diff --git "a/final/trainer_state.json" "b/final/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/final/trainer_state.json"
@@ -0,0 +1,67431 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 9627,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010387451958034694,
+      "grad_norm": 1.3366456031799316,
+      "learning_rate": 9.9999997337695e-05,
+      "loss": 2.4557,
+      "step": 1
+    },
+    {
+      "epoch": 0.00020774903916069389,
+      "grad_norm": 0.7125677466392517,
+      "learning_rate": 9.999998935078033e-05,
+      "loss": 2.1366,
+      "step": 2
+    },
+    {
+      "epoch": 0.0003116235587410408,
+      "grad_norm": 0.5620805621147156,
+      "learning_rate": 9.999997603925677e-05,
+      "loss": 1.9955,
+      "step": 3
+    },
+    {
+      "epoch": 0.00041549807832138777,
+      "grad_norm": 0.44992074370384216,
+      "learning_rate": 9.99999574031258e-05,
+      "loss": 2.1548,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005193725979017347,
+      "grad_norm": 0.4193553626537323,
+      "learning_rate": 9.999993344238936e-05,
+      "loss": 2.0443,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006232471174820816,
+      "grad_norm": 0.4412941336631775,
+      "learning_rate": 9.999990415705005e-05,
+      "loss": 2.0479,
+      "step": 6
+    },
+    {
+      "epoch": 0.0007271216370624286,
+      "grad_norm": 0.40010932087898254,
+      "learning_rate": 9.999986954711094e-05,
+      "loss": 2.0064,
+      "step": 7
+    },
+    {
+      "epoch": 0.0008309961566427755,
+      "grad_norm": 0.4279223680496216,
+      "learning_rate": 9.999982961257574e-05,
+      "loss": 1.8892,
+      "step": 8
+    },
+    {
+      "epoch": 0.0009348706762231225,
+      "grad_norm": 0.4031948149204254,
+      "learning_rate": 9.999978435344872e-05,
+      "loss": 1.9309,
+      "step": 9
+    },
+    {
+      "epoch": 0.0010387451958034693,
+      "grad_norm": 0.34114745259284973,
+      "learning_rate": 9.999973376973466e-05,
+      "loss": 1.6737,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011426197153838163,
+      "grad_norm": 0.4245925545692444,
+      "learning_rate": 9.999967786143895e-05,
+      "loss": 1.8743,
+      "step": 11
+    },
+    {
+      "epoch": 0.0012464942349641633,
+      "grad_norm": 0.4021143317222595,
+      "learning_rate": 9.99996166285676e-05,
+      "loss": 1.8595,
+      "step": 12
+    },
+    {
+      "epoch": 0.0013503687545445102,
+      "grad_norm": 0.38615134358406067,
+      "learning_rate": 9.999955007112706e-05,
+      "loss": 1.8856,
+      "step": 13
+    },
+    {
+      "epoch": 0.0014542432741248572,
+      "grad_norm": 0.3954271078109741,
+      "learning_rate": 9.999947818912448e-05,
+      "loss": 2.0519,
+      "step": 14
+    },
+    {
+      "epoch": 0.0015581177937052041,
+      "grad_norm": 0.3857433795928955,
+      "learning_rate": 9.999940098256747e-05,
+      "loss": 1.8855,
+      "step": 15
+    },
+    {
+      "epoch": 0.001661992313285551,
+      "grad_norm": 0.38353338837623596,
+      "learning_rate": 9.999931845146424e-05,
+      "loss": 2.0608,
+      "step": 16
+    },
+    {
+      "epoch": 0.001765866832865898,
+      "grad_norm": 0.46258458495140076,
+      "learning_rate": 9.999923059582363e-05,
+      "loss": 2.2282,
+      "step": 17
+    },
+    {
+      "epoch": 0.001869741352446245,
+      "grad_norm": 0.40921229124069214,
+      "learning_rate": 9.999913741565496e-05,
+      "loss": 2.0698,
+      "step": 18
+    },
+    {
+      "epoch": 0.0019736158720265917,
+      "grad_norm": 0.39826393127441406,
+      "learning_rate": 9.999903891096817e-05,
+      "loss": 2.0232,
+      "step": 19
+    },
+    {
+      "epoch": 0.0020774903916069387,
+      "grad_norm": 0.4504700005054474,
+      "learning_rate": 9.999893508177373e-05,
+      "loss": 1.8005,
+      "step": 20
+    },
+    {
+      "epoch": 0.0021813649111872857,
+      "grad_norm": 0.35965996980667114,
+      "learning_rate": 9.999882592808274e-05,
+      "loss": 1.8358,
+      "step": 21
+    },
+    {
+      "epoch": 0.0022852394307676326,
+      "grad_norm": 0.36941030621528625,
+      "learning_rate": 9.999871144990677e-05,
+      "loss": 1.8202,
+      "step": 22
+    },
+    {
+      "epoch": 0.0023891139503479796,
+      "grad_norm": 0.3631148338317871,
+      "learning_rate": 9.999859164725804e-05,
+      "loss": 1.9011,
+      "step": 23
+    },
+    {
+      "epoch": 0.0024929884699283265,
