rate": 0.0002,
+      "loss": 2.4407,
+      "step": 19650
+    },
+    {
+      "epoch": 1.464977645305514,
+      "grad_norm": 2.8622679710388184,
+      "learning_rate": 0.0002,
+      "loss": 2.4903,
+      "step": 19660
+    },
+    {
+      "epoch": 1.4657228017883757,
+      "grad_norm": 2.7106306552886963,
+      "learning_rate": 0.0002,
+      "loss": 2.4316,
+      "step": 19670
+    },
+    {
+      "epoch": 1.466467958271237,
+      "grad_norm": 2.3677496910095215,
+      "learning_rate": 0.0002,
+      "loss": 2.4907,
+      "step": 19680
+    },
+    {
+      "epoch": 1.4672131147540983,
+      "grad_norm": 2.7566051483154297,
+      "learning_rate": 0.0002,
+      "loss": 2.5288,
+      "step": 19690
+    },
+    {
+      "epoch": 1.4679582712369599,
+      "grad_norm": 2.2991981506347656,
+      "learning_rate": 0.0002,
+      "loss": 2.4065,
+      "step": 19700
+    },
+    {
+      "epoch": 1.4687034277198212,
+      "grad_norm": 2.579111337661743,
+      "learning_rate": 0.0002,
+      "loss": 2.5721,
+      "step": 19710
+    },
+    {
+      "epoch": 1.4694485842026825,
+      "grad_norm": 2.5662269592285156,
+      "learning_rate": 0.0002,
+      "loss": 2.5849,
+      "step": 19720
+    },
+    {
+      "epoch": 1.470193740685544,
+      "grad_norm": 2.307398796081543,
+      "learning_rate": 0.0002,
+      "loss": 2.3272,
+      "step": 19730
+    },
+    {
+      "epoch": 1.4709388971684054,
+      "grad_norm": 2.695326566696167,
+      "learning_rate": 0.0002,
+      "loss": 2.6097,
+      "step": 19740
+    },
+    {
+      "epoch": 1.4716840536512668,
+      "grad_norm": 2.547471523284912,
+      "learning_rate": 0.0002,
+      "loss": 2.6456,
+      "step": 19750
+    },
+    {
+      "epoch": 1.472429210134128,
+      "grad_norm": 2.5473928451538086,
+      "learning_rate": 0.0002,
+      "loss": 2.6685,
+      "step": 19760
+    },
+    {
+      "epoch": 1.4731743666169896,
+      "grad_norm": 2.6228318214416504,
+      "learning_rate": 0.0002,
+      "loss": 2.3921,
+      "step": 19770
+    },
+    {
+      "epoch": 1.473919523099851,
+      "grad_norm": 2.489250659942627,
+      "learning_rate": 0.0002,
+      "loss": 2.3033,
+      "step": 19780
+    },
+    {
+      "epoch": 1.4746646795827123,
+      "grad_norm": 2.4918441772460938,
+      "learning_rate": 0.0002,
+      "loss": 2.6749,
+      "step": 19790
+    },
+    {
+      "epoch": 1.4754098360655736,
+      "grad_norm": 2.6294076442718506,
+      "learning_rate": 0.0002,
+      "loss": 2.5153,
+      "step": 19800
+    },
+    {
+      "epoch": 1.4761549925484352,
+      "grad_norm": 2.7032785415649414,
+      "learning_rate": 0.0002,
+      "loss": 2.4346,
+      "step": 19810
+    },
+    {
+      "epoch": 1.4769001490312965,
+      "grad_norm": 3.086024761199951,
+      "learning_rate": 0.0002,
+      "loss": 2.3755,
+      "step": 19820
+    },
+    {
+      "epoch": 1.4776453055141578,
+      "grad_norm": 2.7240495681762695,
+      "learning_rate": 0.0002,
+      "loss": 2.4115,
+      "step": 19830
+    },
+    {
+      "epoch": 1.4783904619970194,
+      "grad_norm": 2.3864595890045166,
+      "learning_rate": 0.0002,
+      "loss": 2.4623,
+      "step": 19840
+    },
+    {
+      "epoch": 1.4791356184798807,
+      "grad_norm": 2.517160177230835,
+      "learning_rate": 0.0002,
+      "loss": 2.2625,
+      "step": 19850
+    },
+    {
+      "epoch": 1.479880774962742,
+      "grad_norm": 2.8101651668548584,
+      "learning_rate": 0.0002,
+      "loss": 2.3999,
+      "step": 19860
+    },
+    {
+      "epoch": 1.4806259314456036,
+      "grad_norm": 2.6499056816101074,
+      "learning_rate": 0.0002,
+      "loss": 2.4251,
+      "step": 19870
+    },
+    {
+      "epoch": 1.481371087928465,
+      "grad_norm": 2.563832998275757,
+      "learning_rate": 0.0002,
+      "loss": 2.3608,
+      "step": 19880
+    },
+    {
+      "epoch": 1.4821162444113263,
+      "grad_norm": 2.905278205871582,
+      "learning_rate": 0.0002,
+      "loss": 2.4904,
+      "step": 19890
+    },
+    {
+      "epoch": 1.4828614008941878,
+      "grad_norm": 2.7486226558685303,
+      "learning_rate": 0.0002,
+      "loss": 2.5536,
+      "step": 19900
+    },
+    {
+      "epoch": 1.4836065573770492,
+      "grad_norm": 3.1326828002929688,
+      "learning_rate": 0.0002,
+      "loss": 2.4822,
+      "step": 19910
+    },
+    {
+      "epoch": 1.4843517138599105,
+      "grad_norm": 2.983778476715088,
+      "learning_rate": 0.0002,
+      "loss": 2.5207,
+      "step": 19920
+    },
+    {
+      "epoch": 1.485096870342772,
+      "grad_norm": 2.391115427017212,
+      "learning_rate": 0.0002,
+      "loss": 2.4728,
+      "step": 19930
+    },
+    {
+      "epoch": 1.4858420268256334,
+      "grad_norm": 2.789459466934204,
+      "learning_rate": 0.0002,
+      "loss": 2.3897,
+      "step": 19940
+    },
+    {
+      "epoch": 1.4865871833084947,
+      "grad_norm": 2.1479434967041016,
+      "learning_rate": 0.0002,
+      "loss": 2.4227,
+      "step": 19950
+    },
+    {
+      "epoch": 1.4873323397913563,
+      "grad_norm": 2.5585100650787354,
+      "learning_rate": 0.0002,
+      "loss": 2.5672,
+      "step": 19960
+    },
+    {
+      "epoch": 1.4880774962742176,
+      "grad_norm": 2.8484928607940674,
+      "learning_rate": 0.0002,
+      "loss": 2.5627,
+      "step": 19970
+    },
+    {
+      "epoch": 1.488822652757079,
+      "grad_norm": 2.4347646236419678,
+      "learning_rate": 0.0002,
+      "loss": 2.4704,
+      "step": 19980
+    },
+    {
+      "epoch": 1.4895678092399405,
+      "grad_norm": 2.467780351638794,
+      "learning_rate": 0.0002,
+      "loss": 2.4175,
+      "step": 19990
+    },
+    {
+      "epoch": 1.4903129657228018,
+      "grad_norm": 2.4302515983581543,
+      "learning_rate": 0.0002,
+      "loss": 2.602,
+      "step": 20000
+    },
+    {
+      "epoch": 1.4910581222056631,
+      "grad_norm": 2.5407824516296387,
+      "learning_rate": 0.0002,
+      "loss": 2.3174,
+      "step": 20010
+    },
+    {
+      "epoch": 1.4918032786885247,
+      "grad_norm": 2.509092330932617,
+      "learning_rate": 0.0002,
+      "loss": 2.4737,
+      "step": 20020
+    },
+    {
+      "epoch": 1.492548435171386,
+      "grad_norm": 2.512086868286133,
+      "learning_rate": 0.0002,
+      "loss": 2.4616,
+      "step": 20030
+    },
+    {
+      "epoch": 1.4932935916542474,
+      "grad_norm": 2.5899012088775635,
+      "learning_rate": 0.0002,
+      "loss": 2.6209,
+      "step": 20040
+    },
+    {
+      "epoch": 1.494038748137109,
+      "grad_norm": 2.5134575366973877,
+      "learning_rate": 0.0002,
+      "loss": 2.4823,
+      "step": 20050
+    },
+    {
+      "epoch": 1.4947839046199702,
+      "grad_norm": 2.4115593433380127,
+      "learning_rate": 0.0002,
+      "loss": 2.4584,
+      "step": 20060
+    },
+    {
+      "epoch": 1.4955290611028316,
+      "grad_norm": 2.5678627490997314,
+      "learning_rate": 0.0002,
+      "loss": 2.4916,
+      "step": 20070
+    },
+    {
+      "epoch": 1.4962742175856931,
+      "grad_norm": 3.107654333114624,
+      "learning_rate": 0.0002,
+      "loss": 2.3733,
+      "step": 20080
+    },
+    {
+      "epoch": 1.4970193740685545,
+      "grad_norm": 2.282750368118286,
+      "learning_rate": 0.0002,
+      "loss": 2.4918,
+      "step": 20090
+    },
+    {
+      "epoch": 1.4977645305514158,
+      "grad_norm": 2.589705228805542,
+      "learning_rate": 0.0002,
+      "loss": 2.4723,
+      "step": 20100
+    },
+    {
+      "epoch": 1.4985096870342771,
+      "grad_norm": 2.139538288116455,
+      "learning_rate": 0.0002,
+      "loss": 2.463,
+      "step": 20110
+    },
+    {
+      "epoch": 1.4992548435171387,
+      "grad_norm": 2.6076650619506836,
+      "learning_rate": 0.0002,
+      "loss": 2.4384,
+      "step": 20120
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.653719186782837,
+      "learning_rate": 0.0002,
+      "loss": 2.3014,
+      "step": 20130
+    },
+    {
+      "epoch": 1.5007451564828616,
+      "grad_norm": 3.960010051727295,
+      "learning_rate": 0.0002,
+      "loss": 2.407,
+      "step": 20140
+    },
+    {
+      "epoch": 1.5014903129657227,
+      "grad_norm": 2.3678810596466064,
+      "learning_rate": 0.0002,
+      "loss": 2.3797,
+      "step": 20150
+    },
+    {
+      "epoch": 1.5022354694485842,
+      "grad_norm": 2.61295485496521,
+      "learning_rate": 0.0002,
+      "loss": 2.4066,
+      "step": 20160
+    },
+    {
+      "epoch": 1.5029806259314458,
+      "grad_norm": 2.4393184185028076,
+      "learning_rate": 0.0002,
+      "loss": 2.3545,
+      "step": 20170
+    },
+    {
+      "epoch": 1.5037257824143069,
+      "grad_norm": 2.455565929412842,
+      "learning_rate": 0.0002,
+      "loss": 2.4903,
+      "step": 20180
+    },
+    {
+      "epoch": 1.5044709388971684,
+      "grad_norm": 2.7440402507781982,
+      "learning_rate": 0.0002,
+      "loss": 2.5455,
+      "step": 20190
+    },
+    {
+      "epoch": 1.5052160953800298,
+      "grad_norm": 2.6550514698028564,
+      "learning_rate": 0.0002,
+      "loss": 2.4654,
+      "step": 20200
+    },
+    {
+      "epoch": 1.505961251862891,
+      "grad_norm": 2.5063323974609375,
+      "learning_rate": 0.0002,
+      "loss": 2.4515,
+      "step": 20210
+    },
+    {
+      "epoch": 1.5067064083457526,
+      "grad_norm": 2.9979264736175537,
+      "learning_rate": 0.0002,
+      "loss": 2.3564,
+      "step": 20220
+    },
+    {
+      "epoch": 1.507451564828614,
+      "grad_norm": 2.5606510639190674,
+      "learning_rate": 0.0002,
+      "loss": 2.5883,
+      "step": 20230
+    },
+    {
+      "epoch": 1.5081967213114753,
+      "grad_norm": 2.191575050354004,
+      "learning_rate": 0.0002,
+      "loss": 2.4795,
+      "step": 20240
+    },
+    {
+      "epoch": 1.5089418777943369,
+      "grad_norm": 2.5303499698638916,
+      "learning_rate": 0.0002,
+      "loss": 2.41,
+      "step": 20250
+    },
+    {
+      "epoch": 1.5096870342771982,
+      "grad_norm": 2.602943181991577,
+      "learning_rate": 0.0002,
+      "loss": 2.4533,
+      "step": 20260
+    },
+    {
+      "epoch": 1.5104321907600595,
+      "grad_norm": 2.667649030685425,
+      "learning_rate": 0.0002,
+      "loss": 2.5397,
+      "step": 20270
+    },
+    {
+      "epoch": 1.511177347242921,
+      "grad_norm": 2.7517497539520264,
+      "learning_rate": 0.0002,
+      "loss": 2.5885,
+      "step": 20280
+    },
+    {
+      "epoch": 1.5119225037257824,
+      "grad_norm": 2.7016963958740234,
+      "learning_rate": 0.0002,
+      "loss": 2.2389,
+      "step": 20290
+    },
+    {
+      "epoch": 1.5126676602086437,
+      "grad_norm": 3.5088577270507812,
+      "learning_rate": 0.0002,
+      "loss": 2.5337,
+      "step": 20300
+    },
+    {
+      "epoch": 1.5134128166915053,
+      "grad_norm": 2.009373188018799,
+      "learning_rate": 0.0002,
+      "loss": 2.5013,
+      "step": 20310
+    },
+    {
+      "epoch": 1.5141579731743666,
+      "grad_norm": 3.012927770614624,
+      "learning_rate": 0.0002,
+      "loss": 2.487,
+      "step": 20320
+    },
+    {
+      "epoch": 1.514903129657228,
+      "grad_norm": 2.512377977371216,
+      "learning_rate": 0.0002,
+      "loss": 2.6162,
+      "step": 20330
+    },
+    {
+      "epoch": 1.5156482861400895,
+      "grad_norm": 2.7566356658935547,
+      "learning_rate": 0.0002,
+      "loss": 2.4332,
+      "step": 20340
+    },
+    {
+      "epoch": 1.5163934426229508,
+      "grad_norm": 2.620819330215454,
+      "learning_rate": 0.0002,
+      "loss": 2.451,
+      "step": 20350
+    },
+    {
+      "epoch": 1.5171385991058122,
+      "grad_norm": 2.378424882888794,
+      "learning_rate": 0.0002,
+      "loss": 2.3835,
+      "step": 20360
+    },
+    {
+      "epoch": 1.5178837555886737,
+      "grad_norm": 2.6105260848999023,
+      "learning_rate": 0.0002,
+      "loss": 2.5465,
+      "step": 20370
+    },
+    {
+      "epoch": 1.518628912071535,
+      "grad_norm": 2.4131839275360107,
+      "learning_rate": 0.0002,
+      "loss": 2.3061,
+      "step": 20380
+    },
+    {
+      "epoch": 1.5193740685543964,
+      "grad_norm": 2.658151149749756,
+      "learning_rate": 0.0002,
+      "loss": 2.4822,
+      "step": 20390
+    },
+    {
+      "epoch": 1.520119225037258,
+      "grad_norm": 2.169219732284546,
+      "learning_rate": 0.0002,
+      "loss": 2.4033,
+      "step": 20400
+    },
+    {
+      "epoch": 1.5208643815201193,
+      "grad_norm": 2.769595146179199,
+      "learning_rate": 0.0002,
+      "loss": 2.57,
+      "step": 20410
+    },
+    {
+      "epoch": 1.5216095380029806,
+      "grad_norm": 2.6904990673065186,
+      "learning_rate": 0.0002,
+      "loss": 2.4361,
+      "step": 20420
+    },
+    {
+      "epoch": 1.5223546944858422,
+      "grad_norm": 2.26855206489563,
+      "learning_rate": 0.0002,
+      "loss": 2.4516,
+      "step": 20430
+    },
+    {
+      "epoch": 1.5230998509687033,
+      "grad_norm": 2.5474822521209717,
+      "learning_rate": 0.0002,
+      "loss": 2.4087,
+      "step": 20440
+    },
+    {
+      "epoch": 1.5238450074515648,
+      "grad_norm": 2.5763962268829346,
+      "learning_rate": 0.0002,
+      "loss": 2.5081,
+      "step": 20450
+    },
+    {
+      "epoch": 1.5245901639344264,
+      "grad_norm": 2.8625242710113525,
+      "learning_rate": 0.0002,
+      "loss": 2.5304,
+      "step": 20460
+    },
+    {
+      "epoch": 1.5253353204172875,
+      "grad_norm": 2.5730020999908447,
+      "learning_rate": 0.0002,
+      "loss": 2.5999,
+      "step": 20470
+    },
+    {
+      "epoch": 1.526080476900149,
+      "grad_norm": 2.704777717590332,
+      "learning_rate": 0.0002,
+      "loss": 2.5252,
+      "step": 20480
+    },
+    {
+      "epoch": 1.5268256333830106,
+      "grad_norm": 2.8790910243988037,
+      "learning_rate": 0.0002,
+      "loss": 2.4782,
+      "step": 20490
+    },
+    {
+      "epoch": 1.5275707898658717,
+      "grad_norm": 2.793997287750244,
+      "learning_rate": 0.0002,
+      "loss": 2.5271,
+      "step": 20500
+    },
+    {
+      "epoch": 1.5283159463487332,
+      "grad_norm": 2.4312236309051514,
+      "learning_rate": 0.0002,
+      "loss": 2.7423,
+      "step": 20510
+    },
+    {
+      "epoch": 1.5290611028315948,
+      "grad_norm": 2.508166551589966,
+      "learning_rate": 0.0002,
+      "loss": 2.5464,
+      "step": 20520
+    },
+    {
+      "epoch": 1.529806259314456,
+      "grad_norm": 2.4821505546569824,
+      "learning_rate": 0.0002,
+      "loss": 2.5888,
+      "step": 20530
+    },
+    {
+      "epoch": 1.5305514157973175,
+      "grad_norm": 2.0851380825042725,
+      "learning_rate": 0.0002,
+      "loss": 2.1732,
+      "step": 20540
+    },
+    {
+      "epoch": 1.5312965722801788,
+      "grad_norm": 2.1268575191497803,
+      "learning_rate": 0.0002,
+      "loss": 2.5131,
+      "step": 20550
+    },
+    {
+      "epoch": 1.5320417287630401,
+      "grad_norm": 3.2750792503356934,
+      "learning_rate": 0.0002,
+      "loss": 2.3624,
+      "step": 20560
+    },
+    {
+      "epoch": 1.5327868852459017,
+      "grad_norm": 2.771404266357422,
+      "learning_rate": 0.0002,
+      "loss": 2.2204,
+      "step": 20570
+    },
+    {
+      "epoch": 1.533532041728763,
+      "grad_norm": 3.1132712364196777,
+      "learning_rate": 0.0002,
+      "loss": 2.5077,
+      "step": 20580
+    },
+    {
+      "epoch": 1.5342771982116243,
+      "grad_norm": 2.637928009033203,
+      "learning_rate": 0.0002,
+      "loss": 2.5237,
+      "step": 20590
+    },
+    {
+      "epoch": 1.535022354694486,
+      "grad_norm": 3.4630746841430664,
+      "learning_rate": 0.0002,
+      "loss": 2.4241,
+      "step": 20600
+    },
+    {
+      "epoch": 1.5357675111773472,
+      "grad_norm": 3.054542303085327,
+      "learning_rate": 0.0002,
+      "loss": 2.6337,
+      "step": 20610
+    },
+    {
+      "epoch": 1.5365126676602086,
+      "grad_norm": 2.7666263580322266,
+      "learning_rate": 0.0002,
+      "loss": 2.4202,
+      "step": 20620
+    },
+    {
+      "epoch": 1.53725782414307,
+      "grad_norm": 2.8477747440338135,
+      "learning_rate": 0.0002,
+      "loss": 2.4838,
+      "step": 20630
+    },
+    {
+      "epoch": 1.5380029806259314,
+      "grad_norm": 2.61881422996521,
+      "learning_rate": 0.0002,
+      "loss": 2.2668,
+      "step": 20640
+    },
+    {
+      "epoch": 1.5387481371087928,
+      "grad_norm": 2.792616367340088,
+      "learning_rate": 0.0002,
+      "loss": 2.6354,
+      "step": 20650
+    },
+    {
+      "epoch": 1.5394932935916543,
+      "grad_norm": 2.7468929290771484,
+      "learning_rate": 0.0002,
+      "loss": 2.5306,
+      "step": 20660
+    },
+    {
+      "epoch": 1.5402384500745157,
+      "grad_norm": 2.3328466415405273,
+      "learning_rate": 0.0002,
+      "loss": 2.5575,
+      "step": 20670
+    },
+    {
+      "epoch": 1.540983606557377,
+      "grad_norm": 2.406003475189209,
+      "learning_rate": 0.0002,
+      "loss": 2.4008,
+      "step": 20680
+    },
+    {
+      "epoch": 1.5417287630402385,
+      "grad_norm": 2.558863878250122,
+      "learning_rate": 0.0002,
+      "loss": 2.482,
+      "step": 20690
+    },
+    {
+      "epoch": 1.5424739195230999,
+      "grad_norm": 2.814772605895996,
+      "learning_rate": 0.0002,
+      "loss": 2.4789,
+      "step": 20700
+    },
+    {
+      "epoch": 1.5432190760059612,
+      "grad_norm": 2.6838860511779785,
+      "learning_rate": 0.0002,
+      "loss": 2.5793,
+      "step": 20710
+    },
+    {
+      "epoch": 1.5439642324888228,
+      "grad_norm": 2.651115655899048,
+      "learning_rate": 0.0002,
+      "loss": 2.3765,
+      "step": 20720
+    },
+    {
+      "epoch": 1.544709388971684,
+      "grad_norm": 2.7897510528564453,
+      "learning_rate": 0.0002,
+      "loss": 2.5824,
+      "step": 20730
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 2.5708913803100586,
+      "learning_rate": 0.0002,
+      "loss": 2.3102,
+      "step": 20740
+    },
+    {
+      "epoch": 1.546199701937407,
+      "grad_norm": 2.648242235183716,
+      "learning_rate": 0.0002,
+      "loss": 2.4887,
+      "step": 20750
+    },
+    {
+      "epoch": 1.5469448584202683,
+      "grad_norm": 2.639145612716675,
+      "learning_rate": 0.0002,
+      "loss": 2.3971,
+      "step": 20760
+    },
+    {
+      "epoch": 1.5476900149031296,
+      "grad_norm": 2.5040032863616943,
+      "learning_rate": 0.0002,
+      "loss": 2.5788,
+      "step": 20770
+    },
+    {
+      "epoch": 1.5484351713859912,
+      "grad_norm": 2.368126392364502,
+      "learning_rate": 0.0002,
+      "loss": 2.4651,
+      "step": 20780
+    },
+    {
+      "epoch": 1.5491803278688525,
+      "grad_norm": 2.808847427368164,
+      "learning_rate": 0.0002,
+      "loss": 2.4613,
+      "step": 20790
+    },
+    {
+      "epoch": 1.5499254843517138,
+      "grad_norm": 2.632382392883301,
+      "learning_rate": 0.0002,
+      "loss": 2.5921,
+      "step": 20800
+    },
+    {
+      "epoch": 1.5506706408345754,
+      "grad_norm": 2.4991965293884277,
+      "learning_rate": 0.0002,
+      "loss": 2.6584,
+      "step": 20810
+    },
+    {
+      "epoch": 1.5514157973174365,
+      "grad_norm": 2.2685933113098145,
+      "learning_rate": 0.0002,
+      "loss": 2.6257,
+      "step": 20820
+    },
+    {
+      "epoch": 1.552160953800298,
+      "grad_norm": 2.494006872177124,
+      "learning_rate": 0.0002,
+      "loss": 2.4663,
+      "step": 20830
+    },
+    {
+      "epoch": 1.5529061102831596,
+      "grad_norm": 2.848593235015869,
+      "learning_rate": 0.0002,
+      "loss": 2.6125,
+      "step": 20840
+    },
+    {
+      "epoch": 1.5536512667660207,
+      "grad_norm": 2.5735042095184326,
+      "learning_rate": 0.0002,
+      "loss": 2.3904,
+      "step": 20850
+    },
+    {
+      "epoch": 1.5543964232488823,
+      "grad_norm": 2.3776659965515137,
+      "learning_rate": 0.0002,
+      "loss": 2.397,
+      "step": 20860
+    },
+    {
+      "epoch": 1.5551415797317438,
+      "grad_norm": 2.362820863723755,
+      "learning_rate": 0.0002,
+      "loss": 2.499,
+      "step": 20870
+    },
+    {
+      "epoch": 1.555886736214605,
+      "grad_norm": 2.4711267948150635,
+      "learning_rate": 0.0002,
+      "loss": 2.366,
+      "step": 20880
+    },
+    {
+      "epoch": 1.5566318926974665,
+      "grad_norm": 2.8080945014953613,
+      "learning_rate": 0.0002,
+      "loss": 2.4686,
+      "step": 20890
+    },
+    {
+      "epoch": 1.5573770491803278,
+      "grad_norm": 2.614722967147827,
+      "learning_rate": 0.0002,
+      "loss": 2.4868,
+      "step": 20900
+    },
+    {
+      "epoch": 1.5581222056631892,
+      "grad_norm": 2.4597301483154297,
+      "learning_rate": 0.0002,
+      "loss": 2.5285,
+      "step": 20910
+    },
+    {
+      "epoch": 1.5588673621460507,
+      "grad_norm": 2.947422981262207,
+      "learning_rate": 0.0002,
+      "loss": 2.6405,
+      "step": 20920
+    },
+    {
+      "epoch": 1.559612518628912,
+      "grad_norm": 2.4457414150238037,
+      "learning_rate": 0.0002,
+      "loss": 2.6311,
+      "step": 20930
+    },
+    {
+      "epoch": 1.5603576751117734,
+      "grad_norm": 2.5841987133026123,
+      "learning_rate": 0.0002,
+      "loss": 2.605,
+      "step": 20940
+    },
+    {
+      "epoch": 1.561102831594635,
+      "grad_norm": 2.63436222076416,
+      "learning_rate": 0.0002,
+      "loss": 2.5732,
+      "step": 20950
+    },
+    {
+      "epoch": 1.5618479880774963,
+      "grad_norm": 2.898693799972534,
+      "learning_rate": 0.0002,
+      "loss": 2.5725,
+      "step": 20960
+    },
+    {
+      "epoch": 1.5625931445603576,
+      "grad_norm": 2.2288978099823,
+      "learning_rate": 0.0002,
+      "loss": 2.2325,
+      "step": 20970
+    },
+    {
+      "epoch": 1.5633383010432191,
+      "grad_norm": 2.741015672683716,
+      "learning_rate": 0.0002,
+      "loss": 2.406,
+      "step": 20980
+    },
+    {
+      "epoch": 1.5640834575260805,
+      "grad_norm": 2.3615951538085938,
+      "learning_rate": 0.0002,
+      "loss": 2.4796,
+      "step": 20990
+    },
+    {
+      "epoch": 1.5648286140089418,
+      "grad_norm": 2.332751750946045,
+      "learning_rate": 0.0002,
+      "loss": 2.3425,
+      "step": 21000
+    },
+    {
+      "epoch": 1.5655737704918034,
+      "grad_norm": 2.585552453994751,
+      "learning_rate": 0.0002,
+      "loss": 2.5401,
+      "step": 21010
+    },
+    {
+      "epoch": 1.5663189269746647,
+      "grad_norm": 2.596543550491333,
+      "learning_rate": 0.0002,
+      "loss": 2.2485,
+      "step": 21020
+    },
+    {
+      "epoch": 1.567064083457526,
+      "grad_norm": 2.5111849308013916,
+      "learning_rate": 0.0002,
+      "loss": 2.3883,
+      "step": 21030
+    },
+    {
+      "epoch": 1.5678092399403876,
+      "grad_norm": 2.949110984802246,
+      "learning_rate": 0.0002,
+      "loss": 2.626,
+      "step": 21040
+    },
+    {
+      "epoch": 1.568554396423249,
+      "grad_norm": 2.5220885276794434,
+      "learning_rate": 0.0002,
+      "loss": 2.6748,
+      "step": 21050
+    },
+    {
+      "epoch": 1.5692995529061102,
+      "grad_norm": 2.4245173931121826,
+      "learning_rate": 0.0002,
+      "loss": 2.4853,
+      "step": 21060
+    },
+    {
+      "epoch": 1.5700447093889718,
+      "grad_norm": 2.2944424152374268,
+      "learning_rate": 0.0002,
+      "loss": 2.3863,
+      "step": 21070
+    },
+    {
+      "epoch": 1.5707898658718331,
+      "grad_norm": 2.557868003845215,
+      "learning_rate": 0.0002,
+      "loss": 2.5679,
+      "step": 21080
+    },
+    {
+      "epoch": 1.5715350223546944,
+      "grad_norm": 2.418034315109253,
+      "learning_rate": 0.0002,
+      "loss": 2.3904,
+      "step": 21090
+    },
+    {
+      "epoch": 1.572280178837556,
+      "grad_norm": 2.6931450366973877,
+      "learning_rate": 0.0002,
+      "loss": 2.5675,
+      "step": 21100
+    },
+    {
+      "epoch": 1.5730253353204173,
+      "grad_norm": 2.8950247764587402,
+      "learning_rate": 0.0002,
+      "loss": 2.4783,
+      "step": 21110
+    },
+    {
+      "epoch": 1.5737704918032787,
+      "grad_norm": 2.3878729343414307,
+      "learning_rate": 0.0002,
+      "loss": 2.3789,
+      "step": 21120
+    },
+    {
+      "epoch": 1.5745156482861402,
+      "grad_norm": 2.6210811138153076,
+      "learning_rate": 0.0002,
+      "loss": 2.5938,
+      "step": 21130
+    },
+    {
+      "epoch": 1.5752608047690015,
+      "grad_norm": 2.540480852127075,
+      "learning_rate": 0.0002,
+      "loss": 2.6351,
+      "step": 21140
+    },
+    {
+      "epoch": 1.5760059612518629,
+      "grad_norm": 2.7353644371032715,
+      "learning_rate": 0.0002,
+      "loss": 2.3959,
+      "step": 21150
+    },
+    {
+      "epoch": 1.5767511177347244,
+      "grad_norm": 2.555797576904297,
+      "learning_rate": 0.0002,
+      "loss": 2.7066,
+      "step": 21160
+    },
+    {
+      "epoch": 1.5774962742175855,
+      "grad_norm": 2.2357959747314453,
+      "learning_rate": 0.0002,
+      "loss": 2.4063,
+      "step": 21170
+    },
+    {
+      "epoch": 1.578241430700447,
+      "grad_norm": 2.2032105922698975,
+      "learning_rate": 0.0002,
+      "loss": 2.5256,
+      "step": 21180
+    },
+    {
+      "epoch": 1.5789865871833086,
+      "grad_norm": 2.421905994415283,
+      "learning_rate": 0.0002,
+      "loss": 2.5101,
+      "step": 21190
+    },
+    {
+      "epoch": 1.5797317436661698,
+      "grad_norm": 2.3932013511657715,
+      "learning_rate": 0.0002,
+      "loss": 2.5003,
+      "step": 21200
+    },
+    {
+      "epoch": 1.5804769001490313,
+      "grad_norm": 2.751396656036377,
+      "learning_rate": 0.0002,
+      "loss": 2.6125,
+      "step": 21210
+    },
+    {
+      "epoch": 1.5812220566318929,
+      "grad_norm": 2.415769100189209,
+      "learning_rate": 0.0002,
+      "loss": 2.3935,
+      "step": 21220
+    },
+    {
+      "epoch": 1.581967213114754,
+      "grad_norm": 2.7542152404785156,
+      "learning_rate": 0.0002,
+      "loss": 2.5242,
+      "step": 21230
+    },
+    {
+      "epoch": 1.5827123695976155,
+      "grad_norm": 2.4142708778381348,
+      "learning_rate": 0.0002,
+      "loss": 2.5011,
+      "step": 21240
+    },
+    {
+      "epoch": 1.5834575260804769,
+      "grad_norm": 2.716956615447998,
+      "learning_rate": 0.0002,
+      "loss": 2.5144,
+      "step": 21250
+    },
+    {
+      "epoch": 1.5842026825633382,
+      "grad_norm": 2.8631787300109863,
+      "learning_rate": 0.0002,
+      "loss": 2.5118,
+      "step": 21260
+    },
+    {
+      "epoch": 1.5849478390461997,
+      "grad_norm": 2.676856517791748,
+      "learning_rate": 0.0002,
+      "loss": 2.4845,
+      "step": 21270
+    },
+    {
+      "epoch": 1.585692995529061,
+      "grad_norm": 2.477172374725342,
+      "learning_rate": 0.0002,
+      "loss": 2.5516,
+      "step": 21280
+    },
+    {
+      "epoch": 1.5864381520119224,
+      "grad_norm": 2.3883090019226074,
+      "learning_rate": 0.0002,
+      "loss": 2.5904,
+      "step": 21290
+    },
+    {
+      "epoch": 1.587183308494784,
+      "grad_norm": 2.0615110397338867,
+      "learning_rate": 0.0002,
+      "loss": 2.4803,
+      "step": 21300
+    },
+    {
+      "epoch": 1.5879284649776453,
+      "grad_norm": 2.6857569217681885,
+      "learning_rate": 0.0002,
+      "loss": 2.6149,
+      "step": 21310
+    },
+    {
+      "epoch": 1.5886736214605066,
+      "grad_norm": 2.224165678024292,
+      "learning_rate": 0.0002,
+      "loss": 2.4841,
+      "step": 21320
+    },
+    {
+      "epoch": 1.5894187779433682,
+      "grad_norm": 2.5804173946380615,
+      "learning_rate": 0.0002,
+      "loss": 2.7102,
+      "step": 21330
+    },
+    {
+      "epoch": 1.5901639344262295,
+      "grad_norm": 2.559598445892334,
+      "learning_rate": 0.0002,
+      "loss": 2.3667,
+      "step": 21340
+    },
+    {
+      "epoch": 1.5909090909090908,
+      "grad_norm": 2.5414412021636963,
+      "learning_rate": 0.0002,
+      "loss": 2.44,
+      "step": 21350
+    },
+    {
+      "epoch": 1.5916542473919524,
+      "grad_norm": 2.575948476791382,
+      "learning_rate": 0.0002,
+      "loss": 2.6272,
+      "step": 21360
+    },
+    {
+      "epoch": 1.5923994038748137,
+      "grad_norm": 2.558068037033081,
+      "learning_rate": 0.0002,
+      "loss": 2.2373,
+      "step": 21370
+    },
+    {
+      "epoch": 1.593144560357675,
+      "grad_norm": 2.6954293251037598,
+      "learning_rate": 0.0002,
+      "loss": 2.241,
+      "step": 21380
+    },
+    {
+      "epoch": 1.5938897168405366,
+      "grad_norm": 2.49922251701355,
+      "learning_rate": 0.0002,
+      "loss": 2.5755,
+      "step": 21390
+    },
+    {
+      "epoch": 1.594634873323398,
+      "grad_norm": 3.0789153575897217,
+      "learning_rate": 0.0002,
+      "loss": 2.5445,
+      "step": 21400
+    },
+    {
+      "epoch": 1.5953800298062593,
+      "grad_norm": 2.4592039585113525,
+      "learning_rate": 0.0002,
+      "loss": 2.4833,
+      "step": 21410
+    },
+    {
+      "epoch": 1.5961251862891208,
+      "grad_norm": 2.977599620819092,
+      "learning_rate": 0.0002,
+      "loss": 2.6318,
+      "step": 21420
+    },
+    {
+      "epoch": 1.5968703427719821,
+      "grad_norm": 2.868572950363159,
+      "learning_rate": 0.0002,
+      "loss": 2.6332,
+      "step": 21430
+    },
+    {
+      "epoch": 1.5976154992548435,
+      "grad_norm": 2.8400204181671143,
+      "learning_rate": 0.0002,
+      "loss": 2.6106,
+      "step": 21440
+    },
+    {
+      "epoch": 1.598360655737705,
+      "grad_norm": 3.0851080417633057,
+      "learning_rate": 0.0002,
+      "loss": 2.4113,
+      "step": 21450
+    },
+    {
+      "epoch": 1.5991058122205664,
+      "grad_norm": 2.4475607872009277,
+      "learning_rate": 0.0002,
+      "loss": 2.5,
+      "step": 21460
+    },
+    {
+      "epoch": 1.5998509687034277,
+      "grad_norm": 2.248441696166992,
+      "learning_rate": 0.0002,
+      "loss": 2.2983,
+      "step": 21470
+    },
+    {
+      "epoch": 1.6005961251862892,
+      "grad_norm": 2.1876697540283203,
+      "learning_rate": 0.0002,
+      "loss": 2.5876,
+      "step": 21480
+    },
+    {
+      "epoch": 1.6013412816691506,
+      "grad_norm": 2.766340494155884,
+      "learning_rate": 0.0002,
+      "loss": 2.6814,
+      "step": 21490
+    },
+    {
+      "epoch": 1.602086438152012,
+      "grad_norm": 2.4883816242218018,
+      "learning_rate": 0.0002,
+      "loss": 2.4444,
+      "step": 21500
+    },
+    {
+      "epoch": 1.6028315946348735,
+      "grad_norm": 2.7515087127685547,
+      "learning_rate": 0.0002,
+      "loss": 2.5638,
+      "step": 21510
+    },
+    {
+      "epoch": 1.6035767511177346,
+      "grad_norm": 2.400191068649292,
+      "learning_rate": 0.0002,
+      "loss": 2.5826,
+      "step": 21520
+    },
+    {
+      "epoch": 1.6043219076005961,
+      "grad_norm": 2.65370774269104,
+      "learning_rate": 0.0002,
+      "loss": 2.3286,
+      "step": 21530
+    },
+    {
+      "epoch": 1.6050670640834577,
+      "grad_norm": 2.525686264038086,
+      "learning_rate": 0.0002,
+      "loss": 2.5799,
+      "step": 21540
+    },
+    {
+      "epoch": 1.6058122205663188,
+      "grad_norm": 2.6483981609344482,
+      "learning_rate": 0.0002,
+      "loss": 2.4719,
+      "step": 21550
+    },
+    {
+      "epoch": 1.6065573770491803,
+      "grad_norm": 2.4052860736846924,
+      "learning_rate": 0.0002,
+      "loss": 2.6085,
+      "step": 21560
+    },
+    {
+      "epoch": 1.6073025335320419,
+      "grad_norm": 2.373023271560669,
+      "learning_rate": 0.0002,
+      "loss": 2.5449,
+      "step": 21570
+    },
+    {
+      "epoch": 1.608047690014903,
+      "grad_norm": 2.792961835861206,
+      "learning_rate": 0.0002,
+      "loss": 2.5756,
+      "step": 21580
+    },
+    {
+      "epoch": 1.6087928464977646,
+      "grad_norm": 2.024534225463867,
+      "learning_rate": 0.0002,
+      "loss": 2.4098,
+      "step": 21590
+    },
+    {
+      "epoch": 1.6095380029806259,
+      "grad_norm": 2.550271511077881,
+      "learning_rate": 0.0002,
+      "loss": 2.7197,
+      "step": 21600
+    },
+    {
+      "epoch": 1.6102831594634872,
+      "grad_norm": 2.357914686203003,
+      "learning_rate": 0.0002,
+      "loss": 2.4916,
+      "step": 21610
+    },
+    {
+      "epoch": 1.6110283159463488,
+      "grad_norm": 2.6073484420776367,
+      "learning_rate": 0.0002,
+      "loss": 2.5941,
+      "step": 21620
+    },
+    {
+      "epoch": 1.61177347242921,
+      "grad_norm": 2.532940626144409,
+      "learning_rate": 0.0002,
+      "loss": 2.469,
+      "step": 21630
+    },
+    {
+      "epoch": 1.6125186289120714,
+      "grad_norm": 2.5476560592651367,
+      "learning_rate": 0.0002,
+      "loss": 2.4917,
+      "step": 21640
+    },
+    {
+      "epoch": 1.613263785394933,
+      "grad_norm": 2.505892038345337,
+      "learning_rate": 0.0002,
+      "loss": 2.4284,
+      "step": 21650
+    },
+    {
+      "epoch": 1.6140089418777943,
+      "grad_norm": 2.5390076637268066,
+      "learning_rate": 0.0002,
+      "loss": 2.6434,
+      "step": 21660
+    },
+    {
+      "epoch": 1.6147540983606556,
+      "grad_norm": 1.909812092781067,
+      "learning_rate": 0.0002,
+      "loss": 2.3956,
+      "step": 21670
+    },
+    {
+      "epoch": 1.6154992548435172,
+      "grad_norm": 2.746105670928955,
+      "learning_rate": 0.0002,
+      "loss": 2.6643,
+      "step": 21680
+    },
+    {
+      "epoch": 1.6162444113263785,
+      "grad_norm": 2.7096972465515137,
+      "learning_rate": 0.0002,
+      "loss": 2.3153,
+      "step": 21690
+    },
+    {
+      "epoch": 1.6169895678092399,
+      "grad_norm": 2.1821792125701904,
+      "learning_rate": 0.0002,
+      "loss": 2.4432,
+      "step": 21700
+    },
+    {
+      "epoch": 1.6177347242921014,
+      "grad_norm": 2.201249599456787,
+      "learning_rate": 0.0002,
+      "loss": 2.2265,
+      "step": 21710
+    },
+    {
+      "epoch": 1.6184798807749627,
+      "grad_norm": 2.475764274597168,
+      "learning_rate": 0.0002,
+      "loss": 2.4643,
+      "step": 21720
+    },
+    {
+      "epoch": 1.619225037257824,
+      "grad_norm": 2.8647007942199707,
+      "learning_rate": 0.0002,
+      "loss": 2.4923,
+      "step": 21730
+    },
+    {
+      "epoch": 1.6199701937406856,
+      "grad_norm": 2.39050555229187,
+      "learning_rate": 0.0002,
+      "loss": 2.5228,
+      "step": 21740
+    },
+    {
+      "epoch": 1.620715350223547,
+      "grad_norm": 2.2536306381225586,
+      "learning_rate": 0.0002,
+      "loss": 2.4505,
+      "step": 21750
+    },
+    {
+      "epoch": 1.6214605067064083,
+      "grad_norm": 2.4495081901550293,
+      "learning_rate": 0.0002,
+      "loss": 2.4158,
+      "step": 21760
+    },
+    {
+      "epoch": 1.6222056631892698,
+      "grad_norm": 2.486243486404419,
+      "learning_rate": 0.0002,
+      "loss": 2.5142,
+      "step": 21770
+    },
+    {
+      "epoch": 1.6229508196721312,
+      "grad_norm": 2.2863104343414307,
+      "learning_rate": 0.0002,
+      "loss": 2.5569,
+      "step": 21780
+    },
+    {
+      "epoch": 1.6236959761549925,
+      "grad_norm": 2.1945390701293945,
+      "learning_rate": 0.0002,
+      "loss": 2.362,
+      "step": 21790
+    },
+    {
+      "epoch": 1.624441132637854,
+      "grad_norm": 2.59427809715271,
+      "learning_rate": 0.0002,
+      "loss": 2.5408,
+      "step": 21800
+    },
+    {
+      "epoch": 1.6251862891207154,
+      "grad_norm": 2.5718109607696533,
+      "learning_rate": 0.0002,
+      "loss": 2.628,
+      "step": 21810
+    },
+    {
+      "epoch": 1.6259314456035767,
+      "grad_norm": 2.5133166313171387,
+      "learning_rate": 0.0002,
+      "loss": 2.5077,
+      "step": 21820
+    },
+    {
+      "epoch": 1.6266766020864383,
+      "grad_norm": 2.313368797302246,
+      "learning_rate": 0.0002,
+      "loss": 2.5912,
+      "step": 21830
+    },
+    {
+      "epoch": 1.6274217585692996,
+      "grad_norm": 2.2794532775878906,
+      "learning_rate": 0.0002,
+      "loss": 2.3284,
+      "step": 21840
+    },
+    {
+      "epoch": 1.628166915052161,
+      "grad_norm": 2.508225202560425,
+      "learning_rate": 0.0002,
+      "loss": 2.5363,
+      "step": 21850
+    },
+    {
+      "epoch": 1.6289120715350225,
+      "grad_norm": 2.6698861122131348,
+      "learning_rate": 0.0002,
+      "loss": 2.4758,
+      "step": 21860
+    },
+    {
+      "epoch": 1.6296572280178836,
+      "grad_norm": 2.500823497772217,
+      "learning_rate": 0.0002,
+      "loss": 2.3382,
+      "step": 21870
+    },
+    {
+      "epoch": 1.6304023845007451,
+      "grad_norm": 2.7005372047424316,
+      "learning_rate": 0.0002,
+      "loss": 2.281,
+      "step": 21880
+    },
+    {
+      "epoch": 1.6311475409836067,
+      "grad_norm": 2.6699464321136475,
+      "learning_rate": 0.0002,
+      "loss": 2.5618,
+      "step": 21890
+    },
+    {
+      "epoch": 1.6318926974664678,
+      "grad_norm": 2.852407932281494,
+      "learning_rate": 0.0002,
+      "loss": 2.5993,
+      "step": 21900
+    },
+    {
+      "epoch": 1.6326378539493294,
+      "grad_norm": 2.408608913421631,
+      "learning_rate": 0.0002,
+      "loss": 2.5603,
+      "step": 21910
+    },
+    {
+      "epoch": 1.633383010432191,
+      "grad_norm": 2.828063726425171,
+      "learning_rate": 0.0002,
+      "loss": 2.5029,
+      "step": 21920
+    },
+    {
+      "epoch": 1.634128166915052,
+      "grad_norm": 2.4551167488098145,
+      "learning_rate": 0.0002,
+      "loss": 2.3027,
+      "step": 21930
+    },
+    {
+      "epoch": 1.6348733233979136,
+      "grad_norm": 2.5409445762634277,
+      "learning_rate": 0.0002,
+      "loss": 2.4344,
+      "step": 21940
+    },
+    {
+      "epoch": 1.635618479880775,
+      "grad_norm": 2.7620127201080322,
+      "learning_rate": 0.0002,
+      "loss": 2.6531,
+      "step": 21950
+    },
+    {
+      "epoch": 1.6363636363636362,
+      "grad_norm": 2.699997663497925,
+      "learning_rate": 0.0002,
+      "loss": 2.4624,
+      "step": 21960
+    },
+    {
+      "epoch": 1.6371087928464978,
+      "grad_norm": 2.9198567867279053,
+      "learning_rate": 0.0002,
+      "loss": 2.5269,
+      "step": 21970
+    },
+    {
+      "epoch": 1.6378539493293591,
+      "grad_norm": 2.7812252044677734,
+      "learning_rate": 0.0002,
+      "loss": 2.5048,
+      "step": 21980
+    },
+    {
+      "epoch": 1.6385991058122205,
+      "grad_norm": 2.7388384342193604,
+      "learning_rate": 0.0002,
+      "loss": 2.2389,
+      "step": 21990
+    },
+    {
+      "epoch": 1.639344262295082,
+      "grad_norm": 2.064328670501709,
+      "learning_rate": 0.0002,
+      "loss": 2.4192,
+      "step": 22000
+    },
+    {
+      "epoch": 1.6400894187779433,
+      "grad_norm": 2.4596285820007324,
+      "learning_rate": 0.0002,
+      "loss": 2.543,
+      "step": 22010
+    },
+    {
+      "epoch": 1.6408345752608047,
+      "grad_norm": 2.510390043258667,
+      "learning_rate": 0.0002,
+      "loss": 2.3457,
+      "step": 22020
+    },
+    {
+      "epoch": 1.6415797317436662,
+      "grad_norm": 2.2899527549743652,
+      "learning_rate": 0.0002,
+      "loss": 2.523,
+      "step": 22030
+    },
+    {
+      "epoch": 1.6423248882265276,
+      "grad_norm": 2.8313241004943848,
+      "learning_rate": 0.0002,
+      "loss": 2.4442,
+      "step": 22040
+    },
+    {
+      "epoch": 1.6430700447093889,
+      "grad_norm": 2.6460535526275635,
+      "learning_rate": 0.0002,
+      "loss": 2.4327,
+      "step": 22050
+    },
+    {
+      "epoch": 1.6438152011922504,
+      "grad_norm": 2.5774576663970947,
+      "learning_rate": 0.0002,
+      "loss": 2.6181,
+      "step": 22060
+    },
+    {
+      "epoch": 1.6445603576751118,
+      "grad_norm": 2.517756700515747,
+      "learning_rate": 0.0002,
+      "loss": 2.5624,
+      "step": 22070
+    },
+    {
+      "epoch": 1.645305514157973,
+      "grad_norm": 2.382204055786133,
+      "learning_rate": 0.0002,
+      "loss": 2.6089,
+      "step": 22080
+    },
+    {
+      "epoch": 1.6460506706408347,
+      "grad_norm": 2.5505146980285645,
+      "learning_rate": 0.0002,
+      "loss": 2.5738,
+      "step": 22090
+    },
+    {
+      "epoch": 1.646795827123696,
+      "grad_norm": 3.1729090213775635,
+      "learning_rate": 0.0002,
+      "loss": 2.5716,
+      "step": 22100
+    },
+    {
+      "epoch": 1.6475409836065573,
+      "grad_norm": 2.7010695934295654,
+      "learning_rate": 0.0002,
+      "loss": 2.346,
+      "step": 22110
+    },
+    {
+      "epoch": 1.6482861400894189,
+      "grad_norm": 2.2861719131469727,
+      "learning_rate": 0.0002,
+      "loss": 2.4943,
+      "step": 22120
+    },
+    {
+      "epoch": 1.6490312965722802,
+      "grad_norm": 2.2667644023895264,
+      "learning_rate": 0.0002,
+      "loss": 2.5656,
+      "step": 22130
+    },
+    {
+      "epoch": 1.6497764530551415,
+      "grad_norm": 2.9026103019714355,
+      "learning_rate": 0.0002,
+      "loss": 2.5225,
+      "step": 22140
+    },
+    {
+      "epoch": 1.650521609538003,
+      "grad_norm": 2.4847819805145264,
+      "learning_rate": 0.0002,
+      "loss": 2.5498,
+      "step": 22150
+    },
+    {
+      "epoch": 1.6512667660208644,
+      "grad_norm": 2.3765196800231934,
+      "learning_rate": 0.0002,
+      "loss": 2.4714,
+      "step": 22160
+    },
+    {
+      "epoch": 1.6520119225037257,
+      "grad_norm": 2.203185796737671,
+      "learning_rate": 0.0002,
+      "loss": 2.5005,
+      "step": 22170
+    },
+    {
+      "epoch": 1.6527570789865873,
+      "grad_norm": 2.480102300643921,
+      "learning_rate": 0.0002,
+      "loss": 2.5686,
+      "step": 22180
+    },
+    {
+      "epoch": 1.6535022354694486,
+      "grad_norm": 2.299968957901001,
+      "learning_rate": 0.0002,
+      "loss": 2.5814,
+      "step": 22190
+    },
+    {
+      "epoch": 1.65424739195231,
+      "grad_norm": 3.0211429595947266,
+      "learning_rate": 0.0002,
+      "loss": 2.6236,
+      "step": 22200
+    },
+    {
+      "epoch": 1.6549925484351715,
+      "grad_norm": 2.7016682624816895,
+      "learning_rate": 0.0002,
+      "loss": 2.5385,
+      "step": 22210
+    },
+    {
+      "epoch": 1.6557377049180326,
+      "grad_norm": 2.5749051570892334,
+      "learning_rate": 0.0002,
+      "loss": 2.4144,
+      "step": 22220
+    },
+    {
+      "epoch": 1.6564828614008942,
+      "grad_norm": 2.0706841945648193,
+      "learning_rate": 0.0002,
+      "loss": 2.4751,
+      "step": 22230
+    },
+    {
+      "epoch": 1.6572280178837557,
+      "grad_norm": 2.4944183826446533,
+      "learning_rate": 0.0002,
+      "loss": 2.5371,
+      "step": 22240
+    },
+    {
+      "epoch": 1.6579731743666168,
+      "grad_norm": 2.600158214569092,
+      "learning_rate": 0.0002,
+      "loss": 2.5388,
+      "step": 22250
+    },
+    {
+      "epoch": 1.6587183308494784,
+      "grad_norm": 2.428926706314087,
+      "learning_rate": 0.0002,
+      "loss": 2.4767,
+      "step": 22260
+    },
+    {
+      "epoch": 1.65946348733234,
+      "grad_norm": 2.805734395980835,
+      "learning_rate": 0.0002,
+      "loss": 2.4908,
+      "step": 22270
+    },
+    {
+      "epoch": 1.660208643815201,
+      "grad_norm": 2.7497949600219727,
+      "learning_rate": 0.0002,
+      "loss": 2.4993,
+      "step": 22280
+    },
+    {
+      "epoch": 1.6609538002980626,
+      "grad_norm": 2.559528350830078,
+      "learning_rate": 0.0002,
+      "loss": 2.2774,
+      "step": 22290
+    },
+    {
+      "epoch": 1.661698956780924,
+      "grad_norm": 2.5140788555145264,
+      "learning_rate": 0.0002,
+      "loss": 2.5504,
+      "step": 22300
+    },
+    {
+      "epoch": 1.6624441132637853,
+      "grad_norm": 2.3234243392944336,
+      "learning_rate": 0.0002,
+      "loss": 2.6684,
+      "step": 22310
+    },
+    {
+      "epoch": 1.6631892697466468,
+      "grad_norm": 2.3977694511413574,
+      "learning_rate": 0.0002,
+      "loss": 2.7416,
+      "step": 22320
+    },
+    {
+      "epoch": 1.6639344262295082,
+      "grad_norm": 2.2138288021087646,
+      "learning_rate": 0.0002,
+      "loss": 2.2312,
+      "step": 22330
+    },
+    {
+      "epoch": 1.6646795827123695,
+      "grad_norm": 2.4575119018554688,
+      "learning_rate": 0.0002,
+      "loss": 2.549,
+      "step": 22340
+    },
+    {
+      "epoch": 1.665424739195231,
+      "grad_norm": 2.426833152770996,
+      "learning_rate": 0.0002,
+      "loss": 2.449,
+      "step": 22350
+    },
+    {
+      "epoch": 1.6661698956780924,
+      "grad_norm": 3.055542230606079,
+      "learning_rate": 0.0002,
+      "loss": 2.4273,
+      "step": 22360
+    },
+    {
+      "epoch": 1.6669150521609537,
+      "grad_norm": 2.5469939708709717,
+      "learning_rate": 0.0002,
+      "loss": 2.6418,
+      "step": 22370
+    },
+    {
+      "epoch": 1.6676602086438153,
+      "grad_norm": 2.537905216217041,
+      "learning_rate": 0.0002,
+      "loss": 2.5369,
+      "step": 22380
+    },
+    {
+      "epoch": 1.6684053651266766,
+      "grad_norm": 2.696962833404541,
+      "learning_rate": 0.0002,
+      "loss": 2.6064,
+      "step": 22390
+    },
+    {
+      "epoch": 1.669150521609538,
+      "grad_norm": 2.4675662517547607,
+      "learning_rate": 0.0002,
+      "loss": 2.515,
+      "step": 22400
+    },
+    {
+      "epoch": 1.6698956780923995,
+      "grad_norm": 2.486266851425171,
+      "learning_rate": 0.0002,
+      "loss": 2.4349,
+      "step": 22410
+    },
+    {
+      "epoch": 1.6706408345752608,
+      "grad_norm": 2.529770612716675,
+      "learning_rate": 0.0002,
+      "loss": 2.3911,
+      "step": 22420
+    },
+    {
+      "epoch": 1.6713859910581221,
+      "grad_norm": 2.3345956802368164,
+      "learning_rate": 0.0002,
+      "loss": 2.5192,
+      "step": 22430
+    },
+    {
+      "epoch": 1.6721311475409837,
+      "grad_norm": 2.8987648487091064,
+      "learning_rate": 0.0002,
+      "loss": 2.355,
+      "step": 22440
+    },
+    {
+      "epoch": 1.672876304023845,
+      "grad_norm": 2.4978036880493164,
+      "learning_rate": 0.0002,
+      "loss": 2.5969,
+      "step": 22450
+    },
+    {
+      "epoch": 1.6736214605067063,
+      "grad_norm": 2.2257959842681885,
+      "learning_rate": 0.0002,
+      "loss": 2.5473,
+      "step": 22460
+    },
+    {
+      "epoch": 1.674366616989568,
+      "grad_norm": 2.3007636070251465,
+      "learning_rate": 0.0002,
+      "loss": 2.5016,
+      "step": 22470
+    },
+    {
+      "epoch": 1.6751117734724292,
+      "grad_norm": 2.6325652599334717,
+      "learning_rate": 0.0002,
+      "loss": 2.4148,
+      "step": 22480
+    },
+    {
+      "epoch": 1.6758569299552906,
+      "grad_norm": 2.6435465812683105,
+      "learning_rate": 0.0002,
+      "loss": 2.5434,
+      "step": 22490
+    },
+    {
+      "epoch": 1.6766020864381521,
+      "grad_norm": 2.9889986515045166,
+      "learning_rate": 0.0002,
+      "loss": 2.6347,
+      "step": 22500
+    },
+    {
+      "epoch": 1.6773472429210134,
+      "grad_norm": 3.051703691482544,
+      "learning_rate": 0.0002,
+      "loss": 2.5085,
+      "step": 22510
+    },
+    {
+      "epoch": 1.6780923994038748,
+      "grad_norm": 2.7691986560821533,
+      "learning_rate": 0.0002,
+      "loss": 2.4834,
+      "step": 22520
+    },
+    {
+      "epoch": 1.6788375558867363,
+      "grad_norm": 2.565810441970825,
+      "learning_rate": 0.0002,
+      "loss": 2.4304,
+      "step": 22530
+    },
+    {
+      "epoch": 1.6795827123695977,
+      "grad_norm": 2.5967001914978027,
+      "learning_rate": 0.0002,
+      "loss": 2.4379,
+      "step": 22540
+    },
+    {
+      "epoch": 1.680327868852459,
+      "grad_norm": 1.9632105827331543,
+      "learning_rate": 0.0002,
+      "loss": 2.2073,
+      "step": 22550
+    },
+    {
+      "epoch": 1.6810730253353205,
+      "grad_norm": 2.458916664123535,
+      "learning_rate": 0.0002,
+      "loss": 2.443,
+      "step": 22560
+    },
+    {
+      "epoch": 1.6818181818181817,
+      "grad_norm": 2.67900013923645,
+      "learning_rate": 0.0002,
+      "loss": 2.6422,
+      "step": 22570
+    },
+    {
+      "epoch": 1.6825633383010432,
+      "grad_norm": 2.3862104415893555,
+      "learning_rate": 0.0002,
+      "loss": 2.5824,
+      "step": 22580
+    },
+    {
+      "epoch": 1.6833084947839048,
+      "grad_norm": 2.43456768989563,
+      "learning_rate": 0.0002,
+      "loss": 2.4033,
+      "step": 22590
+    },
+    {
+      "epoch": 1.6840536512667659,
+      "grad_norm": 2.6378731727600098,
+      "learning_rate": 0.0002,
+      "loss": 2.511,
+      "step": 22600
+    },
+    {
+      "epoch": 1.6847988077496274,
+      "grad_norm": 2.612466335296631,
+      "learning_rate": 0.0002,
+      "loss": 2.5721,
+      "step": 22610
+    },
+    {
+      "epoch": 1.685543964232489,
+      "grad_norm": 2.5247201919555664,
+      "learning_rate": 0.0002,
+      "loss": 2.2889,
+      "step": 22620
+    },
+    {
+      "epoch": 1.68628912071535,
+      "grad_norm": 2.7932889461517334,
+      "learning_rate": 0.0002,
+      "loss": 2.5143,
+      "step": 22630
+    },
+    {
+      "epoch": 1.6870342771982116,
+      "grad_norm": 2.531528949737549,
+      "learning_rate": 0.0002,
+      "loss": 2.6344,
+      "step": 22640
+    },
+    {
+      "epoch": 1.687779433681073,
+      "grad_norm": 2.5633575916290283,
+      "learning_rate": 0.0002,
+      "loss": 2.6162,
+      "step": 22650
+    },
+    {
+      "epoch": 1.6885245901639343,
+      "grad_norm": 2.114488124847412,
+      "learning_rate": 0.0002,
+      "loss": 2.5851,
+      "step": 22660
+    },
+    {
+      "epoch": 1.6892697466467959,
+      "grad_norm": 2.478349208831787,
+      "learning_rate": 0.0002,
+      "loss": 2.4753,
+      "step": 22670
+    },
+    {
+      "epoch": 1.6900149031296572,
+      "grad_norm": 2.538219690322876,
+      "learning_rate": 0.0002,
+      "loss": 2.5749,
+      "step": 22680
+    },
+    {
+      "epoch": 1.6907600596125185,
+      "grad_norm": 2.557431221008301,
+      "learning_rate": 0.0002,
+      "loss": 2.5866,
+      "step": 22690
+    },
+    {
+      "epoch": 1.69150521609538,
+      "grad_norm": 2.831338882446289,
+      "learning_rate": 0.0002,
+      "loss": 2.7184,
+      "step": 22700
+    },
+    {
+      "epoch": 1.6922503725782414,
+      "grad_norm": 2.0015451908111572,
+      "learning_rate": 0.0002,
+      "loss": 2.4181,
+      "step": 22710
+    },
+    {
+      "epoch": 1.6929955290611027,
+      "grad_norm": 2.3533787727355957,
+      "learning_rate": 0.0002,
+      "loss": 2.3283,
+      "step": 22720
+    },
+    {
+      "epoch": 1.6937406855439643,
+      "grad_norm": 2.209768295288086,
+      "learning_rate": 0.0002,
+      "loss": 2.5522,
+      "step": 22730
+    },
+    {
+      "epoch": 1.6944858420268256,
+      "grad_norm": 2.5806851387023926,
+      "learning_rate": 0.0002,
+      "loss": 2.6221,
+      "step": 22740
+    },
+    {
+      "epoch": 1.695230998509687,
+      "grad_norm": 2.6998038291931152,
+      "learning_rate": 0.0002,
+      "loss": 2.5525,
+      "step": 22750
+    },
+    {
+      "epoch": 1.6959761549925485,
+      "grad_norm": 2.318673610687256,
+      "learning_rate": 0.0002,
+      "loss": 2.394,
+      "step": 22760
+    },
+    {
+      "epoch": 1.6967213114754098,
+      "grad_norm": 1.767236590385437,
+      "learning_rate": 0.0002,
+      "loss": 2.2679,
+      "step": 22770
+    },
+    {
+      "epoch": 1.6974664679582712,
+      "grad_norm": 2.5516786575317383,
+      "learning_rate": 0.0002,
+      "loss": 2.4601,
+      "step": 22780
+    },
+    {
+      "epoch": 1.6982116244411327,
+      "grad_norm": 2.5691072940826416,
+      "learning_rate": 0.0002,
+      "loss": 2.4903,
+      "step": 22790
+    },
+    {
+      "epoch": 1.698956780923994,
+      "grad_norm": 2.6032519340515137,
+      "learning_rate": 0.0002,
+      "loss": 2.5091,
+      "step": 22800
+    },
+    {
+      "epoch": 1.6997019374068554,
+      "grad_norm": 2.57243013381958,
+      "learning_rate": 0.0002,
+      "loss": 2.3827,
+      "step": 22810
+    },
+    {
+      "epoch": 1.700447093889717,
+      "grad_norm": 2.560321092605591,
+      "learning_rate": 0.0002,
+      "loss": 2.587,
+      "step": 22820
+    },
+    {
+      "epoch": 1.7011922503725783,
+      "grad_norm": 2.3471126556396484,
+      "learning_rate": 0.0002,
+      "loss": 2.5432,
+      "step": 22830
+    },
+    {
+      "epoch": 1.7019374068554396,
+      "grad_norm": 2.7333359718322754,
+      "learning_rate": 0.0002,
+      "loss": 2.4153,
+      "step": 22840
+    },
+    {
+      "epoch": 1.7026825633383011,
+      "grad_norm": 2.481062173843384,
+      "learning_rate": 0.0002,
+      "loss": 2.6483,
+      "step": 22850
+    },
+    {
+      "epoch": 1.7034277198211625,
+      "grad_norm": 2.814061164855957,
+      "learning_rate": 0.0002,
+      "loss": 2.5051,
+      "step": 22860
+    },
+    {
+      "epoch": 1.7041728763040238,
+      "grad_norm": 2.8423242568969727,
+      "learning_rate": 0.0002,
+      "loss": 2.6922,
+      "step": 22870
+    },
+    {
+      "epoch": 1.7049180327868854,
+      "grad_norm": 2.553450107574463,
+      "learning_rate": 0.0002,
+      "loss": 2.5812,
+      "step": 22880
+    },
+    {
+      "epoch": 1.7056631892697467,
+      "grad_norm": 2.6540987491607666,
+      "learning_rate": 0.0002,
+      "loss": 2.5114,
+      "step": 22890
+    },
+    {
+      "epoch": 1.706408345752608,
+      "grad_norm": 2.4300739765167236,
+      "learning_rate": 0.0002,
+      "loss": 2.6493,
+      "step": 22900
+    },
+    {
+      "epoch": 1.7071535022354696,
+      "grad_norm": 2.4339499473571777,
+      "learning_rate": 0.0002,
+      "loss": 2.5105,
+      "step": 22910
+    },
+    {
+      "epoch": 1.7078986587183307,
+      "grad_norm": 2.5802414417266846,
+      "learning_rate": 0.0002,
+      "loss": 2.4289,
+      "step": 22920
+    },
+    {
+      "epoch": 1.7086438152011922,
+      "grad_norm": 2.7700533866882324,
+      "learning_rate": 0.0002,
+      "loss": 2.6301,
+      "step": 22930
+    },
+    {
+      "epoch": 1.7093889716840538,
+      "grad_norm": 2.8011083602905273,
+      "learning_rate": 0.0002,
+      "loss": 2.3991,
+      "step": 22940
+    },
+    {
+      "epoch": 1.710134128166915,
+      "grad_norm": 2.746286630630493,
+      "learning_rate": 0.0002,
+      "loss": 2.396,
+      "step": 22950
+    },
+    {
+      "epoch": 1.7108792846497765,
+      "grad_norm": 2.3285419940948486,
+      "learning_rate": 0.0002,
+      "loss": 2.536,
+      "step": 22960
+    },
+    {
+      "epoch": 1.711624441132638,
+      "grad_norm": 2.7214415073394775,
+      "learning_rate": 0.0002,
+      "loss": 2.5631,
+      "step": 22970
+    },
+    {
+      "epoch": 1.7123695976154991,
+      "grad_norm": 2.2611494064331055,
+      "learning_rate": 0.0002,
+      "loss": 2.4555,
+      "step": 22980
+    },
+    {
+      "epoch": 1.7131147540983607,
+      "grad_norm": 2.569819688796997,
+      "learning_rate": 0.0002,
+      "loss": 2.4973,
+      "step": 22990
+    },
+    {
+      "epoch": 1.713859910581222,
+      "grad_norm": 2.3823273181915283,
+      "learning_rate": 0.0002,
+      "loss": 2.5048,
+      "step": 23000
+    },
+    {
+      "epoch": 1.7146050670640833,
+      "grad_norm": 2.369661569595337,
+      "learning_rate": 0.0002,
+      "loss": 2.351,
+      "step": 23010
+    },
+    {
+      "epoch": 1.7153502235469449,
+      "grad_norm": 2.551729679107666,
+      "learning_rate": 0.0002,
+      "loss": 2.5331,
+      "step": 23020
+    },
+    {
+      "epoch": 1.7160953800298062,
+      "grad_norm": 2.7406301498413086,
+      "learning_rate": 0.0002,
+      "loss": 2.3774,
+      "step": 23030
+    },
+    {
+      "epoch": 1.7168405365126675,
+      "grad_norm": 3.1042418479919434,
+      "learning_rate": 0.0002,
+      "loss": 2.5131,
+      "step": 23040
+    },
+    {
+      "epoch": 1.717585692995529,
+      "grad_norm": 2.6012461185455322,
+      "learning_rate": 0.0002,
+      "loss": 2.635,
+      "step": 23050
+    },
+    {
+      "epoch": 1.7183308494783904,
+      "grad_norm": 2.7491185665130615,
+      "learning_rate": 0.0002,
+      "loss": 2.5063,
+      "step": 23060
+    },
+    {
+      "epoch": 1.7190760059612518,
+      "grad_norm": 2.6317410469055176,
+      "learning_rate": 0.0002,
+      "loss": 2.6614,
+      "step": 23070
+    },
+    {
+      "epoch": 1.7198211624441133,
+      "grad_norm": 2.699126958847046,
+      "learning_rate": 0.0002,
+      "loss": 2.6547,
+      "step": 23080
+    },
+    {
+      "epoch": 1.7205663189269746,
+      "grad_norm": 2.882514715194702,
+      "learning_rate": 0.0002,
+      "loss": 2.5183,
+      "step": 23090
+    },
+    {
+      "epoch": 1.721311475409836,
+      "grad_norm": 2.217384099960327,
+      "learning_rate": 0.0002,
+      "loss": 2.4662,
+      "step": 23100
+    },
+    {
+      "epoch": 1.7220566318926975,
+      "grad_norm": 2.645479679107666,
+      "learning_rate": 0.0002,
+      "loss": 2.5276,
+      "step": 23110
+    },
+    {
+      "epoch": 1.7228017883755589,
+      "grad_norm": 3.0477075576782227,
+      "learning_rate": 0.0002,
+      "loss": 2.5021,
+      "step": 23120
+    },
+    {
+      "epoch": 1.7235469448584202,
+      "grad_norm": 2.56258225440979,
+      "learning_rate": 0.0002,
+      "loss": 2.5864,
+      "step": 23130
+    },
+    {
+      "epoch": 1.7242921013412817,
+      "grad_norm": 2.4579217433929443,
+      "learning_rate": 0.0002,
+      "loss": 2.5179,
+      "step": 23140
+    },
+    {
+      "epoch": 1.725037257824143,
+      "grad_norm": 2.761368989944458,
+      "learning_rate": 0.0002,
+      "loss": 2.4578,
+      "step": 23150
+    },
+    {
+      "epoch": 1.7257824143070044,
+      "grad_norm": 2.7409048080444336,
+      "learning_rate": 0.0002,
+      "loss": 2.5305,
+      "step": 23160
+    },
+    {
+      "epoch": 1.726527570789866,
+      "grad_norm": 2.3490233421325684,
+      "learning_rate": 0.0002,
+      "loss": 2.4945,
+      "step": 23170
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 2.397792339324951,
+      "learning_rate": 0.0002,
+      "loss": 2.5054,
+      "step": 23180
+    },
+    {
+      "epoch": 1.7280178837555886,
+      "grad_norm": 2.8195900917053223,
+      "learning_rate": 0.0002,
+      "loss": 2.5461,
+      "step": 23190
+    },
+    {
+      "epoch": 1.7287630402384502,
+      "grad_norm": 2.2869858741760254,
+      "learning_rate": 0.0002,
+      "loss": 2.4762,
+      "step": 23200
+    },
+    {
+      "epoch": 1.7295081967213115,
+      "grad_norm": 2.136507272720337,
+      "learning_rate": 0.0002,
+      "loss": 2.5111,
+      "step": 23210
+    },
+    {
+      "epoch": 1.7302533532041728,
+      "grad_norm": 2.2523751258850098,
+      "learning_rate": 0.0002,
+      "loss": 2.3777,
+      "step": 23220
+    },
+    {
+      "epoch": 1.7309985096870344,
+      "grad_norm": 2.7627720832824707,
+      "learning_rate": 0.0002,
+      "loss": 2.3795,
+      "step": 23230
+    },
+    {
+      "epoch": 1.7317436661698957,
+      "grad_norm": 2.663520336151123,
+      "learning_rate": 0.0002,
+      "loss": 2.4765,
+      "step": 23240
+    },
+    {
+      "epoch": 1.732488822652757,
+      "grad_norm": 2.7753946781158447,
+      "learning_rate": 0.0002,
+      "loss": 2.6572,
+      "step": 23250
+    },
+    {
+      "epoch": 1.7332339791356186,
+      "grad_norm": 2.697758436203003,
+      "learning_rate": 0.0002,
+      "loss": 2.5019,
+      "step": 23260
+    },
+    {
+      "epoch": 1.7339791356184797,
+      "grad_norm": 2.4751906394958496,
+      "learning_rate": 0.0002,
+      "loss": 2.5409,
+      "step": 23270
+    },
+    {
+      "epoch": 1.7347242921013413,
+      "grad_norm": 2.4185609817504883,
+      "learning_rate": 0.0002,
+      "loss": 2.5397,
+      "step": 23280
+    },
+    {
+      "epoch": 1.7354694485842028,
+      "grad_norm": 2.462890148162842,
+      "learning_rate": 0.0002,
+      "loss": 2.3603,
+      "step": 23290
+    },
+    {
+      "epoch": 1.736214605067064,
+      "grad_norm": 2.575814723968506,
+      "learning_rate": 0.0002,
+      "loss": 2.6263,
+      "step": 23300
+    },
+    {
+      "epoch": 1.7369597615499255,
+      "grad_norm": 2.7433903217315674,
+      "learning_rate": 0.0002,
+      "loss": 2.3266,
+      "step": 23310
+    },
+    {
+      "epoch": 1.737704918032787,
+      "grad_norm": 2.4978883266448975,
+      "learning_rate": 0.0002,
+      "loss": 2.4926,
+      "step": 23320
+    },
+    {
+      "epoch": 1.7384500745156481,
+      "grad_norm": 2.297142744064331,
+      "learning_rate": 0.0002,
+      "loss": 2.6361,
+      "step": 23330
+    },
+    {
+      "epoch": 1.7391952309985097,
+      "grad_norm": 2.436873435974121,
+      "learning_rate": 0.0002,
+      "loss": 2.4963,
+      "step": 23340
+    },
+    {
+      "epoch": 1.7399403874813713,
+      "grad_norm": 2.3684535026550293,
+      "learning_rate": 0.0002,
+      "loss": 2.451,
+      "step": 23350
+    },
+    {
+      "epoch": 1.7406855439642324,
+      "grad_norm": 2.752638101577759,
+      "learning_rate": 0.0002,
+      "loss": 2.3857,
+      "step": 23360
+    },
+    {
+      "epoch": 1.741430700447094,
+      "grad_norm": 3.9165122509002686,
+      "learning_rate": 0.0002,
+      "loss": 2.552,
+      "step": 23370
+    },
+    {
+      "epoch": 1.7421758569299552,
+      "grad_norm": 2.561709403991699,
+      "learning_rate": 0.0002,
+      "loss": 2.5888,
+      "step": 23380
+    },
+    {
+      "epoch": 1.7429210134128166,
+      "grad_norm": 2.5623152256011963,
+      "learning_rate": 0.0002,
+      "loss": 2.3984,
+      "step": 23390
+    },
+    {
+      "epoch": 1.7436661698956781,
+      "grad_norm": 2.3049354553222656,
+      "learning_rate": 0.0002,
+      "loss": 2.3511,
+      "step": 23400
+    },
+    {
+      "epoch": 1.7444113263785395,
+      "grad_norm": 2.6003150939941406,
+      "learning_rate": 0.0002,
+      "loss": 2.4049,
+      "step": 23410
+    },
+    {
+      "epoch": 1.7451564828614008,
+      "grad_norm": 2.4863944053649902,
+      "learning_rate": 0.0002,
+      "loss": 2.3698,
+      "step": 23420
+    },
+    {
+      "epoch": 1.7459016393442623,
+      "grad_norm": 2.313157558441162,
+      "learning_rate": 0.0002,
+      "loss": 2.489,
+      "step": 23430
+    },
+    {
+      "epoch": 1.7466467958271237,
+      "grad_norm": 2.4840941429138184,
+      "learning_rate": 0.0002,
+      "loss": 2.4372,
+      "step": 23440
+    },
+    {
+      "epoch": 1.747391952309985,
+      "grad_norm": 2.6886258125305176,
+      "learning_rate": 0.0002,
+      "loss": 2.4685,
+      "step": 23450
+    },
+    {
+      "epoch": 1.7481371087928466,
+      "grad_norm": 2.4791266918182373,
+      "learning_rate": 0.0002,
+      "loss": 2.4431,
+      "step": 23460
+    },
+    {
+      "epoch": 1.748882265275708,
+      "grad_norm": 2.7931692600250244,
+      "learning_rate": 0.0002,
+      "loss": 2.2753,
+      "step": 23470
+    },
+    {
+      "epoch": 1.7496274217585692,
+      "grad_norm": 2.818281650543213,
+      "learning_rate": 0.0002,
+      "loss": 2.4702,
+      "step": 23480
+    },
+    {
+      "epoch": 1.7503725782414308,
+      "grad_norm": 2.638704299926758,
+      "learning_rate": 0.0002,
+      "loss": 2.3848,
+      "step": 23490
+    },
+    {
+      "epoch": 1.751117734724292,
+      "grad_norm": 2.501591682434082,
+      "learning_rate": 0.0002,
+      "loss": 2.5548,
+      "step": 23500
+    },
+    {
+      "epoch": 1.7518628912071534,
+      "grad_norm": 2.969492197036743,
+      "learning_rate": 0.0002,
+      "loss": 2.5042,
+      "step": 23510
+    },
+    {
+      "epoch": 1.752608047690015,
+      "grad_norm": 2.4512195587158203,
+      "learning_rate": 0.0002,
+      "loss": 2.2923,
+      "step": 23520
+    },
+    {
+      "epoch": 1.7533532041728763,
+      "grad_norm": 2.2900497913360596,
+      "learning_rate": 0.0002,
+      "loss": 2.4102,
+      "step": 23530
+    },
+    {
+      "epoch": 1.7540983606557377,
+      "grad_norm": 2.6613259315490723,
+      "learning_rate": 0.0002,
+      "loss": 2.277,
+      "step": 23540
+    },
+    {
+      "epoch": 1.7548435171385992,
+      "grad_norm": 2.427893877029419,
+      "learning_rate": 0.0002,
+      "loss": 2.5847,
+      "step": 23550
+    },
+    {
+      "epoch": 1.7555886736214605,
+      "grad_norm": 2.631917715072632,
+      "learning_rate": 0.0002,
+      "loss": 2.6333,
+      "step": 23560
+    },
+    {
+      "epoch": 1.7563338301043219,
+      "grad_norm": 2.395526170730591,
+      "learning_rate": 0.0002,
+      "loss": 2.5131,
+      "step": 23570
+    },
+    {
+      "epoch": 1.7570789865871834,
+      "grad_norm": 2.392958641052246,
+      "learning_rate": 0.0002,
+      "loss": 2.5297,
+      "step": 23580
+    },
+    {
+      "epoch": 1.7578241430700448,
+      "grad_norm": 2.660301923751831,
+      "learning_rate": 0.0002,
+      "loss": 2.529,
+      "step": 23590
+    },
+    {
+      "epoch": 1.758569299552906,
+      "grad_norm": 2.4387199878692627,
+      "learning_rate": 0.0002,
+      "loss": 2.5486,
+      "step": 23600
+    },
+    {
+      "epoch": 1.7593144560357676,
+      "grad_norm": 2.463621139526367,
+      "learning_rate": 0.0002,
+      "loss": 2.6065,
+      "step": 23610
+    },
+    {
+      "epoch": 1.7600596125186287,
+      "grad_norm": 2.4696829319000244,
+      "learning_rate": 0.0002,
+      "loss": 2.5175,
+      "step": 23620
+    },
+    {
+      "epoch": 1.7608047690014903,
+      "grad_norm": 2.802701473236084,
+      "learning_rate": 0.0002,
+      "loss": 2.5997,
+      "step": 23630
+    },
+    {
+      "epoch": 1.7615499254843519,
+      "grad_norm": 2.8156678676605225,
+      "learning_rate": 0.0002,
+      "loss": 2.5085,
+      "step": 23640
+    },
+    {
+      "epoch": 1.762295081967213,
+      "grad_norm": 2.4574429988861084,
+      "learning_rate": 0.0002,
+      "loss": 2.5127,
+      "step": 23650
+    },
+    {
+      "epoch": 1.7630402384500745,
+      "grad_norm": 2.26594614982605,
+      "learning_rate": 0.0002,
+      "loss": 2.6283,
+      "step": 23660
+    },
+    {
+      "epoch": 1.763785394932936,
+      "grad_norm": 2.458120107650757,
+      "learning_rate": 0.0002,
+      "loss": 2.4844,
+      "step": 23670
+    },
+    {
+      "epoch": 1.7645305514157972,
+      "grad_norm": 2.4478704929351807,
+      "learning_rate": 0.0002,
+      "loss": 2.596,
+      "step": 23680
+    },
+    {
+      "epoch": 1.7652757078986587,
+      "grad_norm": 2.5973260402679443,
+      "learning_rate": 0.0002,
+      "loss": 2.7246,
+      "step": 23690
+    },
+    {
+      "epoch": 1.7660208643815203,
+      "grad_norm": 2.2895405292510986,
+      "learning_rate": 0.0002,
+      "loss": 2.5442,
+      "step": 23700
+    },
+    {
+      "epoch": 1.7667660208643814,
+      "grad_norm": 3.0286567211151123,
+      "learning_rate": 0.0002,
+      "loss": 2.4685,
+      "step": 23710
+    },
+    {
+      "epoch": 1.767511177347243,
+      "grad_norm": 2.95426869392395,
+      "learning_rate": 0.0002,
+      "loss": 2.6306,
+      "step": 23720
+    },
+    {
+      "epoch": 1.7682563338301043,
+      "grad_norm": 2.4281368255615234,
+      "learning_rate": 0.0002,
+      "loss": 2.4162,
+      "step": 23730
+    },
+    {
+      "epoch": 1.7690014903129656,
+      "grad_norm": 2.274803876876831,
+      "learning_rate": 0.0002,
+      "loss": 2.2333,
+      "step": 23740
+    },
+    {
+      "epoch": 1.7697466467958272,
+      "grad_norm": 2.6567139625549316,
+      "learning_rate": 0.0002,
+      "loss": 2.5369,
+      "step": 23750
+    },
+    {
+      "epoch": 1.7704918032786885,
+      "grad_norm": 2.1148862838745117,
+      "learning_rate": 0.0002,
+      "loss": 2.4882,
+      "step": 23760
+    },
+    {
+      "epoch": 1.7712369597615498,
+      "grad_norm": 2.397024631500244,
+      "learning_rate": 0.0002,
+      "loss": 2.5951,
+      "step": 23770
+    },
+    {
+      "epoch": 1.7719821162444114,
+      "grad_norm": 2.556600570678711,
+      "learning_rate": 0.0002,
+      "loss": 2.4878,
+      "step": 23780
+    },
+    {
+      "epoch": 1.7727272727272727,
+      "grad_norm": 2.361325979232788,
+      "learning_rate": 0.0002,
+      "loss": 2.6025,
+      "step": 23790
+    },
+    {
+      "epoch": 1.773472429210134,
+      "grad_norm": 2.221918821334839,
+      "learning_rate": 0.0002,
+      "loss": 2.4707,
+      "step": 23800
+    },
+    {
+      "epoch": 1.7742175856929956,
+      "grad_norm": 2.5794131755828857,
+      "learning_rate": 0.0002,
+      "loss": 2.3977,
+      "step": 23810
+    },
+    {
+      "epoch": 1.774962742175857,
+      "grad_norm": 2.4929275512695312,
+      "learning_rate": 0.0002,
+      "loss": 2.6056,
+      "step": 23820
+    },
+    {
+      "epoch": 1.7757078986587183,
+      "grad_norm": 2.478788375854492,
+      "learning_rate": 0.0002,
+      "loss": 2.5346,
+      "step": 23830
+    },
+    {
+      "epoch": 1.7764530551415798,
+      "grad_norm": 2.7491824626922607,
+      "learning_rate": 0.0002,
+      "loss": 2.2653,
+      "step": 23840
+    },
+    {
+      "epoch": 1.7771982116244411,
+      "grad_norm": 2.842250108718872,
+      "learning_rate": 0.0002,
+      "loss": 2.5534,
+      "step": 23850
+    },
+    {
+      "epoch": 1.7779433681073025,
+      "grad_norm": 2.78625226020813,
+      "learning_rate": 0.0002,
+      "loss": 2.529,
+      "step": 23860
+    },
+    {
+      "epoch": 1.778688524590164,
+      "grad_norm": 2.5092759132385254,
+      "learning_rate": 0.0002,
+      "loss": 2.3547,
+      "step": 23870
+    },
+    {
+      "epoch": 1.7794336810730254,
+      "grad_norm": 2.5346925258636475,
+      "learning_rate": 0.0002,
+      "loss": 2.5704,
+      "step": 23880
+    },
+    {
+      "epoch": 1.7801788375558867,
+      "grad_norm": 2.1645240783691406,
+      "learning_rate": 0.0002,
+      "loss": 2.4889,
+      "step": 23890
+    },
+    {
+      "epoch": 1.7809239940387482,
+      "grad_norm": 2.653005361557007,
+      "learning_rate": 0.0002,
+      "loss": 2.4614,
+      "step": 23900
+    },
+    {