+      "grad_norm": 0.3719225525856018,
+      "learning_rate": 9.999846652014931e-05,
+      "loss": 1.8046,
+      "step": 24
+    },
+    {
+      "epoch": 0.0025968629895086735,
+      "grad_norm": 0.3886321783065796,
+      "learning_rate": 9.99983360685939e-05,
+      "loss": 1.7041,
+      "step": 25
+    },
+    {
+      "epoch": 0.0027007375090890204,
+      "grad_norm": 0.35539302229881287,
+      "learning_rate": 9.999820029260569e-05,
+      "loss": 1.8635,
+      "step": 26
+    },
+    {
+      "epoch": 0.0028046120286693674,
+      "grad_norm": 0.40370672941207886,
+      "learning_rate": 9.999805919219916e-05,
+      "loss": 2.025,
+      "step": 27
+    },
+    {
+      "epoch": 0.0029084865482497143,
+      "grad_norm": 0.36346563696861267,
+      "learning_rate": 9.999791276738933e-05,
+      "loss": 1.9266,
+      "step": 28
+    },
+    {
+      "epoch": 0.0030123610678300613,
+      "grad_norm": 0.4094865322113037,
+      "learning_rate": 9.999776101819177e-05,
+      "loss": 1.923,
+      "step": 29
+    },
+    {
+      "epoch": 0.0031162355874104083,
+      "grad_norm": 0.33642393350601196,
+      "learning_rate": 9.999760394462268e-05,
+      "loss": 1.876,
+      "step": 30
+    },
+    {
+      "epoch": 0.0032201101069907552,
+      "grad_norm": 0.386692613363266,
+      "learning_rate": 9.999744154669874e-05,
+      "loss": 2.0935,
+      "step": 31
+    },
+    {
+      "epoch": 0.003323984626571102,
+      "grad_norm": 0.32473698258399963,
+      "learning_rate": 9.999727382443728e-05,
+      "loss": 1.7944,
+      "step": 32
+    },
+    {
+      "epoch": 0.003427859146151449,
+      "grad_norm": 0.3738991916179657,
+      "learning_rate": 9.999710077785616e-05,
+      "loss": 2.0774,
+      "step": 33
+    },
+    {
+      "epoch": 0.003531733665731796,
+      "grad_norm": 0.3248012959957123,
+      "learning_rate": 9.99969224069738e-05,
+      "loss": 1.7769,
+      "step": 34
+    },
+    {
+      "epoch": 0.003635608185312143,
+      "grad_norm": 0.3610652685165405,
+      "learning_rate": 9.99967387118092e-05,
+      "loss": 1.9222,
+      "step": 35
+    },
+    {
+      "epoch": 0.00373948270489249,
+      "grad_norm": 0.3891034722328186,
+      "learning_rate": 9.99965496923819e-05,
+      "loss": 2.0162,
+      "step": 36
+    },
+    {
+      "epoch": 0.003843357224472837,
+      "grad_norm": 0.37236684560775757,
+      "learning_rate": 9.999635534871205e-05,
+      "loss": 2.0374,
+      "step": 37
+    },
+    {
+      "epoch": 0.0039472317440531835,
+      "grad_norm": 0.3498350977897644,
+      "learning_rate": 9.999615568082036e-05,
+      "loss": 1.7788,
+      "step": 38
+    },
+    {
+      "epoch": 0.0040511062636335304,
+      "grad_norm": 0.41840648651123047,
+      "learning_rate": 9.999595068872805e-05,
+      "loss": 1.6986,
+      "step": 39
+    },
+    {
+      "epoch": 0.004154980783213877,
+      "grad_norm": 0.3720596134662628,
+      "learning_rate": 9.9995740372457e-05,
+      "loss": 1.9754,
+      "step": 40
+    },
+    {
+      "epoch": 0.004258855302794224,
+      "grad_norm": 0.32730332016944885,
+      "learning_rate": 9.999552473202955e-05,
+      "loss": 1.699,
+      "step": 41
+    },
+    {
+      "epoch": 0.004362729822374571,
+      "grad_norm": 0.36480000615119934,
+      "learning_rate": 9.999530376746873e-05,
+      "loss": 2.1077,
+      "step": 42
+    },
+    {
+      "epoch": 0.004466604341954918,
+      "grad_norm": 0.3458911180496216,
+      "learning_rate": 9.999507747879802e-05,
+      "loss": 1.9693,
+      "step": 43
+    },
+    {
+      "epoch": 0.004570478861535265,
+      "grad_norm": 0.34066182374954224,
+      "learning_rate": 9.999484586604154e-05,
+      "loss": 1.9005,
+      "step": 44
+    },
+    {
+      "epoch": 0.004674353381115612,
+      "grad_norm": 0.34163591265678406,
+      "learning_rate": 9.999460892922394e-05,
+      "loss": 1.8829,
+      "step": 45
+    },
+    {
+      "epoch": 0.004778227900695959,
+      "grad_norm": 0.37044212222099304,
+      "learning_rate": 9.999436666837048e-05,
+      "loss": 1.9172,
+      "step": 46
+    },
+    {
+      "epoch": 0.004882102420276306,
+      "grad_norm": 0.3332674205303192,
+      "learning_rate": 9.999411908350692e-05,
+      "loss": 1.8282,
+      "step": 47
+    },
+    {
+      "epoch": 0.004985976939856653,
+      "grad_norm": 0.3461020886898041,
+      "learning_rate": 9.999386617465966e-05,
+      "loss": 1.8652,
+      "step": 48
+    },
+    {
+      "epoch": 0.005089851459437,
+      "grad_norm": 0.31899985671043396,
+      "learning_rate": 9.999360794185562e-05,
+      "loss": 1.9479,
+      "step": 49
+    },
+    {
+      "epoch": 0.005193725979017347,
+      "grad_norm": 0.3465663194656372,
+      "learning_rate": 9.999334438512228e-05,
+      "loss": 1.8856,
+      "step": 50
+    },
+    {
+      "epoch": 0.005297600498597694,
+      "grad_norm": 0.4567152261734009,
+      "learning_rate": 9.999307550448776e-05,
+      "loss": 2.26,
+      "step": 51
+    },
+    {
+      "epoch": 0.005401475018178041,
+      "grad_norm": 0.36369699239730835,
+      "learning_rate": 9.999280129998063e-05,
+      "loss": 1.9131,
+      "step": 52
+    },
+    {
+      "epoch": 0.005505349537758388,
+      "grad_norm": 0.3090951442718506,
+      "learning_rate": 9.999252177163013e-05,
+      "loss": 1.7429,
+      "step": 53
+    },
+    {
+      "epoch": 0.005609224057338735,
+      "grad_norm": 0.3246748745441437,
+      "learning_rate": 9.999223691946602e-05,
+      "loss": 1.8882,
+      "step": 54
+    },
+    {
+      "epoch": 0.005713098576919082,
+      "grad_norm": 0.3664419949054718,
+      "learning_rate": 9.999194674351864e-05,
+      "loss": 2.2768,
+      "step": 55
+    },
+    {
+      "epoch": 0.005816973096499429,
+      "grad_norm": 0.3436775505542755,
+      "learning_rate": 9.999165124381887e-05,
+      "loss": 1.7684,
+      "step": 56
+    },
+    {
+      "epoch": 0.005920847616079776,
+      "grad_norm": 0.3558364510536194,
+      "learning_rate": 9.99913504203982e-05,
+      "loss": 1.8547,
+      "step": 57
+    },
+    {
+      "epoch": 0.006024722135660123,
+      "grad_norm": 0.3412851393222809,
+      "learning_rate": 9.999104427328865e-05,
+      "loss": 1.9773,
+      "step": 58
+    },
+    {
+      "epoch": 0.0061285966552404696,
+      "grad_norm": 0.3379881978034973,
+      "learning_rate": 9.999073280252283e-05,
+      "loss": 1.9732,
+      "step": 59
+    },
+    {
+      "epoch": 0.0062324711748208165,
+      "grad_norm": 0.38533058762550354,
+      "learning_rate": 9.999041600813393e-05,
+      "loss": 1.9285,
+      "step": 60
+    },
+    {
+      "epoch": 0.0063363456944011635,
+      "grad_norm": 0.3420720398426056,
+      "learning_rate": 9.999009389015564e-05,
+      "loss": 1.9619,
+      "step": 61
+    },
+    {
+      "epoch": 0.0064402202139815104,
+      "grad_norm": 0.3268018662929535,
+      "learning_rate": 9.99897664486223e-05,
+      "loss": 2.0373,
+      "step": 62
+    },
+    {
+      "epoch": 0.006544094733561857,
+      "grad_norm": 0.32013919949531555,
+      "learning_rate": 9.998943368356877e-05,
+      "loss": 1.8389,
+      "step": 63
+    },
+    {
+      "epoch": 0.006647969253142204,
+      "grad_norm": 0.38457608222961426,
+      "learning_rate": 9.998909559503048e-05,
+      "loss": 2.0334,
+      "step": 64
+    },
+    {
+      "epoch": 0.006751843772722551,
+      "grad_norm": 0.35168957710266113,
+      "learning_rate": 9.998875218304345e-05,
+      "loss": 2.0735,
+      "step": 65
+    },
+    {
+      "epoch": 0.006855718292302898,
+      "grad_norm": 0.3252507746219635,
+      "learning_rate": 9.998840344764422e-05,
+      "loss": 1.7135,
+      "step": 66
+    },
+    {
+      "epoch": 0.006959592811883245,
+      "grad_norm": 0.353354275226593,
+      "learning_rate": 9.998804938886997e-05,
+      "loss": 1.9234,
+      "step": 67
+    },
+    {
+      "epoch": 0.007063467331463592,
+      "grad_norm": 0.3150123357772827,
+      "learning_rate": 9.998769000675836e-05,
+      "loss": 1.6909,
+      "step": 68
+    },
+    {
+      "epoch": 0.007167341851043939,
+      "grad_norm": 0.33036699891090393,
+      "learning_rate": 9.99873253013477e-05,
+      "loss": 1.7841,
+      "step": 69
+    },
+    {
+      "epoch": 0.007271216370624286,
+      "grad_norm": 0.3279257118701935,
+      "learning_rate": 9.998695527267681e-05,
+      "loss": 1.8755,
+      "step": 70
+    },
+    {
+      "epoch": 0.007375090890204633,
+      "grad_norm": 0.3096674978733063,
+      "learning_rate": 9.998657992078509e-05,
+      "loss": 1.8373,
+      "step": 71
+    },
+    {
+      "epoch": 0.00747896540978498,
+      "grad_norm": 0.3656545877456665,
+      "learning_rate": 9.998619924571251e-05,
+      "loss": 2.0858,
+      "step": 72
+    },
+    {
+      "epoch": 0.007582839929365327,