+      "epoch": 1.7816691505216096,
+      "grad_norm": 3.1483113765716553,
+      "learning_rate": 0.0002,
+      "loss": 2.5343,
+      "step": 23910
+    },
+    {
+      "epoch": 1.782414307004471,
+      "grad_norm": 2.815312385559082,
+      "learning_rate": 0.0002,
+      "loss": 2.491,
+      "step": 23920
+    },
+    {
+      "epoch": 1.7831594634873325,
+      "grad_norm": 2.5819027423858643,
+      "learning_rate": 0.0002,
+      "loss": 2.4377,
+      "step": 23930
+    },
+    {
+      "epoch": 1.7839046199701938,
+      "grad_norm": 2.691847085952759,
+      "learning_rate": 0.0002,
+      "loss": 2.5363,
+      "step": 23940
+    },
+    {
+      "epoch": 1.7846497764530551,
+      "grad_norm": 3.1236226558685303,
+      "learning_rate": 0.0002,
+      "loss": 2.6327,
+      "step": 23950
+    },
+    {
+      "epoch": 1.7853949329359167,
+      "grad_norm": 2.2760941982269287,
+      "learning_rate": 0.0002,
+      "loss": 2.4321,
+      "step": 23960
+    },
+    {
+      "epoch": 1.786140089418778,
+      "grad_norm": 2.5152201652526855,
+      "learning_rate": 0.0002,
+      "loss": 2.519,
+      "step": 23970
+    },
+    {
+      "epoch": 1.7868852459016393,
+      "grad_norm": 2.2634105682373047,
+      "learning_rate": 0.0002,
+      "loss": 2.4622,
+      "step": 23980
+    },
+    {
+      "epoch": 1.7876304023845009,
+      "grad_norm": 2.6219863891601562,
+      "learning_rate": 0.0002,
+      "loss": 2.6106,
+      "step": 23990
+    },
+    {
+      "epoch": 1.788375558867362,
+      "grad_norm": 3.093374013900757,
+      "learning_rate": 0.0002,
+      "loss": 2.4982,
+      "step": 24000
+    },
+    {
+      "epoch": 1.7891207153502235,
+      "grad_norm": 2.5569963455200195,
+      "learning_rate": 0.0002,
+      "loss": 2.4287,
+      "step": 24010
+    },
+    {
+      "epoch": 1.789865871833085,
+      "grad_norm": 2.514453887939453,
+      "learning_rate": 0.0002,
+      "loss": 2.5517,
+      "step": 24020
+    },
+    {
+      "epoch": 1.7906110283159462,
+      "grad_norm": 2.578756809234619,
+      "learning_rate": 0.0002,
+      "loss": 2.5094,
+      "step": 24030
+    },
+    {
+      "epoch": 1.7913561847988078,
+      "grad_norm": 2.6853983402252197,
+      "learning_rate": 0.0002,
+      "loss": 2.6338,
+      "step": 24040
+    },
+    {
+      "epoch": 1.7921013412816693,
+      "grad_norm": 2.5106022357940674,
+      "learning_rate": 0.0002,
+      "loss": 2.5402,
+      "step": 24050
+    },
+    {
+      "epoch": 1.7928464977645304,
+      "grad_norm": 2.823246479034424,
+      "learning_rate": 0.0002,
+      "loss": 2.321,
+      "step": 24060
+    },
+    {
+      "epoch": 1.793591654247392,
+      "grad_norm": 2.782550096511841,
+      "learning_rate": 0.0002,
+      "loss": 2.6001,
+      "step": 24070
+    },
+    {
+      "epoch": 1.7943368107302533,
+      "grad_norm": 2.8128268718719482,
+      "learning_rate": 0.0002,
+      "loss": 2.3526,
+      "step": 24080
+    },
+    {
+      "epoch": 1.7950819672131146,
+      "grad_norm": 2.4673032760620117,
+      "learning_rate": 0.0002,
+      "loss": 2.2312,
+      "step": 24090
+    },
+    {
+      "epoch": 1.7958271236959762,
+      "grad_norm": 2.503586530685425,
+      "learning_rate": 0.0002,
+      "loss": 2.4475,
+      "step": 24100
+    },
+    {
+      "epoch": 1.7965722801788375,
+      "grad_norm": 2.3996407985687256,
+      "learning_rate": 0.0002,
+      "loss": 2.5589,
+      "step": 24110
+    },
+    {
+      "epoch": 1.7973174366616989,
+      "grad_norm": 2.5348939895629883,
+      "learning_rate": 0.0002,
+      "loss": 2.2477,
+      "step": 24120
+    },
+    {
+      "epoch": 1.7980625931445604,
+      "grad_norm": 2.695479393005371,
+      "learning_rate": 0.0002,
+      "loss": 2.5913,
+      "step": 24130
+    },
+    {
+      "epoch": 1.7988077496274217,
+      "grad_norm": 2.827122211456299,
+      "learning_rate": 0.0002,
+      "loss": 2.3807,
+      "step": 24140
+    },
+    {
+      "epoch": 1.799552906110283,
+      "grad_norm": 2.496100425720215,
+      "learning_rate": 0.0002,
+      "loss": 2.4569,
+      "step": 24150
+    },
+    {
+      "epoch": 1.8002980625931446,
+      "grad_norm": 2.7637364864349365,
+      "learning_rate": 0.0002,
+      "loss": 2.4109,
+      "step": 24160
+    },
+    {
+      "epoch": 1.801043219076006,
+      "grad_norm": 2.378218650817871,
+      "learning_rate": 0.0002,
+      "loss": 2.3436,
+      "step": 24170
+    },
+    {
+      "epoch": 1.8017883755588673,
+      "grad_norm": 2.8666129112243652,
+      "learning_rate": 0.0002,
+      "loss": 2.6058,
+      "step": 24180
+    },
+    {
+      "epoch": 1.8025335320417288,
+      "grad_norm": 2.1839277744293213,
+      "learning_rate": 0.0002,
+      "loss": 2.3226,
+      "step": 24190
+    },
+    {
+      "epoch": 1.8032786885245902,
+      "grad_norm": 2.1608381271362305,
+      "learning_rate": 0.0002,
+      "loss": 2.5368,
+      "step": 24200
+    },
+    {
+      "epoch": 1.8040238450074515,
+      "grad_norm": 2.1796987056732178,
+      "learning_rate": 0.0002,
+      "loss": 2.5402,
+      "step": 24210
+    },
+    {
+      "epoch": 1.804769001490313,
+      "grad_norm": 2.9191579818725586,
+      "learning_rate": 0.0002,
+      "loss": 2.5582,
+      "step": 24220
+    },
+    {
+      "epoch": 1.8055141579731744,
+      "grad_norm": 2.621483087539673,
+      "learning_rate": 0.0002,
+      "loss": 2.4998,
+      "step": 24230
+    },
+    {
+      "epoch": 1.8062593144560357,
+      "grad_norm": 2.5693342685699463,
+      "learning_rate": 0.0002,
+      "loss": 2.6363,
+      "step": 24240
+    },
+    {
+      "epoch": 1.8070044709388973,
+      "grad_norm": 2.3040270805358887,
+      "learning_rate": 0.0002,
+      "loss": 2.3897,
+      "step": 24250
+    },
+    {
+      "epoch": 1.8077496274217586,
+      "grad_norm": 2.5259714126586914,
+      "learning_rate": 0.0002,
+      "loss": 2.3625,
+      "step": 24260
+    },
+    {
+      "epoch": 1.80849478390462,
+      "grad_norm": 2.1863303184509277,
+      "learning_rate": 0.0002,
+      "loss": 2.5026,
+      "step": 24270
+    },
+    {
+      "epoch": 1.8092399403874815,
+      "grad_norm": 2.6622605323791504,
+      "learning_rate": 0.0002,
+      "loss": 2.5436,
+      "step": 24280
+    },
+    {
+      "epoch": 1.8099850968703428,
+      "grad_norm": 2.9189093112945557,
+      "learning_rate": 0.0002,
+      "loss": 2.6779,
+      "step": 24290
+    },
+    {
+      "epoch": 1.8107302533532041,
+      "grad_norm": 2.7089121341705322,
+      "learning_rate": 0.0002,
+      "loss": 2.6477,
+      "step": 24300
+    },
+    {
+      "epoch": 1.8114754098360657,
+      "grad_norm": 2.326711654663086,
+      "learning_rate": 0.0002,
+      "loss": 2.5291,
+      "step": 24310
+    },
+    {
+      "epoch": 1.812220566318927,
+      "grad_norm": 2.570425033569336,
+      "learning_rate": 0.0002,
+      "loss": 2.4593,
+      "step": 24320
+    },
+    {
+      "epoch": 1.8129657228017884,
+      "grad_norm": 3.20462703704834,
+      "learning_rate": 0.0002,
+      "loss": 2.5166,
+      "step": 24330
+    },
+    {
+      "epoch": 1.81371087928465,
+      "grad_norm": 2.2558157444000244,
+      "learning_rate": 0.0002,
+      "loss": 2.5881,
+      "step": 24340
+    },
+    {
+      "epoch": 1.814456035767511,
+      "grad_norm": 2.6994922161102295,
+      "learning_rate": 0.0002,
+      "loss": 2.5089,
+      "step": 24350
+    },
+    {
+      "epoch": 1.8152011922503726,
+      "grad_norm": 2.84627103805542,
+      "learning_rate": 0.0002,
+      "loss": 2.4608,
+      "step": 24360
+    },
+    {
+      "epoch": 1.8159463487332341,
+      "grad_norm": 2.4853367805480957,
+      "learning_rate": 0.0002,
+      "loss": 2.2028,
+      "step": 24370
+    },
+    {
+      "epoch": 1.8166915052160952,
+      "grad_norm": 2.2818758487701416,
+      "learning_rate": 0.0002,
+      "loss": 2.471,
+      "step": 24380
+    },
+    {
+      "epoch": 1.8174366616989568,
+      "grad_norm": 2.9547371864318848,
+      "learning_rate": 0.0002,
+      "loss": 2.5479,
+      "step": 24390
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 2.700378656387329,
+      "learning_rate": 0.0002,
+      "loss": 2.5568,
+      "step": 24400
+    },
+    {
+      "epoch": 1.8189269746646795,
+      "grad_norm": 2.5123701095581055,
+      "learning_rate": 0.0002,
+      "loss": 2.3738,
+      "step": 24410
+    },
+    {
+      "epoch": 1.819672131147541,
+      "grad_norm": 2.8290090560913086,
+      "learning_rate": 0.0002,
+      "loss": 2.5638,
+      "step": 24420
+    },
+    {
+      "epoch": 1.8204172876304023,
+      "grad_norm": 2.747896671295166,
+      "learning_rate": 0.0002,
+      "loss": 2.754,
+      "step": 24430
+    },
+    {
+      "epoch": 1.8211624441132637,
+      "grad_norm": 2.2470264434814453,
+      "learning_rate": 0.0002,
+      "loss": 2.4369,
+      "step": 24440
+    },
+    {
+      "epoch": 1.8219076005961252,
+      "grad_norm": 2.4347453117370605,
+      "learning_rate": 0.0002,
+      "loss": 2.4282,
+      "step": 24450
+    },
+    {
+      "epoch": 1.8226527570789866,
+      "grad_norm": 2.6316144466400146,
+      "learning_rate": 0.0002,
+      "loss": 2.4912,
+      "step": 24460
+    },
+    {
+      "epoch": 1.8233979135618479,
+      "grad_norm": 2.4304425716400146,
+      "learning_rate": 0.0002,
+      "loss": 2.5277,
+      "step": 24470
+    },
+    {
+      "epoch": 1.8241430700447094,
+      "grad_norm": 2.7673425674438477,
+      "learning_rate": 0.0002,
+      "loss": 2.6021,
+      "step": 24480
+    },
+    {
+      "epoch": 1.8248882265275708,
+      "grad_norm": 2.2696149349212646,
+      "learning_rate": 0.0002,
+      "loss": 2.4493,
+      "step": 24490
+    },
+    {
+      "epoch": 1.825633383010432,
+      "grad_norm": 2.58644962310791,
+      "learning_rate": 0.0002,
+      "loss": 2.5297,
+      "step": 24500
+    },
+    {
+      "epoch": 1.8263785394932937,
+      "grad_norm": 3.001873254776001,
+      "learning_rate": 0.0002,
+      "loss": 2.3753,
+      "step": 24510
+    },
+    {
+      "epoch": 1.827123695976155,
+      "grad_norm": 2.724271297454834,
+      "learning_rate": 0.0002,
+      "loss": 2.4127,
+      "step": 24520
+    },
+    {
+      "epoch": 1.8278688524590163,
+      "grad_norm": 2.4911880493164062,
+      "learning_rate": 0.0002,
+      "loss": 2.7232,
+      "step": 24530
+    },
+    {
+      "epoch": 1.8286140089418779,
+      "grad_norm": 2.659914493560791,
+      "learning_rate": 0.0002,
+      "loss": 2.4794,
+      "step": 24540
+    },
+    {
+      "epoch": 1.8293591654247392,
+      "grad_norm": 2.4134860038757324,
+      "learning_rate": 0.0002,
+      "loss": 2.5825,
+      "step": 24550
+    },
+    {
+      "epoch": 1.8301043219076005,
+      "grad_norm": 2.6322641372680664,
+      "learning_rate": 0.0002,
+      "loss": 2.4633,
+      "step": 24560
+    },
+    {
+      "epoch": 1.830849478390462,
+      "grad_norm": 2.942608118057251,
+      "learning_rate": 0.0002,
+      "loss": 2.7876,
+      "step": 24570
+    },
+    {
+      "epoch": 1.8315946348733234,
+      "grad_norm": 2.5157153606414795,
+      "learning_rate": 0.0002,
+      "loss": 2.4832,
+      "step": 24580
+    },
+    {
+      "epoch": 1.8323397913561847,
+      "grad_norm": 2.5834529399871826,
+      "learning_rate": 0.0002,
+      "loss": 2.4591,
+      "step": 24590
+    },
+    {
+      "epoch": 1.8330849478390463,
+      "grad_norm": 2.607112169265747,
+      "learning_rate": 0.0002,
+      "loss": 2.5687,
+      "step": 24600
+    },
+    {
+      "epoch": 1.8338301043219076,
+      "grad_norm": 2.7782909870147705,
+      "learning_rate": 0.0002,
+      "loss": 2.5903,
+      "step": 24610
+    },
+    {
+      "epoch": 1.834575260804769,
+      "grad_norm": 2.6779706478118896,
+      "learning_rate": 0.0002,
+      "loss": 2.561,
+      "step": 24620
+    },
+    {
+      "epoch": 1.8353204172876305,
+      "grad_norm": 2.514186143875122,
+      "learning_rate": 0.0002,
+      "loss": 2.5755,
+      "step": 24630
+    },
+    {
+      "epoch": 1.8360655737704918,
+      "grad_norm": 2.3403496742248535,
+      "learning_rate": 0.0002,
+      "loss": 2.4768,
+      "step": 24640
+    },
+    {
+      "epoch": 1.8368107302533532,
+      "grad_norm": 2.3311824798583984,
+      "learning_rate": 0.0002,
+      "loss": 2.5229,
+      "step": 24650
+    },
+    {
+      "epoch": 1.8375558867362147,
+      "grad_norm": 1.7865532636642456,
+      "learning_rate": 0.0002,
+      "loss": 2.5142,
+      "step": 24660
+    },
+    {
+      "epoch": 1.838301043219076,
+      "grad_norm": 2.4601778984069824,
+      "learning_rate": 0.0002,
+      "loss": 2.5981,
+      "step": 24670
+    },
+    {
+      "epoch": 1.8390461997019374,
+      "grad_norm": 2.6825180053710938,
+      "learning_rate": 0.0002,
+      "loss": 2.5526,
+      "step": 24680
+    },
+    {
+      "epoch": 1.839791356184799,
+      "grad_norm": 2.428622245788574,
+      "learning_rate": 0.0002,
+      "loss": 2.2803,
+      "step": 24690
+    },
+    {
+      "epoch": 1.84053651266766,
+      "grad_norm": 2.353142261505127,
+      "learning_rate": 0.0002,
+      "loss": 2.4896,
+      "step": 24700
+    },
+    {
+      "epoch": 1.8412816691505216,
+      "grad_norm": 2.611560106277466,
+      "learning_rate": 0.0002,
+      "loss": 2.6822,
+      "step": 24710
+    },
+    {
+      "epoch": 1.8420268256333832,
+      "grad_norm": 2.8267176151275635,
+      "learning_rate": 0.0002,
+      "loss": 2.6869,
+      "step": 24720
+    },
+    {
+      "epoch": 1.8427719821162443,
+      "grad_norm": 2.8418166637420654,
+      "learning_rate": 0.0002,
+      "loss": 2.5455,
+      "step": 24730
+    },
+    {
+      "epoch": 1.8435171385991058,
+      "grad_norm": 2.458555221557617,
+      "learning_rate": 0.0002,
+      "loss": 2.6303,
+      "step": 24740
+    },
+    {
+      "epoch": 1.8442622950819674,
+      "grad_norm": 2.2593302726745605,
+      "learning_rate": 0.0002,
+      "loss": 2.5129,
+      "step": 24750
+    },
+    {
+      "epoch": 1.8450074515648285,
+      "grad_norm": 2.7010560035705566,
+      "learning_rate": 0.0002,
+      "loss": 2.3973,
+      "step": 24760
+    },
+    {
+      "epoch": 1.84575260804769,
+      "grad_norm": 2.4872236251831055,
+      "learning_rate": 0.0002,
+      "loss": 2.6178,
+      "step": 24770
+    },
+    {
+      "epoch": 1.8464977645305514,
+      "grad_norm": 2.882812738418579,
+      "learning_rate": 0.0002,
+      "loss": 2.533,
+      "step": 24780
+    },
+    {
+      "epoch": 1.8472429210134127,
+      "grad_norm": 2.354396104812622,
+      "learning_rate": 0.0002,
+      "loss": 2.5111,
+      "step": 24790
+    },
+    {
+      "epoch": 1.8479880774962743,
+      "grad_norm": 2.5242695808410645,
+      "learning_rate": 0.0002,
+      "loss": 2.5467,
+      "step": 24800
+    },
+    {
+      "epoch": 1.8487332339791356,
+      "grad_norm": 2.3869271278381348,
+      "learning_rate": 0.0002,
+      "loss": 2.578,
+      "step": 24810
+    },
+    {
+      "epoch": 1.849478390461997,
+      "grad_norm": 2.6933507919311523,
+      "learning_rate": 0.0002,
+      "loss": 2.6325,
+      "step": 24820
+    },
+    {
+      "epoch": 1.8502235469448585,
+      "grad_norm": 2.445830821990967,
+      "learning_rate": 0.0002,
+      "loss": 2.5435,
+      "step": 24830
+    },
+    {
+      "epoch": 1.8509687034277198,
+      "grad_norm": 2.1250483989715576,
+      "learning_rate": 0.0002,
+      "loss": 2.6424,
+      "step": 24840
+    },
+    {
+      "epoch": 1.8517138599105811,
+      "grad_norm": 2.1476426124572754,
+      "learning_rate": 0.0002,
+      "loss": 2.5599,
+      "step": 24850
+    },
+    {
+      "epoch": 1.8524590163934427,
+      "grad_norm": 2.7636313438415527,
+      "learning_rate": 0.0002,
+      "loss": 2.6807,
+      "step": 24860
+    },
+    {
+      "epoch": 1.853204172876304,
+      "grad_norm": 2.6150963306427,
+      "learning_rate": 0.0002,
+      "loss": 2.5823,
+      "step": 24870
+    },
+    {
+      "epoch": 1.8539493293591653,
+      "grad_norm": 2.9421913623809814,
+      "learning_rate": 0.0002,
+      "loss": 2.5598,
+      "step": 24880
+    },
+    {
+      "epoch": 1.854694485842027,
+      "grad_norm": 2.576680898666382,
+      "learning_rate": 0.0002,
+      "loss": 2.4352,
+      "step": 24890
+    },
+    {
+      "epoch": 1.8554396423248882,
+      "grad_norm": 2.372847557067871,
+      "learning_rate": 0.0002,
+      "loss": 2.4352,
+      "step": 24900
+    },
+    {
+      "epoch": 1.8561847988077496,
+      "grad_norm": 2.7840962409973145,
+      "learning_rate": 0.0002,
+      "loss": 2.4458,
+      "step": 24910
+    },
+    {
+      "epoch": 1.8569299552906111,
+      "grad_norm": 2.608435869216919,
+      "learning_rate": 0.0002,
+      "loss": 2.3279,
+      "step": 24920
+    },
+    {
+      "epoch": 1.8576751117734724,
+      "grad_norm": 2.2083051204681396,
+      "learning_rate": 0.0002,
+      "loss": 2.5685,
+      "step": 24930
+    },
+    {
+      "epoch": 1.8584202682563338,
+      "grad_norm": 2.6166601181030273,
+      "learning_rate": 0.0002,
+      "loss": 2.5726,
+      "step": 24940
+    },
+    {
+      "epoch": 1.8591654247391953,
+      "grad_norm": 2.4802517890930176,
+      "learning_rate": 0.0002,
+      "loss": 2.4114,
+      "step": 24950
+    },
+    {
+      "epoch": 1.8599105812220567,
+      "grad_norm": 2.6344408988952637,
+      "learning_rate": 0.0002,
+      "loss": 2.6164,
+      "step": 24960
+    },
+    {
+      "epoch": 1.860655737704918,
+      "grad_norm": 2.4686670303344727,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 24970
+    },
+    {
+      "epoch": 1.8614008941877795,
+      "grad_norm": 2.556004762649536,
+      "learning_rate": 0.0002,
+      "loss": 2.5358,
+      "step": 24980
+    },
+    {
+      "epoch": 1.8621460506706409,
+      "grad_norm": 2.642232656478882,
+      "learning_rate": 0.0002,
+      "loss": 2.4455,
+      "step": 24990
+    },
+    {
+      "epoch": 1.8628912071535022,
+      "grad_norm": 2.6357104778289795,
+      "learning_rate": 0.0002,
+      "loss": 2.4734,
+      "step": 25000
+    },
+    {
+      "epoch": 1.8636363636363638,
+      "grad_norm": 2.7195780277252197,
+      "learning_rate": 0.0002,
+      "loss": 2.5011,
+      "step": 25010
+    },
+    {
+      "epoch": 1.864381520119225,
+      "grad_norm": 2.306607961654663,
+      "learning_rate": 0.0002,
+      "loss": 2.4745,
+      "step": 25020
+    },
+    {
+      "epoch": 1.8651266766020864,
+      "grad_norm": 2.3458077907562256,
+      "learning_rate": 0.0002,
+      "loss": 2.5006,
+      "step": 25030
+    },
+    {
+      "epoch": 1.865871833084948,
+      "grad_norm": 2.6288564205169678,
+      "learning_rate": 0.0002,
+      "loss": 2.4656,
+      "step": 25040
+    },
+    {
+      "epoch": 1.866616989567809,
+      "grad_norm": 2.5845961570739746,
+      "learning_rate": 0.0002,
+      "loss": 2.4353,
+      "step": 25050
+    },
+    {
+      "epoch": 1.8673621460506706,
+      "grad_norm": 2.3820455074310303,
+      "learning_rate": 0.0002,
+      "loss": 2.5236,
+      "step": 25060
+    },
+    {
+      "epoch": 1.8681073025335322,
+      "grad_norm": 2.8274917602539062,
+      "learning_rate": 0.0002,
+      "loss": 2.5278,
+      "step": 25070
+    },
+    {
+      "epoch": 1.8688524590163933,
+      "grad_norm": 2.551588296890259,
+      "learning_rate": 0.0002,
+      "loss": 2.5355,
+      "step": 25080
+    },
+    {
+      "epoch": 1.8695976154992549,
+      "grad_norm": 2.707473039627075,
+      "learning_rate": 0.0002,
+      "loss": 2.4049,
+      "step": 25090
+    },
+    {
+      "epoch": 1.8703427719821164,
+      "grad_norm": 2.40470290184021,
+      "learning_rate": 0.0002,
+      "loss": 2.5121,
+      "step": 25100
+    },
+    {
+      "epoch": 1.8710879284649775,
+      "grad_norm": 2.507206916809082,
+      "learning_rate": 0.0002,
+      "loss": 2.2627,
+      "step": 25110
+    },
+    {
+      "epoch": 1.871833084947839,
+      "grad_norm": 2.4339914321899414,
+      "learning_rate": 0.0002,
+      "loss": 2.5134,
+      "step": 25120
+    },
+    {
+      "epoch": 1.8725782414307004,
+      "grad_norm": 2.4960076808929443,
+      "learning_rate": 0.0002,
+      "loss": 2.5428,
+      "step": 25130
+    },
+    {
+      "epoch": 1.8733233979135617,
+      "grad_norm": 3.661783218383789,
+      "learning_rate": 0.0002,
+      "loss": 2.6599,
+      "step": 25140
+    },
+    {
+      "epoch": 1.8740685543964233,
+      "grad_norm": 2.513418197631836,
+      "learning_rate": 0.0002,
+      "loss": 2.4901,
+      "step": 25150
+    },
+    {
+      "epoch": 1.8748137108792846,
+      "grad_norm": 2.3090083599090576,
+      "learning_rate": 0.0002,
+      "loss": 2.5672,
+      "step": 25160
+    },
+    {
+      "epoch": 1.875558867362146,
+      "grad_norm": 2.645890951156616,
+      "learning_rate": 0.0002,
+      "loss": 2.5898,
+      "step": 25170
+    },
+    {
+      "epoch": 1.8763040238450075,
+      "grad_norm": 3.336200714111328,
+      "learning_rate": 0.0002,
+      "loss": 2.6214,
+      "step": 25180
+    },
+    {
+      "epoch": 1.8770491803278688,
+      "grad_norm": 2.3214824199676514,
+      "learning_rate": 0.0002,
+      "loss": 2.6054,
+      "step": 25190
+    },
+    {
+      "epoch": 1.8777943368107302,
+      "grad_norm": 2.3875439167022705,
+      "learning_rate": 0.0002,
+      "loss": 2.5562,
+      "step": 25200
+    },
+    {
+      "epoch": 1.8785394932935917,
+      "grad_norm": 2.643754482269287,
+      "learning_rate": 0.0002,
+      "loss": 2.5102,
+      "step": 25210
+    },
+    {
+      "epoch": 1.879284649776453,
+      "grad_norm": 2.7053935527801514,
+      "learning_rate": 0.0002,
+      "loss": 2.3707,
+      "step": 25220
+    },
+    {
+      "epoch": 1.8800298062593144,
+      "grad_norm": 2.6861038208007812,
+      "learning_rate": 0.0002,
+      "loss": 2.7176,
+      "step": 25230
+    },
+    {
+      "epoch": 1.880774962742176,
+      "grad_norm": 2.6391537189483643,
+      "learning_rate": 0.0002,
+      "loss": 2.3108,
+      "step": 25240
+    },
+    {
+      "epoch": 1.8815201192250373,
+      "grad_norm": 2.6725473403930664,
+      "learning_rate": 0.0002,
+      "loss": 2.6606,
+      "step": 25250
+    },
+    {
+      "epoch": 1.8822652757078986,
+      "grad_norm": 2.843163013458252,
+      "learning_rate": 0.0002,
+      "loss": 2.4298,
+      "step": 25260
+    },
+    {
+      "epoch": 1.8830104321907601,
+      "grad_norm": 2.280057430267334,
+      "learning_rate": 0.0002,
+      "loss": 2.5212,
+      "step": 25270
+    },
+    {
+      "epoch": 1.8837555886736215,
+      "grad_norm": 2.221116304397583,
+      "learning_rate": 0.0002,
+      "loss": 2.4512,
+      "step": 25280
+    },
+    {
+      "epoch": 1.8845007451564828,
+      "grad_norm": 2.413114547729492,
+      "learning_rate": 0.0002,
+      "loss": 2.4593,
+      "step": 25290
+    },
+    {
+      "epoch": 1.8852459016393444,
+      "grad_norm": 2.225498914718628,
+      "learning_rate": 0.0002,
+      "loss": 2.3172,
+      "step": 25300
+    },
+    {
+      "epoch": 1.8859910581222057,
+      "grad_norm": 2.6563146114349365,
+      "learning_rate": 0.0002,
+      "loss": 2.5154,
+      "step": 25310
+    },
+    {
+      "epoch": 1.886736214605067,
+      "grad_norm": 2.232198715209961,
+      "learning_rate": 0.0002,
+      "loss": 2.3323,
+      "step": 25320
+    },
+    {
+      "epoch": 1.8874813710879286,
+      "grad_norm": 2.4613037109375,
+      "learning_rate": 0.0002,
+      "loss": 2.5738,
+      "step": 25330
+    },
+    {
+      "epoch": 1.88822652757079,
+      "grad_norm": 2.7953736782073975,
+      "learning_rate": 0.0002,
+      "loss": 2.4914,
+      "step": 25340
+    },
+    {
+      "epoch": 1.8889716840536512,
+      "grad_norm": 2.416280508041382,
+      "learning_rate": 0.0002,
+      "loss": 2.643,
+      "step": 25350
+    },
+    {
+      "epoch": 1.8897168405365128,
+      "grad_norm": 2.5031518936157227,
+      "learning_rate": 0.0002,
+      "loss": 2.3818,
+      "step": 25360
+    },
+    {
+      "epoch": 1.8904619970193741,
+      "grad_norm": 2.5476391315460205,
+      "learning_rate": 0.0002,
+      "loss": 2.6249,
+      "step": 25370
+    },
+    {
+      "epoch": 1.8912071535022354,
+      "grad_norm": 2.5421531200408936,
+      "learning_rate": 0.0002,
+      "loss": 2.6841,
+      "step": 25380
+    },
+    {
+      "epoch": 1.891952309985097,
+      "grad_norm": 2.7282488346099854,
+      "learning_rate": 0.0002,
+      "loss": 2.6655,
+      "step": 25390
+    },
+    {
+      "epoch": 1.8926974664679581,
+      "grad_norm": 2.47148060798645,
+      "learning_rate": 0.0002,
+      "loss": 2.3096,
+      "step": 25400
+    },
+    {
+      "epoch": 1.8934426229508197,
+      "grad_norm": 2.394070625305176,
+      "learning_rate": 0.0002,
+      "loss": 2.6875,
+      "step": 25410
+    },
+    {
+      "epoch": 1.8941877794336812,
+      "grad_norm": 2.7232706546783447,
+      "learning_rate": 0.0002,
+      "loss": 2.5139,
+      "step": 25420
+    },
+    {
+      "epoch": 1.8949329359165423,
+      "grad_norm": 2.7696988582611084,
+      "learning_rate": 0.0002,
+      "loss": 2.5473,
+      "step": 25430
+    },
+    {
+      "epoch": 1.8956780923994039,
+      "grad_norm": 2.484501838684082,
+      "learning_rate": 0.0002,
+      "loss": 2.6315,
+      "step": 25440
+    },
+    {
+      "epoch": 1.8964232488822654,
+      "grad_norm": 2.669006109237671,
+      "learning_rate": 0.0002,
+      "loss": 2.4127,
+      "step": 25450
+    },
+    {
+      "epoch": 1.8971684053651265,
+      "grad_norm": 2.8366858959198,
+      "learning_rate": 0.0002,
+      "loss": 2.6871,
+      "step": 25460
+    },
+    {
+      "epoch": 1.897913561847988,
+      "grad_norm": 2.7077786922454834,
+      "learning_rate": 0.0002,
+      "loss": 2.3913,
+      "step": 25470
+    },
+    {
+      "epoch": 1.8986587183308494,
+      "grad_norm": 2.649768829345703,
+      "learning_rate": 0.0002,
+      "loss": 2.6283,
+      "step": 25480
+    },
+    {
+      "epoch": 1.8994038748137108,
+      "grad_norm": 2.4160945415496826,
+      "learning_rate": 0.0002,
+      "loss": 2.5643,
+      "step": 25490
+    },
+    {
+      "epoch": 1.9001490312965723,
+      "grad_norm": 2.4574332237243652,
+      "learning_rate": 0.0002,
+      "loss": 2.1872,
+      "step": 25500
+    },
+    {
+      "epoch": 1.9008941877794336,
+      "grad_norm": 2.4507484436035156,
+      "learning_rate": 0.0002,
+      "loss": 2.532,
+      "step": 25510
+    },
+    {
+      "epoch": 1.901639344262295,
+      "grad_norm": 1.882723093032837,
+      "learning_rate": 0.0002,
+      "loss": 2.4446,
+      "step": 25520
+    },
+    {
+      "epoch": 1.9023845007451565,
+      "grad_norm": 2.4363670349121094,
+      "learning_rate": 0.0002,
+      "loss": 2.3938,
+      "step": 25530
+    },
+    {
+      "epoch": 1.9031296572280179,
+      "grad_norm": 2.288465738296509,
+      "learning_rate": 0.0002,
+      "loss": 2.5107,
+      "step": 25540
+    },
+    {
+      "epoch": 1.9038748137108792,
+      "grad_norm": 2.611039400100708,
+      "learning_rate": 0.0002,
+      "loss": 2.5329,
+      "step": 25550
+    },
+    {
+      "epoch": 1.9046199701937407,
+      "grad_norm": 2.746476173400879,
+      "learning_rate": 0.0002,
+      "loss": 2.4759,
+      "step": 25560
+    },
+    {
+      "epoch": 1.905365126676602,
+      "grad_norm": 2.628481864929199,
+      "learning_rate": 0.0002,
+      "loss": 2.6533,
+      "step": 25570
+    },
+    {
+      "epoch": 1.9061102831594634,
+      "grad_norm": 2.5151212215423584,
+      "learning_rate": 0.0002,
+      "loss": 2.4787,
+      "step": 25580
+    },
+    {
+      "epoch": 1.906855439642325,
+      "grad_norm": 2.3716254234313965,
+      "learning_rate": 0.0002,
+      "loss": 2.5077,
+      "step": 25590
+    },
+    {
+      "epoch": 1.9076005961251863,
+      "grad_norm": 2.687274932861328,
+      "learning_rate": 0.0002,
+      "loss": 2.3605,
+      "step": 25600
+    },
+    {
+      "epoch": 1.9083457526080476,
+      "grad_norm": 2.3180296421051025,
+      "learning_rate": 0.0002,
+      "loss": 2.4078,
+      "step": 25610
+    },
+    {
+      "epoch": 1.9090909090909092,
+      "grad_norm": 2.033092737197876,
+      "learning_rate": 0.0002,
+      "loss": 2.2995,
+      "step": 25620
+    },
+    {
+      "epoch": 1.9098360655737705,
+      "grad_norm": 2.2614009380340576,
+      "learning_rate": 0.0002,
+      "loss": 2.5323,
+      "step": 25630
+    },
+    {
+      "epoch": 1.9105812220566318,
+      "grad_norm": 2.1245622634887695,
+      "learning_rate": 0.0002,
+      "loss": 2.4616,
+      "step": 25640
+    },
+    {
+      "epoch": 1.9113263785394934,
+      "grad_norm": 2.5503997802734375,
+      "learning_rate": 0.0002,
+      "loss": 2.6637,
+      "step": 25650
+    },
+    {
+      "epoch": 1.9120715350223547,
+      "grad_norm": 2.7903940677642822,
+      "learning_rate": 0.0002,
+      "loss": 2.6795,
+      "step": 25660
+    },
+    {
+      "epoch": 1.912816691505216,
+      "grad_norm": 2.427300453186035,
+      "learning_rate": 0.0002,
+      "loss": 2.4091,
+      "step": 25670
+    },
+    {
+      "epoch": 1.9135618479880776,
+      "grad_norm": 2.535790205001831,
+      "learning_rate": 0.0002,
+      "loss": 2.3897,
+      "step": 25680
+    },
+    {
+      "epoch": 1.914307004470939,
+      "grad_norm": 2.60248064994812,
+      "learning_rate": 0.0002,
+      "loss": 2.5952,
+      "step": 25690
+    },
+    {
+      "epoch": 1.9150521609538003,
+      "grad_norm": 2.7798526287078857,
+      "learning_rate": 0.0002,
+      "loss": 2.4811,
+      "step": 25700
+    },
+    {
+      "epoch": 1.9157973174366618,
+      "grad_norm": 2.0604312419891357,
+      "learning_rate": 0.0002,
+      "loss": 2.6361,
+      "step": 25710
+    },
+    {
+      "epoch": 1.9165424739195231,
+      "grad_norm": 2.663926124572754,
+      "learning_rate": 0.0002,
+      "loss": 2.4111,
+      "step": 25720
+    },
+    {
+      "epoch": 1.9172876304023845,
+      "grad_norm": 2.741672992706299,
+      "learning_rate": 0.0002,
+      "loss": 2.391,
+      "step": 25730
+    },
+    {
+      "epoch": 1.918032786885246,
+      "grad_norm": 2.5494937896728516,
+      "learning_rate": 0.0002,
+      "loss": 2.497,
+      "step": 25740
+    },
+    {
+      "epoch": 1.9187779433681071,
+      "grad_norm": 2.7732324600219727,
+      "learning_rate": 0.0002,
+      "loss": 2.3377,
+      "step": 25750
+    },
+    {
+      "epoch": 1.9195230998509687,
+      "grad_norm": 2.3186347484588623,
+      "learning_rate": 0.0002,
+      "loss": 2.6408,
+      "step": 25760
+    },
+    {
+      "epoch": 1.9202682563338302,
+      "grad_norm": 2.6721131801605225,
+      "learning_rate": 0.0002,
+      "loss": 2.5869,
+      "step": 25770
+    },
+    {
+      "epoch": 1.9210134128166914,
+      "grad_norm": 2.4322500228881836,
+      "learning_rate": 0.0002,
+      "loss": 2.4635,
+      "step": 25780
+    },
+    {
+      "epoch": 1.921758569299553,
+      "grad_norm": 2.7271809577941895,
+      "learning_rate": 0.0002,
+      "loss": 2.6242,
+      "step": 25790
+    },
+    {
+      "epoch": 1.9225037257824145,
+      "grad_norm": 2.2470762729644775,
+      "learning_rate": 0.0002,
+      "loss": 2.5125,
+      "step": 25800
+    },
+    {
+      "epoch": 1.9232488822652756,
+      "grad_norm": 2.861534833908081,
+      "learning_rate": 0.0002,
+      "loss": 2.5176,
+      "step": 25810
+    },
+    {
+      "epoch": 1.9239940387481371,
+      "grad_norm": 2.7653250694274902,
+      "learning_rate": 0.0002,
+      "loss": 2.4254,
+      "step": 25820
+    },
+    {
+      "epoch": 1.9247391952309985,
+      "grad_norm": 2.630427122116089,
+      "learning_rate": 0.0002,
+      "loss": 2.4022,
+      "step": 25830
+    },
+    {
+      "epoch": 1.9254843517138598,
+      "grad_norm": 2.5337681770324707,
+      "learning_rate": 0.0002,
+      "loss": 2.497,
+      "step": 25840
+    },
+    {
+      "epoch": 1.9262295081967213,
+      "grad_norm": 2.4699361324310303,
+      "learning_rate": 0.0002,
+      "loss": 2.6214,
+      "step": 25850
+    },
+    {
+      "epoch": 1.9269746646795827,
+      "grad_norm": 1.9499725103378296,
+      "learning_rate": 0.0002,
+      "loss": 2.4598,
+      "step": 25860
+    },
+    {
+      "epoch": 1.927719821162444,
+      "grad_norm": 2.4422264099121094,
+      "learning_rate": 0.0002,
+      "loss": 2.4889,
+      "step": 25870
+    },
+    {
+      "epoch": 1.9284649776453056,
+      "grad_norm": 2.2586450576782227,
+      "learning_rate": 0.0002,
+      "loss": 2.5206,
+      "step": 25880
+    },
+    {
+      "epoch": 1.9292101341281669,
+      "grad_norm": 2.6377463340759277,
+      "learning_rate": 0.0002,
+      "loss": 2.3293,
+      "step": 25890
+    },
+    {
+      "epoch": 1.9299552906110282,
+      "grad_norm": 2.7808732986450195,
+      "learning_rate": 0.0002,
+      "loss": 2.5287,
+      "step": 25900
+    },
+    {
+      "epoch": 1.9307004470938898,
+      "grad_norm": 3.0270299911499023,
+      "learning_rate": 0.0002,
+      "loss": 2.5512,
+      "step": 25910
+    },
+    {
+      "epoch": 1.931445603576751,
+      "grad_norm": 2.4635636806488037,
+      "learning_rate": 0.0002,
+      "loss": 2.5143,
+      "step": 25920
+    },
+    {
+      "epoch": 1.9321907600596124,
+      "grad_norm": 3.6546430587768555,
+      "learning_rate": 0.0002,
+      "loss": 2.4352,
+      "step": 25930
+    },
+    {
+      "epoch": 1.932935916542474,
+      "grad_norm": 2.299994707107544,
+      "learning_rate": 0.0002,
+      "loss": 2.4868,
+      "step": 25940
+    },
+    {
+      "epoch": 1.9336810730253353,
+      "grad_norm": 2.5070691108703613,
+      "learning_rate": 0.0002,
+      "loss": 2.5456,
+      "step": 25950
+    },
+    {
+      "epoch": 1.9344262295081966,
+      "grad_norm": 2.400691509246826,
+      "learning_rate": 0.0002,
+      "loss": 2.6418,
+      "step": 25960
+    },
+    {
+      "epoch": 1.9351713859910582,
+      "grad_norm": 2.6068758964538574,
+      "learning_rate": 0.0002,
+      "loss": 2.5137,
+      "step": 25970
+    },
+    {
+      "epoch": 1.9359165424739195,
+      "grad_norm": 2.5092999935150146,
+      "learning_rate": 0.0002,
+      "loss": 2.5346,
+      "step": 25980
+    },
+    {
+      "epoch": 1.9366616989567809,
+      "grad_norm": 2.6502528190612793,
+      "learning_rate": 0.0002,
+      "loss": 2.6351,
+      "step": 25990
+    },
+    {
+      "epoch": 1.9374068554396424,
+      "grad_norm": 2.406677722930908,
+      "learning_rate": 0.0002,
+      "loss": 2.4647,
+      "step": 26000
+    },
+    {
+      "epoch": 1.9381520119225037,
+      "grad_norm": 2.532393217086792,
+      "learning_rate": 0.0002,
+      "loss": 2.4782,
+      "step": 26010
+    },
+    {
+      "epoch": 1.938897168405365,
+      "grad_norm": 2.3414087295532227,
+      "learning_rate": 0.0002,
+      "loss": 2.4257,
+      "step": 26020
+    },
+    {
+      "epoch": 1.9396423248882266,
+      "grad_norm": 2.684752941131592,
+      "learning_rate": 0.0002,
+      "loss": 2.3947,
+      "step": 26030
+    },
+    {
+      "epoch": 1.940387481371088,
+      "grad_norm": 2.432410955429077,
+      "learning_rate": 0.0002,
+      "loss": 2.623,
+      "step": 26040
+    },
+    {
+      "epoch": 1.9411326378539493,
+      "grad_norm": 3.1778249740600586,
+      "learning_rate": 0.0002,
+      "loss": 2.4038,
+      "step": 26050
+    },
+    {
+      "epoch": 1.9418777943368108,
+      "grad_norm": 2.836688280105591,
+      "learning_rate": 0.0002,
+      "loss": 2.5303,
+      "step": 26060
+    },
+    {
+      "epoch": 1.9426229508196722,
+      "grad_norm": 2.4091341495513916,
+      "learning_rate": 0.0002,
+      "loss": 2.4783,
+      "step": 26070
+    },
+    {
+      "epoch": 1.9433681073025335,
+      "grad_norm": 2.5459396839141846,
+      "learning_rate": 0.0002,
+      "loss": 2.5745,
+      "step": 26080
+    },
+    {
+      "epoch": 1.944113263785395,
+      "grad_norm": 2.6074795722961426,
+      "learning_rate": 0.0002,
+      "loss": 2.5106,
+      "step": 26090
+    },
+    {
+      "epoch": 1.9448584202682562,
+      "grad_norm": 2.8824479579925537,
+      "learning_rate": 0.0002,
+      "loss": 2.4438,
+      "step": 26100
+    },
+    {
+      "epoch": 1.9456035767511177,
+      "grad_norm": 2.3653016090393066,
+      "learning_rate": 0.0002,
+      "loss": 2.5536,
+      "step": 26110
+    },
+    {
+      "epoch": 1.9463487332339793,
+      "grad_norm": 2.3197124004364014,
+      "learning_rate": 0.0002,
+      "loss": 2.4542,
+      "step": 26120
+    },
+    {
+      "epoch": 1.9470938897168404,
+      "grad_norm": 2.626267194747925,
+      "learning_rate": 0.0002,
+      "loss": 2.4828,
+      "step": 26130
+    },
+    {
+      "epoch": 1.947839046199702,
+      "grad_norm": 2.914698362350464,
+      "learning_rate": 0.0002,
+      "loss": 2.5634,
+      "step": 26140
+    },
+    {
+      "epoch": 1.9485842026825635,
+      "grad_norm": 2.8641860485076904,
+      "learning_rate": 0.0002,
+      "loss": 2.6208,
+      "step": 26150
+    },
+    {
+      "epoch": 1.9493293591654246,
+      "grad_norm": 2.823122262954712,
+      "learning_rate": 0.0002,
+      "loss": 2.5639,
+      "step": 26160
+    },
+    {
+      "epoch": 1.9500745156482862,
+      "grad_norm": 2.152498722076416,
+      "learning_rate": 0.0002,
+      "loss": 2.5226,
+      "step": 26170
+    },
+    {
+      "epoch": 1.9508196721311475,
+      "grad_norm": 2.7207183837890625,
+      "learning_rate": 0.0002,
+      "loss": 2.4951,
+      "step": 26180
+    },
+    {
+      "epoch": 1.9515648286140088,
+      "grad_norm": 2.4285998344421387,
+      "learning_rate": 0.0002,
+      "loss": 2.5207,
+      "step": 26190
+    },
+    {
+      "epoch": 1.9523099850968704,
+      "grad_norm": 2.4538893699645996,
+      "learning_rate": 0.0002,
+      "loss": 2.4034,
+      "step": 26200
+    },
+    {
+      "epoch": 1.9530551415797317,
+      "grad_norm": 2.513335704803467,
+      "learning_rate": 0.0002,
+      "loss": 2.4913,
+      "step": 26210
+    },
+    {
+      "epoch": 1.953800298062593,
+      "grad_norm": 2.8927149772644043,
+      "learning_rate": 0.0002,
+      "loss": 2.5541,
+      "step": 26220
+    },
+    {
+      "epoch": 1.9545454545454546,
+      "grad_norm": 2.610621452331543,
+      "learning_rate": 0.0002,
+      "loss": 2.3861,
+      "step": 26230
+    },
+    {
+      "epoch": 1.955290611028316,
+      "grad_norm": 2.7638297080993652,
+      "learning_rate": 0.0002,
+      "loss": 2.452,
+      "step": 26240
+    },
+    {
+      "epoch": 1.9560357675111772,
+      "grad_norm": 2.722166061401367,
+      "learning_rate": 0.0002,
+      "loss": 2.6235,
+      "step": 26250
+    },
+    {
+      "epoch": 1.9567809239940388,
+      "grad_norm": 2.614750862121582,
+      "learning_rate": 0.0002,
+      "loss": 2.4784,
+      "step": 26260
+    },
+    {
+      "epoch": 1.9575260804769001,
+      "grad_norm": 2.5103914737701416,
+      "learning_rate": 0.0002,
+      "loss": 2.5742,
+      "step": 26270
+    },
+    {
+      "epoch": 1.9582712369597615,
+      "grad_norm": 2.72584867477417,
+      "learning_rate": 0.0002,
+      "loss": 2.379,
+      "step": 26280
+    },
+    {
+      "epoch": 1.959016393442623,
+      "grad_norm": 2.343048095703125,
+      "learning_rate": 0.0002,
+      "loss": 2.5236,
+      "step": 26290
+    },
+    {
+      "epoch": 1.9597615499254843,
+      "grad_norm": 2.357050895690918,
+      "learning_rate": 0.0002,
+      "loss": 2.4987,
+      "step": 26300
+    },
+    {
+      "epoch": 1.9605067064083457,
+      "grad_norm": 2.71625018119812,
+      "learning_rate": 0.0002,
+      "loss": 2.596,
+      "step": 26310
+    },
+    {
+      "epoch": 1.9612518628912072,
+      "grad_norm": 2.790226459503174,
+      "learning_rate": 0.0002,
+      "loss": 2.4313,
+      "step": 26320
+    },
+    {
+      "epoch": 1.9619970193740686,
+      "grad_norm": 2.4327445030212402,
+      "learning_rate": 0.0002,
+      "loss": 2.5478,
+      "step": 26330
+    },
+    {
+      "epoch": 1.96274217585693,
+      "grad_norm": 2.637118339538574,
+      "learning_rate": 0.0002,
+      "loss": 2.638,
+      "step": 26340
+    },
+    {
+      "epoch": 1.9634873323397914,
+      "grad_norm": 2.398437023162842,
+      "learning_rate": 0.0002,
+      "loss": 2.3303,
+      "step": 26350
+    },
+    {
+      "epoch": 1.9642324888226528,
+      "grad_norm": 2.131314277648926,
+      "learning_rate": 0.0002,
+      "loss": 2.2881,
+      "step": 26360
+    },
+    {
+      "epoch": 1.964977645305514,
+      "grad_norm": 2.55584716796875,
+      "learning_rate": 0.0002,
+      "loss": 2.5168,
+      "step": 26370
+    },
+    {
+      "epoch": 1.9657228017883757,
+      "grad_norm": 2.5380923748016357,
+      "learning_rate": 0.0002,
+      "loss": 2.362,
+      "step": 26380
+    },
+    {
+      "epoch": 1.966467958271237,
+      "grad_norm": 2.3342056274414062,
+      "learning_rate": 0.0002,
+      "loss": 2.6316,
+      "step": 26390
+    },
+    {
+      "epoch": 1.9672131147540983,
+      "grad_norm": 2.483621835708618,
+      "learning_rate": 0.0002,
+      "loss": 2.4311,
+      "step": 26400
+    },
+    {
+      "epoch": 1.9679582712369599,
+      "grad_norm": 3.239633560180664,
+      "learning_rate": 0.0002,
+      "loss": 2.4224,
+      "step": 26410
+    },
+    {
+      "epoch": 1.9687034277198212,
+      "grad_norm": 2.420544147491455,
+      "learning_rate": 0.0002,
+      "loss": 2.4733,
+      "step": 26420
+    },
+    {
+      "epoch": 1.9694485842026825,
+      "grad_norm": 2.3765010833740234,
+      "learning_rate": 0.0002,
+      "loss": 2.519,
+      "step": 26430
+    },
+    {
+      "epoch": 1.970193740685544,
+      "grad_norm": 2.2879652976989746,
+      "learning_rate": 0.0002,
+      "loss": 2.5747,
+      "step": 26440
+    },
+    {
+      "epoch": 1.9709388971684052,
+      "grad_norm": 2.4816901683807373,
+      "learning_rate": 0.0002,
+      "loss": 2.566,
+      "step": 26450
+    },
+    {
+      "epoch": 1.9716840536512668,
+      "grad_norm": 2.6422929763793945,
+      "learning_rate": 0.0002,
+      "loss": 2.4609,
+      "step": 26460
+    },
+    {
+      "epoch": 1.9724292101341283,
+      "grad_norm": 3.031636953353882,
+      "learning_rate": 0.0002,
+      "loss": 2.4401,
+      "step": 26470
+    },
+    {
+      "epoch": 1.9731743666169894,
+      "grad_norm": 1.9829246997833252,
+      "learning_rate": 0.0002,
+      "loss": 2.4296,
+      "step": 26480
+    },
+    {
+      "epoch": 1.973919523099851,
+      "grad_norm": 2.687513828277588,
+      "learning_rate": 0.0002,
+      "loss": 2.5862,
+      "step": 26490
+    },
+    {
+      "epoch": 1.9746646795827125,
+      "grad_norm": 2.684131145477295,
+      "learning_rate": 0.0002,
+      "loss": 2.6292,
+      "step": 26500
+    },
+    {
+      "epoch": 1.9754098360655736,
+      "grad_norm": 2.090724468231201,
+      "learning_rate": 0.0002,
+      "loss": 2.3809,
+      "step": 26510
+    },
+    {
+      "epoch": 1.9761549925484352,
+      "grad_norm": 2.503957748413086,
+      "learning_rate": 0.0002,
+      "loss": 2.662,
+      "step": 26520
+    },
+    {
+      "epoch": 1.9769001490312967,
+      "grad_norm": 2.3615572452545166,
+      "learning_rate": 0.0002,
+      "loss": 2.5989,
+      "step": 26530
+    },
+    {
+      "epoch": 1.9776453055141578,
+      "grad_norm": 2.630366086959839,
+      "learning_rate": 0.0002,
+      "loss": 2.5245,
+      "step": 26540
+    },
+    {
+      "epoch": 1.9783904619970194,
+      "grad_norm": 2.7511889934539795,
+      "learning_rate": 0.0002,
+      "loss": 2.4983,
+      "step": 26550
+    },
+    {
+      "epoch": 1.9791356184798807,
+      "grad_norm": 2.7013278007507324,
+      "learning_rate": 0.0002,
+      "loss": 2.6212,
+      "step": 26560
+    },
+    {
+      "epoch": 1.979880774962742,
+      "grad_norm": 2.5615251064300537,
+      "learning_rate": 0.0002,
+      "loss": 2.5408,
+      "step": 26570
+    },
+    {
+      "epoch": 1.9806259314456036,
+      "grad_norm": 2.737309217453003,
+      "learning_rate": 0.0002,
+      "loss": 2.5117,
+      "step": 26580
+    },
+    {
+      "epoch": 1.981371087928465,
+      "grad_norm": 2.0868828296661377,
+      "learning_rate": 0.0002,
+      "loss": 2.3501,
+      "step": 26590
+    },
+    {
+      "epoch": 1.9821162444113263,
+      "grad_norm": 2.7384564876556396,
+      "learning_rate": 0.0002,
+      "loss": 2.4741,
+      "step": 26600
+    },
+    {
+      "epoch": 1.9828614008941878,
+      "grad_norm": 2.5381107330322266,
+      "learning_rate": 0.0002,
+      "loss": 2.5183,
+      "step": 26610
+    },
+    {
+      "epoch": 1.9836065573770492,
+      "grad_norm": 2.441673994064331,
+      "learning_rate": 0.0002,
+      "loss": 2.6689,
+      "step": 26620
+    },
+    {
+      "epoch": 1.9843517138599105,
+      "grad_norm": 2.331639528274536,
+      "learning_rate": 0.0002,
+      "loss": 2.4582,
+      "step": 26630
+    },
+    {
+      "epoch": 1.985096870342772,
+      "grad_norm": 2.777003288269043,
+      "learning_rate": 0.0002,
+      "loss": 2.4258,
+      "step": 26640
+    },
+    {
+      "epoch": 1.9858420268256334,
+      "grad_norm": 2.382567882537842,
+      "learning_rate": 0.0002,
+      "loss": 2.4606,
+      "step": 26650
+    },
+    {
+      "epoch": 1.9865871833084947,
+      "grad_norm": 2.5854926109313965,
+      "learning_rate": 0.0002,
+      "loss": 2.558,
+      "step": 26660
+    },
+    {
+      "epoch": 1.9873323397913563,
+      "grad_norm": 2.8911728858947754,
+      "learning_rate": 0.0002,
+      "loss": 2.6694,
+      "step": 26670
+    },
+    {
+      "epoch": 1.9880774962742176,
+      "grad_norm": 2.597140073776245,
+      "learning_rate": 0.0002,
+      "loss": 2.5155,
+      "step": 26680
+    },
+    {
+      "epoch": 1.988822652757079,
+      "grad_norm": 2.733752489089966,
+      "learning_rate": 0.0002,
+      "loss": 2.7112,
+      "step": 26690
+    },
+    {
+      "epoch": 1.9895678092399405,
+      "grad_norm": 2.311286687850952,
+      "learning_rate": 0.0002,
+      "loss": 2.3415,
+      "step": 26700
+    },
+    {
+      "epoch": 1.9903129657228018,
+      "grad_norm": 2.7273147106170654,
+      "learning_rate": 0.0002,
+      "loss": 2.6524,
+      "step": 26710
+    },
+    {
+      "epoch": 1.9910581222056631,
+      "grad_norm": 2.367246389389038,
+      "learning_rate": 0.0002,
+      "loss": 2.5106,
+      "step": 26720
+    },
+    {
+      "epoch": 1.9918032786885247,
+      "grad_norm": 2.5700063705444336,
+      "learning_rate": 0.0002,
+      "loss": 2.5936,
+      "step": 26730
+    },
+    {
+      "epoch": 1.992548435171386,
+      "grad_norm": 2.5638208389282227,
+      "learning_rate": 0.0002,
+      "loss": 2.6146,
+      "step": 26740
+    },
+    {
+      "epoch": 1.9932935916542474,
+      "grad_norm": 1.9451677799224854,
+      "learning_rate": 0.0002,
+      "loss": 2.4673,
+      "step": 26750
+    },
+    {
+      "epoch": 1.994038748137109,
+      "grad_norm": 2.4559242725372314,
+      "learning_rate": 0.0002,
+      "loss": 2.5946,
+      "step": 26760
+    },
+    {
+      "epoch": 1.9947839046199702,
+      "grad_norm": 2.445701837539673,
+      "learning_rate": 0.0002,
+      "loss": 2.7494,
+      "step": 26770
+    },
+    {
+      "epoch": 1.9955290611028316,
+      "grad_norm": 2.668653964996338,
+      "learning_rate": 0.0002,
+      "loss": 2.3548,
+      "step": 26780
+    },
+    {
+      "epoch": 1.9962742175856931,
+      "grad_norm": 2.523761749267578,
+      "learning_rate": 0.0002,
+      "loss": 2.4606,
+      "step": 26790
+    },
+    {
+      "epoch": 1.9970193740685542,
+      "grad_norm": 2.4471657276153564,
+      "learning_rate": 0.0002,
+      "loss": 2.4458,
+      "step": 26800
+    },
+    {
+      "epoch": 1.9977645305514158,
+      "grad_norm": 2.2107632160186768,
+      "learning_rate": 0.0002,
+      "loss": 2.5186,
+      "step": 26810
+    },
+    {
+      "epoch": 1.9985096870342773,
+      "grad_norm": 2.4634511470794678,
+      "learning_rate": 0.0002,
+      "loss": 2.5912,
+      "step": 26820
+    },
+    {
+      "epoch": 1.9992548435171384,
+      "grad_norm": 2.3914170265197754,
+      "learning_rate": 0.0002,
+      "loss": 2.4112,
+      "step": 26830
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.793278455734253,
+      "learning_rate": 0.0002,
+      "loss": 2.4186,
+      "step": 26840
+    },
+    {
+      "epoch": 2.0,
+      "eval_runtime": 2778.8528,
+      "eval_samples_per_second": 4.829,
+      "eval_steps_per_second": 0.604,
+      "step": 26840
+    },
+    {
+      "epoch": 2.0007451564828616,
+      "grad_norm": 2.427016258239746,
+      "learning_rate": 0.0002,
+      "loss": 2.2428,
+      "step": 26850
+    },
+    {
+      "epoch": 2.0014903129657227,
+      "grad_norm": 2.445763111114502,
+      "learning_rate": 0.0002,
+      "loss": 2.4872,
+      "step": 26860
+    },
+    {
+      "epoch": 2.002235469448584,
+      "grad_norm": 2.4963574409484863,
+      "learning_rate": 0.0002,
+      "loss": 2.4217,
+      "step": 26870
+    },
+    {
+      "epoch": 2.0029806259314458,
+      "grad_norm": 3.122708320617676,
+      "learning_rate": 0.0002,
+      "loss": 2.482,
+      "step": 26880
+    },
+    {
+      "epoch": 2.003725782414307,
+      "grad_norm": 2.769930124282837,
+      "learning_rate": 0.0002,
+      "loss": 2.3721,
+      "step": 26890
+    },
+    {
+      "epoch": 2.0044709388971684,
+      "grad_norm": 2.6897122859954834,
+      "learning_rate": 0.0002,
+      "loss": 2.2882,
+      "step": 26900
+    },
+    {
+      "epoch": 2.00521609538003,
+      "grad_norm": 2.7737324237823486,
+      "learning_rate": 0.0002,
+      "loss": 2.4593,
+      "step": 26910
+    },
+    {
+      "epoch": 2.005961251862891,
+      "grad_norm": 2.649350643157959,
+      "learning_rate": 0.0002,
+      "loss": 2.5489,
+      "step": 26920
+    },
+    {
+      "epoch": 2.0067064083457526,
+      "grad_norm": 2.721299886703491,
+      "learning_rate": 0.0002,
+      "loss": 2.4069,
+      "step": 26930
+    },
+    {
+      "epoch": 2.007451564828614,
+      "grad_norm": 2.5351850986480713,
+      "learning_rate": 0.0002,
+      "loss": 2.3394,
+      "step": 26940
+    },
+    {
+      "epoch": 2.0081967213114753,
+      "grad_norm": 2.717834949493408,
+      "learning_rate": 0.0002,
+      "loss": 2.505,
+      "step": 26950
+    },
+    {
+      "epoch": 2.008941877794337,
+      "grad_norm": 2.9618914127349854,
+      "learning_rate": 0.0002,
+      "loss": 2.4794,
+      "step": 26960
+    },
+    {
+      "epoch": 2.0096870342771984,
+      "grad_norm": 2.4195728302001953,
+      "learning_rate": 0.0002,
+      "loss": 2.3318,
+      "step": 26970
+    },
+    {
+      "epoch": 2.0104321907600595,
+      "grad_norm": 3.099935531616211,
+      "learning_rate": 0.0002,
+      "loss": 2.2687,
+      "step": 26980
+    },
+    {
+      "epoch": 2.011177347242921,
+      "grad_norm": 2.5944323539733887,
+      "learning_rate": 0.0002,
+      "loss": 2.5544,
+      "step": 26990
+    },
+    {
+      "epoch": 2.0119225037257826,
+      "grad_norm": 2.521864175796509,
+      "learning_rate": 0.0002,
+      "loss": 2.4157,
+      "step": 27000
+    },
+    {
+      "epoch": 2.0126676602086437,
+      "grad_norm": 2.8137166500091553,
+      "learning_rate": 0.0002,
+      "loss": 2.2001,
+      "step": 27010
+    },
+    {
+      "epoch": 2.0134128166915053,
+      "grad_norm": 2.695870876312256,
+      "learning_rate": 0.0002,
+      "loss": 2.3827,
+      "step": 27020
+    },
+    {
+      "epoch": 2.0141579731743664,
+      "grad_norm": 2.4329819679260254,
+      "learning_rate": 0.0002,
+      "loss": 2.2124,
+      "step": 27030
+    },
+    {
+      "epoch": 2.014903129657228,
+      "grad_norm": 2.539754629135132,
+      "learning_rate": 0.0002,
+      "loss": 2.3536,
+      "step": 27040
+    },
+    {
+      "epoch": 2.0156482861400895,
+      "grad_norm": 2.557807207107544,
+      "learning_rate": 0.0002,
+      "loss": 2.3424,
+      "step": 27050
+    },
+    {
+      "epoch": 2.0163934426229506,
+      "grad_norm": 3.0624873638153076,
+      "learning_rate": 0.0002,
+      "loss": 2.5087,
+      "step": 27060
+    },
+    {
+      "epoch": 2.017138599105812,
+      "grad_norm": 2.791489601135254,
+      "learning_rate": 0.0002,
+      "loss": 2.515,
+      "step": 27070
+    },
+    {
+      "epoch": 2.0178837555886737,
+      "grad_norm": 2.6769886016845703,
+      "learning_rate": 0.0002,
+      "loss": 2.5366,
+      "step": 27080
+    },
+    {
+      "epoch": 2.018628912071535,
+      "grad_norm": 2.861119270324707,
+      "learning_rate": 0.0002,
+      "loss": 2.5285,
+      "step": 27090
+    },
+    {
+      "epoch": 2.0193740685543964,
+      "grad_norm": 2.2920444011688232,
+      "learning_rate": 0.0002,
+      "loss": 2.3931,
+      "step": 27100
+    },
+    {
+      "epoch": 2.020119225037258,
+      "grad_norm": 2.878737211227417,
+      "learning_rate": 0.0002,
+      "loss": 2.2288,
+      "step": 27110
+    },
+    {
+      "epoch": 2.020864381520119,
+      "grad_norm": 2.782578706741333,
+      "learning_rate": 0.0002,
+      "loss": 2.2606,
+      "step": 27120
+    },
+    {
+      "epoch": 2.0216095380029806,
+      "grad_norm": 2.6214089393615723,
+      "learning_rate": 0.0002,
+      "loss": 2.4091,
+      "step": 27130
+    },
+    {
+      "epoch": 2.022354694485842,
+      "grad_norm": 2.628899335861206,
+      "learning_rate": 0.0002,
+      "loss": 2.3615,
+      "step": 27140
+    },
+    {
+      "epoch": 2.0230998509687033,
+      "grad_norm": 2.4732677936553955,
+      "learning_rate": 0.0002,
+      "loss": 2.2547,
+      "step": 27150
+    },
+    {
+      "epoch": 2.023845007451565,
+      "grad_norm": 2.5701940059661865,
+      "learning_rate": 0.0002,
+      "loss": 2.3523,
+      "step": 27160
+    },
+    {
+      "epoch": 2.0245901639344264,
+      "grad_norm": 2.6380484104156494,
+      "learning_rate": 0.0002,
+      "loss": 2.2513,
+      "step": 27170
+    },
+    {
+      "epoch": 2.0253353204172875,
+      "grad_norm": 2.5996925830841064,
+      "learning_rate": 0.0002,
+      "loss": 2.3843,
+      "step": 27180
+    },
+    {
+      "epoch": 2.026080476900149,
+      "grad_norm": 2.727452516555786,
+      "learning_rate": 0.0002,
+      "loss": 2.2922,
+      "step": 27190
+    },
+    {
+      "epoch": 2.0268256333830106,
+      "grad_norm": 2.9830238819122314,
+      "learning_rate": 0.0002,
+      "loss": 2.313,
+      "step": 27200
+    },
+    {
+      "epoch": 2.0275707898658717,
+      "grad_norm": 2.7571492195129395,
+      "learning_rate": 0.0002,
+      "loss": 2.3851,
+      "step": 27210
+    },
+    {
+      "epoch": 2.0283159463487332,
+      "grad_norm": 2.3931894302368164,
+      "learning_rate": 0.0002,
+      "loss": 2.0832,
+      "step": 27220
+    },
+    {
+      "epoch": 2.029061102831595,
+      "grad_norm": 2.732896566390991,
+      "learning_rate": 0.0002,
+      "loss": 2.4687,
+      "step": 27230
+    },
+    {
+      "epoch": 2.029806259314456,
+      "grad_norm": 2.637854814529419,
+      "learning_rate": 0.0002,
+      "loss": 2.487,
+      "step": 27240
+    },
+    {
+      "epoch": 2.0305514157973175,
+      "grad_norm": 2.2118914127349854,
+      "learning_rate": 0.0002,
+      "loss": 2.3466,
+      "step": 27250
+    },
+    {
+      "epoch": 2.031296572280179,
+      "grad_norm": 2.7009260654449463,
+      "learning_rate": 0.0002,
+      "loss": 2.3756,
+      "step": 27260
+    },
+    {
+      "epoch": 2.03204172876304,
+      "grad_norm": 2.3376128673553467,
+      "learning_rate": 0.0002,
+      "loss": 2.4311,
+      "step": 27270
+    },
+    {
+      "epoch": 2.0327868852459017,
+      "grad_norm": 2.7543601989746094,
+      "learning_rate": 0.0002,
+      "loss": 2.2645,
+      "step": 27280
+    },
+    {
+      "epoch": 2.0335320417287632,
+      "grad_norm": 2.2563414573669434,
+      "learning_rate": 0.0002,
+      "loss": 2.3356,
+      "step": 27290
+    },
+    {
+      "epoch": 2.0342771982116243,
+      "grad_norm": 2.4909472465515137,
+      "learning_rate": 0.0002,
+      "loss": 2.4706,
+      "step": 27300
+    },
+    {
+      "epoch": 2.035022354694486,
+      "grad_norm": 2.689833164215088,
+      "learning_rate": 0.0002,
+      "loss": 2.5033,
+      "step": 27310
+    },
+    {
+      "epoch": 2.0357675111773474,
+      "grad_norm": 2.7386205196380615,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 27320
+    },
+    {
+      "epoch": 2.0365126676602086,
+      "grad_norm": 2.537919759750366,
+      "learning_rate": 0.0002,
+      "loss": 2.4454,
+      "step": 27330
+    },
+    {
+      "epoch": 2.03725782414307,
+      "grad_norm": 2.522033214569092,
+      "learning_rate": 0.0002,
+      "loss": 2.2687,
+      "step": 27340
+    },
+    {
+      "epoch": 2.0380029806259317,
+      "grad_norm": 2.6754488945007324,
+      "learning_rate": 0.0002,
+      "loss": 2.2479,
+      "step": 27350
+    },
+    {
+      "epoch": 2.0387481371087928,
+      "grad_norm": 2.134866952896118,
+      "learning_rate": 0.0002,
+      "loss": 2.3477,
+      "step": 27360
+    },
+    {
+      "epoch": 2.0394932935916543,
+      "grad_norm": 2.3635213375091553,
+      "learning_rate": 0.0002,
+      "loss": 2.4031,
+      "step": 27370
+    },
+    {
+      "epoch": 2.0402384500745154,
+      "grad_norm": 2.4664597511291504,
+      "learning_rate": 0.0002,
+      "loss": 2.5241,
+      "step": 27380
+    },
+    {
+      "epoch": 2.040983606557377,
+      "grad_norm": 2.7248127460479736,
+      "learning_rate": 0.0002,
+      "loss": 2.4092,
+      "step": 27390
+    },
+    {
+      "epoch": 2.0417287630402385,
+      "grad_norm": 2.75724458694458,
+      "learning_rate": 0.0002,
+      "loss": 2.4693,
+      "step": 27400
+    },
+    {
+      "epoch": 2.0424739195230996,
+      "grad_norm": 2.580315113067627,
+      "learning_rate": 0.0002,
+      "loss": 2.4291,
+      "step": 27410
+    },
+    {
+      "epoch": 2.043219076005961,
+      "grad_norm": 3.07570743560791,
+      "learning_rate": 0.0002,
+      "loss": 2.4226,
+      "step": 27420
+    },
+    {
+      "epoch": 2.0439642324888228,
+      "grad_norm": 2.4915974140167236,
+      "learning_rate": 0.0002,
+      "loss": 2.4481,
+      "step": 27430
+    },
+    {
+      "epoch": 2.044709388971684,
+      "grad_norm": 2.7171037197113037,
+      "learning_rate": 0.0002,
+      "loss": 2.2091,
+      "step": 27440
+    },
+    {
+      "epoch": 2.0454545454545454,
+      "grad_norm": 3.0568337440490723,
+      "learning_rate": 0.0002,
+      "loss": 2.3815,
+      "step": 27450
+    },
+    {
+      "epoch": 2.046199701937407,
+      "grad_norm": 2.659975051879883,
+      "learning_rate": 0.0002,
+      "loss": 2.4473,
+      "step": 27460
+    },
+    {
+      "epoch": 2.046944858420268,
+      "grad_norm": 2.7344188690185547,
+      "learning_rate": 0.0002,
+      "loss": 2.429,
+      "step": 27470
+    },
+    {
+      "epoch": 2.0476900149031296,
+      "grad_norm": 2.360349416732788,
+      "learning_rate": 0.0002,
+      "loss": 2.2687,
+      "step": 27480
+    },
+    {
+      "epoch": 2.048435171385991,
+      "grad_norm": 2.6498806476593018,
+      "learning_rate": 0.0002,
+      "loss": 2.4305,
+      "step": 27490
+    },
+    {
+      "epoch": 2.0491803278688523,
+      "grad_norm": 2.6238315105438232,
+      "learning_rate": 0.0002,
+      "loss": 2.3188,
+      "step": 27500
+    },
+    {
+      "epoch": 2.049925484351714,
+      "grad_norm": 2.6448466777801514,
+      "learning_rate": 0.0002,
+      "loss": 2.286,
+      "step": 27510
+    },
+    {
+      "epoch": 2.0506706408345754,
+      "grad_norm": 2.3168444633483887,
+      "learning_rate": 0.0002,
+      "loss": 2.2408,
+      "step": 27520
+    },
+    {
+      "epoch": 2.0514157973174365,
+      "grad_norm": 2.7542989253997803,
+      "learning_rate": 0.0002,
+      "loss": 2.2652,
+      "step": 27530
+    },
+    {
+      "epoch": 2.052160953800298,
+      "grad_norm": 2.794304132461548,
+      "learning_rate": 0.0002,
+      "loss": 2.3382,
+      "step": 27540
+    },
+    {
+      "epoch": 2.0529061102831596,
+      "grad_norm": 2.4550607204437256,
+      "learning_rate": 0.0002,
+      "loss": 2.5653,
+      "step": 27550
+    },
+    {
+      "epoch": 2.0536512667660207,
+      "grad_norm": 2.7925267219543457,
+      "learning_rate": 0.0002,
+      "loss": 2.3883,
+      "step": 27560
+    },
+    {
+      "epoch": 2.0543964232488823,
+      "grad_norm": 2.732017993927002,
+      "learning_rate": 0.0002,
+      "loss": 2.5619,
+      "step": 27570
+    },
+    {
+      "epoch": 2.055141579731744,
+      "grad_norm": 2.7166635990142822,
+      "learning_rate": 0.0002,
+      "loss": 2.3683,
+      "step": 27580
+    },
+    {
+      "epoch": 2.055886736214605,
+      "grad_norm": 2.5912301540374756,
+      "learning_rate": 0.0002,
+      "loss": 2.3318,
+      "step": 27590
+    },
+    {
+      "epoch": 2.0566318926974665,
+      "grad_norm": 2.5270578861236572,
+      "learning_rate": 0.0002,
+      "loss": 2.2477,
+      "step": 27600
+    },
+    {
+      "epoch": 2.057377049180328,
+      "grad_norm": 2.8963217735290527,
+      "learning_rate": 0.0002,
+      "loss": 2.3905,
+      "step": 27610
+    },
+    {
+      "epoch": 2.058122205663189,
+      "grad_norm": 2.306488513946533,
+      "learning_rate": 0.0002,
+      "loss": 2.3994,
+      "step": 27620
+    },
+    {
+      "epoch": 2.0588673621460507,
+      "grad_norm": 2.74562406539917,
+      "learning_rate": 0.0002,
+      "loss": 2.2255,
+      "step": 27630
+    },
+    {
+      "epoch": 2.0596125186289123,
+      "grad_norm": 2.907052755355835,
+      "learning_rate": 0.0002,
+      "loss": 2.4501,
+      "step": 27640
+    },
+    {
+      "epoch": 2.0603576751117734,
+      "grad_norm": 2.6865978240966797,
+      "learning_rate": 0.0002,
+      "loss": 2.4872,
+      "step": 27650
+    },
+    {
+      "epoch": 2.061102831594635,
+      "grad_norm": 2.829334259033203,
+      "learning_rate": 0.0002,
+      "loss": 2.3885,
+      "step": 27660
+    },
+    {
+      "epoch": 2.0618479880774965,
+      "grad_norm": 2.276557445526123,
+      "learning_rate": 0.0002,
+      "loss": 2.2833,
+      "step": 27670
+    },
+    {
+      "epoch": 2.0625931445603576,
+      "grad_norm": 2.7757015228271484,
+      "learning_rate": 0.0002,
+      "loss": 2.4336,
+      "step": 27680
+    },
+    {
+      "epoch": 2.063338301043219,
+      "grad_norm": 2.630664587020874,
+      "learning_rate": 0.0002,
+      "loss": 2.466,
+      "step": 27690
+    },
+    {
+      "epoch": 2.0640834575260807,
+      "grad_norm": 2.500433921813965,
+      "learning_rate": 0.0002,
+      "loss": 2.4625,
+      "step": 27700
+    },
+    {
+      "epoch": 2.064828614008942,
+      "grad_norm": 2.5673367977142334,
+      "learning_rate": 0.0002,
+      "loss": 2.3708,
+      "step": 27710
+    },
+    {
+      "epoch": 2.0655737704918034,
+      "grad_norm": 2.798462390899658,
+      "learning_rate": 0.0002,
+      "loss": 2.5025,
+      "step": 27720
+    },
+    {
+      "epoch": 2.066318926974665,
+      "grad_norm": 2.660876989364624,
+      "learning_rate": 0.0002,
+      "loss": 2.3465,
+      "step": 27730
+    },
+    {
+      "epoch": 2.067064083457526,
+      "grad_norm": 2.681070566177368,
+      "learning_rate": 0.0002,
+      "loss": 2.4037,
+      "step": 27740
+    },
+    {
+      "epoch": 2.0678092399403876,
+      "grad_norm": 2.4489269256591797,
+      "learning_rate": 0.0002,
+      "loss": 2.3239,
+      "step": 27750
+    },
+    {
+      "epoch": 2.0685543964232487,
+      "grad_norm": 2.980785608291626,
+      "learning_rate": 0.0002,
+      "loss": 2.4702,
+      "step": 27760
+    },
+    {
+      "epoch": 2.0692995529061102,
+      "grad_norm": 2.7260890007019043,
+      "learning_rate": 0.0002,
+      "loss": 2.3616,
+      "step": 27770
+    },
+    {
+      "epoch": 2.070044709388972,
+      "grad_norm": 2.714730739593506,
+      "learning_rate": 0.0002,
+      "loss": 2.3833,
+      "step": 27780
+    },
+    {
+      "epoch": 2.070789865871833,
+      "grad_norm": 2.3693652153015137,
+      "learning_rate": 0.0002,
+      "loss": 2.2471,
+      "step": 27790
+    },
+    {
+      "epoch": 2.0715350223546944,
+      "grad_norm": 2.6847126483917236,
+      "learning_rate": 0.0002,
+      "loss": 2.5155,
+      "step": 27800
+    },
+    {
+      "epoch": 2.072280178837556,
+      "grad_norm": 3.285604476928711,
+      "learning_rate": 0.0002,
+      "loss": 2.5073,
+      "step": 27810
+    },
+    {
+      "epoch": 2.073025335320417,
+      "grad_norm": 2.9383273124694824,
+      "learning_rate": 0.0002,
+      "loss": 2.401,
+      "step": 27820
+    },
+    {
+      "epoch": 2.0737704918032787,
+      "grad_norm": 2.5305216312408447,
+      "learning_rate": 0.0002,
+      "loss": 2.49,
+      "step": 27830
+    },
+    {
+      "epoch": 2.07451564828614,
+      "grad_norm": 3.088163375854492,
+      "learning_rate": 0.0002,
+      "loss": 2.3074,
+      "step": 27840
+    },
+    {
+      "epoch": 2.0752608047690013,
+      "grad_norm": 2.9205684661865234,
+      "learning_rate": 0.0002,
+      "loss": 2.3076,
+      "step": 27850
+    },
+    {
+      "epoch": 2.076005961251863,
+      "grad_norm": 2.644702196121216,
+      "learning_rate": 0.0002,
+      "loss": 2.3757,
+      "step": 27860
+    },
+    {
+      "epoch": 2.0767511177347244,
+      "grad_norm": 2.7440295219421387,
+      "learning_rate": 0.0002,
+      "loss": 2.4924,
+      "step": 27870
+    },
+    {
+      "epoch": 2.0774962742175855,
+      "grad_norm": 2.949817657470703,
+      "learning_rate": 0.0002,
+      "loss": 2.343,
+      "step": 27880
+    },
+    {
+      "epoch": 2.078241430700447,
+      "grad_norm": 2.5574519634246826,
+      "learning_rate": 0.0002,
+      "loss": 2.4203,
+      "step": 27890
+    },
+    {
+      "epoch": 2.0789865871833086,
+      "grad_norm": 2.6132795810699463,
+      "learning_rate": 0.0002,
+      "loss": 2.5225,
+      "step": 27900
+    },
+    {
+      "epoch": 2.0797317436661698,
+      "grad_norm": 2.1226987838745117,
+      "learning_rate": 0.0002,
+      "loss": 2.1986,
+      "step": 27910
+    },
+    {
+      "epoch": 2.0804769001490313,
+      "grad_norm": 2.5743627548217773,
+      "learning_rate": 0.0002,
+      "loss": 2.3704,
+      "step": 27920
+    },
+    {
+      "epoch": 2.081222056631893,
+      "grad_norm": 2.4740493297576904,
+      "learning_rate": 0.0002,
+      "loss": 2.2522,
+      "step": 27930
+    },
+    {
+      "epoch": 2.081967213114754,
+      "grad_norm": 2.5864386558532715,
+      "learning_rate": 0.0002,
+      "loss": 2.4904,
+      "step": 27940
+    },
+    {
+      "epoch": 2.0827123695976155,
+      "grad_norm": 2.6593196392059326,
+      "learning_rate": 0.0002,
+      "loss": 2.2939,
+      "step": 27950
+    },
+    {
+      "epoch": 2.083457526080477,
+      "grad_norm": 2.5160462856292725,
+      "learning_rate": 0.0002,
+      "loss": 2.5326,
+      "step": 27960
+    },
+    {
+      "epoch": 2.084202682563338,
+      "grad_norm": 2.8797519207000732,
+      "learning_rate": 0.0002,
+      "loss": 2.4039,
+      "step": 27970
+    },
+    {
+      "epoch": 2.0849478390461997,
+      "grad_norm": 2.435398578643799,
+      "learning_rate": 0.0002,
+      "loss": 2.4237,
+      "step": 27980
+    },
+    {
+      "epoch": 2.0856929955290613,
+      "grad_norm": 2.6593923568725586,
+      "learning_rate": 0.0002,
+      "loss": 2.5967,
+      "step": 27990
+    },
+    {
+      "epoch": 2.0864381520119224,
+      "grad_norm": 2.205998182296753,
+      "learning_rate": 0.0002,
+      "loss": 2.2983,
+      "step": 28000
+    },
+    {
+      "epoch": 2.087183308494784,
+      "grad_norm": 2.313530445098877,
+      "learning_rate": 0.0002,
+      "loss": 2.291,
+      "step": 28010
+    },
+    {
+      "epoch": 2.0879284649776455,
+      "grad_norm": 2.9315454959869385,
+      "learning_rate": 0.0002,
+      "loss": 2.311,
+      "step": 28020
+    },
+    {
+      "epoch": 2.0886736214605066,
+      "grad_norm": 2.7529959678649902,
+      "learning_rate": 0.0002,
+      "loss": 2.2777,
+      "step": 28030
+    },
+    {
+      "epoch": 2.089418777943368,
+      "grad_norm": 2.668069362640381,
+      "learning_rate": 0.0002,
+      "loss": 2.6066,
+      "step": 28040
+    },
+    {
+      "epoch": 2.0901639344262297,
+      "grad_norm": 2.849188804626465,
+      "learning_rate": 0.0002,
+      "loss": 2.535,
+      "step": 28050
+    },
+    {
+      "epoch": 2.090909090909091,
+      "grad_norm": 2.613250970840454,
+      "learning_rate": 0.0002,
+      "loss": 2.2796,
+      "step": 28060
+    },
+    {
+      "epoch": 2.0916542473919524,
+      "grad_norm": 2.3394582271575928,
+      "learning_rate": 0.0002,
+      "loss": 2.4765,
+      "step": 28070
+    },
+    {
+      "epoch": 2.092399403874814,
+      "grad_norm": 2.5623340606689453,
+      "learning_rate": 0.0002,
+      "loss": 2.4527,
+      "step": 28080
+    },
+    {
+      "epoch": 2.093144560357675,
+      "grad_norm": 2.4075920581817627,
+      "learning_rate": 0.0002,
+      "loss": 2.2952,
+      "step": 28090
+    },
+    {
+      "epoch": 2.0938897168405366,
+      "grad_norm": 2.763248920440674,
+      "learning_rate": 0.0002,
+      "loss": 2.2659,
+      "step": 28100
+    },
+    {
+      "epoch": 2.0946348733233977,
+      "grad_norm": 2.7755019664764404,
+      "learning_rate": 0.0002,
+      "loss": 2.622,
+      "step": 28110
+    },
+    {
+      "epoch": 2.0953800298062593,
+      "grad_norm": 2.6472527980804443,
+      "learning_rate": 0.0002,
+      "loss": 2.4746,
+      "step": 28120
+    },
+    {
+      "epoch": 2.096125186289121,
+      "grad_norm": 3.476677417755127,
+      "learning_rate": 0.0002,
+      "loss": 2.3413,
+      "step": 28130
+    },
+    {
+      "epoch": 2.096870342771982,
+      "grad_norm": 2.5230743885040283,
+      "learning_rate": 0.0002,
+      "loss": 2.3565,
+      "step": 28140
+    },
+    {
+      "epoch": 2.0976154992548435,
+      "grad_norm": 2.7483949661254883,
+      "learning_rate": 0.0002,