+      "grad_norm": 0.31613457202911377,
+      "learning_rate": 9.998581324749964e-05,
+      "loss": 1.7906,
+      "step": 73
+    },
+    {
+      "epoch": 0.007686714448945674,
+      "grad_norm": 0.3255898356437683,
+      "learning_rate": 9.998542192618755e-05,
+      "loss": 1.8037,
+      "step": 74
+    },
+    {
+      "epoch": 0.007790588968526021,
+      "grad_norm": 0.33904996514320374,
+      "learning_rate": 9.998502528181793e-05,
+      "loss": 1.6548,
+      "step": 75
+    },
+    {
+      "epoch": 0.007894463488106367,
+      "grad_norm": 0.3505510985851288,
+      "learning_rate": 9.998462331443301e-05,
+      "loss": 2.0898,
+      "step": 76
+    },
+    {
+      "epoch": 0.007998338007686715,
+      "grad_norm": 0.3301371932029724,
+      "learning_rate": 9.99842160240756e-05,
+      "loss": 1.9709,
+      "step": 77
+    },
+    {
+      "epoch": 0.008102212527267061,
+      "grad_norm": 0.3287624418735504,
+      "learning_rate": 9.998380341078909e-05,
+      "loss": 1.6921,
+      "step": 78
+    },
+    {
+      "epoch": 0.008206087046847409,
+      "grad_norm": 0.3255762755870819,
+      "learning_rate": 9.99833854746174e-05,
+      "loss": 1.8009,
+      "step": 79
+    },
+    {
+      "epoch": 0.008309961566427755,
+      "grad_norm": 0.3210470378398895,
+      "learning_rate": 9.998296221560504e-05,
+      "loss": 1.9141,
+      "step": 80
+    },
+    {
+      "epoch": 0.008413836086008103,
+      "grad_norm": 0.3226737082004547,
+      "learning_rate": 9.998253363379708e-05,
+      "loss": 1.8365,
+      "step": 81
+    },
+    {
+      "epoch": 0.008517710605588449,
+      "grad_norm": 0.3540826439857483,
+      "learning_rate": 9.998209972923918e-05,
+      "loss": 2.0948,
+      "step": 82
+    },
+    {
+      "epoch": 0.008621585125168797,
+      "grad_norm": 0.3137127161026001,
+      "learning_rate": 9.998166050197751e-05,
+      "loss": 1.8404,
+      "step": 83
+    },
+    {
+      "epoch": 0.008725459644749143,
+      "grad_norm": 0.3495839536190033,
+      "learning_rate": 9.998121595205889e-05,
+      "loss": 1.8201,
+      "step": 84
+    },
+    {
+      "epoch": 0.00882933416432949,
+      "grad_norm": 0.3384789824485779,
+      "learning_rate": 9.998076607953064e-05,
+      "loss": 1.9134,
+      "step": 85
+    },
+    {
+      "epoch": 0.008933208683909837,
+      "grad_norm": 0.3618725836277008,
+      "learning_rate": 9.998031088444067e-05,
+      "loss": 1.9558,
+      "step": 86
+    },
+    {
+      "epoch": 0.009037083203490184,
+      "grad_norm": 0.370078444480896,
+      "learning_rate": 9.997985036683744e-05,
+      "loss": 2.045,
+      "step": 87
+    },
+    {
+      "epoch": 0.00914095772307053,
+      "grad_norm": 0.32593002915382385,
+      "learning_rate": 9.997938452677e-05,
+      "loss": 1.7929,
+      "step": 88
+    },
+    {
+      "epoch": 0.009244832242650878,
+      "grad_norm": 0.3358675241470337,
+      "learning_rate": 9.997891336428797e-05,
+      "loss": 1.8742,
+      "step": 89
+    },
+    {
+      "epoch": 0.009348706762231224,
+      "grad_norm": 0.35875624418258667,
+      "learning_rate": 9.997843687944152e-05,
+      "loss": 1.9634,
+      "step": 90
+    },
+    {
+      "epoch": 0.009452581281811572,
+      "grad_norm": 0.3237409293651581,
+      "learning_rate": 9.997795507228139e-05,
+      "loss": 1.8137,
+      "step": 91
+    },
+    {
+      "epoch": 0.009556455801391918,
+      "grad_norm": 0.3336571455001831,
+      "learning_rate": 9.997746794285887e-05,
+      "loss": 1.9342,
+      "step": 92
+    },
+    {
+      "epoch": 0.009660330320972266,
+      "grad_norm": 0.3358410894870758,
+      "learning_rate": 9.997697549122586e-05,
+      "loss": 1.8772,
+      "step": 93
+    },
+    {
+      "epoch": 0.009764204840552612,
+      "grad_norm": 0.3093462884426117,
+      "learning_rate": 9.997647771743482e-05,
+      "loss": 1.8048,
+      "step": 94
+    },
+    {
+      "epoch": 0.00986807936013296,
+      "grad_norm": 0.34004223346710205,
+      "learning_rate": 9.997597462153871e-05,
+      "loss": 1.8228,
+      "step": 95
+    },
+    {
+      "epoch": 0.009971953879713306,
+      "grad_norm": 0.33898910880088806,
+      "learning_rate": 9.997546620359114e-05,
+      "loss": 1.8203,
+      "step": 96
+    },
+    {
+      "epoch": 0.010075828399293654,
+      "grad_norm": 0.3222779929637909,
+      "learning_rate": 9.997495246364624e-05,
+      "loss": 1.7667,