+      "loss": 2.3682,
+      "step": 28150
+    },
+    {
+      "epoch": 2.098360655737705,
+      "grad_norm": 2.6513829231262207,
+      "learning_rate": 0.0002,
+      "loss": 2.2721,
+      "step": 28160
+    },
+    {
+      "epoch": 2.099105812220566,
+      "grad_norm": 2.6120636463165283,
+      "learning_rate": 0.0002,
+      "loss": 2.6904,
+      "step": 28170
+    },
+    {
+      "epoch": 2.0998509687034277,
+      "grad_norm": 2.4628775119781494,
+      "learning_rate": 0.0002,
+      "loss": 2.4378,
+      "step": 28180
+    },
+    {
+      "epoch": 2.1005961251862892,
+      "grad_norm": 2.5905325412750244,
+      "learning_rate": 0.0002,
+      "loss": 2.2657,
+      "step": 28190
+    },
+    {
+      "epoch": 2.1013412816691504,
+      "grad_norm": 2.9600167274475098,
+      "learning_rate": 0.0002,
+      "loss": 2.2855,
+      "step": 28200
+    },
+    {
+      "epoch": 2.102086438152012,
+      "grad_norm": 3.109748363494873,
+      "learning_rate": 0.0002,
+      "loss": 2.4388,
+      "step": 28210
+    },
+    {
+      "epoch": 2.1028315946348735,
+      "grad_norm": 2.639881134033203,
+      "learning_rate": 0.0002,
+      "loss": 2.4958,
+      "step": 28220
+    },
+    {
+      "epoch": 2.1035767511177346,
+      "grad_norm": 2.901447057723999,
+      "learning_rate": 0.0002,
+      "loss": 2.6251,
+      "step": 28230
+    },
+    {
+      "epoch": 2.104321907600596,
+      "grad_norm": 2.4745752811431885,
+      "learning_rate": 0.0002,
+      "loss": 2.5932,
+      "step": 28240
+    },
+    {
+      "epoch": 2.1050670640834577,
+      "grad_norm": 2.363933801651001,
+      "learning_rate": 0.0002,
+      "loss": 2.3875,
+      "step": 28250
+    },
+    {
+      "epoch": 2.105812220566319,
+      "grad_norm": 2.6219677925109863,
+      "learning_rate": 0.0002,
+      "loss": 2.3415,
+      "step": 28260
+    },
+    {
+      "epoch": 2.1065573770491803,
+      "grad_norm": 2.6669254302978516,
+      "learning_rate": 0.0002,
+      "loss": 2.5103,
+      "step": 28270
+    },
+    {
+      "epoch": 2.107302533532042,
+      "grad_norm": 2.7276668548583984,
+      "learning_rate": 0.0002,
+      "loss": 2.2377,
+      "step": 28280
+    },
+    {
+      "epoch": 2.108047690014903,
+      "grad_norm": 2.473850727081299,
+      "learning_rate": 0.0002,
+      "loss": 2.4347,
+      "step": 28290
+    },
+    {
+      "epoch": 2.1087928464977646,
+      "grad_norm": 2.89056396484375,
+      "learning_rate": 0.0002,
+      "loss": 2.3286,
+      "step": 28300
+    },
+    {
+      "epoch": 2.109538002980626,
+      "grad_norm": 2.4817709922790527,
+      "learning_rate": 0.0002,
+      "loss": 2.5692,
+      "step": 28310
+    },
+    {
+      "epoch": 2.110283159463487,
+      "grad_norm": 1.887121319770813,
+      "learning_rate": 0.0002,
+      "loss": 2.2894,
+      "step": 28320
+    },
+    {
+      "epoch": 2.1110283159463488,
+      "grad_norm": 2.799809694290161,
+      "learning_rate": 0.0002,
+      "loss": 2.4199,
+      "step": 28330
+    },
+    {
+      "epoch": 2.1117734724292103,
+      "grad_norm": 2.481428861618042,
+      "learning_rate": 0.0002,
+      "loss": 2.3954,
+      "step": 28340
+    },
+    {
+      "epoch": 2.1125186289120714,
+      "grad_norm": 2.573892831802368,
+      "learning_rate": 0.0002,
+      "loss": 2.4221,
+      "step": 28350
+    },
+    {
+      "epoch": 2.113263785394933,
+      "grad_norm": 2.6230032444000244,
+      "learning_rate": 0.0002,
+      "loss": 2.3498,
+      "step": 28360
+    },
+    {
+      "epoch": 2.1140089418777945,
+      "grad_norm": 2.3305563926696777,
+      "learning_rate": 0.0002,
+      "loss": 2.4446,
+      "step": 28370
+    },
+    {
+      "epoch": 2.1147540983606556,
+      "grad_norm": 2.297919750213623,
+      "learning_rate": 0.0002,
+      "loss": 2.452,
+      "step": 28380
+    },
+    {
+      "epoch": 2.115499254843517,
+      "grad_norm": 3.0434772968292236,
+      "learning_rate": 0.0002,
+      "loss": 2.3133,
+      "step": 28390
+    },
+    {
+      "epoch": 2.1162444113263787,
+      "grad_norm": 3.005995750427246,
+      "learning_rate": 0.0002,
+      "loss": 2.3543,
+      "step": 28400
+    },
+    {
+      "epoch": 2.11698956780924,
+      "grad_norm": 2.65181565284729,
+      "learning_rate": 0.0002,
+      "loss": 2.2972,
+      "step": 28410
+    },
+    {
+      "epoch": 2.1177347242921014,
+      "grad_norm": 2.5356462001800537,
+      "learning_rate": 0.0002,
+      "loss": 2.3888,
+      "step": 28420
+    },
+    {
+      "epoch": 2.118479880774963,
+      "grad_norm": 2.4722752571105957,
+      "learning_rate": 0.0002,
+      "loss": 2.4412,
+      "step": 28430
+    },
+    {
+      "epoch": 2.119225037257824,
+      "grad_norm": 2.7374267578125,
+      "learning_rate": 0.0002,
+      "loss": 2.5805,
+      "step": 28440
+    },
+    {
+      "epoch": 2.1199701937406856,
+      "grad_norm": 2.93241548538208,
+      "learning_rate": 0.0002,
+      "loss": 2.5802,
+      "step": 28450
+    },
+    {
+      "epoch": 2.1207153502235467,
+      "grad_norm": 2.6112005710601807,
+      "learning_rate": 0.0002,
+      "loss": 2.485,
+      "step": 28460
+    },
+    {
+      "epoch": 2.1214605067064083,
+      "grad_norm": 2.2507147789001465,
+      "learning_rate": 0.0002,
+      "loss": 2.3001,
+      "step": 28470
+    },
+    {
+      "epoch": 2.12220566318927,
+      "grad_norm": 2.8266193866729736,
+      "learning_rate": 0.0002,
+      "loss": 2.3678,
+      "step": 28480
+    },
+    {
+      "epoch": 2.122950819672131,
+      "grad_norm": 2.5680441856384277,
+      "learning_rate": 0.0002,
+      "loss": 2.5827,
+      "step": 28490
+    },
+    {
+      "epoch": 2.1236959761549925,
+      "grad_norm": 2.623478889465332,
+      "learning_rate": 0.0002,
+      "loss": 2.3191,
+      "step": 28500
+    },
+    {
+      "epoch": 2.124441132637854,
+      "grad_norm": 2.591749906539917,
+      "learning_rate": 0.0002,
+      "loss": 2.4871,
+      "step": 28510
+    },
+    {
+      "epoch": 2.125186289120715,
+      "grad_norm": 2.7326436042785645,
+      "learning_rate": 0.0002,
+      "loss": 2.3359,
+      "step": 28520
+    },
+    {
+      "epoch": 2.1259314456035767,
+      "grad_norm": 3.0076076984405518,
+      "learning_rate": 0.0002,
+      "loss": 2.3586,
+      "step": 28530
+    },
+    {
+      "epoch": 2.1266766020864383,
+      "grad_norm": 2.5301334857940674,
+      "learning_rate": 0.0002,
+      "loss": 2.4504,
+      "step": 28540
+    },
+    {
+      "epoch": 2.1274217585692994,
+      "grad_norm": 2.3998358249664307,
+      "learning_rate": 0.0002,
+      "loss": 2.2544,
+      "step": 28550
+    },
+    {
+      "epoch": 2.128166915052161,
+      "grad_norm": 2.7175776958465576,
+      "learning_rate": 0.0002,
+      "loss": 2.582,
+      "step": 28560
+    },
+    {
+      "epoch": 2.1289120715350225,
+      "grad_norm": 2.4228036403656006,
+      "learning_rate": 0.0002,
+      "loss": 2.3565,
+      "step": 28570
+    },
+    {
+      "epoch": 2.1296572280178836,
+      "grad_norm": 2.6200060844421387,
+      "learning_rate": 0.0002,
+      "loss": 2.1041,
+      "step": 28580
+    },
+    {
+      "epoch": 2.130402384500745,
+      "grad_norm": 2.493910312652588,
+      "learning_rate": 0.0002,
+      "loss": 2.3126,
+      "step": 28590
+    },
+    {
+      "epoch": 2.1311475409836067,
+      "grad_norm": 3.021311044692993,
+      "learning_rate": 0.0002,
+      "loss": 2.5049,
+      "step": 28600
+    },
+    {
+      "epoch": 2.131892697466468,
+      "grad_norm": 2.4209420680999756,
+      "learning_rate": 0.0002,
+      "loss": 2.3074,
+      "step": 28610
+    },
+    {
+      "epoch": 2.1326378539493294,
+      "grad_norm": 2.2820701599121094,
+      "learning_rate": 0.0002,
+      "loss": 2.3154,
+      "step": 28620
+    },
+    {
+      "epoch": 2.133383010432191,
+      "grad_norm": 2.4041130542755127,
+      "learning_rate": 0.0002,
+      "loss": 2.4206,
+      "step": 28630
+    },
+    {
+      "epoch": 2.134128166915052,
+      "grad_norm": 2.7592175006866455,
+      "learning_rate": 0.0002,
+      "loss": 2.4367,
+      "step": 28640
+    },
+    {
+      "epoch": 2.1348733233979136,
+      "grad_norm": 2.982658863067627,
+      "learning_rate": 0.0002,
+      "loss": 2.4851,
+      "step": 28650
+    },
+    {
+      "epoch": 2.135618479880775,
+      "grad_norm": 2.370884656906128,
+      "learning_rate": 0.0002,
+      "loss": 2.3921,
+      "step": 28660
+    },
+    {
+      "epoch": 2.1363636363636362,
+      "grad_norm": 2.121638298034668,
+      "learning_rate": 0.0002,
+      "loss": 2.4153,
+      "step": 28670
+    },
+    {
+      "epoch": 2.137108792846498,
+      "grad_norm": 2.6304523944854736,
+      "learning_rate": 0.0002,
+      "loss": 2.5541,
+      "step": 28680
+    },
+    {
+      "epoch": 2.1378539493293593,
+      "grad_norm": 2.711233139038086,
+      "learning_rate": 0.0002,
+      "loss": 2.3325,
+      "step": 28690
+    },
+    {
+      "epoch": 2.1385991058122205,
+      "grad_norm": 2.832350969314575,
+      "learning_rate": 0.0002,
+      "loss": 2.4829,
+      "step": 28700
+    },
+    {
+      "epoch": 2.139344262295082,
+      "grad_norm": 2.737445831298828,
+      "learning_rate": 0.0002,
+      "loss": 2.5671,
+      "step": 28710
+    },
+    {
+      "epoch": 2.1400894187779436,
+      "grad_norm": 2.456326484680176,
+      "learning_rate": 0.0002,
+      "loss": 2.4368,
+      "step": 28720
+    },
+    {
+      "epoch": 2.1408345752608047,
+      "grad_norm": 2.3802084922790527,
+      "learning_rate": 0.0002,
+      "loss": 2.5299,
+      "step": 28730
+    },
+    {
+      "epoch": 2.1415797317436662,
+      "grad_norm": 2.521446943283081,
+      "learning_rate": 0.0002,
+      "loss": 2.51,
+      "step": 28740
+    },
+    {
+      "epoch": 2.1423248882265273,
+      "grad_norm": 2.6229171752929688,
+      "learning_rate": 0.0002,
+      "loss": 2.4292,
+      "step": 28750
+    },
+    {
+      "epoch": 2.143070044709389,
+      "grad_norm": 2.573530435562134,
+      "learning_rate": 0.0002,
+      "loss": 2.5116,
+      "step": 28760
+    },
+    {
+      "epoch": 2.1438152011922504,
+      "grad_norm": 2.3284008502960205,
+      "learning_rate": 0.0002,
+      "loss": 2.4296,
+      "step": 28770
+    },
+    {
+      "epoch": 2.144560357675112,
+      "grad_norm": 2.8102099895477295,
+      "learning_rate": 0.0002,
+      "loss": 2.4006,
+      "step": 28780
+    },
+    {
+      "epoch": 2.145305514157973,
+      "grad_norm": 2.7067055702209473,
+      "learning_rate": 0.0002,
+      "loss": 2.5078,
+      "step": 28790
+    },
+    {
+      "epoch": 2.1460506706408347,
+      "grad_norm": 2.4002151489257812,
+      "learning_rate": 0.0002,
+      "loss": 2.6035,
+      "step": 28800
+    },
+    {
+      "epoch": 2.1467958271236958,
+      "grad_norm": 2.9272592067718506,
+      "learning_rate": 0.0002,
+      "loss": 2.6574,
+      "step": 28810
+    },
+    {
+      "epoch": 2.1475409836065573,
+      "grad_norm": 2.3437740802764893,
+      "learning_rate": 0.0002,
+      "loss": 2.3623,
+      "step": 28820
+    },
+    {
+      "epoch": 2.148286140089419,
+      "grad_norm": 2.795177459716797,
+      "learning_rate": 0.0002,
+      "loss": 2.4582,
+      "step": 28830
+    },
+    {
+      "epoch": 2.14903129657228,
+      "grad_norm": 2.934332847595215,
+      "learning_rate": 0.0002,
+      "loss": 2.5275,
+      "step": 28840
+    },
+    {
+      "epoch": 2.1497764530551415,
+      "grad_norm": 2.446051597595215,
+      "learning_rate": 0.0002,
+      "loss": 2.3525,
+      "step": 28850
+    },
+    {
+      "epoch": 2.150521609538003,
+      "grad_norm": 2.6848883628845215,
+      "learning_rate": 0.0002,
+      "loss": 2.5435,
+      "step": 28860
+    },
+    {
+      "epoch": 2.151266766020864,
+      "grad_norm": 2.80637788772583,
+      "learning_rate": 0.0002,
+      "loss": 2.3997,
+      "step": 28870
+    },
+    {
+      "epoch": 2.1520119225037257,
+      "grad_norm": 2.7611582279205322,
+      "learning_rate": 0.0002,
+      "loss": 2.4226,
+      "step": 28880
+    },
+    {
+      "epoch": 2.1527570789865873,
+      "grad_norm": 3.1131093502044678,
+      "learning_rate": 0.0002,
+      "loss": 2.2822,
+      "step": 28890
+    },
+    {
+      "epoch": 2.1535022354694484,
+      "grad_norm": 2.372504949569702,
+      "learning_rate": 0.0002,
+      "loss": 2.3701,
+      "step": 28900
+    },
+    {
+      "epoch": 2.15424739195231,
+      "grad_norm": 2.3762528896331787,
+      "learning_rate": 0.0002,
+      "loss": 2.4664,
+      "step": 28910
+    },
+    {
+      "epoch": 2.1549925484351715,
+      "grad_norm": 2.665152072906494,
+      "learning_rate": 0.0002,
+      "loss": 2.5005,
+      "step": 28920
+    },
+    {
+      "epoch": 2.1557377049180326,
+      "grad_norm": 2.8510541915893555,
+      "learning_rate": 0.0002,
+      "loss": 2.5503,
+      "step": 28930
+    },
+    {
+      "epoch": 2.156482861400894,
+      "grad_norm": 2.647361993789673,
+      "learning_rate": 0.0002,
+      "loss": 2.2592,
+      "step": 28940
+    },
+    {
+      "epoch": 2.1572280178837557,
+      "grad_norm": 2.626680612564087,
+      "learning_rate": 0.0002,
+      "loss": 2.4075,
+      "step": 28950
+    },
+    {
+      "epoch": 2.157973174366617,
+      "grad_norm": 3.47946834564209,
+      "learning_rate": 0.0002,
+      "loss": 2.5345,
+      "step": 28960
+    },
+    {
+      "epoch": 2.1587183308494784,
+      "grad_norm": 2.671481132507324,
+      "learning_rate": 0.0002,
+      "loss": 2.2957,
+      "step": 28970
+    },
+    {
+      "epoch": 2.15946348733234,
+      "grad_norm": 2.696821689605713,
+      "learning_rate": 0.0002,
+      "loss": 2.3831,
+      "step": 28980
+    },
+    {
+      "epoch": 2.160208643815201,
+      "grad_norm": 2.814481019973755,
+      "learning_rate": 0.0002,
+      "loss": 2.4499,
+      "step": 28990
+    },
+    {
+      "epoch": 2.1609538002980626,
+      "grad_norm": 2.933293342590332,
+      "learning_rate": 0.0002,
+      "loss": 2.3267,
+      "step": 29000
+    },
+    {
+      "epoch": 2.161698956780924,
+      "grad_norm": 2.5142786502838135,
+      "learning_rate": 0.0002,
+      "loss": 2.349,
+      "step": 29010
+    },
+    {
+      "epoch": 2.1624441132637853,
+      "grad_norm": 2.8400211334228516,
+      "learning_rate": 0.0002,
+      "loss": 2.4176,
+      "step": 29020
+    },
+    {
+      "epoch": 2.163189269746647,
+      "grad_norm": 2.3746659755706787,
+      "learning_rate": 0.0002,
+      "loss": 2.4822,
+      "step": 29030
+    },
+    {
+      "epoch": 2.1639344262295084,
+      "grad_norm": 2.615368604660034,
+      "learning_rate": 0.0002,
+      "loss": 2.4141,
+      "step": 29040
+    },
+    {
+      "epoch": 2.1646795827123695,
+      "grad_norm": 2.766857385635376,
+      "learning_rate": 0.0002,
+      "loss": 2.5287,
+      "step": 29050
+    },
+    {
+      "epoch": 2.165424739195231,
+      "grad_norm": 2.6013858318328857,
+      "learning_rate": 0.0002,
+      "loss": 2.5461,
+      "step": 29060
+    },
+    {
+      "epoch": 2.1661698956780926,
+      "grad_norm": 2.6237449645996094,
+      "learning_rate": 0.0002,
+      "loss": 2.4418,
+      "step": 29070
+    },
+    {
+      "epoch": 2.1669150521609537,
+      "grad_norm": 2.9323627948760986,
+      "learning_rate": 0.0002,
+      "loss": 2.504,
+      "step": 29080
+    },
+    {
+      "epoch": 2.1676602086438153,
+      "grad_norm": 2.950990915298462,
+      "learning_rate": 0.0002,
+      "loss": 2.2819,
+      "step": 29090
+    },
+    {
+      "epoch": 2.168405365126677,
+      "grad_norm": 2.595555305480957,
+      "learning_rate": 0.0002,
+      "loss": 2.4392,
+      "step": 29100
+    },
+    {
+      "epoch": 2.169150521609538,
+      "grad_norm": 2.9946279525756836,
+      "learning_rate": 0.0002,
+      "loss": 2.538,
+      "step": 29110
+    },
+    {
+      "epoch": 2.1698956780923995,
+      "grad_norm": 2.4822959899902344,
+      "learning_rate": 0.0002,
+      "loss": 2.329,
+      "step": 29120
+    },
+    {
+      "epoch": 2.170640834575261,
+      "grad_norm": 2.95805287361145,
+      "learning_rate": 0.0002,
+      "loss": 2.5216,
+      "step": 29130
+    },
+    {
+      "epoch": 2.171385991058122,
+      "grad_norm": 2.540740966796875,
+      "learning_rate": 0.0002,
+      "loss": 2.3522,
+      "step": 29140
+    },
+    {
+      "epoch": 2.1721311475409837,
+      "grad_norm": 2.614377737045288,
+      "learning_rate": 0.0002,
+      "loss": 2.4523,
+      "step": 29150
+    },
+    {
+      "epoch": 2.172876304023845,
+      "grad_norm": 2.5126142501831055,
+      "learning_rate": 0.0002,
+      "loss": 2.4256,
+      "step": 29160
+    },
+    {
+      "epoch": 2.1736214605067063,
+      "grad_norm": 2.8608033657073975,
+      "learning_rate": 0.0002,
+      "loss": 2.3527,
+      "step": 29170
+    },
+    {
+      "epoch": 2.174366616989568,
+      "grad_norm": 2.863196849822998,
+      "learning_rate": 0.0002,
+      "loss": 2.3969,
+      "step": 29180
+    },
+    {
+      "epoch": 2.175111773472429,
+      "grad_norm": 2.6187872886657715,
+      "learning_rate": 0.0002,
+      "loss": 2.569,
+      "step": 29190
+    },
+    {
+      "epoch": 2.1758569299552906,
+      "grad_norm": 2.520378351211548,
+      "learning_rate": 0.0002,
+      "loss": 2.4414,
+      "step": 29200
+    },
+    {
+      "epoch": 2.176602086438152,
+      "grad_norm": 2.438509464263916,
+      "learning_rate": 0.0002,
+      "loss": 2.5206,
+      "step": 29210
+    },
+    {
+      "epoch": 2.1773472429210132,
+      "grad_norm": 2.899704694747925,
+      "learning_rate": 0.0002,
+      "loss": 2.4476,
+      "step": 29220
+    },
+    {
+      "epoch": 2.178092399403875,
+      "grad_norm": 2.8411431312561035,
+      "learning_rate": 0.0002,
+      "loss": 2.3632,
+      "step": 29230
+    },
+    {
+      "epoch": 2.1788375558867363,
+      "grad_norm": 3.046539783477783,
+      "learning_rate": 0.0002,
+      "loss": 2.4188,
+      "step": 29240
+    },
+    {
+      "epoch": 2.1795827123695974,
+      "grad_norm": 2.8408050537109375,
+      "learning_rate": 0.0002,
+      "loss": 2.5058,
+      "step": 29250
+    },
+    {
+      "epoch": 2.180327868852459,
+      "grad_norm": 1.9899271726608276,
+      "learning_rate": 0.0002,
+      "loss": 2.3903,
+      "step": 29260
+    },
+    {
+      "epoch": 2.1810730253353205,
+      "grad_norm": 3.0790789127349854,
+      "learning_rate": 0.0002,
+      "loss": 2.4835,
+      "step": 29270
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 2.649176836013794,
+      "learning_rate": 0.0002,
+      "loss": 2.381,
+      "step": 29280
+    },
+    {
+      "epoch": 2.182563338301043,
+      "grad_norm": 2.7816390991210938,
+      "learning_rate": 0.0002,
+      "loss": 2.5655,
+      "step": 29290
+    },
+    {
+      "epoch": 2.1833084947839048,
+      "grad_norm": 2.8856558799743652,
+      "learning_rate": 0.0002,
+      "loss": 2.4248,
+      "step": 29300
+    },
+    {
+      "epoch": 2.184053651266766,
+      "grad_norm": 2.5775249004364014,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 29310
+    },
+    {
+      "epoch": 2.1847988077496274,
+      "grad_norm": 2.680647134780884,
+      "learning_rate": 0.0002,
+      "loss": 2.4791,
+      "step": 29320
+    },
+    {
+      "epoch": 2.185543964232489,
+      "grad_norm": 2.7082343101501465,
+      "learning_rate": 0.0002,
+      "loss": 2.2716,
+      "step": 29330
+    },
+    {
+      "epoch": 2.18628912071535,
+      "grad_norm": 2.76863956451416,
+      "learning_rate": 0.0002,
+      "loss": 2.3834,
+      "step": 29340
+    },
+    {
+      "epoch": 2.1870342771982116,
+      "grad_norm": 2.652808666229248,
+      "learning_rate": 0.0002,
+      "loss": 2.5556,
+      "step": 29350
+    },
+    {
+      "epoch": 2.187779433681073,
+      "grad_norm": 2.7211995124816895,
+      "learning_rate": 0.0002,
+      "loss": 2.428,
+      "step": 29360
+    },
+    {
+      "epoch": 2.1885245901639343,
+      "grad_norm": 2.6517481803894043,
+      "learning_rate": 0.0002,
+      "loss": 2.3506,
+      "step": 29370
+    },
+    {
+      "epoch": 2.189269746646796,
+      "grad_norm": 2.680163621902466,
+      "learning_rate": 0.0002,
+      "loss": 2.3581,
+      "step": 29380
+    },
+    {
+      "epoch": 2.1900149031296574,
+      "grad_norm": 2.6171483993530273,
+      "learning_rate": 0.0002,
+      "loss": 2.3894,
+      "step": 29390
+    },
+    {
+      "epoch": 2.1907600596125185,
+      "grad_norm": 2.645461320877075,
+      "learning_rate": 0.0002,
+      "loss": 2.4044,
+      "step": 29400
+    },
+    {
+      "epoch": 2.19150521609538,
+      "grad_norm": 2.607429265975952,
+      "learning_rate": 0.0002,
+      "loss": 2.3481,
+      "step": 29410
+    },
+    {
+      "epoch": 2.1922503725782416,
+      "grad_norm": 2.634819269180298,
+      "learning_rate": 0.0002,
+      "loss": 2.3877,
+      "step": 29420
+    },
+    {
+      "epoch": 2.1929955290611027,
+      "grad_norm": 2.682586193084717,
+      "learning_rate": 0.0002,
+      "loss": 2.3752,
+      "step": 29430
+    },
+    {
+      "epoch": 2.1937406855439643,
+      "grad_norm": 2.5644547939300537,
+      "learning_rate": 0.0002,
+      "loss": 2.4439,
+      "step": 29440
+    },
+    {
+      "epoch": 2.194485842026826,
+      "grad_norm": 2.639521837234497,
+      "learning_rate": 0.0002,
+      "loss": 2.4411,
+      "step": 29450
+    },
+    {
+      "epoch": 2.195230998509687,
+      "grad_norm": 2.455165147781372,
+      "learning_rate": 0.0002,
+      "loss": 2.6128,
+      "step": 29460
+    },
+    {
+      "epoch": 2.1959761549925485,
+      "grad_norm": 2.800022602081299,
+      "learning_rate": 0.0002,
+      "loss": 2.4411,
+      "step": 29470
+    },
+    {
+      "epoch": 2.19672131147541,
+      "grad_norm": 2.4787418842315674,
+      "learning_rate": 0.0002,
+      "loss": 2.4363,
+      "step": 29480
+    },
+    {
+      "epoch": 2.197466467958271,
+      "grad_norm": 2.7749505043029785,
+      "learning_rate": 0.0002,
+      "loss": 2.4775,
+      "step": 29490
+    },
+    {
+      "epoch": 2.1982116244411327,
+      "grad_norm": 2.472622871398926,
+      "learning_rate": 0.0002,
+      "loss": 2.3768,
+      "step": 29500
+    },
+    {
+      "epoch": 2.198956780923994,
+      "grad_norm": 2.560708999633789,
+      "learning_rate": 0.0002,
+      "loss": 2.4336,
+      "step": 29510
+    },
+    {
+      "epoch": 2.1997019374068554,
+      "grad_norm": 2.5658538341522217,
+      "learning_rate": 0.0002,
+      "loss": 2.389,
+      "step": 29520
+    },
+    {
+      "epoch": 2.200447093889717,
+      "grad_norm": 2.6552112102508545,
+      "learning_rate": 0.0002,
+      "loss": 2.1017,
+      "step": 29530
+    },
+    {
+      "epoch": 2.201192250372578,
+      "grad_norm": 2.23706316947937,
+      "learning_rate": 0.0002,
+      "loss": 2.4461,
+      "step": 29540
+    },
+    {
+      "epoch": 2.2019374068554396,
+      "grad_norm": 2.763774871826172,
+      "learning_rate": 0.0002,
+      "loss": 2.4351,
+      "step": 29550
+    },
+    {
+      "epoch": 2.202682563338301,
+      "grad_norm": 2.5139219760894775,
+      "learning_rate": 0.0002,
+      "loss": 2.351,
+      "step": 29560
+    },
+    {
+      "epoch": 2.2034277198211623,
+      "grad_norm": 2.893038272857666,
+      "learning_rate": 0.0002,
+      "loss": 2.4692,
+      "step": 29570
+    },
+    {
+      "epoch": 2.204172876304024,
+      "grad_norm": 2.818425178527832,
+      "learning_rate": 0.0002,
+      "loss": 2.4447,
+      "step": 29580
+    },
+    {
+      "epoch": 2.2049180327868854,
+      "grad_norm": 2.331808567047119,
+      "learning_rate": 0.0002,
+      "loss": 2.4306,
+      "step": 29590
+    },
+    {
+      "epoch": 2.2056631892697465,
+      "grad_norm": 3.2033214569091797,
+      "learning_rate": 0.0002,
+      "loss": 2.5642,
+      "step": 29600
+    },
+    {
+      "epoch": 2.206408345752608,
+      "grad_norm": 2.41534686088562,
+      "learning_rate": 0.0002,
+      "loss": 2.3699,
+      "step": 29610
+    },
+    {
+      "epoch": 2.2071535022354696,
+      "grad_norm": 2.8032174110412598,
+      "learning_rate": 0.0002,
+      "loss": 2.6087,
+      "step": 29620
+    },
+    {
+      "epoch": 2.2078986587183307,
+      "grad_norm": 2.4665184020996094,
+      "learning_rate": 0.0002,
+      "loss": 2.4702,
+      "step": 29630
+    },
+    {
+      "epoch": 2.2086438152011922,
+      "grad_norm": 2.474245309829712,
+      "learning_rate": 0.0002,
+      "loss": 2.392,
+      "step": 29640
+    },
+    {
+      "epoch": 2.209388971684054,
+      "grad_norm": 2.9560275077819824,
+      "learning_rate": 0.0002,
+      "loss": 2.5555,
+      "step": 29650
+    },
+    {
+      "epoch": 2.210134128166915,
+      "grad_norm": 2.4646859169006348,
+      "learning_rate": 0.0002,
+      "loss": 2.3858,
+      "step": 29660
+    },
+    {
+      "epoch": 2.2108792846497765,
+      "grad_norm": 2.8833131790161133,
+      "learning_rate": 0.0002,
+      "loss": 2.3451,
+      "step": 29670
+    },
+    {
+      "epoch": 2.211624441132638,
+      "grad_norm": 2.540769577026367,
+      "learning_rate": 0.0002,
+      "loss": 2.256,
+      "step": 29680
+    },
+    {
+      "epoch": 2.212369597615499,
+      "grad_norm": 3.0184593200683594,
+      "learning_rate": 0.0002,
+      "loss": 2.4306,
+      "step": 29690
+    },
+    {
+      "epoch": 2.2131147540983607,
+      "grad_norm": 2.6296427249908447,
+      "learning_rate": 0.0002,
+      "loss": 2.5813,
+      "step": 29700
+    },
+    {
+      "epoch": 2.2138599105812222,
+      "grad_norm": 2.667278528213501,
+      "learning_rate": 0.0002,
+      "loss": 2.4874,
+      "step": 29710
+    },
+    {
+      "epoch": 2.2146050670640833,
+      "grad_norm": 2.660968065261841,
+      "learning_rate": 0.0002,
+      "loss": 2.3787,
+      "step": 29720
+    },
+    {
+      "epoch": 2.215350223546945,
+      "grad_norm": 2.5494511127471924,
+      "learning_rate": 0.0002,
+      "loss": 2.4766,
+      "step": 29730
+    },
+    {
+      "epoch": 2.2160953800298064,
+      "grad_norm": 2.503387928009033,
+      "learning_rate": 0.0002,
+      "loss": 2.4909,
+      "step": 29740
+    },
+    {
+      "epoch": 2.2168405365126675,
+      "grad_norm": 2.6121363639831543,
+      "learning_rate": 0.0002,
+      "loss": 2.5237,
+      "step": 29750
+    },
+    {
+      "epoch": 2.217585692995529,
+      "grad_norm": 2.4190587997436523,
+      "learning_rate": 0.0002,
+      "loss": 2.3553,
+      "step": 29760
+    },
+    {
+      "epoch": 2.2183308494783907,
+      "grad_norm": 2.8600144386291504,
+      "learning_rate": 0.0002,
+      "loss": 2.4106,
+      "step": 29770
+    },
+    {
+      "epoch": 2.2190760059612518,
+      "grad_norm": 3.615830898284912,
+      "learning_rate": 0.0002,
+      "loss": 2.5998,
+      "step": 29780
+    },
+    {
+      "epoch": 2.2198211624441133,
+      "grad_norm": 3.1991500854492188,
+      "learning_rate": 0.0002,
+      "loss": 2.6738,
+      "step": 29790
+    },
+    {
+      "epoch": 2.220566318926975,
+      "grad_norm": 2.7967700958251953,
+      "learning_rate": 0.0002,
+      "loss": 2.4236,
+      "step": 29800
+    },
+    {
+      "epoch": 2.221311475409836,
+      "grad_norm": 2.68129301071167,
+      "learning_rate": 0.0002,
+      "loss": 2.4228,
+      "step": 29810
+    },
+    {
+      "epoch": 2.2220566318926975,
+      "grad_norm": 2.7020351886749268,
+      "learning_rate": 0.0002,
+      "loss": 2.3199,
+      "step": 29820
+    },
+    {
+      "epoch": 2.222801788375559,
+      "grad_norm": 2.417415142059326,
+      "learning_rate": 0.0002,
+      "loss": 2.344,
+      "step": 29830
+    },
+    {
+      "epoch": 2.22354694485842,
+      "grad_norm": 2.5789639949798584,
+      "learning_rate": 0.0002,
+      "loss": 2.3901,
+      "step": 29840
+    },
+    {
+      "epoch": 2.2242921013412817,
+      "grad_norm": 2.623487710952759,
+      "learning_rate": 0.0002,
+      "loss": 2.3944,
+      "step": 29850
+    },
+    {
+      "epoch": 2.225037257824143,
+      "grad_norm": 2.554654598236084,
+      "learning_rate": 0.0002,
+      "loss": 2.3396,
+      "step": 29860
+    },
+    {
+      "epoch": 2.2257824143070044,
+      "grad_norm": 2.624058485031128,
+      "learning_rate": 0.0002,
+      "loss": 2.222,
+      "step": 29870
+    },
+    {
+      "epoch": 2.226527570789866,
+      "grad_norm": 2.9472296237945557,
+      "learning_rate": 0.0002,
+      "loss": 2.5653,
+      "step": 29880
+    },
+    {
+      "epoch": 2.227272727272727,
+      "grad_norm": 2.7747509479522705,
+      "learning_rate": 0.0002,
+      "loss": 2.3922,
+      "step": 29890
+    },
+    {
+      "epoch": 2.2280178837555886,
+      "grad_norm": 2.717242956161499,
+      "learning_rate": 0.0002,
+      "loss": 2.3198,
+      "step": 29900
+    },
+    {
+      "epoch": 2.22876304023845,
+      "grad_norm": 2.829017400741577,
+      "learning_rate": 0.0002,
+      "loss": 2.2318,
+      "step": 29910
+    },
+    {
+      "epoch": 2.2295081967213113,
+      "grad_norm": 2.65274715423584,
+      "learning_rate": 0.0002,
+      "loss": 2.3987,
+      "step": 29920
+    },
+    {
+      "epoch": 2.230253353204173,
+      "grad_norm": 2.738203287124634,
+      "learning_rate": 0.0002,
+      "loss": 2.2976,
+      "step": 29930
+    },
+    {
+      "epoch": 2.2309985096870344,
+      "grad_norm": 2.4520084857940674,
+      "learning_rate": 0.0002,
+      "loss": 2.5559,
+      "step": 29940
+    },
+    {
+      "epoch": 2.2317436661698955,
+      "grad_norm": 2.829719066619873,
+      "learning_rate": 0.0002,
+      "loss": 2.3981,
+      "step": 29950
+    },
+    {
+      "epoch": 2.232488822652757,
+      "grad_norm": 2.8684730529785156,
+      "learning_rate": 0.0002,
+      "loss": 2.4195,
+      "step": 29960
+    },
+    {
+      "epoch": 2.2332339791356186,
+      "grad_norm": 2.597276449203491,
+      "learning_rate": 0.0002,
+      "loss": 2.4955,
+      "step": 29970
+    },
+    {
+      "epoch": 2.2339791356184797,
+      "grad_norm": 2.678597927093506,
+      "learning_rate": 0.0002,
+      "loss": 2.3952,
+      "step": 29980
+    },
+    {
+      "epoch": 2.2347242921013413,
+      "grad_norm": 2.6332530975341797,
+      "learning_rate": 0.0002,
+      "loss": 2.6308,
+      "step": 29990
+    },
+    {
+      "epoch": 2.235469448584203,
+      "grad_norm": 2.6793925762176514,
+      "learning_rate": 0.0002,
+      "loss": 2.3281,
+      "step": 30000
+    },
+    {
+      "epoch": 2.236214605067064,
+      "grad_norm": 2.714738607406616,
+      "learning_rate": 0.0002,
+      "loss": 2.4091,
+      "step": 30010
+    },
+    {
+      "epoch": 2.2369597615499255,
+      "grad_norm": 2.5853700637817383,
+      "learning_rate": 0.0002,
+      "loss": 2.4243,
+      "step": 30020
+    },
+    {
+      "epoch": 2.237704918032787,
+      "grad_norm": 2.972341775894165,
+      "learning_rate": 0.0002,
+      "loss": 2.4984,
+      "step": 30030
+    },
+    {
+      "epoch": 2.238450074515648,
+      "grad_norm": 2.5449442863464355,
+      "learning_rate": 0.0002,
+      "loss": 2.4433,
+      "step": 30040
+    },
+    {
+      "epoch": 2.2391952309985097,
+      "grad_norm": 2.6521236896514893,
+      "learning_rate": 0.0002,
+      "loss": 2.5521,
+      "step": 30050
+    },
+    {
+      "epoch": 2.2399403874813713,
+      "grad_norm": 2.510643482208252,
+      "learning_rate": 0.0002,
+      "loss": 2.5315,
+      "step": 30060
+    },
+    {
+      "epoch": 2.2406855439642324,
+      "grad_norm": 2.896329641342163,
+      "learning_rate": 0.0002,
+      "loss": 2.4374,
+      "step": 30070
+    },
+    {
+      "epoch": 2.241430700447094,
+      "grad_norm": 2.6130764484405518,
+      "learning_rate": 0.0002,
+      "loss": 2.4338,
+      "step": 30080
+    },
+    {
+      "epoch": 2.2421758569299555,
+      "grad_norm": 2.5632150173187256,
+      "learning_rate": 0.0002,
+      "loss": 2.4224,
+      "step": 30090
+    },
+    {
+      "epoch": 2.2429210134128166,
+      "grad_norm": 2.6753768920898438,
+      "learning_rate": 0.0002,
+      "loss": 2.4606,
+      "step": 30100
+    },
+    {
+      "epoch": 2.243666169895678,
+      "grad_norm": 2.60931396484375,
+      "learning_rate": 0.0002,
+      "loss": 2.4416,
+      "step": 30110
+    },
+    {
+      "epoch": 2.2444113263785397,
+      "grad_norm": 2.428415536880493,
+      "learning_rate": 0.0002,
+      "loss": 2.596,
+      "step": 30120
+    },
+    {
+      "epoch": 2.245156482861401,
+      "grad_norm": 2.97552752494812,
+      "learning_rate": 0.0002,
+      "loss": 2.6441,
+      "step": 30130
+    },
+    {
+      "epoch": 2.2459016393442623,
+      "grad_norm": 2.513252019882202,
+      "learning_rate": 0.0002,
+      "loss": 2.5045,
+      "step": 30140
+    },
+    {
+      "epoch": 2.246646795827124,
+      "grad_norm": 2.529350996017456,
+      "learning_rate": 0.0002,
+      "loss": 2.2703,
+      "step": 30150
+    },
+    {
+      "epoch": 2.247391952309985,
+      "grad_norm": 2.729717254638672,
+      "learning_rate": 0.0002,
+      "loss": 2.5772,
+      "step": 30160
+    },
+    {
+      "epoch": 2.2481371087928466,
+      "grad_norm": 2.8661959171295166,
+      "learning_rate": 0.0002,
+      "loss": 2.5031,
+      "step": 30170
+    },
+    {
+      "epoch": 2.248882265275708,
+      "grad_norm": 2.6999034881591797,
+      "learning_rate": 0.0002,
+      "loss": 2.4389,
+      "step": 30180
+    },
+    {
+      "epoch": 2.2496274217585692,
+      "grad_norm": 3.331155300140381,
+      "learning_rate": 0.0002,
+      "loss": 2.5102,
+      "step": 30190
+    },
+    {
+      "epoch": 2.2503725782414308,
+      "grad_norm": 2.8043971061706543,
+      "learning_rate": 0.0002,
+      "loss": 2.3276,
+      "step": 30200
+    },
+    {
+      "epoch": 2.251117734724292,
+      "grad_norm": 2.57314133644104,
+      "learning_rate": 0.0002,
+      "loss": 2.3967,
+      "step": 30210
+    },
+    {
+      "epoch": 2.2518628912071534,
+      "grad_norm": 2.738511085510254,
+      "learning_rate": 0.0002,
+      "loss": 2.5336,
+      "step": 30220
+    },
+    {
+      "epoch": 2.252608047690015,
+      "grad_norm": 2.7028000354766846,
+      "learning_rate": 0.0002,
+      "loss": 2.4197,
+      "step": 30230
+    },
+    {
+      "epoch": 2.2533532041728765,
+      "grad_norm": 1.9753464460372925,
+      "learning_rate": 0.0002,
+      "loss": 2.2315,
+      "step": 30240
+    },
+    {
+      "epoch": 2.2540983606557377,
+      "grad_norm": 2.5343260765075684,
+      "learning_rate": 0.0002,
+      "loss": 2.656,
+      "step": 30250
+    },
+    {
+      "epoch": 2.254843517138599,
+      "grad_norm": 3.3517582416534424,
+      "learning_rate": 0.0002,
+      "loss": 2.5387,
+      "step": 30260
+    },
+    {
+      "epoch": 2.2555886736214603,
+      "grad_norm": 2.795591115951538,
+      "learning_rate": 0.0002,
+      "loss": 2.428,
+      "step": 30270
+    },
+    {
+      "epoch": 2.256333830104322,
+      "grad_norm": 2.458268404006958,
+      "learning_rate": 0.0002,
+      "loss": 2.6109,
+      "step": 30280
+    },
+    {
+      "epoch": 2.2570789865871834,
+      "grad_norm": 2.965975761413574,
+      "learning_rate": 0.0002,
+      "loss": 2.265,
+      "step": 30290
+    },
+    {
+      "epoch": 2.2578241430700445,
+      "grad_norm": 2.6871607303619385,
+      "learning_rate": 0.0002,
+      "loss": 2.3368,
+      "step": 30300
+    },
+    {
+      "epoch": 2.258569299552906,
+      "grad_norm": 2.760650157928467,
+      "learning_rate": 0.0002,
+      "loss": 2.4479,
+      "step": 30310
+    },
+    {
+      "epoch": 2.2593144560357676,
+      "grad_norm": 2.5536069869995117,
+      "learning_rate": 0.0002,
+      "loss": 2.3555,
+      "step": 30320
+    },
+    {
+      "epoch": 2.2600596125186287,
+      "grad_norm": 2.6912643909454346,
+      "learning_rate": 0.0002,
+      "loss": 2.4761,
+      "step": 30330
+    },
+    {
+      "epoch": 2.2608047690014903,
+      "grad_norm": 2.6149418354034424,
+      "learning_rate": 0.0002,
+      "loss": 2.3776,
+      "step": 30340
+    },
+    {
+      "epoch": 2.261549925484352,
+      "grad_norm": 2.6882479190826416,
+      "learning_rate": 0.0002,
+      "loss": 2.4329,
+      "step": 30350
+    },
+    {
+      "epoch": 2.262295081967213,
+      "grad_norm": 2.7375595569610596,
+      "learning_rate": 0.0002,
+      "loss": 2.4161,
+      "step": 30360
+    },
+    {
+      "epoch": 2.2630402384500745,
+      "grad_norm": 2.4539854526519775,
+      "learning_rate": 0.0002,
+      "loss": 2.4432,
+      "step": 30370
+    },
+    {
+      "epoch": 2.263785394932936,
+      "grad_norm": 2.7913661003112793,
+      "learning_rate": 0.0002,
+      "loss": 2.6048,
+      "step": 30380
+    },
+    {
+      "epoch": 2.264530551415797,
+      "grad_norm": 2.822239875793457,
+      "learning_rate": 0.0002,
+      "loss": 2.6286,
+      "step": 30390
+    },
+    {
+      "epoch": 2.2652757078986587,
+      "grad_norm": 2.82560658454895,
+      "learning_rate": 0.0002,
+      "loss": 2.2957,
+      "step": 30400
+    },
+    {
+      "epoch": 2.2660208643815203,
+      "grad_norm": 2.30305814743042,
+      "learning_rate": 0.0002,
+      "loss": 2.5297,
+      "step": 30410
+    },
+    {
+      "epoch": 2.2667660208643814,
+      "grad_norm": 2.366816520690918,
+      "learning_rate": 0.0002,
+      "loss": 2.3713,
+      "step": 30420
+    },
+    {
+      "epoch": 2.267511177347243,
+      "grad_norm": 2.849090814590454,
+      "learning_rate": 0.0002,
+      "loss": 2.2744,
+      "step": 30430
+    },
+    {
+      "epoch": 2.2682563338301045,
+      "grad_norm": 2.4802587032318115,
+      "learning_rate": 0.0002,
+      "loss": 2.2494,
+      "step": 30440
+    },
+    {
+      "epoch": 2.2690014903129656,
+      "grad_norm": 2.9670379161834717,
+      "learning_rate": 0.0002,
+      "loss": 2.5435,
+      "step": 30450
+    },
+    {
+      "epoch": 2.269746646795827,
+      "grad_norm": 2.596202850341797,
+      "learning_rate": 0.0002,
+      "loss": 2.428,
+      "step": 30460
+    },
+    {
+      "epoch": 2.2704918032786887,
+      "grad_norm": 2.6651835441589355,
+      "learning_rate": 0.0002,
+      "loss": 2.3748,
+      "step": 30470
+    },
+    {
+      "epoch": 2.27123695976155,
+      "grad_norm": 2.2997448444366455,
+      "learning_rate": 0.0002,
+      "loss": 2.5528,
+      "step": 30480
+    },
+    {
+      "epoch": 2.2719821162444114,
+      "grad_norm": 2.415158271789551,
+      "learning_rate": 0.0002,
+      "loss": 2.3274,
+      "step": 30490
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "grad_norm": 2.361969470977783,
+      "learning_rate": 0.0002,
+      "loss": 2.5814,
+      "step": 30500
+    },
+    {
+      "epoch": 2.273472429210134,
+      "grad_norm": 2.340508222579956,
+      "learning_rate": 0.0002,
+      "loss": 2.2694,
+      "step": 30510
+    },
+    {
+      "epoch": 2.2742175856929956,
+      "grad_norm": 2.5094423294067383,
+      "learning_rate": 0.0002,
+      "loss": 2.4632,
+      "step": 30520
+    },
+    {
+      "epoch": 2.274962742175857,
+      "grad_norm": 2.8565590381622314,
+      "learning_rate": 0.0002,
+      "loss": 2.2868,
+      "step": 30530
+    },
+    {
+      "epoch": 2.2757078986587183,
+      "grad_norm": 2.701770067214966,
+      "learning_rate": 0.0002,
+      "loss": 2.4865,
+      "step": 30540
+    },
+    {
+      "epoch": 2.27645305514158,
+      "grad_norm": 2.518066644668579,
+      "learning_rate": 0.0002,
+      "loss": 2.5408,
+      "step": 30550
+    },
+    {
+      "epoch": 2.277198211624441,
+      "grad_norm": 2.743431806564331,
+      "learning_rate": 0.0002,
+      "loss": 2.4585,
+      "step": 30560
+    },
+    {
+      "epoch": 2.2779433681073025,
+      "grad_norm": 2.8521339893341064,
+      "learning_rate": 0.0002,
+      "loss": 2.3772,
+      "step": 30570
+    },
+    {
+      "epoch": 2.278688524590164,
+      "grad_norm": 2.967759370803833,
+      "learning_rate": 0.0002,
+      "loss": 2.5984,
+      "step": 30580
+    },
+    {
+      "epoch": 2.2794336810730256,
+      "grad_norm": 2.4297335147857666,
+      "learning_rate": 0.0002,
+      "loss": 2.4854,
+      "step": 30590
+    },
+    {
+      "epoch": 2.2801788375558867,
+      "grad_norm": 2.547301769256592,
+      "learning_rate": 0.0002,
+      "loss": 2.478,
+      "step": 30600
+    },
+    {
+      "epoch": 2.2809239940387482,
+      "grad_norm": 2.791649341583252,
+      "learning_rate": 0.0002,
+      "loss": 2.5139,
+      "step": 30610
+    },
+    {
+      "epoch": 2.2816691505216093,
+      "grad_norm": 2.7930760383605957,
+      "learning_rate": 0.0002,
+      "loss": 2.5727,
+      "step": 30620
+    },
+    {
+      "epoch": 2.282414307004471,
+      "grad_norm": 2.5463852882385254,
+      "learning_rate": 0.0002,
+      "loss": 2.5221,
+      "step": 30630
+    },
+    {
+      "epoch": 2.2831594634873325,
+      "grad_norm": 2.653203248977661,
+      "learning_rate": 0.0002,
+      "loss": 2.3023,
+      "step": 30640
+    },
+    {
+      "epoch": 2.2839046199701936,
+      "grad_norm": 3.0527825355529785,
+      "learning_rate": 0.0002,
+      "loss": 2.2436,
+      "step": 30650
+    },
+    {
+      "epoch": 2.284649776453055,
+      "grad_norm": 2.5038249492645264,
+      "learning_rate": 0.0002,
+      "loss": 2.5085,
+      "step": 30660
+    },
+    {
+      "epoch": 2.2853949329359167,
+      "grad_norm": 2.571519374847412,
+      "learning_rate": 0.0002,
+      "loss": 2.2828,
+      "step": 30670
+    },
+    {
+      "epoch": 2.2861400894187778,
+      "grad_norm": 2.742832660675049,
+      "learning_rate": 0.0002,
+      "loss": 2.4795,
+      "step": 30680
+    },
+    {
+      "epoch": 2.2868852459016393,
+      "grad_norm": 2.4409167766571045,
+      "learning_rate": 0.0002,
+      "loss": 2.4449,
+      "step": 30690
+    },
+    {
+      "epoch": 2.287630402384501,
+      "grad_norm": 2.7374167442321777,
+      "learning_rate": 0.0002,
+      "loss": 2.2682,
+      "step": 30700
+    },
+    {
+      "epoch": 2.288375558867362,
+      "grad_norm": 2.6040146350860596,
+      "learning_rate": 0.0002,
+      "loss": 2.4203,
+      "step": 30710
+    },
+    {
+      "epoch": 2.2891207153502235,
+      "grad_norm": 2.70377779006958,
+      "learning_rate": 0.0002,
+      "loss": 2.3425,
+      "step": 30720
+    },
+    {
+      "epoch": 2.289865871833085,
+      "grad_norm": 2.382824420928955,
+      "learning_rate": 0.0002,
+      "loss": 2.3675,
+      "step": 30730
+    },
+    {
+      "epoch": 2.290611028315946,
+      "grad_norm": 2.6663031578063965,
+      "learning_rate": 0.0002,
+      "loss": 2.5198,
+      "step": 30740
+    },
+    {
+      "epoch": 2.2913561847988078,
+      "grad_norm": 2.4763283729553223,
+      "learning_rate": 0.0002,
+      "loss": 2.3904,
+      "step": 30750
+    },
+    {
+      "epoch": 2.2921013412816693,
+      "grad_norm": 2.601113796234131,
+      "learning_rate": 0.0002,
+      "loss": 2.5542,
+      "step": 30760
+    },
+    {
+      "epoch": 2.2928464977645304,
+      "grad_norm": 2.841099739074707,
+      "learning_rate": 0.0002,
+      "loss": 2.5474,
+      "step": 30770
+    },
+    {
+      "epoch": 2.293591654247392,
+      "grad_norm": 2.850001573562622,
+      "learning_rate": 0.0002,
+      "loss": 2.4976,
+      "step": 30780
+    },
+    {
+      "epoch": 2.2943368107302535,
+      "grad_norm": 2.646151304244995,
+      "learning_rate": 0.0002,
+      "loss": 2.4373,
+      "step": 30790
+    },
+    {
+      "epoch": 2.2950819672131146,
+      "grad_norm": 2.71795916557312,
+      "learning_rate": 0.0002,
+      "loss": 2.3082,
+      "step": 30800
+    },
+    {
+      "epoch": 2.295827123695976,
+      "grad_norm": 2.5677895545959473,
+      "learning_rate": 0.0002,
+      "loss": 2.3645,
+      "step": 30810
+    },
+    {
+      "epoch": 2.2965722801788377,
+      "grad_norm": 2.6931703090667725,
+      "learning_rate": 0.0002,
+      "loss": 2.5397,
+      "step": 30820
+    },
+    {
+      "epoch": 2.297317436661699,
+      "grad_norm": 2.990161418914795,
+      "learning_rate": 0.0002,
+      "loss": 2.5421,
+      "step": 30830
+    },
+    {
+      "epoch": 2.2980625931445604,
+      "grad_norm": 2.717191696166992,
+      "learning_rate": 0.0002,
+      "loss": 2.4656,
+      "step": 30840
+    },
+    {
+      "epoch": 2.2988077496274215,
+      "grad_norm": 3.0563786029815674,
+      "learning_rate": 0.0002,
+      "loss": 2.5854,
+      "step": 30850
+    },
+    {
+      "epoch": 2.299552906110283,
+      "grad_norm": 2.8930904865264893,
+      "learning_rate": 0.0002,
+      "loss": 2.3787,
+      "step": 30860
+    },
+    {
+      "epoch": 2.3002980625931446,
+      "grad_norm": 2.5913820266723633,
+      "learning_rate": 0.0002,
+      "loss": 2.3729,
+      "step": 30870
+    },
+    {
+      "epoch": 2.301043219076006,
+      "grad_norm": 2.668519973754883,
+      "learning_rate": 0.0002,
+      "loss": 2.2982,
+      "step": 30880
+    },
+    {
+      "epoch": 2.3017883755588673,
+      "grad_norm": 2.6482491493225098,
+      "learning_rate": 0.0002,
+      "loss": 2.4505,
+      "step": 30890
+    },
+    {
+      "epoch": 2.302533532041729,
+      "grad_norm": 2.7581119537353516,
+      "learning_rate": 0.0002,
+      "loss": 2.5304,
+      "step": 30900
+    },
+    {
+      "epoch": 2.30327868852459,
+      "grad_norm": 2.807406425476074,
+      "learning_rate": 0.0002,
+      "loss": 2.6199,
+      "step": 30910
+    },
+    {
+      "epoch": 2.3040238450074515,
+      "grad_norm": 2.669703245162964,
+      "learning_rate": 0.0002,
+      "loss": 2.3566,
+      "step": 30920
+    },
+    {
+      "epoch": 2.304769001490313,
+      "grad_norm": 2.4919867515563965,
+      "learning_rate": 0.0002,
+      "loss": 2.4593,
+      "step": 30930
+    },
+    {
+      "epoch": 2.3055141579731746,
+      "grad_norm": 2.781989574432373,
+      "learning_rate": 0.0002,
+      "loss": 2.3244,
+      "step": 30940
+    },
+    {
+      "epoch": 2.3062593144560357,
+      "grad_norm": 2.7712695598602295,
+      "learning_rate": 0.0002,
+      "loss": 2.3854,
+      "step": 30950
+    },
+    {
+      "epoch": 2.3070044709388973,
+      "grad_norm": 2.8249754905700684,
+      "learning_rate": 0.0002,
+      "loss": 2.6018,
+      "step": 30960
+    },
+    {
+      "epoch": 2.3077496274217584,
+      "grad_norm": 2.443105459213257,
+      "learning_rate": 0.0002,
+      "loss": 2.2767,
+      "step": 30970
+    },
+    {
+      "epoch": 2.30849478390462,
+      "grad_norm": 2.589513063430786,
+      "learning_rate": 0.0002,
+      "loss": 2.5455,
+      "step": 30980
+    },
+    {
+      "epoch": 2.3092399403874815,
+      "grad_norm": 2.7437689304351807,
+      "learning_rate": 0.0002,
+      "loss": 2.7474,
+      "step": 30990
+    },
+    {
+      "epoch": 2.3099850968703426,
+      "grad_norm": 2.4255599975585938,
+      "learning_rate": 0.0002,
+      "loss": 2.3217,
+      "step": 31000
+    },
+    {
+      "epoch": 2.310730253353204,
+      "grad_norm": 2.56787371635437,
+      "learning_rate": 0.0002,
+      "loss": 2.3108,
+      "step": 31010
+    },
+    {
+      "epoch": 2.3114754098360657,
+      "grad_norm": 2.7535786628723145,
+      "learning_rate": 0.0002,
+      "loss": 2.4737,
+      "step": 31020
+    },
+    {
+      "epoch": 2.312220566318927,
+      "grad_norm": 2.5729820728302,
+      "learning_rate": 0.0002,
+      "loss": 2.3027,
+      "step": 31030
+    },
+    {
+      "epoch": 2.3129657228017884,
+      "grad_norm": 2.6444244384765625,
+      "learning_rate": 0.0002,
+      "loss": 2.4639,
+      "step": 31040
+    },
+    {
+      "epoch": 2.31371087928465,
+      "grad_norm": 2.6988863945007324,
+      "learning_rate": 0.0002,
+      "loss": 2.4188,
+      "step": 31050
+    },
+    {
+      "epoch": 2.314456035767511,
+      "grad_norm": 3.3972041606903076,
+      "learning_rate": 0.0002,
+      "loss": 2.639,
+      "step": 31060
+    },
+    {
+      "epoch": 2.3152011922503726,
+      "grad_norm": 2.741131544113159,
+      "learning_rate": 0.0002,
+      "loss": 2.484,
+      "step": 31070
+    },
+    {
+      "epoch": 2.315946348733234,
+      "grad_norm": 2.761235475540161,
+      "learning_rate": 0.0002,
+      "loss": 2.4808,
+      "step": 31080
+    },
+    {
+      "epoch": 2.3166915052160952,
+      "grad_norm": 2.5519707202911377,
+      "learning_rate": 0.0002,
+      "loss": 2.5373,
+      "step": 31090
+    },
+    {
+      "epoch": 2.317436661698957,
+      "grad_norm": 2.4959230422973633,
+      "learning_rate": 0.0002,
+      "loss": 2.5488,
+      "step": 31100
+    },
+    {
+      "epoch": 2.3181818181818183,
+      "grad_norm": 2.7000091075897217,
+      "learning_rate": 0.0002,
+      "loss": 2.2065,
+      "step": 31110
+    },
+    {
+      "epoch": 2.3189269746646795,
+      "grad_norm": 2.97749662399292,
+      "learning_rate": 0.0002,
+      "loss": 2.3805,
+      "step": 31120
+    },
+    {
+      "epoch": 2.319672131147541,
+      "grad_norm": 2.6675100326538086,
+      "learning_rate": 0.0002,
+      "loss": 2.5932,
+      "step": 31130
+    },
+    {
+      "epoch": 2.3204172876304026,
+      "grad_norm": 2.4122798442840576,
+      "learning_rate": 0.0002,
+      "loss": 2.47,
+      "step": 31140
+    },
+    {
+      "epoch": 2.3211624441132637,
+      "grad_norm": 2.4250221252441406,
+      "learning_rate": 0.0002,
+      "loss": 2.3444,
+      "step": 31150
+    },
+    {
+      "epoch": 2.321907600596125,
+      "grad_norm": 2.625974178314209,
+      "learning_rate": 0.0002,
+      "loss": 2.6722,
+      "step": 31160
+    },
+    {
+      "epoch": 2.3226527570789868,
+      "grad_norm": 2.5637755393981934,
+      "learning_rate": 0.0002,
+      "loss": 2.4901,
+      "step": 31170
+    },
+    {
+      "epoch": 2.323397913561848,
+      "grad_norm": 2.491208076477051,
+      "learning_rate": 0.0002,
+      "loss": 2.3771,
+      "step": 31180
+    },
+    {
+      "epoch": 2.3241430700447094,
+      "grad_norm": 2.751556634902954,
+      "learning_rate": 0.0002,
+      "loss": 2.3967,
+      "step": 31190
+    },
+    {
+      "epoch": 2.3248882265275705,
+      "grad_norm": 2.622948169708252,
+      "learning_rate": 0.0002,
+      "loss": 2.4714,
+      "step": 31200
+    },
+    {
+      "epoch": 2.325633383010432,
+      "grad_norm": 2.500760078430176,
+      "learning_rate": 0.0002,
+      "loss": 2.4578,
+      "step": 31210
+    },
+    {
+      "epoch": 2.3263785394932937,
+      "grad_norm": 2.7233376502990723,
+      "learning_rate": 0.0002,
+      "loss": 2.5258,
+      "step": 31220
+    },
+    {
+      "epoch": 2.327123695976155,
+      "grad_norm": 2.32476806640625,
+      "learning_rate": 0.0002,
+      "loss": 2.431,
+      "step": 31230
+    },
+    {
+      "epoch": 2.3278688524590163,
+      "grad_norm": 2.283064603805542,
+      "learning_rate": 0.0002,
+      "loss": 2.2992,
+      "step": 31240
+    },
+    {
+      "epoch": 2.328614008941878,
+      "grad_norm": 2.767536163330078,
+      "learning_rate": 0.0002,
+      "loss": 2.4628,
+      "step": 31250
+    },
+    {
+      "epoch": 2.329359165424739,
+      "grad_norm": 3.01369047164917,
+      "learning_rate": 0.0002,
+      "loss": 2.3962,
+      "step": 31260
+    },
+    {
+      "epoch": 2.3301043219076005,
+      "grad_norm": 2.545297622680664,
+      "learning_rate": 0.0002,
+      "loss": 2.5169,
+      "step": 31270
+    },
+    {
+      "epoch": 2.330849478390462,
+      "grad_norm": 2.610537528991699,
+      "learning_rate": 0.0002,
+      "loss": 2.5468,
+      "step": 31280
+    },
+    {
+      "epoch": 2.3315946348733236,
+      "grad_norm": 2.7675936222076416,
+      "learning_rate": 0.0002,
+      "loss": 2.2562,
+      "step": 31290
+    },
+    {
+      "epoch": 2.3323397913561847,
+      "grad_norm": 2.5579588413238525,
+      "learning_rate": 0.0002,
+      "loss": 2.5384,
+      "step": 31300
+    },
+    {
+      "epoch": 2.3330849478390463,
+      "grad_norm": 2.905510902404785,
+      "learning_rate": 0.0002,
+      "loss": 2.3248,
+      "step": 31310
+    },
+    {
+      "epoch": 2.3338301043219074,
+      "grad_norm": 2.6101415157318115,
+      "learning_rate": 0.0002,
+      "loss": 2.5987,
+      "step": 31320
+    },
+    {
+      "epoch": 2.334575260804769,
+      "grad_norm": 2.5008299350738525,
+      "learning_rate": 0.0002,
+      "loss": 2.2796,
+      "step": 31330
+    },
+    {
+      "epoch": 2.3353204172876305,
+      "grad_norm": 2.7054800987243652,
+      "learning_rate": 0.0002,
+      "loss": 2.2972,
+      "step": 31340
+    },
+    {
+      "epoch": 2.3360655737704916,
+      "grad_norm": 2.9480528831481934,
+      "learning_rate": 0.0002,
+      "loss": 2.624,
+      "step": 31350
+    },
+    {
+      "epoch": 2.336810730253353,
+      "grad_norm": 2.9812073707580566,
+      "learning_rate": 0.0002,
+      "loss": 2.5909,
+      "step": 31360
+    },
+    {
+      "epoch": 2.3375558867362147,
+      "grad_norm": 2.557450294494629,
+      "learning_rate": 0.0002,
+      "loss": 2.3268,
+      "step": 31370
+    },
+    {
+      "epoch": 2.338301043219076,
+      "grad_norm": 2.770230770111084,
+      "learning_rate": 0.0002,
+      "loss": 2.4681,
+      "step": 31380
+    },
+    {
+      "epoch": 2.3390461997019374,
+      "grad_norm": 2.71527099609375,
+      "learning_rate": 0.0002,
+      "loss": 2.4039,
+      "step": 31390
+    },
+    {
+      "epoch": 2.339791356184799,
+      "grad_norm": 2.635856866836548,
+      "learning_rate": 0.0002,
+      "loss": 2.4766,
+      "step": 31400
+    },
+    {
+      "epoch": 2.34053651266766,
+      "grad_norm": 2.7687203884124756,
+      "learning_rate": 0.0002,
+      "loss": 2.4795,
+      "step": 31410
+    },
+    {
+      "epoch": 2.3412816691505216,
+      "grad_norm": 2.6635115146636963,
+      "learning_rate": 0.0002,
+      "loss": 2.3895,
+      "step": 31420
+    },
+    {
+      "epoch": 2.342026825633383,
+      "grad_norm": 2.6172454357147217,
+      "learning_rate": 0.0002,
+      "loss": 2.7309,
+      "step": 31430
+    },
+    {
+      "epoch": 2.3427719821162443,
+      "grad_norm": 3.0446231365203857,
+      "learning_rate": 0.0002,
+      "loss": 2.3944,
+      "step": 31440
+    },
+    {
+      "epoch": 2.343517138599106,
+      "grad_norm": 2.6833817958831787,
+      "learning_rate": 0.0002,
+      "loss": 2.412,
+      "step": 31450
+    },
+    {
+      "epoch": 2.3442622950819674,
+      "grad_norm": 2.333240509033203,
+      "learning_rate": 0.0002,
+      "loss": 2.5243,
+      "step": 31460
+    },
+    {
+      "epoch": 2.3450074515648285,
+      "grad_norm": 2.985318422317505,
+      "learning_rate": 0.0002,
+      "loss": 2.4594,
+      "step": 31470
+    },
+    {
+      "epoch": 2.34575260804769,
+      "grad_norm": 2.4295361042022705,
+      "learning_rate": 0.0002,
+      "loss": 2.3086,
+      "step": 31480
+    },
+    {
+      "epoch": 2.3464977645305516,
+      "grad_norm": 2.663954734802246,
+      "learning_rate": 0.0002,
+      "loss": 2.5138,
+      "step": 31490
+    },
+    {
+      "epoch": 2.3472429210134127,
+      "grad_norm": 2.842383861541748,
+      "learning_rate": 0.0002,
+      "loss": 2.649,
+      "step": 31500
+    },
+    {
+      "epoch": 2.3479880774962743,
+      "grad_norm": 2.4584267139434814,
+      "learning_rate": 0.0002,
+      "loss": 2.2672,
+      "step": 31510
+    },
+    {
+      "epoch": 2.348733233979136,
+      "grad_norm": 2.6671886444091797,
+      "learning_rate": 0.0002,
+      "loss": 2.3535,
+      "step": 31520
+    },
+    {
+      "epoch": 2.349478390461997,
+      "grad_norm": 2.6735708713531494,
+      "learning_rate": 0.0002,
+      "loss": 2.4926,
+      "step": 31530
+    },
+    {
+      "epoch": 2.3502235469448585,
+      "grad_norm": 2.579631805419922,
+      "learning_rate": 0.0002,
+      "loss": 2.1678,
+      "step": 31540
+    },
+    {
+      "epoch": 2.3509687034277196,
+      "grad_norm": 2.5396575927734375,
+      "learning_rate": 0.0002,
+      "loss": 2.4117,
+      "step": 31550
+    },
+    {
+      "epoch": 2.351713859910581,
+      "grad_norm": 2.7359862327575684,
+      "learning_rate": 0.0002,
+      "loss": 2.459,
+      "step": 31560
+    },
+    {
+      "epoch": 2.3524590163934427,
+      "grad_norm": 2.8085885047912598,
+      "learning_rate": 0.0002,
+      "loss": 2.3537,
+      "step": 31570
+    },
+    {
+      "epoch": 2.3532041728763042,
+      "grad_norm": 3.020338773727417,
+      "learning_rate": 0.0002,
+      "loss": 2.5794,
+      "step": 31580
+    },
+    {
+      "epoch": 2.3539493293591653,
+      "grad_norm": 2.755391836166382,
+      "learning_rate": 0.0002,
+      "loss": 2.3765,
+      "step": 31590
+    },
+    {
+      "epoch": 2.354694485842027,
+      "grad_norm": 2.8029239177703857,
+      "learning_rate": 0.0002,
+      "loss": 2.3851,
+      "step": 31600
+    },
+    {
+      "epoch": 2.355439642324888,
+      "grad_norm": 2.442831516265869,
+      "learning_rate": 0.0002,
+      "loss": 2.426,
+      "step": 31610
+    },
+    {
+      "epoch": 2.3561847988077496,
+      "grad_norm": 2.6783061027526855,
+      "learning_rate": 0.0002,
+      "loss": 2.5379,
+      "step": 31620
+    },
+    {
+      "epoch": 2.356929955290611,
+      "grad_norm": 2.8545401096343994,
+      "learning_rate": 0.0002,
+      "loss": 2.3918,
+      "step": 31630
+    },
+    {
+      "epoch": 2.3576751117734727,
+      "grad_norm": 2.6297447681427,
+      "learning_rate": 0.0002,
+      "loss": 2.5717,
+      "step": 31640
+    },
+    {
+      "epoch": 2.3584202682563338,
+      "grad_norm": 2.3793184757232666,
+      "learning_rate": 0.0002,
+      "loss": 2.4338,
+      "step": 31650
+    },
+    {
+      "epoch": 2.3591654247391953,
+      "grad_norm": 2.4211175441741943,
+      "learning_rate": 0.0002,
+      "loss": 2.5335,
+      "step": 31660
+    },
+    {
+      "epoch": 2.3599105812220564,
+      "grad_norm": 2.7338030338287354,
+      "learning_rate": 0.0002,
+      "loss": 2.4089,
+      "step": 31670
+    },
+    {
+      "epoch": 2.360655737704918,
+      "grad_norm": 2.5373005867004395,
+      "learning_rate": 0.0002,
+      "loss": 2.5354,
+      "step": 31680
+    },
+    {
+      "epoch": 2.3614008941877795,
+      "grad_norm": 2.932400703430176,
+      "learning_rate": 0.0002,
+      "loss": 2.4069,
+      "step": 31690
+    },
+    {
+      "epoch": 2.3621460506706407,
+      "grad_norm": 2.6089446544647217,
+      "learning_rate": 0.0002,
+      "loss": 2.4368,
+      "step": 31700
+    },
+    {
+      "epoch": 2.362891207153502,
+      "grad_norm": 2.8511080741882324,
+      "learning_rate": 0.0002,
+      "loss": 2.3916,
+      "step": 31710
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 3.1267642974853516,
+      "learning_rate": 0.0002,
+      "loss": 2.5248,
+      "step": 31720
+    },
+    {
+      "epoch": 2.364381520119225,
+      "grad_norm": 2.741081953048706,
+      "learning_rate": 0.0002,
+      "loss": 2.3826,
+      "step": 31730
+    },
+    {
+      "epoch": 2.3651266766020864,
+      "grad_norm": 2.7541043758392334,
+      "learning_rate": 0.0002,
+      "loss": 2.4853,
+      "step": 31740
+    },
+    {
+      "epoch": 2.365871833084948,
+      "grad_norm": 2.6684322357177734,
+      "learning_rate": 0.0002,
+      "loss": 2.2149,
+      "step": 31750
+    },
+    {
+      "epoch": 2.366616989567809,
+      "grad_norm": 2.5884592533111572,
+      "learning_rate": 0.0002,
+      "loss": 2.5272,
+      "step": 31760
+    },
+    {
+      "epoch": 2.3673621460506706,
+      "grad_norm": 2.832197666168213,
+      "learning_rate": 0.0002,
+      "loss": 2.548,
+      "step": 31770
+    },
+    {
+      "epoch": 2.368107302533532,
+      "grad_norm": 2.574288845062256,
+      "learning_rate": 0.0002,
+      "loss": 2.333,
+      "step": 31780
+    },
+    {
+      "epoch": 2.3688524590163933,
+      "grad_norm": 2.654503107070923,
+      "learning_rate": 0.0002,
+      "loss": 2.4165,
+      "step": 31790
+    },
+    {
+      "epoch": 2.369597615499255,
+      "grad_norm": 2.2965807914733887,
+      "learning_rate": 0.0002,
+      "loss": 2.5123,
+      "step": 31800
+    },
+    {
+      "epoch": 2.3703427719821164,
+      "grad_norm": 2.757343292236328,
+      "learning_rate": 0.0002,
+      "loss": 2.5358,
+      "step": 31810
+    },
+    {
+      "epoch": 2.3710879284649775,
+      "grad_norm": 2.404266119003296,
+      "learning_rate": 0.0002,
+      "loss": 2.4523,
+      "step": 31820
+    },
+    {
+      "epoch": 2.371833084947839,
+      "grad_norm": 2.5798537731170654,
+      "learning_rate": 0.0002,
+      "loss": 2.5129,
+      "step": 31830
+    },
+    {
+      "epoch": 2.3725782414307006,
+      "grad_norm": 2.3345487117767334,
+      "learning_rate": 0.0002,
+      "loss": 2.4274,
+      "step": 31840
+    },
+    {
+      "epoch": 2.3733233979135617,
+      "grad_norm": 2.4548275470733643,
+      "learning_rate": 0.0002,
+      "loss": 2.6288,
+      "step": 31850
+    },
+    {
+      "epoch": 2.3740685543964233,
+      "grad_norm": 2.830592632293701,
+      "learning_rate": 0.0002,
+      "loss": 2.4632,
+      "step": 31860
+    },
+    {
+      "epoch": 2.374813710879285,
+      "grad_norm": 2.5509464740753174,
+      "learning_rate": 0.0002,
+      "loss": 2.4051,
+      "step": 31870
+    },
+    {
+      "epoch": 2.375558867362146,
+      "grad_norm": 2.498534679412842,
+      "learning_rate": 0.0002,
+      "loss": 2.4469,
+      "step": 31880
+    },
+    {
+      "epoch": 2.3763040238450075,
+      "grad_norm": 2.4056124687194824,
+      "learning_rate": 0.0002,
+      "loss": 2.4079,
+      "step": 31890
+    },
+    {
+      "epoch": 2.3770491803278686,
+      "grad_norm": 2.7303104400634766,
+      "learning_rate": 0.0002,
+      "loss": 2.3214,
+      "step": 31900
+    },
+    {
+      "epoch": 2.37779433681073,
+      "grad_norm": 2.8305602073669434,
+      "learning_rate": 0.0002,
+      "loss": 2.4894,
+      "step": 31910
+    },
+    {
+      "epoch": 2.3785394932935917,
+      "grad_norm": 3.2949180603027344,
+      "learning_rate": 0.0002,
+      "loss": 2.4071,
+      "step": 31920
+    },
+    {
+      "epoch": 2.3792846497764533,
+      "grad_norm": 2.639760971069336,
+      "learning_rate": 0.0002,
+      "loss": 2.5399,
+      "step": 31930
+    },
+    {
+      "epoch": 2.3800298062593144,
+      "grad_norm": 2.7378196716308594,
+      "learning_rate": 0.0002,
+      "loss": 2.4692,
+      "step": 31940
+    },
+    {
+      "epoch": 2.380774962742176,
+      "grad_norm": 2.9000864028930664,
+      "learning_rate": 0.0002,
+      "loss": 2.6328,
+      "step": 31950
+    },
+    {
+      "epoch": 2.381520119225037,
+      "grad_norm": 2.8372485637664795,
+      "learning_rate": 0.0002,
+      "loss": 2.5593,
+      "step": 31960
+    },
+    {
+      "epoch": 2.3822652757078986,
+      "grad_norm": 2.391453504562378,
+      "learning_rate": 0.0002,
+      "loss": 2.4111,
+      "step": 31970
+    },
+    {
+      "epoch": 2.38301043219076,
+      "grad_norm": 2.5468664169311523,
+      "learning_rate": 0.0002,
+      "loss": 2.5173,
+      "step": 31980
+    },
+    {
+      "epoch": 2.3837555886736217,
+      "grad_norm": 2.59735107421875,
+      "learning_rate": 0.0002,
+      "loss": 2.5609,
+      "step": 31990
+    },
+    {
+      "epoch": 2.384500745156483,
+      "grad_norm": 2.841322660446167,
+      "learning_rate": 0.0002,
+      "loss": 2.4528,
+      "step": 32000
+    },
+    {
+      "epoch": 2.3852459016393444,
+      "grad_norm": 2.7470717430114746,
+      "learning_rate": 0.0002,
+      "loss": 2.6136,
+      "step": 32010
+    },
+    {
+      "epoch": 2.3859910581222055,
+      "grad_norm": 3.1498608589172363,
+      "learning_rate": 0.0002,
+      "loss": 2.4515,
+      "step": 32020
+    },
+    {
+      "epoch": 2.386736214605067,
+      "grad_norm": 2.6419332027435303,
+      "learning_rate": 0.0002,
+      "loss": 2.6439,
+      "step": 32030
+    },
+    {
+      "epoch": 2.3874813710879286,
+      "grad_norm": 2.7486846446990967,
+      "learning_rate": 0.0002,
+      "loss": 2.3877,
+      "step": 32040
+    },
+    {
+      "epoch": 2.3882265275707897,
+      "grad_norm": 2.418884515762329,
+      "learning_rate": 0.0002,
+      "loss": 2.3455,
+      "step": 32050
+    },
+    {
+      "epoch": 2.3889716840536512,
+      "grad_norm": 2.8571934700012207,
+      "learning_rate": 0.0002,
+      "loss": 2.5127,
+      "step": 32060
+    },
+    {
+      "epoch": 2.389716840536513,
+      "grad_norm": 2.683866262435913,
+      "learning_rate": 0.0002,
+      "loss": 2.604,
+      "step": 32070
+    },
+    {
+      "epoch": 2.390461997019374,
+      "grad_norm": 2.7434074878692627,
+      "learning_rate": 0.0002,
+      "loss": 2.4338,
+      "step": 32080
+    },
+    {
+      "epoch": 2.3912071535022354,
+      "grad_norm": 2.5270578861236572,
+      "learning_rate": 0.0002,
+      "loss": 2.555,
+      "step": 32090
+    },
+    {
+      "epoch": 2.391952309985097,
+      "grad_norm": 2.526337146759033,
+      "learning_rate": 0.0002,
+      "loss": 2.5349,
+      "step": 32100
+    },
+    {
+      "epoch": 2.392697466467958,
+      "grad_norm": 2.61208176612854,
+      "learning_rate": 0.0002,
+      "loss": 2.634,
+      "step": 32110
+    },
+    {
+      "epoch": 2.3934426229508197,
+      "grad_norm": 2.6599786281585693,
+      "learning_rate": 0.0002,
+      "loss": 2.2547,
+      "step": 32120
+    },
+    {
+      "epoch": 2.394187779433681,
+      "grad_norm": 2.4096555709838867,
+      "learning_rate": 0.0002,
+      "loss": 2.4555,
+      "step": 32130
+    },
+    {
+      "epoch": 2.3949329359165423,
+      "grad_norm": 2.4262518882751465,
+      "learning_rate": 0.0002,
+      "loss": 2.5405,
+      "step": 32140
+    },
+    {
+      "epoch": 2.395678092399404,
+      "grad_norm": 2.738983392715454,
+      "learning_rate": 0.0002,
+      "loss": 2.5729,
+      "step": 32150
+    },
+    {
+      "epoch": 2.3964232488822654,
+      "grad_norm": 2.9784836769104004,
+      "learning_rate": 0.0002,
+      "loss": 2.5058,
+      "step": 32160
+    },
+    {
+      "epoch": 2.3971684053651265,
+      "grad_norm": 2.204904556274414,
+      "learning_rate": 0.0002,
+      "loss": 2.4696,
+      "step": 32170
+    },
+    {
+      "epoch": 2.397913561847988,
+      "grad_norm": 2.905687093734741,
+      "learning_rate": 0.0002,
+      "loss": 2.5,
+      "step": 32180
+    },
+    {
+      "epoch": 2.3986587183308496,
+      "grad_norm": 2.540802240371704,
+      "learning_rate": 0.0002,
+      "loss": 2.7352,
+      "step": 32190
+    },
+    {
+      "epoch": 2.3994038748137108,
+      "grad_norm": 2.4641029834747314,
+      "learning_rate": 0.0002,
+      "loss": 2.2883,
+      "step": 32200
+    },
+    {
+      "epoch": 2.4001490312965723,
+      "grad_norm": 2.787086248397827,
+      "learning_rate": 0.0002,
+      "loss": 2.3687,
+      "step": 32210
+    },
+    {
+      "epoch": 2.400894187779434,
+      "grad_norm": 2.6492245197296143,
+      "learning_rate": 0.0002,
+      "loss": 2.5022,
+      "step": 32220
+    },
+    {
+      "epoch": 2.401639344262295,
+      "grad_norm": 2.8292956352233887,
+      "learning_rate": 0.0002,
+      "loss": 2.5244,
+      "step": 32230
+    },
+    {
+      "epoch": 2.4023845007451565,
+      "grad_norm": 2.6720879077911377,
+      "learning_rate": 0.0002,
+      "loss": 2.3531,
+      "step": 32240
+    },
+    {
+      "epoch": 2.4031296572280176,
+      "grad_norm": 2.6397554874420166,
+      "learning_rate": 0.0002,
+      "loss": 2.6169,
+      "step": 32250
+    },
+    {
+      "epoch": 2.403874813710879,
+      "grad_norm": 2.494509696960449,
+      "learning_rate": 0.0002,
+      "loss": 2.6833,
+      "step": 32260
+    },
+    {
+      "epoch": 2.4046199701937407,
+      "grad_norm": 2.6733009815216064,
+      "learning_rate": 0.0002,
+      "loss": 2.5063,
+      "step": 32270
+    },
+    {
+      "epoch": 2.4053651266766023,
+      "grad_norm": 2.522325277328491,
+      "learning_rate": 0.0002,
+      "loss": 2.4202,
+      "step": 32280
+    },
+    {
+      "epoch": 2.4061102831594634,
+      "grad_norm": 2.934260845184326,
+      "learning_rate": 0.0002,
+      "loss": 2.6828,
+      "step": 32290
+    },
+    {
+      "epoch": 2.406855439642325,
+      "grad_norm": 2.7040140628814697,
+      "learning_rate": 0.0002,
+      "loss": 2.4384,
+      "step": 32300
+    },
+    {
+      "epoch": 2.407600596125186,
+      "grad_norm": 2.9310309886932373,
+      "learning_rate": 0.0002,
+      "loss": 2.5099,
+      "step": 32310
+    },
+    {
+      "epoch": 2.4083457526080476,
+      "grad_norm": 2.7560479640960693,
+      "learning_rate": 0.0002,
+      "loss": 2.4807,
+      "step": 32320
+    },
+    {
+      "epoch": 2.409090909090909,
+      "grad_norm": 2.6352450847625732,
+      "learning_rate": 0.0002,
+      "loss": 2.5188,
+      "step": 32330
+    },
+    {
+      "epoch": 2.4098360655737707,
+      "grad_norm": 2.613154888153076,
+      "learning_rate": 0.0002,
+      "loss": 2.6614,
+      "step": 32340
+    },
+    {
+      "epoch": 2.410581222056632,
+      "grad_norm": 2.5213232040405273,
+      "learning_rate": 0.0002,
+      "loss": 2.6163,
+      "step": 32350
+    },
+    {
+      "epoch": 2.4113263785394934,
+      "grad_norm": 2.757983684539795,
+      "learning_rate": 0.0002,
+      "loss": 2.4484,
+      "step": 32360
+    },
+    {
+      "epoch": 2.4120715350223545,
+      "grad_norm": 2.594177007675171,
+      "learning_rate": 0.0002,
+      "loss": 2.4938,
+      "step": 32370
+    },
+    {
+      "epoch": 2.412816691505216,
+      "grad_norm": 2.2750751972198486,
+      "learning_rate": 0.0002,
+      "loss": 2.3242,
+      "step": 32380
+    },
+    {
+      "epoch": 2.4135618479880776,
+      "grad_norm": 3.5427920818328857,
+      "learning_rate": 0.0002,
+      "loss": 2.6142,
+      "step": 32390
+    },
+    {
+      "epoch": 2.4143070044709387,
+      "grad_norm": 2.4356930255889893,
+      "learning_rate": 0.0002,
+      "loss": 2.2886,
+      "step": 32400
+    },
+    {
+      "epoch": 2.4150521609538003,
+      "grad_norm": 3.0683324337005615,
+      "learning_rate": 0.0002,
+      "loss": 2.3534,
+      "step": 32410
+    },