+      "step": 97
+    },
+    {
+      "epoch": 0.010179702918874,
+      "grad_norm": 0.3132547438144684,
+      "learning_rate": 9.997443340175872e-05,
+      "loss": 1.7176,
+      "step": 98
+    },
+    {
+      "epoch": 0.010283577438454348,
+      "grad_norm": 0.30829909443855286,
+      "learning_rate": 9.997390901798386e-05,
+      "loss": 1.7196,
+      "step": 99
+    },
+    {
+      "epoch": 0.010387451958034694,
+      "grad_norm": 0.3239877223968506,
+      "learning_rate": 9.997337931237751e-05,
+      "loss": 1.6993,
+      "step": 100
+    },
+    {
+      "epoch": 0.010491326477615042,
+      "grad_norm": 0.34558477997779846,
+      "learning_rate": 9.997284428499605e-05,
+      "loss": 1.9082,
+      "step": 101
+    },
+    {
+      "epoch": 0.010595200997195388,
+      "grad_norm": 0.34653279185295105,
+      "learning_rate": 9.99723039358965e-05,
+      "loss": 1.8484,
+      "step": 102
+    },
+    {
+      "epoch": 0.010699075516775736,
+      "grad_norm": 0.4017353057861328,
+      "learning_rate": 9.997175826513637e-05,
+      "loss": 2.0352,
+      "step": 103
+    },
+    {
+      "epoch": 0.010802950036356082,
+      "grad_norm": 0.3651774227619171,
+      "learning_rate": 9.997120727277378e-05,
+      "loss": 2.1032,
+      "step": 104
+    },
+    {
+      "epoch": 0.01090682455593643,
+      "grad_norm": 0.32084140181541443,
+      "learning_rate": 9.997065095886741e-05,
+      "loss": 1.7161,
+      "step": 105
+    },
+    {
+      "epoch": 0.011010699075516776,
+      "grad_norm": 0.3476540446281433,
+      "learning_rate": 9.997008932347649e-05,
+      "loss": 1.8586,
+      "step": 106
+    },
+    {
+      "epoch": 0.011114573595097123,
+      "grad_norm": 0.3458161950111389,
+      "learning_rate": 9.996952236666084e-05,
+      "loss": 1.8134,
+      "step": 107
+    },
+    {
+      "epoch": 0.01121844811467747,
+      "grad_norm": 0.3564072549343109,
+      "learning_rate": 9.996895008848085e-05,
+      "loss": 1.9237,
+      "step": 108
+    },
+    {
+      "epoch": 0.011322322634257817,
+      "grad_norm": 0.33512404561042786,
+      "learning_rate": 9.996837248899741e-05,
+      "loss": 1.8897,
+      "step": 109
+    },
+    {
+      "epoch": 0.011426197153838163,
+      "grad_norm": 0.36306363344192505,
+      "learning_rate": 9.99677895682721e-05,
+      "loss": 1.9856,
+      "step": 110
+    },
+    {
+      "epoch": 0.011530071673418511,
+      "grad_norm": 0.32571837306022644,
+      "learning_rate": 9.996720132636697e-05,
+      "loss": 1.6574,
+      "step": 111
+    },
+    {
+      "epoch": 0.011633946192998857,
+      "grad_norm": 0.3236237168312073,
+      "learning_rate": 9.996660776334464e-05,
+      "loss": 1.7604,
+      "step": 112
+    },
+    {
+      "epoch": 0.011737820712579203,
+      "grad_norm": 0.34493985772132874,
+      "learning_rate": 9.996600887926834e-05,
+      "loss": 1.766,
+      "step": 113
+    },
+    {
+      "epoch": 0.011841695232159551,
+      "grad_norm": 0.34191280603408813,
+      "learning_rate": 9.996540467420186e-05,
+      "loss": 1.7565,
+      "step": 114
+    },
+    {
+      "epoch": 0.011945569751739897,
+      "grad_norm": 0.33236685395240784,
+      "learning_rate": 9.996479514820952e-05,
+      "loss": 1.8874,
+      "step": 115
+    },
+    {
+      "epoch": 0.012049444271320245,
+      "grad_norm": 0.37998783588409424,
+      "learning_rate": 9.996418030135622e-05,
+      "loss": 2.0761,
+      "step": 116
+    },
+    {
+      "epoch": 0.012153318790900591,
+      "grad_norm": 0.34933921694755554,
+      "learning_rate": 9.996356013370747e-05,
+      "loss": 2.0269,
+      "step": 117
+    },
+    {
+      "epoch": 0.012257193310480939,
+      "grad_norm": 0.3566916286945343,
+      "learning_rate": 9.99629346453293e-05,
+      "loss": 1.7982,
+      "step": 118
+    },
+    {
+      "epoch": 0.012361067830061285,
+      "grad_norm": 0.36033695936203003,
+      "learning_rate": 9.996230383628831e-05,
+      "loss": 1.9449,
+      "step": 119
+    },
+    {
+      "epoch": 0.012464942349641633,
+      "grad_norm": 0.34493738412857056,
+      "learning_rate": 9.996166770665167e-05,
+      "loss": 1.9453,
+      "step": 120
+    },
+    {
+      "epoch": 0.012568816869221979,
+      "grad_norm": 0.3142033815383911,
+      "learning_rate": 9.996102625648715e-05,
+      "loss": 1.7562,
+      "step": 121
+    },
+    {
+      "epoch": 0.012672691388802327,