+    {
+      "epoch": 2.415797317436662,
+      "grad_norm": 2.464430809020996,
+      "learning_rate": 0.0002,
+      "loss": 2.3722,
+      "step": 32420
+    },
+    {
+      "epoch": 2.416542473919523,
+      "grad_norm": 2.803317070007324,
+      "learning_rate": 0.0002,
+      "loss": 2.4745,
+      "step": 32430
+    },
+    {
+      "epoch": 2.4172876304023845,
+      "grad_norm": 2.4788084030151367,
+      "learning_rate": 0.0002,
+      "loss": 2.4501,
+      "step": 32440
+    },
+    {
+      "epoch": 2.418032786885246,
+      "grad_norm": 2.686638832092285,
+      "learning_rate": 0.0002,
+      "loss": 2.5698,
+      "step": 32450
+    },
+    {
+      "epoch": 2.418777943368107,
+      "grad_norm": 3.0381052494049072,
+      "learning_rate": 0.0002,
+      "loss": 2.5183,
+      "step": 32460
+    },
+    {
+      "epoch": 2.4195230998509687,
+      "grad_norm": 2.5471017360687256,
+      "learning_rate": 0.0002,
+      "loss": 2.5867,
+      "step": 32470
+    },
+    {
+      "epoch": 2.4202682563338302,
+      "grad_norm": 2.2520577907562256,
+      "learning_rate": 0.0002,
+      "loss": 2.3733,
+      "step": 32480
+    },
+    {
+      "epoch": 2.4210134128166914,
+      "grad_norm": 2.5006518363952637,
+      "learning_rate": 0.0002,
+      "loss": 2.5879,
+      "step": 32490
+    },
+    {
+      "epoch": 2.421758569299553,
+      "grad_norm": 2.8727335929870605,
+      "learning_rate": 0.0002,
+      "loss": 2.489,
+      "step": 32500
+    },
+    {
+      "epoch": 2.4225037257824145,
+      "grad_norm": 2.567643165588379,
+      "learning_rate": 0.0002,
+      "loss": 2.4855,
+      "step": 32510
+    },
+    {
+      "epoch": 2.4232488822652756,
+      "grad_norm": 2.561478614807129,
+      "learning_rate": 0.0002,
+      "loss": 2.4656,
+      "step": 32520
+    },
+    {
+      "epoch": 2.423994038748137,
+      "grad_norm": 2.8764541149139404,
+      "learning_rate": 0.0002,
+      "loss": 2.2183,
+      "step": 32530
+    },
+    {
+      "epoch": 2.4247391952309987,
+      "grad_norm": 2.419424057006836,
+      "learning_rate": 0.0002,
+      "loss": 2.3354,
+      "step": 32540
+    },
+    {
+      "epoch": 2.42548435171386,
+      "grad_norm": 2.3081166744232178,
+      "learning_rate": 0.0002,
+      "loss": 2.3535,
+      "step": 32550
+    },
+    {
+      "epoch": 2.4262295081967213,
+      "grad_norm": 2.464271068572998,
+      "learning_rate": 0.0002,
+      "loss": 2.437,
+      "step": 32560
+    },
+    {
+      "epoch": 2.426974664679583,
+      "grad_norm": 2.919722080230713,
+      "learning_rate": 0.0002,
+      "loss": 2.6343,
+      "step": 32570
+    },
+    {
+      "epoch": 2.427719821162444,
+      "grad_norm": 2.682070016860962,
+      "learning_rate": 0.0002,
+      "loss": 2.453,
+      "step": 32580
+    },
+    {
+      "epoch": 2.4284649776453056,
+      "grad_norm": 2.3702690601348877,
+      "learning_rate": 0.0002,
+      "loss": 2.3294,
+      "step": 32590
+    },
+    {
+      "epoch": 2.429210134128167,
+      "grad_norm": 2.7352712154388428,
+      "learning_rate": 0.0002,
+      "loss": 2.6568,
+      "step": 32600
+    },
+    {
+      "epoch": 2.429955290611028,
+      "grad_norm": 2.396092653274536,
+      "learning_rate": 0.0002,
+      "loss": 2.5299,
+      "step": 32610
+    },
+    {
+      "epoch": 2.4307004470938898,
+      "grad_norm": 2.298727512359619,
+      "learning_rate": 0.0002,
+      "loss": 2.4894,
+      "step": 32620
+    },
+    {
+      "epoch": 2.4314456035767513,
+      "grad_norm": 2.5034990310668945,
+      "learning_rate": 0.0002,
+      "loss": 2.5429,
+      "step": 32630
+    },
+    {
+      "epoch": 2.4321907600596124,
+      "grad_norm": 2.662663221359253,
+      "learning_rate": 0.0002,
+      "loss": 2.5597,
+      "step": 32640
+    },
+    {
+      "epoch": 2.432935916542474,
+      "grad_norm": 2.516818046569824,
+      "learning_rate": 0.0002,
+      "loss": 2.4248,
+      "step": 32650
+    },
+    {
+      "epoch": 2.433681073025335,
+      "grad_norm": 2.416969060897827,
+      "learning_rate": 0.0002,
+      "loss": 2.5677,
+      "step": 32660
+    },
+    {
+      "epoch": 2.4344262295081966,
+      "grad_norm": 2.575011968612671,
+      "learning_rate": 0.0002,
+      "loss": 2.5937,
+      "step": 32670
+    },
+    {
+      "epoch": 2.435171385991058,
+      "grad_norm": 2.692478895187378,
+      "learning_rate": 0.0002,
+      "loss": 2.5584,
+      "step": 32680
+    },
+    {
+      "epoch": 2.4359165424739198,
+      "grad_norm": 3.400285243988037,
+      "learning_rate": 0.0002,
+      "loss": 2.415,
+      "step": 32690
+    },
+    {
+      "epoch": 2.436661698956781,
+      "grad_norm": 3.0482239723205566,
+      "learning_rate": 0.0002,
+      "loss": 2.6148,
+      "step": 32700
+    },
+    {
+      "epoch": 2.4374068554396424,
+      "grad_norm": 2.66845965385437,
+      "learning_rate": 0.0002,
+      "loss": 2.4034,
+      "step": 32710
+    },
+    {
+      "epoch": 2.4381520119225035,
+      "grad_norm": 2.6935739517211914,
+      "learning_rate": 0.0002,
+      "loss": 2.1729,
+      "step": 32720
+    },
+    {
+      "epoch": 2.438897168405365,
+      "grad_norm": 2.8061110973358154,
+      "learning_rate": 0.0002,
+      "loss": 2.3881,
+      "step": 32730
+    },
+    {
+      "epoch": 2.4396423248882266,
+      "grad_norm": 2.6738741397857666,
+      "learning_rate": 0.0002,
+      "loss": 2.4427,
+      "step": 32740
+    },
+    {
+      "epoch": 2.4403874813710877,
+      "grad_norm": 2.7789554595947266,
+      "learning_rate": 0.0002,
+      "loss": 2.3962,
+      "step": 32750
+    },
+    {
+      "epoch": 2.4411326378539493,
+      "grad_norm": 2.6854817867279053,
+      "learning_rate": 0.0002,
+      "loss": 2.4369,
+      "step": 32760
+    },
+    {
+      "epoch": 2.441877794336811,
+      "grad_norm": 2.570359706878662,
+      "learning_rate": 0.0002,
+      "loss": 2.4015,
+      "step": 32770
+    },
+    {
+      "epoch": 2.442622950819672,
+      "grad_norm": 2.7053561210632324,
+      "learning_rate": 0.0002,
+      "loss": 2.4896,
+      "step": 32780
+    },
+    {
+      "epoch": 2.4433681073025335,
+      "grad_norm": 2.5930299758911133,
+      "learning_rate": 0.0002,
+      "loss": 2.4455,
+      "step": 32790
+    },
+    {
+      "epoch": 2.444113263785395,
+      "grad_norm": 2.827674627304077,
+      "learning_rate": 0.0002,
+      "loss": 2.4475,
+      "step": 32800
+    },
+    {
+      "epoch": 2.444858420268256,
+      "grad_norm": 2.686424732208252,
+      "learning_rate": 0.0002,
+      "loss": 2.4927,
+      "step": 32810
+    },
+    {
+      "epoch": 2.4456035767511177,
+      "grad_norm": 2.965540647506714,
+      "learning_rate": 0.0002,
+      "loss": 2.673,
+      "step": 32820
+    },
+    {
+      "epoch": 2.4463487332339793,
+      "grad_norm": 2.8941571712493896,
+      "learning_rate": 0.0002,
+      "loss": 2.5195,
+      "step": 32830
+    },
+    {
+      "epoch": 2.4470938897168404,
+      "grad_norm": 3.103050947189331,
+      "learning_rate": 0.0002,
+      "loss": 2.5864,
+      "step": 32840
+    },
+    {
+      "epoch": 2.447839046199702,
+      "grad_norm": 2.7334511280059814,
+      "learning_rate": 0.0002,
+      "loss": 2.4724,
+      "step": 32850
+    },
+    {
+      "epoch": 2.4485842026825635,
+      "grad_norm": 2.449275493621826,
+      "learning_rate": 0.0002,
+      "loss": 2.5776,
+      "step": 32860
+    },
+    {
+      "epoch": 2.4493293591654246,
+      "grad_norm": 2.5807056427001953,
+      "learning_rate": 0.0002,
+      "loss": 2.536,
+      "step": 32870
+    },
+    {
+      "epoch": 2.450074515648286,
+      "grad_norm": 2.7957377433776855,
+      "learning_rate": 0.0002,
+      "loss": 2.6551,
+      "step": 32880
+    },
+    {
+      "epoch": 2.4508196721311477,
+      "grad_norm": 2.858997106552124,
+      "learning_rate": 0.0002,
+      "loss": 2.3838,
+      "step": 32890
+    },
+    {
+      "epoch": 2.451564828614009,
+      "grad_norm": 2.8242721557617188,
+      "learning_rate": 0.0002,
+      "loss": 2.3843,
+      "step": 32900
+    },
+    {
+      "epoch": 2.4523099850968704,
+      "grad_norm": 2.914250612258911,
+      "learning_rate": 0.0002,
+      "loss": 2.3497,
+      "step": 32910
+    },
+    {
+      "epoch": 2.453055141579732,
+      "grad_norm": 2.325657606124878,
+      "learning_rate": 0.0002,
+      "loss": 2.5284,
+      "step": 32920
+    },
+    {
+      "epoch": 2.453800298062593,
+      "grad_norm": 2.4877982139587402,
+      "learning_rate": 0.0002,
+      "loss": 2.5986,
+      "step": 32930
+    },
+    {
+      "epoch": 2.4545454545454546,
+      "grad_norm": 2.6177945137023926,
+      "learning_rate": 0.0002,
+      "loss": 2.5368,
+      "step": 32940
+    },
+    {
+      "epoch": 2.455290611028316,
+      "grad_norm": 2.1501779556274414,
+      "learning_rate": 0.0002,
+      "loss": 2.3093,
+      "step": 32950
+    },
+    {
+      "epoch": 2.4560357675111772,
+      "grad_norm": 2.546349048614502,
+      "learning_rate": 0.0002,
+      "loss": 2.3231,
+      "step": 32960
+    },
+    {
+      "epoch": 2.456780923994039,
+      "grad_norm": 2.3199493885040283,
+      "learning_rate": 0.0002,
+      "loss": 2.5672,
+      "step": 32970
+    },
+    {
+      "epoch": 2.4575260804769004,
+      "grad_norm": 2.532883405685425,
+      "learning_rate": 0.0002,
+      "loss": 2.2814,
+      "step": 32980
+    },
+    {
+      "epoch": 2.4582712369597615,
+      "grad_norm": 2.9254093170166016,
+      "learning_rate": 0.0002,
+      "loss": 2.5414,
+      "step": 32990
+    },
+    {
+      "epoch": 2.459016393442623,
+      "grad_norm": 2.5555853843688965,
+      "learning_rate": 0.0002,
+      "loss": 2.5877,
+      "step": 33000
+    },
+    {
+      "epoch": 2.459761549925484,
+      "grad_norm": 2.520555019378662,
+      "learning_rate": 0.0002,
+      "loss": 2.4599,
+      "step": 33010
+    },
+    {
+      "epoch": 2.4605067064083457,
+      "grad_norm": 2.652587413787842,
+      "learning_rate": 0.0002,
+      "loss": 2.5855,
+      "step": 33020
+    },
+    {
+      "epoch": 2.4612518628912072,
+      "grad_norm": 2.7828288078308105,
+      "learning_rate": 0.0002,
+      "loss": 2.5077,
+      "step": 33030
+    },
+    {
+      "epoch": 2.461997019374069,
+      "grad_norm": 2.4250879287719727,
+      "learning_rate": 0.0002,
+      "loss": 2.6537,
+      "step": 33040
+    },
+    {
+      "epoch": 2.46274217585693,
+      "grad_norm": 2.376924753189087,
+      "learning_rate": 0.0002,
+      "loss": 2.5013,
+      "step": 33050
+    },
+    {
+      "epoch": 2.4634873323397914,
+      "grad_norm": 2.6665396690368652,
+      "learning_rate": 0.0002,
+      "loss": 2.452,
+      "step": 33060
+    },
+    {
+      "epoch": 2.4642324888226526,
+      "grad_norm": 2.920982599258423,
+      "learning_rate": 0.0002,
+      "loss": 2.4782,
+      "step": 33070
+    },
+    {
+      "epoch": 2.464977645305514,
+      "grad_norm": 1.9443694353103638,
+      "learning_rate": 0.0002,
+      "loss": 2.2695,
+      "step": 33080
+    },
+    {
+      "epoch": 2.4657228017883757,
+      "grad_norm": 2.8950319290161133,
+      "learning_rate": 0.0002,
+      "loss": 2.6325,
+      "step": 33090
+    },
+    {
+      "epoch": 2.4664679582712368,
+      "grad_norm": 3.0620553493499756,
+      "learning_rate": 0.0002,
+      "loss": 2.3013,
+      "step": 33100
+    },
+    {
+      "epoch": 2.4672131147540983,
+      "grad_norm": 2.653696060180664,
+      "learning_rate": 0.0002,
+      "loss": 2.5691,
+      "step": 33110
+    },
+    {
+      "epoch": 2.46795827123696,
+      "grad_norm": 2.347299337387085,
+      "learning_rate": 0.0002,
+      "loss": 2.2995,
+      "step": 33120
+    },
+    {
+      "epoch": 2.468703427719821,
+      "grad_norm": 2.8021976947784424,
+      "learning_rate": 0.0002,
+      "loss": 2.6018,
+      "step": 33130
+    },
+    {
+      "epoch": 2.4694485842026825,
+      "grad_norm": 1.9987612962722778,
+      "learning_rate": 0.0002,
+      "loss": 2.4832,
+      "step": 33140
+    },
+    {
+      "epoch": 2.470193740685544,
+      "grad_norm": 2.7334144115448,
+      "learning_rate": 0.0002,
+      "loss": 2.4264,
+      "step": 33150
+    },
+    {
+      "epoch": 2.470938897168405,
+      "grad_norm": 2.6771295070648193,
+      "learning_rate": 0.0002,
+      "loss": 2.5609,
+      "step": 33160
+    },
+    {
+      "epoch": 2.4716840536512668,
+      "grad_norm": 2.628952741622925,
+      "learning_rate": 0.0002,
+      "loss": 2.5076,
+      "step": 33170
+    },
+    {
+      "epoch": 2.4724292101341283,
+      "grad_norm": 2.5400376319885254,
+      "learning_rate": 0.0002,
+      "loss": 2.6434,
+      "step": 33180
+    },
+    {
+      "epoch": 2.4731743666169894,
+      "grad_norm": 2.332078695297241,
+      "learning_rate": 0.0002,
+      "loss": 2.4067,
+      "step": 33190
+    },
+    {
+      "epoch": 2.473919523099851,
+      "grad_norm": 2.7730212211608887,
+      "learning_rate": 0.0002,
+      "loss": 2.4936,
+      "step": 33200
+    },
+    {
+      "epoch": 2.4746646795827125,
+      "grad_norm": 2.5415468215942383,
+      "learning_rate": 0.0002,
+      "loss": 2.6113,
+      "step": 33210
+    },
+    {
+      "epoch": 2.4754098360655736,
+      "grad_norm": 2.6376805305480957,
+      "learning_rate": 0.0002,
+      "loss": 2.4445,
+      "step": 33220
+    },
+    {
+      "epoch": 2.476154992548435,
+      "grad_norm": 2.808635711669922,
+      "learning_rate": 0.0002,
+      "loss": 2.4696,
+      "step": 33230
+    },
+    {
+      "epoch": 2.4769001490312967,
+      "grad_norm": 2.3896541595458984,
+      "learning_rate": 0.0002,
+      "loss": 2.316,
+      "step": 33240
+    },
+    {
+      "epoch": 2.477645305514158,
+      "grad_norm": 2.628732919692993,
+      "learning_rate": 0.0002,
+      "loss": 2.661,
+      "step": 33250
+    },
+    {
+      "epoch": 2.4783904619970194,
+      "grad_norm": 2.627577543258667,
+      "learning_rate": 0.0002,
+      "loss": 2.5131,
+      "step": 33260
+    },
+    {
+      "epoch": 2.479135618479881,
+      "grad_norm": 2.3876442909240723,
+      "learning_rate": 0.0002,
+      "loss": 2.3684,
+      "step": 33270
+    },
+    {
+      "epoch": 2.479880774962742,
+      "grad_norm": 2.680107355117798,
+      "learning_rate": 0.0002,
+      "loss": 2.3554,
+      "step": 33280
+    },
+    {
+      "epoch": 2.4806259314456036,
+      "grad_norm": 2.5032846927642822,
+      "learning_rate": 0.0002,
+      "loss": 2.4945,
+      "step": 33290
+    },
+    {
+      "epoch": 2.481371087928465,
+      "grad_norm": 2.6249349117279053,
+      "learning_rate": 0.0002,
+      "loss": 2.574,
+      "step": 33300
+    },
+    {
+      "epoch": 2.4821162444113263,
+      "grad_norm": 2.3198981285095215,
+      "learning_rate": 0.0002,
+      "loss": 2.2761,
+      "step": 33310
+    },
+    {
+      "epoch": 2.482861400894188,
+      "grad_norm": 2.745832920074463,
+      "learning_rate": 0.0002,
+      "loss": 2.2473,
+      "step": 33320
+    },
+    {
+      "epoch": 2.4836065573770494,
+      "grad_norm": 2.7251551151275635,
+      "learning_rate": 0.0002,
+      "loss": 2.5399,
+      "step": 33330
+    },
+    {
+      "epoch": 2.4843517138599105,
+      "grad_norm": 2.507437229156494,
+      "learning_rate": 0.0002,
+      "loss": 2.4234,
+      "step": 33340
+    },
+    {
+      "epoch": 2.485096870342772,
+      "grad_norm": 2.7611029148101807,
+      "learning_rate": 0.0002,
+      "loss": 2.4828,
+      "step": 33350
+    },
+    {
+      "epoch": 2.485842026825633,
+      "grad_norm": 2.7167751789093018,
+      "learning_rate": 0.0002,
+      "loss": 2.3991,
+      "step": 33360
+    },
+    {
+      "epoch": 2.4865871833084947,
+      "grad_norm": 2.4960172176361084,
+      "learning_rate": 0.0002,
+      "loss": 2.3095,
+      "step": 33370
+    },
+    {
+      "epoch": 2.4873323397913563,
+      "grad_norm": 2.461411237716675,
+      "learning_rate": 0.0002,
+      "loss": 2.4042,
+      "step": 33380
+    },
+    {
+      "epoch": 2.488077496274218,
+      "grad_norm": 2.855609655380249,
+      "learning_rate": 0.0002,
+      "loss": 2.6609,
+      "step": 33390
+    },
+    {
+      "epoch": 2.488822652757079,
+      "grad_norm": 2.677143096923828,
+      "learning_rate": 0.0002,
+      "loss": 2.4067,
+      "step": 33400
+    },
+    {
+      "epoch": 2.4895678092399405,
+      "grad_norm": 2.7277297973632812,
+      "learning_rate": 0.0002,
+      "loss": 2.4223,
+      "step": 33410
+    },
+    {
+      "epoch": 2.4903129657228016,
+      "grad_norm": 2.5944976806640625,
+      "learning_rate": 0.0002,
+      "loss": 2.622,
+      "step": 33420
+    },
+    {
+      "epoch": 2.491058122205663,
+      "grad_norm": 2.656810998916626,
+      "learning_rate": 0.0002,
+      "loss": 2.5017,
+      "step": 33430
+    },
+    {
+      "epoch": 2.4918032786885247,
+      "grad_norm": 2.643364429473877,
+      "learning_rate": 0.0002,
+      "loss": 2.5823,
+      "step": 33440
+    },
+    {
+      "epoch": 2.492548435171386,
+      "grad_norm": 2.4666240215301514,
+      "learning_rate": 0.0002,
+      "loss": 2.4402,
+      "step": 33450
+    },
+    {
+      "epoch": 2.4932935916542474,
+      "grad_norm": 2.453465700149536,
+      "learning_rate": 0.0002,
+      "loss": 2.6123,
+      "step": 33460
+    },
+    {
+      "epoch": 2.494038748137109,
+      "grad_norm": 2.574659824371338,
+      "learning_rate": 0.0002,
+      "loss": 2.642,
+      "step": 33470
+    },
+    {
+      "epoch": 2.49478390461997,
+      "grad_norm": 2.6516501903533936,
+      "learning_rate": 0.0002,
+      "loss": 2.3554,
+      "step": 33480
+    },
+    {
+      "epoch": 2.4955290611028316,
+      "grad_norm": 2.5587661266326904,
+      "learning_rate": 0.0002,
+      "loss": 2.3586,
+      "step": 33490
+    },
+    {
+      "epoch": 2.496274217585693,
+      "grad_norm": 2.8648366928100586,
+      "learning_rate": 0.0002,
+      "loss": 2.392,
+      "step": 33500
+    },
+    {
+      "epoch": 2.4970193740685542,
+      "grad_norm": 2.6029820442199707,
+      "learning_rate": 0.0002,
+      "loss": 2.4047,
+      "step": 33510
+    },
+    {
+      "epoch": 2.497764530551416,
+      "grad_norm": 3.100987672805786,
+      "learning_rate": 0.0002,
+      "loss": 2.4915,
+      "step": 33520
+    },
+    {
+      "epoch": 2.4985096870342773,
+      "grad_norm": 2.6375701427459717,
+      "learning_rate": 0.0002,
+      "loss": 2.3645,
+      "step": 33530
+    },
+    {
+      "epoch": 2.4992548435171384,
+      "grad_norm": 2.5169084072113037,
+      "learning_rate": 0.0002,
+      "loss": 2.3556,
+      "step": 33540
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.6907997131347656,
+      "learning_rate": 0.0002,
+      "loss": 2.4695,
+      "step": 33550
+    },
+    {
+      "epoch": 2.5007451564828616,
+      "grad_norm": 2.838801622390747,
+      "learning_rate": 0.0002,
+      "loss": 2.6255,
+      "step": 33560
+    },
+    {
+      "epoch": 2.5014903129657227,
+      "grad_norm": 2.6259567737579346,
+      "learning_rate": 0.0002,
+      "loss": 2.232,
+      "step": 33570
+    },
+    {
+      "epoch": 2.502235469448584,
+      "grad_norm": 2.487612247467041,
+      "learning_rate": 0.0002,
+      "loss": 2.4681,
+      "step": 33580
+    },
+    {
+      "epoch": 2.5029806259314458,
+      "grad_norm": 2.7123425006866455,
+      "learning_rate": 0.0002,
+      "loss": 2.4245,
+      "step": 33590
+    },
+    {
+      "epoch": 2.503725782414307,
+      "grad_norm": 2.5011355876922607,
+      "learning_rate": 0.0002,
+      "loss": 2.3437,
+      "step": 33600
+    },
+    {
+      "epoch": 2.5044709388971684,
+      "grad_norm": 2.726006507873535,
+      "learning_rate": 0.0002,
+      "loss": 2.5336,
+      "step": 33610
+    },
+    {
+      "epoch": 2.50521609538003,
+      "grad_norm": 2.6538584232330322,
+      "learning_rate": 0.0002,
+      "loss": 2.5752,
+      "step": 33620
+    },
+    {
+      "epoch": 2.505961251862891,
+      "grad_norm": 2.8791120052337646,
+      "learning_rate": 0.0002,
+      "loss": 2.5624,
+      "step": 33630
+    },
+    {
+      "epoch": 2.5067064083457526,
+      "grad_norm": 2.7906670570373535,
+      "learning_rate": 0.0002,
+      "loss": 2.566,
+      "step": 33640
+    },
+    {
+      "epoch": 2.5074515648286138,
+      "grad_norm": 2.980694532394409,
+      "learning_rate": 0.0002,
+      "loss": 2.6706,
+      "step": 33650
+    },
+    {
+      "epoch": 2.5081967213114753,
+      "grad_norm": 2.641268014907837,
+      "learning_rate": 0.0002,
+      "loss": 2.4101,
+      "step": 33660
+    },
+    {
+      "epoch": 2.508941877794337,
+      "grad_norm": 2.835050344467163,
+      "learning_rate": 0.0002,
+      "loss": 2.4747,
+      "step": 33670
+    },
+    {
+      "epoch": 2.5096870342771984,
+      "grad_norm": 2.5962984561920166,
+      "learning_rate": 0.0002,
+      "loss": 2.3435,
+      "step": 33680
+    },
+    {
+      "epoch": 2.5104321907600595,
+      "grad_norm": 2.6460578441619873,
+      "learning_rate": 0.0002,
+      "loss": 2.4282,
+      "step": 33690
+    },
+    {
+      "epoch": 2.511177347242921,
+      "grad_norm": 2.415022373199463,
+      "learning_rate": 0.0002,
+      "loss": 2.3002,
+      "step": 33700
+    },
+    {
+      "epoch": 2.511922503725782,
+      "grad_norm": 2.628546953201294,
+      "learning_rate": 0.0002,
+      "loss": 2.6371,
+      "step": 33710
+    },
+    {
+      "epoch": 2.5126676602086437,
+      "grad_norm": 2.697800636291504,
+      "learning_rate": 0.0002,
+      "loss": 2.302,
+      "step": 33720
+    },
+    {
+      "epoch": 2.5134128166915053,
+      "grad_norm": 2.8529772758483887,
+      "learning_rate": 0.0002,
+      "loss": 2.5254,
+      "step": 33730
+    },
+    {
+      "epoch": 2.514157973174367,
+      "grad_norm": 2.5316379070281982,
+      "learning_rate": 0.0002,
+      "loss": 2.4038,
+      "step": 33740
+    },
+    {
+      "epoch": 2.514903129657228,
+      "grad_norm": 2.350818634033203,
+      "learning_rate": 0.0002,
+      "loss": 2.3349,
+      "step": 33750
+    },
+    {
+      "epoch": 2.5156482861400895,
+      "grad_norm": 2.5626444816589355,
+      "learning_rate": 0.0002,
+      "loss": 2.5337,
+      "step": 33760
+    },
+    {
+      "epoch": 2.5163934426229506,
+      "grad_norm": 2.1654140949249268,
+      "learning_rate": 0.0002,
+      "loss": 2.3723,
+      "step": 33770
+    },
+    {
+      "epoch": 2.517138599105812,
+      "grad_norm": 3.0190258026123047,
+      "learning_rate": 0.0002,
+      "loss": 2.5135,
+      "step": 33780
+    },
+    {
+      "epoch": 2.5178837555886737,
+      "grad_norm": 2.6092681884765625,
+      "learning_rate": 0.0002,
+      "loss": 2.4508,
+      "step": 33790
+    },
+    {
+      "epoch": 2.5186289120715353,
+      "grad_norm": 3.0868630409240723,
+      "learning_rate": 0.0002,
+      "loss": 2.4954,
+      "step": 33800
+    },
+    {
+      "epoch": 2.5193740685543964,
+      "grad_norm": 2.437244176864624,
+      "learning_rate": 0.0002,
+      "loss": 2.5313,
+      "step": 33810
+    },
+    {
+      "epoch": 2.520119225037258,
+      "grad_norm": 2.7770519256591797,
+      "learning_rate": 0.0002,
+      "loss": 2.4047,
+      "step": 33820
+    },
+    {
+      "epoch": 2.520864381520119,
+      "grad_norm": 2.540032386779785,
+      "learning_rate": 0.0002,
+      "loss": 2.3,
+      "step": 33830
+    },
+    {
+      "epoch": 2.5216095380029806,
+      "grad_norm": 2.8454859256744385,
+      "learning_rate": 0.0002,
+      "loss": 2.4243,
+      "step": 33840
+    },
+    {
+      "epoch": 2.522354694485842,
+      "grad_norm": 2.4793384075164795,
+      "learning_rate": 0.0002,
+      "loss": 2.441,
+      "step": 33850
+    },
+    {
+      "epoch": 2.5230998509687033,
+      "grad_norm": 2.884309768676758,
+      "learning_rate": 0.0002,
+      "loss": 2.6058,
+      "step": 33860
+    },
+    {
+      "epoch": 2.523845007451565,
+      "grad_norm": 2.188300371170044,
+      "learning_rate": 0.0002,
+      "loss": 2.2676,
+      "step": 33870
+    },
+    {
+      "epoch": 2.5245901639344264,
+      "grad_norm": 2.1281039714813232,
+      "learning_rate": 0.0002,
+      "loss": 2.4825,
+      "step": 33880
+    },
+    {
+      "epoch": 2.5253353204172875,
+      "grad_norm": 2.872929096221924,
+      "learning_rate": 0.0002,
+      "loss": 2.6046,
+      "step": 33890
+    },
+    {
+      "epoch": 2.526080476900149,
+      "grad_norm": 2.7643649578094482,
+      "learning_rate": 0.0002,
+      "loss": 2.5626,
+      "step": 33900
+    },
+    {
+      "epoch": 2.5268256333830106,
+      "grad_norm": 2.42836856842041,
+      "learning_rate": 0.0002,
+      "loss": 2.5225,
+      "step": 33910
+    },
+    {
+      "epoch": 2.5275707898658717,
+      "grad_norm": 2.7934534549713135,
+      "learning_rate": 0.0002,
+      "loss": 2.6089,
+      "step": 33920
+    },
+    {
+      "epoch": 2.5283159463487332,
+      "grad_norm": 2.9219272136688232,
+      "learning_rate": 0.0002,
+      "loss": 2.4704,
+      "step": 33930
+    },
+    {
+      "epoch": 2.529061102831595,
+      "grad_norm": 2.470702648162842,
+      "learning_rate": 0.0002,
+      "loss": 2.5143,
+      "step": 33940
+    },
+    {
+      "epoch": 2.529806259314456,
+      "grad_norm": 2.40101957321167,
+      "learning_rate": 0.0002,
+      "loss": 2.4033,
+      "step": 33950
+    },
+    {
+      "epoch": 2.5305514157973175,
+      "grad_norm": 2.4224092960357666,
+      "learning_rate": 0.0002,
+      "loss": 2.3719,
+      "step": 33960
+    },
+    {
+      "epoch": 2.531296572280179,
+      "grad_norm": 2.8269007205963135,
+      "learning_rate": 0.0002,
+      "loss": 2.6289,
+      "step": 33970
+    },
+    {
+      "epoch": 2.53204172876304,
+      "grad_norm": 2.590324878692627,
+      "learning_rate": 0.0002,
+      "loss": 2.3846,
+      "step": 33980
+    },
+    {
+      "epoch": 2.5327868852459017,
+      "grad_norm": 2.609064817428589,
+      "learning_rate": 0.0002,
+      "loss": 2.5036,
+      "step": 33990
+    },
+    {
+      "epoch": 2.533532041728763,
+      "grad_norm": 2.3954341411590576,
+      "learning_rate": 0.0002,
+      "loss": 2.3507,
+      "step": 34000
+    },
+    {
+      "epoch": 2.5342771982116243,
+      "grad_norm": 2.4784460067749023,
+      "learning_rate": 0.0002,
+      "loss": 2.5389,
+      "step": 34010
+    },
+    {
+      "epoch": 2.535022354694486,
+      "grad_norm": 2.3329291343688965,
+      "learning_rate": 0.0002,
+      "loss": 2.4564,
+      "step": 34020
+    },
+    {
+      "epoch": 2.5357675111773474,
+      "grad_norm": 2.564966917037964,
+      "learning_rate": 0.0002,
+      "loss": 2.5259,
+      "step": 34030
+    },
+    {
+      "epoch": 2.5365126676602086,
+      "grad_norm": 2.7907464504241943,
+      "learning_rate": 0.0002,
+      "loss": 2.5996,
+      "step": 34040
+    },
+    {
+      "epoch": 2.53725782414307,
+      "grad_norm": 2.61496901512146,
+      "learning_rate": 0.0002,
+      "loss": 2.4294,
+      "step": 34050
+    },
+    {
+      "epoch": 2.538002980625931,
+      "grad_norm": 2.5736818313598633,
+      "learning_rate": 0.0002,
+      "loss": 2.2709,
+      "step": 34060
+    },
+    {
+      "epoch": 2.5387481371087928,
+      "grad_norm": 2.3524608612060547,
+      "learning_rate": 0.0002,
+      "loss": 2.2737,
+      "step": 34070
+    },
+    {
+      "epoch": 2.5394932935916543,
+      "grad_norm": 2.7142202854156494,
+      "learning_rate": 0.0002,
+      "loss": 2.4911,
+      "step": 34080
+    },
+    {
+      "epoch": 2.540238450074516,
+      "grad_norm": 2.64286470413208,
+      "learning_rate": 0.0002,
+      "loss": 2.5254,
+      "step": 34090
+    },
+    {
+      "epoch": 2.540983606557377,
+      "grad_norm": 3.114713191986084,
+      "learning_rate": 0.0002,
+      "loss": 2.5674,
+      "step": 34100
+    },
+    {
+      "epoch": 2.5417287630402385,
+      "grad_norm": 3.056455373764038,
+      "learning_rate": 0.0002,
+      "loss": 2.4114,
+      "step": 34110
+    },
+    {
+      "epoch": 2.5424739195230996,
+      "grad_norm": 2.2923319339752197,
+      "learning_rate": 0.0002,
+      "loss": 2.372,
+      "step": 34120
+    },
+    {
+      "epoch": 2.543219076005961,
+      "grad_norm": 3.0862503051757812,
+      "learning_rate": 0.0002,
+      "loss": 2.5038,
+      "step": 34130
+    },
+    {
+      "epoch": 2.5439642324888228,
+      "grad_norm": 2.7312214374542236,
+      "learning_rate": 0.0002,
+      "loss": 2.7273,
+      "step": 34140
+    },
+    {
+      "epoch": 2.5447093889716843,
+      "grad_norm": 2.667236089706421,
+      "learning_rate": 0.0002,
+      "loss": 2.5896,
+      "step": 34150
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 2.5944461822509766,
+      "learning_rate": 0.0002,
+      "loss": 2.6472,
+      "step": 34160
+    },
+    {
+      "epoch": 2.546199701937407,
+      "grad_norm": 2.375399589538574,
+      "learning_rate": 0.0002,
+      "loss": 2.5534,
+      "step": 34170
+    },
+    {
+      "epoch": 2.546944858420268,
+      "grad_norm": 2.3310904502868652,
+      "learning_rate": 0.0002,
+      "loss": 2.3438,
+      "step": 34180
+    },
+    {
+      "epoch": 2.5476900149031296,
+      "grad_norm": 2.3396718502044678,
+      "learning_rate": 0.0002,
+      "loss": 2.629,
+      "step": 34190
+    },
+    {
+      "epoch": 2.548435171385991,
+      "grad_norm": 2.717461109161377,
+      "learning_rate": 0.0002,
+      "loss": 2.5391,
+      "step": 34200
+    },
+    {
+      "epoch": 2.5491803278688527,
+      "grad_norm": 2.5795769691467285,
+      "learning_rate": 0.0002,
+      "loss": 2.4663,
+      "step": 34210
+    },
+    {
+      "epoch": 2.549925484351714,
+      "grad_norm": 2.31074857711792,
+      "learning_rate": 0.0002,
+      "loss": 2.3499,
+      "step": 34220
+    },
+    {
+      "epoch": 2.5506706408345754,
+      "grad_norm": 2.2600789070129395,
+      "learning_rate": 0.0002,
+      "loss": 2.4089,
+      "step": 34230
+    },
+    {
+      "epoch": 2.5514157973174365,
+      "grad_norm": 2.5004208087921143,
+      "learning_rate": 0.0002,
+      "loss": 2.4895,
+      "step": 34240
+    },
+    {
+      "epoch": 2.552160953800298,
+      "grad_norm": 2.4714882373809814,
+      "learning_rate": 0.0002,
+      "loss": 2.3912,
+      "step": 34250
+    },
+    {
+      "epoch": 2.5529061102831596,
+      "grad_norm": 2.5733115673065186,
+      "learning_rate": 0.0002,
+      "loss": 2.4612,
+      "step": 34260
+    },
+    {
+      "epoch": 2.5536512667660207,
+      "grad_norm": 2.537883996963501,
+      "learning_rate": 0.0002,
+      "loss": 2.5025,
+      "step": 34270
+    },
+    {
+      "epoch": 2.5543964232488823,
+      "grad_norm": 2.867398262023926,
+      "learning_rate": 0.0002,
+      "loss": 2.4764,
+      "step": 34280
+    },
+    {
+      "epoch": 2.555141579731744,
+      "grad_norm": 2.3895552158355713,
+      "learning_rate": 0.0002,
+      "loss": 2.6755,
+      "step": 34290
+    },
+    {
+      "epoch": 2.555886736214605,
+      "grad_norm": 2.6244559288024902,
+      "learning_rate": 0.0002,
+      "loss": 2.5904,
+      "step": 34300
+    },
+    {
+      "epoch": 2.5566318926974665,
+      "grad_norm": 3.0970191955566406,
+      "learning_rate": 0.0002,
+      "loss": 2.4423,
+      "step": 34310
+    },
+    {
+      "epoch": 2.557377049180328,
+      "grad_norm": 2.543386459350586,
+      "learning_rate": 0.0002,
+      "loss": 2.4975,
+      "step": 34320
+    },
+    {
+      "epoch": 2.558122205663189,
+      "grad_norm": 2.6352686882019043,
+      "learning_rate": 0.0002,
+      "loss": 2.5865,
+      "step": 34330
+    },
+    {
+      "epoch": 2.5588673621460507,
+      "grad_norm": 3.1985583305358887,
+      "learning_rate": 0.0002,
+      "loss": 2.6501,
+      "step": 34340
+    },
+    {
+      "epoch": 2.559612518628912,
+      "grad_norm": 2.206559896469116,
+      "learning_rate": 0.0002,
+      "loss": 2.5045,
+      "step": 34350
+    },
+    {
+      "epoch": 2.5603576751117734,
+      "grad_norm": 2.626335620880127,
+      "learning_rate": 0.0002,
+      "loss": 2.4685,
+      "step": 34360
+    },
+    {
+      "epoch": 2.561102831594635,
+      "grad_norm": 2.7654409408569336,
+      "learning_rate": 0.0002,
+      "loss": 2.2565,
+      "step": 34370
+    },
+    {
+      "epoch": 2.5618479880774965,
+      "grad_norm": 2.5158398151397705,
+      "learning_rate": 0.0002,
+      "loss": 2.5521,
+      "step": 34380
+    },
+    {
+      "epoch": 2.5625931445603576,
+      "grad_norm": 3.0391643047332764,
+      "learning_rate": 0.0002,
+      "loss": 2.5027,
+      "step": 34390
+    },
+    {
+      "epoch": 2.563338301043219,
+      "grad_norm": 2.9316203594207764,
+      "learning_rate": 0.0002,
+      "loss": 2.3832,
+      "step": 34400
+    },
+    {
+      "epoch": 2.5640834575260802,
+      "grad_norm": 2.54579496383667,
+      "learning_rate": 0.0002,
+      "loss": 2.4208,
+      "step": 34410
+    },
+    {
+      "epoch": 2.564828614008942,
+      "grad_norm": 2.8894309997558594,
+      "learning_rate": 0.0002,
+      "loss": 2.4221,
+      "step": 34420
+    },
+    {
+      "epoch": 2.5655737704918034,
+      "grad_norm": 2.742469072341919,
+      "learning_rate": 0.0002,
+      "loss": 2.5528,
+      "step": 34430
+    },
+    {
+      "epoch": 2.566318926974665,
+      "grad_norm": 2.7238597869873047,
+      "learning_rate": 0.0002,
+      "loss": 2.5734,
+      "step": 34440
+    },
+    {
+      "epoch": 2.567064083457526,
+      "grad_norm": 2.892765998840332,
+      "learning_rate": 0.0002,
+      "loss": 2.3945,
+      "step": 34450
+    },
+    {
+      "epoch": 2.5678092399403876,
+      "grad_norm": 2.4451138973236084,
+      "learning_rate": 0.0002,
+      "loss": 2.5518,
+      "step": 34460
+    },
+    {
+      "epoch": 2.5685543964232487,
+      "grad_norm": 2.573063850402832,
+      "learning_rate": 0.0002,
+      "loss": 2.4919,
+      "step": 34470
+    },
+    {
+      "epoch": 2.5692995529061102,
+      "grad_norm": 2.7531449794769287,
+      "learning_rate": 0.0002,
+      "loss": 2.5301,
+      "step": 34480
+    },
+    {
+      "epoch": 2.570044709388972,
+      "grad_norm": 2.6223232746124268,
+      "learning_rate": 0.0002,
+      "loss": 2.6511,
+      "step": 34490
+    },
+    {
+      "epoch": 2.5707898658718333,
+      "grad_norm": 2.8373148441314697,
+      "learning_rate": 0.0002,
+      "loss": 2.3971,
+      "step": 34500
+    },
+    {
+      "epoch": 2.5715350223546944,
+      "grad_norm": 2.770490884780884,
+      "learning_rate": 0.0002,
+      "loss": 2.5108,
+      "step": 34510
+    },
+    {
+      "epoch": 2.572280178837556,
+      "grad_norm": 2.5854387283325195,
+      "learning_rate": 0.0002,
+      "loss": 2.5076,
+      "step": 34520
+    },
+    {
+      "epoch": 2.573025335320417,
+      "grad_norm": 2.411576986312866,
+      "learning_rate": 0.0002,
+      "loss": 2.6648,
+      "step": 34530
+    },
+    {
+      "epoch": 2.5737704918032787,
+      "grad_norm": 2.5123982429504395,
+      "learning_rate": 0.0002,
+      "loss": 2.5907,
+      "step": 34540
+    },
+    {
+      "epoch": 2.57451564828614,
+      "grad_norm": 2.831042528152466,
+      "learning_rate": 0.0002,
+      "loss": 2.5356,
+      "step": 34550
+    },
+    {
+      "epoch": 2.5752608047690018,
+      "grad_norm": 2.643927574157715,
+      "learning_rate": 0.0002,
+      "loss": 2.4261,
+      "step": 34560
+    },
+    {
+      "epoch": 2.576005961251863,
+      "grad_norm": 2.698838949203491,
+      "learning_rate": 0.0002,
+      "loss": 2.4362,
+      "step": 34570
+    },
+    {
+      "epoch": 2.5767511177347244,
+      "grad_norm": 2.549325704574585,
+      "learning_rate": 0.0002,
+      "loss": 2.5073,
+      "step": 34580
+    },
+    {
+      "epoch": 2.5774962742175855,
+      "grad_norm": 2.3992767333984375,
+      "learning_rate": 0.0002,
+      "loss": 2.468,
+      "step": 34590
+    },
+    {
+      "epoch": 2.578241430700447,
+      "grad_norm": 3.1624796390533447,
+      "learning_rate": 0.0002,
+      "loss": 2.5036,
+      "step": 34600
+    },
+    {
+      "epoch": 2.5789865871833086,
+      "grad_norm": 2.8962011337280273,
+      "learning_rate": 0.0002,
+      "loss": 2.5846,
+      "step": 34610
+    },
+    {
+      "epoch": 2.5797317436661698,
+      "grad_norm": 2.2955880165100098,
+      "learning_rate": 0.0002,
+      "loss": 2.5806,
+      "step": 34620
+    },
+    {
+      "epoch": 2.5804769001490313,
+      "grad_norm": 2.375588893890381,
+      "learning_rate": 0.0002,
+      "loss": 2.4193,
+      "step": 34630
+    },
+    {
+      "epoch": 2.581222056631893,
+      "grad_norm": 2.7803938388824463,
+      "learning_rate": 0.0002,
+      "loss": 2.4429,
+      "step": 34640
+    },
+    {
+      "epoch": 2.581967213114754,
+      "grad_norm": 2.5117294788360596,
+      "learning_rate": 0.0002,
+      "loss": 2.459,
+      "step": 34650
+    },
+    {
+      "epoch": 2.5827123695976155,
+      "grad_norm": 2.65371036529541,
+      "learning_rate": 0.0002,
+      "loss": 2.5789,
+      "step": 34660
+    },
+    {
+      "epoch": 2.583457526080477,
+      "grad_norm": 2.6229374408721924,
+      "learning_rate": 0.0002,
+      "loss": 2.6575,
+      "step": 34670
+    },
+    {
+      "epoch": 2.584202682563338,
+      "grad_norm": 3.0383689403533936,
+      "learning_rate": 0.0002,
+      "loss": 2.5808,
+      "step": 34680
+    },
+    {
+      "epoch": 2.5849478390461997,
+      "grad_norm": 2.6986963748931885,
+      "learning_rate": 0.0002,
+      "loss": 2.3615,
+      "step": 34690
+    },
+    {
+      "epoch": 2.585692995529061,
+      "grad_norm": 2.7407946586608887,
+      "learning_rate": 0.0002,
+      "loss": 2.1574,
+      "step": 34700
+    },
+    {
+      "epoch": 2.5864381520119224,
+      "grad_norm": 2.9318466186523438,
+      "learning_rate": 0.0002,
+      "loss": 2.6112,
+      "step": 34710
+    },
+    {
+      "epoch": 2.587183308494784,
+      "grad_norm": 2.139223337173462,
+      "learning_rate": 0.0002,
+      "loss": 2.4195,
+      "step": 34720
+    },
+    {
+      "epoch": 2.5879284649776455,
+      "grad_norm": 2.886835813522339,
+      "learning_rate": 0.0002,
+      "loss": 2.4984,
+      "step": 34730
+    },
+    {
+      "epoch": 2.5886736214605066,
+      "grad_norm": 2.40415358543396,
+      "learning_rate": 0.0002,
+      "loss": 2.5024,
+      "step": 34740
+    },
+    {
+      "epoch": 2.589418777943368,
+      "grad_norm": 2.612410306930542,
+      "learning_rate": 0.0002,
+      "loss": 2.4832,
+      "step": 34750
+    },
+    {
+      "epoch": 2.5901639344262293,
+      "grad_norm": 2.2832210063934326,
+      "learning_rate": 0.0002,
+      "loss": 2.4649,
+      "step": 34760
+    },
+    {
+      "epoch": 2.590909090909091,
+      "grad_norm": 2.763620615005493,
+      "learning_rate": 0.0002,
+      "loss": 2.3836,
+      "step": 34770
+    },
+    {
+      "epoch": 2.5916542473919524,
+      "grad_norm": 2.6266095638275146,
+      "learning_rate": 0.0002,
+      "loss": 2.4563,
+      "step": 34780
+    },
+    {
+      "epoch": 2.592399403874814,
+      "grad_norm": 2.604623317718506,
+      "learning_rate": 0.0002,
+      "loss": 2.5474,
+      "step": 34790
+    },
+    {
+      "epoch": 2.593144560357675,
+      "grad_norm": 2.5370583534240723,
+      "learning_rate": 0.0002,
+      "loss": 2.4926,
+      "step": 34800
+    },
+    {
+      "epoch": 2.5938897168405366,
+      "grad_norm": 2.7688779830932617,
+      "learning_rate": 0.0002,
+      "loss": 2.5277,
+      "step": 34810
+    },
+    {
+      "epoch": 2.5946348733233977,
+      "grad_norm": 1.6677604913711548,
+      "learning_rate": 0.0002,
+      "loss": 2.4469,
+      "step": 34820
+    },
+    {
+      "epoch": 2.5953800298062593,
+      "grad_norm": 2.6081464290618896,
+      "learning_rate": 0.0002,
+      "loss": 2.6667,
+      "step": 34830
+    },
+    {
+      "epoch": 2.596125186289121,
+      "grad_norm": 2.504087209701538,
+      "learning_rate": 0.0002,
+      "loss": 2.4811,
+      "step": 34840
+    },
+    {
+      "epoch": 2.5968703427719824,
+      "grad_norm": 2.514352798461914,
+      "learning_rate": 0.0002,
+      "loss": 2.4831,
+      "step": 34850
+    },
+    {
+      "epoch": 2.5976154992548435,
+      "grad_norm": 2.6763663291931152,
+      "learning_rate": 0.0002,
+      "loss": 2.4431,
+      "step": 34860
+    },
+    {
+      "epoch": 2.598360655737705,
+      "grad_norm": 2.6285643577575684,
+      "learning_rate": 0.0002,
+      "loss": 2.6107,
+      "step": 34870
+    },
+    {
+      "epoch": 2.599105812220566,
+      "grad_norm": 2.520859956741333,
+      "learning_rate": 0.0002,
+      "loss": 2.5125,
+      "step": 34880
+    },
+    {
+      "epoch": 2.5998509687034277,
+      "grad_norm": 2.3472795486450195,
+      "learning_rate": 0.0002,
+      "loss": 2.308,
+      "step": 34890
+    },
+    {
+      "epoch": 2.6005961251862892,
+      "grad_norm": 2.735682249069214,
+      "learning_rate": 0.0002,
+      "loss": 2.5909,
+      "step": 34900
+    },
+    {
+      "epoch": 2.601341281669151,
+      "grad_norm": 3.270073652267456,
+      "learning_rate": 0.0002,
+      "loss": 2.6196,
+      "step": 34910
+    },
+    {
+      "epoch": 2.602086438152012,
+      "grad_norm": 2.6279194355010986,
+      "learning_rate": 0.0002,
+      "loss": 2.519,
+      "step": 34920
+    },
+    {
+      "epoch": 2.6028315946348735,
+      "grad_norm": 2.125945568084717,
+      "learning_rate": 0.0002,
+      "loss": 2.4861,
+      "step": 34930
+    },
+    {
+      "epoch": 2.6035767511177346,
+      "grad_norm": 2.817042589187622,
+      "learning_rate": 0.0002,
+      "loss": 2.4631,
+      "step": 34940
+    },
+    {
+      "epoch": 2.604321907600596,
+      "grad_norm": 2.75313138961792,
+      "learning_rate": 0.0002,
+      "loss": 2.304,
+      "step": 34950
+    },
+    {
+      "epoch": 2.6050670640834577,
+      "grad_norm": 2.5719428062438965,
+      "learning_rate": 0.0002,
+      "loss": 2.4361,
+      "step": 34960
+    },
+    {
+      "epoch": 2.605812220566319,
+      "grad_norm": 2.4278576374053955,
+      "learning_rate": 0.0002,
+      "loss": 2.5518,
+      "step": 34970
+    },
+    {
+      "epoch": 2.6065573770491803,
+      "grad_norm": 2.707906484603882,
+      "learning_rate": 0.0002,
+      "loss": 2.5591,
+      "step": 34980
+    },
+    {
+      "epoch": 2.607302533532042,
+      "grad_norm": 3.9694321155548096,
+      "learning_rate": 0.0002,
+      "loss": 2.5416,
+      "step": 34990
+    },
+    {
+      "epoch": 2.608047690014903,
+      "grad_norm": 2.7127926349639893,
+      "learning_rate": 0.0002,
+      "loss": 2.3842,
+      "step": 35000
+    },
+    {
+      "epoch": 2.6087928464977646,
+      "grad_norm": 2.6059675216674805,
+      "learning_rate": 0.0002,
+      "loss": 2.4429,
+      "step": 35010
+    },
+    {
+      "epoch": 2.609538002980626,
+      "grad_norm": 2.09088134765625,
+      "learning_rate": 0.0002,
+      "loss": 2.3938,
+      "step": 35020
+    },
+    {
+      "epoch": 2.610283159463487,
+      "grad_norm": 3.14701509475708,
+      "learning_rate": 0.0002,
+      "loss": 2.3627,
+      "step": 35030
+    },
+    {
+      "epoch": 2.6110283159463488,
+      "grad_norm": 2.85632061958313,
+      "learning_rate": 0.0002,
+      "loss": 2.6743,
+      "step": 35040
+    },
+    {
+      "epoch": 2.61177347242921,
+      "grad_norm": 2.6178882122039795,
+      "learning_rate": 0.0002,
+      "loss": 2.5535,
+      "step": 35050
+    },
+    {
+      "epoch": 2.6125186289120714,
+      "grad_norm": 2.384420394897461,
+      "learning_rate": 0.0002,
+      "loss": 2.5122,
+      "step": 35060
+    },
+    {
+      "epoch": 2.613263785394933,
+      "grad_norm": 2.4369149208068848,
+      "learning_rate": 0.0002,
+      "loss": 2.6472,
+      "step": 35070
+    },
+    {
+      "epoch": 2.6140089418777945,
+      "grad_norm": 2.6348259449005127,
+      "learning_rate": 0.0002,
+      "loss": 2.4447,
+      "step": 35080
+    },
+    {
+      "epoch": 2.6147540983606556,
+      "grad_norm": 2.780752420425415,
+      "learning_rate": 0.0002,
+      "loss": 2.5351,
+      "step": 35090
+    },
+    {
+      "epoch": 2.615499254843517,
+      "grad_norm": 2.450267791748047,
+      "learning_rate": 0.0002,
+      "loss": 2.3382,
+      "step": 35100
+    },
+    {
+      "epoch": 2.6162444113263783,
+      "grad_norm": 2.720033645629883,
+      "learning_rate": 0.0002,
+      "loss": 2.3684,
+      "step": 35110
+    },
+    {
+      "epoch": 2.61698956780924,
+      "grad_norm": 2.995755910873413,
+      "learning_rate": 0.0002,
+      "loss": 2.4707,
+      "step": 35120
+    },
+    {
+      "epoch": 2.6177347242921014,
+      "grad_norm": 2.7804386615753174,
+      "learning_rate": 0.0002,
+      "loss": 2.5992,
+      "step": 35130
+    },
+    {
+      "epoch": 2.618479880774963,
+      "grad_norm": 2.687592029571533,
+      "learning_rate": 0.0002,
+      "loss": 2.6438,
+      "step": 35140
+    },
+    {
+      "epoch": 2.619225037257824,
+      "grad_norm": 2.72942852973938,
+      "learning_rate": 0.0002,
+      "loss": 2.5031,
+      "step": 35150
+    },
+    {
+      "epoch": 2.6199701937406856,
+      "grad_norm": 2.743436813354492,
+      "learning_rate": 0.0002,
+      "loss": 2.5397,
+      "step": 35160
+    },
+    {
+      "epoch": 2.6207153502235467,
+      "grad_norm": 2.544156789779663,
+      "learning_rate": 0.0002,
+      "loss": 2.579,
+      "step": 35170
+    },
+    {
+      "epoch": 2.6214605067064083,
+      "grad_norm": 2.2848076820373535,
+      "learning_rate": 0.0002,
+      "loss": 2.42,
+      "step": 35180
+    },
+    {
+      "epoch": 2.62220566318927,
+      "grad_norm": 2.5742909908294678,
+      "learning_rate": 0.0002,
+      "loss": 2.4444,
+      "step": 35190
+    },
+    {
+      "epoch": 2.6229508196721314,
+      "grad_norm": 2.081449270248413,
+      "learning_rate": 0.0002,
+      "loss": 2.4044,
+      "step": 35200
+    },
+    {
+      "epoch": 2.6236959761549925,
+      "grad_norm": 2.488731622695923,
+      "learning_rate": 0.0002,
+      "loss": 2.5191,
+      "step": 35210
+    },
+    {
+      "epoch": 2.624441132637854,
+      "grad_norm": 2.7537927627563477,
+      "learning_rate": 0.0002,
+      "loss": 2.6245,
+      "step": 35220
+    },
+    {
+      "epoch": 2.625186289120715,
+      "grad_norm": 2.9869134426116943,
+      "learning_rate": 0.0002,
+      "loss": 2.4964,
+      "step": 35230
+    },
+    {
+      "epoch": 2.6259314456035767,
+      "grad_norm": 2.343207836151123,
+      "learning_rate": 0.0002,
+      "loss": 2.3746,
+      "step": 35240
+    },
+    {
+      "epoch": 2.6266766020864383,
+      "grad_norm": 2.5939278602600098,
+      "learning_rate": 0.0002,
+      "loss": 2.5991,
+      "step": 35250
+    },
+    {
+      "epoch": 2.6274217585693,
+      "grad_norm": 2.727090835571289,
+      "learning_rate": 0.0002,
+      "loss": 2.2542,
+      "step": 35260
+    },
+    {
+      "epoch": 2.628166915052161,
+      "grad_norm": 2.4267163276672363,
+      "learning_rate": 0.0002,
+      "loss": 2.3105,
+      "step": 35270
+    },
+    {
+      "epoch": 2.6289120715350225,
+      "grad_norm": 2.565988540649414,
+      "learning_rate": 0.0002,
+      "loss": 2.6,
+      "step": 35280
+    },
+    {
+      "epoch": 2.6296572280178836,
+      "grad_norm": 2.4375131130218506,
+      "learning_rate": 0.0002,
+      "loss": 2.5983,
+      "step": 35290
+    },
+    {
+      "epoch": 2.630402384500745,
+      "grad_norm": 2.661345958709717,
+      "learning_rate": 0.0002,
+      "loss": 2.3076,
+      "step": 35300
+    },
+    {
+      "epoch": 2.6311475409836067,
+      "grad_norm": 2.9548776149749756,
+      "learning_rate": 0.0002,
+      "loss": 2.5269,
+      "step": 35310
+    },
+    {
+      "epoch": 2.631892697466468,
+      "grad_norm": 2.780071496963501,
+      "learning_rate": 0.0002,
+      "loss": 2.5164,
+      "step": 35320
+    },
+    {
+      "epoch": 2.6326378539493294,
+      "grad_norm": 3.2674901485443115,
+      "learning_rate": 0.0002,
+      "loss": 2.3561,
+      "step": 35330
+    },
+    {
+      "epoch": 2.633383010432191,
+      "grad_norm": 2.833890438079834,
+      "learning_rate": 0.0002,
+      "loss": 2.2951,
+      "step": 35340
+    },
+    {
+      "epoch": 2.634128166915052,
+      "grad_norm": 2.463898181915283,
+      "learning_rate": 0.0002,
+      "loss": 2.5373,
+      "step": 35350
+    },
+    {
+      "epoch": 2.6348733233979136,
+      "grad_norm": 2.6374974250793457,
+      "learning_rate": 0.0002,
+      "loss": 2.5007,
+      "step": 35360
+    },
+    {
+      "epoch": 2.635618479880775,
+      "grad_norm": 2.711229085922241,
+      "learning_rate": 0.0002,
+      "loss": 2.4471,
+      "step": 35370
+    },
+    {
+      "epoch": 2.6363636363636362,
+      "grad_norm": 2.583221435546875,
+      "learning_rate": 0.0002,
+      "loss": 2.5363,
+      "step": 35380
+    },
+    {
+      "epoch": 2.637108792846498,
+      "grad_norm": 2.668238878250122,
+      "learning_rate": 0.0002,
+      "loss": 2.5197,
+      "step": 35390
+    },
+    {
+      "epoch": 2.637853949329359,
+      "grad_norm": 2.6465303897857666,
+      "learning_rate": 0.0002,
+      "loss": 2.2952,
+      "step": 35400
+    },
+    {
+      "epoch": 2.6385991058122205,
+      "grad_norm": 2.5099635124206543,
+      "learning_rate": 0.0002,
+      "loss": 2.547,
+      "step": 35410
+    },
+    {
+      "epoch": 2.639344262295082,
+      "grad_norm": 2.978407621383667,
+      "learning_rate": 0.0002,
+      "loss": 2.563,
+      "step": 35420
+    },
+    {
+      "epoch": 2.6400894187779436,
+      "grad_norm": 2.6208372116088867,
+      "learning_rate": 0.0002,
+      "loss": 2.6188,
+      "step": 35430
+    },
+    {
+      "epoch": 2.6408345752608047,
+      "grad_norm": 3.47013258934021,
+      "learning_rate": 0.0002,
+      "loss": 2.5536,
+      "step": 35440
+    },
+    {
+      "epoch": 2.6415797317436662,
+      "grad_norm": 2.438246726989746,
+      "learning_rate": 0.0002,
+      "loss": 2.385,
+      "step": 35450
+    },
+    {
+      "epoch": 2.6423248882265273,
+      "grad_norm": 2.5819106101989746,
+      "learning_rate": 0.0002,
+      "loss": 2.5517,
+      "step": 35460
+    },
+    {
+      "epoch": 2.643070044709389,
+      "grad_norm": 2.8637759685516357,
+      "learning_rate": 0.0002,
+      "loss": 2.48,
+      "step": 35470
+    },
+    {
+      "epoch": 2.6438152011922504,
+      "grad_norm": 2.3931562900543213,
+      "learning_rate": 0.0002,
+      "loss": 2.3955,
+      "step": 35480
+    },
+    {
+      "epoch": 2.644560357675112,
+      "grad_norm": 3.022334575653076,
+      "learning_rate": 0.0002,
+      "loss": 2.3759,
+      "step": 35490
+    },
+    {
+      "epoch": 2.645305514157973,
+      "grad_norm": 2.471989154815674,
+      "learning_rate": 0.0002,
+      "loss": 2.4271,
+      "step": 35500
+    },
+    {
+      "epoch": 2.6460506706408347,
+      "grad_norm": 2.57979679107666,
+      "learning_rate": 0.0002,
+      "loss": 2.5865,
+      "step": 35510
+    },
+    {
+      "epoch": 2.6467958271236958,
+      "grad_norm": 2.5882067680358887,
+      "learning_rate": 0.0002,
+      "loss": 2.5651,
+      "step": 35520
+    },
+    {
+      "epoch": 2.6475409836065573,
+      "grad_norm": 2.924689531326294,
+      "learning_rate": 0.0002,
+      "loss": 2.5692,
+      "step": 35530
+    },
+    {
+      "epoch": 2.648286140089419,
+      "grad_norm": 2.678907632827759,
+      "learning_rate": 0.0002,
+      "loss": 2.3231,
+      "step": 35540
+    },
+    {
+      "epoch": 2.6490312965722804,
+      "grad_norm": 2.9073657989501953,
+      "learning_rate": 0.0002,
+      "loss": 2.4733,
+      "step": 35550
+    },
+    {
+      "epoch": 2.6497764530551415,
+      "grad_norm": 2.6203291416168213,
+      "learning_rate": 0.0002,
+      "loss": 2.4252,
+      "step": 35560
+    },
+    {
+      "epoch": 2.650521609538003,
+      "grad_norm": 2.8620171546936035,
+      "learning_rate": 0.0002,
+      "loss": 2.5345,
+      "step": 35570
+    },
+    {
+      "epoch": 2.651266766020864,
+      "grad_norm": 3.0765066146850586,
+      "learning_rate": 0.0002,
+      "loss": 2.5476,
+      "step": 35580
+    },
+    {
+      "epoch": 2.6520119225037257,
+      "grad_norm": 2.6432547569274902,
+      "learning_rate": 0.0002,
+      "loss": 2.6744,
+      "step": 35590
+    },
+    {
+      "epoch": 2.6527570789865873,
+      "grad_norm": 2.4183008670806885,
+      "learning_rate": 0.0002,
+      "loss": 2.4373,
+      "step": 35600
+    },
+    {
+      "epoch": 2.653502235469449,
+      "grad_norm": 2.6401169300079346,
+      "learning_rate": 0.0002,
+      "loss": 2.6115,
+      "step": 35610
+    },
+    {
+      "epoch": 2.65424739195231,
+      "grad_norm": 2.336107015609741,
+      "learning_rate": 0.0002,
+      "loss": 2.5014,
+      "step": 35620
+    },
+    {
+      "epoch": 2.6549925484351715,
+      "grad_norm": 2.7248358726501465,
+      "learning_rate": 0.0002,
+      "loss": 2.3869,
+      "step": 35630
+    },
+    {
+      "epoch": 2.6557377049180326,
+      "grad_norm": 2.5932023525238037,
+      "learning_rate": 0.0002,
+      "loss": 2.6997,
+      "step": 35640
+    },
+    {
+      "epoch": 2.656482861400894,
+      "grad_norm": 2.6041550636291504,
+      "learning_rate": 0.0002,
+      "loss": 2.4897,
+      "step": 35650
+    },
+    {
+      "epoch": 2.6572280178837557,
+      "grad_norm": 2.6969892978668213,
+      "learning_rate": 0.0002,
+      "loss": 2.564,
+      "step": 35660
+    },
+    {
+      "epoch": 2.657973174366617,
+      "grad_norm": 2.610666275024414,
+      "learning_rate": 0.0002,
+      "loss": 2.6471,
+      "step": 35670
+    },
+    {
+      "epoch": 2.6587183308494784,
+      "grad_norm": 2.631580114364624,
+      "learning_rate": 0.0002,
+      "loss": 2.5464,
+      "step": 35680
+    },
+    {
+      "epoch": 2.65946348733234,
+      "grad_norm": 2.7000410556793213,
+      "learning_rate": 0.0002,
+      "loss": 2.5847,
+      "step": 35690
+    },
+    {
+      "epoch": 2.660208643815201,
+      "grad_norm": 2.5906224250793457,
+      "learning_rate": 0.0002,
+      "loss": 2.5968,
+      "step": 35700
+    },
+    {
+      "epoch": 2.6609538002980626,
+      "grad_norm": 2.6354689598083496,
+      "learning_rate": 0.0002,
+      "loss": 2.6701,
+      "step": 35710
+    },
+    {
+      "epoch": 2.661698956780924,
+      "grad_norm": 2.598720073699951,
+      "learning_rate": 0.0002,
+      "loss": 2.5396,
+      "step": 35720
+    },
+    {
+      "epoch": 2.6624441132637853,
+      "grad_norm": 2.6841044425964355,
+      "learning_rate": 0.0002,
+      "loss": 2.5725,
+      "step": 35730
+    },
+    {
+      "epoch": 2.663189269746647,
+      "grad_norm": 2.7917165756225586,
+      "learning_rate": 0.0002,
+      "loss": 2.5492,
+      "step": 35740
+    },
+    {
+      "epoch": 2.663934426229508,
+      "grad_norm": 2.852966547012329,
+      "learning_rate": 0.0002,
+      "loss": 2.5787,
+      "step": 35750
+    },
+    {
+      "epoch": 2.6646795827123695,
+      "grad_norm": 2.4196488857269287,
+      "learning_rate": 0.0002,
+      "loss": 2.6144,
+      "step": 35760
+    },
+    {
+      "epoch": 2.665424739195231,
+      "grad_norm": 2.710623264312744,
+      "learning_rate": 0.0002,
+      "loss": 2.669,
+      "step": 35770
+    },
+    {
+      "epoch": 2.6661698956780926,
+      "grad_norm": 2.8702971935272217,
+      "learning_rate": 0.0002,
+      "loss": 2.3902,
+      "step": 35780
+    },
+    {
+      "epoch": 2.6669150521609537,
+      "grad_norm": 2.8416409492492676,
+      "learning_rate": 0.0002,
+      "loss": 2.6465,
+      "step": 35790
+    },
+    {
+      "epoch": 2.6676602086438153,
+      "grad_norm": 2.761198043823242,
+      "learning_rate": 0.0002,
+      "loss": 2.3041,
+      "step": 35800
+    },
+    {
+      "epoch": 2.6684053651266764,
+      "grad_norm": 2.7360453605651855,
+      "learning_rate": 0.0002,
+      "loss": 2.3746,
+      "step": 35810
+    },
+    {
+      "epoch": 2.669150521609538,
+      "grad_norm": 2.5194859504699707,
+      "learning_rate": 0.0002,
+      "loss": 2.5007,
+      "step": 35820
+    },
+    {
+      "epoch": 2.6698956780923995,
+      "grad_norm": 2.272688388824463,
+      "learning_rate": 0.0002,
+      "loss": 2.3409,
+      "step": 35830
+    },
+    {
+      "epoch": 2.670640834575261,
+      "grad_norm": 2.8286778926849365,
+      "learning_rate": 0.0002,
+      "loss": 2.4279,
+      "step": 35840
+    },
+    {
+      "epoch": 2.671385991058122,
+      "grad_norm": 2.7673115730285645,
+      "learning_rate": 0.0002,
+      "loss": 2.6094,
+      "step": 35850
+    },
+    {
+      "epoch": 2.6721311475409837,
+      "grad_norm": 2.627052068710327,
+      "learning_rate": 0.0002,
+      "loss": 2.4081,
+      "step": 35860
+    },
+    {
+      "epoch": 2.672876304023845,
+      "grad_norm": 2.9847543239593506,
+      "learning_rate": 0.0002,
+      "loss": 2.5277,
+      "step": 35870
+    },
+    {
+      "epoch": 2.6736214605067063,
+      "grad_norm": 2.7096521854400635,
+      "learning_rate": 0.0002,
+      "loss": 2.4191,
+      "step": 35880
+    },
+    {
+      "epoch": 2.674366616989568,
+      "grad_norm": 2.454817056655884,
+      "learning_rate": 0.0002,
+      "loss": 2.5331,
+      "step": 35890
+    },
+    {
+      "epoch": 2.6751117734724295,
+      "grad_norm": 2.617335557937622,
+      "learning_rate": 0.0002,
+      "loss": 2.2716,
+      "step": 35900
+    },
+    {
+      "epoch": 2.6758569299552906,
+      "grad_norm": 2.6219711303710938,
+      "learning_rate": 0.0002,
+      "loss": 2.4334,
+      "step": 35910
+    },
+    {
+      "epoch": 2.676602086438152,
+      "grad_norm": 2.2985072135925293,
+      "learning_rate": 0.0002,
+      "loss": 2.5012,
+      "step": 35920
+    },
+    {
+      "epoch": 2.6773472429210132,
+      "grad_norm": 2.7559444904327393,
+      "learning_rate": 0.0002,
+      "loss": 2.5972,
+      "step": 35930
+    },
+    {
+      "epoch": 2.678092399403875,
+      "grad_norm": 2.789966583251953,
+      "learning_rate": 0.0002,
+      "loss": 2.4791,
+      "step": 35940
+    },
+    {
+      "epoch": 2.6788375558867363,
+      "grad_norm": 2.9208645820617676,
+      "learning_rate": 0.0002,
+      "loss": 2.5503,
+      "step": 35950
+    },
+    {
+      "epoch": 2.679582712369598,
+      "grad_norm": 2.563593864440918,
+      "learning_rate": 0.0002,
+      "loss": 2.6192,
+      "step": 35960
+    },
+    {
+      "epoch": 2.680327868852459,
+      "grad_norm": 2.91192364692688,
+      "learning_rate": 0.0002,
+      "loss": 2.3414,
+      "step": 35970
+    },
+    {
+      "epoch": 2.6810730253353205,
+      "grad_norm": 2.6559500694274902,
+      "learning_rate": 0.0002,
+      "loss": 2.3,
+      "step": 35980
+    },
+    {
+      "epoch": 2.6818181818181817,
+      "grad_norm": 2.29866886138916,
+      "learning_rate": 0.0002,
+      "loss": 2.463,
+      "step": 35990
+    },
+    {
+      "epoch": 2.682563338301043,
+      "grad_norm": 2.475616216659546,
+      "learning_rate": 0.0002,
+      "loss": 2.456,
+      "step": 36000
+    },
+    {
+      "epoch": 2.6833084947839048,
+      "grad_norm": 2.6832971572875977,
+      "learning_rate": 0.0002,
+      "loss": 2.5678,
+      "step": 36010
+    },
+    {
+      "epoch": 2.684053651266766,
+      "grad_norm": 2.717376708984375,
+      "learning_rate": 0.0002,
+      "loss": 2.5248,
+      "step": 36020
+    },
+    {
+      "epoch": 2.6847988077496274,
+      "grad_norm": 3.192349910736084,
+      "learning_rate": 0.0002,
+      "loss": 2.3875,
+      "step": 36030
+    },
+    {
+      "epoch": 2.685543964232489,
+      "grad_norm": 2.5880353450775146,
+      "learning_rate": 0.0002,
+      "loss": 2.5578,
+      "step": 36040
+    },
+    {
+      "epoch": 2.68628912071535,
+      "grad_norm": 2.367215394973755,
+      "learning_rate": 0.0002,
+      "loss": 2.553,
+      "step": 36050
+    },
+    {
+      "epoch": 2.6870342771982116,
+      "grad_norm": 2.3446109294891357,
+      "learning_rate": 0.0002,
+      "loss": 2.5302,
+      "step": 36060
+    },
+    {
+      "epoch": 2.687779433681073,
+      "grad_norm": 2.9561381340026855,
+      "learning_rate": 0.0002,
+      "loss": 2.7431,
+      "step": 36070
+    },
+    {
+      "epoch": 2.6885245901639343,
+      "grad_norm": 2.4364163875579834,
+      "learning_rate": 0.0002,
+      "loss": 2.4944,
+      "step": 36080
+    },
+    {
+      "epoch": 2.689269746646796,
+      "grad_norm": 2.623122215270996,
+      "learning_rate": 0.0002,
+      "loss": 2.5464,
+      "step": 36090
+    },
+    {
+      "epoch": 2.690014903129657,
+      "grad_norm": 2.6617910861968994,
+      "learning_rate": 0.0002,
+      "loss": 2.5365,
+      "step": 36100
+    },
+    {
+      "epoch": 2.6907600596125185,
+      "grad_norm": 2.458702564239502,
+      "learning_rate": 0.0002,
+      "loss": 2.525,
+      "step": 36110
+    },
+    {
+      "epoch": 2.69150521609538,
+      "grad_norm": 2.783048629760742,
+      "learning_rate": 0.0002,
+      "loss": 2.4975,
+      "step": 36120
+    },
+    {
+      "epoch": 2.6922503725782416,
+      "grad_norm": 3.2591092586517334,
+      "learning_rate": 0.0002,
+      "loss": 2.5694,
+      "step": 36130
+    },
+    {
+      "epoch": 2.6929955290611027,
+      "grad_norm": 2.77760910987854,
+      "learning_rate": 0.0002,
+      "loss": 2.4527,
+      "step": 36140
+    },
+    {
+      "epoch": 2.6937406855439643,
+      "grad_norm": 2.6666505336761475,
+      "learning_rate": 0.0002,
+      "loss": 2.5507,
+      "step": 36150
+    },
+    {
+      "epoch": 2.6944858420268254,
+      "grad_norm": 2.61444354057312,
+      "learning_rate": 0.0002,
+      "loss": 2.4439,
+      "step": 36160
+    },
+    {
+      "epoch": 2.695230998509687,
+      "grad_norm": 2.2256479263305664,
+      "learning_rate": 0.0002,
+      "loss": 2.3892,
+      "step": 36170
+    },
+    {
+      "epoch": 2.6959761549925485,
+      "grad_norm": 2.4084644317626953,
+      "learning_rate": 0.0002,
+      "loss": 2.5677,
+      "step": 36180
+    },
+    {
+      "epoch": 2.69672131147541,
+      "grad_norm": 3.0166308879852295,
+      "learning_rate": 0.0002,
+      "loss": 2.4645,
+      "step": 36190
+    },
+    {
+      "epoch": 2.697466467958271,
+      "grad_norm": 2.671741008758545,
+      "learning_rate": 0.0002,
+      "loss": 2.6156,
+      "step": 36200
+    },
+    {
+      "epoch": 2.6982116244411327,
+      "grad_norm": 2.723830223083496,
+      "learning_rate": 0.0002,
+      "loss": 2.4664,
+      "step": 36210
+    },
+    {
+      "epoch": 2.698956780923994,
+      "grad_norm": 2.855217695236206,
+      "learning_rate": 0.0002,
+      "loss": 2.481,
+      "step": 36220
+    },
+    {
+      "epoch": 2.6997019374068554,
+      "grad_norm": 2.242652416229248,
+      "learning_rate": 0.0002,
+      "loss": 2.3911,
+      "step": 36230
+    },
+    {
+      "epoch": 2.700447093889717,
+      "grad_norm": 2.4699301719665527,
+      "learning_rate": 0.0002,
+      "loss": 2.5208,
+      "step": 36240
+    },
+    {
+      "epoch": 2.7011922503725785,
+      "grad_norm": 2.6784889698028564,
+      "learning_rate": 0.0002,
+      "loss": 2.2951,
+      "step": 36250
+    },
+    {
+      "epoch": 2.7019374068554396,
+      "grad_norm": 2.4282853603363037,
+      "learning_rate": 0.0002,
+      "loss": 2.6418,
+      "step": 36260
+    },
+    {
+      "epoch": 2.702682563338301,
+      "grad_norm": 2.9294357299804688,
+      "learning_rate": 0.0002,
+      "loss": 2.6716,
+      "step": 36270
+    },
+    {
+      "epoch": 2.7034277198211623,
+      "grad_norm": 2.729619264602661,
+      "learning_rate": 0.0002,
+      "loss": 2.4832,
+      "step": 36280
+    },
+    {
+      "epoch": 2.704172876304024,
+      "grad_norm": 2.574918031692505,
+      "learning_rate": 0.0002,
+      "loss": 2.4267,
+      "step": 36290
+    },
+    {
+      "epoch": 2.7049180327868854,
+      "grad_norm": 2.5243308544158936,
+      "learning_rate": 0.0002,
+      "loss": 2.4349,
+      "step": 36300
+    },
+    {
+      "epoch": 2.705663189269747,
+      "grad_norm": 2.761279821395874,
+      "learning_rate": 0.0002,
+      "loss": 2.5064,
+      "step": 36310
+    },
+    {
+      "epoch": 2.706408345752608,
+      "grad_norm": 2.487800121307373,
+      "learning_rate": 0.0002,
+      "loss": 2.481,
+      "step": 36320
+    },
+    {
+      "epoch": 2.7071535022354696,
+      "grad_norm": 2.707909107208252,
+      "learning_rate": 0.0002,
+      "loss": 2.5022,
+      "step": 36330
+    },
+    {
+      "epoch": 2.7078986587183307,
+      "grad_norm": 2.6682002544403076,
+      "learning_rate": 0.0002,
+      "loss": 2.5912,
+      "step": 36340
+    },
+    {
+      "epoch": 2.7086438152011922,
+      "grad_norm": 2.786775588989258,
+      "learning_rate": 0.0002,
+      "loss": 2.4986,
+      "step": 36350
+    },
+    {
+      "epoch": 2.709388971684054,
+      "grad_norm": 2.84505295753479,
+      "learning_rate": 0.0002,
+      "loss": 2.4681,
+      "step": 36360
+    },
+    {
+      "epoch": 2.710134128166915,
+      "grad_norm": 2.364448308944702,
+      "learning_rate": 0.0002,
+      "loss": 2.5899,
+      "step": 36370
+    },
+    {
+      "epoch": 2.7108792846497765,
+      "grad_norm": 2.7492170333862305,
+      "learning_rate": 0.0002,
+      "loss": 2.3887,
+      "step": 36380
+    },
+    {
+      "epoch": 2.711624441132638,
+      "grad_norm": 2.7889928817749023,
+      "learning_rate": 0.0002,
+      "loss": 2.5625,
+      "step": 36390
+    },
+    {
+      "epoch": 2.712369597615499,
+      "grad_norm": 2.5459094047546387,
+      "learning_rate": 0.0002,
+      "loss": 2.6412,
+      "step": 36400
+    },
+    {
+      "epoch": 2.7131147540983607,
+      "grad_norm": 2.648048162460327,
+      "learning_rate": 0.0002,
+      "loss": 2.5925,
+      "step": 36410
+    },
+    {
+      "epoch": 2.7138599105812222,
+      "grad_norm": 2.7833564281463623,
+      "learning_rate": 0.0002,
+      "loss": 2.3773,
+      "step": 36420
+    },
+    {
+      "epoch": 2.7146050670640833,
+      "grad_norm": 2.9601666927337646,
+      "learning_rate": 0.0002,
+      "loss": 2.7538,
+      "step": 36430
+    },
+    {
+      "epoch": 2.715350223546945,
+      "grad_norm": 3.0814125537872314,
+      "learning_rate": 0.0002,
+      "loss": 2.6718,
+      "step": 36440
+    },
+    {
+      "epoch": 2.716095380029806,
+      "grad_norm": 2.4058563709259033,
+      "learning_rate": 0.0002,
+      "loss": 2.6188,
+      "step": 36450
+    },
+    {
+      "epoch": 2.7168405365126675,
+      "grad_norm": 2.7348439693450928,
+      "learning_rate": 0.0002,
+      "loss": 2.6306,
+      "step": 36460
+    },
+    {
+      "epoch": 2.717585692995529,
+      "grad_norm": 2.448627471923828,
+      "learning_rate": 0.0002,
+      "loss": 2.5885,
+      "step": 36470
+    },
+    {
+      "epoch": 2.7183308494783907,
+      "grad_norm": 2.4670491218566895,
+      "learning_rate": 0.0002,
+      "loss": 2.44,
+      "step": 36480
+    },
+    {
+      "epoch": 2.7190760059612518,
+      "grad_norm": 2.3683083057403564,
+      "learning_rate": 0.0002,
+      "loss": 2.5304,
+      "step": 36490
+    },
+    {
+      "epoch": 2.7198211624441133,
+      "grad_norm": 2.6678662300109863,
+      "learning_rate": 0.0002,
+      "loss": 2.6104,
+      "step": 36500
+    },
+    {
+      "epoch": 2.7205663189269744,
+      "grad_norm": 2.6376142501831055,
+      "learning_rate": 0.0002,
+      "loss": 2.5446,
+      "step": 36510
+    },
+    {
+      "epoch": 2.721311475409836,
+      "grad_norm": 2.51070237159729,
+      "learning_rate": 0.0002,
+      "loss": 2.5938,
+      "step": 36520
+    },
+    {
+      "epoch": 2.7220566318926975,
+      "grad_norm": 2.3976802825927734,
+      "learning_rate": 0.0002,
+      "loss": 2.3664,
+      "step": 36530
+    },
+    {
+      "epoch": 2.722801788375559,
+      "grad_norm": 2.557739496231079,
+      "learning_rate": 0.0002,
+      "loss": 2.6302,
+      "step": 36540
+    },
+    {
+      "epoch": 2.72354694485842,
+      "grad_norm": 2.489729404449463,
+      "learning_rate": 0.0002,
+      "loss": 2.325,
+      "step": 36550
+    },
+    {
+      "epoch": 2.7242921013412817,
+      "grad_norm": 2.4591825008392334,
+      "learning_rate": 0.0002,
+      "loss": 2.4624,
+      "step": 36560
+    },
+    {
+      "epoch": 2.725037257824143,
+      "grad_norm": 2.7376599311828613,
+      "learning_rate": 0.0002,
+      "loss": 2.3916,
+      "step": 36570
+    },
+    {
+      "epoch": 2.7257824143070044,
+      "grad_norm": 2.656623125076294,
+      "learning_rate": 0.0002,
+      "loss": 2.48,
+      "step": 36580
+    },
+    {
+      "epoch": 2.726527570789866,
+      "grad_norm": 2.71795392036438,
+      "learning_rate": 0.0002,
+      "loss": 2.5828,
+      "step": 36590
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 2.759921073913574,
+      "learning_rate": 0.0002,
+      "loss": 2.5211,
+      "step": 36600
+    },
+    {
+      "epoch": 2.7280178837555886,
+      "grad_norm": 2.558744430541992,
+      "learning_rate": 0.0002,
+      "loss": 2.5161,
+      "step": 36610
+    },
+    {
+      "epoch": 2.72876304023845,
+      "grad_norm": 2.899496078491211,
+      "learning_rate": 0.0002,
+      "loss": 2.3968,
+      "step": 36620
+    },
+    {
+      "epoch": 2.7295081967213113,
+      "grad_norm": 2.5663375854492188,
+      "learning_rate": 0.0002,
+      "loss": 2.4568,
+      "step": 36630
+    },
+    {
+      "epoch": 2.730253353204173,
+      "grad_norm": 2.4600677490234375,
+      "learning_rate": 0.0002,
+      "loss": 2.4366,
+      "step": 36640
+    },
+    {
+      "epoch": 2.7309985096870344,
+      "grad_norm": 2.5599944591522217,
+      "learning_rate": 0.0002,
+      "loss": 2.3192,
+      "step": 36650
+    },
+    {
+      "epoch": 2.731743666169896,
+      "grad_norm": 2.5939974784851074,
+      "learning_rate": 0.0002,
+      "loss": 2.6118,
+      "step": 36660
+    },
+    {
+      "epoch": 2.732488822652757,
+      "grad_norm": 2.5912539958953857,
+      "learning_rate": 0.0002,
+      "loss": 2.5891,
+      "step": 36670
+    },
+    {
+      "epoch": 2.7332339791356186,