+      "grad_norm": 0.3889004588127136,
+      "learning_rate": 9.996037948586305e-05,
+      "loss": 1.9439,
+      "step": 122
+    },
+    {
+      "epoch": 0.012776565908382673,
+      "grad_norm": 0.32579320669174194,
+      "learning_rate": 9.995972739484822e-05,
+      "loss": 1.7729,
+      "step": 123
+    },
+    {
+      "epoch": 0.012880440427963021,
+      "grad_norm": 0.3976999521255493,
+      "learning_rate": 9.995906998351215e-05,
+      "loss": 2.246,
+      "step": 124
+    },
+    {
+      "epoch": 0.012984314947543367,
+      "grad_norm": 0.38258862495422363,
+      "learning_rate": 9.99584072519248e-05,
+      "loss": 1.9098,
+      "step": 125
+    },
+    {
+      "epoch": 0.013088189467123715,
+      "grad_norm": 0.3394373655319214,
+      "learning_rate": 9.995773920015679e-05,
+      "loss": 1.7569,
+      "step": 126
+    },
+    {
+      "epoch": 0.013192063986704061,
+      "grad_norm": 0.3687582015991211,
+      "learning_rate": 9.995706582827922e-05,
+      "loss": 2.0098,
+      "step": 127
+    },
+    {
+      "epoch": 0.013295938506284409,
+      "grad_norm": 0.36636659502983093,
+      "learning_rate": 9.995638713636382e-05,
+      "loss": 1.9661,
+      "step": 128
+    },
+    {
+      "epoch": 0.013399813025864755,
+      "grad_norm": 0.33423036336898804,
+      "learning_rate": 9.995570312448289e-05,
+      "loss": 1.768,
+      "step": 129
+    },
+    {
+      "epoch": 0.013503687545445103,
+      "grad_norm": 0.34405717253685,
+      "learning_rate": 9.995501379270922e-05,
+      "loss": 1.649,
+      "step": 130
+    },
+    {
+      "epoch": 0.013607562065025449,
+      "grad_norm": 0.352104514837265,
+      "learning_rate": 9.995431914111624e-05,
+      "loss": 1.904,
+      "step": 131
+    },
+    {
+      "epoch": 0.013711436584605797,
+      "grad_norm": 0.34278222918510437,
+      "learning_rate": 9.995361916977795e-05,
+      "loss": 1.9214,
+      "step": 132
+    },
+    {
+      "epoch": 0.013815311104186143,
+      "grad_norm": 0.3315337300300598,
+      "learning_rate": 9.995291387876886e-05,
+      "loss": 1.9647,
+      "step": 133
+    },
+    {
+      "epoch": 0.01391918562376649,
+      "grad_norm": 0.349161297082901,
+      "learning_rate": 9.99522032681641e-05,
+      "loss": 1.8797,
+      "step": 134
+    },
+    {
+      "epoch": 0.014023060143346837,
+      "grad_norm": 0.32067403197288513,
+      "learning_rate": 9.995148733803933e-05,
+      "loss": 1.9065,
+      "step": 135
+    },
+    {
+      "epoch": 0.014126934662927184,
+      "grad_norm": 0.336671382188797,
+      "learning_rate": 9.995076608847078e-05,
+      "loss": 1.7649,
+      "step": 136
+    },
+    {
+      "epoch": 0.01423080918250753,
+      "grad_norm": 0.3314901292324066,
+      "learning_rate": 9.99500395195353e-05,
+      "loss": 1.7482,
+      "step": 137
+    },
+    {
+      "epoch": 0.014334683702087878,
+      "grad_norm": 0.3493605852127075,
+      "learning_rate": 9.994930763131022e-05,
+      "loss": 2.0021,
+      "step": 138
+    },
+    {
+      "epoch": 0.014438558221668224,
+      "grad_norm": 0.37969377636909485,
+      "learning_rate": 9.99485704238735e-05,
+      "loss": 1.9293,
+      "step": 139
+    },
+    {
+      "epoch": 0.014542432741248572,
+      "grad_norm": 0.31969866156578064,
+      "learning_rate": 9.994782789730363e-05,
+      "loss": 1.789,
+      "step": 140
+    },
+    {
+      "epoch": 0.014646307260828918,
+      "grad_norm": 0.3660503625869751,
+      "learning_rate": 9.994708005167972e-05,
+      "loss": 2.0092,
+      "step": 141
+    },
+    {
+      "epoch": 0.014750181780409266,
+      "grad_norm": 0.346720427274704,
+      "learning_rate": 9.994632688708138e-05,
+      "loss": 1.9393,
+      "step": 142
+    },
+    {
+      "epoch": 0.014854056299989612,
+      "grad_norm": 0.3383733630180359,
+      "learning_rate": 9.994556840358882e-05,
+      "loss": 1.7967,
+      "step": 143
+    },
+    {
+      "epoch": 0.01495793081956996,
+      "grad_norm": 0.3345518708229065,
+      "learning_rate": 9.994480460128282e-05,
+      "loss": 1.8901,
+      "step": 144
+    },
+    {
+      "epoch": 0.015061805339150306,
+      "grad_norm": 0.34996676445007324,
+      "learning_rate": 9.99440354802447e-05,
+      "loss": 1.8006,
+      "step": 145
+    },
+    {
+      "epoch": 0.015165679858730654,
+      "grad_norm": 0.34575289487838745,
+      "learning_rate": 9.994326104055639e-05,
+      "loss": 1.9967,
+      "step": 146
+    },
+    {
+      "epoch": 0.015269554378311,
+      "grad_norm": 0.39413705468177795,
+      "learning_rate": 9.994248128230035e-05,
+      "loss": 1.9811,
+      "step": 147
+    },
+    {
+      "epoch": 0.015373428897891348,
+      "grad_norm": 0.3441784083843231,
+      "learning_rate": 9.994169620555962e-05,
+      "loss": 1.7302,
+      "step": 148
+    },
+    {
+      "epoch": 0.015477303417471694,
+      "grad_norm": 0.32033270597457886,
+      "learning_rate": 9.994090581041781e-05,
+      "loss": 1.8822,
+      "step": 149
+    },
+    {
+      "epoch": 0.015581177937052042,
+      "grad_norm": 0.37517550587654114,
+      "learning_rate": 9.994011009695908e-05,
+      "loss": 2.0916,
+      "step": 150
+    },
+    {
+      "epoch": 0.01568505245663239,
+      "grad_norm": 0.3081076145172119,
+      "learning_rate": 9.993930906526818e-05,
+      "loss": 1.6987,
+      "step": 151
+    },
+    {
+      "epoch": 0.015788926976212734,
+      "grad_norm": 0.31799814105033875,
+      "learning_rate": 9.993850271543039e-05,
+      "loss": 1.6714,
+      "step": 152
+    },
+    {
+      "epoch": 0.015892801495793082,
+      "grad_norm": 0.3602600693702698,
+      "learning_rate": 9.99376910475316e-05,
+      "loss": 2.0209,
+      "step": 153
+    },
+    {
+      "epoch": 0.01599667601537343,
+      "grad_norm": 0.31287044286727905,
+      "learning_rate": 9.993687406165826e-05,
+      "loss": 1.6677,
+      "step": 154
+    },
+    {
+      "epoch": 0.016100550534953777,
+      "grad_norm": 0.34423595666885376,
+      "learning_rate": 9.993605175789733e-05,
+      "loss": 1.863,
+      "step": 155
+    },
+    {
+      "epoch": 0.016204425054534122,
+      "grad_norm": 0.3550304174423218,
+      "learning_rate": 9.993522413633643e-05,
+      "loss": 1.9312,
+      "step": 156
+    },
+    {
+      "epoch": 0.01630829957411447,
+      "grad_norm": 0.3672383725643158,
+      "learning_rate": 9.993439119706364e-05,
+      "loss": 2.0446,
+      "step": 157
+    },
+    {
+      "epoch": 0.016412174093694817,
+      "grad_norm": 0.32233211398124695,
+      "learning_rate": 9.993355294016771e-05,
+      "loss": 1.7413,
+      "step": 158
+    },
+    {
+      "epoch": 0.016516048613275165,
+      "grad_norm": 0.33033961057662964,
+      "learning_rate": 9.993270936573788e-05,
+      "loss": 1.8413,
+      "step": 159
+    },
+    {
+      "epoch": 0.01661992313285551,
+      "grad_norm": 0.3530619442462921,
+      "learning_rate": 9.9931860473864e-05,
+      "loss": 1.9083,
+      "step": 160
+    },
+    {
+      "epoch": 0.016723797652435857,
+      "grad_norm": 0.33920934796333313,
+      "learning_rate": 9.993100626463646e-05,
+      "loss": 1.8213,
+      "step": 161
+    },
+    {
+      "epoch": 0.016827672172016205,
+      "grad_norm": 0.3833538889884949,
+      "learning_rate": 9.993014673814624e-05,
+      "loss": 2.1405,
+      "step": 162
+    },
+    {
+      "epoch": 0.016931546691596553,
+      "grad_norm": 0.35124388337135315,
+      "learning_rate": 9.992928189448484e-05,
+      "loss": 1.8398,
+      "step": 163
+    },
+    {
+      "epoch": 0.017035421211176897,
+      "grad_norm": 0.34892386198043823,
+      "learning_rate": 9.992841173374441e-05,
+      "loss": 1.8151,
+      "step": 164
+    },
+    {
+      "epoch": 0.017139295730757245,
+      "grad_norm": 0.33861371874809265,
+      "learning_rate": 9.992753625601756e-05,
+      "loss": 1.7062,
+      "step": 165
+    },
+    {
+      "epoch": 0.017243170250337593,
+      "grad_norm": 0.34012502431869507,
+      "learning_rate": 9.992665546139757e-05,
+      "loss": 1.6682,
+      "step": 166
+    },
+    {
+      "epoch": 0.017347044769917937,
+      "grad_norm": 0.38989123702049255,
+      "learning_rate": 9.992576934997819e-05,
+      "loss": 1.9738,
+      "step": 167
+    },
+    {
+      "epoch": 0.017450919289498285,
+      "grad_norm": 0.37011152505874634,
+      "learning_rate": 9.992487792185383e-05,
+      "loss": 1.9593,
+      "step": 168
+    },
+    {
+      "epoch": 0.017554793809078633,
+      "grad_norm": 0.341496080160141,
+      "learning_rate": 9.992398117711941e-05,
+      "loss": 1.8838,
+      "step": 169
+    },
+    {
+      "epoch": 0.01765866832865898,
+      "grad_norm": 0.3466763496398926,
+      "learning_rate": 9.99230791158704e-05,
+      "loss": 1.7701,
+      "step": 170
+    },
+    {
+      "epoch": 0.017762542848239325,