+      "grad_norm": 3.277555465698242,
+      "learning_rate": 0.0002,
+      "loss": 2.3884,
+      "step": 36680
+    },
+    {
+      "epoch": 2.7339791356184797,
+      "grad_norm": 2.8053581714630127,
+      "learning_rate": 0.0002,
+      "loss": 2.4551,
+      "step": 36690
+    },
+    {
+      "epoch": 2.7347242921013413,
+      "grad_norm": 2.5936853885650635,
+      "learning_rate": 0.0002,
+      "loss": 2.513,
+      "step": 36700
+    },
+    {
+      "epoch": 2.735469448584203,
+      "grad_norm": 2.418405294418335,
+      "learning_rate": 0.0002,
+      "loss": 2.2918,
+      "step": 36710
+    },
+    {
+      "epoch": 2.736214605067064,
+      "grad_norm": 2.522139072418213,
+      "learning_rate": 0.0002,
+      "loss": 2.4492,
+      "step": 36720
+    },
+    {
+      "epoch": 2.7369597615499255,
+      "grad_norm": 2.8760268688201904,
+      "learning_rate": 0.0002,
+      "loss": 2.4819,
+      "step": 36730
+    },
+    {
+      "epoch": 2.737704918032787,
+      "grad_norm": 2.54500150680542,
+      "learning_rate": 0.0002,
+      "loss": 2.4401,
+      "step": 36740
+    },
+    {
+      "epoch": 2.738450074515648,
+      "grad_norm": 3.1572742462158203,
+      "learning_rate": 0.0002,
+      "loss": 2.3991,
+      "step": 36750
+    },
+    {
+      "epoch": 2.7391952309985097,
+      "grad_norm": 2.786902666091919,
+      "learning_rate": 0.0002,
+      "loss": 2.5974,
+      "step": 36760
+    },
+    {
+      "epoch": 2.7399403874813713,
+      "grad_norm": 2.5400989055633545,
+      "learning_rate": 0.0002,
+      "loss": 2.5479,
+      "step": 36770
+    },
+    {
+      "epoch": 2.7406855439642324,
+      "grad_norm": 2.4169130325317383,
+      "learning_rate": 0.0002,
+      "loss": 2.4016,
+      "step": 36780
+    },
+    {
+      "epoch": 2.741430700447094,
+      "grad_norm": 2.450286865234375,
+      "learning_rate": 0.0002,
+      "loss": 2.5699,
+      "step": 36790
+    },
+    {
+      "epoch": 2.742175856929955,
+      "grad_norm": 2.757652759552002,
+      "learning_rate": 0.0002,
+      "loss": 2.5433,
+      "step": 36800
+    },
+    {
+      "epoch": 2.7429210134128166,
+      "grad_norm": 2.3215696811676025,
+      "learning_rate": 0.0002,
+      "loss": 2.4423,
+      "step": 36810
+    },
+    {
+      "epoch": 2.743666169895678,
+      "grad_norm": 2.422499895095825,
+      "learning_rate": 0.0002,
+      "loss": 2.2945,
+      "step": 36820
+    },
+    {
+      "epoch": 2.7444113263785397,
+      "grad_norm": 2.4259049892425537,
+      "learning_rate": 0.0002,
+      "loss": 2.5629,
+      "step": 36830
+    },
+    {
+      "epoch": 2.745156482861401,
+      "grad_norm": 2.6686112880706787,
+      "learning_rate": 0.0002,
+      "loss": 2.4427,
+      "step": 36840
+    },
+    {
+      "epoch": 2.7459016393442623,
+      "grad_norm": 2.5597991943359375,
+      "learning_rate": 0.0002,
+      "loss": 2.4121,
+      "step": 36850
+    },
+    {
+      "epoch": 2.7466467958271235,
+      "grad_norm": 2.713035821914673,
+      "learning_rate": 0.0002,
+      "loss": 2.5649,
+      "step": 36860
+    },
+    {
+      "epoch": 2.747391952309985,
+      "grad_norm": 2.730883836746216,
+      "learning_rate": 0.0002,
+      "loss": 2.3433,
+      "step": 36870
+    },
+    {
+      "epoch": 2.7481371087928466,
+      "grad_norm": 2.833503007888794,
+      "learning_rate": 0.0002,
+      "loss": 2.4411,
+      "step": 36880
+    },
+    {
+      "epoch": 2.748882265275708,
+      "grad_norm": 2.7715904712677,
+      "learning_rate": 0.0002,
+      "loss": 2.4082,
+      "step": 36890
+    },
+    {
+      "epoch": 2.7496274217585692,
+      "grad_norm": 2.716165065765381,
+      "learning_rate": 0.0002,
+      "loss": 2.3817,
+      "step": 36900
+    },
+    {
+      "epoch": 2.7503725782414308,
+      "grad_norm": 2.7706308364868164,
+      "learning_rate": 0.0002,
+      "loss": 2.5381,
+      "step": 36910
+    },
+    {
+      "epoch": 2.751117734724292,
+      "grad_norm": 2.569395065307617,
+      "learning_rate": 0.0002,
+      "loss": 2.4863,
+      "step": 36920
+    },
+    {
+      "epoch": 2.7518628912071534,
+      "grad_norm": 2.589057207107544,
+      "learning_rate": 0.0002,
+      "loss": 2.5399,
+      "step": 36930
+    },
+    {
+      "epoch": 2.752608047690015,
+      "grad_norm": 2.632874011993408,
+      "learning_rate": 0.0002,
+      "loss": 2.4628,
+      "step": 36940
+    },
+    {
+      "epoch": 2.7533532041728765,
+      "grad_norm": 3.023503541946411,
+      "learning_rate": 0.0002,
+      "loss": 2.5736,
+      "step": 36950
+    },
+    {
+      "epoch": 2.7540983606557377,
+      "grad_norm": 2.538649559020996,
+      "learning_rate": 0.0002,
+      "loss": 2.4441,
+      "step": 36960
+    },
+    {
+      "epoch": 2.754843517138599,
+      "grad_norm": 2.65378737449646,
+      "learning_rate": 0.0002,
+      "loss": 2.5479,
+      "step": 36970
+    },
+    {
+      "epoch": 2.7555886736214603,
+      "grad_norm": 2.514577865600586,
+      "learning_rate": 0.0002,
+      "loss": 2.3853,
+      "step": 36980
+    },
+    {
+      "epoch": 2.756333830104322,
+      "grad_norm": 2.3040006160736084,
+      "learning_rate": 0.0002,
+      "loss": 2.5896,
+      "step": 36990
+    },
+    {
+      "epoch": 2.7570789865871834,
+      "grad_norm": 2.7404825687408447,
+      "learning_rate": 0.0002,
+      "loss": 2.7121,
+      "step": 37000
+    },
+    {
+      "epoch": 2.757824143070045,
+      "grad_norm": 2.6444921493530273,
+      "learning_rate": 0.0002,
+      "loss": 2.4906,
+      "step": 37010
+    },
+    {
+      "epoch": 2.758569299552906,
+      "grad_norm": 2.368401288986206,
+      "learning_rate": 0.0002,
+      "loss": 2.413,
+      "step": 37020
+    },
+    {
+      "epoch": 2.7593144560357676,
+      "grad_norm": 2.6097865104675293,
+      "learning_rate": 0.0002,
+      "loss": 2.567,
+      "step": 37030
+    },
+    {
+      "epoch": 2.7600596125186287,
+      "grad_norm": 2.694190740585327,
+      "learning_rate": 0.0002,
+      "loss": 2.57,
+      "step": 37040
+    },
+    {
+      "epoch": 2.7608047690014903,
+      "grad_norm": 2.5604560375213623,
+      "learning_rate": 0.0002,
+      "loss": 2.4598,
+      "step": 37050
+    },
+    {
+      "epoch": 2.761549925484352,
+      "grad_norm": 2.912165641784668,
+      "learning_rate": 0.0002,
+      "loss": 2.6524,
+      "step": 37060
+    },
+    {
+      "epoch": 2.762295081967213,
+      "grad_norm": 3.2894179821014404,
+      "learning_rate": 0.0002,
+      "loss": 2.5217,
+      "step": 37070
+    },
+    {
+      "epoch": 2.7630402384500745,
+      "grad_norm": 2.828571319580078,
+      "learning_rate": 0.0002,
+      "loss": 2.5196,
+      "step": 37080
+    },
+    {
+      "epoch": 2.763785394932936,
+      "grad_norm": 2.4312257766723633,
+      "learning_rate": 0.0002,
+      "loss": 2.3419,
+      "step": 37090
+    },
+    {
+      "epoch": 2.764530551415797,
+      "grad_norm": 2.4520037174224854,
+      "learning_rate": 0.0002,
+      "loss": 2.4494,
+      "step": 37100
+    },
+    {
+      "epoch": 2.7652757078986587,
+      "grad_norm": 2.708139181137085,
+      "learning_rate": 0.0002,
+      "loss": 2.5631,
+      "step": 37110
+    },
+    {
+      "epoch": 2.7660208643815203,
+      "grad_norm": 2.8655529022216797,
+      "learning_rate": 0.0002,
+      "loss": 2.5352,
+      "step": 37120
+    },
+    {
+      "epoch": 2.7667660208643814,
+      "grad_norm": 2.484017848968506,
+      "learning_rate": 0.0002,
+      "loss": 2.4615,
+      "step": 37130
+    },
+    {
+      "epoch": 2.767511177347243,
+      "grad_norm": 2.4956047534942627,
+      "learning_rate": 0.0002,
+      "loss": 2.5465,
+      "step": 37140
+    },
+    {
+      "epoch": 2.768256333830104,
+      "grad_norm": 2.5548791885375977,
+      "learning_rate": 0.0002,
+      "loss": 2.4609,
+      "step": 37150
+    },
+    {
+      "epoch": 2.7690014903129656,
+      "grad_norm": 2.561833381652832,
+      "learning_rate": 0.0002,
+      "loss": 2.5994,
+      "step": 37160
+    },
+    {
+      "epoch": 2.769746646795827,
+      "grad_norm": 2.9793152809143066,
+      "learning_rate": 0.0002,
+      "loss": 2.432,
+      "step": 37170
+    },
+    {
+      "epoch": 2.7704918032786887,
+      "grad_norm": 2.6743390560150146,
+      "learning_rate": 0.0002,
+      "loss": 2.4788,
+      "step": 37180
+    },
+    {
+      "epoch": 2.77123695976155,
+      "grad_norm": 2.461435317993164,
+      "learning_rate": 0.0002,
+      "loss": 2.5217,
+      "step": 37190
+    },
+    {
+      "epoch": 2.7719821162444114,
+      "grad_norm": 2.5211734771728516,
+      "learning_rate": 0.0002,
+      "loss": 2.613,
+      "step": 37200
+    },
+    {
+      "epoch": 2.7727272727272725,
+      "grad_norm": 2.824679136276245,
+      "learning_rate": 0.0002,
+      "loss": 2.6382,
+      "step": 37210
+    },
+    {
+      "epoch": 2.773472429210134,
+      "grad_norm": 2.458425521850586,
+      "learning_rate": 0.0002,
+      "loss": 2.4033,
+      "step": 37220
+    },
+    {
+      "epoch": 2.7742175856929956,
+      "grad_norm": 2.7605643272399902,
+      "learning_rate": 0.0002,
+      "loss": 2.4153,
+      "step": 37230
+    },
+    {
+      "epoch": 2.774962742175857,
+      "grad_norm": 2.3235936164855957,
+      "learning_rate": 0.0002,
+      "loss": 2.5107,
+      "step": 37240
+    },
+    {
+      "epoch": 2.7757078986587183,
+      "grad_norm": 2.2972285747528076,
+      "learning_rate": 0.0002,
+      "loss": 2.5371,
+      "step": 37250
+    },
+    {
+      "epoch": 2.77645305514158,
+      "grad_norm": 2.7939791679382324,
+      "learning_rate": 0.0002,
+      "loss": 2.3748,
+      "step": 37260
+    },
+    {
+      "epoch": 2.777198211624441,
+      "grad_norm": 2.5067827701568604,
+      "learning_rate": 0.0002,
+      "loss": 2.3598,
+      "step": 37270
+    },
+    {
+      "epoch": 2.7779433681073025,
+      "grad_norm": 2.909909248352051,
+      "learning_rate": 0.0002,
+      "loss": 2.5853,
+      "step": 37280
+    },
+    {
+      "epoch": 2.778688524590164,
+      "grad_norm": 2.569460868835449,
+      "learning_rate": 0.0002,
+      "loss": 2.5193,
+      "step": 37290
+    },
+    {
+      "epoch": 2.7794336810730256,
+      "grad_norm": 3.214890956878662,
+      "learning_rate": 0.0002,
+      "loss": 2.4651,
+      "step": 37300
+    },
+    {
+      "epoch": 2.7801788375558867,
+      "grad_norm": 2.867858648300171,
+      "learning_rate": 0.0002,
+      "loss": 2.6508,
+      "step": 37310
+    },
+    {
+      "epoch": 2.7809239940387482,
+      "grad_norm": 2.7677152156829834,
+      "learning_rate": 0.0002,
+      "loss": 2.5517,
+      "step": 37320
+    },
+    {
+      "epoch": 2.7816691505216093,
+      "grad_norm": 2.633157253265381,
+      "learning_rate": 0.0002,
+      "loss": 2.4063,
+      "step": 37330
+    },
+    {
+      "epoch": 2.782414307004471,
+      "grad_norm": 2.4681646823883057,
+      "learning_rate": 0.0002,
+      "loss": 2.4698,
+      "step": 37340
+    },
+    {
+      "epoch": 2.7831594634873325,
+      "grad_norm": 2.595750570297241,
+      "learning_rate": 0.0002,
+      "loss": 2.398,
+      "step": 37350
+    },
+    {
+      "epoch": 2.783904619970194,
+      "grad_norm": 2.4892077445983887,
+      "learning_rate": 0.0002,
+      "loss": 2.5829,
+      "step": 37360
+    },
+    {
+      "epoch": 2.784649776453055,
+      "grad_norm": 2.4976730346679688,
+      "learning_rate": 0.0002,
+      "loss": 2.3726,
+      "step": 37370
+    },
+    {
+      "epoch": 2.7853949329359167,
+      "grad_norm": 2.330193519592285,
+      "learning_rate": 0.0002,
+      "loss": 2.4521,
+      "step": 37380
+    },
+    {
+      "epoch": 2.7861400894187778,
+      "grad_norm": 2.6689720153808594,
+      "learning_rate": 0.0002,
+      "loss": 2.414,
+      "step": 37390
+    },
+    {
+      "epoch": 2.7868852459016393,
+      "grad_norm": 2.6920714378356934,
+      "learning_rate": 0.0002,
+      "loss": 2.5038,
+      "step": 37400
+    },
+    {
+      "epoch": 2.787630402384501,
+      "grad_norm": 2.4694406986236572,
+      "learning_rate": 0.0002,
+      "loss": 2.3686,
+      "step": 37410
+    },
+    {
+      "epoch": 2.788375558867362,
+      "grad_norm": 2.5774457454681396,
+      "learning_rate": 0.0002,
+      "loss": 2.3996,
+      "step": 37420
+    },
+    {
+      "epoch": 2.7891207153502235,
+      "grad_norm": 2.640498638153076,
+      "learning_rate": 0.0002,
+      "loss": 2.4829,
+      "step": 37430
+    },
+    {
+      "epoch": 2.789865871833085,
+      "grad_norm": 2.770134925842285,
+      "learning_rate": 0.0002,
+      "loss": 2.5329,
+      "step": 37440
+    },
+    {
+      "epoch": 2.790611028315946,
+      "grad_norm": 2.9241816997528076,
+      "learning_rate": 0.0002,
+      "loss": 2.4869,
+      "step": 37450
+    },
+    {
+      "epoch": 2.7913561847988078,
+      "grad_norm": 2.629659652709961,
+      "learning_rate": 0.0002,
+      "loss": 2.4289,
+      "step": 37460
+    },
+    {
+      "epoch": 2.7921013412816693,
+      "grad_norm": 2.6241800785064697,
+      "learning_rate": 0.0002,
+      "loss": 2.5179,
+      "step": 37470
+    },
+    {
+      "epoch": 2.7928464977645304,
+      "grad_norm": 2.5617072582244873,
+      "learning_rate": 0.0002,
+      "loss": 2.5042,
+      "step": 37480
+    },
+    {
+      "epoch": 2.793591654247392,
+      "grad_norm": 2.659287452697754,
+      "learning_rate": 0.0002,
+      "loss": 2.4369,
+      "step": 37490
+    },
+    {
+      "epoch": 2.794336810730253,
+      "grad_norm": 2.6287639141082764,
+      "learning_rate": 0.0002,
+      "loss": 2.4064,
+      "step": 37500
+    },
+    {
+      "epoch": 2.7950819672131146,
+      "grad_norm": 2.7507472038269043,
+      "learning_rate": 0.0002,
+      "loss": 2.4495,
+      "step": 37510
+    },
+    {
+      "epoch": 2.795827123695976,
+      "grad_norm": 2.509035348892212,
+      "learning_rate": 0.0002,
+      "loss": 2.4707,
+      "step": 37520
+    },
+    {
+      "epoch": 2.7965722801788377,
+      "grad_norm": 2.2699403762817383,
+      "learning_rate": 0.0002,
+      "loss": 2.3909,
+      "step": 37530
+    },
+    {
+      "epoch": 2.797317436661699,
+      "grad_norm": 2.630939245223999,
+      "learning_rate": 0.0002,
+      "loss": 2.4802,
+      "step": 37540
+    },
+    {
+      "epoch": 2.7980625931445604,
+      "grad_norm": 3.045865058898926,
+      "learning_rate": 0.0002,
+      "loss": 2.4884,
+      "step": 37550
+    },
+    {
+      "epoch": 2.7988077496274215,
+      "grad_norm": 2.486050605773926,
+      "learning_rate": 0.0002,
+      "loss": 2.4905,
+      "step": 37560
+    },
+    {
+      "epoch": 2.799552906110283,
+      "grad_norm": 2.318683624267578,
+      "learning_rate": 0.0002,
+      "loss": 2.2404,
+      "step": 37570
+    },
+    {
+      "epoch": 2.8002980625931446,
+      "grad_norm": 2.7278268337249756,
+      "learning_rate": 0.0002,
+      "loss": 2.5899,
+      "step": 37580
+    },
+    {
+      "epoch": 2.801043219076006,
+      "grad_norm": 2.8368096351623535,
+      "learning_rate": 0.0002,
+      "loss": 2.5105,
+      "step": 37590
+    },
+    {
+      "epoch": 2.8017883755588673,
+      "grad_norm": 2.7143986225128174,
+      "learning_rate": 0.0002,
+      "loss": 2.4156,
+      "step": 37600
+    },
+    {
+      "epoch": 2.802533532041729,
+      "grad_norm": 2.6735024452209473,
+      "learning_rate": 0.0002,
+      "loss": 2.3569,
+      "step": 37610
+    },
+    {
+      "epoch": 2.80327868852459,
+      "grad_norm": 2.700565814971924,
+      "learning_rate": 0.0002,
+      "loss": 2.485,
+      "step": 37620
+    },
+    {
+      "epoch": 2.8040238450074515,
+      "grad_norm": 2.5607948303222656,
+      "learning_rate": 0.0002,
+      "loss": 2.5495,
+      "step": 37630
+    },
+    {
+      "epoch": 2.804769001490313,
+      "grad_norm": 2.044367790222168,
+      "learning_rate": 0.0002,
+      "loss": 2.452,
+      "step": 37640
+    },
+    {
+      "epoch": 2.8055141579731746,
+      "grad_norm": 2.569173574447632,
+      "learning_rate": 0.0002,
+      "loss": 2.554,
+      "step": 37650
+    },
+    {
+      "epoch": 2.8062593144560357,
+      "grad_norm": 2.6920878887176514,
+      "learning_rate": 0.0002,
+      "loss": 2.5635,
+      "step": 37660
+    },
+    {
+      "epoch": 2.8070044709388973,
+      "grad_norm": 2.5670292377471924,
+      "learning_rate": 0.0002,
+      "loss": 2.4568,
+      "step": 37670
+    },
+    {
+      "epoch": 2.8077496274217584,
+      "grad_norm": 2.1942081451416016,
+      "learning_rate": 0.0002,
+      "loss": 2.4599,
+      "step": 37680
+    },
+    {
+      "epoch": 2.80849478390462,
+      "grad_norm": 2.5298988819122314,
+      "learning_rate": 0.0002,
+      "loss": 2.4536,
+      "step": 37690
+    },
+    {
+      "epoch": 2.8092399403874815,
+      "grad_norm": 2.82232666015625,
+      "learning_rate": 0.0002,
+      "loss": 2.4608,
+      "step": 37700
+    },
+    {
+      "epoch": 2.809985096870343,
+      "grad_norm": 2.4265050888061523,
+      "learning_rate": 0.0002,
+      "loss": 2.4723,
+      "step": 37710
+    },
+    {
+      "epoch": 2.810730253353204,
+      "grad_norm": 2.705826759338379,
+      "learning_rate": 0.0002,
+      "loss": 2.4147,
+      "step": 37720
+    },
+    {
+      "epoch": 2.8114754098360657,
+      "grad_norm": 2.584636926651001,
+      "learning_rate": 0.0002,
+      "loss": 2.5809,
+      "step": 37730
+    },
+    {
+      "epoch": 2.812220566318927,
+      "grad_norm": 2.8928425312042236,
+      "learning_rate": 0.0002,
+      "loss": 2.7184,
+      "step": 37740
+    },
+    {
+      "epoch": 2.8129657228017884,
+      "grad_norm": 2.567809581756592,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 37750
+    },
+    {
+      "epoch": 2.81371087928465,
+      "grad_norm": 2.8919975757598877,
+      "learning_rate": 0.0002,
+      "loss": 2.4801,
+      "step": 37760
+    },
+    {
+      "epoch": 2.814456035767511,
+      "grad_norm": 2.455564022064209,
+      "learning_rate": 0.0002,
+      "loss": 2.5779,
+      "step": 37770
+    },
+    {
+      "epoch": 2.8152011922503726,
+      "grad_norm": 2.659975528717041,
+      "learning_rate": 0.0002,
+      "loss": 2.6608,
+      "step": 37780
+    },
+    {
+      "epoch": 2.815946348733234,
+      "grad_norm": 2.6775760650634766,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 37790
+    },
+    {
+      "epoch": 2.8166915052160952,
+      "grad_norm": 2.86912202835083,
+      "learning_rate": 0.0002,
+      "loss": 2.3338,
+      "step": 37800
+    },
+    {
+      "epoch": 2.817436661698957,
+      "grad_norm": 2.8781628608703613,
+      "learning_rate": 0.0002,
+      "loss": 2.6059,
+      "step": 37810
+    },
+    {
+      "epoch": 2.8181818181818183,
+      "grad_norm": 2.688969135284424,
+      "learning_rate": 0.0002,
+      "loss": 2.792,
+      "step": 37820
+    },
+    {
+      "epoch": 2.8189269746646795,
+      "grad_norm": 2.80010986328125,
+      "learning_rate": 0.0002,
+      "loss": 2.3393,
+      "step": 37830
+    },
+    {
+      "epoch": 2.819672131147541,
+      "grad_norm": 2.4589450359344482,
+      "learning_rate": 0.0002,
+      "loss": 2.5836,
+      "step": 37840
+    },
+    {
+      "epoch": 2.820417287630402,
+      "grad_norm": 2.3914148807525635,
+      "learning_rate": 0.0002,
+      "loss": 2.5583,
+      "step": 37850
+    },
+    {
+      "epoch": 2.8211624441132637,
+      "grad_norm": 2.734759569168091,
+      "learning_rate": 0.0002,
+      "loss": 2.624,
+      "step": 37860
+    },
+    {
+      "epoch": 2.821907600596125,
+      "grad_norm": 2.7147603034973145,
+      "learning_rate": 0.0002,
+      "loss": 2.6149,
+      "step": 37870
+    },
+    {
+      "epoch": 2.8226527570789868,
+      "grad_norm": 2.4781367778778076,
+      "learning_rate": 0.0002,
+      "loss": 2.5671,
+      "step": 37880
+    },
+    {
+      "epoch": 2.823397913561848,
+      "grad_norm": 2.6984505653381348,
+      "learning_rate": 0.0002,
+      "loss": 2.5764,
+      "step": 37890
+    },
+    {
+      "epoch": 2.8241430700447094,
+      "grad_norm": 2.6264562606811523,
+      "learning_rate": 0.0002,
+      "loss": 2.272,
+      "step": 37900
+    },
+    {
+      "epoch": 2.8248882265275705,
+      "grad_norm": 2.978095054626465,
+      "learning_rate": 0.0002,
+      "loss": 2.483,
+      "step": 37910
+    },
+    {
+      "epoch": 2.825633383010432,
+      "grad_norm": 2.6886980533599854,
+      "learning_rate": 0.0002,
+      "loss": 2.6149,
+      "step": 37920
+    },
+    {
+      "epoch": 2.8263785394932937,
+      "grad_norm": 2.677035331726074,
+      "learning_rate": 0.0002,
+      "loss": 2.6447,
+      "step": 37930
+    },
+    {
+      "epoch": 2.827123695976155,
+      "grad_norm": 3.3771278858184814,
+      "learning_rate": 0.0002,
+      "loss": 2.7219,
+      "step": 37940
+    },
+    {
+      "epoch": 2.8278688524590163,
+      "grad_norm": 1.9655869007110596,
+      "learning_rate": 0.0002,
+      "loss": 2.4848,
+      "step": 37950
+    },
+    {
+      "epoch": 2.828614008941878,
+      "grad_norm": 2.6271440982818604,
+      "learning_rate": 0.0002,
+      "loss": 2.4456,
+      "step": 37960
+    },
+    {
+      "epoch": 2.829359165424739,
+      "grad_norm": 2.5585412979125977,
+      "learning_rate": 0.0002,
+      "loss": 2.7289,
+      "step": 37970
+    },
+    {
+      "epoch": 2.8301043219076005,
+      "grad_norm": 2.735304117202759,
+      "learning_rate": 0.0002,
+      "loss": 2.4382,
+      "step": 37980
+    },
+    {
+      "epoch": 2.830849478390462,
+      "grad_norm": 2.609837055206299,
+      "learning_rate": 0.0002,
+      "loss": 2.4991,
+      "step": 37990
+    },
+    {
+      "epoch": 2.8315946348733236,
+      "grad_norm": 2.7447588443756104,
+      "learning_rate": 0.0002,
+      "loss": 2.5325,
+      "step": 38000
+    },
+    {
+      "epoch": 2.8323397913561847,
+      "grad_norm": 2.400519847869873,
+      "learning_rate": 0.0002,
+      "loss": 2.6211,
+      "step": 38010
+    },
+    {
+      "epoch": 2.8330849478390463,
+      "grad_norm": 2.8218953609466553,
+      "learning_rate": 0.0002,
+      "loss": 2.6096,
+      "step": 38020
+    },
+    {
+      "epoch": 2.8338301043219074,
+      "grad_norm": 2.7308266162872314,
+      "learning_rate": 0.0002,
+      "loss": 2.557,
+      "step": 38030
+    },
+    {
+      "epoch": 2.834575260804769,
+      "grad_norm": 2.5944695472717285,
+      "learning_rate": 0.0002,
+      "loss": 2.5554,
+      "step": 38040
+    },
+    {
+      "epoch": 2.8353204172876305,
+      "grad_norm": 2.427563428878784,
+      "learning_rate": 0.0002,
+      "loss": 2.3361,
+      "step": 38050
+    },
+    {
+      "epoch": 2.836065573770492,
+      "grad_norm": 2.4476137161254883,
+      "learning_rate": 0.0002,
+      "loss": 2.4115,
+      "step": 38060
+    },
+    {
+      "epoch": 2.836810730253353,
+      "grad_norm": 3.2641639709472656,
+      "learning_rate": 0.0002,
+      "loss": 2.4905,
+      "step": 38070
+    },
+    {
+      "epoch": 2.8375558867362147,
+      "grad_norm": 2.3455142974853516,
+      "learning_rate": 0.0002,
+      "loss": 2.4734,
+      "step": 38080
+    },
+    {
+      "epoch": 2.838301043219076,
+      "grad_norm": 2.834339141845703,
+      "learning_rate": 0.0002,
+      "loss": 2.5312,
+      "step": 38090
+    },
+    {
+      "epoch": 2.8390461997019374,
+      "grad_norm": 2.8637094497680664,
+      "learning_rate": 0.0002,
+      "loss": 2.4174,
+      "step": 38100
+    },
+    {
+      "epoch": 2.839791356184799,
+      "grad_norm": 3.0524630546569824,
+      "learning_rate": 0.0002,
+      "loss": 2.5523,
+      "step": 38110
+    },
+    {
+      "epoch": 2.84053651266766,
+      "grad_norm": 2.649812936782837,
+      "learning_rate": 0.0002,
+      "loss": 2.2958,
+      "step": 38120
+    },
+    {
+      "epoch": 2.8412816691505216,
+      "grad_norm": 2.645263433456421,
+      "learning_rate": 0.0002,
+      "loss": 2.4867,
+      "step": 38130
+    },
+    {
+      "epoch": 2.842026825633383,
+      "grad_norm": 2.0363569259643555,
+      "learning_rate": 0.0002,
+      "loss": 2.3287,
+      "step": 38140
+    },
+    {
+      "epoch": 2.8427719821162443,
+      "grad_norm": 2.7432138919830322,
+      "learning_rate": 0.0002,
+      "loss": 2.6093,
+      "step": 38150
+    },
+    {
+      "epoch": 2.843517138599106,
+      "grad_norm": 2.721869945526123,
+      "learning_rate": 0.0002,
+      "loss": 2.6309,
+      "step": 38160
+    },
+    {
+      "epoch": 2.8442622950819674,
+      "grad_norm": 2.674832582473755,
+      "learning_rate": 0.0002,
+      "loss": 2.3478,
+      "step": 38170
+    },
+    {
+      "epoch": 2.8450074515648285,
+      "grad_norm": 2.252639055252075,
+      "learning_rate": 0.0002,
+      "loss": 2.3417,
+      "step": 38180
+    },
+    {
+      "epoch": 2.84575260804769,
+      "grad_norm": 2.549644708633423,
+      "learning_rate": 0.0002,
+      "loss": 2.3126,
+      "step": 38190
+    },
+    {
+      "epoch": 2.846497764530551,
+      "grad_norm": 2.77371883392334,
+      "learning_rate": 0.0002,
+      "loss": 2.6122,
+      "step": 38200
+    },
+    {
+      "epoch": 2.8472429210134127,
+      "grad_norm": 2.7712173461914062,
+      "learning_rate": 0.0002,
+      "loss": 2.4918,
+      "step": 38210
+    },
+    {
+      "epoch": 2.8479880774962743,
+      "grad_norm": 2.5276310443878174,
+      "learning_rate": 0.0002,
+      "loss": 2.443,
+      "step": 38220
+    },
+    {
+      "epoch": 2.848733233979136,
+      "grad_norm": 2.3760945796966553,
+      "learning_rate": 0.0002,
+      "loss": 2.4374,
+      "step": 38230
+    },
+    {
+      "epoch": 2.849478390461997,
+      "grad_norm": 2.5871782302856445,
+      "learning_rate": 0.0002,
+      "loss": 2.571,
+      "step": 38240
+    },
+    {
+      "epoch": 2.8502235469448585,
+      "grad_norm": 2.721224069595337,
+      "learning_rate": 0.0002,
+      "loss": 2.582,
+      "step": 38250
+    },
+    {
+      "epoch": 2.8509687034277196,
+      "grad_norm": 2.55556058883667,
+      "learning_rate": 0.0002,
+      "loss": 2.3885,
+      "step": 38260
+    },
+    {
+      "epoch": 2.851713859910581,
+      "grad_norm": 2.8083126544952393,
+      "learning_rate": 0.0002,
+      "loss": 2.4455,
+      "step": 38270
+    },
+    {
+      "epoch": 2.8524590163934427,
+      "grad_norm": 2.5174221992492676,
+      "learning_rate": 0.0002,
+      "loss": 2.5159,
+      "step": 38280
+    },
+    {
+      "epoch": 2.8532041728763042,
+      "grad_norm": 2.380772590637207,
+      "learning_rate": 0.0002,
+      "loss": 2.6566,
+      "step": 38290
+    },
+    {
+      "epoch": 2.8539493293591653,
+      "grad_norm": 2.3448495864868164,
+      "learning_rate": 0.0002,
+      "loss": 2.4935,
+      "step": 38300
+    },
+    {
+      "epoch": 2.854694485842027,
+      "grad_norm": 2.500901699066162,
+      "learning_rate": 0.0002,
+      "loss": 2.3849,
+      "step": 38310
+    },
+    {
+      "epoch": 2.855439642324888,
+      "grad_norm": 2.658292293548584,
+      "learning_rate": 0.0002,
+      "loss": 2.3852,
+      "step": 38320
+    },
+    {
+      "epoch": 2.8561847988077496,
+      "grad_norm": 2.5677103996276855,
+      "learning_rate": 0.0002,
+      "loss": 2.467,
+      "step": 38330
+    },
+    {
+      "epoch": 2.856929955290611,
+      "grad_norm": 2.411125898361206,
+      "learning_rate": 0.0002,
+      "loss": 2.4616,
+      "step": 38340
+    },
+    {
+      "epoch": 2.8576751117734727,
+      "grad_norm": 2.651226043701172,
+      "learning_rate": 0.0002,
+      "loss": 2.4536,
+      "step": 38350
+    },
+    {
+      "epoch": 2.8584202682563338,
+      "grad_norm": 2.6440541744232178,
+      "learning_rate": 0.0002,
+      "loss": 2.349,
+      "step": 38360
+    },
+    {
+      "epoch": 2.8591654247391953,
+      "grad_norm": 2.4641504287719727,
+      "learning_rate": 0.0002,
+      "loss": 2.5248,
+      "step": 38370
+    },
+    {
+      "epoch": 2.8599105812220564,
+      "grad_norm": 2.716660499572754,
+      "learning_rate": 0.0002,
+      "loss": 2.606,
+      "step": 38380
+    },
+    {
+      "epoch": 2.860655737704918,
+      "grad_norm": 2.255239963531494,
+      "learning_rate": 0.0002,
+      "loss": 2.4059,
+      "step": 38390
+    },
+    {
+      "epoch": 2.8614008941877795,
+      "grad_norm": 2.145322322845459,
+      "learning_rate": 0.0002,
+      "loss": 2.4944,
+      "step": 38400
+    },
+    {
+      "epoch": 2.862146050670641,
+      "grad_norm": 2.833284378051758,
+      "learning_rate": 0.0002,
+      "loss": 2.5779,
+      "step": 38410
+    },
+    {
+      "epoch": 2.862891207153502,
+      "grad_norm": 2.970714569091797,
+      "learning_rate": 0.0002,
+      "loss": 2.4606,
+      "step": 38420
+    },
+    {
+      "epoch": 2.8636363636363638,
+      "grad_norm": 2.7407591342926025,
+      "learning_rate": 0.0002,
+      "loss": 2.5038,
+      "step": 38430
+    },
+    {
+      "epoch": 2.864381520119225,
+      "grad_norm": 2.734581708908081,
+      "learning_rate": 0.0002,
+      "loss": 2.4065,
+      "step": 38440
+    },
+    {
+      "epoch": 2.8651266766020864,
+      "grad_norm": 2.5742697715759277,
+      "learning_rate": 0.0002,
+      "loss": 2.359,
+      "step": 38450
+    },
+    {
+      "epoch": 2.865871833084948,
+      "grad_norm": 2.6709296703338623,
+      "learning_rate": 0.0002,
+      "loss": 2.4409,
+      "step": 38460
+    },
+    {
+      "epoch": 2.866616989567809,
+      "grad_norm": 2.3500664234161377,
+      "learning_rate": 0.0002,
+      "loss": 2.3481,
+      "step": 38470
+    },
+    {
+      "epoch": 2.8673621460506706,
+      "grad_norm": 2.6254522800445557,
+      "learning_rate": 0.0002,
+      "loss": 2.6385,
+      "step": 38480
+    },
+    {
+      "epoch": 2.868107302533532,
+      "grad_norm": 2.2674224376678467,
+      "learning_rate": 0.0002,
+      "loss": 2.5171,
+      "step": 38490
+    },
+    {
+      "epoch": 2.8688524590163933,
+      "grad_norm": 2.6358728408813477,
+      "learning_rate": 0.0002,
+      "loss": 2.649,
+      "step": 38500
+    },
+    {
+      "epoch": 2.869597615499255,
+      "grad_norm": 2.7351884841918945,
+      "learning_rate": 0.0002,
+      "loss": 2.6956,
+      "step": 38510
+    },
+    {
+      "epoch": 2.8703427719821164,
+      "grad_norm": 2.099317789077759,
+      "learning_rate": 0.0002,
+      "loss": 2.4007,
+      "step": 38520
+    },
+    {
+      "epoch": 2.8710879284649775,
+      "grad_norm": 2.7157390117645264,
+      "learning_rate": 0.0002,
+      "loss": 2.6054,
+      "step": 38530
+    },
+    {
+      "epoch": 2.871833084947839,
+      "grad_norm": 2.5156712532043457,
+      "learning_rate": 0.0002,
+      "loss": 2.4476,
+      "step": 38540
+    },
+    {
+      "epoch": 2.8725782414307,
+      "grad_norm": 2.307621955871582,
+      "learning_rate": 0.0002,
+      "loss": 2.3631,
+      "step": 38550
+    },
+    {
+      "epoch": 2.8733233979135617,
+      "grad_norm": 2.7251169681549072,
+      "learning_rate": 0.0002,
+      "loss": 2.4591,
+      "step": 38560
+    },
+    {
+      "epoch": 2.8740685543964233,
+      "grad_norm": 2.367175340652466,
+      "learning_rate": 0.0002,
+      "loss": 2.5013,
+      "step": 38570
+    },
+    {
+      "epoch": 2.874813710879285,
+      "grad_norm": 3.0403735637664795,
+      "learning_rate": 0.0002,
+      "loss": 2.4797,
+      "step": 38580
+    },
+    {
+      "epoch": 2.875558867362146,
+      "grad_norm": 2.884767770767212,
+      "learning_rate": 0.0002,
+      "loss": 2.159,
+      "step": 38590
+    },
+    {
+      "epoch": 2.8763040238450075,
+      "grad_norm": 2.2452404499053955,
+      "learning_rate": 0.0002,
+      "loss": 2.4509,
+      "step": 38600
+    },
+    {
+      "epoch": 2.8770491803278686,
+      "grad_norm": 2.496917486190796,
+      "learning_rate": 0.0002,
+      "loss": 2.5203,
+      "step": 38610
+    },
+    {
+      "epoch": 2.87779433681073,
+      "grad_norm": 2.7510409355163574,
+      "learning_rate": 0.0002,
+      "loss": 2.5307,
+      "step": 38620
+    },
+    {
+      "epoch": 2.8785394932935917,
+      "grad_norm": 2.698000907897949,
+      "learning_rate": 0.0002,
+      "loss": 2.625,
+      "step": 38630
+    },
+    {
+      "epoch": 2.8792846497764533,
+      "grad_norm": 2.3727385997772217,
+      "learning_rate": 0.0002,
+      "loss": 2.612,
+      "step": 38640
+    },
+    {
+      "epoch": 2.8800298062593144,
+      "grad_norm": 2.5714304447174072,
+      "learning_rate": 0.0002,
+      "loss": 2.5383,
+      "step": 38650
+    },
+    {
+      "epoch": 2.880774962742176,
+      "grad_norm": 2.4994068145751953,
+      "learning_rate": 0.0002,
+      "loss": 2.3407,
+      "step": 38660
+    },
+    {
+      "epoch": 2.881520119225037,
+      "grad_norm": 2.728253126144409,
+      "learning_rate": 0.0002,
+      "loss": 2.4754,
+      "step": 38670
+    },
+    {
+      "epoch": 2.8822652757078986,
+      "grad_norm": 2.7636911869049072,
+      "learning_rate": 0.0002,
+      "loss": 2.5834,
+      "step": 38680
+    },
+    {
+      "epoch": 2.88301043219076,
+      "grad_norm": 2.6368253231048584,
+      "learning_rate": 0.0002,
+      "loss": 2.526,
+      "step": 38690
+    },
+    {
+      "epoch": 2.8837555886736217,
+      "grad_norm": 2.8573157787323,
+      "learning_rate": 0.0002,
+      "loss": 2.5318,
+      "step": 38700
+    },
+    {
+      "epoch": 2.884500745156483,
+      "grad_norm": 2.7631237506866455,
+      "learning_rate": 0.0002,
+      "loss": 2.4429,
+      "step": 38710
+    },
+    {
+      "epoch": 2.8852459016393444,
+      "grad_norm": 2.6455793380737305,
+      "learning_rate": 0.0002,
+      "loss": 2.5363,
+      "step": 38720
+    },
+    {
+      "epoch": 2.8859910581222055,
+      "grad_norm": 2.3246517181396484,
+      "learning_rate": 0.0002,
+      "loss": 2.379,
+      "step": 38730
+    },
+    {
+      "epoch": 2.886736214605067,
+      "grad_norm": 2.690464735031128,
+      "learning_rate": 0.0002,
+      "loss": 2.6405,
+      "step": 38740
+    },
+    {
+      "epoch": 2.8874813710879286,
+      "grad_norm": 2.527547597885132,
+      "learning_rate": 0.0002,
+      "loss": 2.6174,
+      "step": 38750
+    },
+    {
+      "epoch": 2.88822652757079,
+      "grad_norm": 2.4176671504974365,
+      "learning_rate": 0.0002,
+      "loss": 2.4597,
+      "step": 38760
+    },
+    {
+      "epoch": 2.8889716840536512,
+      "grad_norm": 2.7073466777801514,
+      "learning_rate": 0.0002,
+      "loss": 2.5065,
+      "step": 38770
+    },
+    {
+      "epoch": 2.889716840536513,
+      "grad_norm": 2.439682722091675,
+      "learning_rate": 0.0002,
+      "loss": 2.3043,
+      "step": 38780
+    },
+    {
+      "epoch": 2.890461997019374,
+      "grad_norm": 2.4189679622650146,
+      "learning_rate": 0.0002,
+      "loss": 2.4287,
+      "step": 38790
+    },
+    {
+      "epoch": 2.8912071535022354,
+      "grad_norm": 2.4912519454956055,
+      "learning_rate": 0.0002,
+      "loss": 2.4595,
+      "step": 38800
+    },
+    {
+      "epoch": 2.891952309985097,
+      "grad_norm": 2.4591786861419678,
+      "learning_rate": 0.0002,
+      "loss": 2.5589,
+      "step": 38810
+    },
+    {
+      "epoch": 2.892697466467958,
+      "grad_norm": 2.469432830810547,
+      "learning_rate": 0.0002,
+      "loss": 2.5173,
+      "step": 38820
+    },
+    {
+      "epoch": 2.8934426229508197,
+      "grad_norm": 2.421070098876953,
+      "learning_rate": 0.0002,
+      "loss": 2.4513,
+      "step": 38830
+    },
+    {
+      "epoch": 2.894187779433681,
+      "grad_norm": 2.4555258750915527,
+      "learning_rate": 0.0002,
+      "loss": 2.6341,
+      "step": 38840
+    },
+    {
+      "epoch": 2.8949329359165423,
+      "grad_norm": 2.3373961448669434,
+      "learning_rate": 0.0002,
+      "loss": 2.5724,
+      "step": 38850
+    },
+    {
+      "epoch": 2.895678092399404,
+      "grad_norm": 2.7534141540527344,
+      "learning_rate": 0.0002,
+      "loss": 2.4472,
+      "step": 38860
+    },
+    {
+      "epoch": 2.8964232488822654,
+      "grad_norm": 2.53497052192688,
+      "learning_rate": 0.0002,
+      "loss": 2.1598,
+      "step": 38870
+    },
+    {
+      "epoch": 2.8971684053651265,
+      "grad_norm": 2.5229415893554688,
+      "learning_rate": 0.0002,
+      "loss": 2.581,
+      "step": 38880
+    },
+    {
+      "epoch": 2.897913561847988,
+      "grad_norm": 2.3807928562164307,
+      "learning_rate": 0.0002,
+      "loss": 2.4987,
+      "step": 38890
+    },
+    {
+      "epoch": 2.898658718330849,
+      "grad_norm": 2.803424596786499,
+      "learning_rate": 0.0002,
+      "loss": 2.6143,
+      "step": 38900
+    },
+    {
+      "epoch": 2.8994038748137108,
+      "grad_norm": 2.6139283180236816,
+      "learning_rate": 0.0002,
+      "loss": 2.403,
+      "step": 38910
+    },
+    {
+      "epoch": 2.9001490312965723,
+      "grad_norm": 2.9887192249298096,
+      "learning_rate": 0.0002,
+      "loss": 2.6108,
+      "step": 38920
+    },
+    {
+      "epoch": 2.900894187779434,
+      "grad_norm": 2.6957292556762695,
+      "learning_rate": 0.0002,
+      "loss": 2.4734,
+      "step": 38930
+    },
+    {
+      "epoch": 2.901639344262295,
+      "grad_norm": 2.6717658042907715,
+      "learning_rate": 0.0002,
+      "loss": 2.5778,
+      "step": 38940
+    },
+    {
+      "epoch": 2.9023845007451565,
+      "grad_norm": 2.75540828704834,
+      "learning_rate": 0.0002,
+      "loss": 2.5948,
+      "step": 38950
+    },
+    {
+      "epoch": 2.9031296572280176,
+      "grad_norm": 2.5205585956573486,
+      "learning_rate": 0.0002,
+      "loss": 2.5108,
+      "step": 38960
+    },
+    {
+      "epoch": 2.903874813710879,
+      "grad_norm": 2.2808632850646973,
+      "learning_rate": 0.0002,
+      "loss": 2.5035,
+      "step": 38970
+    },
+    {
+      "epoch": 2.9046199701937407,
+      "grad_norm": 2.8300139904022217,
+      "learning_rate": 0.0002,
+      "loss": 2.6883,
+      "step": 38980
+    },
+    {
+      "epoch": 2.9053651266766023,
+      "grad_norm": 2.3492043018341064,
+      "learning_rate": 0.0002,
+      "loss": 2.517,
+      "step": 38990
+    },
+    {
+      "epoch": 2.9061102831594634,
+      "grad_norm": 2.677483081817627,
+      "learning_rate": 0.0002,
+      "loss": 2.5173,
+      "step": 39000
+    },
+    {
+      "epoch": 2.906855439642325,
+      "grad_norm": 2.076521635055542,
+      "learning_rate": 0.0002,
+      "loss": 2.3087,
+      "step": 39010
+    },
+    {
+      "epoch": 2.907600596125186,
+      "grad_norm": 2.9172234535217285,
+      "learning_rate": 0.0002,
+      "loss": 2.4004,
+      "step": 39020
+    },
+    {
+      "epoch": 2.9083457526080476,
+      "grad_norm": 2.752596616744995,
+      "learning_rate": 0.0002,
+      "loss": 2.4206,
+      "step": 39030
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 2.586287260055542,
+      "learning_rate": 0.0002,
+      "loss": 2.5331,
+      "step": 39040
+    },
+    {
+      "epoch": 2.9098360655737707,
+      "grad_norm": 2.6482787132263184,
+      "learning_rate": 0.0002,
+      "loss": 2.5963,
+      "step": 39050
+    },
+    {
+      "epoch": 2.910581222056632,
+      "grad_norm": 2.513068914413452,
+      "learning_rate": 0.0002,
+      "loss": 2.6465,
+      "step": 39060
+    },
+    {
+      "epoch": 2.9113263785394934,
+      "grad_norm": 2.8123700618743896,
+      "learning_rate": 0.0002,
+      "loss": 2.5256,
+      "step": 39070
+    },
+    {
+      "epoch": 2.9120715350223545,
+      "grad_norm": 2.57496976852417,
+      "learning_rate": 0.0002,
+      "loss": 2.4337,
+      "step": 39080
+    },
+    {
+      "epoch": 2.912816691505216,
+      "grad_norm": 2.719500780105591,
+      "learning_rate": 0.0002,
+      "loss": 2.672,
+      "step": 39090
+    },
+    {
+      "epoch": 2.9135618479880776,
+      "grad_norm": 2.6564457416534424,
+      "learning_rate": 0.0002,
+      "loss": 2.5579,
+      "step": 39100
+    },
+    {
+      "epoch": 2.914307004470939,
+      "grad_norm": 2.656972646713257,
+      "learning_rate": 0.0002,
+      "loss": 2.532,
+      "step": 39110
+    },
+    {
+      "epoch": 2.9150521609538003,
+      "grad_norm": 2.5278375148773193,
+      "learning_rate": 0.0002,
+      "loss": 2.3322,
+      "step": 39120
+    },
+    {
+      "epoch": 2.915797317436662,
+      "grad_norm": 2.6466145515441895,
+      "learning_rate": 0.0002,
+      "loss": 2.5936,
+      "step": 39130
+    },
+    {
+      "epoch": 2.916542473919523,
+      "grad_norm": 2.8328826427459717,
+      "learning_rate": 0.0002,
+      "loss": 2.5601,
+      "step": 39140
+    },
+    {
+      "epoch": 2.9172876304023845,
+      "grad_norm": 2.4768784046173096,
+      "learning_rate": 0.0002,
+      "loss": 2.4364,
+      "step": 39150
+    },
+    {
+      "epoch": 2.918032786885246,
+      "grad_norm": 3.1645164489746094,
+      "learning_rate": 0.0002,
+      "loss": 2.5898,
+      "step": 39160
+    },
+    {
+      "epoch": 2.918777943368107,
+      "grad_norm": 2.8019328117370605,
+      "learning_rate": 0.0002,
+      "loss": 2.5901,
+      "step": 39170
+    },
+    {
+      "epoch": 2.9195230998509687,
+      "grad_norm": 3.203110933303833,
+      "learning_rate": 0.0002,
+      "loss": 2.4048,
+      "step": 39180
+    },
+    {
+      "epoch": 2.9202682563338302,
+      "grad_norm": 3.061889410018921,
+      "learning_rate": 0.0002,
+      "loss": 2.6781,
+      "step": 39190
+    },
+    {
+      "epoch": 2.9210134128166914,
+      "grad_norm": 2.735222101211548,
+      "learning_rate": 0.0002,
+      "loss": 2.5379,
+      "step": 39200
+    },
+    {
+      "epoch": 2.921758569299553,
+      "grad_norm": 2.0281805992126465,
+      "learning_rate": 0.0002,
+      "loss": 2.4894,
+      "step": 39210
+    },
+    {
+      "epoch": 2.9225037257824145,
+      "grad_norm": 2.89827561378479,
+      "learning_rate": 0.0002,
+      "loss": 2.3229,
+      "step": 39220
+    },
+    {
+      "epoch": 2.9232488822652756,
+      "grad_norm": 2.8243911266326904,
+      "learning_rate": 0.0002,
+      "loss": 2.2712,
+      "step": 39230
+    },
+    {
+      "epoch": 2.923994038748137,
+      "grad_norm": 2.6554248332977295,
+      "learning_rate": 0.0002,
+      "loss": 2.4221,
+      "step": 39240
+    },
+    {
+      "epoch": 2.9247391952309982,
+      "grad_norm": 2.6550304889678955,
+      "learning_rate": 0.0002,
+      "loss": 2.5087,
+      "step": 39250
+    },
+    {
+      "epoch": 2.92548435171386,
+      "grad_norm": 2.154710054397583,
+      "learning_rate": 0.0002,
+      "loss": 2.4395,
+      "step": 39260
+    },
+    {
+      "epoch": 2.9262295081967213,
+      "grad_norm": 2.5351250171661377,
+      "learning_rate": 0.0002,
+      "loss": 2.7179,
+      "step": 39270
+    },
+    {
+      "epoch": 2.926974664679583,
+      "grad_norm": 2.533205986022949,
+      "learning_rate": 0.0002,
+      "loss": 2.5654,
+      "step": 39280
+    },
+    {
+      "epoch": 2.927719821162444,
+      "grad_norm": 2.13325572013855,
+      "learning_rate": 0.0002,
+      "loss": 2.6483,
+      "step": 39290
+    },
+    {
+      "epoch": 2.9284649776453056,
+      "grad_norm": 2.66166353225708,
+      "learning_rate": 0.0002,
+      "loss": 2.5622,
+      "step": 39300
+    },
+    {
+      "epoch": 2.9292101341281667,
+      "grad_norm": 2.33381724357605,
+      "learning_rate": 0.0002,
+      "loss": 2.555,
+      "step": 39310
+    },
+    {
+      "epoch": 2.929955290611028,
+      "grad_norm": 2.7781875133514404,
+      "learning_rate": 0.0002,
+      "loss": 2.6041,
+      "step": 39320
+    },
+    {
+      "epoch": 2.9307004470938898,
+      "grad_norm": 2.579591989517212,
+      "learning_rate": 0.0002,
+      "loss": 2.4904,
+      "step": 39330
+    },
+    {
+      "epoch": 2.9314456035767513,
+      "grad_norm": 2.5462992191314697,
+      "learning_rate": 0.0002,
+      "loss": 2.6305,
+      "step": 39340
+    },
+    {
+      "epoch": 2.9321907600596124,
+      "grad_norm": 2.6036219596862793,
+      "learning_rate": 0.0002,
+      "loss": 2.6255,
+      "step": 39350
+    },
+    {
+      "epoch": 2.932935916542474,
+      "grad_norm": 2.4048843383789062,
+      "learning_rate": 0.0002,
+      "loss": 2.4814,
+      "step": 39360
+    },
+    {
+      "epoch": 2.933681073025335,
+      "grad_norm": 2.5451865196228027,
+      "learning_rate": 0.0002,
+      "loss": 2.4788,
+      "step": 39370
+    },
+    {
+      "epoch": 2.9344262295081966,
+      "grad_norm": 2.7191903591156006,
+      "learning_rate": 0.0002,
+      "loss": 2.5476,
+      "step": 39380
+    },
+    {
+      "epoch": 2.935171385991058,
+      "grad_norm": 2.3997859954833984,
+      "learning_rate": 0.0002,
+      "loss": 2.6285,
+      "step": 39390
+    },
+    {
+      "epoch": 2.9359165424739198,
+      "grad_norm": 2.5516610145568848,
+      "learning_rate": 0.0002,
+      "loss": 2.5219,
+      "step": 39400
+    },
+    {
+      "epoch": 2.936661698956781,
+      "grad_norm": 2.5959670543670654,
+      "learning_rate": 0.0002,
+      "loss": 2.5965,
+      "step": 39410
+    },
+    {
+      "epoch": 2.9374068554396424,
+      "grad_norm": 2.9444220066070557,
+      "learning_rate": 0.0002,
+      "loss": 2.5746,
+      "step": 39420
+    },
+    {
+      "epoch": 2.9381520119225035,
+      "grad_norm": 2.497292995452881,
+      "learning_rate": 0.0002,
+      "loss": 2.5914,
+      "step": 39430
+    },
+    {
+      "epoch": 2.938897168405365,
+      "grad_norm": 2.524833917617798,
+      "learning_rate": 0.0002,
+      "loss": 2.581,
+      "step": 39440
+    },
+    {
+      "epoch": 2.9396423248882266,
+      "grad_norm": 2.8581857681274414,
+      "learning_rate": 0.0002,
+      "loss": 2.5388,
+      "step": 39450
+    },
+    {
+      "epoch": 2.940387481371088,
+      "grad_norm": 2.8836469650268555,
+      "learning_rate": 0.0002,
+      "loss": 2.5906,
+      "step": 39460
+    },
+    {
+      "epoch": 2.9411326378539493,
+      "grad_norm": 2.530555248260498,
+      "learning_rate": 0.0002,
+      "loss": 2.6937,
+      "step": 39470
+    },
+    {
+      "epoch": 2.941877794336811,
+      "grad_norm": 2.8848953247070312,
+      "learning_rate": 0.0002,
+      "loss": 2.4355,
+      "step": 39480
+    },
+    {
+      "epoch": 2.942622950819672,
+      "grad_norm": 2.4546191692352295,
+      "learning_rate": 0.0002,
+      "loss": 2.5684,
+      "step": 39490
+    },
+    {
+      "epoch": 2.9433681073025335,
+      "grad_norm": 3.093285083770752,
+      "learning_rate": 0.0002,
+      "loss": 2.4457,
+      "step": 39500
+    },
+    {
+      "epoch": 2.944113263785395,
+      "grad_norm": 2.597832679748535,
+      "learning_rate": 0.0002,
+      "loss": 2.402,
+      "step": 39510
+    },
+    {
+      "epoch": 2.944858420268256,
+      "grad_norm": 2.5404412746429443,
+      "learning_rate": 0.0002,
+      "loss": 2.5724,
+      "step": 39520
+    },
+    {
+      "epoch": 2.9456035767511177,
+      "grad_norm": 3.019563674926758,
+      "learning_rate": 0.0002,
+      "loss": 2.3523,
+      "step": 39530
+    },
+    {
+      "epoch": 2.9463487332339793,
+      "grad_norm": 3.1213128566741943,
+      "learning_rate": 0.0002,
+      "loss": 2.6512,
+      "step": 39540
+    },
+    {
+      "epoch": 2.9470938897168404,
+      "grad_norm": 2.413450002670288,
+      "learning_rate": 0.0002,
+      "loss": 2.4045,
+      "step": 39550
+    },
+    {
+      "epoch": 2.947839046199702,
+      "grad_norm": 2.7260587215423584,
+      "learning_rate": 0.0002,
+      "loss": 2.6666,
+      "step": 39560
+    },
+    {
+      "epoch": 2.9485842026825635,
+      "grad_norm": 3.4645745754241943,
+      "learning_rate": 0.0002,
+      "loss": 2.6023,
+      "step": 39570
+    },
+    {
+      "epoch": 2.9493293591654246,
+      "grad_norm": 2.8158528804779053,
+      "learning_rate": 0.0002,
+      "loss": 2.4214,
+      "step": 39580
+    },
+    {
+      "epoch": 2.950074515648286,
+      "grad_norm": 2.712031841278076,
+      "learning_rate": 0.0002,
+      "loss": 2.5187,
+      "step": 39590
+    },
+    {
+      "epoch": 2.9508196721311473,
+      "grad_norm": 2.5256295204162598,
+      "learning_rate": 0.0002,
+      "loss": 2.4577,
+      "step": 39600
+    },
+    {
+      "epoch": 2.951564828614009,
+      "grad_norm": 2.72895884513855,
+      "learning_rate": 0.0002,
+      "loss": 2.5944,
+      "step": 39610
+    },
+    {
+      "epoch": 2.9523099850968704,
+      "grad_norm": 2.040031909942627,
+      "learning_rate": 0.0002,
+      "loss": 2.4583,
+      "step": 39620
+    },
+    {
+      "epoch": 2.953055141579732,
+      "grad_norm": 2.666433811187744,
+      "learning_rate": 0.0002,
+      "loss": 2.5275,
+      "step": 39630
+    },
+    {
+      "epoch": 2.953800298062593,
+      "grad_norm": 2.6652700901031494,
+      "learning_rate": 0.0002,
+      "loss": 2.3713,
+      "step": 39640
+    },
+    {
+      "epoch": 2.9545454545454546,
+      "grad_norm": 2.743901014328003,
+      "learning_rate": 0.0002,
+      "loss": 2.5227,
+      "step": 39650
+    },
+    {
+      "epoch": 2.9552906110283157,
+      "grad_norm": 2.845050811767578,
+      "learning_rate": 0.0002,
+      "loss": 2.5615,
+      "step": 39660
+    },
+    {
+      "epoch": 2.9560357675111772,
+      "grad_norm": 2.8907158374786377,
+      "learning_rate": 0.0002,
+      "loss": 2.5205,
+      "step": 39670
+    },
+    {
+      "epoch": 2.956780923994039,
+      "grad_norm": 2.6992757320404053,
+      "learning_rate": 0.0002,
+      "loss": 2.4661,
+      "step": 39680
+    },
+    {
+      "epoch": 2.9575260804769004,
+      "grad_norm": 2.9309325218200684,
+      "learning_rate": 0.0002,
+      "loss": 2.6417,
+      "step": 39690
+    },
+    {
+      "epoch": 2.9582712369597615,
+      "grad_norm": 2.801201581954956,
+      "learning_rate": 0.0002,
+      "loss": 2.495,
+      "step": 39700
+    },
+    {
+      "epoch": 2.959016393442623,
+      "grad_norm": 2.763702869415283,
+      "learning_rate": 0.0002,
+      "loss": 2.4012,
+      "step": 39710
+    },
+    {
+      "epoch": 2.959761549925484,
+      "grad_norm": 2.6679039001464844,
+      "learning_rate": 0.0002,
+      "loss": 2.6976,
+      "step": 39720
+    },
+    {
+      "epoch": 2.9605067064083457,
+      "grad_norm": 2.567474603652954,
+      "learning_rate": 0.0002,
+      "loss": 2.6038,
+      "step": 39730
+    },
+    {
+      "epoch": 2.9612518628912072,
+      "grad_norm": 3.114344596862793,
+      "learning_rate": 0.0002,
+      "loss": 2.4868,
+      "step": 39740
+    },
+    {
+      "epoch": 2.961997019374069,
+      "grad_norm": 2.5168049335479736,
+      "learning_rate": 0.0002,
+      "loss": 2.5279,
+      "step": 39750
+    },
+    {
+      "epoch": 2.96274217585693,
+      "grad_norm": 3.0251882076263428,
+      "learning_rate": 0.0002,
+      "loss": 2.6403,
+      "step": 39760
+    },
+    {
+      "epoch": 2.9634873323397914,
+      "grad_norm": 2.606680393218994,
+      "learning_rate": 0.0002,
+      "loss": 2.6361,
+      "step": 39770
+    },
+    {
+      "epoch": 2.9642324888226526,
+      "grad_norm": 2.699082612991333,
+      "learning_rate": 0.0002,
+      "loss": 2.5908,
+      "step": 39780
+    },
+    {
+      "epoch": 2.964977645305514,
+      "grad_norm": 2.8698062896728516,
+      "learning_rate": 0.0002,
+      "loss": 2.4495,
+      "step": 39790
+    },
+    {
+      "epoch": 2.9657228017883757,
+      "grad_norm": 2.898927927017212,
+      "learning_rate": 0.0002,
+      "loss": 2.4352,
+      "step": 39800
+    },
+    {
+      "epoch": 2.966467958271237,
+      "grad_norm": 2.3014628887176514,
+      "learning_rate": 0.0002,
+      "loss": 2.5411,
+      "step": 39810
+    },
+    {
+      "epoch": 2.9672131147540983,
+      "grad_norm": 2.9208085536956787,
+      "learning_rate": 0.0002,
+      "loss": 2.4678,
+      "step": 39820
+    },
+    {
+      "epoch": 2.96795827123696,
+      "grad_norm": 2.51240873336792,
+      "learning_rate": 0.0002,
+      "loss": 2.101,
+      "step": 39830
+    },
+    {
+      "epoch": 2.968703427719821,
+      "grad_norm": 2.7985494136810303,
+      "learning_rate": 0.0002,
+      "loss": 2.6076,
+      "step": 39840
+    },
+    {
+      "epoch": 2.9694485842026825,
+      "grad_norm": 2.5904476642608643,
+      "learning_rate": 0.0002,
+      "loss": 2.4389,
+      "step": 39850
+    },
+    {
+      "epoch": 2.970193740685544,
+      "grad_norm": 2.6392509937286377,
+      "learning_rate": 0.0002,
+      "loss": 2.3973,
+      "step": 39860
+    },
+    {
+      "epoch": 2.970938897168405,
+      "grad_norm": 1.774827241897583,
+      "learning_rate": 0.0002,
+      "loss": 2.4029,
+      "step": 39870
+    },
+    {
+      "epoch": 2.9716840536512668,
+      "grad_norm": 2.54217529296875,
+      "learning_rate": 0.0002,
+      "loss": 2.6573,
+      "step": 39880
+    },
+    {
+      "epoch": 2.9724292101341283,
+      "grad_norm": 2.6354312896728516,
+      "learning_rate": 0.0002,
+      "loss": 2.3592,
+      "step": 39890
+    },
+    {
+      "epoch": 2.9731743666169894,
+      "grad_norm": 2.6568002700805664,
+      "learning_rate": 0.0002,
+      "loss": 2.5426,
+      "step": 39900
+    },
+    {
+      "epoch": 2.973919523099851,
+      "grad_norm": 2.660454273223877,
+      "learning_rate": 0.0002,
+      "loss": 2.4449,
+      "step": 39910
+    },
+    {
+      "epoch": 2.9746646795827125,
+      "grad_norm": 2.9466612339019775,
+      "learning_rate": 0.0002,
+      "loss": 2.48,
+      "step": 39920
+    },
+    {
+      "epoch": 2.9754098360655736,
+      "grad_norm": 2.5808732509613037,
+      "learning_rate": 0.0002,
+      "loss": 2.6554,
+      "step": 39930
+    },
+    {
+      "epoch": 2.976154992548435,
+      "grad_norm": 2.8035290241241455,
+      "learning_rate": 0.0002,
+      "loss": 2.8002,
+      "step": 39940
+    },
+    {
+      "epoch": 2.9769001490312967,
+      "grad_norm": 2.615812301635742,
+      "learning_rate": 0.0002,
+      "loss": 2.4047,
+      "step": 39950
+    },
+    {
+      "epoch": 2.977645305514158,
+      "grad_norm": 2.588972330093384,
+      "learning_rate": 0.0002,
+      "loss": 2.495,
+      "step": 39960
+    },
+    {
+      "epoch": 2.9783904619970194,
+      "grad_norm": 2.669950008392334,
+      "learning_rate": 0.0002,
+      "loss": 2.5453,
+      "step": 39970
+    },
+    {
+      "epoch": 2.979135618479881,
+      "grad_norm": 2.609656572341919,
+      "learning_rate": 0.0002,
+      "loss": 2.549,
+      "step": 39980
+    },
+    {
+      "epoch": 2.979880774962742,
+      "grad_norm": 2.9678401947021484,
+      "learning_rate": 0.0002,
+      "loss": 2.5067,
+      "step": 39990
+    },
+    {
+      "epoch": 2.9806259314456036,
+      "grad_norm": 1.8499410152435303,
+      "learning_rate": 0.0002,
+      "loss": 2.5054,
+      "step": 40000
+    },
+    {
+      "epoch": 2.9813710879284647,
+      "grad_norm": 2.4903769493103027,
+      "learning_rate": 0.0002,
+      "loss": 2.4195,
+      "step": 40010
+    },
+    {
+      "epoch": 2.9821162444113263,
+      "grad_norm": 2.87322998046875,
+      "learning_rate": 0.0002,
+      "loss": 2.4816,
+      "step": 40020
+    },
+    {
+      "epoch": 2.982861400894188,
+      "grad_norm": 2.6124825477600098,
+      "learning_rate": 0.0002,
+      "loss": 2.3755,
+      "step": 40030
+    },
+    {
+      "epoch": 2.9836065573770494,
+      "grad_norm": 2.674150228500366,
+      "learning_rate": 0.0002,
+      "loss": 2.4891,
+      "step": 40040
+    },
+    {
+      "epoch": 2.9843517138599105,
+      "grad_norm": 2.499284267425537,
+      "learning_rate": 0.0002,
+      "loss": 2.5349,
+      "step": 40050
+    },
+    {
+      "epoch": 2.985096870342772,
+      "grad_norm": 2.7705399990081787,
+      "learning_rate": 0.0002,
+      "loss": 2.4778,
+      "step": 40060
+    },
+    {
+      "epoch": 2.985842026825633,
+      "grad_norm": 2.2329294681549072,
+      "learning_rate": 0.0002,
+      "loss": 2.4104,
+      "step": 40070
+    },
+    {
+      "epoch": 2.9865871833084947,
+      "grad_norm": 2.8790485858917236,
+      "learning_rate": 0.0002,
+      "loss": 2.4753,
+      "step": 40080
+    },
+    {
+      "epoch": 2.9873323397913563,
+      "grad_norm": 2.469627618789673,
+      "learning_rate": 0.0002,
+      "loss": 2.6872,
+      "step": 40090
+    },
+    {
+      "epoch": 2.988077496274218,
+      "grad_norm": 2.38393235206604,
+      "learning_rate": 0.0002,
+      "loss": 2.6073,
+      "step": 40100
+    },
+    {
+      "epoch": 2.988822652757079,
+      "grad_norm": 2.2364330291748047,
+      "learning_rate": 0.0002,
+      "loss": 2.4443,
+      "step": 40110
+    },
+    {
+      "epoch": 2.9895678092399405,
+      "grad_norm": 2.5776116847991943,
+      "learning_rate": 0.0002,
+      "loss": 2.532,
+      "step": 40120
+    },
+    {
+      "epoch": 2.9903129657228016,
+      "grad_norm": 2.5935330390930176,
+      "learning_rate": 0.0002,
+      "loss": 2.4889,
+      "step": 40130
+    },
+    {
+      "epoch": 2.991058122205663,
+      "grad_norm": 2.6229746341705322,
+      "learning_rate": 0.0002,
+      "loss": 2.5562,
+      "step": 40140
+    },
+    {
+      "epoch": 2.9918032786885247,
+      "grad_norm": 2.5080349445343018,
+      "learning_rate": 0.0002,
+      "loss": 2.4328,
+      "step": 40150
+    },
+    {
+      "epoch": 2.9925484351713862,
+      "grad_norm": 2.5937001705169678,
+      "learning_rate": 0.0002,
+      "loss": 2.591,
+      "step": 40160
+    },
+    {
+      "epoch": 2.9932935916542474,
+      "grad_norm": 2.3092591762542725,
+      "learning_rate": 0.0002,
+      "loss": 2.4414,
+      "step": 40170
+    },
+    {
+      "epoch": 2.994038748137109,
+      "grad_norm": 2.848226308822632,
+      "learning_rate": 0.0002,
+      "loss": 2.4109,
+      "step": 40180
+    },
+    {
+      "epoch": 2.99478390461997,
+      "grad_norm": 2.3122453689575195,
+      "learning_rate": 0.0002,
+      "loss": 2.3277,
+      "step": 40190
+    },
+    {
+      "epoch": 2.9955290611028316,
+      "grad_norm": 2.513367176055908,
+      "learning_rate": 0.0002,
+      "loss": 2.5037,
+      "step": 40200
+    },
+    {
+      "epoch": 2.996274217585693,
+      "grad_norm": 2.57222843170166,
+      "learning_rate": 0.0002,
+      "loss": 2.5487,
+      "step": 40210
+    },
+    {
+      "epoch": 2.9970193740685542,
+      "grad_norm": 2.354405164718628,
+      "learning_rate": 0.0002,
+      "loss": 2.386,
+      "step": 40220
+    },
+    {
+      "epoch": 2.997764530551416,
+      "grad_norm": 2.828702688217163,
+      "learning_rate": 0.0002,
+      "loss": 2.3264,
+      "step": 40230
+    },
+    {
+      "epoch": 2.9985096870342773,
+      "grad_norm": 2.779327392578125,
+      "learning_rate": 0.0002,
+      "loss": 2.3778,
+      "step": 40240
+    },
+    {
+      "epoch": 2.9992548435171384,
+      "grad_norm": 2.4226698875427246,
+      "learning_rate": 0.0002,
+      "loss": 2.1855,
+      "step": 40250
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.918992042541504,
+      "learning_rate": 0.0002,
+      "loss": 2.4347,
+      "step": 40260
+    },
+    {
+      "epoch": 3.0,
+      "eval_runtime": 2763.3152,
+      "eval_samples_per_second": 4.856,
+      "eval_steps_per_second": 0.607,
+      "step": 40260
+    },
+    {
+      "epoch": 3.0007451564828616,
+      "grad_norm": 2.5024285316467285,
+      "learning_rate": 0.0002,
+      "loss": 2.3297,
+      "step": 40270
+    },
+    {
+      "epoch": 3.0014903129657227,
+      "grad_norm": 2.8226559162139893,
+      "learning_rate": 0.0002,
+      "loss": 2.3661,
+      "step": 40280
+    },
+    {
+      "epoch": 3.002235469448584,
+      "grad_norm": 2.758744716644287,
+      "learning_rate": 0.0002,
+      "loss": 2.4495,
+      "step": 40290
+    },
+    {
+      "epoch": 3.0029806259314458,
+      "grad_norm": 2.426999568939209,
+      "learning_rate": 0.0002,
+      "loss": 2.2649,
+      "step": 40300
+    },
+    {
+      "epoch": 3.003725782414307,
+      "grad_norm": 2.6783957481384277,
+      "learning_rate": 0.0002,
+      "loss": 2.3838,
+      "step": 40310
+    },
+    {
+      "epoch": 3.0044709388971684,
+      "grad_norm": 3.014652967453003,
+      "learning_rate": 0.0002,
+      "loss": 2.1913,
+      "step": 40320
+    },
+    {
+      "epoch": 3.00521609538003,
+      "grad_norm": 2.7266387939453125,
+      "learning_rate": 0.0002,
+      "loss": 2.3369,
+      "step": 40330
+    },
+    {
+      "epoch": 3.005961251862891,
+      "grad_norm": 2.7019906044006348,
+      "learning_rate": 0.0002,
+      "loss": 2.1718,
+      "step": 40340
+    },
+    {
+      "epoch": 3.0067064083457526,
+      "grad_norm": 2.5368382930755615,
+      "learning_rate": 0.0002,
+      "loss": 2.4286,
+      "step": 40350
+    },
+    {
+      "epoch": 3.007451564828614,
+      "grad_norm": 2.9685909748077393,
+      "learning_rate": 0.0002,
+      "loss": 2.3881,
+      "step": 40360
+    },
+    {
+      "epoch": 3.0081967213114753,
+      "grad_norm": 2.656423330307007,
+      "learning_rate": 0.0002,
+      "loss": 2.3198,
+      "step": 40370
+    },
+    {
+      "epoch": 3.008941877794337,
+      "grad_norm": 2.7926347255706787,
+      "learning_rate": 0.0002,
+      "loss": 2.2615,
+      "step": 40380
+    },
+    {
+      "epoch": 3.0096870342771984,
+      "grad_norm": 3.065396308898926,
+      "learning_rate": 0.0002,
+      "loss": 2.4156,
+      "step": 40390
+    },
+    {
+      "epoch": 3.0104321907600595,
+      "grad_norm": 2.8727266788482666,
+      "learning_rate": 0.0002,
+      "loss": 2.429,
+      "step": 40400
+    },
+    {
+      "epoch": 3.011177347242921,
+      "grad_norm": 2.9066572189331055,
+      "learning_rate": 0.0002,
+      "loss": 2.3824,
+      "step": 40410
+    },
+    {
+      "epoch": 3.0119225037257826,
+      "grad_norm": 2.813734769821167,
+      "learning_rate": 0.0002,
+      "loss": 2.2325,
+      "step": 40420
+    },
+    {
+      "epoch": 3.0126676602086437,
+      "grad_norm": 2.808375120162964,
+      "learning_rate": 0.0002,
+      "loss": 2.3175,
+      "step": 40430
+    },
+    {
+      "epoch": 3.0134128166915053,
+      "grad_norm": 2.6436479091644287,
+      "learning_rate": 0.0002,
+      "loss": 2.4796,
+      "step": 40440
+    },
+    {
+      "epoch": 3.0141579731743664,
+      "grad_norm": 2.557705879211426,
+      "learning_rate": 0.0002,
+      "loss": 2.5264,
+      "step": 40450
+    },
+    {
+      "epoch": 3.014903129657228,
+      "grad_norm": 2.7825839519500732,
+      "learning_rate": 0.0002,
+      "loss": 2.4484,
+      "step": 40460
+    },
+    {
+      "epoch": 3.0156482861400895,
+      "grad_norm": 2.4266088008880615,
+      "learning_rate": 0.0002,
+      "loss": 2.4146,
+      "step": 40470
+    },
+    {
+      "epoch": 3.0163934426229506,
+      "grad_norm": 2.8366827964782715,
+      "learning_rate": 0.0002,
+      "loss": 2.3706,
+      "step": 40480
+    },
+    {
+      "epoch": 3.017138599105812,
+      "grad_norm": 2.837256669998169,
+      "learning_rate": 0.0002,
+      "loss": 2.4929,
+      "step": 40490
+    },
+    {
+      "epoch": 3.0178837555886737,
+      "grad_norm": 2.7179813385009766,
+      "learning_rate": 0.0002,
+      "loss": 2.365,
+      "step": 40500
+    },
+    {
+      "epoch": 3.018628912071535,
+      "grad_norm": 2.3554391860961914,
+      "learning_rate": 0.0002,
+      "loss": 2.2821,
+      "step": 40510
+    },
+    {
+      "epoch": 3.0193740685543964,
+      "grad_norm": 2.696993112564087,
+      "learning_rate": 0.0002,
+      "loss": 2.5171,
+      "step": 40520
+    },
+    {
+      "epoch": 3.020119225037258,
+      "grad_norm": 2.1214194297790527,
+      "learning_rate": 0.0002,
+      "loss": 2.2159,
+      "step": 40530
+    },
+    {
+      "epoch": 3.020864381520119,
+      "grad_norm": 2.8695015907287598,
+      "learning_rate": 0.0002,
+      "loss": 2.3214,
+      "step": 40540
+    },
+    {
+      "epoch": 3.0216095380029806,
+      "grad_norm": 2.551729679107666,
+      "learning_rate": 0.0002,
+      "loss": 2.2563,
+      "step": 40550
+    },
+    {
+      "epoch": 3.022354694485842,
+      "grad_norm": 2.3707385063171387,
+      "learning_rate": 0.0002,
+      "loss": 2.4613,
+      "step": 40560
+    },
+    {
+      "epoch": 3.0230998509687033,
+      "grad_norm": 2.9850101470947266,
+      "learning_rate": 0.0002,
+      "loss": 2.37,
+      "step": 40570
+    },
+    {
+      "epoch": 3.023845007451565,
+      "grad_norm": 2.5724594593048096,
+      "learning_rate": 0.0002,
+      "loss": 2.3877,
+      "step": 40580
+    },
+    {
+      "epoch": 3.0245901639344264,
+      "grad_norm": 2.7039642333984375,
+      "learning_rate": 0.0002,
+      "loss": 2.2501,
+      "step": 40590
+    },
+    {
+      "epoch": 3.0253353204172875,
+      "grad_norm": 2.7257907390594482,
+      "learning_rate": 0.0002,
+      "loss": 2.3512,
+      "step": 40600
+    },
+    {
+      "epoch": 3.026080476900149,
+      "grad_norm": 2.788710355758667,
+      "learning_rate": 0.0002,
+      "loss": 2.3724,
+      "step": 40610
+    },
+    {
+      "epoch": 3.0268256333830106,
+      "grad_norm": 2.3942348957061768,
+      "learning_rate": 0.0002,
+      "loss": 2.288,
+      "step": 40620
+    },
+    {
+      "epoch": 3.0275707898658717,
+      "grad_norm": 2.6751253604888916,
+      "learning_rate": 0.0002,
+      "loss": 2.4059,
+      "step": 40630
+    },
+    {
+      "epoch": 3.0283159463487332,
+      "grad_norm": 3.1774163246154785,
+      "learning_rate": 0.0002,
+      "loss": 2.4086,
+      "step": 40640
+    },
+    {
+      "epoch": 3.029061102831595,
+      "grad_norm": 2.69118595123291,
+      "learning_rate": 0.0002,
+      "loss": 2.4347,
+      "step": 40650
+    },
+    {
+      "epoch": 3.029806259314456,
+      "grad_norm": 2.844470262527466,
+      "learning_rate": 0.0002,
+      "loss": 2.3638,
+      "step": 40660
+    },
+    {
+      "epoch": 3.0305514157973175,
+      "grad_norm": 2.662158966064453,
+      "learning_rate": 0.0002,
+      "loss": 2.3734,
+      "step": 40670
+    },
+    {
+      "epoch": 3.031296572280179,
+      "grad_norm": 2.9637556076049805,
+      "learning_rate": 0.0002,
+      "loss": 2.3224,
+      "step": 40680
+    },
+    {
+      "epoch": 3.03204172876304,
+      "grad_norm": 2.62660813331604,
+      "learning_rate": 0.0002,
+      "loss": 2.4436,
+      "step": 40690
+    },
+    {
+      "epoch": 3.0327868852459017,
+      "grad_norm": 2.9725089073181152,
+      "learning_rate": 0.0002,
+      "loss": 2.5596,
+      "step": 40700
+    },
+    {
+      "epoch": 3.0335320417287632,
+      "grad_norm": 2.9883205890655518,
+      "learning_rate": 0.0002,
+      "loss": 2.2181,
+      "step": 40710
+    },
+    {
+      "epoch": 3.0342771982116243,
+      "grad_norm": 2.632417917251587,
+      "learning_rate": 0.0002,
+      "loss": 2.3767,
+      "step": 40720
+    },
+    {
+      "epoch": 3.035022354694486,
+      "grad_norm": 2.510411262512207,
+      "learning_rate": 0.0002,
+      "loss": 2.3296,
+      "step": 40730
+    },
+    {
+      "epoch": 3.0357675111773474,
+      "grad_norm": 2.5027692317962646,
+      "learning_rate": 0.0002,
+      "loss": 2.2664,
+      "step": 40740
+    },
+    {
+      "epoch": 3.0365126676602086,
+      "grad_norm": 2.8952243328094482,
+      "learning_rate": 0.0002,
+      "loss": 2.3159,
+      "step": 40750
+    },
+    {
+      "epoch": 3.03725782414307,
+      "grad_norm": 2.4545536041259766,
+      "learning_rate": 0.0002,
+      "loss": 2.5346,
+      "step": 40760
+    },
+    {
+      "epoch": 3.0380029806259317,
+      "grad_norm": 2.8499226570129395,
+      "learning_rate": 0.0002,
+      "loss": 2.3085,
+      "step": 40770
+    },
+    {
+      "epoch": 3.0387481371087928,
+      "grad_norm": 3.025698184967041,
+      "learning_rate": 0.0002,
+      "loss": 2.5594,
+      "step": 40780
+    },
+    {
+      "epoch": 3.0394932935916543,
+      "grad_norm": 2.6746528148651123,
+      "learning_rate": 0.0002,
+      "loss": 2.4489,
+      "step": 40790
+    },
+    {
+      "epoch": 3.0402384500745154,
+      "grad_norm": 3.0967001914978027,
+      "learning_rate": 0.0002,
+      "loss": 2.2821,
+      "step": 40800
+    },
+    {
+      "epoch": 3.040983606557377,
+      "grad_norm": 2.5189292430877686,
+      "learning_rate": 0.0002,
+      "loss": 2.1087,
+      "step": 40810
+    },
+    {
+      "epoch": 3.0417287630402385,
+      "grad_norm": 2.407550573348999,
+      "learning_rate": 0.0002,
+      "loss": 2.4971,
+      "step": 40820
+    },
+    {
+      "epoch": 3.0424739195230996,
+      "grad_norm": 2.7270169258117676,
+      "learning_rate": 0.0002,
+      "loss": 2.3941,
+      "step": 40830
+    },
+    {
+      "epoch": 3.043219076005961,
+      "grad_norm": 2.8451573848724365,
+      "learning_rate": 0.0002,
+      "loss": 2.4935,
+      "step": 40840
+    },
+    {
+      "epoch": 3.0439642324888228,
+      "grad_norm": 2.520591974258423,
+      "learning_rate": 0.0002,
+      "loss": 2.4202,
+      "step": 40850
+    },
+    {
+      "epoch": 3.044709388971684,
+      "grad_norm": 2.5729188919067383,
+      "learning_rate": 0.0002,
+      "loss": 2.3634,
+      "step": 40860
+    },
+    {
+      "epoch": 3.0454545454545454,
+      "grad_norm": 2.5250637531280518,
+      "learning_rate": 0.0002,
+      "loss": 2.3602,
+      "step": 40870
+    },
+    {
+      "epoch": 3.046199701937407,
+      "grad_norm": 2.7925519943237305,
+      "learning_rate": 0.0002,
+      "loss": 2.4729,
+      "step": 40880
+    },
+    {
+      "epoch": 3.046944858420268,
+      "grad_norm": 2.4078192710876465,
+      "learning_rate": 0.0002,
+      "loss": 2.4195,
+      "step": 40890
+    },
+    {
+      "epoch": 3.0476900149031296,
+      "grad_norm": 2.648886203765869,
+      "learning_rate": 0.0002,
+      "loss": 2.4189,
+      "step": 40900
+    },
+    {
+      "epoch": 3.048435171385991,
+      "grad_norm": 2.811570167541504,
+      "learning_rate": 0.0002,
+      "loss": 2.3038,
+      "step": 40910
+    },
+    {
+      "epoch": 3.0491803278688523,
+      "grad_norm": 2.7872612476348877,
+      "learning_rate": 0.0002,
+      "loss": 2.3317,
+      "step": 40920
+    },
+    {
+      "epoch": 3.049925484351714,
+      "grad_norm": 2.933014154434204,
+      "learning_rate": 0.0002,
+      "loss": 2.398,
+      "step": 40930
+    },
+    {
+      "epoch": 3.0506706408345754,
+      "grad_norm": 2.8727405071258545,
+      "learning_rate": 0.0002,
+      "loss": 2.2654,
+      "step": 40940
+    },
+    {
+      "epoch": 3.0514157973174365,
+      "grad_norm": 2.7194955348968506,
+      "learning_rate": 0.0002,
+      "loss": 2.5749,
+      "step": 40950
+    },
+    {
+      "epoch": 3.052160953800298,
+      "grad_norm": 2.616337776184082,
+      "learning_rate": 0.0002,
+      "loss": 2.4542,
+      "step": 40960
+    },
+    {
+      "epoch": 3.0529061102831596,
+      "grad_norm": 2.4622654914855957,
+      "learning_rate": 0.0002,
+      "loss": 2.3492,
+      "step": 40970
+    },
+    {
+      "epoch": 3.0536512667660207,
+      "grad_norm": 2.7078351974487305,
+      "learning_rate": 0.0002,
+      "loss": 2.3654,
+      "step": 40980
+    },
+    {
+      "epoch": 3.0543964232488823,
+      "grad_norm": 2.5035488605499268,
+      "learning_rate": 0.0002,
+      "loss": 2.4756,
+      "step": 40990
+    },
+    {
+      "epoch": 3.055141579731744,
+      "grad_norm": 2.7884819507598877,
+      "learning_rate": 0.0002,
+      "loss": 2.3663,
+      "step": 41000
+    },
+    {
+      "epoch": 3.055886736214605,
+      "grad_norm": 2.593498945236206,
+      "learning_rate": 0.0002,
+      "loss": 2.3312,
+      "step": 41010
+    },
+    {
+      "epoch": 3.0566318926974665,
+      "grad_norm": 2.6816306114196777,
+      "learning_rate": 0.0002,
+      "loss": 2.4835,
+      "step": 41020
+    },
+    {
+      "epoch": 3.057377049180328,
+      "grad_norm": 2.6609058380126953,
+      "learning_rate": 0.0002,
+      "loss": 2.4931,
+      "step": 41030
+    },
+    {
+      "epoch": 3.058122205663189,
+      "grad_norm": 2.7509803771972656,
+      "learning_rate": 0.0002,
+      "loss": 2.3721,
+      "step": 41040
+    },
+    {
+      "epoch": 3.0588673621460507,
+      "grad_norm": 2.3781402111053467,
+      "learning_rate": 0.0002,
+      "loss": 2.3468,
+      "step": 41050
+    },
+    {
+      "epoch": 3.0596125186289123,
+      "grad_norm": 2.922632932662964,
+      "learning_rate": 0.0002,
+      "loss": 2.525,
+      "step": 41060
+    },
+    {
+      "epoch": 3.0603576751117734,
+      "grad_norm": 2.792450189590454,
+      "learning_rate": 0.0002,
+      "loss": 2.2651,
+      "step": 41070
+    },
+    {
+      "epoch": 3.061102831594635,
+      "grad_norm": 2.9025795459747314,
+      "learning_rate": 0.0002,
+      "loss": 2.4843,
+      "step": 41080
+    },
+    {
+      "epoch": 3.0618479880774965,
+      "grad_norm": 2.6068289279937744,
+      "learning_rate": 0.0002,
+      "loss": 2.4153,
+      "step": 41090
+    },
+    {
+      "epoch": 3.0625931445603576,
+      "grad_norm": 2.5537047386169434,
+      "learning_rate": 0.0002,
+      "loss": 2.4612,
+      "step": 41100
+    },
+    {
+      "epoch": 3.063338301043219,
+      "grad_norm": 2.9191601276397705,
+      "learning_rate": 0.0002,
+      "loss": 2.5681,
+      "step": 41110
+    },
+    {
+      "epoch": 3.0640834575260807,
+      "grad_norm": 2.4749388694763184,
+      "learning_rate": 0.0002,
+      "loss": 2.2255,
+      "step": 41120
+    },
+    {
+      "epoch": 3.064828614008942,
+      "grad_norm": 3.2458298206329346,
+      "learning_rate": 0.0002,
+      "loss": 2.3751,
+      "step": 41130
+    },
+    {
+      "epoch": 3.0655737704918034,
+      "grad_norm": 2.4875826835632324,
+      "learning_rate": 0.0002,
+      "loss": 2.4063,
+      "step": 41140
+    },
+    {
+      "epoch": 3.066318926974665,
+      "grad_norm": 2.788435697555542,
+      "learning_rate": 0.0002,
+      "loss": 2.4848,
+      "step": 41150
+    },
+    {
+      "epoch": 3.067064083457526,
+      "grad_norm": 2.731872320175171,
+      "learning_rate": 0.0002,
+      "loss": 2.4219,
+      "step": 41160
+    },
+    {
+      "epoch": 3.0678092399403876,
+      "grad_norm": 2.5522799491882324,
+      "learning_rate": 0.0002,
+      "loss": 2.2948,
+      "step": 41170
+    },
+    {
+      "epoch": 3.0685543964232487,
+      "grad_norm": 2.494781255722046,
+      "learning_rate": 0.0002,
+      "loss": 2.4205,
+      "step": 41180
+    },
+    {
+      "epoch": 3.0692995529061102,
+      "grad_norm": 2.828057289123535,
+      "learning_rate": 0.0002,
+      "loss": 2.4637,
+      "step": 41190
+    },
+    {
+      "epoch": 3.070044709388972,
+      "grad_norm": 2.98608660697937,
+      "learning_rate": 0.0002,
+      "loss": 2.4205,
+      "step": 41200
+    },
+    {
+      "epoch": 3.070789865871833,
+      "grad_norm": 2.9125189781188965,
+      "learning_rate": 0.0002,
+      "loss": 2.3945,
+      "step": 41210
+    },
+    {
+      "epoch": 3.0715350223546944,
+      "grad_norm": 2.599982261657715,
+      "learning_rate": 0.0002,
+      "loss": 2.4643,
+      "step": 41220
+    },
+    {
+      "epoch": 3.072280178837556,
+      "grad_norm": 2.4792168140411377,
+      "learning_rate": 0.0002,
+      "loss": 2.368,
+      "step": 41230
+    },
+    {
+      "epoch": 3.073025335320417,
+      "grad_norm": 2.604639768600464,
+      "learning_rate": 0.0002,
+      "loss": 2.3853,
+      "step": 41240
+    },
+    {
+      "epoch": 3.0737704918032787,
+      "grad_norm": 2.617868185043335,
+      "learning_rate": 0.0002,
+      "loss": 2.342,
+      "step": 41250
+    },
+    {
+      "epoch": 3.07451564828614,
+      "grad_norm": 2.1132025718688965,
+      "learning_rate": 0.0002,
+      "loss": 2.2172,
+      "step": 41260
+    },
+    {
+      "epoch": 3.0752608047690013,
+      "grad_norm": 2.4275896549224854,
+      "learning_rate": 0.0002,
+      "loss": 2.0727,
+      "step": 41270
+    },
+    {
+      "epoch": 3.076005961251863,
+      "grad_norm": 2.285426616668701,
+      "learning_rate": 0.0002,
+      "loss": 2.2816,
+      "step": 41280
+    },
+    {
+      "epoch": 3.0767511177347244,
+      "grad_norm": 2.9328603744506836,
+      "learning_rate": 0.0002,
+      "loss": 2.3676,
+      "step": 41290
+    },
+    {
+      "epoch": 3.0774962742175855,
+      "grad_norm": 2.247255325317383,
+      "learning_rate": 0.0002,
+      "loss": 2.2789,
+      "step": 41300
+    },
+    {
+      "epoch": 3.078241430700447,
+      "grad_norm": 2.8094482421875,
+      "learning_rate": 0.0002,
+      "loss": 2.5418,
+      "step": 41310
+    },
+    {
+      "epoch": 3.0789865871833086,
+      "grad_norm": 2.7178444862365723,
+      "learning_rate": 0.0002,
+      "loss": 2.4174,
+      "step": 41320
+    },
+    {
+      "epoch": 3.0797317436661698,
+      "grad_norm": 3.3467769622802734,
+      "learning_rate": 0.0002,
+      "loss": 2.459,
+      "step": 41330
+    },
+    {
+      "epoch": 3.0804769001490313,
+      "grad_norm": 2.7289459705352783,
+      "learning_rate": 0.0002,
+      "loss": 2.3743,
+      "step": 41340
+    },
+    {
+      "epoch": 3.081222056631893,
+      "grad_norm": 3.2940895557403564,
+      "learning_rate": 0.0002,
+      "loss": 2.4498,
+      "step": 41350
+    },
+    {
+      "epoch": 3.081967213114754,
+      "grad_norm": 2.498302459716797,
+      "learning_rate": 0.0002,
+      "loss": 2.4309,
+      "step": 41360
+    },
+    {
+      "epoch": 3.0827123695976155,
+      "grad_norm": 2.9865078926086426,
+      "learning_rate": 0.0002,
+      "loss": 2.585,
+      "step": 41370
+    },
+    {
+      "epoch": 3.083457526080477,
+      "grad_norm": 3.016395092010498,
+      "learning_rate": 0.0002,
+      "loss": 2.6466,
+      "step": 41380
+    },
+    {
+      "epoch": 3.084202682563338,
+      "grad_norm": 2.7781736850738525,
+      "learning_rate": 0.0002,
+      "loss": 2.3803,
+      "step": 41390
+    },
+    {
+      "epoch": 3.0849478390461997,
+      "grad_norm": 2.8058199882507324,
+      "learning_rate": 0.0002,
+      "loss": 2.3414,
+      "step": 41400
+    },
+    {
+      "epoch": 3.0856929955290613,
+      "grad_norm": 2.8715202808380127,
+      "learning_rate": 0.0002,
+      "loss": 2.2693,
+      "step": 41410
+    },
+    {
+      "epoch": 3.0864381520119224,
+      "grad_norm": 2.7370269298553467,
+      "learning_rate": 0.0002,
+      "loss": 2.4274,
+      "step": 41420
+    },
+    {
+      "epoch": 3.087183308494784,
+      "grad_norm": 2.570589065551758,
+      "learning_rate": 0.0002,
+      "loss": 2.5473,
+      "step": 41430
+    },
+    {
+      "epoch": 3.0879284649776455,
+      "grad_norm": 2.858039617538452,
+      "learning_rate": 0.0002,
+      "loss": 2.4307,
+      "step": 41440
+    },
+    {
+      "epoch": 3.0886736214605066,
+      "grad_norm": 2.4094789028167725,
+      "learning_rate": 0.0002,
+      "loss": 2.3369,
+      "step": 41450
+    },
+    {
+      "epoch": 3.089418777943368,
+      "grad_norm": 2.665861129760742,
+      "learning_rate": 0.0002,
+      "loss": 2.4736,
+      "step": 41460
+    },
+    {
+      "epoch": 3.0901639344262297,
+      "grad_norm": 2.7448792457580566,
+      "learning_rate": 0.0002,
+      "loss": 2.2552,
+      "step": 41470
+    },
+    {
+      "epoch": 3.090909090909091,
+      "grad_norm": 3.320862293243408,
+      "learning_rate": 0.0002,
+      "loss": 2.361,
+      "step": 41480
+    },
+    {
+      "epoch": 3.0916542473919524,
+      "grad_norm": 2.623382568359375,
+      "learning_rate": 0.0002,
+      "loss": 2.2576,
+      "step": 41490
+    },
+    {
+      "epoch": 3.092399403874814,
+      "grad_norm": 2.757305383682251,
+      "learning_rate": 0.0002,
+      "loss": 2.4842,
+      "step": 41500
+    },
+    {
+      "epoch": 3.093144560357675,
+      "grad_norm": 3.2521588802337646,
+      "learning_rate": 0.0002,
+      "loss": 2.4941,
+      "step": 41510
+    },
+    {
+      "epoch": 3.0938897168405366,
+      "grad_norm": 2.6293327808380127,
+      "learning_rate": 0.0002,
+      "loss": 2.3642,
+      "step": 41520
+    },
+    {
+      "epoch": 3.0946348733233977,
+      "grad_norm": 2.7597970962524414,
+      "learning_rate": 0.0002,
+      "loss": 2.4894,
+      "step": 41530
+    },
+    {
+      "epoch": 3.0953800298062593,
+      "grad_norm": 3.117297887802124,
+      "learning_rate": 0.0002,
+      "loss": 2.5951,
+      "step": 41540
+    },
+    {
+      "epoch": 3.096125186289121,
+      "grad_norm": 2.641249656677246,
+      "learning_rate": 0.0002,
+      "loss": 2.2928,
+      "step": 41550
+    },
+    {
+      "epoch": 3.096870342771982,
+      "grad_norm": 3.0518288612365723,
+      "learning_rate": 0.0002,
+      "loss": 2.3452,
+      "step": 41560
+    },
+    {
+      "epoch": 3.0976154992548435,
+      "grad_norm": 2.6108360290527344,
+      "learning_rate": 0.0002,
+      "loss": 2.2999,
+      "step": 41570
+    },
+    {
+      "epoch": 3.098360655737705,
+      "grad_norm": 3.087437629699707,
+      "learning_rate": 0.0002,
+      "loss": 2.4094,
+      "step": 41580
+    },
+    {
+      "epoch": 3.099105812220566,
+      "grad_norm": 2.4869635105133057,
+      "learning_rate": 0.0002,
+      "loss": 2.4677,
+      "step": 41590
+    },
+    {
+      "epoch": 3.0998509687034277,
+      "grad_norm": 3.1462340354919434,
+      "learning_rate": 0.0002,
+      "loss": 2.5562,
+      "step": 41600
+    },
+    {
+      "epoch": 3.1005961251862892,
+      "grad_norm": 2.7001521587371826,
+      "learning_rate": 0.0002,
+      "loss": 2.4251,
+      "step": 41610
+    },
+    {
+      "epoch": 3.1013412816691504,
+      "grad_norm": 3.1031558513641357,
+      "learning_rate": 0.0002,
+      "loss": 2.492,
+      "step": 41620
+    },
+    {
+      "epoch": 3.102086438152012,
+      "grad_norm": 2.6600401401519775,
+      "learning_rate": 0.0002,
+      "loss": 2.4346,
+      "step": 41630
+    },
+    {
+      "epoch": 3.1028315946348735,
+      "grad_norm": 2.673090934753418,
+      "learning_rate": 0.0002,
+      "loss": 2.3663,
+      "step": 41640
+    },
+    {
+      "epoch": 3.1035767511177346,
+      "grad_norm": 2.6423120498657227,
+      "learning_rate": 0.0002,
+      "loss": 2.486,
+      "step": 41650
+    },
+    {
+      "epoch": 3.104321907600596,
+      "grad_norm": 2.6316373348236084,
+      "learning_rate": 0.0002,
+      "loss": 2.5157,
+      "step": 41660
+    },
+    {
+      "epoch": 3.1050670640834577,
+      "grad_norm": 2.9219932556152344,
+      "learning_rate": 0.0002,
+      "loss": 2.2875,
+      "step": 41670
+    },
+    {
+      "epoch": 3.105812220566319,
+      "grad_norm": 3.2191224098205566,
+      "learning_rate": 0.0002,
+      "loss": 2.3327,
+      "step": 41680
+    },
+    {
+      "epoch": 3.1065573770491803,
+      "grad_norm": 2.6431362628936768,
+      "learning_rate": 0.0002,
+      "loss": 2.4269,
+      "step": 41690
+    },
+    {
+      "epoch": 3.107302533532042,
+      "grad_norm": 2.56723952293396,
+      "learning_rate": 0.0002,
+      "loss": 2.3461,
+      "step": 41700
+    },
+    {
+      "epoch": 3.108047690014903,
+      "grad_norm": 3.1009645462036133,
+      "learning_rate": 0.0002,
+      "loss": 2.62,
+      "step": 41710
+    },
+    {
+      "epoch": 3.1087928464977646,
+      "grad_norm": 2.805692672729492,
+      "learning_rate": 0.0002,
+      "loss": 2.2908,
+      "step": 41720
+    },
+    {
+      "epoch": 3.109538002980626,
+      "grad_norm": 2.93380069732666,
+      "learning_rate": 0.0002,
+      "loss": 2.4379,
+      "step": 41730
+    },
+    {
+      "epoch": 3.110283159463487,
+      "grad_norm": 2.5862841606140137,
+      "learning_rate": 0.0002,
+      "loss": 2.3883,
+      "step": 41740
+    },
+    {
+      "epoch": 3.1110283159463488,
+      "grad_norm": 2.6500649452209473,
+      "learning_rate": 0.0002,
+      "loss": 2.3901,
+      "step": 41750
+    },
+    {
+      "epoch": 3.1117734724292103,
+      "grad_norm": 2.738546371459961,
+      "learning_rate": 0.0002,
+      "loss": 2.4566,
+      "step": 41760
+    },
+    {
+      "epoch": 3.1125186289120714,
+      "grad_norm": 2.519313335418701,
+      "learning_rate": 0.0002,
+      "loss": 2.4855,
+      "step": 41770
+    },
+    {
+      "epoch": 3.113263785394933,
+      "grad_norm": 2.5396151542663574,
+      "learning_rate": 0.0002,
+      "loss": 2.4722,
+      "step": 41780
+    },
+    {
+      "epoch": 3.1140089418777945,
+      "grad_norm": 2.6931238174438477,
+      "learning_rate": 0.0002,
+      "loss": 2.5307,
+      "step": 41790
+    },
+    {
+      "epoch": 3.1147540983606556,
+      "grad_norm": 2.5568339824676514,
+      "learning_rate": 0.0002,
+      "loss": 2.3558,
+      "step": 41800
+    },
+    {
+      "epoch": 3.115499254843517,
+      "grad_norm": 2.8651018142700195,
+      "learning_rate": 0.0002,
+      "loss": 2.3666,
+      "step": 41810
+    },
+    {
+      "epoch": 3.1162444113263787,
+      "grad_norm": 2.6825473308563232,
+      "learning_rate": 0.0002,
+      "loss": 2.4556,
+      "step": 41820
+    },
+    {
+      "epoch": 3.11698956780924,
+      "grad_norm": 2.327749490737915,
+      "learning_rate": 0.0002,
+      "loss": 2.3705,
+      "step": 41830
+    },
+    {
+      "epoch": 3.1177347242921014,
+      "grad_norm": 2.5712668895721436,
+      "learning_rate": 0.0002,
+      "loss": 2.5078,
+      "step": 41840
+    },
+    {
+      "epoch": 3.118479880774963,
+      "grad_norm": 2.770862102508545,
+      "learning_rate": 0.0002,
+      "loss": 2.4223,
+      "step": 41850
+    },
+    {
+      "epoch": 3.119225037257824,
+      "grad_norm": 2.7376694679260254,
+      "learning_rate": 0.0002,
+      "loss": 2.0986,
+      "step": 41860
+    },
+    {
+      "epoch": 3.1199701937406856,
+      "grad_norm": 2.3979244232177734,
+      "learning_rate": 0.0002,
+      "loss": 2.3534,
+      "step": 41870
+    },
+    {
+      "epoch": 3.1207153502235467,
+      "grad_norm": 3.0536673069000244,
+      "learning_rate": 0.0002,
+      "loss": 2.4107,
+      "step": 41880
+    },
+    {
+      "epoch": 3.1214605067064083,
+      "grad_norm": 2.726121187210083,
+      "learning_rate": 0.0002,
+      "loss": 2.3308,
+      "step": 41890
+    },
+    {
+      "epoch": 3.12220566318927,
+      "grad_norm": 2.5417113304138184,
+      "learning_rate": 0.0002,
+      "loss": 2.3748,
+      "step": 41900
+    },
+    {
+      "epoch": 3.122950819672131,
+      "grad_norm": 2.813762664794922,
+      "learning_rate": 0.0002,
+      "loss": 2.4617,
+      "step": 41910
+    },
+    {
+      "epoch": 3.1236959761549925,
+      "grad_norm": 2.67824649810791,
+      "learning_rate": 0.0002,
+      "loss": 2.391,
+      "step": 41920
+    },
+    {
+      "epoch": 3.124441132637854,
+      "grad_norm": 2.5998127460479736,
+      "learning_rate": 0.0002,
+      "loss": 2.417,
+      "step": 41930
+    },
+    {
+      "epoch": 3.125186289120715,
+      "grad_norm": 2.6877424716949463,
+      "learning_rate": 0.0002,
+      "loss": 2.154,
+      "step": 41940
+    },
+    {
+      "epoch": 3.1259314456035767,
+      "grad_norm": 2.592630386352539,
+      "learning_rate": 0.0002,
+      "loss": 2.2968,
+      "step": 41950
+    },
+    {
+      "epoch": 3.1266766020864383,
+      "grad_norm": 2.4909517765045166,
+      "learning_rate": 0.0002,
+      "loss": 2.2333,
+      "step": 41960
+    },
+    {
+      "epoch": 3.1274217585692994,
+      "grad_norm": 2.7702407836914062,
+      "learning_rate": 0.0002,
+      "loss": 2.329,
+      "step": 41970
+    },
+    {
+      "epoch": 3.128166915052161,
+      "grad_norm": 3.2048144340515137,
+      "learning_rate": 0.0002,
+      "loss": 2.5344,
+      "step": 41980
+    },
+    {
+      "epoch": 3.1289120715350225,
+      "grad_norm": 2.452103614807129,
+      "learning_rate": 0.0002,
+      "loss": 2.3947,
+      "step": 41990
+    },
+    {
+      "epoch": 3.1296572280178836,
+      "grad_norm": 3.0650296211242676,
+      "learning_rate": 0.0002,
+      "loss": 2.6355,
+      "step": 42000
+    },
+    {
+      "epoch": 3.130402384500745,
+      "grad_norm": 3.1083576679229736,
+      "learning_rate": 0.0002,
+      "loss": 2.3897,
+      "step": 42010
+    },
+    {
+      "epoch": 3.1311475409836067,
+      "grad_norm": 2.4616737365722656,
+      "learning_rate": 0.0002,
+      "loss": 2.4941,
+      "step": 42020
+    },
+    {
+      "epoch": 3.131892697466468,
+      "grad_norm": 2.4813618659973145,
+      "learning_rate": 0.0002,
+      "loss": 2.2795,
+      "step": 42030
+    },
+    {
+      "epoch": 3.1326378539493294,
+      "grad_norm": 2.4839460849761963,
+      "learning_rate": 0.0002,
+      "loss": 2.4753,
+      "step": 42040
+    },
+    {
+      "epoch": 3.133383010432191,
+      "grad_norm": 2.5165300369262695,
+      "learning_rate": 0.0002,
+      "loss": 2.4168,
+      "step": 42050
+    },
+    {
+      "epoch": 3.134128166915052,
+      "grad_norm": 3.2887141704559326,
+      "learning_rate": 0.0002,
+      "loss": 2.2558,
+      "step": 42060
+    },
+    {
+      "epoch": 3.1348733233979136,
+      "grad_norm": 2.8972301483154297,
+      "learning_rate": 0.0002,
+      "loss": 2.5248,
+      "step": 42070
+    },
+    {
+      "epoch": 3.135618479880775,
+      "grad_norm": 2.6605656147003174,
+      "learning_rate": 0.0002,
+      "loss": 2.3511,
+      "step": 42080
+    },
+    {
+      "epoch": 3.1363636363636362,
+      "grad_norm": 2.9574508666992188,
+      "learning_rate": 0.0002,
+      "loss": 2.5589,
+      "step": 42090
+    },
+    {
+      "epoch": 3.137108792846498,
+      "grad_norm": 2.4659640789031982,
+      "learning_rate": 0.0002,
+      "loss": 2.4887,
+      "step": 42100
+    },
+    {
+      "epoch": 3.1378539493293593,
+      "grad_norm": 2.799456834793091,
+      "learning_rate": 0.0002,
+      "loss": 2.4112,
+      "step": 42110
+    },
+    {
+      "epoch": 3.1385991058122205,
+      "grad_norm": 2.674593925476074,
+      "learning_rate": 0.0002,
+      "loss": 2.472,
+      "step": 42120
+    },
+    {
+      "epoch": 3.139344262295082,
+      "grad_norm": 2.741276979446411,
+      "learning_rate": 0.0002,
+      "loss": 2.438,
+      "step": 42130
+    },
+    {
+      "epoch": 3.1400894187779436,
+      "grad_norm": 3.0024397373199463,
+      "learning_rate": 0.0002,
+      "loss": 2.4534,
+      "step": 42140
+    },
+    {
+      "epoch": 3.1408345752608047,
+      "grad_norm": 2.393557548522949,
+      "learning_rate": 0.0002,
+      "loss": 2.4285,
+      "step": 42150
+    },
+    {
+      "epoch": 3.1415797317436662,
+      "grad_norm": 2.408137798309326,
+      "learning_rate": 0.0002,
+      "loss": 2.2754,
+      "step": 42160
+    },
+    {
+      "epoch": 3.1423248882265273,
+      "grad_norm": 2.6676998138427734,
+      "learning_rate": 0.0002,
+      "loss": 2.4991,
+      "step": 42170
+    },
+    {
+      "epoch": 3.143070044709389,
+      "grad_norm": 2.7383525371551514,
+      "learning_rate": 0.0002,
+      "loss": 2.4517,
+      "step": 42180
+    },
+    {
+      "epoch": 3.1438152011922504,
+      "grad_norm": 3.209341526031494,
+      "learning_rate": 0.0002,
+      "loss": 2.4987,
+      "step": 42190
+    },
+    {
+      "epoch": 3.144560357675112,
+      "grad_norm": 2.7689366340637207,
+      "learning_rate": 0.0002,
+      "loss": 2.3316,
+      "step": 42200
+    },
+    {
+      "epoch": 3.145305514157973,
+      "grad_norm": 2.7870845794677734,
+      "learning_rate": 0.0002,
+      "loss": 2.4651,
+      "step": 42210
+    },
+    {
+      "epoch": 3.1460506706408347,
+      "grad_norm": 2.8037571907043457,
+      "learning_rate": 0.0002,
+      "loss": 2.4491,
+      "step": 42220
+    },
+    {
+      "epoch": 3.1467958271236958,
+      "grad_norm": 2.640016555786133,
+      "learning_rate": 0.0002,
+      "loss": 2.2967,
+      "step": 42230
+    },
+    {
+      "epoch": 3.1475409836065573,
+      "grad_norm": 2.958155870437622,
+      "learning_rate": 0.0002,
+      "loss": 2.1981,
+      "step": 42240
+    },
+    {
+      "epoch": 3.148286140089419,
+      "grad_norm": 3.091360569000244,
+      "learning_rate": 0.0002,
+      "loss": 2.5147,
+      "step": 42250
+    },
+    {
+      "epoch": 3.14903129657228,
+      "grad_norm": 3.4414288997650146,
+      "learning_rate": 0.0002,
+      "loss": 2.4308,
+      "step": 42260
+    },
+    {
+      "epoch": 3.1497764530551415,
+      "grad_norm": 3.074572801589966,
+      "learning_rate": 0.0002,
+      "loss": 2.5425,
+      "step": 42270
+    },
+    {
+      "epoch": 3.150521609538003,
+      "grad_norm": 2.8518035411834717,
+      "learning_rate": 0.0002,
+      "loss": 2.5321,
+      "step": 42280
+    },
+    {
+      "epoch": 3.151266766020864,
+      "grad_norm": 2.689629077911377,
+      "learning_rate": 0.0002,
+      "loss": 2.527,
+      "step": 42290
+    },
+    {
+      "epoch": 3.1520119225037257,
+      "grad_norm": 2.6013498306274414,
+      "learning_rate": 0.0002,
+      "loss": 2.3945,
+      "step": 42300
+    },
+    {
+      "epoch": 3.1527570789865873,
+      "grad_norm": 2.6785435676574707,
+      "learning_rate": 0.0002,
+      "loss": 2.5889,
+      "step": 42310
+    },
+    {
+      "epoch": 3.1535022354694484,
+      "grad_norm": 2.7143375873565674,
+      "learning_rate": 0.0002,
+      "loss": 2.4548,
+      "step": 42320
+    },
+    {
+      "epoch": 3.15424739195231,
+      "grad_norm": 2.794013261795044,
+      "learning_rate": 0.0002,
+      "loss": 2.4599,
+      "step": 42330
+    },
+    {
+      "epoch": 3.1549925484351715,
+      "grad_norm": 2.9435579776763916,
+      "learning_rate": 0.0002,
+      "loss": 2.4458,
+      "step": 42340
+    },
+    {
+      "epoch": 3.1557377049180326,
+      "grad_norm": 2.210059642791748,
+      "learning_rate": 0.0002,
+      "loss": 2.4752,
+      "step": 42350
+    },
+    {
+      "epoch": 3.156482861400894,
+      "grad_norm": 3.0881826877593994,
+      "learning_rate": 0.0002,
+      "loss": 2.3679,
+      "step": 42360
+    },
+    {
+      "epoch": 3.1572280178837557,
+      "grad_norm": 2.6826508045196533,
+      "learning_rate": 0.0002,
+      "loss": 2.3735,
+      "step": 42370
+    },
+    {
+      "epoch": 3.157973174366617,
+      "grad_norm": 2.657874822616577,
+      "learning_rate": 0.0002,
+      "loss": 2.3878,
+      "step": 42380
+    },
+    {
+      "epoch": 3.1587183308494784,
+      "grad_norm": 2.6119277477264404,
+      "learning_rate": 0.0002,
+      "loss": 2.4191,
+      "step": 42390
+    },
+    {
+      "epoch": 3.15946348733234,
+      "grad_norm": 2.929684638977051,
+      "learning_rate": 0.0002,
+      "loss": 2.3763,
+      "step": 42400
+    },
+    {
+      "epoch": 3.160208643815201,
+      "grad_norm": 2.660856246948242,
+      "learning_rate": 0.0002,
+      "loss": 2.4806,
+      "step": 42410
+    },
+    {
+      "epoch": 3.1609538002980626,
+      "grad_norm": 2.3593862056732178,
+      "learning_rate": 0.0002,
+      "loss": 2.2756,
+      "step": 42420
+    },
+    {
+      "epoch": 3.161698956780924,
+      "grad_norm": 3.1583991050720215,
+      "learning_rate": 0.0002,
+      "loss": 2.4429,
+      "step": 42430
+    },
+    {
+      "epoch": 3.1624441132637853,
+      "grad_norm": 2.3144760131835938,
+      "learning_rate": 0.0002,
+      "loss": 2.3234,
+      "step": 42440
+    },
+    {
+      "epoch": 3.163189269746647,
+      "grad_norm": 2.6270995140075684,
+      "learning_rate": 0.0002,
+      "loss": 2.4479,
+      "step": 42450
+    },
+    {
+      "epoch": 3.1639344262295084,
+      "grad_norm": 2.9927475452423096,
+      "learning_rate": 0.0002,
+      "loss": 2.4985,
+      "step": 42460
+    },
+    {
+      "epoch": 3.1646795827123695,
+      "grad_norm": 2.8037357330322266,
+      "learning_rate": 0.0002,
+      "loss": 2.3792,
+      "step": 42470
+    },
+    {
+      "epoch": 3.165424739195231,
+      "grad_norm": 2.522001028060913,
+      "learning_rate": 0.0002,
+      "loss": 2.5035,
+      "step": 42480
+    },
+    {
+      "epoch": 3.1661698956780926,
+      "grad_norm": 2.7954230308532715,
+      "learning_rate": 0.0002,
+      "loss": 2.5186,
+      "step": 42490
+    },
+    {
+      "epoch": 3.1669150521609537,
+      "grad_norm": 2.842353105545044,
+      "learning_rate": 0.0002,
+      "loss": 2.3828,
+      "step": 42500
+    },
+    {
+      "epoch": 3.1676602086438153,
+      "grad_norm": 3.6176884174346924,
+      "learning_rate": 0.0002,
+      "loss": 2.4387,
+      "step": 42510
+    },
+    {
+      "epoch": 3.168405365126677,
+      "grad_norm": 2.5962915420532227,
+      "learning_rate": 0.0002,
+      "loss": 2.3441,
+      "step": 42520
+    },
+    {
+      "epoch": 3.169150521609538,
+      "grad_norm": 2.856750726699829,
+      "learning_rate": 0.0002,
+      "loss": 2.4871,
+      "step": 42530
+    },
+    {
+      "epoch": 3.1698956780923995,
+      "grad_norm": 2.5241379737854004,
+      "learning_rate": 0.0002,
+      "loss": 2.4527,
+      "step": 42540
+    },
+    {
+      "epoch": 3.170640834575261,
+      "grad_norm": 3.149193525314331,
+      "learning_rate": 0.0002,
+      "loss": 2.4242,
+      "step": 42550
+    },
+    {
+      "epoch": 3.171385991058122,
+      "grad_norm": 2.600130081176758,
+      "learning_rate": 0.0002,
+      "loss": 2.4865,
+      "step": 42560
+    },
+    {
+      "epoch": 3.1721311475409837,
+      "grad_norm": 2.598031520843506,
+      "learning_rate": 0.0002,
+      "loss": 2.4508,
+      "step": 42570
+    },
+    {
+      "epoch": 3.172876304023845,
+      "grad_norm": 2.6382107734680176,
+      "learning_rate": 0.0002,
+      "loss": 2.4075,
+      "step": 42580
+    },
+    {
+      "epoch": 3.1736214605067063,
+      "grad_norm": 2.6536340713500977,
+      "learning_rate": 0.0002,
+      "loss": 2.2179,
+      "step": 42590
+    },
+    {
+      "epoch": 3.174366616989568,
+      "grad_norm": 2.2409815788269043,
+      "learning_rate": 0.0002,
+      "loss": 2.4891,
+      "step": 42600
+    },
+    {
+      "epoch": 3.175111773472429,
+      "grad_norm": 2.8284525871276855,
+      "learning_rate": 0.0002,
+      "loss": 2.4104,
+      "step": 42610
+    },
+    {
+      "epoch": 3.1758569299552906,
+      "grad_norm": 2.8395347595214844,
+      "learning_rate": 0.0002,
+      "loss": 2.4732,
+      "step": 42620
+    },
+    {
+      "epoch": 3.176602086438152,
+      "grad_norm": 2.7434206008911133,
+      "learning_rate": 0.0002,
+      "loss": 2.5753,
+      "step": 42630
+    },
+    {
+      "epoch": 3.1773472429210132,
+      "grad_norm": 2.5308117866516113,
+      "learning_rate": 0.0002,
+      "loss": 2.4494,
+      "step": 42640
+    },
+    {
+      "epoch": 3.178092399403875,
+      "grad_norm": 2.765371322631836,
+      "learning_rate": 0.0002,
+      "loss": 2.2449,
+      "step": 42650
+    },
+    {
+      "epoch": 3.1788375558867363,
+      "grad_norm": 2.844912052154541,
+      "learning_rate": 0.0002,
+      "loss": 2.5911,
+      "step": 42660
+    },
+    {
+      "epoch": 3.1795827123695974,
+      "grad_norm": 2.811023712158203,
+      "learning_rate": 0.0002,
+      "loss": 2.2395,
+      "step": 42670
+    },
+    {
+      "epoch": 3.180327868852459,
+      "grad_norm": 2.2983829975128174,
+      "learning_rate": 0.0002,
+      "loss": 2.2522,
+      "step": 42680
+    },
+    {
+      "epoch": 3.1810730253353205,
+      "grad_norm": 2.5697834491729736,
+      "learning_rate": 0.0002,
+      "loss": 2.4592,
+      "step": 42690
+    },
+    {
+      "epoch": 3.1818181818181817,
+      "grad_norm": 2.6632912158966064,
+      "learning_rate": 0.0002,
+      "loss": 2.3589,
+      "step": 42700
+    },
+    {
+      "epoch": 3.182563338301043,
+      "grad_norm": 2.8424110412597656,
+      "learning_rate": 0.0002,
+      "loss": 2.4862,
+      "step": 42710
+    },
+    {
+      "epoch": 3.1833084947839048,
+      "grad_norm": 2.80009126663208,
+      "learning_rate": 0.0002,
+      "loss": 2.3216,
+      "step": 42720
+    },
+    {
+      "epoch": 3.184053651266766,
+      "grad_norm": 2.937347173690796,
+      "learning_rate": 0.0002,
+      "loss": 2.5931,
+      "step": 42730
+    },
+    {
+      "epoch": 3.1847988077496274,
+      "grad_norm": 3.170665740966797,
+      "learning_rate": 0.0002,
+      "loss": 2.3354,
+      "step": 42740
+    },
+    {
+      "epoch": 3.185543964232489,
+      "grad_norm": 2.7834672927856445,
+      "learning_rate": 0.0002,
+      "loss": 2.3822,
+      "step": 42750
+    },
+    {
+      "epoch": 3.18628912071535,
+      "grad_norm": 2.834599733352661,
+      "learning_rate": 0.0002,
+      "loss": 2.5047,
+      "step": 42760
+    },
+    {
+      "epoch": 3.1870342771982116,
+      "grad_norm": 2.9384005069732666,
+      "learning_rate": 0.0002,
+      "loss": 2.4439,
+      "step": 42770
+    },
+    {
+      "epoch": 3.187779433681073,
+      "grad_norm": 2.7702577114105225,
+      "learning_rate": 0.0002,
+      "loss": 2.4609,
+      "step": 42780
+    },
+    {
+      "epoch": 3.1885245901639343,
+      "grad_norm": 2.811617612838745,
+      "learning_rate": 0.0002,
+      "loss": 2.5652,
+      "step": 42790
+    },
+    {
+      "epoch": 3.189269746646796,
+      "grad_norm": 2.532407283782959,
+      "learning_rate": 0.0002,
+      "loss": 2.3831,
+      "step": 42800
+    },
+    {
+      "epoch": 3.1900149031296574,
+      "grad_norm": 2.6427395343780518,
+      "learning_rate": 0.0002,
+      "loss": 2.5223,
+      "step": 42810
+    },
+    {
+      "epoch": 3.1907600596125185,
+      "grad_norm": 2.6416213512420654,
+      "learning_rate": 0.0002,
+      "loss": 2.2635,
+      "step": 42820
+    },
+    {
+      "epoch": 3.19150521609538,
+      "grad_norm": 2.5367422103881836,
+      "learning_rate": 0.0002,
+      "loss": 2.2955,
+      "step": 42830
+    },
+    {
+      "epoch": 3.1922503725782416,
+      "grad_norm": 2.593743085861206,
+      "learning_rate": 0.0002,
+      "loss": 2.5609,
+      "step": 42840
+    },
+    {
+      "epoch": 3.1929955290611027,
+      "grad_norm": 2.4231393337249756,
+      "learning_rate": 0.0002,
+      "loss": 2.2702,
+      "step": 42850
+    },
+    {
+      "epoch": 3.1937406855439643,
+      "grad_norm": 2.6415021419525146,
+      "learning_rate": 0.0002,
+      "loss": 2.4803,
+      "step": 42860
+    },
+    {
+      "epoch": 3.194485842026826,
+      "grad_norm": 3.0961923599243164,
+      "learning_rate": 0.0002,
+      "loss": 2.5815,
+      "step": 42870
+    },
+    {
+      "epoch": 3.195230998509687,
+      "grad_norm": 2.677943229675293,
+      "learning_rate": 0.0002,
+      "loss": 2.4246,
+      "step": 42880
+    },
+    {
+      "epoch": 3.1959761549925485,
+      "grad_norm": 2.9092931747436523,
+      "learning_rate": 0.0002,
+      "loss": 2.4639,
+      "step": 42890
+    },
+    {
+      "epoch": 3.19672131147541,
+      "grad_norm": 2.8990132808685303,
+      "learning_rate": 0.0002,
+      "loss": 2.5542,
+      "step": 42900
+    },
+    {
+      "epoch": 3.197466467958271,
+      "grad_norm": 2.2910470962524414,
+      "learning_rate": 0.0002,
+      "loss": 2.5081,
+      "step": 42910
+    },
+    {
+      "epoch": 3.1982116244411327,
+      "grad_norm": 2.7519373893737793,
+      "learning_rate": 0.0002,
+      "loss": 2.2503,
+      "step": 42920
+    },
+    {
+      "epoch": 3.198956780923994,
+      "grad_norm": 2.626312017440796,
+      "learning_rate": 0.0002,
+      "loss": 2.3392,
+      "step": 42930
+    },
+    {
+      "epoch": 3.1997019374068554,
+      "grad_norm": 2.6032698154449463,
+      "learning_rate": 0.0002,
+      "loss": 2.4991,
+      "step": 42940
+    },
+    {
+      "epoch": 3.200447093889717,
+      "grad_norm": 2.6117911338806152,
+      "learning_rate": 0.0002,
+      "loss": 2.6928,
+      "step": 42950
+    },
+    {
+      "epoch": 3.201192250372578,
+      "grad_norm": 2.7212131023406982,
+      "learning_rate": 0.0002,
+      "loss": 2.0789,
+      "step": 42960
+    },
+    {
+      "epoch": 3.2019374068554396,
+      "grad_norm": 2.6502857208251953,
+      "learning_rate": 0.0002,
+      "loss": 2.5775,
+      "step": 42970
+    },
+    {
+      "epoch": 3.202682563338301,
+      "grad_norm": 2.800010919570923,
+      "learning_rate": 0.0002,
+      "loss": 2.4754,
+      "step": 42980
+    },
+    {
+      "epoch": 3.2034277198211623,
+      "grad_norm": 2.5369884967803955,
+      "learning_rate": 0.0002,
+      "loss": 2.4423,
+      "step": 42990
+    },
+    {
+      "epoch": 3.204172876304024,
+      "grad_norm": 3.0359115600585938,
+      "learning_rate": 0.0002,
+      "loss": 2.4593,
+      "step": 43000
+    },
+    {
+      "epoch": 3.2049180327868854,
+      "grad_norm": 2.6096441745758057,
+      "learning_rate": 0.0002,
+      "loss": 2.3148,
+      "step": 43010
+    },
+    {
+      "epoch": 3.2056631892697465,
+      "grad_norm": 3.0144314765930176,
+      "learning_rate": 0.0002,
+      "loss": 2.3979,
+      "step": 43020
+    },
+    {
+      "epoch": 3.206408345752608,
+      "grad_norm": 2.607226610183716,
+      "learning_rate": 0.0002,
+      "loss": 2.5429,
+      "step": 43030
+    },
+    {
+      "epoch": 3.2071535022354696,
+      "grad_norm": 3.2121124267578125,
+      "learning_rate": 0.0002,
+      "loss": 2.5202,
+      "step": 43040
+    },
+    {
+      "epoch": 3.2078986587183307,
+      "grad_norm": 2.5059688091278076,
+      "learning_rate": 0.0002,
+      "loss": 2.4008,
+      "step": 43050
+    },
+    {
+      "epoch": 3.2086438152011922,
+      "grad_norm": 2.7579286098480225,
+      "learning_rate": 0.0002,
+      "loss": 2.4386,
+      "step": 43060
+    },
+    {
+      "epoch": 3.209388971684054,
+      "grad_norm": 2.520742654800415,
+      "learning_rate": 0.0002,
+      "loss": 2.3709,
+      "step": 43070
+    },
+    {
+      "epoch": 3.210134128166915,
+      "grad_norm": 2.72102952003479,
+      "learning_rate": 0.0002,
+      "loss": 2.1754,
+      "step": 43080
+    },
+    {
+      "epoch": 3.2108792846497765,
+      "grad_norm": 2.840904474258423,
+      "learning_rate": 0.0002,
+      "loss": 2.5463,
+      "step": 43090
+    },
+    {
+      "epoch": 3.211624441132638,
+      "grad_norm": 2.4814138412475586,
+      "learning_rate": 0.0002,
+      "loss": 2.3215,
+      "step": 43100
+    },
+    {
+      "epoch": 3.212369597615499,
+      "grad_norm": 2.4107296466827393,
+      "learning_rate": 0.0002,
+      "loss": 2.4029,
+      "step": 43110
+    },
+    {
+      "epoch": 3.2131147540983607,
+      "grad_norm": 2.8633036613464355,
+      "learning_rate": 0.0002,
+      "loss": 2.2321,
+      "step": 43120
+    },
+    {
+      "epoch": 3.2138599105812222,
+      "grad_norm": 2.631967067718506,
+      "learning_rate": 0.0002,
+      "loss": 2.2269,
+      "step": 43130
+    },
+    {
+      "epoch": 3.2146050670640833,
+      "grad_norm": 2.8928747177124023,
+      "learning_rate": 0.0002,
+      "loss": 2.4295,
+      "step": 43140
+    },
+    {
+      "epoch": 3.215350223546945,
+      "grad_norm": 2.65671443939209,
+      "learning_rate": 0.0002,
+      "loss": 2.4534,
+      "step": 43150
+    },
+    {
+      "epoch": 3.2160953800298064,
+      "grad_norm": 2.9897003173828125,
+      "learning_rate": 0.0002,
+      "loss": 2.5198,
+      "step": 43160
+    },
+    {
+      "epoch": 3.2168405365126675,
+      "grad_norm": 2.7278244495391846,
+      "learning_rate": 0.0002,
+      "loss": 2.4829,
+      "step": 43170
+    },
+    {
+      "epoch": 3.217585692995529,
+      "grad_norm": 2.805060625076294,
+      "learning_rate": 0.0002,
+      "loss": 2.3381,
+      "step": 43180
+    },
+    {
+      "epoch": 3.2183308494783907,
+      "grad_norm": 2.806382656097412,
+      "learning_rate": 0.0002,
+      "loss": 2.3296,
+      "step": 43190
+    },
+    {
+      "epoch": 3.2190760059612518,
+      "grad_norm": 3.0220911502838135,
+      "learning_rate": 0.0002,
+      "loss": 2.514,
+      "step": 43200
+    },
+    {
+      "epoch": 3.2198211624441133,
+      "grad_norm": 2.6727559566497803,
+      "learning_rate": 0.0002,
+      "loss": 2.4302,
+      "step": 43210
+    },
+    {
+      "epoch": 3.220566318926975,
+      "grad_norm": 2.733079433441162,
+      "learning_rate": 0.0002,
+      "loss": 2.5175,
+      "step": 43220
+    },
+    {
+      "epoch": 3.221311475409836,
+      "grad_norm": 2.369826555252075,
+      "learning_rate": 0.0002,
+      "loss": 2.1816,
+      "step": 43230
+    },
+    {
+      "epoch": 3.2220566318926975,
+      "grad_norm": 2.4770312309265137,
+      "learning_rate": 0.0002,
+      "loss": 2.6042,
+      "step": 43240
+    },
+    {
+      "epoch": 3.222801788375559,
+      "grad_norm": 2.658463954925537,
+      "learning_rate": 0.0002,
+      "loss": 2.3651,
+      "step": 43250
+    },
+    {
+      "epoch": 3.22354694485842,
+      "grad_norm": 2.8133113384246826,
+      "learning_rate": 0.0002,
+      "loss": 2.585,
+      "step": 43260
+    },
+    {
+      "epoch": 3.2242921013412817,
+      "grad_norm": 2.6553611755371094,
+      "learning_rate": 0.0002,
+      "loss": 2.3669,
+      "step": 43270
+    },
+    {
+      "epoch": 3.225037257824143,
+      "grad_norm": 2.66679048538208,
+      "learning_rate": 0.0002,
+      "loss": 2.6358,
+      "step": 43280
+    },
+    {
+      "epoch": 3.2257824143070044,
+      "grad_norm": 2.5681023597717285,
+      "learning_rate": 0.0002,
+      "loss": 2.3225,
+      "step": 43290
+    },
+    {
+      "epoch": 3.226527570789866,
+      "grad_norm": 3.2260591983795166,
+      "learning_rate": 0.0002,
+      "loss": 2.4329,
+      "step": 43300
+    },
+    {
+      "epoch": 3.227272727272727,
+      "grad_norm": 2.7905280590057373,
+      "learning_rate": 0.0002,
+      "loss": 2.4839,
+      "step": 43310
+    },
+    {
+      "epoch": 3.2280178837555886,
+      "grad_norm": 3.111307382583618,
+      "learning_rate": 0.0002,
+      "loss": 2.2841,
+      "step": 43320
+    },
+    {
+      "epoch": 3.22876304023845,
+      "grad_norm": 2.8080315589904785,
+      "learning_rate": 0.0002,
+      "loss": 2.4087,
+      "step": 43330
+    },
+    {
+      "epoch": 3.2295081967213113,
+      "grad_norm": 2.8237740993499756,
+      "learning_rate": 0.0002,
+      "loss": 2.4916,
+      "step": 43340
+    },
+    {
+      "epoch": 3.230253353204173,
+      "grad_norm": 2.3456661701202393,
+      "learning_rate": 0.0002,
+      "loss": 2.3679,
+      "step": 43350
+    },
+    {
+      "epoch": 3.2309985096870344,
+      "grad_norm": 2.5504918098449707,
+      "learning_rate": 0.0002,
+      "loss": 2.4728,
+      "step": 43360
+    },
+    {
+      "epoch": 3.2317436661698955,
+      "grad_norm": 2.742598295211792,
+      "learning_rate": 0.0002,
+      "loss": 2.5054,
+      "step": 43370
+    },
+    {
+      "epoch": 3.232488822652757,
+      "grad_norm": 2.7378361225128174,
+      "learning_rate": 0.0002,
+      "loss": 2.4379,
+      "step": 43380
+    },
+    {
+      "epoch": 3.2332339791356186,
+      "grad_norm": 2.609257936477661,
+      "learning_rate": 0.0002,
+      "loss": 2.5783,
+      "step": 43390
+    },
+    {
+      "epoch": 3.2339791356184797,
+      "grad_norm": 2.4723076820373535,
+      "learning_rate": 0.0002,
+      "loss": 2.4199,
+      "step": 43400
+    },
+    {
+      "epoch": 3.2347242921013413,
+      "grad_norm": 2.965663194656372,
+      "learning_rate": 0.0002,
+      "loss": 2.4963,
+      "step": 43410
+    },
+    {
+      "epoch": 3.235469448584203,
+      "grad_norm": 2.788440704345703,
+      "learning_rate": 0.0002,
+      "loss": 2.4947,
+      "step": 43420
+    },
+    {
+      "epoch": 3.236214605067064,
+      "grad_norm": 2.882891893386841,
+      "learning_rate": 0.0002,
+      "loss": 2.4839,
+      "step": 43430
+    },
+    {
+      "epoch": 3.2369597615499255,
+      "grad_norm": 2.6174352169036865,
+      "learning_rate": 0.0002,
+      "loss": 2.332,
+      "step": 43440
+    },
+    {
+      "epoch": 3.237704918032787,
+      "grad_norm": 2.465498685836792,
+      "learning_rate": 0.0002,
+      "loss": 2.2894,
+      "step": 43450
+    },
+    {
+      "epoch": 3.238450074515648,
+      "grad_norm": 2.473252773284912,
+      "learning_rate": 0.0002,
+      "loss": 2.4923,
+      "step": 43460
+    },
+    {
+      "epoch": 3.2391952309985097,
+      "grad_norm": 2.660106897354126,
+      "learning_rate": 0.0002,
+      "loss": 2.3363,
+      "step": 43470
+    },
+    {
+      "epoch": 3.2399403874813713,
+      "grad_norm": 2.8930771350860596,
+      "learning_rate": 0.0002,
+      "loss": 2.5892,
+      "step": 43480
+    },
+    {
+      "epoch": 3.2406855439642324,
+      "grad_norm": 2.360569953918457,
+      "learning_rate": 0.0002,
+      "loss": 2.4016,
+      "step": 43490
+    },
+    {
+      "epoch": 3.241430700447094,
+      "grad_norm": 2.7150566577911377,
+      "learning_rate": 0.0002,
+      "loss": 2.1884,
+      "step": 43500
+    },
+    {
+      "epoch": 3.2421758569299555,
+      "grad_norm": 3.106088876724243,
+      "learning_rate": 0.0002,
+      "loss": 2.5107,
+      "step": 43510
+    },
+    {
+      "epoch": 3.2429210134128166,
+      "grad_norm": 2.799938201904297,
+      "learning_rate": 0.0002,
+      "loss": 2.5407,
+      "step": 43520
+    },
+    {
+      "epoch": 3.243666169895678,
+      "grad_norm": 2.440535306930542,
+      "learning_rate": 0.0002,
+      "loss": 2.5892,
+      "step": 43530
+    },
+    {
+      "epoch": 3.2444113263785397,
+      "grad_norm": 2.66219162940979,
+      "learning_rate": 0.0002,
+      "loss": 2.3643,
+      "step": 43540
+    },
+    {
+      "epoch": 3.245156482861401,
+      "grad_norm": 2.7151572704315186,
+      "learning_rate": 0.0002,
+      "loss": 2.3527,
+      "step": 43550
+    },
+    {
+      "epoch": 3.2459016393442623,
+      "grad_norm": 3.217926025390625,
+      "learning_rate": 0.0002,
+      "loss": 2.3433,
+      "step": 43560
+    },
+    {
+      "epoch": 3.246646795827124,
+      "grad_norm": 2.634303569793701,
+      "learning_rate": 0.0002,
+      "loss": 2.3928,
+      "step": 43570
+    },
+    {
+      "epoch": 3.247391952309985,
+      "grad_norm": 2.9070944786071777,
+      "learning_rate": 0.0002,
+      "loss": 2.4142,
+      "step": 43580
+    },
+    {
+      "epoch": 3.2481371087928466,
+      "grad_norm": 3.3297340869903564,
+      "learning_rate": 0.0002,
+      "loss": 2.4862,
+      "step": 43590
+    },
+    {
+      "epoch": 3.248882265275708,
+      "grad_norm": 2.632563591003418,
+      "learning_rate": 0.0002,
+      "loss": 2.5569,
+      "step": 43600
+    },
+    {
+      "epoch": 3.2496274217585692,
+      "grad_norm": 3.3055930137634277,
+      "learning_rate": 0.0002,
+      "loss": 2.4756,
+      "step": 43610
+    },
+    {
+      "epoch": 3.2503725782414308,
+      "grad_norm": 2.854285955429077,
+      "learning_rate": 0.0002,
+      "loss": 2.5673,
+      "step": 43620
+    },
+    {
+      "epoch": 3.251117734724292,
+      "grad_norm": 2.5545718669891357,
+      "learning_rate": 0.0002,
+      "loss": 2.4522,
+      "step": 43630
+    },
+    {
+      "epoch": 3.2518628912071534,
+      "grad_norm": 2.046332836151123,
+      "learning_rate": 0.0002,
+      "loss": 2.2818,
+      "step": 43640
+    },
+    {
+      "epoch": 3.252608047690015,
+      "grad_norm": 2.8132576942443848,
+      "learning_rate": 0.0002,
+      "loss": 2.4598,
+      "step": 43650
+    },
+    {
+      "epoch": 3.2533532041728765,
+      "grad_norm": 3.0209426879882812,
+      "learning_rate": 0.0002,
+      "loss": 2.5342,
+      "step": 43660
+    },
+    {
+      "epoch": 3.2540983606557377,
+      "grad_norm": 2.7068018913269043,
+      "learning_rate": 0.0002,
+      "loss": 2.3042,
+      "step": 43670
+    },
+    {
+      "epoch": 3.254843517138599,
+      "grad_norm": 2.8782992362976074,
+      "learning_rate": 0.0002,
+      "loss": 2.4472,
+      "step": 43680
+    },
+    {
+      "epoch": 3.2555886736214603,
+      "grad_norm": 2.6785826683044434,
+      "learning_rate": 0.0002,
+      "loss": 2.4746,
+      "step": 43690
+    },
+    {
+      "epoch": 3.256333830104322,
+      "grad_norm": 2.71146821975708,
+      "learning_rate": 0.0002,
+      "loss": 2.5153,
+      "step": 43700
+    },
+    {
+      "epoch": 3.2570789865871834,
+      "grad_norm": 2.8306570053100586,
+      "learning_rate": 0.0002,
+      "loss": 2.3877,
+      "step": 43710
+    },
+    {
+      "epoch": 3.2578241430700445,
+      "grad_norm": 2.9262166023254395,
+      "learning_rate": 0.0002,
+      "loss": 2.5242,
+      "step": 43720
+    },
+    {
+      "epoch": 3.258569299552906,
+      "grad_norm": 2.766948938369751,
+      "learning_rate": 0.0002,
+      "loss": 2.3074,
+      "step": 43730
+    },
+    {
+      "epoch": 3.2593144560357676,
+      "grad_norm": 2.673241376876831,
+      "learning_rate": 0.0002,
+      "loss": 2.4664,
+      "step": 43740
+    },
+    {
+      "epoch": 3.2600596125186287,
+      "grad_norm": 2.917844295501709,
+      "learning_rate": 0.0002,
+      "loss": 2.3575,
+      "step": 43750
+    },
+    {
+      "epoch": 3.2608047690014903,
+      "grad_norm": 2.763775587081909,
+      "learning_rate": 0.0002,
+      "loss": 2.4072,
+      "step": 43760
+    },
+    {
+      "epoch": 3.261549925484352,
+      "grad_norm": 2.166604518890381,
+      "learning_rate": 0.0002,
+      "loss": 2.3257,
+      "step": 43770
+    },
+    {
+      "epoch": 3.262295081967213,
+      "grad_norm": 2.9719974994659424,
+      "learning_rate": 0.0002,
+      "loss": 2.3358,
+      "step": 43780
+    },
+    {
+      "epoch": 3.2630402384500745,
+      "grad_norm": 3.200921058654785,
+      "learning_rate": 0.0002,
+      "loss": 2.5461,
+      "step": 43790
+    },
+    {
+      "epoch": 3.263785394932936,
+      "grad_norm": 2.6361167430877686,
+      "learning_rate": 0.0002,
+      "loss": 2.6172,
+      "step": 43800
+    },
+    {
+      "epoch": 3.264530551415797,
+      "grad_norm": 2.4378700256347656,
+      "learning_rate": 0.0002,
+      "loss": 2.4782,
+      "step": 43810
+    },
+    {
+      "epoch": 3.2652757078986587,
+      "grad_norm": 2.8756327629089355,
+      "learning_rate": 0.0002,
+      "loss": 2.6013,
+      "step": 43820
+    },
+    {
+      "epoch": 3.2660208643815203,
+      "grad_norm": 2.74397611618042,
+      "learning_rate": 0.0002,
+      "loss": 2.6196,
+      "step": 43830
+    },
+    {
+      "epoch": 3.2667660208643814,
+      "grad_norm": 2.8466079235076904,
+      "learning_rate": 0.0002,
+      "loss": 2.4868,
+      "step": 43840
+    },
+    {
+      "epoch": 3.267511177347243,
+      "grad_norm": 2.6650137901306152,
+      "learning_rate": 0.0002,
+      "loss": 2.461,
+      "step": 43850
+    },
+    {
+      "epoch": 3.2682563338301045,
+      "grad_norm": 2.747535228729248,
+      "learning_rate": 0.0002,
+      "loss": 2.3765,
+      "step": 43860
+    },
+    {
+      "epoch": 3.2690014903129656,
+      "grad_norm": 2.7829294204711914,
+      "learning_rate": 0.0002,
+      "loss": 2.4128,
+      "step": 43870
+    },
+    {
+      "epoch": 3.269746646795827,
+      "grad_norm": 2.3735265731811523,
+      "learning_rate": 0.0002,
+      "loss": 2.5265,
+      "step": 43880
+    },
+    {
+      "epoch": 3.2704918032786887,
+      "grad_norm": 2.747103214263916,
+      "learning_rate": 0.0002,
+      "loss": 2.5282,
+      "step": 43890
+    },
+    {
+      "epoch": 3.27123695976155,
+      "grad_norm": 2.760833740234375,
+      "learning_rate": 0.0002,
+      "loss": 2.4452,
+      "step": 43900
+    },
+    {
+      "epoch": 3.2719821162444114,
+      "grad_norm": 2.7421011924743652,
+      "learning_rate": 0.0002,
+      "loss": 2.4462,
+      "step": 43910
+    },
+    {
+      "epoch": 3.2727272727272725,
+      "grad_norm": 2.950409412384033,
+      "learning_rate": 0.0002,
+      "loss": 2.5802,
+      "step": 43920
+    },
+    {
+      "epoch": 3.273472429210134,
+      "grad_norm": 2.5211143493652344,
+      "learning_rate": 0.0002,
+      "loss": 2.3641,
+      "step": 43930
+    },
+    {
+      "epoch": 3.2742175856929956,
+      "grad_norm": 3.1419506072998047,
+      "learning_rate": 0.0002,
+      "loss": 2.4876,
+      "step": 43940
+    },
+    {
+      "epoch": 3.274962742175857,
+      "grad_norm": 2.7190613746643066,
+      "learning_rate": 0.0002,
+      "loss": 2.4003,
+      "step": 43950
+    },
+    {
+      "epoch": 3.2757078986587183,
+      "grad_norm": 2.7831997871398926,
+      "learning_rate": 0.0002,
+      "loss": 2.4456,
+      "step": 43960
+    },
+    {
+      "epoch": 3.27645305514158,
+      "grad_norm": 2.8021488189697266,
+      "learning_rate": 0.0002,
+      "loss": 2.3307,
+      "step": 43970
+    },
+    {
+      "epoch": 3.277198211624441,
+      "grad_norm": 2.8400144577026367,
+      "learning_rate": 0.0002,
+      "loss": 2.4423,
+      "step": 43980
+    },
+    {
+      "epoch": 3.2779433681073025,
+      "grad_norm": 2.451012134552002,
+      "learning_rate": 0.0002,
+      "loss": 2.3158,
+      "step": 43990
+    },
+    {
+      "epoch": 3.278688524590164,
+      "grad_norm": 2.821969985961914,
+      "learning_rate": 0.0002,
+      "loss": 2.318,
+      "step": 44000
+    },
+    {
+      "epoch": 3.2794336810730256,
+      "grad_norm": 2.844849109649658,
+      "learning_rate": 0.0002,
+      "loss": 2.4889,
+      "step": 44010
+    },
+    {
+      "epoch": 3.2801788375558867,
+      "grad_norm": 2.737499237060547,
+      "learning_rate": 0.0002,
+      "loss": 2.4165,
+      "step": 44020
+    },
+    {
+      "epoch": 3.2809239940387482,
+      "grad_norm": 2.477479934692383,
+      "learning_rate": 0.0002,
+      "loss": 2.4084,
+      "step": 44030
+    },
+    {
+      "epoch": 3.2816691505216093,
+      "grad_norm": 2.523198366165161,
+      "learning_rate": 0.0002,
+      "loss": 2.4999,
+      "step": 44040
+    },
+    {
+      "epoch": 3.282414307004471,
+      "grad_norm": 2.695603609085083,
+      "learning_rate": 0.0002,
+      "loss": 2.3774,
+      "step": 44050
+    },
+    {
+      "epoch": 3.2831594634873325,
+      "grad_norm": 3.03360652923584,
+      "learning_rate": 0.0002,
+      "loss": 2.2998,
+      "step": 44060
+    },
+    {
+      "epoch": 3.2839046199701936,
+      "grad_norm": 2.572307825088501,
+      "learning_rate": 0.0002,
+      "loss": 2.3314,
+      "step": 44070
+    },
+    {
+      "epoch": 3.284649776453055,
+      "grad_norm": 2.493046283721924,
+      "learning_rate": 0.0002,
+      "loss": 2.5724,
+      "step": 44080
+    },
+    {
+      "epoch": 3.2853949329359167,
+      "grad_norm": 2.80645489692688,
+      "learning_rate": 0.0002,
+      "loss": 2.4075,
+      "step": 44090
+    },
+    {
+      "epoch": 3.2861400894187778,
+      "grad_norm": 2.564210891723633,
+      "learning_rate": 0.0002,
+      "loss": 2.4697,
+      "step": 44100
+    },
+    {
+      "epoch": 3.2868852459016393,
+      "grad_norm": 2.9380991458892822,
+      "learning_rate": 0.0002,
+      "loss": 2.4112,
+      "step": 44110
+    },
+    {
+      "epoch": 3.287630402384501,
+      "grad_norm": 2.7709860801696777,
+      "learning_rate": 0.0002,
+      "loss": 2.5449,
+      "step": 44120
+    },
+    {
+      "epoch": 3.288375558867362,
+      "grad_norm": 2.371518611907959,
+      "learning_rate": 0.0002,
+      "loss": 2.2587,
+      "step": 44130
+    },
+    {
+      "epoch": 3.2891207153502235,
+      "grad_norm": 2.40743350982666,
+      "learning_rate": 0.0002,
+      "loss": 2.4445,
+      "step": 44140
+    },
+    {
+      "epoch": 3.289865871833085,
+      "grad_norm": 3.0384521484375,
+      "learning_rate": 0.0002,
+      "loss": 2.3444,
+      "step": 44150
+    },
+    {
+      "epoch": 3.290611028315946,
+      "grad_norm": 2.498652935028076,
+      "learning_rate": 0.0002,
+      "loss": 2.4962,
+      "step": 44160
+    },
+    {
+      "epoch": 3.2913561847988078,
+      "grad_norm": 3.0309088230133057,
+      "learning_rate": 0.0002,
+      "loss": 2.4336,
+      "step": 44170
+    },
+    {
+      "epoch": 3.2921013412816693,
+      "grad_norm": 2.711760997772217,
+      "learning_rate": 0.0002,
+      "loss": 2.3514,
+      "step": 44180
+    },
+    {
+      "epoch": 3.2928464977645304,
+      "grad_norm": 2.884289264678955,
+      "learning_rate": 0.0002,
+      "loss": 2.5209,
+      "step": 44190
+    },
+    {
+      "epoch": 3.293591654247392,
+      "grad_norm": 2.5556466579437256,
+      "learning_rate": 0.0002,
+      "loss": 2.4795,
+      "step": 44200
+    },
+    {
+      "epoch": 3.2943368107302535,
+      "grad_norm": 2.5224225521087646,
+      "learning_rate": 0.0002,
+      "loss": 2.3033,
+      "step": 44210
+    },
+    {
+      "epoch": 3.2950819672131146,
+      "grad_norm": 2.700383424758911,
+      "learning_rate": 0.0002,
+      "loss": 2.4198,
+      "step": 44220
+    },
+    {
+      "epoch": 3.295827123695976,
+      "grad_norm": 2.7120893001556396,
+      "learning_rate": 0.0002,
+      "loss": 2.3386,
+      "step": 44230
+    },
+    {
+      "epoch": 3.2965722801788377,
+      "grad_norm": 2.6876585483551025,
+      "learning_rate": 0.0002,
+      "loss": 2.3047,
+      "step": 44240
+    },
+    {
+      "epoch": 3.297317436661699,
+      "grad_norm": 2.7172040939331055,
+      "learning_rate": 0.0002,
+      "loss": 2.5499,
+      "step": 44250
+    },
+    {
+      "epoch": 3.2980625931445604,
+      "grad_norm": 2.645282506942749,
+      "learning_rate": 0.0002,
+      "loss": 2.4294,
+      "step": 44260
+    },
+    {
+      "epoch": 3.2988077496274215,
+      "grad_norm": 2.8003482818603516,
+      "learning_rate": 0.0002,
+      "loss": 2.4373,
+      "step": 44270
+    },
+    {
+      "epoch": 3.299552906110283,
+      "grad_norm": 2.8162996768951416,
+      "learning_rate": 0.0002,
+      "loss": 2.4679,
+      "step": 44280
+    },
+    {
+      "epoch": 3.3002980625931446,
+      "grad_norm": 2.7134716510772705,
+      "learning_rate": 0.0002,
+      "loss": 2.3455,
+      "step": 44290
+    },
+    {
+      "epoch": 3.301043219076006,
+      "grad_norm": 2.854041337966919,
+      "learning_rate": 0.0002,
+      "loss": 2.5039,
+      "step": 44300
+    },
+    {
+      "epoch": 3.3017883755588673,
+      "grad_norm": 2.794156789779663,
+      "learning_rate": 0.0002,
+      "loss": 2.6137,
+      "step": 44310
+    },
+    {
+      "epoch": 3.302533532041729,
+      "grad_norm": 2.626420021057129,
+      "learning_rate": 0.0002,
+      "loss": 2.5824,
+      "step": 44320
+    },
+    {
+      "epoch": 3.30327868852459,
+      "grad_norm": 2.318218946456909,
+      "learning_rate": 0.0002,
+      "loss": 2.6067,
+      "step": 44330
+    },
+    {
+      "epoch": 3.3040238450074515,
+      "grad_norm": 2.7970023155212402,
+      "learning_rate": 0.0002,
+      "loss": 2.4504,
+      "step": 44340
+    },
+    {
+      "epoch": 3.304769001490313,
+      "grad_norm": 2.773334503173828,
+      "learning_rate": 0.0002,
+      "loss": 2.6143,
+      "step": 44350
+    },
+    {
+      "epoch": 3.3055141579731746,
+      "grad_norm": 2.5117685794830322,
+      "learning_rate": 0.0002,
+      "loss": 2.2903,
+      "step": 44360
+    },
+    {
+      "epoch": 3.3062593144560357,
+      "grad_norm": 2.768136501312256,
+      "learning_rate": 0.0002,
+      "loss": 2.4649,
+      "step": 44370
+    },
+    {
+      "epoch": 3.3070044709388973,
+      "grad_norm": 2.2928473949432373,
+      "learning_rate": 0.0002,
+      "loss": 2.4589,
+      "step": 44380
+    },
+    {
+      "epoch": 3.3077496274217584,
+      "grad_norm": 2.4605629444122314,
+      "learning_rate": 0.0002,
+      "loss": 2.324,
+      "step": 44390
+    },
+    {
+      "epoch": 3.30849478390462,
+      "grad_norm": 2.7441422939300537,
+      "learning_rate": 0.0002,
+      "loss": 2.333,
+      "step": 44400
+    },
+    {
+      "epoch": 3.3092399403874815,
+      "grad_norm": 2.4191830158233643,
+      "learning_rate": 0.0002,
+      "loss": 2.5422,
+      "step": 44410
+    },
+    {
+      "epoch": 3.3099850968703426,
+      "grad_norm": 2.452444314956665,
+      "learning_rate": 0.0002,
+      "loss": 2.377,
+      "step": 44420
+    },
+    {
+      "epoch": 3.310730253353204,
+      "grad_norm": 2.9849095344543457,
+      "learning_rate": 0.0002,
+      "loss": 2.5103,
+      "step": 44430
+    },
+    {
+      "epoch": 3.3114754098360657,
+      "grad_norm": 2.7864060401916504,
+      "learning_rate": 0.0002,
+      "loss": 2.4538,
+      "step": 44440
+    },
+    {
+      "epoch": 3.312220566318927,
+      "grad_norm": 2.8097054958343506,
+      "learning_rate": 0.0002,
+      "loss": 2.4729,
+      "step": 44450
+    },
+    {
+      "epoch": 3.3129657228017884,
+      "grad_norm": 2.5728414058685303,
+      "learning_rate": 0.0002,
+      "loss": 2.4751,
+      "step": 44460
+    },
+    {
+      "epoch": 3.31371087928465,
+      "grad_norm": 2.7810518741607666,
+      "learning_rate": 0.0002,
+      "loss": 2.5308,
+      "step": 44470
+    },
+    {
+      "epoch": 3.314456035767511,
+      "grad_norm": 2.845767021179199,
+      "learning_rate": 0.0002,
+      "loss": 2.4061,
+      "step": 44480
+    },
+    {
+      "epoch": 3.3152011922503726,
+      "grad_norm": 2.721355438232422,
+      "learning_rate": 0.0002,
+      "loss": 2.5016,
+      "step": 44490
+    },
+    {
+      "epoch": 3.315946348733234,
+      "grad_norm": 2.6908435821533203,
+      "learning_rate": 0.0002,
+      "loss": 2.576,
+      "step": 44500
+    },
+    {
+      "epoch": 3.3166915052160952,
+      "grad_norm": 2.6813907623291016,
+      "learning_rate": 0.0002,
+      "loss": 2.6311,
+      "step": 44510
+    },
+    {
+      "epoch": 3.317436661698957,
+      "grad_norm": 3.2077786922454834,
+      "learning_rate": 0.0002,
+      "loss": 2.5925,
+      "step": 44520
+    },
+    {
+      "epoch": 3.3181818181818183,
+      "grad_norm": 2.6613762378692627,
+      "learning_rate": 0.0002,
+      "loss": 2.506,
+      "step": 44530
+    },
+    {
+      "epoch": 3.3189269746646795,
+      "grad_norm": 2.0632989406585693,
+      "learning_rate": 0.0002,
+      "loss": 2.3812,
+      "step": 44540
+    },
+    {
+      "epoch": 3.319672131147541,
+      "grad_norm": 2.6608712673187256,
+      "learning_rate": 0.0002,
+      "loss": 2.319,
+      "step": 44550
+    },
+    {
+      "epoch": 3.3204172876304026,
+      "grad_norm": 3.02986216545105,
+      "learning_rate": 0.0002,
+      "loss": 2.2761,
+      "step": 44560
+    },
+    {
+      "epoch": 3.3211624441132637,
+      "grad_norm": 2.6622636318206787,
+      "learning_rate": 0.0002,
+      "loss": 2.4661,
+      "step": 44570
+    },
+    {
+      "epoch": 3.321907600596125,
+      "grad_norm": 2.753340721130371,
+      "learning_rate": 0.0002,
+      "loss": 2.4398,
+      "step": 44580
+    },
+    {
+      "epoch": 3.3226527570789868,
+      "grad_norm": 2.8317463397979736,
+      "learning_rate": 0.0002,
+      "loss": 2.4028,
+      "step": 44590
+    },
+    {
+      "epoch": 3.323397913561848,
+      "grad_norm": 2.6618127822875977,
+      "learning_rate": 0.0002,
+      "loss": 2.469,
+      "step": 44600
+    },
+    {
+      "epoch": 3.3241430700447094,
+      "grad_norm": 2.7465057373046875,
+      "learning_rate": 0.0002,
+      "loss": 2.6312,
+      "step": 44610
+    },
+    {
+      "epoch": 3.3248882265275705,
+      "grad_norm": 2.6905951499938965,
+      "learning_rate": 0.0002,
+      "loss": 2.5652,
+      "step": 44620
+    },
+    {
+      "epoch": 3.325633383010432,
+      "grad_norm": 2.609553337097168,
+      "learning_rate": 0.0002,
+      "loss": 2.4733,
+      "step": 44630
+    },
+    {
+      "epoch": 3.3263785394932937,
+      "grad_norm": 3.0250022411346436,
+      "learning_rate": 0.0002,
+      "loss": 2.2852,
+      "step": 44640
+    },
+    {
+      "epoch": 3.327123695976155,
+      "grad_norm": 2.7133560180664062,
+      "learning_rate": 0.0002,
+      "loss": 2.6097,
+      "step": 44650
+    },
+    {
+      "epoch": 3.3278688524590163,
+      "grad_norm": 2.479445695877075,
+      "learning_rate": 0.0002,
+      "loss": 2.4231,
+      "step": 44660
+    },
+    {
+      "epoch": 3.328614008941878,
+      "grad_norm": 2.7370223999023438,
+      "learning_rate": 0.0002,
+      "loss": 2.359,
+      "step": 44670
+    },
+    {
+      "epoch": 3.329359165424739,
+      "grad_norm": 2.5486905574798584,
+      "learning_rate": 0.0002,
+      "loss": 2.5112,
+      "step": 44680
+    },
+    {
+      "epoch": 3.3301043219076005,
+      "grad_norm": 2.0843801498413086,
+      "learning_rate": 0.0002,
+      "loss": 2.4368,
+      "step": 44690
+    },
+    {
+      "epoch": 3.330849478390462,
+      "grad_norm": 2.692003011703491,
+      "learning_rate": 0.0002,
+      "loss": 2.4591,
+      "step": 44700
+    },
+    {
+      "epoch": 3.3315946348733236,
+      "grad_norm": 2.4572720527648926,
+      "learning_rate": 0.0002,
+      "loss": 2.521,
+      "step": 44710
+    },
+    {
+      "epoch": 3.3323397913561847,
+      "grad_norm": 2.675017833709717,
+      "learning_rate": 0.0002,
+      "loss": 2.5644,
+      "step": 44720
+    },
+    {
+      "epoch": 3.3330849478390463,
+      "grad_norm": 2.6414129734039307,
+      "learning_rate": 0.0002,
+      "loss": 2.601,
+      "step": 44730
+    },
+    {
+      "epoch": 3.3338301043219074,
+      "grad_norm": 2.5128352642059326,
+      "learning_rate": 0.0002,
+      "loss": 2.4304,
+      "step": 44740
+    },
+    {
+      "epoch": 3.334575260804769,
+      "grad_norm": 2.5863492488861084,
+      "learning_rate": 0.0002,
+      "loss": 2.3684,
+      "step": 44750
+    },
+    {
+      "epoch": 3.3353204172876305,
+      "grad_norm": 2.2314746379852295,
+      "learning_rate": 0.0002,
+      "loss": 2.5097,
+      "step": 44760
+    },
+    {
+      "epoch": 3.3360655737704916,
+      "grad_norm": 2.5924713611602783,
+      "learning_rate": 0.0002,
+      "loss": 2.5148,
+      "step": 44770
+    },
+    {
+      "epoch": 3.336810730253353,
+      "grad_norm": 2.298295497894287,
+      "learning_rate": 0.0002,
+      "loss": 2.3235,
+      "step": 44780
+    },
+    {
+      "epoch": 3.3375558867362147,
+      "grad_norm": 2.701958417892456,
+      "learning_rate": 0.0002,
+      "loss": 2.4991,
+      "step": 44790
+    },
+    {
+      "epoch": 3.338301043219076,
+      "grad_norm": 2.766066074371338,
+      "learning_rate": 0.0002,
+      "loss": 2.4484,
+      "step": 44800
+    },
+    {
+      "epoch": 3.3390461997019374,
+      "grad_norm": 2.684408187866211,
+      "learning_rate": 0.0002,
+      "loss": 2.5959,
+      "step": 44810
+    },
+    {
+      "epoch": 3.339791356184799,
+      "grad_norm": 2.5606040954589844,
+      "learning_rate": 0.0002,
+      "loss": 2.3181,
+      "step": 44820
+    },
+    {
+      "epoch": 3.34053651266766,
+      "grad_norm": 2.261032819747925,
+      "learning_rate": 0.0002,
+      "loss": 2.3889,
+      "step": 44830
+    },
+    {
+      "epoch": 3.3412816691505216,
+      "grad_norm": 2.6086981296539307,
+      "learning_rate": 0.0002,
+      "loss": 2.3264,
+      "step": 44840
+    },
+    {
+      "epoch": 3.342026825633383,
+      "grad_norm": 2.7089896202087402,
+      "learning_rate": 0.0002,
+      "loss": 2.4724,
+      "step": 44850
+    },
+    {
+      "epoch": 3.3427719821162443,
+      "grad_norm": 2.4596564769744873,
+      "learning_rate": 0.0002,
+      "loss": 2.3453,
+      "step": 44860
+    },
+    {
+      "epoch": 3.343517138599106,
+      "grad_norm": 3.0206546783447266,
+      "learning_rate": 0.0002,
+      "loss": 2.274,
+      "step": 44870
+    },
+    {
+      "epoch": 3.3442622950819674,
+      "grad_norm": 2.696455717086792,
+      "learning_rate": 0.0002,
+      "loss": 2.5678,
+      "step": 44880
+    },
+    {
+      "epoch": 3.3450074515648285,
+      "grad_norm": 2.755286693572998,
+      "learning_rate": 0.0002,
+      "loss": 2.6242,
+      "step": 44890
+    },
+    {
+      "epoch": 3.34575260804769,
+      "grad_norm": 2.4545297622680664,
+      "learning_rate": 0.0002,
+      "loss": 2.4866,
+      "step": 44900
+    },
+    {
+      "epoch": 3.3464977645305516,
+      "grad_norm": 1.9408307075500488,
+      "learning_rate": 0.0002,
+      "loss": 2.2509,
+      "step": 44910
+    },
+    {
+      "epoch": 3.3472429210134127,
+      "grad_norm": 2.969522714614868,
+      "learning_rate": 0.0002,
+      "loss": 2.5758,
+      "step": 44920
+    },
+    {
+      "epoch": 3.3479880774962743,
+      "grad_norm": 2.4225828647613525,
+      "learning_rate": 0.0002,
+      "loss": 2.2521,
+      "step": 44930
+    },
+    {
+      "epoch": 3.348733233979136,
+      "grad_norm": 2.567896842956543,
+      "learning_rate": 0.0002,
+      "loss": 2.5203,
+      "step": 44940
+    },
+    {
+      "epoch": 3.349478390461997,
+      "grad_norm": 2.6760551929473877,
+      "learning_rate": 0.0002,
+      "loss": 2.4264,
+      "step": 44950
+    },
+    {
+      "epoch": 3.3502235469448585,
+      "grad_norm": 2.5255861282348633,
+      "learning_rate": 0.0002,
+      "loss": 2.5705,
+      "step": 44960
+    },
+    {
+      "epoch": 3.3509687034277196,
+      "grad_norm": 2.5304393768310547,
+      "learning_rate": 0.0002,
+      "loss": 2.3158,
+      "step": 44970
+    },
+    {
+      "epoch": 3.351713859910581,
+      "grad_norm": 2.5755674839019775,
+      "learning_rate": 0.0002,
+      "loss": 2.4185,
+      "step": 44980
+    },
+    {
+      "epoch": 3.3524590163934427,
+      "grad_norm": 2.782089948654175,
+      "learning_rate": 0.0002,
+      "loss": 2.5002,
+      "step": 44990
+    },
+    {
+      "epoch": 3.3532041728763042,
+      "grad_norm": 2.9056670665740967,
+      "learning_rate": 0.0002,
+      "loss": 2.5654,
+      "step": 45000
+    },
+    {
+      "epoch": 3.3539493293591653,
+      "grad_norm": 3.098062038421631,
+      "learning_rate": 0.0002,
+      "loss": 2.4097,
+      "step": 45010
+    },
+    {
+      "epoch": 3.354694485842027,
+      "grad_norm": 2.8015189170837402,
+      "learning_rate": 0.0002,
+      "loss": 2.4039,
+      "step": 45020
+    },
+    {
+      "epoch": 3.355439642324888,
+      "grad_norm": 2.5579729080200195,
+      "learning_rate": 0.0002,
+      "loss": 2.4874,
+      "step": 45030
+    },
+    {
+      "epoch": 3.3561847988077496,
+      "grad_norm": 2.528961181640625,
+      "learning_rate": 0.0002,
+      "loss": 2.4139,
+      "step": 45040
+    },
+    {
+      "epoch": 3.356929955290611,
+      "grad_norm": 2.753173828125,
+      "learning_rate": 0.0002,
+      "loss": 2.2572,
+      "step": 45050
+    },
+    {
+      "epoch": 3.3576751117734727,
+      "grad_norm": 2.733590841293335,
+      "learning_rate": 0.0002,
+      "loss": 2.571,
+      "step": 45060
+    },
+    {
+      "epoch": 3.3584202682563338,
+      "grad_norm": 2.6290042400360107,
+      "learning_rate": 0.0002,
+      "loss": 2.5703,
+      "step": 45070
+    },
+    {
+      "epoch": 3.3591654247391953,
+      "grad_norm": 2.5969390869140625,
+      "learning_rate": 0.0002,
+      "loss": 2.3944,
+      "step": 45080
+    },
+    {
+      "epoch": 3.3599105812220564,
+      "grad_norm": 2.6536154747009277,
+      "learning_rate": 0.0002,
+      "loss": 2.4431,
+      "step": 45090
+    },
+    {
+      "epoch": 3.360655737704918,
+      "grad_norm": 2.6232528686523438,
+      "learning_rate": 0.0002,
+      "loss": 2.502,
+      "step": 45100
+    },
+    {
+      "epoch": 3.3614008941877795,
+      "grad_norm": 2.800161361694336,
+      "learning_rate": 0.0002,
+      "loss": 2.3932,
+      "step": 45110
+    },
+    {
+      "epoch": 3.3621460506706407,
+      "grad_norm": 2.5421383380889893,
+      "learning_rate": 0.0002,
+      "loss": 2.527,
+      "step": 45120
+    },
+    {
+      "epoch": 3.362891207153502,
+      "grad_norm": 2.968223810195923,
+      "learning_rate": 0.0002,
+      "loss": 2.6396,
+      "step": 45130
+    },
+    {
+      "epoch": 3.3636363636363638,
+      "grad_norm": 2.4359583854675293,
+      "learning_rate": 0.0002,
+      "loss": 2.3573,
+      "step": 45140
+    },
+    {
+      "epoch": 3.364381520119225,
+      "grad_norm": 2.3761839866638184,
+      "learning_rate": 0.0002,
+      "loss": 2.6412,
+      "step": 45150
+    },
+    {
+      "epoch": 3.3651266766020864,
+      "grad_norm": 2.7938098907470703,
+      "learning_rate": 0.0002,
+      "loss": 2.5008,
+      "step": 45160
+    },
+    {
+      "epoch": 3.365871833084948,
+      "grad_norm": 2.7078990936279297,
+      "learning_rate": 0.0002,
+      "loss": 2.627,
+      "step": 45170
+    },
+    {
+      "epoch": 3.366616989567809,
+      "grad_norm": 2.781672477722168,
+      "learning_rate": 0.0002,
+      "loss": 2.5451,
+      "step": 45180
+    },
+    {
+      "epoch": 3.3673621460506706,
+      "grad_norm": 2.7269418239593506,
+      "learning_rate": 0.0002,
+      "loss": 2.3616,
+      "step": 45190
+    },
+    {
+      "epoch": 3.368107302533532,
+      "grad_norm": 2.7787158489227295,
+      "learning_rate": 0.0002,
+      "loss": 2.4846,
+      "step": 45200
+    },
+    {
+      "epoch": 3.3688524590163933,
+      "grad_norm": 2.631345748901367,
+      "learning_rate": 0.0002,
+      "loss": 2.556,
+      "step": 45210
+    },
+    {
+      "epoch": 3.369597615499255,
+      "grad_norm": 2.876192092895508,
+      "learning_rate": 0.0002,
+      "loss": 2.5152,
+      "step": 45220
+    },
+    {
+      "epoch": 3.3703427719821164,
+      "grad_norm": 2.6566519737243652,
+      "learning_rate": 0.0002,
+      "loss": 2.6127,
+      "step": 45230
+    },
+    {
+      "epoch": 3.3710879284649775,
+      "grad_norm": 3.0558583736419678,
+      "learning_rate": 0.0002,
+      "loss": 2.5733,
+      "step": 45240
+    },
+    {
+      "epoch": 3.371833084947839,
+      "grad_norm": 2.9002668857574463,
+      "learning_rate": 0.0002,
+      "loss": 2.5031,
+      "step": 45250
+    },
+    {
+      "epoch": 3.3725782414307006,
+      "grad_norm": 2.6893272399902344,
+      "learning_rate": 0.0002,
+      "loss": 2.5278,
+      "step": 45260
+    },
+    {
+      "epoch": 3.3733233979135617,
+      "grad_norm": 2.7107250690460205,
+      "learning_rate": 0.0002,
+      "loss": 2.6275,
+      "step": 45270
+    },
+    {
+      "epoch": 3.3740685543964233,
+      "grad_norm": 2.816999912261963,
+      "learning_rate": 0.0002,
+      "loss": 2.5468,
+      "step": 45280
+    },
+    {
+      "epoch": 3.374813710879285,
+      "grad_norm": 2.670982599258423,
+      "learning_rate": 0.0002,
+      "loss": 2.4967,
+      "step": 45290
+    },
+    {
+      "epoch": 3.375558867362146,
+      "grad_norm": 2.483621835708618,
+      "learning_rate": 0.0002,
+      "loss": 2.2782,
+      "step": 45300
+    },
+    {
+      "epoch": 3.3763040238450075,
+      "grad_norm": 2.733739137649536,
+      "learning_rate": 0.0002,
+      "loss": 2.3869,
+      "step": 45310
+    },
+    {
+      "epoch": 3.3770491803278686,
+      "grad_norm": 2.6917941570281982,
+      "learning_rate": 0.0002,
+      "loss": 2.4299,
+      "step": 45320
+    },
+    {
+      "epoch": 3.37779433681073,
+      "grad_norm": 2.675069570541382,
+      "learning_rate": 0.0002,
+      "loss": 2.5749,
+      "step": 45330
+    },
+    {
+      "epoch": 3.3785394932935917,
+      "grad_norm": 2.6892285346984863,
+      "learning_rate": 0.0002,
+      "loss": 2.387,
+      "step": 45340
+    },
+    {
+      "epoch": 3.3792846497764533,
+      "grad_norm": 3.1701433658599854,
+      "learning_rate": 0.0002,
+      "loss": 2.5472,
+      "step": 45350
+    },
+    {
+      "epoch": 3.3800298062593144,
+      "grad_norm": 2.8027126789093018,
+      "learning_rate": 0.0002,
+      "loss": 2.4505,
+      "step": 45360
+    },
+    {
+      "epoch": 3.380774962742176,
+      "grad_norm": 3.0367190837860107,
+      "learning_rate": 0.0002,
+      "loss": 2.6113,
+      "step": 45370
+    },
+    {
+      "epoch": 3.381520119225037,
+      "grad_norm": 2.723576545715332,
+      "learning_rate": 0.0002,
+      "loss": 2.3816,
+      "step": 45380
+    },
+    {
+      "epoch": 3.3822652757078986,
+      "grad_norm": 2.6705434322357178,
+      "learning_rate": 0.0002,
+      "loss": 2.4231,
+      "step": 45390
+    },
+    {
+      "epoch": 3.38301043219076,
+      "grad_norm": 2.584972381591797,
+      "learning_rate": 0.0002,
+      "loss": 2.3369,
+      "step": 45400
+    },
+    {
+      "epoch": 3.3837555886736217,
+      "grad_norm": 2.5815274715423584,
+      "learning_rate": 0.0002,
+      "loss": 2.3362,
+      "step": 45410
+    },
+    {
+      "epoch": 3.384500745156483,
+      "grad_norm": 2.7649471759796143,
+      "learning_rate": 0.0002,
+      "loss": 2.3817,
+      "step": 45420
+    },
+    {
+      "epoch": 3.3852459016393444,
+      "grad_norm": 2.435465097427368,
+      "learning_rate": 0.0002,
+      "loss": 2.5519,
+      "step": 45430
+    },
+    {
+      "epoch": 3.3859910581222055,
+      "grad_norm": 2.7562053203582764,
+      "learning_rate": 0.0002,
+      "loss": 2.4509,
+      "step": 45440
+    },
+    {
+      "epoch": 3.386736214605067,
+      "grad_norm": 2.829549789428711,
+      "learning_rate": 0.0002,
+      "loss": 2.4191,
+      "step": 45450
+    },
+    {
+      "epoch": 3.3874813710879286,
+      "grad_norm": 2.9970972537994385,
+      "learning_rate": 0.0002,
+      "loss": 2.5063,
+      "step": 45460
+    },
+    {
+      "epoch": 3.3882265275707897,
+      "grad_norm": 3.018434524536133,
+      "learning_rate": 0.0002,
+      "loss": 2.4994,
+      "step": 45470
+    },
+    {
+      "epoch": 3.3889716840536512,
+      "grad_norm": 2.6649057865142822,
+      "learning_rate": 0.0002,
+      "loss": 2.421,
+      "step": 45480
+    },
+    {
+      "epoch": 3.389716840536513,
+      "grad_norm": 2.2870116233825684,
+      "learning_rate": 0.0002,
+      "loss": 2.5703,
+      "step": 45490
+    },
+    {
+      "epoch": 3.390461997019374,
+      "grad_norm": 2.1558291912078857,
+      "learning_rate": 0.0002,
+      "loss": 2.618,
+      "step": 45500
+    },
+    {
+      "epoch": 3.3912071535022354,
+      "grad_norm": 2.7797162532806396,
+      "learning_rate": 0.0002,
+      "loss": 2.404,
+      "step": 45510
+    },
+    {
+      "epoch": 3.391952309985097,
+      "grad_norm": 3.0129196643829346,
+      "learning_rate": 0.0002,
+      "loss": 2.3708,
+      "step": 45520
+    },
+    {
+      "epoch": 3.392697466467958,
+      "grad_norm": 2.453110694885254,
+      "learning_rate": 0.0002,
+      "loss": 2.5923,
+      "step": 45530
+    },
+    {
+      "epoch": 3.3934426229508197,
+      "grad_norm": 2.845862865447998,
+      "learning_rate": 0.0002,
+      "loss": 2.5138,
+      "step": 45540
+    },
+    {
+      "epoch": 3.394187779433681,
+      "grad_norm": 3.013814687728882,
+      "learning_rate": 0.0002,
+      "loss": 2.6143,
+      "step": 45550
+    },
+    {
+      "epoch": 3.3949329359165423,
+      "grad_norm": 2.650480270385742,
+      "learning_rate": 0.0002,
+      "loss": 2.6253,
+      "step": 45560
+    },
+    {
+      "epoch": 3.395678092399404,
+      "grad_norm": 2.6634345054626465,
+      "learning_rate": 0.0002,
+      "loss": 2.4054,
+      "step": 45570
+    },
+    {
+      "epoch": 3.3964232488822654,
+      "grad_norm": 2.5562565326690674,
+      "learning_rate": 0.0002,
+      "loss": 2.3089,
+      "step": 45580
+    },
+    {
+      "epoch": 3.3971684053651265,
+      "grad_norm": 2.798459768295288,
+      "learning_rate": 0.0002,
+      "loss": 2.3853,
+      "step": 45590
+    },
+    {
+      "epoch": 3.397913561847988,
+      "grad_norm": 2.3773062229156494,
+      "learning_rate": 0.0002,
+      "loss": 2.5745,
+      "step": 45600
+    },
+    {
+      "epoch": 3.3986587183308496,
+      "grad_norm": 2.7547972202301025,
+      "learning_rate": 0.0002,
+      "loss": 2.4046,
+      "step": 45610
+    },
+    {
+      "epoch": 3.3994038748137108,
+      "grad_norm": 2.549546003341675,
+      "learning_rate": 0.0002,
+      "loss": 2.6018,
+      "step": 45620
+    },
+    {
+      "epoch": 3.4001490312965723,
+      "grad_norm": 2.604703903198242,
+      "learning_rate": 0.0002,
+      "loss": 2.5778,
+      "step": 45630
+    },
+    {
+      "epoch": 3.400894187779434,
+      "grad_norm": 2.515007495880127,
+      "learning_rate": 0.0002,
+      "loss": 2.4235,
+      "step": 45640
+    },
+    {
+      "epoch": 3.401639344262295,
+      "grad_norm": 3.1332290172576904,
+      "learning_rate": 0.0002,
+      "loss": 2.5225,
+      "step": 45650
+    },
+    {
+      "epoch": 3.4023845007451565,
+      "grad_norm": 2.7233545780181885,
+      "learning_rate": 0.0002,
+      "loss": 2.5562,
+      "step": 45660
+    },
+    {
+      "epoch": 3.4031296572280176,
+      "grad_norm": 2.413351058959961,
+      "learning_rate": 0.0002,
+      "loss": 2.4198,
+      "step": 45670
+    },
+    {
+      "epoch": 3.403874813710879,
+      "grad_norm": 2.706968307495117,
+      "learning_rate": 0.0002,
+      "loss": 2.44,
+      "step": 45680
+    },
+    {
+      "epoch": 3.4046199701937407,
+      "grad_norm": 2.8159170150756836,
+      "learning_rate": 0.0002,
+      "loss": 2.499,
+      "step": 45690
+    },
+    {
+      "epoch": 3.4053651266766023,
+      "grad_norm": 2.412567615509033,
+      "learning_rate": 0.0002,
+      "loss": 2.4767,
+      "step": 45700
+    },
+    {
+      "epoch": 3.4061102831594634,
+      "grad_norm": 2.582634210586548,
+      "learning_rate": 0.0002,
+      "loss": 2.3958,
+      "step": 45710
+    },
+    {
+      "epoch": 3.406855439642325,
+      "grad_norm": 2.5285251140594482,
+      "learning_rate": 0.0002,
+      "loss": 2.332,
+      "step": 45720
+    },
+    {
+      "epoch": 3.407600596125186,
+      "grad_norm": 2.7439749240875244,
+      "learning_rate": 0.0002,
+      "loss": 2.3391,
+      "step": 45730
+    },
+    {
+      "epoch": 3.4083457526080476,
+      "grad_norm": 2.710297107696533,
+      "learning_rate": 0.0002,
+      "loss": 2.5074,
+      "step": 45740
+    },
+    {
+      "epoch": 3.409090909090909,
+      "grad_norm": 2.586346387863159,
+      "learning_rate": 0.0002,
+      "loss": 2.5411,
+      "step": 45750
+    },
+    {
+      "epoch": 3.4098360655737707,
+      "grad_norm": 2.2955081462860107,
+      "learning_rate": 0.0002,
+      "loss": 2.5316,
+      "step": 45760
+    },
+    {
+      "epoch": 3.410581222056632,
+      "grad_norm": 2.482043504714966,
+      "learning_rate": 0.0002,
+      "loss": 2.5112,
+      "step": 45770
+    },
+    {
+      "epoch": 3.4113263785394934,
+      "grad_norm": 2.5250322818756104,
+      "learning_rate": 0.0002,
+      "loss": 2.4582,
+      "step": 45780
+    },
+    {
+      "epoch": 3.4120715350223545,
+      "grad_norm": 3.7372679710388184,
+      "learning_rate": 0.0002,
+      "loss": 2.6059,
+      "step": 45790
+    },
+    {
+      "epoch": 3.412816691505216,
+      "grad_norm": 2.6757683753967285,
+      "learning_rate": 0.0002,
+      "loss": 2.421,
+      "step": 45800
+    },
+    {
+      "epoch": 3.4135618479880776,
+      "grad_norm": 2.1661860942840576,
+      "learning_rate": 0.0002,
+      "loss": 2.243,
+      "step": 45810
+    },
+    {
+      "epoch": 3.4143070044709387,
+      "grad_norm": 2.95503830909729,
+      "learning_rate": 0.0002,
+      "loss": 2.5708,
+      "step": 45820
+    },
+    {
+      "epoch": 3.4150521609538003,
+      "grad_norm": 2.38747239112854,
+      "learning_rate": 0.0002,
+      "loss": 2.4036,
+      "step": 45830
+    },
+    {
+      "epoch": 3.415797317436662,
+      "grad_norm": 2.2388579845428467,
+      "learning_rate": 0.0002,
+      "loss": 2.4164,
+      "step": 45840
+    },
+    {
+      "epoch": 3.416542473919523,
+      "grad_norm": 2.6116092205047607,
+      "learning_rate": 0.0002,
+      "loss": 2.6254,
+      "step": 45850
+    },
+    {
+      "epoch": 3.4172876304023845,
+      "grad_norm": 2.408705472946167,
+      "learning_rate": 0.0002,
+      "loss": 2.6464,
+      "step": 45860
+    },
+    {
+      "epoch": 3.418032786885246,
+      "grad_norm": 2.4243757724761963,
+      "learning_rate": 0.0002,
+      "loss": 2.6336,
+      "step": 45870
+    },
+    {
+      "epoch": 3.418777943368107,
+      "grad_norm": 2.9750640392303467,
+      "learning_rate": 0.0002,
+      "loss": 2.4791,
+      "step": 45880
+    },
+    {
+      "epoch": 3.4195230998509687,
+      "grad_norm": 2.836698055267334,
+      "learning_rate": 0.0002,
+      "loss": 2.6173,
+      "step": 45890
+    },
+    {
+      "epoch": 3.4202682563338302,
+      "grad_norm": 2.734327793121338,
+      "learning_rate": 0.0002,
+      "loss": 2.3956,
+      "step": 45900
+    },
+    {
+      "epoch": 3.4210134128166914,
+      "grad_norm": 2.6513071060180664,
+      "learning_rate": 0.0002,
+      "loss": 2.5761,
+      "step": 45910
+    },
+    {
+      "epoch": 3.421758569299553,
+      "grad_norm": 2.8348026275634766,
+      "learning_rate": 0.0002,
+      "loss": 2.4796,
+      "step": 45920
+    },
+    {
+      "epoch": 3.4225037257824145,
+      "grad_norm": 2.5720503330230713,
+      "learning_rate": 0.0002,
+      "loss": 2.6315,
+      "step": 45930
+    },
+    {
+      "epoch": 3.4232488822652756,
+      "grad_norm": 2.497056245803833,
+      "learning_rate": 0.0002,
+      "loss": 2.4317,
+      "step": 45940
+    },
+    {
+      "epoch": 3.423994038748137,
+      "grad_norm": 2.849952459335327,
+      "learning_rate": 0.0002,
+      "loss": 2.4361,
+      "step": 45950
+    },
+    {
+      "epoch": 3.4247391952309987,
+      "grad_norm": 2.823962688446045,
+      "learning_rate": 0.0002,
+      "loss": 2.6059,
+      "step": 45960
+    },
+    {
+      "epoch": 3.42548435171386,
+      "grad_norm": 2.484067440032959,
+      "learning_rate": 0.0002,
+      "loss": 2.3086,
+      "step": 45970
+    },
+    {
+      "epoch": 3.4262295081967213,
+      "grad_norm": 3.0087854862213135,
+      "learning_rate": 0.0002,
+      "loss": 2.3751,
+      "step": 45980
+    },
+    {
+      "epoch": 3.426974664679583,
+      "grad_norm": 2.5996415615081787,
+      "learning_rate": 0.0002,
+      "loss": 2.5007,
+      "step": 45990
+    },
+    {
+      "epoch": 3.427719821162444,
+      "grad_norm": 2.758151054382324,
+      "learning_rate": 0.0002,
+      "loss": 2.4652,
+      "step": 46000
+    },
+    {
+      "epoch": 3.4284649776453056,
+      "grad_norm": 2.789275646209717,
+      "learning_rate": 0.0002,
+      "loss": 2.645,
+      "step": 46010
+    },
+    {
+      "epoch": 3.429210134128167,
+      "grad_norm": 2.859830141067505,
+      "learning_rate": 0.0002,
+      "loss": 2.5338,
+      "step": 46020
+    },
+    {
+      "epoch": 3.429955290611028,
+      "grad_norm": 2.25303053855896,
+      "learning_rate": 0.0002,
+      "loss": 2.2158,
+      "step": 46030
+    },
+    {
+      "epoch": 3.4307004470938898,
+      "grad_norm": 3.3532886505126953,
+      "learning_rate": 0.0002,
+      "loss": 2.335,
+      "step": 46040
+    },
+    {
+      "epoch": 3.4314456035767513,
+      "grad_norm": 2.6834700107574463,
+      "learning_rate": 0.0002,
+      "loss": 2.662,
+      "step": 46050
+    },
+    {
+      "epoch": 3.4321907600596124,
+      "grad_norm": 2.6147358417510986,
+      "learning_rate": 0.0002,
+      "loss": 2.4504,
+      "step": 46060
+    },
+    {
+      "epoch": 3.432935916542474,
+      "grad_norm": 2.483168363571167,
+      "learning_rate": 0.0002,
+      "loss": 2.241,
+      "step": 46070
+    },
+    {
+      "epoch": 3.433681073025335,
+      "grad_norm": 2.5020689964294434,
+      "learning_rate": 0.0002,
+      "loss": 2.2708,
+      "step": 46080
+    },
+    {
+      "epoch": 3.4344262295081966,
+      "grad_norm": 2.6769652366638184,
+      "learning_rate": 0.0002,
+      "loss": 2.4544,
+      "step": 46090
+    },
+    {
+      "epoch": 3.435171385991058,
+      "grad_norm": 2.70047926902771,
+      "learning_rate": 0.0002,
+      "loss": 2.5237,
+      "step": 46100
+    },
+    {
+      "epoch": 3.4359165424739198,
+      "grad_norm": 2.588719367980957,
+      "learning_rate": 0.0002,
+      "loss": 2.4592,
+      "step": 46110
+    },
+    {
+      "epoch": 3.436661698956781,
+      "grad_norm": 2.694007158279419,
+      "learning_rate": 0.0002,
+      "loss": 2.5565,
+      "step": 46120
+    },
+    {
+      "epoch": 3.4374068554396424,
+      "grad_norm": 2.5668649673461914,
+      "learning_rate": 0.0002,
+      "loss": 2.414,
+      "step": 46130
+    },
+    {
+      "epoch": 3.4381520119225035,
+      "grad_norm": 2.8535616397857666,
+      "learning_rate": 0.0002,
+      "loss": 2.5938,
+      "step": 46140
+    },
+    {
+      "epoch": 3.438897168405365,
+      "grad_norm": 2.704505443572998,
+      "learning_rate": 0.0002,
+      "loss": 2.488,
+      "step": 46150
+    },
+    {
+      "epoch": 3.4396423248882266,
+      "grad_norm": 2.4701411724090576,
+      "learning_rate": 0.0002,
+      "loss": 2.3884,
+      "step": 46160
+    },
+    {
+      "epoch": 3.4403874813710877,
+      "grad_norm": 2.898048162460327,
+      "learning_rate": 0.0002,
+      "loss": 2.5974,
+      "step": 46170
+    },
+    {
+      "epoch": 3.4411326378539493,
+      "grad_norm": 2.504690170288086,
+      "learning_rate": 0.0002,
+      "loss": 2.5142,
+      "step": 46180
+    },
+    {
+      "epoch": 3.441877794336811,
+      "grad_norm": 2.4398272037506104,
+      "learning_rate": 0.0002,
+      "loss": 2.3432,
+      "step": 46190
+    },
+    {
+      "epoch": 3.442622950819672,
+      "grad_norm": 2.6902806758880615,
+      "learning_rate": 0.0002,
+      "loss": 2.5055,
+      "step": 46200
+    },
+    {
+      "epoch": 3.4433681073025335,
+      "grad_norm": 3.127070426940918,
+      "learning_rate": 0.0002,
+      "loss": 2.3998,
+      "step": 46210
+    },
+    {
+      "epoch": 3.444113263785395,
+      "grad_norm": 2.8394556045532227,
+      "learning_rate": 0.0002,
+      "loss": 2.5031,
+      "step": 46220
+    },
+    {
+      "epoch": 3.444858420268256,
+      "grad_norm": 2.2959282398223877,
+      "learning_rate": 0.0002,
+      "loss": 2.4069,
+      "step": 46230
+    },
+    {
+      "epoch": 3.4456035767511177,
+      "grad_norm": 2.881746530532837,
+      "learning_rate": 0.0002,
+      "loss": 2.4572,
+      "step": 46240
+    },
+    {
+      "epoch": 3.4463487332339793,
+      "grad_norm": 2.929504871368408,
+      "learning_rate": 0.0002,
+      "loss": 2.2962,
+      "step": 46250
+    },
+    {
+      "epoch": 3.4470938897168404,
+      "grad_norm": 2.570570945739746,
+      "learning_rate": 0.0002,
+      "loss": 2.3323,
+      "step": 46260
+    },
+    {
+      "epoch": 3.447839046199702,
+      "grad_norm": 2.8684487342834473,
+      "learning_rate": 0.0002,
+      "loss": 2.5943,
+      "step": 46270
+    },
+    {
+      "epoch": 3.4485842026825635,
+      "grad_norm": 2.15544056892395,
+      "learning_rate": 0.0002,
+      "loss": 2.4587,
+      "step": 46280
+    },
+    {
+      "epoch": 3.4493293591654246,
+      "grad_norm": 2.5828864574432373,
+      "learning_rate": 0.0002,
+      "loss": 2.5151,
+      "step": 46290
+    },
+    {
+      "epoch": 3.450074515648286,
+      "grad_norm": 2.892545461654663,
+      "learning_rate": 0.0002,
+      "loss": 2.4213,
+      "step": 46300
+    },
+    {
+      "epoch": 3.4508196721311477,
+      "grad_norm": 2.867642879486084,
+      "learning_rate": 0.0002,
+      "loss": 2.4857,
+      "step": 46310
+    },
+    {
+      "epoch": 3.451564828614009,
+      "grad_norm": 3.811511278152466,
+      "learning_rate": 0.0002,
+      "loss": 2.5187,
+      "step": 46320
+    },
+    {
+      "epoch": 3.4523099850968704,
+      "grad_norm": 1.9235666990280151,
+      "learning_rate": 0.0002,
+      "loss": 2.2494,
+      "step": 46330
+    },
+    {
+      "epoch": 3.453055141579732,
+      "grad_norm": 2.4899580478668213,
+      "learning_rate": 0.0002,
+      "loss": 2.3877,
+      "step": 46340
+    },
+    {
+      "epoch": 3.453800298062593,
+      "grad_norm": 2.7740957736968994,
+      "learning_rate": 0.0002,
+      "loss": 2.5916,
+      "step": 46350
+    },
+    {
+      "epoch": 3.4545454545454546,
+      "grad_norm": 3.30594539642334,
+      "learning_rate": 0.0002,
+      "loss": 2.5507,
+      "step": 46360
+    },
+    {
+      "epoch": 3.455290611028316,
+      "grad_norm": 2.3682942390441895,
+      "learning_rate": 0.0002,
+      "loss": 2.5072,
+      "step": 46370
+    },
+    {
+      "epoch": 3.4560357675111772,
+      "grad_norm": 2.640885829925537,
+      "learning_rate": 0.0002,
+      "loss": 2.5571,
+      "step": 46380
+    },
+    {
+      "epoch": 3.456780923994039,
+      "grad_norm": 2.810642719268799,
+      "learning_rate": 0.0002,
+      "loss": 2.4887,
+      "step": 46390
+    },
+    {
+      "epoch": 3.4575260804769004,
+      "grad_norm": 2.425133228302002,
+      "learning_rate": 0.0002,
+      "loss": 2.3365,
+      "step": 46400
+    },
+    {
+      "epoch": 3.4582712369597615,
+      "grad_norm": 2.6030843257904053,
+      "learning_rate": 0.0002,
+      "loss": 2.475,
+      "step": 46410
+    },
+    {
+      "epoch": 3.459016393442623,
+      "grad_norm": 2.4549689292907715,
+      "learning_rate": 0.0002,
+      "loss": 2.4761,
+      "step": 46420
+    },
+    {
+      "epoch": 3.459761549925484,
+      "grad_norm": 2.5607504844665527,
+      "learning_rate": 0.0002,
+      "loss": 2.4621,
+      "step": 46430
+    },
+    {
+      "epoch": 3.4605067064083457,
+      "grad_norm": 2.6467771530151367,
+      "learning_rate": 0.0002,
+      "loss": 2.5109,
+      "step": 46440
+    },
+    {
+      "epoch": 3.4612518628912072,
+      "grad_norm": 2.1029367446899414,
+      "learning_rate": 0.0002,
+      "loss": 2.4962,
+      "step": 46450
+    },
+    {
+      "epoch": 3.461997019374069,
+      "grad_norm": 2.5813140869140625,
+      "learning_rate": 0.0002,
+      "loss": 2.6477,
+      "step": 46460
+    },
+    {
+      "epoch": 3.46274217585693,
+      "grad_norm": 2.906320571899414,
+      "learning_rate": 0.0002,
+      "loss": 2.5958,
+      "step": 46470
+    },
+    {
+      "epoch": 3.4634873323397914,
+      "grad_norm": 2.431466579437256,
+      "learning_rate": 0.0002,
+      "loss": 2.5054,
+      "step": 46480
+    },
+    {
+      "epoch": 3.4642324888226526,
+      "grad_norm": 2.375662088394165,
+      "learning_rate": 0.0002,
+      "loss": 2.4448,
+      "step": 46490
+    },
+    {
+      "epoch": 3.464977645305514,
+      "grad_norm": 2.675147771835327,
+      "learning_rate": 0.0002,
+      "loss": 2.4303,
+      "step": 46500
+    },
+    {
+      "epoch": 3.4657228017883757,
+      "grad_norm": 2.9978744983673096,
+      "learning_rate": 0.0002,
+      "loss": 2.3751,
+      "step": 46510
+    },
+    {
+      "epoch": 3.4664679582712368,
+      "grad_norm": 2.483147621154785,
+      "learning_rate": 0.0002,
+      "loss": 2.4115,
+      "step": 46520
+    },
+    {
+      "epoch": 3.4672131147540983,
+      "grad_norm": 2.3755555152893066,
+      "learning_rate": 0.0002,
+      "loss": 2.2329,
+      "step": 46530
+    },
+    {
+      "epoch": 3.46795827123696,
+      "grad_norm": 2.9722342491149902,
+      "learning_rate": 0.0002,
+      "loss": 2.66,
+      "step": 46540
+    },
+    {
+      "epoch": 3.468703427719821,
+      "grad_norm": 2.3765087127685547,
+      "learning_rate": 0.0002,
+      "loss": 2.3346,
+      "step": 46550
+    },
+    {
+      "epoch": 3.4694485842026825,
+      "grad_norm": 3.10359263420105,
+      "learning_rate": 0.0002,
+      "loss": 2.4465,
+      "step": 46560
+    },
+    {
+      "epoch": 3.470193740685544,
+      "grad_norm": 2.499055862426758,
+      "learning_rate": 0.0002,
+      "loss": 2.4433,
+      "step": 46570
+    },
+    {
+      "epoch": 3.470938897168405,
+      "grad_norm": 2.9596831798553467,
+      "learning_rate": 0.0002,
+      "loss": 2.6495,
+      "step": 46580
+    },
+    {
+      "epoch": 3.4716840536512668,
+      "grad_norm": 2.490100622177124,
+      "learning_rate": 0.0002,
+      "loss": 2.4167,
+      "step": 46590
+    },
+    {
+      "epoch": 3.4724292101341283,
+      "grad_norm": 2.3789775371551514,
+      "learning_rate": 0.0002,
+      "loss": 2.3696,
+      "step": 46600
+    },
+    {
+      "epoch": 3.4731743666169894,
+      "grad_norm": 2.522531032562256,
+      "learning_rate": 0.0002,
+      "loss": 2.416,
+      "step": 46610
+    },
+    {
+      "epoch": 3.473919523099851,
+      "grad_norm": 1.9312344789505005,
+      "learning_rate": 0.0002,
+      "loss": 2.3704,
+      "step": 46620
+    },
+    {
+      "epoch": 3.4746646795827125,
+      "grad_norm": 2.6317503452301025,
+      "learning_rate": 0.0002,
+      "loss": 2.4834,
+      "step": 46630
+    },
+    {
+      "epoch": 3.4754098360655736,
+      "grad_norm": 2.854759693145752,
+      "learning_rate": 0.0002,
+      "loss": 2.3015,
+      "step": 46640
+    },
+    {
+      "epoch": 3.476154992548435,
+      "grad_norm": 2.843125581741333,
+      "learning_rate": 0.0002,
+      "loss": 2.6554,
+      "step": 46650
+    },
+    {
+      "epoch": 3.4769001490312967,
+      "grad_norm": 2.7469377517700195,
+      "learning_rate": 0.0002,
+      "loss": 2.5061,
+      "step": 46660
+    },
+    {
+      "epoch": 3.477645305514158,
+      "grad_norm": 2.6205174922943115,
+      "learning_rate": 0.0002,
+      "loss": 2.256,
+      "step": 46670
+    },
+    {
+      "epoch": 3.4783904619970194,
+      "grad_norm": 2.8084607124328613,
+      "learning_rate": 0.0002,
+      "loss": 2.5537,
+      "step": 46680
+    },
+    {
+      "epoch": 3.479135618479881,
+      "grad_norm": 2.556391716003418,
+      "learning_rate": 0.0002,
+      "loss": 2.3825,
+      "step": 46690
+    },
+    {
+      "epoch": 3.479880774962742,
+      "grad_norm": 2.547045946121216,
+      "learning_rate": 0.0002,
+      "loss": 2.5505,
+      "step": 46700
+    },
+    {
+      "epoch": 3.4806259314456036,
+      "grad_norm": 2.9189865589141846,
+      "learning_rate": 0.0002,
+      "loss": 2.3604,
+      "step": 46710
+    },
+    {
+      "epoch": 3.481371087928465,
+      "grad_norm": 2.6150434017181396,
+      "learning_rate": 0.0002,
+      "loss": 2.5506,
+      "step": 46720
+    },
+    {
+      "epoch": 3.4821162444113263,
+      "grad_norm": 2.9110896587371826,
+      "learning_rate": 0.0002,
+      "loss": 2.5469,
+      "step": 46730
+    },
+    {
+      "epoch": 3.482861400894188,
+      "grad_norm": 2.4996886253356934,
+      "learning_rate": 0.0002,
+      "loss": 2.4058,
+      "step": 46740
+    },
+    {
+      "epoch": 3.4836065573770494,
+      "grad_norm": 2.6568424701690674,
+      "learning_rate": 0.0002,
+      "loss": 2.4281,
+      "step": 46750
+    },
+    {
+      "epoch": 3.4843517138599105,
+      "grad_norm": 2.790637254714966,
+      "learning_rate": 0.0002,
+      "loss": 2.2597,
+      "step": 46760
+    },
+    {
+      "epoch": 3.485096870342772,
+      "grad_norm": 2.712463855743408,
+      "learning_rate": 0.0002,
+      "loss": 2.6115,
+      "step": 46770
+    },
+    {
+      "epoch": 3.485842026825633,
+      "grad_norm": 2.838874340057373,
+      "learning_rate": 0.0002,
+      "loss": 2.5807,
+      "step": 46780
+    },
+    {
+      "epoch": 3.4865871833084947,
+      "grad_norm": 2.435051441192627,
+      "learning_rate": 0.0002,
+      "loss": 2.2417,
+      "step": 46790
+    },
+    {
+      "epoch": 3.4873323397913563,
+      "grad_norm": 2.4676756858825684,
+      "learning_rate": 0.0002,
+      "loss": 2.4674,
+      "step": 46800
+    },
+    {
+      "epoch": 3.488077496274218,
+      "grad_norm": 2.871751070022583,
+      "learning_rate": 0.0002,
+      "loss": 2.5139,
+      "step": 46810
+    },
+    {
+      "epoch": 3.488822652757079,
+      "grad_norm": 2.740755319595337,
+      "learning_rate": 0.0002,
+      "loss": 2.459,
+      "step": 46820
+    },
+    {
+      "epoch": 3.4895678092399405,
+      "grad_norm": 2.719113349914551,
+      "learning_rate": 0.0002,
+      "loss": 2.2784,
+      "step": 46830
+    },
+    {
+      "epoch": 3.4903129657228016,
+      "grad_norm": 2.9216134548187256,
+      "learning_rate": 0.0002,
+      "loss": 2.572,
+      "step": 46840
+    },
+    {
+      "epoch": 3.491058122205663,
+      "grad_norm": 2.4576525688171387,
+      "learning_rate": 0.0002,
+      "loss": 2.2147,
+      "step": 46850
+    },
+    {
+      "epoch": 3.4918032786885247,
+      "grad_norm": 2.628617525100708,
+      "learning_rate": 0.0002,
+      "loss": 2.6005,
+      "step": 46860
+    },
+    {
+      "epoch": 3.492548435171386,
+      "grad_norm": 2.8615686893463135,
+      "learning_rate": 0.0002,
+      "loss": 2.4068,
+      "step": 46870
+    },
+    {
+      "epoch": 3.4932935916542474,
+      "grad_norm": 2.8863871097564697,
+      "learning_rate": 0.0002,
+      "loss": 2.5519,
+      "step": 46880
+    },
+    {
+      "epoch": 3.494038748137109,
+      "grad_norm": 2.529972553253174,
+      "learning_rate": 0.0002,
+      "loss": 2.5626,
+      "step": 46890
+    },
+    {
+      "epoch": 3.49478390461997,
+      "grad_norm": 2.534062385559082,
+      "learning_rate": 0.0002,
+      "loss": 2.4962,
+      "step": 46900
+    },
+    {
+      "epoch": 3.4955290611028316,
+      "grad_norm": 2.641014575958252,
+      "learning_rate": 0.0002,
+      "loss": 2.3222,
+      "step": 46910
+    },
+    {
+      "epoch": 3.496274217585693,
+      "grad_norm": 2.8244290351867676,
+      "learning_rate": 0.0002,
+      "loss": 2.4623,
+      "step": 46920
+    },
+    {
+      "epoch": 3.4970193740685542,
+      "grad_norm": 2.8250889778137207,
+      "learning_rate": 0.0002,
+      "loss": 2.5361,
+      "step": 46930
+    },
+    {
+      "epoch": 3.497764530551416,
+      "grad_norm": 3.2792413234710693,
+      "learning_rate": 0.0002,
+      "loss": 2.6656,
+      "step": 46940
+    },
+    {
+      "epoch": 3.4985096870342773,
+      "grad_norm": 2.25415301322937,
+      "learning_rate": 0.0002,
+      "loss": 2.4449,
+      "step": 46950
+    },
+    {
+      "epoch": 3.4992548435171384,
+      "grad_norm": 2.867708921432495,
+      "learning_rate": 0.0002,
+      "loss": 2.3822,
+      "step": 46960
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.8867411613464355,
+      "learning_rate": 0.0002,
+      "loss": 2.4606,
+      "step": 46970
+    },
+    {
+      "epoch": 3.5007451564828616,
+      "grad_norm": 2.8825507164001465,
+      "learning_rate": 0.0002,
+      "loss": 2.5032,
+      "step": 46980
+    },
+    {
+      "epoch": 3.5014903129657227,
+      "grad_norm": 3.2540154457092285,
+      "learning_rate": 0.0002,
+      "loss": 2.4558,
+      "step": 46990
+    },
+    {
+      "epoch": 3.502235469448584,
+      "grad_norm": 2.831401824951172,
+      "learning_rate": 0.0002,
+      "loss": 2.54,
+      "step": 47000
+    },
+    {
+      "epoch": 3.5029806259314458,
+      "grad_norm": 2.632877826690674,
+      "learning_rate": 0.0002,
+      "loss": 2.3713,
+      "step": 47010
+    },
+    {
+      "epoch": 3.503725782414307,
+      "grad_norm": 2.836522340774536,
+      "learning_rate": 0.0002,
+      "loss": 2.5631,
+      "step": 47020
+    },
+    {
+      "epoch": 3.5044709388971684,
+      "grad_norm": 2.5134811401367188,
+      "learning_rate": 0.0002,
+      "loss": 2.4654,
+      "step": 47030
+    },
+    {
+      "epoch": 3.50521609538003,
+      "grad_norm": 2.8926382064819336,
+      "learning_rate": 0.0002,
+      "loss": 2.525,
+      "step": 47040
+    },
+    {
+      "epoch": 3.505961251862891,
+      "grad_norm": 2.3543901443481445,
+      "learning_rate": 0.0002,
+      "loss": 2.3846,
+      "step": 47050
+    },
+    {
+      "epoch": 3.5067064083457526,
+      "grad_norm": 2.8736624717712402,
+      "learning_rate": 0.0002,
+      "loss": 2.608,
+      "step": 47060
+    },
+    {
+      "epoch": 3.5074515648286138,
+      "grad_norm": 2.5337979793548584,
+      "learning_rate": 0.0002,
+      "loss": 2.5153,
+      "step": 47070
+    },
+    {
+      "epoch": 3.5081967213114753,
+      "grad_norm": 2.630679130554199,
+      "learning_rate": 0.0002,
+      "loss": 2.6418,
+      "step": 47080
+    },
+    {
+      "epoch": 3.508941877794337,
+      "grad_norm": 2.820647954940796,
+      "learning_rate": 0.0002,
+      "loss": 2.5799,
+      "step": 47090
+    },
+    {
+      "epoch": 3.5096870342771984,
+      "grad_norm": 2.707824230194092,
+      "learning_rate": 0.0002,
+      "loss": 2.3407,
+      "step": 47100
+    },
+    {
+      "epoch": 3.5104321907600595,
+      "grad_norm": 2.6139681339263916,
+      "learning_rate": 0.0002,
+      "loss": 2.5448,
+      "step": 47110
+    },
+    {
+      "epoch": 3.511177347242921,
+      "grad_norm": 2.887916326522827,
+      "learning_rate": 0.0002,
+      "loss": 2.4502,
+      "step": 47120
+    },
+    {
+      "epoch": 3.511922503725782,
+      "grad_norm": 2.730395555496216,
+      "learning_rate": 0.0002,
+      "loss": 2.4459,
+      "step": 47130
+    },
+    {
+      "epoch": 3.5126676602086437,
+      "grad_norm": 2.41025447845459,
+      "learning_rate": 0.0002,
+      "loss": 2.5996,
+      "step": 47140
+    },
+    {
+      "epoch": 3.5134128166915053,
+      "grad_norm": 2.7070274353027344,
+      "learning_rate": 0.0002,
+      "loss": 2.5929,
+      "step": 47150
+    },
+    {
+      "epoch": 3.514157973174367,
+      "grad_norm": 2.963433265686035,
+      "learning_rate": 0.0002,
+      "loss": 2.5335,
+      "step": 47160
+    },
+    {
+      "epoch": 3.514903129657228,
+      "grad_norm": 2.7727930545806885,
+      "learning_rate": 0.0002,
+      "loss": 2.5952,
+      "step": 47170
+    },
+    {
+      "epoch": 3.5156482861400895,
+      "grad_norm": 2.868478536605835,
+      "learning_rate": 0.0002,
+      "loss": 2.4693,
+      "step": 47180
+    },
+    {
+      "epoch": 3.5163934426229506,
+      "grad_norm": 2.550471305847168,
+      "learning_rate": 0.0002,
+      "loss": 2.6188,
+      "step": 47190
+    },
+    {
+      "epoch": 3.517138599105812,
+      "grad_norm": 2.536785125732422,
+      "learning_rate": 0.0002,
+      "loss": 2.513,
+      "step": 47200
+    },
+    {
+      "epoch": 3.5178837555886737,
+      "grad_norm": 2.470963716506958,
+      "learning_rate": 0.0002,
+      "loss": 2.4627,
+      "step": 47210
+    },
+    {
+      "epoch": 3.5186289120715353,
+      "grad_norm": 2.5856993198394775,
+      "learning_rate": 0.0002,
+      "loss": 2.4106,
+      "step": 47220
+    },
+    {
+      "epoch": 3.5193740685543964,
+      "grad_norm": 2.3822851181030273,
+      "learning_rate": 0.0002,
+      "loss": 2.4622,
+      "step": 47230
+    },
+    {
+      "epoch": 3.520119225037258,
+      "grad_norm": 2.6513640880584717,
+      "learning_rate": 0.0002,
+      "loss": 2.3524,
+      "step": 47240
+    },
+    {
+      "epoch": 3.520864381520119,
+      "grad_norm": 2.421938419342041,
+      "learning_rate": 0.0002,
+      "loss": 2.5163,
+      "step": 47250
+    },
+    {
+      "epoch": 3.5216095380029806,
+      "grad_norm": 2.7255730628967285,
+      "learning_rate": 0.0002,
+      "loss": 2.4664,
+      "step": 47260
+    },
+    {
+      "epoch": 3.522354694485842,
+      "grad_norm": 2.8659775257110596,
+      "learning_rate": 0.0002,
+      "loss": 2.5989,
+      "step": 47270
+    },
+    {
+      "epoch": 3.5230998509687033,
+      "grad_norm": 2.8360795974731445,
+      "learning_rate": 0.0002,
+      "loss": 2.5135,
+      "step": 47280
+    },
+    {
+      "epoch": 3.523845007451565,
+      "grad_norm": 2.400752544403076,
+      "learning_rate": 0.0002,
+      "loss": 2.2798,
+      "step": 47290
+    },
+    {
+      "epoch": 3.5245901639344264,
+      "grad_norm": 2.741732120513916,
+      "learning_rate": 0.0002,
+      "loss": 2.5847,
+      "step": 47300
+    },
+    {
+      "epoch": 3.5253353204172875,
+      "grad_norm": 2.8098509311676025,
+      "learning_rate": 0.0002,
+      "loss": 2.4662,
+      "step": 47310
+    },
+    {
+      "epoch": 3.526080476900149,
+      "grad_norm": 2.437133312225342,
+      "learning_rate": 0.0002,
+      "loss": 2.528,
+      "step": 47320
+    },
+    {
+      "epoch": 3.5268256333830106,
+      "grad_norm": 2.986236572265625,
+      "learning_rate": 0.0002,
+      "loss": 2.5172,
+      "step": 47330
+    },
+    {
+      "epoch": 3.5275707898658717,
+      "grad_norm": 2.7789604663848877,
+      "learning_rate": 0.0002,
+      "loss": 2.6075,
+      "step": 47340
+    },
+    {
+      "epoch": 3.5283159463487332,
+      "grad_norm": 2.8106555938720703,
+      "learning_rate": 0.0002,
+      "loss": 2.3869,
+      "step": 47350
+    },
+    {
+      "epoch": 3.529061102831595,
+      "grad_norm": 2.428912878036499,
+      "learning_rate": 0.0002,
+      "loss": 2.5395,
+      "step": 47360
+    },
+    {
+      "epoch": 3.529806259314456,
+      "grad_norm": 2.528440237045288,
+      "learning_rate": 0.0002,
+      "loss": 2.3945,
+      "step": 47370
+    },
+    {
+      "epoch": 3.5305514157973175,
+      "grad_norm": 2.547029972076416,
+      "learning_rate": 0.0002,
+      "loss": 2.4825,
+      "step": 47380
+    },
+    {
+      "epoch": 3.531296572280179,
+      "grad_norm": 2.567345142364502,
+      "learning_rate": 0.0002,
+      "loss": 2.4825,
+      "step": 47390
+    },
+    {
+      "epoch": 3.53204172876304,
+      "grad_norm": 2.5996034145355225,
+      "learning_rate": 0.0002,
+      "loss": 2.5573,
+      "step": 47400
+    },
+    {
+      "epoch": 3.5327868852459017,
+      "grad_norm": 2.373216390609741,
+      "learning_rate": 0.0002,
+      "loss": 2.432,
+      "step": 47410
+    },
+    {
+      "epoch": 3.533532041728763,
+      "grad_norm": 2.4746549129486084,
+      "learning_rate": 0.0002,
+      "loss": 2.6546,
+      "step": 47420
+    },
+    {
+      "epoch": 3.5342771982116243,
+      "grad_norm": 2.6175284385681152,
+      "learning_rate": 0.0002,
+      "loss": 2.5659,
+      "step": 47430
+    },
+    {
+      "epoch": 3.535022354694486,
+      "grad_norm": 2.7137603759765625,
+      "learning_rate": 0.0002,
+      "loss": 2.5188,
+      "step": 47440
+    },
+    {
+      "epoch": 3.5357675111773474,
+      "grad_norm": 2.842714548110962,
+      "learning_rate": 0.0002,
+      "loss": 2.637,
+      "step": 47450
+    },
+    {
+      "epoch": 3.5365126676602086,
+      "grad_norm": 2.494274139404297,
+      "learning_rate": 0.0002,
+      "loss": 2.4436,
+      "step": 47460
+    },
+    {
+      "epoch": 3.53725782414307,
+      "grad_norm": 2.0920238494873047,
+      "learning_rate": 0.0002,
+      "loss": 2.3784,
+      "step": 47470
+    },
+    {
+      "epoch": 3.538002980625931,
+      "grad_norm": 2.5640475749969482,
+      "learning_rate": 0.0002,
+      "loss": 2.5922,
+      "step": 47480
+    },
+    {
+      "epoch": 3.5387481371087928,
+      "grad_norm": 2.3668644428253174,
+      "learning_rate": 0.0002,
+      "loss": 2.4113,
+      "step": 47490
+    },
+    {
+      "epoch": 3.5394932935916543,
+      "grad_norm": 2.7140488624572754,
+      "learning_rate": 0.0002,
+      "loss": 2.6541,
+      "step": 47500
+    },
+    {
+      "epoch": 3.540238450074516,
+      "grad_norm": 2.655503511428833,
+      "learning_rate": 0.0002,
+      "loss": 2.6018,
+      "step": 47510
+    },
+    {
+      "epoch": 3.540983606557377,
+      "grad_norm": 2.689578056335449,
+      "learning_rate": 0.0002,
+      "loss": 2.4236,
+      "step": 47520
+    },
+    {
+      "epoch": 3.5417287630402385,
+      "grad_norm": 2.8998870849609375,
+      "learning_rate": 0.0002,
+      "loss": 2.4237,
+      "step": 47530
+    },
+    {
+      "epoch": 3.5424739195230996,
+      "grad_norm": 2.749411106109619,
+      "learning_rate": 0.0002,
+      "loss": 2.3785,
+      "step": 47540
+    },
+    {
+      "epoch": 3.543219076005961,
+      "grad_norm": 2.72086238861084,
+      "learning_rate": 0.0002,
+      "loss": 2.3861,
+      "step": 47550
+    },
+    {
+      "epoch": 3.5439642324888228,
+      "grad_norm": 2.7207980155944824,
+      "learning_rate": 0.0002,
+      "loss": 2.6309,
+      "step": 47560
+    },
+    {
+      "epoch": 3.5447093889716843,
+      "grad_norm": 2.5211992263793945,
+      "learning_rate": 0.0002,
+      "loss": 2.6079,
+      "step": 47570
+    },
+    {
+      "epoch": 3.5454545454545454,
+      "grad_norm": 2.592176914215088,
+      "learning_rate": 0.0002,
+      "loss": 2.6657,
+      "step": 47580
+    },
+    {
+      "epoch": 3.546199701937407,
+      "grad_norm": 2.651425838470459,
+      "learning_rate": 0.0002,
+      "loss": 2.432,
+      "step": 47590
+    },
+    {
+      "epoch": 3.546944858420268,
+      "grad_norm": 2.9301748275756836,
+      "learning_rate": 0.0002,
+      "loss": 2.6111,
+      "step": 47600
+    },
+    {
+      "epoch": 3.5476900149031296,
+      "grad_norm": 2.4946770668029785,
+      "learning_rate": 0.0002,
+      "loss": 2.5022,
+      "step": 47610
+    },
+    {
+      "epoch": 3.548435171385991,
+      "grad_norm": 2.465942144393921,
+      "learning_rate": 0.0002,
+      "loss": 2.4284,
+      "step": 47620
+    },
+    {
+      "epoch": 3.5491803278688527,
+      "grad_norm": 2.899252414703369,
+      "learning_rate": 0.0002,
+      "loss": 2.203,
+      "step": 47630
+    },
+    {
+      "epoch": 3.549925484351714,
+      "grad_norm": 2.9829139709472656,
+      "learning_rate": 0.0002,
+      "loss": 2.4265,
+      "step": 47640
+    },
+    {
+      "epoch": 3.5506706408345754,
+      "grad_norm": 3.0602529048919678,
+      "learning_rate": 0.0002,
+      "loss": 2.4456,
+      "step": 47650
+    },
+    {
+      "epoch": 3.5514157973174365,
+      "grad_norm": 2.6664137840270996,
+      "learning_rate": 0.0002,
+      "loss": 2.4857,
+      "step": 47660
+    },
+    {
+      "epoch": 3.552160953800298,
+      "grad_norm": 2.6339783668518066,
+      "learning_rate": 0.0002,
+      "loss": 2.4292,
+      "step": 47670
+    },
+    {
+      "epoch": 3.5529061102831596,
+      "grad_norm": 2.6717617511749268,
+      "learning_rate": 0.0002,
+      "loss": 2.7226,
+      "step": 47680
+    },
+    {
+      "epoch": 3.5536512667660207,
+      "grad_norm": 2.6578307151794434,
+      "learning_rate": 0.0002,
+      "loss": 2.5738,
+      "step": 47690
+    },
+    {
+      "epoch": 3.5543964232488823,
+      "grad_norm": 3.1757662296295166,
+      "learning_rate": 0.0002,
+      "loss": 2.3902,
+      "step": 47700
+    },
+    {
+      "epoch": 3.555141579731744,
+      "grad_norm": 2.820986032485962,
+      "learning_rate": 0.0002,
+      "loss": 2.5761,
+      "step": 47710
+    },
+    {
+      "epoch": 3.555886736214605,
+      "grad_norm": 2.5975351333618164,
+      "learning_rate": 0.0002,
+      "loss": 2.4182,
+      "step": 47720
+    },
+    {
+      "epoch": 3.5566318926974665,
+      "grad_norm": 2.1608307361602783,
+      "learning_rate": 0.0002,
+      "loss": 2.4476,
+      "step": 47730
+    },
+    {
+      "epoch": 3.557377049180328,
+      "grad_norm": 2.800523519515991,
+      "learning_rate": 0.0002,
+      "loss": 2.5175,
+      "step": 47740
+    },
+    {
+      "epoch": 3.558122205663189,
+      "grad_norm": 2.303067207336426,
+      "learning_rate": 0.0002,
+      "loss": 2.3914,
+      "step": 47750
+    },
+    {
+      "epoch": 3.5588673621460507,
+      "grad_norm": 2.3079776763916016,
+      "learning_rate": 0.0002,
+      "loss": 2.4874,
+      "step": 47760
+    },
+    {
+      "epoch": 3.559612518628912,
+      "grad_norm": 3.0130832195281982,
+      "learning_rate": 0.0002,
+      "loss": 2.6741,
+      "step": 47770
+    },
+    {
+      "epoch": 3.5603576751117734,
+      "grad_norm": 2.6367170810699463,
+      "learning_rate": 0.0002,
+      "loss": 2.3853,
+      "step": 47780
+    },
+    {
+      "epoch": 3.561102831594635,
+      "grad_norm": 2.8482704162597656,
+      "learning_rate": 0.0002,
+      "loss": 2.4991,
+      "step": 47790
+    },
+    {
+      "epoch": 3.5618479880774965,
+      "grad_norm": 2.4554948806762695,
+      "learning_rate": 0.0002,
+      "loss": 2.4502,
+      "step": 47800
+    },
+    {
+      "epoch": 3.5625931445603576,
+      "grad_norm": 2.6724441051483154,
+      "learning_rate": 0.0002,
+      "loss": 2.6487,
+      "step": 47810
+    },
+    {
+      "epoch": 3.563338301043219,
+      "grad_norm": 2.8899343013763428,
+      "learning_rate": 0.0002,
+      "loss": 2.6546,
+      "step": 47820
+    },
+    {
+      "epoch": 3.5640834575260802,
+      "grad_norm": 3.299750804901123,
+      "learning_rate": 0.0002,
+      "loss": 2.6725,
+      "step": 47830
+    },
+    {
+      "epoch": 3.564828614008942,
+      "grad_norm": 2.936067819595337,
+      "learning_rate": 0.0002,
+      "loss": 2.5936,
+      "step": 47840
+    },
+    {
+      "epoch": 3.5655737704918034,
+      "grad_norm": 2.8651251792907715,
+      "learning_rate": 0.0002,
+      "loss": 2.5748,
+      "step": 47850
+    },
+    {
+      "epoch": 3.566318926974665,
+      "grad_norm": 2.662198066711426,
+      "learning_rate": 0.0002,
+      "loss": 2.5171,
+      "step": 47860
+    },
+    {
+      "epoch": 3.567064083457526,
+      "grad_norm": 2.711688995361328,
+      "learning_rate": 0.0002,
+      "loss": 2.542,
+      "step": 47870
+    },
+    {
+      "epoch": 3.5678092399403876,
+      "grad_norm": 2.569240093231201,
+      "learning_rate": 0.0002,
+      "loss": 2.5765,
+      "step": 47880
+    },
+    {
+      "epoch": 3.5685543964232487,
+      "grad_norm": 3.035104990005493,
+      "learning_rate": 0.0002,
+      "loss": 2.6961,
+      "step": 47890
+    },
+    {
+      "epoch": 3.5692995529061102,
+      "grad_norm": 2.7538046836853027,
+      "learning_rate": 0.0002,
+      "loss": 2.4598,
+      "step": 47900
+    },
+    {
+      "epoch": 3.570044709388972,
+      "grad_norm": 1.9985271692276,
+      "learning_rate": 0.0002,
+      "loss": 2.3044,
+      "step": 47910
+    },
+    {
+      "epoch": 3.5707898658718333,
+      "grad_norm": 2.6245272159576416,
+      "learning_rate": 0.0002,
+      "loss": 2.6054,
+      "step": 47920
+    },
+    {
+      "epoch": 3.5715350223546944,
+      "grad_norm": 2.6232852935791016,
+      "learning_rate": 0.0002,
+      "loss": 2.5152,
+      "step": 47930
+    },
+    {
+      "epoch": 3.572280178837556,
+      "grad_norm": 2.8082447052001953,
+      "learning_rate": 0.0002,
+      "loss": 2.4981,
+      "step": 47940
+    },
+    {
+      "epoch": 3.573025335320417,
+      "grad_norm": 2.4905755519866943,
+      "learning_rate": 0.0002,
+      "loss": 2.6797,
+      "step": 47950
+    },
+    {
+      "epoch": 3.5737704918032787,
+      "grad_norm": 2.3986294269561768,
+      "learning_rate": 0.0002,
+      "loss": 2.3939,
+      "step": 47960
+    },
+    {
+      "epoch": 3.57451564828614,
+      "grad_norm": 2.645498037338257,
+      "learning_rate": 0.0002,
+      "loss": 2.5753,
+      "step": 47970
+    },
+    {
+      "epoch": 3.5752608047690018,
+      "grad_norm": 2.494290828704834,
+      "learning_rate": 0.0002,
+      "loss": 2.3679,
+      "step": 47980
+    },
+    {
+      "epoch": 3.576005961251863,
+      "grad_norm": 2.542083740234375,
+      "learning_rate": 0.0002,
+      "loss": 2.53,
+      "step": 47990
+    },
+    {
+      "epoch": 3.5767511177347244,
+      "grad_norm": 2.529242753982544,
+      "learning_rate": 0.0002,
+      "loss": 2.5242,
+      "step": 48000
+    },
+    {
+      "epoch": 3.5774962742175855,
+      "grad_norm": 2.7906339168548584,
+      "learning_rate": 0.0002,
+      "loss": 2.4736,
+      "step": 48010
+    },
+    {
+      "epoch": 3.578241430700447,
+      "grad_norm": 2.607820510864258,
+      "learning_rate": 0.0002,
+      "loss": 2.5297,
+      "step": 48020
+    },
+    {
+      "epoch": 3.5789865871833086,
+      "grad_norm": 2.3591980934143066,
+      "learning_rate": 0.0002,
+      "loss": 2.6442,
+      "step": 48030
+    },
+    {
+      "epoch": 3.5797317436661698,
+      "grad_norm": 2.4848647117614746,
+      "learning_rate": 0.0002,
+      "loss": 2.4396,
+      "step": 48040
+    },
+    {
+      "epoch": 3.5804769001490313,
+      "grad_norm": 2.30979323387146,
+      "learning_rate": 0.0002,
+      "loss": 2.4167,
+      "step": 48050
+    },
+    {
+      "epoch": 3.581222056631893,
+      "grad_norm": 2.738574981689453,
+      "learning_rate": 0.0002,
+      "loss": 2.5074,
+      "step": 48060
+    },
+    {
+      "epoch": 3.581967213114754,
+      "grad_norm": 2.5661182403564453,
+      "learning_rate": 0.0002,
+      "loss": 2.3834,
+      "step": 48070
+    },
+    {
+      "epoch": 3.5827123695976155,
+      "grad_norm": 2.47078013420105,
+      "learning_rate": 0.0002,
+      "loss": 2.5172,
+      "step": 48080
+    },
+    {
+      "epoch": 3.583457526080477,
+      "grad_norm": 2.635650157928467,
+      "learning_rate": 0.0002,
+      "loss": 2.5179,
+      "step": 48090
+    },
+    {
+      "epoch": 3.584202682563338,
+      "grad_norm": 2.555915117263794,
+      "learning_rate": 0.0002,
+      "loss": 2.3665,
+      "step": 48100
+    },
+    {
+      "epoch": 3.5849478390461997,
+      "grad_norm": 2.8681654930114746,
+      "learning_rate": 0.0002,
+      "loss": 2.3979,
+      "step": 48110
+    },
+    {
+      "epoch": 3.585692995529061,
+      "grad_norm": 2.8930814266204834,
+      "learning_rate": 0.0002,
+      "loss": 2.5871,
+      "step": 48120
+    },
+    {
+      "epoch": 3.5864381520119224,
+      "grad_norm": 2.3275046348571777,
+      "learning_rate": 0.0002,
+      "loss": 2.4286,
+      "step": 48130
+    },
+    {
+      "epoch": 3.587183308494784,
+      "grad_norm": 2.7267794609069824,
+      "learning_rate": 0.0002,
+      "loss": 2.6153,
+      "step": 48140
+    },
+    {
+      "epoch": 3.5879284649776455,
+      "grad_norm": 2.8347225189208984,
+      "learning_rate": 0.0002,
+      "loss": 2.4797,
+      "step": 48150
+    },
+    {
+      "epoch": 3.5886736214605066,
+      "grad_norm": 2.776869058609009,
+      "learning_rate": 0.0002,
+      "loss": 2.4546,
+      "step": 48160
+    },
+    {
+      "epoch": 3.589418777943368,
+      "grad_norm": 2.580460548400879,
+      "learning_rate": 0.0002,
+      "loss": 2.4808,
+      "step": 48170
+    },
+    {
+      "epoch": 3.5901639344262293,
+      "grad_norm": 2.462087869644165,
+      "learning_rate": 0.0002,
+      "loss": 2.4598,
+      "step": 48180
+    },
+    {
+      "epoch": 3.590909090909091,
+      "grad_norm": 2.611598491668701,
+      "learning_rate": 0.0002,
+      "loss": 2.3196,
+      "step": 48190
+    },
+    {
+      "epoch": 3.5916542473919524,
+      "grad_norm": 2.7346081733703613,
+      "learning_rate": 0.0002,
+      "loss": 2.6164,
+      "step": 48200
+    },
+    {
+      "epoch": 3.592399403874814,
+      "grad_norm": 2.7156240940093994,
+      "learning_rate": 0.0002,
+      "loss": 2.5327,
+      "step": 48210
+    },
+    {
+      "epoch": 3.593144560357675,
+      "grad_norm": 2.703343629837036,
+      "learning_rate": 0.0002,
+      "loss": 2.4702,
+      "step": 48220
+    },
+    {
+      "epoch": 3.5938897168405366,
+      "grad_norm": 3.1075315475463867,
+      "learning_rate": 0.0002,
+      "loss": 2.5256,
+      "step": 48230
+    },
+    {
+      "epoch": 3.5946348733233977,
+      "grad_norm": 3.027200937271118,
+      "learning_rate": 0.0002,
+      "loss": 2.4926,
+      "step": 48240
+    },
+    {
+      "epoch": 3.5953800298062593,
+      "grad_norm": 2.59000825881958,
+      "learning_rate": 0.0002,
+      "loss": 2.0317,
+      "step": 48250
+    },
+    {
+      "epoch": 3.596125186289121,
+      "grad_norm": 2.602517604827881,
+      "learning_rate": 0.0002,
+      "loss": 2.3541,
+      "step": 48260
+    },
+    {
+      "epoch": 3.5968703427719824,
+      "grad_norm": 2.692680835723877,
+      "learning_rate": 0.0002,
+      "loss": 2.4874,
+      "step": 48270
+    },
+    {
+      "epoch": 3.5976154992548435,
+      "grad_norm": 1.833561658859253,
+      "learning_rate": 0.0002,
+      "loss": 2.3711,
+      "step": 48280
+    },
+    {
+      "epoch": 3.598360655737705,
+      "grad_norm": 3.1557459831237793,
+      "learning_rate": 0.0002,
+      "loss": 2.5707,
+      "step": 48290
+    },
+    {
+      "epoch": 3.599105812220566,
+      "grad_norm": 2.482980728149414,
+      "learning_rate": 0.0002,
+      "loss": 2.4062,
+      "step": 48300
+    },
+    {
+      "epoch": 3.5998509687034277,
+      "grad_norm": 2.730379104614258,
+      "learning_rate": 0.0002,
+      "loss": 2.6087,
+      "step": 48310
+    },
+    {
+      "epoch": 3.6005961251862892,
+      "grad_norm": 2.50185227394104,
+      "learning_rate": 0.0002,
+      "loss": 2.4351,
+      "step": 48320
+    },
+    {
+      "epoch": 3.601341281669151,
+      "grad_norm": 2.4742274284362793,
+      "learning_rate": 0.0002,
+      "loss": 2.3752,
+      "step": 48330
+    },
+    {
+      "epoch": 3.602086438152012,
+      "grad_norm": 2.677802801132202,
+      "learning_rate": 0.0002,
+      "loss": 2.4924,
+      "step": 48340
+    },
+    {
+      "epoch": 3.6028315946348735,
+      "grad_norm": 2.7922158241271973,
+      "learning_rate": 0.0002,
+      "loss": 2.5855,
+      "step": 48350
+    },
+    {
+      "epoch": 3.6035767511177346,
+      "grad_norm": 3.383920431137085,
+      "learning_rate": 0.0002,
+      "loss": 2.5071,
+      "step": 48360
+    },
+    {
+      "epoch": 3.604321907600596,
+      "grad_norm": 2.9359383583068848,
+      "learning_rate": 0.0002,
+      "loss": 2.361,
+      "step": 48370
+    },
+    {
+      "epoch": 3.6050670640834577,
+      "grad_norm": 2.7278084754943848,
+      "learning_rate": 0.0002,
+      "loss": 2.641,
+      "step": 48380
+    },
+    {
+      "epoch": 3.605812220566319,
+      "grad_norm": 2.810471296310425,
+      "learning_rate": 0.0002,
+      "loss": 2.6345,
+      "step": 48390
+    },
+    {
+      "epoch": 3.6065573770491803,
+      "grad_norm": 2.4108266830444336,
+      "learning_rate": 0.0002,
+      "loss": 2.5544,
+      "step": 48400
+    },
+    {
+      "epoch": 3.607302533532042,
+      "grad_norm": 2.8318967819213867,
+      "learning_rate": 0.0002,
+      "loss": 2.52,
+      "step": 48410
+    },
+    {
+      "epoch": 3.608047690014903,
+      "grad_norm": 2.8543920516967773,
+      "learning_rate": 0.0002,
+      "loss": 2.5537,
+      "step": 48420
+    },
+    {
+      "epoch": 3.6087928464977646,
+      "grad_norm": 2.791773796081543,
+      "learning_rate": 0.0002,
+      "loss": 2.4955,
+      "step": 48430
+    },
+    {
+      "epoch": 3.609538002980626,
+      "grad_norm": 2.8440630435943604,
+      "learning_rate": 0.0002,
+      "loss": 2.4265,
+      "step": 48440
+    },
+    {
+      "epoch": 3.610283159463487,
+      "grad_norm": 2.4981586933135986,
+      "learning_rate": 0.0002,
+      "loss": 2.4647,
+      "step": 48450
+    },
+    {
+      "epoch": 3.6110283159463488,
+      "grad_norm": 2.399179458618164,
+      "learning_rate": 0.0002,
+      "loss": 2.5682,
+      "step": 48460
+    },
+    {
+      "epoch": 3.61177347242921,
+      "grad_norm": 2.578258991241455,
+      "learning_rate": 0.0002,
+      "loss": 2.5542,
+      "step": 48470
+    },
+    {
+      "epoch": 3.6125186289120714,
+      "grad_norm": 2.8405141830444336,
+      "learning_rate": 0.0002,
+      "loss": 2.5277,
+      "step": 48480
+    },
+    {
+      "epoch": 3.613263785394933,
+      "grad_norm": 2.530799150466919,
+      "learning_rate": 0.0002,
+      "loss": 2.644,
+      "step": 48490
+    },
+    {
+      "epoch": 3.6140089418777945,
+      "grad_norm": 2.7181057929992676,
+      "learning_rate": 0.0002,
+      "loss": 2.4457,
+      "step": 48500
+    },
+    {
+      "epoch": 3.6147540983606556,
+      "grad_norm": 2.577636480331421,
+      "learning_rate": 0.0002,
+      "loss": 2.4226,
+      "step": 48510
+    },
+    {
+      "epoch": 3.615499254843517,
+      "grad_norm": 3.4074344635009766,
+      "learning_rate": 0.0002,
+      "loss": 2.5066,
+      "step": 48520
+    },
+    {
+      "epoch": 3.6162444113263783,
+      "grad_norm": 2.69319748878479,
+      "learning_rate": 0.0002,
+      "loss": 2.5189,
+      "step": 48530
+    },
+    {
+      "epoch": 3.61698956780924,
+      "grad_norm": 2.8684957027435303,
+      "learning_rate": 0.0002,
+      "loss": 2.5323,
+      "step": 48540
+    },
+    {
+      "epoch": 3.6177347242921014,
+      "grad_norm": 2.828727960586548,
+      "learning_rate": 0.0002,
+      "loss": 2.5495,
+      "step": 48550
+    },
+    {
+      "epoch": 3.618479880774963,
+      "grad_norm": 2.4730985164642334,
+      "learning_rate": 0.0002,
+      "loss": 2.5354,
+      "step": 48560
+    },
+    {
+      "epoch": 3.619225037257824,
+      "grad_norm": 2.840999126434326,
+      "learning_rate": 0.0002,
+      "loss": 2.5194,
+      "step": 48570
+    },
+    {
+      "epoch": 3.6199701937406856,
+      "grad_norm": 2.5371737480163574,
+      "learning_rate": 0.0002,
+      "loss": 2.4109,
+      "step": 48580
+    },
+    {
+      "epoch": 3.6207153502235467,
+      "grad_norm": 3.4489569664001465,
+      "learning_rate": 0.0002,
+      "loss": 2.6168,
+      "step": 48590
+    },
+    {
+      "epoch": 3.6214605067064083,
+      "grad_norm": 2.614786386489868,
+      "learning_rate": 0.0002,
+      "loss": 2.4671,
+      "step": 48600
+    },
+    {
+      "epoch": 3.62220566318927,
+      "grad_norm": 2.5180749893188477,
+      "learning_rate": 0.0002,
+      "loss": 2.4171,
+      "step": 48610
+    },
+    {
+      "epoch": 3.6229508196721314,
+      "grad_norm": 2.856128692626953,
+      "learning_rate": 0.0002,
+      "loss": 2.3804,
+      "step": 48620
+    },
+    {
+      "epoch": 3.6236959761549925,
+      "grad_norm": 2.196049213409424,
+      "learning_rate": 0.0002,
+      "loss": 2.4736,
+      "step": 48630
+    },
+    {
+      "epoch": 3.624441132637854,
+      "grad_norm": 2.728996515274048,
+      "learning_rate": 0.0002,
+      "loss": 2.4175,
+      "step": 48640
+    },
+    {
+      "epoch": 3.625186289120715,
+      "grad_norm": 2.3340585231781006,
+      "learning_rate": 0.0002,
+      "loss": 2.4509,
+      "step": 48650
+    },
+    {
+      "epoch": 3.6259314456035767,
+      "grad_norm": 2.537264823913574,
+      "learning_rate": 0.0002,
+      "loss": 2.3519,
+      "step": 48660
+    },
+    {
+      "epoch": 3.6266766020864383,
+      "grad_norm": 2.4871232509613037,
+      "learning_rate": 0.0002,
+      "loss": 2.3988,
+      "step": 48670
+    },
+    {
+      "epoch": 3.6274217585693,
+      "grad_norm": 2.5547709465026855,
+      "learning_rate": 0.0002,
+      "loss": 2.2731,
+      "step": 48680
+    },
+    {
+      "epoch": 3.628166915052161,
+      "grad_norm": 2.8591160774230957,
+      "learning_rate": 0.0002,
+      "loss": 2.6361,
+      "step": 48690
+    },
+    {
+      "epoch": 3.6289120715350225,
+      "grad_norm": 2.216677665710449,
+      "learning_rate": 0.0002,
+      "loss": 2.3692,
+      "step": 48700
+    },
+    {
+      "epoch": 3.6296572280178836,
+      "grad_norm": 2.721316337585449,
+      "learning_rate": 0.0002,
+      "loss": 2.4691,
+      "step": 48710
+    },
+    {
+      "epoch": 3.630402384500745,
+      "grad_norm": 2.8119468688964844,
+      "learning_rate": 0.0002,
+      "loss": 2.397,
+      "step": 48720
+    },
+    {
+      "epoch": 3.6311475409836067,
+      "grad_norm": 3.7255992889404297,
+      "learning_rate": 0.0002,
+      "loss": 2.461,
+      "step": 48730
+    },
+    {
+      "epoch": 3.631892697466468,
+      "grad_norm": 2.680361032485962,
+      "learning_rate": 0.0002,
+      "loss": 2.2952,
+      "step": 48740
+    },
+    {
+      "epoch": 3.6326378539493294,
+      "grad_norm": 2.1289005279541016,
+      "learning_rate": 0.0002,
+      "loss": 2.4144,
+      "step": 48750
+    },
+    {
+      "epoch": 3.633383010432191,
+      "grad_norm": 2.259798765182495,
+      "learning_rate": 0.0002,
+      "loss": 2.6487,
+      "step": 48760
+    },
+    {
+      "epoch": 3.634128166915052,
+      "grad_norm": 2.293454170227051,
+      "learning_rate": 0.0002,
+      "loss": 2.5502,
+      "step": 48770
+    },
+    {
+      "epoch": 3.6348733233979136,
+      "grad_norm": 2.636841297149658,
+      "learning_rate": 0.0002,
+      "loss": 2.5529,
+      "step": 48780
+    },
+    {
+      "epoch": 3.635618479880775,
+      "grad_norm": 2.310394287109375,
+      "learning_rate": 0.0002,
+      "loss": 2.4676,
+      "step": 48790
+    },
+    {
+      "epoch": 3.6363636363636362,
+      "grad_norm": 2.545966625213623,
+      "learning_rate": 0.0002,
+      "loss": 2.4672,
+      "step": 48800
+    },
+    {
+      "epoch": 3.637108792846498,
+      "grad_norm": 2.9500443935394287,
+      "learning_rate": 0.0002,
+      "loss": 2.4827,
+      "step": 48810
+    },
+    {
+      "epoch": 3.637853949329359,
+      "grad_norm": 2.543187141418457,
+      "learning_rate": 0.0002,
+      "loss": 2.4132,
+      "step": 48820
+    },
+    {
+      "epoch": 3.6385991058122205,
+      "grad_norm": 2.2847654819488525,
+      "learning_rate": 0.0002,
+      "loss": 2.4768,
+      "step": 48830
+    },
+    {
+      "epoch": 3.639344262295082,
+      "grad_norm": 2.72090482711792,
+      "learning_rate": 0.0002,
+      "loss": 2.5947,
+      "step": 48840
+    },
+    {
+      "epoch": 3.6400894187779436,
+      "grad_norm": 2.562582015991211,
+      "learning_rate": 0.0002,
+      "loss": 2.5658,
+      "step": 48850
+    },
+    {
+      "epoch": 3.6408345752608047,
+      "grad_norm": 2.780815839767456,
+      "learning_rate": 0.0002,
+      "loss": 2.5017,
+      "step": 48860
+    },
+    {
+      "epoch": 3.6415797317436662,
+      "grad_norm": 2.9853780269622803,
+      "learning_rate": 0.0002,
+      "loss": 2.6693,
+      "step": 48870
+    },
+    {
+      "epoch": 3.6423248882265273,
+      "grad_norm": 2.79302716255188,
+      "learning_rate": 0.0002,
+      "loss": 2.4589,
+      "step": 48880
+    },
+    {
+      "epoch": 3.643070044709389,
+      "grad_norm": 2.7018563747406006,
+      "learning_rate": 0.0002,
+      "loss": 2.3826,
+      "step": 48890
+    },
+    {
+      "epoch": 3.6438152011922504,
+      "grad_norm": 2.699528694152832,
+      "learning_rate": 0.0002,
+      "loss": 2.3877,
+      "step": 48900
+    },
+    {
+      "epoch": 3.644560357675112,
+      "grad_norm": 2.4166371822357178,
+      "learning_rate": 0.0002,
+      "loss": 2.5869,
+      "step": 48910
+    },
+    {
+      "epoch": 3.645305514157973,
+      "grad_norm": 2.5733654499053955,
+      "learning_rate": 0.0002,
+      "loss": 2.5201,
+      "step": 48920
+    },
+    {
+      "epoch": 3.6460506706408347,
+      "grad_norm": 2.7172818183898926,
+      "learning_rate": 0.0002,
+      "loss": 2.4568,
+      "step": 48930
+    },
+    {
+      "epoch": 3.6467958271236958,
+      "grad_norm": 2.5636215209960938,
+      "learning_rate": 0.0002,
+      "loss": 2.5941,
+      "step": 48940
+    },
+    {
+      "epoch": 3.6475409836065573,
+      "grad_norm": 2.592777729034424,
+      "learning_rate": 0.0002,
+      "loss": 2.3906,
+      "step": 48950
+    },
+    {
+      "epoch": 3.648286140089419,
+      "grad_norm": 2.5706088542938232,
+      "learning_rate": 0.0002,
+      "loss": 2.4504,
+      "step": 48960
+    },
+    {
+      "epoch": 3.6490312965722804,
+      "grad_norm": 2.48825740814209,
+      "learning_rate": 0.0002,
+      "loss": 2.4888,
+      "step": 48970
+    },
+    {
+      "epoch": 3.6497764530551415,
+      "grad_norm": 2.727311372756958,
+      "learning_rate": 0.0002,
+      "loss": 2.4781,
+      "step": 48980
+    },
+    {
+      "epoch": 3.650521609538003,
+      "grad_norm": 2.3753762245178223,
+      "learning_rate": 0.0002,
+      "loss": 2.504,
+      "step": 48990
+    },
+    {
+      "epoch": 3.651266766020864,
+      "grad_norm": 2.639061450958252,
+      "learning_rate": 0.0002,
+      "loss": 2.5593,
+      "step": 49000
+    },
+    {
+      "epoch": 3.6520119225037257,
+      "grad_norm": 2.736478090286255,
+      "learning_rate": 0.0002,
+      "loss": 2.5617,
+      "step": 49010
+    },
+    {
+      "epoch": 3.6527570789865873,
+      "grad_norm": 2.6167259216308594,
+      "learning_rate": 0.0002,
+      "loss": 2.7529,
+      "step": 49020
+    },
+    {
+      "epoch": 3.653502235469449,
+      "grad_norm": 2.2042272090911865,
+      "learning_rate": 0.0002,
+      "loss": 2.4023,
+      "step": 49030
+    },
+    {
+      "epoch": 3.65424739195231,
+      "grad_norm": 2.191154956817627,
+      "learning_rate": 0.0002,
+      "loss": 2.5161,
+      "step": 49040
+    },
+    {
+      "epoch": 3.6549925484351715,
+      "grad_norm": 2.672147035598755,
+      "learning_rate": 0.0002,
+      "loss": 2.4459,
+      "step": 49050
+    },
+    {
+      "epoch": 3.6557377049180326,
+      "grad_norm": 2.6667487621307373,
+      "learning_rate": 0.0002,
+      "loss": 2.5033,
+      "step": 49060
+    },
+    {
+      "epoch": 3.656482861400894,
+      "grad_norm": 2.488171100616455,
+      "learning_rate": 0.0002,
+      "loss": 2.3307,
+      "step": 49070
+    },
+    {
+      "epoch": 3.6572280178837557,
+      "grad_norm": 2.9280152320861816,
+      "learning_rate": 0.0002,
+      "loss": 2.5912,
+      "step": 49080
+    },
+    {
+      "epoch": 3.657973174366617,
+      "grad_norm": 2.767899513244629,
+      "learning_rate": 0.0002,
+      "loss": 2.5289,
+      "step": 49090
+    },
+    {
+      "epoch": 3.6587183308494784,
+      "grad_norm": 2.595731019973755,
+      "learning_rate": 0.0002,
+      "loss": 2.6172,
+      "step": 49100
+    },
+    {
+      "epoch": 3.65946348733234,
+      "grad_norm": 2.531205177307129,
+      "learning_rate": 0.0002,
+      "loss": 2.5384,
+      "step": 49110
+    },
+    {
+      "epoch": 3.660208643815201,
+      "grad_norm": 2.5499446392059326,
+      "learning_rate": 0.0002,
+      "loss": 2.4912,
+      "step": 49120
+    },
+    {
+      "epoch": 3.6609538002980626,
+      "grad_norm": 2.692110300064087,
+      "learning_rate": 0.0002,
+      "loss": 2.6736,
+      "step": 49130
+    },
+    {
+      "epoch": 3.661698956780924,
+      "grad_norm": 2.5231432914733887,
+      "learning_rate": 0.0002,
+      "loss": 2.4245,
+      "step": 49140
+    },
+    {
+      "epoch": 3.6624441132637853,
+      "grad_norm": 2.7942330837249756,
+      "learning_rate": 0.0002,
+      "loss": 2.5642,
+      "step": 49150
+    },
+    {
+      "epoch": 3.663189269746647,
+      "grad_norm": 3.3361740112304688,
+      "learning_rate": 0.0002,
+      "loss": 2.5024,
+      "step": 49160
+    },
+    {
+      "epoch": 3.663934426229508,
+      "grad_norm": 3.006382465362549,
+      "learning_rate": 0.0002,
+      "loss": 2.5587,
+      "step": 49170
+    },
+    {
+      "epoch": 3.6646795827123695,
+      "grad_norm": 2.637521743774414,
+      "learning_rate": 0.0002,
+      "loss": 2.5168,
+      "step": 49180
+    },
+    {
+      "epoch": 3.665424739195231,
+      "grad_norm": 2.8171842098236084,
+      "learning_rate": 0.0002,
+      "loss": 2.6647,
+      "step": 49190
+    },
+    {
+      "epoch": 3.6661698956780926,
+      "grad_norm": 2.3812084197998047,
+      "learning_rate": 0.0002,
+      "loss": 2.2236,
+      "step": 49200
+    },
+    {
+      "epoch": 3.6669150521609537,
+      "grad_norm": 2.4592621326446533,
+      "learning_rate": 0.0002,
+      "loss": 2.3351,
+      "step": 49210
+    },
+    {
+      "epoch": 3.6676602086438153,
+      "grad_norm": 2.5426881313323975,
+      "learning_rate": 0.0002,
+      "loss": 2.5694,
+      "step": 49220
+    },
+    {
+      "epoch": 3.6684053651266764,
+      "grad_norm": 2.749530553817749,
+      "learning_rate": 0.0002,
+      "loss": 2.4955,
+      "step": 49230
+    },
+    {
+      "epoch": 3.669150521609538,
+      "grad_norm": 3.183913230895996,
+      "learning_rate": 0.0002,
+      "loss": 2.7195,
+      "step": 49240
+    },
+    {
+      "epoch": 3.6698956780923995,
+      "grad_norm": 2.6278300285339355,
+      "learning_rate": 0.0002,
+      "loss": 2.5673,
+      "step": 49250
+    },
+    {
+      "epoch": 3.670640834575261,
+      "grad_norm": 2.996420383453369,
+      "learning_rate": 0.0002,
+      "loss": 2.5843,
+      "step": 49260
+    },
+    {
+      "epoch": 3.671385991058122,
+      "grad_norm": 2.6500403881073,
+      "learning_rate": 0.0002,
+      "loss": 2.6122,
+      "step": 49270
+    },
+    {
+      "epoch": 3.6721311475409837,
+      "grad_norm": 3.047901153564453,
+      "learning_rate": 0.0002,
+      "loss": 2.4905,
+      "step": 49280
+    },
+    {
+      "epoch": 3.672876304023845,
+      "grad_norm": 2.8154449462890625,
+      "learning_rate": 0.0002,
+      "loss": 2.4668,
+      "step": 49290
+    },
+    {
+      "epoch": 3.6736214605067063,
+      "grad_norm": 2.3584389686584473,
+      "learning_rate": 0.0002,
+      "loss": 2.3198,
+      "step": 49300
+    },
+    {
+      "epoch": 3.674366616989568,
+      "grad_norm": 2.881648063659668,
+      "learning_rate": 0.0002,
+      "loss": 2.4651,
+      "step": 49310
+    },
+    {
+      "epoch": 3.6751117734724295,
+      "grad_norm": 2.5896761417388916,
+      "learning_rate": 0.0002,
+      "loss": 2.241,
+      "step": 49320
+    },
+    {
+      "epoch": 3.6758569299552906,
+      "grad_norm": 2.812044858932495,
+      "learning_rate": 0.0002,
+      "loss": 2.5172,
+      "step": 49330
+    },
+    {
+      "epoch": 3.676602086438152,
+      "grad_norm": 2.66312575340271,
+      "learning_rate": 0.0002,
+      "loss": 2.4133,
+      "step": 49340
+    },
+    {
+      "epoch": 3.6773472429210132,
+      "grad_norm": 2.790090799331665,
+      "learning_rate": 0.0002,
+      "loss": 2.709,
+      "step": 49350
+    },
+    {
+      "epoch": 3.678092399403875,
+      "grad_norm": 2.4893035888671875,
+      "learning_rate": 0.0002,
+      "loss": 2.4402,
+      "step": 49360
+    },
+    {
+      "epoch": 3.6788375558867363,
+      "grad_norm": 2.754606246948242,
+      "learning_rate": 0.0002,
+      "loss": 2.4593,
+      "step": 49370
+    },
+    {
+      "epoch": 3.679582712369598,
+      "grad_norm": 2.878617763519287,
+      "learning_rate": 0.0002,
+      "loss": 2.6367,
+      "step": 49380
+    },
+    {
+      "epoch": 3.680327868852459,
+      "grad_norm": 2.854499101638794,
+      "learning_rate": 0.0002,
+      "loss": 2.4631,
+      "step": 49390
+    },
+    {
+      "epoch": 3.6810730253353205,
+      "grad_norm": 2.849299669265747,
+      "learning_rate": 0.0002,
+      "loss": 2.465,
+      "step": 49400
+    },
+    {
+      "epoch": 3.6818181818181817,
+      "grad_norm": 2.570016860961914,
+      "learning_rate": 0.0002,
+      "loss": 2.4901,
+      "step": 49410
+    },
+    {
+      "epoch": 3.682563338301043,
+      "grad_norm": 2.6923599243164062,
+      "learning_rate": 0.0002,
+      "loss": 2.4682,
+      "step": 49420
+    },
+    {
+      "epoch": 3.6833084947839048,
+      "grad_norm": 2.5435593128204346,
+      "learning_rate": 0.0002,
+      "loss": 2.5621,
+      "step": 49430
+    },
+    {
+      "epoch": 3.684053651266766,
+      "grad_norm": 2.8608274459838867,
+      "learning_rate": 0.0002,
+      "loss": 2.586,
+      "step": 49440
+    },
+    {
+      "epoch": 3.6847988077496274,
+      "grad_norm": 2.5079166889190674,
+      "learning_rate": 0.0002,
+      "loss": 2.4451,
+      "step": 49450
+    },
+    {
+      "epoch": 3.685543964232489,
+      "grad_norm": 2.874114990234375,
+      "learning_rate": 0.0002,
+      "loss": 2.446,
+      "step": 49460
+    },
+    {
+      "epoch": 3.68628912071535,
+      "grad_norm": 2.727252960205078,
+      "learning_rate": 0.0002,
+      "loss": 2.4395,
+      "step": 49470
+    },
+    {
+      "epoch": 3.6870342771982116,
+      "grad_norm": 2.734837532043457,
+      "learning_rate": 0.0002,
+      "loss": 2.4344,
+      "step": 49480
+    },
+    {
+      "epoch": 3.687779433681073,
+      "grad_norm": 2.5745677947998047,
+      "learning_rate": 0.0002,
+      "loss": 2.3926,
+      "step": 49490
+    },
+    {
+      "epoch": 3.6885245901639343,
+      "grad_norm": 2.9296648502349854,
+      "learning_rate": 0.0002,
+      "loss": 2.458,
+      "step": 49500
+    },
+    {
+      "epoch": 3.689269746646796,
+      "grad_norm": 2.666250705718994,
+      "learning_rate": 0.0002,
+      "loss": 2.5119,
+      "step": 49510
+    },
+    {
+      "epoch": 3.690014903129657,
+      "grad_norm": 2.7720589637756348,
+      "learning_rate": 0.0002,
+      "loss": 2.4189,
+      "step": 49520
+    },
+    {
+      "epoch": 3.6907600596125185,
+      "grad_norm": 2.5077097415924072,
+      "learning_rate": 0.0002,
+      "loss": 2.486,
+      "step": 49530
+    },
+    {
+      "epoch": 3.69150521609538,
+      "grad_norm": 2.7415547370910645,
+      "learning_rate": 0.0002,
+      "loss": 2.4732,
+      "step": 49540
+    },
+    {
+      "epoch": 3.6922503725782416,
+      "grad_norm": 3.023948907852173,
+      "learning_rate": 0.0002,
+      "loss": 2.4565,
+      "step": 49550
+    },
+    {
+      "epoch": 3.6929955290611027,
+      "grad_norm": 2.926791191101074,
+      "learning_rate": 0.0002,
+      "loss": 2.3835,
+      "step": 49560
+    },
+    {
+      "epoch": 3.6937406855439643,
+      "grad_norm": 2.64837646484375,
+      "learning_rate": 0.0002,
+      "loss": 2.5395,
+      "step": 49570
+    },
+    {
+      "epoch": 3.6944858420268254,
+      "grad_norm": 3.1777780055999756,
+      "learning_rate": 0.0002,
+      "loss": 2.6367,
+      "step": 49580
+    },
+    {
+      "epoch": 3.695230998509687,
+      "grad_norm": 2.567836284637451,
+      "learning_rate": 0.0002,
+      "loss": 2.7127,
+      "step": 49590
+    },
+    {
+      "epoch": 3.6959761549925485,
+      "grad_norm": 2.6760616302490234,
+      "learning_rate": 0.0002,
+      "loss": 2.5427,
+      "step": 49600
+    },
+    {
+      "epoch": 3.69672131147541,
+      "grad_norm": 2.6477506160736084,
+      "learning_rate": 0.0002,
+      "loss": 2.446,
+      "step": 49610
+    },
+    {
+      "epoch": 3.697466467958271,
+      "grad_norm": 2.555225372314453,
+      "learning_rate": 0.0002,
+      "loss": 2.4988,
+      "step": 49620
+    },
+    {
+      "epoch": 3.6982116244411327,
+      "grad_norm": 2.3987672328948975,
+      "learning_rate": 0.0002,
+      "loss": 2.476,
+      "step": 49630
+    },
+    {
+      "epoch": 3.698956780923994,
+      "grad_norm": 2.7095890045166016,
+      "learning_rate": 0.0002,
+      "loss": 2.5311,
+      "step": 49640
+    },
+    {
+      "epoch": 3.6997019374068554,
+      "grad_norm": 2.7306811809539795,
+      "learning_rate": 0.0002,
+      "loss": 2.5193,
+      "step": 49650
+    },
+    {
+      "epoch": 3.700447093889717,
+      "grad_norm": 2.5883631706237793,
+      "learning_rate": 0.0002,
+      "loss": 2.5261,
+      "step": 49660
+    },
+    {
+      "epoch": 3.7011922503725785,
+      "grad_norm": 2.3863844871520996,
+      "learning_rate": 0.0002,
+      "loss": 2.4421,
+      "step": 49670
+    },
+    {
+      "epoch": 3.7019374068554396,
+      "grad_norm": 2.6333467960357666,
+      "learning_rate": 0.0002,
+      "loss": 2.5035,
+      "step": 49680
+    },
+    {
+      "epoch": 3.702682563338301,
+      "grad_norm": 2.6204724311828613,
+      "learning_rate": 0.0002,
+      "loss": 2.3942,
+      "step": 49690
+    },
+    {
+      "epoch": 3.7034277198211623,
+      "grad_norm": 2.580791711807251,
+      "learning_rate": 0.0002,
+      "loss": 2.205,
+      "step": 49700
+    },
+    {
+      "epoch": 3.704172876304024,
+      "grad_norm": 2.720482587814331,
+      "learning_rate": 0.0002,
+      "loss": 2.4407,
+      "step": 49710
+    },
+    {
+      "epoch": 3.7049180327868854,
+      "grad_norm": 2.664541006088257,
+      "learning_rate": 0.0002,
+      "loss": 2.4447,
+      "step": 49720
+    },
+    {
+      "epoch": 3.705663189269747,
+      "grad_norm": 2.4946305751800537,
+      "learning_rate": 0.0002,
+      "loss": 2.5258,
+      "step": 49730
+    },
+    {
+      "epoch": 3.706408345752608,
+      "grad_norm": 2.704585552215576,
+      "learning_rate": 0.0002,
+      "loss": 2.3665,
+      "step": 49740
+    },
+    {
+      "epoch": 3.7071535022354696,
+      "grad_norm": 2.7640819549560547,
+      "learning_rate": 0.0002,
+      "loss": 2.5606,
+      "step": 49750
+    },
+    {
+      "epoch": 3.7078986587183307,
+      "grad_norm": 2.631479263305664,
+      "learning_rate": 0.0002,
+      "loss": 2.5774,
+      "step": 49760
+    },
+    {
+      "epoch": 3.7086438152011922,
+      "grad_norm": 2.800203800201416,
+      "learning_rate": 0.0002,
+      "loss": 2.4952,
+      "step": 49770
+    },
+    {
+      "epoch": 3.709388971684054,
+      "grad_norm": 2.66166090965271,
+      "learning_rate": 0.0002,
+      "loss": 2.5133,
+      "step": 49780
+    },
+    {
+      "epoch": 3.710134128166915,
+      "grad_norm": 2.272296905517578,
+      "learning_rate": 0.0002,
+      "loss": 2.5408,
+      "step": 49790
+    },
+    {
+      "epoch": 3.7108792846497765,
+      "grad_norm": 2.591278076171875,
+      "learning_rate": 0.0002,
+      "loss": 2.5519,
+      "step": 49800
+    },
+    {
+      "epoch": 3.711624441132638,
+      "grad_norm": 2.4364070892333984,
+      "learning_rate": 0.0002,
+      "loss": 2.4554,
+      "step": 49810
+    },
+    {
+      "epoch": 3.712369597615499,
+      "grad_norm": 2.7269444465637207,
+      "learning_rate": 0.0002,
+      "loss": 2.4394,
+      "step": 49820
+    },
+    {
+      "epoch": 3.7131147540983607,
+      "grad_norm": 2.7368903160095215,
+      "learning_rate": 0.0002,
+      "loss": 2.5778,
+      "step": 49830
+    },
+    {
+      "epoch": 3.7138599105812222,
+      "grad_norm": 2.6224026679992676,
+      "learning_rate": 0.0002,
+      "loss": 2.4692,
+      "step": 49840
+    },
+    {
+      "epoch": 3.7146050670640833,
+      "grad_norm": 2.544132709503174,
+      "learning_rate": 0.0002,
+      "loss": 2.5929,
+      "step": 49850
+    },
+    {
+      "epoch": 3.715350223546945,
+      "grad_norm": 2.141615629196167,
+      "learning_rate": 0.0002,
+      "loss": 2.4091,
+      "step": 49860
+    },
+    {
+      "epoch": 3.716095380029806,
+      "grad_norm": 2.357640027999878,
+      "learning_rate": 0.0002,
+      "loss": 2.5352,
+      "step": 49870
+    },
+    {
+      "epoch": 3.7168405365126675,
+      "grad_norm": 2.8612751960754395,
+      "learning_rate": 0.0002,
+      "loss": 2.6656,
+      "step": 49880
+    },
+    {
+      "epoch": 3.717585692995529,
+      "grad_norm": 2.558424472808838,
+      "learning_rate": 0.0002,
+      "loss": 2.4203,
+      "step": 49890
+    },
+    {
+      "epoch": 3.7183308494783907,
+      "grad_norm": 2.3906445503234863,
+      "learning_rate": 0.0002,
+      "loss": 2.4509,
+      "step": 49900
+    },
+    {
+      "epoch": 3.7190760059612518,
+      "grad_norm": 2.647514581680298,
+      "learning_rate": 0.0002,
+      "loss": 2.5821,
+      "step": 49910
+    },
+    {
+      "epoch": 3.7198211624441133,
+      "grad_norm": 2.710005283355713,
+      "learning_rate": 0.0002,
+      "loss": 2.5298,
+      "step": 49920
+    },
+    {
+      "epoch": 3.7205663189269744,
+      "grad_norm": 2.5558953285217285,
+      "learning_rate": 0.0002,
+      "loss": 2.4721,
+      "step": 49930
+    },
+    {
+      "epoch": 3.721311475409836,
+      "grad_norm": 2.446017265319824,
+      "learning_rate": 0.0002,
+      "loss": 2.4778,
+      "step": 49940
+    },
+    {
+      "epoch": 3.7220566318926975,
+      "grad_norm": 3.0394492149353027,
+      "learning_rate": 0.0002,
+      "loss": 2.3313,
+      "step": 49950
+    },
+    {
+      "epoch": 3.722801788375559,
+      "grad_norm": 2.5545358657836914,
+      "learning_rate": 0.0002,
+      "loss": 2.6545,
+      "step": 49960
+    },
+    {
+      "epoch": 3.72354694485842,
+      "grad_norm": 2.7570934295654297,
+      "learning_rate": 0.0002,
+      "loss": 2.4509,
+      "step": 49970
+    },
+    {
+      "epoch": 3.7242921013412817,
+      "grad_norm": 2.592240571975708,
+      "learning_rate": 0.0002,
+      "loss": 2.4852,
+      "step": 49980
+    },
+    {
+      "epoch": 3.725037257824143,
+      "grad_norm": 2.751447916030884,
+      "learning_rate": 0.0002,
+      "loss": 2.3889,
+      "step": 49990
+    },
+    {
+      "epoch": 3.7257824143070044,
+      "grad_norm": 2.723025321960449,
+      "learning_rate": 0.0002,
+      "loss": 2.3713,
+      "step": 50000
+    },
+    {
+      "epoch": 3.726527570789866,
+      "grad_norm": 2.759801149368286,
+      "learning_rate": 0.0002,
+      "loss": 2.507,
+      "step": 50010
+    },
+    {
+      "epoch": 3.7272727272727275,
+      "grad_norm": 2.721895933151245,
+      "learning_rate": 0.0002,
+      "loss": 2.4909,
+      "step": 50020
+    },
+    {
+      "epoch": 3.7280178837555886,
+      "grad_norm": 2.8816866874694824,
+      "learning_rate": 0.0002,
+      "loss": 2.5981,
+      "step": 50030
+    },
+    {
+      "epoch": 3.72876304023845,
+      "grad_norm": 2.328068256378174,
+      "learning_rate": 0.0002,
+      "loss": 2.5311,
+      "step": 50040
+    },
+    {
+      "epoch": 3.7295081967213113,
+      "grad_norm": 2.826514482498169,
+      "learning_rate": 0.0002,
+      "loss": 2.4776,
+      "step": 50050
+    },
+    {
+      "epoch": 3.730253353204173,
+      "grad_norm": 2.855038642883301,
+      "learning_rate": 0.0002,
+      "loss": 2.5622,
+      "step": 50060
+    },
+    {
+      "epoch": 3.7309985096870344,
+      "grad_norm": 2.6433498859405518,
+      "learning_rate": 0.0002,
+      "loss": 2.4781,
+      "step": 50070
+    },
+    {
+      "epoch": 3.731743666169896,
+      "grad_norm": 2.565962076187134,
+      "learning_rate": 0.0002,
+      "loss": 2.5446,
+      "step": 50080
+    },
+    {
+      "epoch": 3.732488822652757,
+      "grad_norm": 2.8265738487243652,
+      "learning_rate": 0.0002,
+      "loss": 2.534,
+      "step": 50090
+    },
+    {
+      "epoch": 3.7332339791356186,
+      "grad_norm": 2.9599409103393555,
+      "learning_rate": 0.0002,
+      "loss": 2.4206,
+      "step": 50100
+    },
+    {
+      "epoch": 3.7339791356184797,
+      "grad_norm": 2.669797897338867,
+      "learning_rate": 0.0002,
+      "loss": 2.6088,
+      "step": 50110
+    },
+    {
+      "epoch": 3.7347242921013413,
+      "grad_norm": 2.8583147525787354,
+      "learning_rate": 0.0002,
+      "loss": 2.5827,
+      "step": 50120
+    },
+    {
+      "epoch": 3.735469448584203,
+      "grad_norm": 2.217470407485962,
+      "learning_rate": 0.0002,
+      "loss": 2.4562,
+      "step": 50130
+    },
+    {
+      "epoch": 3.736214605067064,
+      "grad_norm": 3.1019763946533203,
+      "learning_rate": 0.0002,
+      "loss": 2.4064,
+      "step": 50140
+    },
+    {
+      "epoch": 3.7369597615499255,
+      "grad_norm": 2.963712692260742,
+      "learning_rate": 0.0002,
+      "loss": 2.5403,
+      "step": 50150
+    },
+    {
+      "epoch": 3.737704918032787,
+      "grad_norm": 3.015799045562744,
+      "learning_rate": 0.0002,
+      "loss": 2.6094,
+      "step": 50160
+    },
+    {
+      "epoch": 3.738450074515648,
+      "grad_norm": 2.435353994369507,
+      "learning_rate": 0.0002,
+      "loss": 2.4104,
+      "step": 50170
+    },
+    {
+      "epoch": 3.7391952309985097,
+      "grad_norm": 2.5293540954589844,
+      "learning_rate": 0.0002,
+      "loss": 2.5578,
+      "step": 50180
+    },
+    {
+      "epoch": 3.7399403874813713,
+      "grad_norm": 2.1452317237854004,
+      "learning_rate": 0.0002,
+      "loss": 2.4667,
+      "step": 50190
+    },
+    {
+      "epoch": 3.7406855439642324,
+      "grad_norm": 2.675736665725708,
+      "learning_rate": 0.0002,
+      "loss": 2.6145,
+      "step": 50200
+    },
+    {
+      "epoch": 3.741430700447094,
+      "grad_norm": 2.4674720764160156,
+      "learning_rate": 0.0002,
+      "loss": 2.3061,
+      "step": 50210
+    },
+    {
+      "epoch": 3.742175856929955,
+      "grad_norm": 2.5854485034942627,
+      "learning_rate": 0.0002,
+      "loss": 2.6369,
+      "step": 50220
+    },
+    {
+      "epoch": 3.7429210134128166,
+      "grad_norm": 2.5845370292663574,
+      "learning_rate": 0.0002,
+      "loss": 2.4753,
+      "step": 50230
+    },
+    {
+      "epoch": 3.743666169895678,
+      "grad_norm": 2.4826760292053223,
+      "learning_rate": 0.0002,
+      "loss": 2.4353,
+      "step": 50240
+    },
+    {
+      "epoch": 3.7444113263785397,
+      "grad_norm": 2.491818428039551,
+      "learning_rate": 0.0002,
+      "loss": 2.3542,
+      "step": 50250
+    },
+    {
+      "epoch": 3.745156482861401,
+      "grad_norm": 2.7059671878814697,
+      "learning_rate": 0.0002,
+      "loss": 2.3786,
+      "step": 50260
+    },
+    {
+      "epoch": 3.7459016393442623,
+      "grad_norm": 2.488055944442749,
+      "learning_rate": 0.0002,
+      "loss": 2.3972,
+      "step": 50270
+    },
+    {
+      "epoch": 3.7466467958271235,
+      "grad_norm": 2.8024628162384033,
+      "learning_rate": 0.0002,
+      "loss": 2.4418,
+      "step": 50280
+    },
+    {
+      "epoch": 3.747391952309985,
+      "grad_norm": 2.639986991882324,
+      "learning_rate": 0.0002,
+      "loss": 2.5661,
+      "step": 50290
+    },
+    {
+      "epoch": 3.7481371087928466,
+      "grad_norm": 2.477285623550415,
+      "learning_rate": 0.0002,
+      "loss": 2.4308,
+      "step": 50300
+    },
+    {
+      "epoch": 3.748882265275708,
+      "grad_norm": 2.709949254989624,
+      "learning_rate": 0.0002,
+      "loss": 2.6701,
+      "step": 50310
+    },
+    {
+      "epoch": 3.7496274217585692,
+      "grad_norm": 2.567444324493408,
+      "learning_rate": 0.0002,
+      "loss": 2.6772,
+      "step": 50320
+    },
+    {
+      "epoch": 3.7503725782414308,
+      "grad_norm": 2.8398115634918213,
+      "learning_rate": 0.0002,
+      "loss": 2.5978,
+      "step": 50330
+    },
+    {
+      "epoch": 3.751117734724292,
+      "grad_norm": 2.623926877975464,
+      "learning_rate": 0.0002,
+      "loss": 2.4706,
+      "step": 50340
+    },
+    {
+      "epoch": 3.7518628912071534,
+      "grad_norm": 2.404278039932251,
+      "learning_rate": 0.0002,
+      "loss": 2.6895,
+      "step": 50350
+    },
+    {
+      "epoch": 3.752608047690015,
+      "grad_norm": 2.811429023742676,
+      "learning_rate": 0.0002,
+      "loss": 2.3915,
+      "step": 50360
+    },
+    {
+      "epoch": 3.7533532041728765,
+      "grad_norm": 2.551865339279175,
+      "learning_rate": 0.0002,
+      "loss": 2.3965,
+      "step": 50370
+    },
+    {
+      "epoch": 3.7540983606557377,
+      "grad_norm": 3.0696215629577637,
+      "learning_rate": 0.0002,
+      "loss": 2.4566,
+      "step": 50380
+    },
+    {
+      "epoch": 3.754843517138599,
+      "grad_norm": 2.6258938312530518,
+      "learning_rate": 0.0002,
+      "loss": 2.564,
+      "step": 50390
+    },
+    {
+      "epoch": 3.7555886736214603,
+      "grad_norm": 2.8297739028930664,
+      "learning_rate": 0.0002,
+      "loss": 2.7246,
+      "step": 50400
+    },
+    {
+      "epoch": 3.756333830104322,
+      "grad_norm": 2.9615511894226074,
+      "learning_rate": 0.0002,
+      "loss": 2.6202,
+      "step": 50410
+    },
+    {
+      "epoch": 3.7570789865871834,
+      "grad_norm": 2.3409204483032227,
+      "learning_rate": 0.0002,
+      "loss": 2.5359,
+      "step": 50420
+    },
+    {
+      "epoch": 3.757824143070045,
+      "grad_norm": 2.545732021331787,
+      "learning_rate": 0.0002,
+      "loss": 2.5465,
+      "step": 50430
+    },
+    {
+      "epoch": 3.758569299552906,
+      "grad_norm": 2.5160508155822754,
+      "learning_rate": 0.0002,
+      "loss": 2.5209,
+      "step": 50440
+    },
+    {
+      "epoch": 3.7593144560357676,
+      "grad_norm": 2.9805400371551514,
+      "learning_rate": 0.0002,
+      "loss": 2.4882,
+      "step": 50450
+    },
+    {
+      "epoch": 3.7600596125186287,
+      "grad_norm": 2.4430508613586426,
+      "learning_rate": 0.0002,
+      "loss": 2.5019,
+      "step": 50460
+    },
+    {
+      "epoch": 3.7608047690014903,
+      "grad_norm": 2.3823506832122803,
+      "learning_rate": 0.0002,
+      "loss": 2.3845,
+      "step": 50470
+    },
+    {
+      "epoch": 3.761549925484352,
+      "grad_norm": 2.54958176612854,
+      "learning_rate": 0.0002,
+      "loss": 2.4257,
+      "step": 50480
+    },
+    {
+      "epoch": 3.762295081967213,
+      "grad_norm": 2.992579460144043,
+      "learning_rate": 0.0002,
+      "loss": 2.5295,
+      "step": 50490
+    },
+    {
+      "epoch": 3.7630402384500745,
+      "grad_norm": 2.7364273071289062,
+      "learning_rate": 0.0002,
+      "loss": 2.7259,
+      "step": 50500
+    },
+    {
+      "epoch": 3.763785394932936,
+      "grad_norm": 2.527580499649048,
+      "learning_rate": 0.0002,
+      "loss": 2.5548,
+      "step": 50510
+    },
+    {
+      "epoch": 3.764530551415797,
+      "grad_norm": 2.6764330863952637,
+      "learning_rate": 0.0002,
+      "loss": 2.4295,
+      "step": 50520
+    },
+    {
+      "epoch": 3.7652757078986587,
+      "grad_norm": 2.7412283420562744,
+      "learning_rate": 0.0002,
+      "loss": 2.5146,
+      "step": 50530
+    },
+    {
+      "epoch": 3.7660208643815203,
+      "grad_norm": 2.5869789123535156,
+      "learning_rate": 0.0002,
+      "loss": 2.5206,
+      "step": 50540
+    },
+    {
+      "epoch": 3.7667660208643814,
+      "grad_norm": 2.71201491355896,
+      "learning_rate": 0.0002,
+      "loss": 2.4387,
+      "step": 50550
+    },
+    {
+      "epoch": 3.767511177347243,
+      "grad_norm": 2.881373167037964,
+      "learning_rate": 0.0002,
+      "loss": 2.5828,
+      "step": 50560
+    },
+    {
+      "epoch": 3.768256333830104,
+      "grad_norm": 2.6753411293029785,
+      "learning_rate": 0.0002,
+      "loss": 2.6769,
+      "step": 50570
+    },
+    {
+      "epoch": 3.7690014903129656,
+      "grad_norm": 2.626857042312622,
+      "learning_rate": 0.0002,
+      "loss": 2.5313,
+      "step": 50580
+    },
+    {
+      "epoch": 3.769746646795827,
+      "grad_norm": 2.0912129878997803,
+      "learning_rate": 0.0002,
+      "loss": 2.1949,
+      "step": 50590
+    },
+    {
+      "epoch": 3.7704918032786887,
+      "grad_norm": 2.5169942378997803,
+      "learning_rate": 0.0002,
+      "loss": 2.4955,
+      "step": 50600
+    },
+    {
+      "epoch": 3.77123695976155,
+      "grad_norm": 2.7082998752593994,
+      "learning_rate": 0.0002,
+      "loss": 2.5901,
+      "step": 50610
+    },
+    {
+      "epoch": 3.7719821162444114,
+      "grad_norm": 2.681004762649536,
+      "learning_rate": 0.0002,
+      "loss": 2.5101,
+      "step": 50620
+    },
+    {
+      "epoch": 3.7727272727272725,
+      "grad_norm": 2.7208945751190186,
+      "learning_rate": 0.0002,
+      "loss": 2.4098,
+      "step": 50630
+    },
+    {
+      "epoch": 3.773472429210134,
+      "grad_norm": 2.369579553604126,
+      "learning_rate": 0.0002,
+      "loss": 2.4761,
+      "step": 50640
+    },
+    {
+      "epoch": 3.7742175856929956,
+      "grad_norm": 2.708627223968506,
+      "learning_rate": 0.0002,
+      "loss": 2.5875,
+      "step": 50650
+    },
+    {
+      "epoch": 3.774962742175857,
+      "grad_norm": 2.6574110984802246,
+      "learning_rate": 0.0002,
+      "loss": 2.4478,
+      "step": 50660
+    },
+    {
+      "epoch": 3.7757078986587183,
+      "grad_norm": 2.582005739212036,
+      "learning_rate": 0.0002,
+      "loss": 2.4235,
+      "step": 50670
+    },
+    {
+      "epoch": 3.77645305514158,
+      "grad_norm": 2.9919252395629883,
+      "learning_rate": 0.0002,
+      "loss": 2.5915,
+      "step": 50680
+    },
+    {
+      "epoch": 3.777198211624441,
+      "grad_norm": 2.489131212234497,
+      "learning_rate": 0.0002,
+      "loss": 2.5323,
+      "step": 50690
+    },
+    {
+      "epoch": 3.7779433681073025,
+      "grad_norm": 3.1119613647460938,
+      "learning_rate": 0.0002,
+      "loss": 2.4534,
+      "step": 50700
+    },
+    {
+      "epoch": 3.778688524590164,
+      "grad_norm": 2.639336585998535,
+      "learning_rate": 0.0002,
+      "loss": 2.5858,
+      "step": 50710
+    },
+    {
+      "epoch": 3.7794336810730256,
+      "grad_norm": 2.6446197032928467,
+      "learning_rate": 0.0002,
+      "loss": 2.4894,
+      "step": 50720
+    },
+    {
+      "epoch": 3.7801788375558867,
+      "grad_norm": 2.721428871154785,
+      "learning_rate": 0.0002,
+      "loss": 2.6664,
+      "step": 50730
+    },
+    {
+      "epoch": 3.7809239940387482,
+      "grad_norm": 3.0219340324401855,
+      "learning_rate": 0.0002,
+      "loss": 2.5447,
+      "step": 50740
+    },
+    {
+      "epoch": 3.7816691505216093,
+      "grad_norm": 2.371525764465332,
+      "learning_rate": 0.0002,
+      "loss": 2.2999,
+      "step": 50750
+    },
+    {
+      "epoch": 3.782414307004471,
+      "grad_norm": 3.023408889770508,
+      "learning_rate": 0.0002,
+      "loss": 2.4849,
+      "step": 50760
+    },
+    {
+      "epoch": 3.7831594634873325,
+      "grad_norm": 2.1331787109375,
+      "learning_rate": 0.0002,
+      "loss": 2.5371,
+      "step": 50770
+    },
+    {
+      "epoch": 3.783904619970194,
+      "grad_norm": 2.6607348918914795,
+      "learning_rate": 0.0002,
+      "loss": 2.5061,
+      "step": 50780
+    },
+    {
+      "epoch": 3.784649776453055,
+      "grad_norm": 2.316195011138916,
+      "learning_rate": 0.0002,
+      "loss": 2.5484,
+      "step": 50790
+    },
+    {
+      "epoch": 3.7853949329359167,
+      "grad_norm": 2.7560927867889404,
+      "learning_rate": 0.0002,
+      "loss": 2.4545,
+      "step": 50800
+    },
+    {
+      "epoch": 3.7861400894187778,
+      "grad_norm": 2.3662703037261963,
+      "learning_rate": 0.0002,
+      "loss": 2.5492,
+      "step": 50810
+    },
+    {
+      "epoch": 3.7868852459016393,
+      "grad_norm": 2.9224653244018555,
+      "learning_rate": 0.0002,
+      "loss": 2.6868,
+      "step": 50820
+    },
+    {
+      "epoch": 3.787630402384501,
+      "grad_norm": 2.669923782348633,
+      "learning_rate": 0.0002,
+      "loss": 2.3945,
+      "step": 50830
+    },
+    {
+      "epoch": 3.788375558867362,
+      "grad_norm": 2.4683589935302734,
+      "learning_rate": 0.0002,
+      "loss": 2.6,
+      "step": 50840
+    },
+    {
+      "epoch": 3.7891207153502235,
+      "grad_norm": 2.4863712787628174,
+      "learning_rate": 0.0002,
+      "loss": 2.6107,
+      "step": 50850
+    },
+    {
+      "epoch": 3.789865871833085,
+      "grad_norm": 2.4746665954589844,
+      "learning_rate": 0.0002,
+      "loss": 2.421,
+      "step": 50860
+    },
+    {
+      "epoch": 3.790611028315946,
+      "grad_norm": 2.53080153465271,
+      "learning_rate": 0.0002,
+      "loss": 2.2403,
+      "step": 50870
+    },
+    {
+      "epoch": 3.7913561847988078,
+      "grad_norm": 2.8547794818878174,
+      "learning_rate": 0.0002,
+      "loss": 2.4001,
+      "step": 50880
+    },
+    {
+      "epoch": 3.7921013412816693,
+      "grad_norm": 2.5060040950775146,
+      "learning_rate": 0.0002,
+      "loss": 2.4906,
+      "step": 50890
+    },
+    {
+      "epoch": 3.7928464977645304,
+      "grad_norm": 2.542915105819702,
+      "learning_rate": 0.0002,
+      "loss": 2.3689,
+      "step": 50900
+    },
+    {
+      "epoch": 3.793591654247392,
+      "grad_norm": 2.546762704849243,
+      "learning_rate": 0.0002,
+      "loss": 2.5893,
+      "step": 50910
+    },
+    {
+      "epoch": 3.794336810730253,
+      "grad_norm": 2.9480457305908203,
+      "learning_rate": 0.0002,
+      "loss": 2.4305,
+      "step": 50920
+    },
+    {
+      "epoch": 3.7950819672131146,
+      "grad_norm": 2.469069719314575,
+      "learning_rate": 0.0002,
+      "loss": 2.3765,
+      "step": 50930
+    },
+    {
+      "epoch": 3.795827123695976,
+      "grad_norm": 2.714322328567505,
+      "learning_rate": 0.0002,
+      "loss": 2.5366,
+      "step": 50940
+    },
+    {
+      "epoch": 3.7965722801788377,
+      "grad_norm": 2.614262580871582,
+      "learning_rate": 0.0002,
+      "loss": 2.721,
+      "step": 50950
+    },
+    {
+      "epoch": 3.797317436661699,
+      "grad_norm": 2.443918466567993,
+      "learning_rate": 0.0002,
+      "loss": 2.3982,
+      "step": 50960
+    },
+    {
+      "epoch": 3.7980625931445604,
+      "grad_norm": 2.338960647583008,
+      "learning_rate": 0.0002,
+      "loss": 2.3152,
+      "step": 50970
+    },
+    {
+      "epoch": 3.7988077496274215,
+      "grad_norm": 2.9768331050872803,
+      "learning_rate": 0.0002,
+      "loss": 2.4752,
+      "step": 50980
+    },
+    {
+      "epoch": 3.799552906110283,
+      "grad_norm": 2.6073148250579834,
+      "learning_rate": 0.0002,
+      "loss": 2.4918,
+      "step": 50990
+    },
+    {
+      "epoch": 3.8002980625931446,
+      "grad_norm": 2.7576828002929688,
+      "learning_rate": 0.0002,
+      "loss": 2.5247,
+      "step": 51000
+    },
+    {
+      "epoch": 3.801043219076006,
+      "grad_norm": 2.710273504257202,
+      "learning_rate": 0.0002,
+      "loss": 2.6895,
+      "step": 51010
+    },
+    {
+      "epoch": 3.8017883755588673,
+      "grad_norm": 2.608431339263916,
+      "learning_rate": 0.0002,
+      "loss": 2.4041,
+      "step": 51020
+    },
+    {
+      "epoch": 3.802533532041729,
+      "grad_norm": 2.511503219604492,
+      "learning_rate": 0.0002,
+      "loss": 2.4616,
+      "step": 51030
+    },
+    {
+      "epoch": 3.80327868852459,
+      "grad_norm": 2.6001126766204834,
+      "learning_rate": 0.0002,
+      "loss": 2.467,
+      "step": 51040
+    },
+    {
+      "epoch": 3.8040238450074515,
+      "grad_norm": 2.7922747135162354,
+      "learning_rate": 0.0002,
+      "loss": 2.5395,
+      "step": 51050
+    },
+    {
+      "epoch": 3.804769001490313,
+      "grad_norm": 2.260394334793091,
+      "learning_rate": 0.0002,
+      "loss": 2.2577,
+      "step": 51060
+    },
+    {
+      "epoch": 3.8055141579731746,
+      "grad_norm": 2.6235592365264893,
+      "learning_rate": 0.0002,
+      "loss": 2.3754,
+      "step": 51070
+    },
+    {
+      "epoch": 3.8062593144560357,
+      "grad_norm": 2.6814324855804443,
+      "learning_rate": 0.0002,
+      "loss": 2.495,
+      "step": 51080
+    },
+    {
+      "epoch": 3.8070044709388973,
+      "grad_norm": 1.8012224435806274,
+      "learning_rate": 0.0002,
+      "loss": 2.4035,
+      "step": 51090
+    },
+    {
+      "epoch": 3.8077496274217584,
+      "grad_norm": 2.7128658294677734,
+      "learning_rate": 0.0002,
+      "loss": 2.6043,
+      "step": 51100
+    },
+    {
+      "epoch": 3.80849478390462,
+      "grad_norm": 2.600168466567993,
+      "learning_rate": 0.0002,
+      "loss": 2.6245,
+      "step": 51110
+    },
+    {
+      "epoch": 3.8092399403874815,
+      "grad_norm": 2.9342854022979736,
+      "learning_rate": 0.0002,
+      "loss": 2.5065,
+      "step": 51120
+    },
+    {
+      "epoch": 3.809985096870343,
+      "grad_norm": 2.6882317066192627,
+      "learning_rate": 0.0002,
+      "loss": 2.5987,
+      "step": 51130
+    },
+    {
+      "epoch": 3.810730253353204,
+      "grad_norm": 2.672231674194336,
+      "learning_rate": 0.0002,
+      "loss": 2.5405,
+      "step": 51140
+    },
+    {
+      "epoch": 3.8114754098360657,
+      "grad_norm": 2.7051048278808594,
+      "learning_rate": 0.0002,
+      "loss": 2.5367,
+      "step": 51150
+    },
+    {
+      "epoch": 3.812220566318927,
+      "grad_norm": 2.8326873779296875,
+      "learning_rate": 0.0002,
+      "loss": 2.6169,
+      "step": 51160
+    },
+    {
+      "epoch": 3.8129657228017884,
+      "grad_norm": 2.9274744987487793,
+      "learning_rate": 0.0002,
+      "loss": 2.6881,
+      "step": 51170
+    },
+    {
+      "epoch": 3.81371087928465,
+      "grad_norm": 2.338207721710205,
+      "learning_rate": 0.0002,
+      "loss": 2.5462,
+      "step": 51180
+    },
+    {
+      "epoch": 3.814456035767511,
+      "grad_norm": 2.5919198989868164,
+      "learning_rate": 0.0002,
+      "loss": 2.2888,
+      "step": 51190
+    },
+    {
+      "epoch": 3.8152011922503726,
+      "grad_norm": 2.5705173015594482,
+      "learning_rate": 0.0002,
+      "loss": 2.5768,
+      "step": 51200
+    },
+    {
+      "epoch": 3.815946348733234,
+      "grad_norm": 2.6229724884033203,
+      "learning_rate": 0.0002,
+      "loss": 2.4647,
+      "step": 51210
+    },
+    {
+      "epoch": 3.8166915052160952,
+      "grad_norm": 2.2743210792541504,
+      "learning_rate": 0.0002,
+      "loss": 2.3109,
+      "step": 51220
+    },
+    {
+      "epoch": 3.817436661698957,
+      "grad_norm": 2.2738277912139893,
+      "learning_rate": 0.0002,
+      "loss": 2.5418,
+      "step": 51230
+    },
+    {
+      "epoch": 3.8181818181818183,
+      "grad_norm": 2.817492961883545,
+      "learning_rate": 0.0002,
+      "loss": 2.5758,
+      "step": 51240
+    },
+    {
+      "epoch": 3.8189269746646795,
+      "grad_norm": 2.5765511989593506,
+      "learning_rate": 0.0002,
+      "loss": 2.428,
+      "step": 51250
+    },
+    {
+      "epoch": 3.819672131147541,
+      "grad_norm": 2.62100887298584,
+      "learning_rate": 0.0002,
+      "loss": 2.2739,
+      "step": 51260
+    },
+    {
+      "epoch": 3.820417287630402,
+      "grad_norm": 2.599499225616455,
+      "learning_rate": 0.0002,
+      "loss": 2.6626,
+      "step": 51270
+    },
+    {
+      "epoch": 3.8211624441132637,
+      "grad_norm": 3.347637176513672,
+      "learning_rate": 0.0002,
+      "loss": 2.441,
+      "step": 51280
+    },
+    {
+      "epoch": 3.821907600596125,
+      "grad_norm": 2.842421293258667,
+      "learning_rate": 0.0002,
+      "loss": 2.5185,
+      "step": 51290
+    },
+    {
+      "epoch": 3.8226527570789868,
+      "grad_norm": 2.7661526203155518,
+      "learning_rate": 0.0002,
+      "loss": 2.5964,
+      "step": 51300
+    },
+    {
+      "epoch": 3.823397913561848,
+      "grad_norm": 2.7636051177978516,
+      "learning_rate": 0.0002,
+      "loss": 2.6437,
+      "step": 51310
+    },
+    {
+      "epoch": 3.8241430700447094,
+      "grad_norm": 2.534313678741455,
+      "learning_rate": 0.0002,
+      "loss": 2.4475,
+      "step": 51320
+    },
+    {
+      "epoch": 3.8248882265275705,
+      "grad_norm": 1.8897809982299805,
+      "learning_rate": 0.0002,
+      "loss": 2.1495,
+      "step": 51330
+    },
+    {
+      "epoch": 3.825633383010432,
+      "grad_norm": 2.465552806854248,
+      "learning_rate": 0.0002,
+      "loss": 2.4313,
+      "step": 51340
+    },
+    {
+      "epoch": 3.8263785394932937,
+      "grad_norm": 2.8640079498291016,
+      "learning_rate": 0.0002,
+      "loss": 2.6838,
+      "step": 51350
+    },
+    {
+      "epoch": 3.827123695976155,
+      "grad_norm": 1.6258368492126465,
+      "learning_rate": 0.0002,
+      "loss": 2.3335,
+      "step": 51360
+    },
+    {
+      "epoch": 3.8278688524590163,
+      "grad_norm": 2.6712746620178223,
+      "learning_rate": 0.0002,
+      "loss": 2.5755,
+      "step": 51370
+    },
+    {
+      "epoch": 3.828614008941878,
+      "grad_norm": 2.809267282485962,
+      "learning_rate": 0.0002,
+      "loss": 2.4954,
+      "step": 51380
+    },
+    {
+      "epoch": 3.829359165424739,
+      "grad_norm": 2.894386053085327,
+      "learning_rate": 0.0002,
+      "loss": 2.4332,
+      "step": 51390
+    },
+    {
+      "epoch": 3.8301043219076005,
+      "grad_norm": 2.8683619499206543,
+      "learning_rate": 0.0002,
+      "loss": 2.4112,
+      "step": 51400
+    },
+    {
+      "epoch": 3.830849478390462,
+      "grad_norm": 2.383283853530884,
+      "learning_rate": 0.0002,
+      "loss": 2.5905,
+      "step": 51410
+    },
+    {
+      "epoch": 3.8315946348733236,
+      "grad_norm": 2.6616315841674805,
+      "learning_rate": 0.0002,
+      "loss": 2.416,
+      "step": 51420
+    },
+    {
+      "epoch": 3.8323397913561847,
+      "grad_norm": 3.09716796875,
+      "learning_rate": 0.0002,
+      "loss": 2.5738,
+      "step": 51430
+    },
+    {
+      "epoch": 3.8330849478390463,
+      "grad_norm": 2.646336793899536,
+      "learning_rate": 0.0002,
+      "loss": 2.5713,
+      "step": 51440
+    },
+    {
+      "epoch": 3.8338301043219074,
+      "grad_norm": 2.847592353820801,
+      "learning_rate": 0.0002,
+      "loss": 2.3993,
+      "step": 51450
+    },
+    {
+      "epoch": 3.834575260804769,
+      "grad_norm": 2.573282480239868,
+      "learning_rate": 0.0002,
+      "loss": 2.4144,
+      "step": 51460
+    },
+    {
+      "epoch": 3.8353204172876305,
+      "grad_norm": 2.621114492416382,
+      "learning_rate": 0.0002,
+      "loss": 2.418,
+      "step": 51470
+    },
+    {
+      "epoch": 3.836065573770492,
+      "grad_norm": 2.757758855819702,
+      "learning_rate": 0.0002,
+      "loss": 2.3968,
+      "step": 51480
+    },
+    {
+      "epoch": 3.836810730253353,
+      "grad_norm": 2.8561811447143555,
+      "learning_rate": 0.0002,
+      "loss": 2.4822,
+      "step": 51490
+    },
+    {
+      "epoch": 3.8375558867362147,
+      "grad_norm": 2.84928035736084,
+      "learning_rate": 0.0002,
+      "loss": 2.2998,
+      "step": 51500
+    },
+    {
+      "epoch": 3.838301043219076,
+      "grad_norm": 2.6037542819976807,
+      "learning_rate": 0.0002,
+      "loss": 2.4761,
+      "step": 51510
+    },
+    {
+      "epoch": 3.8390461997019374,
+      "grad_norm": 3.122798442840576,
+      "learning_rate": 0.0002,
+      "loss": 2.6203,
+      "step": 51520
+    },
+    {
+      "epoch": 3.839791356184799,
+      "grad_norm": 2.524787664413452,
+      "learning_rate": 0.0002,
+      "loss": 2.4676,
+      "step": 51530
+    },
+    {
+      "epoch": 3.84053651266766,
+      "grad_norm": 2.385199785232544,
+      "learning_rate": 0.0002,
+      "loss": 2.4442,
+      "step": 51540
+    },
+    {
+      "epoch": 3.8412816691505216,
+      "grad_norm": 2.4337494373321533,
+      "learning_rate": 0.0002,
+      "loss": 2.5004,
+      "step": 51550
+    },
+    {
+      "epoch": 3.842026825633383,
+      "grad_norm": 2.2421205043792725,
+      "learning_rate": 0.0002,
+      "loss": 2.362,
+      "step": 51560
+    },
+    {
+      "epoch": 3.8427719821162443,
+      "grad_norm": 2.355182647705078,
+      "learning_rate": 0.0002,
+      "loss": 2.4003,
+      "step": 51570
+    },
+    {
+      "epoch": 3.843517138599106,
+      "grad_norm": 2.7259955406188965,
+      "learning_rate": 0.0002,
+      "loss": 2.362,
+      "step": 51580
+    },
+    {
+      "epoch": 3.8442622950819674,
+      "grad_norm": 2.447803258895874,
+      "learning_rate": 0.0002,
+      "loss": 2.4592,
+      "step": 51590
+    },
+    {
+      "epoch": 3.8450074515648285,
+      "grad_norm": 3.1207966804504395,
+      "learning_rate": 0.0002,
+      "loss": 2.3862,
+      "step": 51600
+    },
+    {
+      "epoch": 3.84575260804769,
+      "grad_norm": 3.0702199935913086,
+      "learning_rate": 0.0002,
+      "loss": 2.58,
+      "step": 51610
+    },
+    {
+      "epoch": 3.846497764530551,
+      "grad_norm": 2.7334744930267334,
+      "learning_rate": 0.0002,
+      "loss": 2.5699,
+      "step": 51620
+    },
+    {
+      "epoch": 3.8472429210134127,
+      "grad_norm": 2.2549920082092285,
+      "learning_rate": 0.0002,
+      "loss": 2.3604,
+      "step": 51630
+    },
+    {
+      "epoch": 3.8479880774962743,
+      "grad_norm": 2.400271415710449,
+      "learning_rate": 0.0002,
+      "loss": 2.6297,
+      "step": 51640
+    },
+    {
+      "epoch": 3.848733233979136,
+      "grad_norm": 2.1789305210113525,
+      "learning_rate": 0.0002,
+      "loss": 2.5277,
+      "step": 51650
+    },
+    {
+      "epoch": 3.849478390461997,
+      "grad_norm": 2.701901435852051,
+      "learning_rate": 0.0002,
+      "loss": 2.4133,
+      "step": 51660
+    },
+    {
+      "epoch": 3.8502235469448585,
+      "grad_norm": 2.7472198009490967,
+      "learning_rate": 0.0002,
+      "loss": 2.3137,
+      "step": 51670
+    },
+    {
+      "epoch": 3.8509687034277196,
+      "grad_norm": 2.6537559032440186,
+      "learning_rate": 0.0002,
+      "loss": 2.5885,
+      "step": 51680
+    },
+    {
+      "epoch": 3.851713859910581,
+      "grad_norm": 2.521488666534424,
+      "learning_rate": 0.0002,
+      "loss": 2.5628,
+      "step": 51690
+    },
+    {
+      "epoch": 3.8524590163934427,
+      "grad_norm": 2.5507121086120605,
+      "learning_rate": 0.0002,
+      "loss": 2.4122,
+      "step": 51700
+    },
+    {
+      "epoch": 3.8532041728763042,
+      "grad_norm": 2.589373826980591,
+      "learning_rate": 0.0002,
+      "loss": 2.6058,
+      "step": 51710
+    },
+    {
+      "epoch": 3.8539493293591653,
+      "grad_norm": 2.9518160820007324,
+      "learning_rate": 0.0002,
+      "loss": 2.5514,
+      "step": 51720
+    },
+    {
+      "epoch": 3.854694485842027,
+      "grad_norm": 2.976022243499756,
+      "learning_rate": 0.0002,
+      "loss": 2.556,
+      "step": 51730
+    },
+    {
+      "epoch": 3.855439642324888,
+      "grad_norm": 2.435248374938965,
+      "learning_rate": 0.0002,
+      "loss": 2.4229,
+      "step": 51740
+    },
+    {
+      "epoch": 3.8561847988077496,
+      "grad_norm": 2.6801443099975586,
+      "learning_rate": 0.0002,
+      "loss": 2.4686,
+      "step": 51750
+    },
+    {
+      "epoch": 3.856929955290611,
+      "grad_norm": 1.9394075870513916,
+      "learning_rate": 0.0002,
+      "loss": 2.2974,
+      "step": 51760
+    },
+    {
+      "epoch": 3.8576751117734727,
+      "grad_norm": 2.496823310852051,
+      "learning_rate": 0.0002,
+      "loss": 2.3496,
+      "step": 51770
+    },
+    {
+      "epoch": 3.8584202682563338,
+      "grad_norm": 3.03865122795105,
+      "learning_rate": 0.0002,
+      "loss": 2.2881,
+      "step": 51780
+    },
+    {
+      "epoch": 3.8591654247391953,
+      "grad_norm": 2.5730419158935547,
+      "learning_rate": 0.0002,
+      "loss": 2.392,
+      "step": 51790
+    },
+    {
+      "epoch": 3.8599105812220564,
+      "grad_norm": 2.6450493335723877,
+      "learning_rate": 0.0002,
+      "loss": 2.5092,
+      "step": 51800
+    },
+    {
+      "epoch": 3.860655737704918,
+      "grad_norm": 2.5688185691833496,
+      "learning_rate": 0.0002,
+      "loss": 2.366,
+      "step": 51810
+    },
+    {
+      "epoch": 3.8614008941877795,
+      "grad_norm": 2.4910011291503906,
+      "learning_rate": 0.0002,
+      "loss": 2.4483,
+      "step": 51820
+    },
+    {
+      "epoch": 3.862146050670641,
+      "grad_norm": 3.078287363052368,
+      "learning_rate": 0.0002,
+      "loss": 2.6476,
+      "step": 51830
+    },
+    {
+      "epoch": 3.862891207153502,
+      "grad_norm": 2.2944555282592773,
+      "learning_rate": 0.0002,
+      "loss": 2.5578,
+      "step": 51840
+    },
+    {
+      "epoch": 3.8636363636363638,
+      "grad_norm": 2.777221202850342,
+      "learning_rate": 0.0002,
+      "loss": 2.4892,
+      "step": 51850
+    },
+    {
+      "epoch": 3.864381520119225,
+      "grad_norm": 2.459630250930786,
+      "learning_rate": 0.0002,
+      "loss": 2.5713,
+      "step": 51860
+    },
+    {
+      "epoch": 3.8651266766020864,
+      "grad_norm": 3.0153188705444336,
+      "learning_rate": 0.0002,
+      "loss": 2.6051,
+      "step": 51870
+    },
+    {
+      "epoch": 3.865871833084948,
+      "grad_norm": 2.6499626636505127,
+      "learning_rate": 0.0002,
+      "loss": 2.5795,
+      "step": 51880
+    },
+    {
+      "epoch": 3.866616989567809,
+      "grad_norm": 2.9734129905700684,
+      "learning_rate": 0.0002,
+      "loss": 2.5746,
+      "step": 51890
+    },
+    {
+      "epoch": 3.8673621460506706,
+      "grad_norm": 2.2308945655822754,
+      "learning_rate": 0.0002,
+      "loss": 2.4392,
+      "step": 51900
+    },
+    {
+      "epoch": 3.868107302533532,
+      "grad_norm": 2.5618507862091064,
+      "learning_rate": 0.0002,
+      "loss": 2.4746,
+      "step": 51910
+    },
+    {
+      "epoch": 3.8688524590163933,
+      "grad_norm": 2.3245627880096436,
+      "learning_rate": 0.0002,
+      "loss": 2.4258,
+      "step": 51920
+    },
+    {
+      "epoch": 3.869597615499255,
+      "grad_norm": 2.2224037647247314,
+      "learning_rate": 0.0002,
+      "loss": 2.5592,
+      "step": 51930
+    },
+    {
+      "epoch": 3.8703427719821164,
+      "grad_norm": 3.254966974258423,
+      "learning_rate": 0.0002,
+      "loss": 2.5822,
+      "step": 51940
+    },
+    {
+      "epoch": 3.8710879284649775,
+      "grad_norm": 2.469639778137207,
+      "learning_rate": 0.0002,
+      "loss": 2.4718,
+      "step": 51950
+    },
+    {
+      "epoch": 3.871833084947839,
+      "grad_norm": 2.6414148807525635,
+      "learning_rate": 0.0002,
+      "loss": 2.5728,
+      "step": 51960
+    },
+    {
+      "epoch": 3.8725782414307,
+      "grad_norm": 2.460594892501831,
+      "learning_rate": 0.0002,
+      "loss": 2.3055,
+      "step": 51970
+    },
+    {
+      "epoch": 3.8733233979135617,
+      "grad_norm": 2.6977572441101074,
+      "learning_rate": 0.0002,
+      "loss": 2.5869,
+      "step": 51980
+    },
+    {
+      "epoch": 3.8740685543964233,
+      "grad_norm": 2.5198538303375244,
+      "learning_rate": 0.0002,
+      "loss": 2.3507,
+      "step": 51990
+    },
+    {
+      "epoch": 3.874813710879285,
+      "grad_norm": 2.7170722484588623,
+      "learning_rate": 0.0002,
+      "loss": 2.6903,
+      "step": 52000
+    },
+    {
+      "epoch": 3.875558867362146,
+      "grad_norm": 2.7183640003204346,
+      "learning_rate": 0.0002,
+      "loss": 2.6414,
+      "step": 52010
+    },
+    {
+      "epoch": 3.8763040238450075,
+      "grad_norm": 2.5241034030914307,
+      "learning_rate": 0.0002,
+      "loss": 2.2518,
+      "step": 52020
+    },
+    {
+      "epoch": 3.8770491803278686,
+      "grad_norm": 2.6409521102905273,
+      "learning_rate": 0.0002,
+      "loss": 2.4558,
+      "step": 52030
+    },
+    {
+      "epoch": 3.87779433681073,
+      "grad_norm": 2.5587220191955566,
+      "learning_rate": 0.0002,
+      "loss": 2.3879,
+      "step": 52040
+    },
+    {
+      "epoch": 3.8785394932935917,
+      "grad_norm": 2.8432421684265137,
+      "learning_rate": 0.0002,
+      "loss": 2.6653,
+      "step": 52050
+    },
+    {
+      "epoch": 3.8792846497764533,
+      "grad_norm": 3.088017225265503,
+      "learning_rate": 0.0002,
+      "loss": 2.5824,
+      "step": 52060
+    },
+    {
+      "epoch": 3.8800298062593144,
+      "grad_norm": 2.6075172424316406,
+      "learning_rate": 0.0002,
+      "loss": 2.2853,
+      "step": 52070
+    },
+    {
+      "epoch": 3.880774962742176,
+      "grad_norm": 2.3779594898223877,
+      "learning_rate": 0.0002,
+      "loss": 2.5866,
+      "step": 52080
+    },
+    {
+      "epoch": 3.881520119225037,
+      "grad_norm": 2.727888584136963,
+      "learning_rate": 0.0002,
+      "loss": 2.6519,
+      "step": 52090
+    },
+    {
+      "epoch": 3.8822652757078986,
+      "grad_norm": 3.6070122718811035,
+      "learning_rate": 0.0002,
+      "loss": 2.4691,
+      "step": 52100
+    },
+    {
+      "epoch": 3.88301043219076,
+      "grad_norm": 2.8092737197875977,
+      "learning_rate": 0.0002,
+      "loss": 2.5209,
+      "step": 52110
+    },
+    {
+      "epoch": 3.8837555886736217,
+      "grad_norm": 2.367863416671753,
+      "learning_rate": 0.0002,
+      "loss": 2.3449,
+      "step": 52120
+    },
+    {
+      "epoch": 3.884500745156483,
+      "grad_norm": 2.1919851303100586,
+      "learning_rate": 0.0002,
+      "loss": 2.6012,
+      "step": 52130
+    },
+    {
+      "epoch": 3.8852459016393444,
+      "grad_norm": 2.689823865890503,
+      "learning_rate": 0.0002,
+      "loss": 2.5961,
+      "step": 52140
+    },
+    {
+      "epoch": 3.8859910581222055,
+      "grad_norm": 2.8224596977233887,
+      "learning_rate": 0.0002,
+      "loss": 2.4739,
+      "step": 52150
+    },
+    {
+      "epoch": 3.886736214605067,
+      "grad_norm": 2.892871856689453,
+      "learning_rate": 0.0002,
+      "loss": 2.6218,
+      "step": 52160
+    },
+    {
+      "epoch": 3.8874813710879286,
+      "grad_norm": 2.5560922622680664,
+      "learning_rate": 0.0002,
+      "loss": 2.4748,
+      "step": 52170
+    },
+    {
+      "epoch": 3.88822652757079,
+      "grad_norm": 2.412459135055542,
+      "learning_rate": 0.0002,
+      "loss": 2.5312,
+      "step": 52180
+    },
+    {
+      "epoch": 3.8889716840536512,
+      "grad_norm": 2.6065220832824707,
+      "learning_rate": 0.0002,
+      "loss": 2.3348,
+      "step": 52190
+    },
+    {
+      "epoch": 3.889716840536513,
+      "grad_norm": 2.771390199661255,
+      "learning_rate": 0.0002,
+      "loss": 2.4942,
+      "step": 52200
+    },
+    {
+      "epoch": 3.890461997019374,
+      "grad_norm": 2.511683702468872,
+      "learning_rate": 0.0002,
+      "loss": 2.5767,
+      "step": 52210
+    },
+    {
+      "epoch": 3.8912071535022354,
+      "grad_norm": 2.926457643508911,
+      "learning_rate": 0.0002,
+      "loss": 2.4977,
+      "step": 52220
+    },
+    {
+      "epoch": 3.891952309985097,
+      "grad_norm": 2.466374397277832,
+      "learning_rate": 0.0002,
+      "loss": 2.5909,
+      "step": 52230
+    },
+    {
+      "epoch": 3.892697466467958,
+      "grad_norm": 2.454491376876831,
+      "learning_rate": 0.0002,
+      "loss": 2.2894,
+      "step": 52240
+    },
+    {
+      "epoch": 3.8934426229508197,
+      "grad_norm": 2.548574209213257,
+      "learning_rate": 0.0002,
+      "loss": 2.4649,
+      "step": 52250
+    },
+    {
+      "epoch": 3.894187779433681,
+      "grad_norm": 2.4088494777679443,
+      "learning_rate": 0.0002,
+      "loss": 2.1674,
+      "step": 52260
+    },
+    {
+      "epoch": 3.8949329359165423,
+      "grad_norm": 2.7018039226531982,
+      "learning_rate": 0.0002,
+      "loss": 2.5642,
+      "step": 52270
+    },
+    {
+      "epoch": 3.895678092399404,
+      "grad_norm": 2.7482492923736572,
+      "learning_rate": 0.0002,
+      "loss": 2.4304,
+      "step": 52280
+    },
+    {
+      "epoch": 3.8964232488822654,
+      "grad_norm": 2.329071521759033,
+      "learning_rate": 0.0002,
+      "loss": 2.4925,
+      "step": 52290
+    },
+    {
+      "epoch": 3.8971684053651265,
+      "grad_norm": 2.745163917541504,
+      "learning_rate": 0.0002,
+      "loss": 2.465,
+      "step": 52300
+    },
+    {
+      "epoch": 3.897913561847988,
+      "grad_norm": 2.767301559448242,
+      "learning_rate": 0.0002,
+      "loss": 2.5761,
+      "step": 52310
+    },
+    {
+      "epoch": 3.898658718330849,
+      "grad_norm": 2.6464600563049316,
+      "learning_rate": 0.0002,
+      "loss": 2.3783,
+      "step": 52320
+    },
+    {
+      "epoch": 3.8994038748137108,
+      "grad_norm": 2.457310199737549,
+      "learning_rate": 0.0002,
+      "loss": 2.0868,
+      "step": 52330
+    },
+    {
+      "epoch": 3.9001490312965723,
+      "grad_norm": 2.3915839195251465,
+      "learning_rate": 0.0002,
+      "loss": 2.4427,
+      "step": 52340
+    },
+    {
+      "epoch": 3.900894187779434,
+      "grad_norm": 2.6058714389801025,
+      "learning_rate": 0.0002,
+      "loss": 2.4721,
+      "step": 52350
+    },
+    {
+      "epoch": 3.901639344262295,
+      "grad_norm": 3.087757110595703,
+      "learning_rate": 0.0002,
+      "loss": 2.567,
+      "step": 52360
+    },
+    {
+      "epoch": 3.9023845007451565,
+      "grad_norm": 2.904249429702759,
+      "learning_rate": 0.0002,
+      "loss": 2.4766,
+      "step": 52370
+    },
+    {
+      "epoch": 3.9031296572280176,
+      "grad_norm": 2.673194169998169,
+      "learning_rate": 0.0002,
+      "loss": 2.6861,
+      "step": 52380
+    },
+    {
+      "epoch": 3.903874813710879,
+      "grad_norm": 2.4433786869049072,
+      "learning_rate": 0.0002,
+      "loss": 2.4555,
+      "step": 52390
+    },
+    {
+      "epoch": 3.9046199701937407,
+      "grad_norm": 2.482530117034912,
+      "learning_rate": 0.0002,
+      "loss": 2.5579,
+      "step": 52400
+    },
+    {
+      "epoch": 3.9053651266766023,
+      "grad_norm": 2.7197022438049316,
+      "learning_rate": 0.0002,
+      "loss": 2.4326,
+      "step": 52410
+    },
+    {
+      "epoch": 3.9061102831594634,
+      "grad_norm": 2.534710645675659,
+      "learning_rate": 0.0002,
+      "loss": 2.7553,
+      "step": 52420
+    },
+    {
+      "epoch": 3.906855439642325,
+      "grad_norm": 2.1356122493743896,
+      "learning_rate": 0.0002,
+      "loss": 2.2721,
+      "step": 52430
+    },
+    {
+      "epoch": 3.907600596125186,
+      "grad_norm": 2.3630571365356445,
+      "learning_rate": 0.0002,
+      "loss": 2.5597,
+      "step": 52440
+    },
+    {
+      "epoch": 3.9083457526080476,
+      "grad_norm": 2.6591246128082275,
+      "learning_rate": 0.0002,
+      "loss": 2.4836,
+      "step": 52450
+    },
+    {
+      "epoch": 3.909090909090909,
+      "grad_norm": 2.511927604675293,
+      "learning_rate": 0.0002,
+      "loss": 2.6249,
+      "step": 52460
+    },
+    {
+      "epoch": 3.9098360655737707,
+      "grad_norm": 2.8676466941833496,
+      "learning_rate": 0.0002,
+      "loss": 2.4936,
+      "step": 52470
+    },
+    {
+      "epoch": 3.910581222056632,
+      "grad_norm": 2.7902708053588867,
+      "learning_rate": 0.0002,
+      "loss": 2.5414,
+      "step": 52480
+    },
+    {
+      "epoch": 3.9113263785394934,
+      "grad_norm": 2.6967098712921143,
+      "learning_rate": 0.0002,
+      "loss": 2.5556,
+      "step": 52490
+    },
+    {
+      "epoch": 3.9120715350223545,
+      "grad_norm": 2.569732904434204,
+      "learning_rate": 0.0002,
+      "loss": 2.4874,
+      "step": 52500
+    },
+    {
+      "epoch": 3.912816691505216,
+      "grad_norm": 2.595181465148926,
+      "learning_rate": 0.0002,
+      "loss": 2.5599,
+      "step": 52510
+    },
+    {
+      "epoch": 3.9135618479880776,
+      "grad_norm": 2.3675436973571777,
+      "learning_rate": 0.0002,
+      "loss": 2.6449,
+      "step": 52520
+    },
+    {
+      "epoch": 3.914307004470939,
+      "grad_norm": 2.709636688232422,
+      "learning_rate": 0.0002,
+      "loss": 2.5443,
+      "step": 52530
+    },
+    {
+      "epoch": 3.9150521609538003,
+      "grad_norm": 2.7258338928222656,
+      "learning_rate": 0.0002,
+      "loss": 2.3612,
+      "step": 52540
+    },
+    {
+      "epoch": 3.915797317436662,
+      "grad_norm": 3.1132118701934814,
+      "learning_rate": 0.0002,
+      "loss": 2.3272,
+      "step": 52550
+    },
+    {
+      "epoch": 3.916542473919523,
+      "grad_norm": 2.628192186355591,
+      "learning_rate": 0.0002,
+      "loss": 2.4815,
+      "step": 52560
+    },
+    {
+      "epoch": 3.9172876304023845,
+      "grad_norm": 2.5047247409820557,
+      "learning_rate": 0.0002,
+      "loss": 2.4923,
+      "step": 52570
+    },
+    {
+      "epoch": 3.918032786885246,
+      "grad_norm": 2.726745843887329,
+      "learning_rate": 0.0002,
+      "loss": 2.4188,
+      "step": 52580
+    },
+    {
+      "epoch": 3.918777943368107,
+      "grad_norm": 2.729163646697998,
+      "learning_rate": 0.0002,
+      "loss": 2.5176,
+      "step": 52590
+    },
+    {
+      "epoch": 3.9195230998509687,
+      "grad_norm": 2.8599231243133545,
+      "learning_rate": 0.0002,
+      "loss": 2.5171,
+      "step": 52600
+    },
+    {
+      "epoch": 3.9202682563338302,
+      "grad_norm": 2.2114243507385254,
+      "learning_rate": 0.0002,
+      "loss": 2.5073,
+      "step": 52610
+    },
+    {
+      "epoch": 3.9210134128166914,
+      "grad_norm": 2.164640426635742,
+      "learning_rate": 0.0002,
+      "loss": 2.6172,
+      "step": 52620
+    },
+    {
+      "epoch": 3.921758569299553,
+      "grad_norm": 2.7725958824157715,
+      "learning_rate": 0.0002,
+      "loss": 2.5268,
+      "step": 52630
+    },
+    {
+      "epoch": 3.9225037257824145,
+      "grad_norm": 2.6083829402923584,
+      "learning_rate": 0.0002,
+      "loss": 2.4921,
+      "step": 52640
+    },
+    {
+      "epoch": 3.9232488822652756,
+      "grad_norm": 2.948575496673584,
+      "learning_rate": 0.0002,
+      "loss": 2.5642,
+      "step": 52650
+    },
+    {
+      "epoch": 3.923994038748137,
+      "grad_norm": 2.494868755340576,
+      "learning_rate": 0.0002,
+      "loss": 2.4473,
+      "step": 52660
+    },
+    {
+      "epoch": 3.9247391952309982,
+      "grad_norm": 3.170365333557129,
+      "learning_rate": 0.0002,
+      "loss": 2.542,
+      "step": 52670
+    },
+    {
+      "epoch": 3.92548435171386,
+      "grad_norm": 2.6471188068389893,
+      "learning_rate": 0.0002,
+      "loss": 2.436,
+      "step": 52680
+    },
+    {
+      "epoch": 3.9262295081967213,
+      "grad_norm": 2.868462562561035,
+      "learning_rate": 0.0002,
+      "loss": 2.4097,
+      "step": 52690
+    },
+    {
+      "epoch": 3.926974664679583,
+      "grad_norm": 2.8815088272094727,
+      "learning_rate": 0.0002,
+      "loss": 2.5023,
+      "step": 52700
+    },
+    {
+      "epoch": 3.927719821162444,
+      "grad_norm": 2.6703245639801025,
+      "learning_rate": 0.0002,
+      "loss": 2.5889,
+      "step": 52710
+    },
+    {
+      "epoch": 3.9284649776453056,
+      "grad_norm": 2.9111456871032715,
+      "learning_rate": 0.0002,
+      "loss": 2.3997,
+      "step": 52720
+    },
+    {
+      "epoch": 3.9292101341281667,
+      "grad_norm": 3.0684051513671875,
+      "learning_rate": 0.0002,
+      "loss": 2.379,
+      "step": 52730
+    },
+    {
+      "epoch": 3.929955290611028,
+      "grad_norm": 2.901998996734619,
+      "learning_rate": 0.0002,
+      "loss": 2.6043,
+      "step": 52740
+    },
+    {
+      "epoch": 3.9307004470938898,
+      "grad_norm": 3.4542715549468994,
+      "learning_rate": 0.0002,
+      "loss": 2.2915,
+      "step": 52750
+    },
+    {
+      "epoch": 3.9314456035767513,
+      "grad_norm": 2.3926374912261963,
+      "learning_rate": 0.0002,
+      "loss": 2.2738,
+      "step": 52760
+    },
+    {
+      "epoch": 3.9321907600596124,
+      "grad_norm": 2.790262222290039,
+      "learning_rate": 0.0002,
+      "loss": 2.4374,
+      "step": 52770
+    },
+    {
+      "epoch": 3.932935916542474,
+      "grad_norm": 3.287041664123535,
+      "learning_rate": 0.0002,
+      "loss": 2.3272,
+      "step": 52780
+    },
+    {
+      "epoch": 3.933681073025335,
+      "grad_norm": 2.6632344722747803,
+      "learning_rate": 0.0002,
+      "loss": 2.619,
+      "step": 52790
+    },
+    {
+      "epoch": 3.9344262295081966,
+      "grad_norm": 2.1901233196258545,
+      "learning_rate": 0.0002,
+      "loss": 2.3979,
+      "step": 52800
+    },
+    {
+      "epoch": 3.935171385991058,
+      "grad_norm": 2.72855544090271,
+      "learning_rate": 0.0002,
+      "loss": 2.5868,
+      "step": 52810
+    },
+    {
+      "epoch": 3.9359165424739198,
+      "grad_norm": 2.598860502243042,
+      "learning_rate": 0.0002,
+      "loss": 2.4699,
+      "step": 52820
+    },
+    {
+      "epoch": 3.936661698956781,
+      "grad_norm": 2.5119919776916504,
+      "learning_rate": 0.0002,
+      "loss": 2.3506,
+      "step": 52830
+    },
+    {
+      "epoch": 3.9374068554396424,
+      "grad_norm": 2.60629940032959,
+      "learning_rate": 0.0002,
+      "loss": 2.5943,
+      "step": 52840
+    },
+    {
+      "epoch": 3.9381520119225035,
+      "grad_norm": 2.4986929893493652,
+      "learning_rate": 0.0002,
+      "loss": 2.4542,
+      "step": 52850
+    },
+    {
+      "epoch": 3.938897168405365,
+      "grad_norm": 2.564370632171631,
+      "learning_rate": 0.0002,
+      "loss": 2.4587,
+      "step": 52860
+    },
+    {
+      "epoch": 3.9396423248882266,
+      "grad_norm": 3.0465457439422607,
+      "learning_rate": 0.0002,
+      "loss": 2.5511,
+      "step": 52870
+    },
+    {
+      "epoch": 3.940387481371088,
+      "grad_norm": 2.802629232406616,
+      "learning_rate": 0.0002,
+      "loss": 2.5935,
+      "step": 52880
+    },
+    {
+      "epoch": 3.9411326378539493,
+      "grad_norm": 2.5224101543426514,
+      "learning_rate": 0.0002,
+      "loss": 2.4176,
+      "step": 52890
+    },
+    {
+      "epoch": 3.941877794336811,
+      "grad_norm": 2.601170301437378,
+      "learning_rate": 0.0002,
+      "loss": 2.4259,
+      "step": 52900
+    },
+    {
+      "epoch": 3.942622950819672,
+      "grad_norm": 2.618849515914917,
+      "learning_rate": 0.0002,
+      "loss": 2.5917,
+      "step": 52910
+    },
+    {
+      "epoch": 3.9433681073025335,
+      "grad_norm": 2.3083279132843018,
+      "learning_rate": 0.0002,
+      "loss": 2.4501,
+      "step": 52920
+    },
+    {
+      "epoch": 3.944113263785395,
+      "grad_norm": 2.783717155456543,
+      "learning_rate": 0.0002,
+      "loss": 2.6085,
+      "step": 52930
+    },
+    {
+      "epoch": 3.944858420268256,
+      "grad_norm": 2.4879956245422363,
+      "learning_rate": 0.0002,
+      "loss": 2.3325,
+      "step": 52940
+    },
+    {
+      "epoch": 3.9456035767511177,
+      "grad_norm": 2.765442132949829,
+      "learning_rate": 0.0002,
+      "loss": 2.5773,
+      "step": 52950
+    },
+    {
+      "epoch": 3.9463487332339793,
+      "grad_norm": 2.7897586822509766,
+      "learning_rate": 0.0002,
+      "loss": 2.4599,
+      "step": 52960
+    },
+    {
+      "epoch": 3.9470938897168404,
+      "grad_norm": 2.905817747116089,
+      "learning_rate": 0.0002,
+      "loss": 2.475,
+      "step": 52970
+    },
+    {
+      "epoch": 3.947839046199702,
+      "grad_norm": 2.8461194038391113,
+      "learning_rate": 0.0002,
+      "loss": 2.6288,
+      "step": 52980
+    },
+    {
+      "epoch": 3.9485842026825635,
+      "grad_norm": 2.5353071689605713,
+      "learning_rate": 0.0002,
+      "loss": 2.4977,
+      "step": 52990
+    },
+    {
+      "epoch": 3.9493293591654246,
+      "grad_norm": 2.5417919158935547,
+      "learning_rate": 0.0002,
+      "loss": 2.4847,
+      "step": 53000
+    },
+    {
+      "epoch": 3.950074515648286,
+      "grad_norm": 2.5590386390686035,
+      "learning_rate": 0.0002,
+      "loss": 2.4574,
+      "step": 53010
+    },
+    {
+      "epoch": 3.9508196721311473,
+      "grad_norm": 2.5833194255828857,
+      "learning_rate": 0.0002,
+      "loss": 2.1748,
+      "step": 53020
+    },
+    {
+      "epoch": 3.951564828614009,
+      "grad_norm": 2.8139731884002686,
+      "learning_rate": 0.0002,
+      "loss": 2.5048,
+      "step": 53030
+    },
+    {
+      "epoch": 3.9523099850968704,
+      "grad_norm": 2.520247459411621,
+      "learning_rate": 0.0002,
+      "loss": 2.5423,
+      "step": 53040
+    },
+    {
+      "epoch": 3.953055141579732,
+      "grad_norm": 2.8328166007995605,
+      "learning_rate": 0.0002,
+      "loss": 2.7476,
+      "step": 53050
+    },
+    {
+      "epoch": 3.953800298062593,
+      "grad_norm": 2.491218328475952,
+      "learning_rate": 0.0002,
+      "loss": 2.5617,
+      "step": 53060
+    },
+    {
+      "epoch": 3.9545454545454546,
+      "grad_norm": 2.5774359703063965,
+      "learning_rate": 0.0002,
+      "loss": 2.4691,
+      "step": 53070
+    },
+    {
+      "epoch": 3.9552906110283157,
+      "grad_norm": 2.8533051013946533,
+      "learning_rate": 0.0002,
+      "loss": 2.7179,
+      "step": 53080
+    },
+    {
+      "epoch": 3.9560357675111772,
+      "grad_norm": 2.5596423149108887,
+      "learning_rate": 0.0002,
+      "loss": 2.5002,
+      "step": 53090
+    },
+    {
+      "epoch": 3.956780923994039,
+      "grad_norm": 2.4977810382843018,
+      "learning_rate": 0.0002,
+      "loss": 2.5137,
+      "step": 53100
+    },
+    {
+      "epoch": 3.9575260804769004,
+      "grad_norm": 2.417628526687622,
+      "learning_rate": 0.0002,
+      "loss": 2.4268,
+      "step": 53110
+    },
+    {
+      "epoch": 3.9582712369597615,
+      "grad_norm": 2.7879631519317627,
+      "learning_rate": 0.0002,
+      "loss": 2.6652,
+      "step": 53120
+    },
+    {
+      "epoch": 3.959016393442623,
+      "grad_norm": 2.732900857925415,
+      "learning_rate": 0.0002,
+      "loss": 2.4687,
+      "step": 53130
+    },
+    {
+      "epoch": 3.959761549925484,
+      "grad_norm": 2.5908401012420654,
+      "learning_rate": 0.0002,
+      "loss": 2.4273,
+      "step": 53140
+    },
+    {
+      "epoch": 3.9605067064083457,
+      "grad_norm": 3.021219253540039,
+      "learning_rate": 0.0002,
+      "loss": 2.4994,
+      "step": 53150
+    },
+    {
+      "epoch": 3.9612518628912072,
+      "grad_norm": 2.95470929145813,
+      "learning_rate": 0.0002,
+      "loss": 2.4703,
+      "step": 53160
+    },
+    {
+      "epoch": 3.961997019374069,
+      "grad_norm": 2.4512953758239746,
+      "learning_rate": 0.0002,
+      "loss": 2.5641,
+      "step": 53170
+    },
+    {
+      "epoch": 3.96274217585693,
+      "grad_norm": 2.5563712120056152,
+      "learning_rate": 0.0002,
+      "loss": 2.5442,
+      "step": 53180
+    },
+    {
+      "epoch": 3.9634873323397914,
+      "grad_norm": 2.6706528663635254,
+      "learning_rate": 0.0002,
+      "loss": 2.5561,
+      "step": 53190
+    },
+    {
+      "epoch": 3.9642324888226526,
+      "grad_norm": 2.7212600708007812,
+      "learning_rate": 0.0002,
+      "loss": 2.4722,
+      "step": 53200
+    },
+    {
+      "epoch": 3.964977645305514,
+      "grad_norm": 2.527886390686035,
+      "learning_rate": 0.0002,
+      "loss": 2.4376,
+      "step": 53210
+    },
+    {
+      "epoch": 3.9657228017883757,
+      "grad_norm": 2.4174673557281494,
+      "learning_rate": 0.0002,
+      "loss": 2.5337,
+      "step": 53220
+    },
+    {
+      "epoch": 3.966467958271237,
+      "grad_norm": 3.1026086807250977,
+      "learning_rate": 0.0002,
+      "loss": 2.4212,
+      "step": 53230
+    },
+    {
+      "epoch": 3.9672131147540983,
+      "grad_norm": 2.696936845779419,
+      "learning_rate": 0.0002,
+      "loss": 2.4115,
+      "step": 53240
+    },
+    {
+      "epoch": 3.96795827123696,
+      "grad_norm": 2.7065093517303467,
+      "learning_rate": 0.0002,
+      "loss": 2.575,
+      "step": 53250
+    },
+    {
+      "epoch": 3.968703427719821,
+      "grad_norm": 2.599903106689453,
+      "learning_rate": 0.0002,
+      "loss": 2.4136,
+      "step": 53260
+    },
+    {
+      "epoch": 3.9694485842026825,
+      "grad_norm": 2.7001261711120605,
+      "learning_rate": 0.0002,
+      "loss": 2.4217,
+      "step": 53270
+    },
+    {
+      "epoch": 3.970193740685544,
+      "grad_norm": 2.7080583572387695,
+      "learning_rate": 0.0002,
+      "loss": 2.6516,
+      "step": 53280
+    },
+    {
+      "epoch": 3.970938897168405,
+      "grad_norm": 2.549535036087036,
+      "learning_rate": 0.0002,
+      "loss": 2.4606,
+      "step": 53290
+    },
+    {
+      "epoch": 3.9716840536512668,
+      "grad_norm": 3.246875047683716,
+      "learning_rate": 0.0002,
+      "loss": 2.6515,
+      "step": 53300
+    },
+    {
+      "epoch": 3.9724292101341283,
+      "grad_norm": 2.5322539806365967,
+      "learning_rate": 0.0002,
+      "loss": 2.4486,
+      "step": 53310
+    },
+    {
+      "epoch": 3.9731743666169894,
+      "grad_norm": 2.2940611839294434,
+      "learning_rate": 0.0002,
+      "loss": 2.358,
+      "step": 53320
+    },
+    {
+      "epoch": 3.973919523099851,
+      "grad_norm": 2.6890695095062256,
+      "learning_rate": 0.0002,
+      "loss": 2.5776,
+      "step": 53330
+    },
+    {
+      "epoch": 3.9746646795827125,
+      "grad_norm": 2.7836878299713135,
+      "learning_rate": 0.0002,
+      "loss": 2.3885,
+      "step": 53340
+    },
+    {
+      "epoch": 3.9754098360655736,
+      "grad_norm": 2.572012186050415,
+      "learning_rate": 0.0002,
+      "loss": 2.3641,
+      "step": 53350
+    },
+    {
+      "epoch": 3.976154992548435,
+      "grad_norm": 2.936228036880493,
+      "learning_rate": 0.0002,
+      "loss": 2.4106,
+      "step": 53360
+    },
+    {
+      "epoch": 3.9769001490312967,
+      "grad_norm": 2.7177422046661377,
+      "learning_rate": 0.0002,
+      "loss": 2.6147,
+      "step": 53370
+    },
+    {
+      "epoch": 3.977645305514158,
+      "grad_norm": 2.8363125324249268,
+      "learning_rate": 0.0002,
+      "loss": 2.5204,
+      "step": 53380
+    },
+    {
+      "epoch": 3.9783904619970194,
+      "grad_norm": 2.6096811294555664,
+      "learning_rate": 0.0002,
+      "loss": 2.3503,
+      "step": 53390
+    },
+    {
+      "epoch": 3.979135618479881,
+      "grad_norm": 2.363093852996826,
+      "learning_rate": 0.0002,
+      "loss": 2.7152,
+      "step": 53400
+    },
+    {
+      "epoch": 3.979880774962742,
+      "grad_norm": 2.8034679889678955,
+      "learning_rate": 0.0002,
+      "loss": 2.5279,
+      "step": 53410
+    },
+    {
+      "epoch": 3.9806259314456036,
+      "grad_norm": 2.614542245864868,
+      "learning_rate": 0.0002,
+      "loss": 2.429,
+      "step": 53420
+    },
+    {
+      "epoch": 3.9813710879284647,
+      "grad_norm": 2.7522165775299072,
+      "learning_rate": 0.0002,
+      "loss": 2.547,
+      "step": 53430
+    },
+    {
+      "epoch": 3.9821162444113263,
+      "grad_norm": 2.595388412475586,
+      "learning_rate": 0.0002,
+      "loss": 2.566,
+      "step": 53440
+    },
+    {
+      "epoch": 3.982861400894188,
+      "grad_norm": 2.463608741760254,
+      "learning_rate": 0.0002,
+      "loss": 2.4284,
+      "step": 53450
+    },
+    {
+      "epoch": 3.9836065573770494,
+      "grad_norm": 2.377199172973633,
+      "learning_rate": 0.0002,
+      "loss": 2.5547,
+      "step": 53460
+    },
+    {
+      "epoch": 3.9843517138599105,
+      "grad_norm": 2.985874652862549,
+      "learning_rate": 0.0002,
+      "loss": 2.5976,
+      "step": 53470
+    },
+    {
+      "epoch": 3.985096870342772,
+      "grad_norm": 2.601820707321167,
+      "learning_rate": 0.0002,
+      "loss": 2.5576,
+      "step": 53480
+    },
+    {
+      "epoch": 3.985842026825633,
+      "grad_norm": 3.1383955478668213,
+      "learning_rate": 0.0002,
+      "loss": 2.4143,
+      "step": 53490
+    },
+    {
+      "epoch": 3.9865871833084947,
+      "grad_norm": 2.46256685256958,
+      "learning_rate": 0.0002,
+      "loss": 2.6677,
+      "step": 53500
+    },
+    {
+      "epoch": 3.9873323397913563,
+      "grad_norm": 2.43536376953125,
+      "learning_rate": 0.0002,
+      "loss": 2.526,
+      "step": 53510
+    },
+    {
+      "epoch": 3.988077496274218,
+      "grad_norm": 2.3847920894622803,
+      "learning_rate": 0.0002,
+      "loss": 2.6184,
+      "step": 53520
+    },
+    {
+      "epoch": 3.988822652757079,
+      "grad_norm": 2.5792295932769775,
+      "learning_rate": 0.0002,
+      "loss": 2.6166,
+      "step": 53530
+    },
+    {
+      "epoch": 3.9895678092399405,
+      "grad_norm": 2.792299509048462,
+      "learning_rate": 0.0002,
+      "loss": 2.6626,
+      "step": 53540
+    },
+    {
+      "epoch": 3.9903129657228016,
+      "grad_norm": 2.445161819458008,
+      "learning_rate": 0.0002,
+      "loss": 2.5005,
+      "step": 53550
+    },
+    {
+      "epoch": 3.991058122205663,
+      "grad_norm": 3.263998031616211,
+      "learning_rate": 0.0002,
+      "loss": 2.5312,
+      "step": 53560
+    },
+    {
+      "epoch": 3.9918032786885247,
+      "grad_norm": 2.557817220687866,
+      "learning_rate": 0.0002,
+      "loss": 2.503,
+      "step": 53570
+    },
+    {
+      "epoch": 3.9925484351713862,
+      "grad_norm": 2.727743625640869,
+      "learning_rate": 0.0002,
+      "loss": 2.6169,
+      "step": 53580
+    },
+    {
+      "epoch": 3.9932935916542474,
+      "grad_norm": 2.6156795024871826,
+      "learning_rate": 0.0002,
+      "loss": 2.5591,
+      "step": 53590
+    },
+    {
+      "epoch": 3.994038748137109,
+      "grad_norm": 2.661540985107422,
+      "learning_rate": 0.0002,
+      "loss": 2.5086,
+      "step": 53600
+    },
+    {
+      "epoch": 3.99478390461997,
+      "grad_norm": 2.503107786178589,
+      "learning_rate": 0.0002,
+      "loss": 2.5765,
+      "step": 53610
+    },
+    {
+      "epoch": 3.9955290611028316,
+      "grad_norm": 2.6578187942504883,
+      "learning_rate": 0.0002,
+      "loss": 2.4175,
+      "step": 53620
+    },
+    {
+      "epoch": 3.996274217585693,
+      "grad_norm": 2.996702194213867,
+      "learning_rate": 0.0002,
+      "loss": 2.694,
+      "step": 53630
+    },
+    {
+      "epoch": 3.9970193740685542,
+      "grad_norm": 2.9681408405303955,
+      "learning_rate": 0.0002,
+      "loss": 2.4948,
+      "step": 53640
+    },
+    {
+      "epoch": 3.997764530551416,
+      "grad_norm": 2.7573134899139404,
+      "learning_rate": 0.0002,
+      "loss": 2.6957,
+      "step": 53650
+    },
+    {
+      "epoch": 3.9985096870342773,
+      "grad_norm": 2.6993305683135986,
+      "learning_rate": 0.0002,
+      "loss": 2.588,
+      "step": 53660
+    },
+    {
+      "epoch": 3.9992548435171384,
+      "grad_norm": 2.6153762340545654,
+      "learning_rate": 0.0002,
+      "loss": 2.5999,
+      "step": 53670
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.1838414669036865,
+      "learning_rate": 0.0002,
+      "loss": 2.4747,
+      "step": 53680
+    },
+    {
+      "epoch": 4.0,
+      "eval_runtime": 2866.1708,
+      "eval_samples_per_second": 4.682,
+      "eval_steps_per_second": 0.585,
+      "step": 53680
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 80520,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "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": 4.734753409106688e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}