+      "grad_norm": 0.36252373456954956,
+      "learning_rate": 9.992217173820288e-05,
+      "loss": 1.8656,
+      "step": 171
+    },
+    {
+      "epoch": 0.017866417367819673,
+      "grad_norm": 0.3116392195224762,
+      "learning_rate": 9.992125904421348e-05,
+      "loss": 1.7344,
+      "step": 172
+    },
+    {
+      "epoch": 0.01797029188740002,
+      "grad_norm": 0.3929993808269501,
+      "learning_rate": 9.992034103399939e-05,
+      "loss": 1.6855,
+      "step": 173
+    },
+    {
+      "epoch": 0.01807416640698037,
+      "grad_norm": 0.3422747552394867,
+      "learning_rate": 9.991941770765838e-05,
+      "loss": 1.8554,
+      "step": 174
+    },
+    {
+      "epoch": 0.018178040926560713,
+      "grad_norm": 0.3650684058666229,
+      "learning_rate": 9.991848906528876e-05,
+      "loss": 1.9407,
+      "step": 175
+    },
+    {
+      "epoch": 0.01828191544614106,
+      "grad_norm": 0.35484546422958374,
+      "learning_rate": 9.991755510698944e-05,
+      "loss": 1.8553,
+      "step": 176
+    },
+    {
+      "epoch": 0.01838578996572141,
+      "grad_norm": 0.3734646737575531,
+      "learning_rate": 9.991661583285987e-05,
+      "loss": 1.77,
+      "step": 177
+    },
+    {
+      "epoch": 0.018489664485301757,
+      "grad_norm": 0.3550145924091339,
+      "learning_rate": 9.991567124300009e-05,
+      "loss": 1.8825,
+      "step": 178
+    },
+    {
+      "epoch": 0.0185935390048821,
+      "grad_norm": 0.3555574119091034,
+      "learning_rate": 9.991472133751067e-05,
+      "loss": 1.9889,
+      "step": 179
+    },
+    {
+      "epoch": 0.01869741352446245,
+      "grad_norm": 0.32128405570983887,
+      "learning_rate": 9.991376611649279e-05,
+      "loss": 1.683,
+      "step": 180
+    },
+    {
+      "epoch": 0.018801288044042797,
+      "grad_norm": 0.3892490863800049,
+      "learning_rate": 9.991280558004815e-05,
+      "loss": 1.9933,
+      "step": 181
+    },
+    {
+      "epoch": 0.018905162563623144,
+      "grad_norm": 0.32436904311180115,
+      "learning_rate": 9.991183972827904e-05,
+      "loss": 1.8897,
+      "step": 182
+    },
+    {
+      "epoch": 0.01900903708320349,
+      "grad_norm": 0.3443160355091095,
+      "learning_rate": 9.991086856128833e-05,
+      "loss": 1.6934,
+      "step": 183
+    },
+    {
+      "epoch": 0.019112911602783837,
+      "grad_norm": 0.3968258202075958,
+      "learning_rate": 9.990989207917944e-05,
+      "loss": 1.9867,
+      "step": 184
+    },
+    {
+      "epoch": 0.019216786122364184,
+      "grad_norm": 0.3794998526573181,
+      "learning_rate": 9.990891028205636e-05,
+      "loss": 2.0397,
+      "step": 185
+    },
+    {
+      "epoch": 0.019320660641944532,
+      "grad_norm": 0.3027797043323517,
+      "learning_rate": 9.990792317002364e-05,
+      "loss": 1.6816,
+      "step": 186
+    },
+    {
+      "epoch": 0.019424535161524877,
+      "grad_norm": 0.3459644913673401,
+      "learning_rate": 9.99069307431864e-05,
+      "loss": 1.8124,
+      "step": 187
+    },
+    {
+      "epoch": 0.019528409681105224,
+      "grad_norm": 0.3326679766178131,
+      "learning_rate": 9.990593300165033e-05,
+      "loss": 1.6452,
+      "step": 188
+    },
+    {
+      "epoch": 0.019632284200685572,
+      "grad_norm": 0.3605174422264099,
+      "learning_rate": 9.990492994552167e-05,
+      "loss": 2.088,
+      "step": 189
+    },
+    {
+      "epoch": 0.01973615872026592,
+      "grad_norm": 0.3342396914958954,
+      "learning_rate": 9.990392157490724e-05,
+      "loss": 1.9396,
+      "step": 190
+    },
+    {
+      "epoch": 0.019840033239846264,
+      "grad_norm": 0.37349754571914673,
+      "learning_rate": 9.990290788991443e-05,
+      "loss": 1.9863,
+      "step": 191
+    },
+    {
+      "epoch": 0.019943907759426612,
+      "grad_norm": 0.3778427541255951,
+      "learning_rate": 9.99018888906512e-05,
+      "loss": 1.902,
+      "step": 192
+    },
+    {
+      "epoch": 0.02004778227900696,
+      "grad_norm": 0.3246352970600128,
+      "learning_rate": 9.990086457722604e-05,
+      "loss": 1.8484,
+      "step": 193
+    },
+    {
+      "epoch": 0.020151656798587308,
+      "grad_norm": 0.37238767743110657,
+      "learning_rate": 9.989983494974805e-05,
+      "loss": 1.9543,
+      "step": 194
+    },
+    {
+      "epoch": 0.020255531318167652,
+      "grad_norm": 0.37035781145095825,
+      "learning_rate": 9.989880000832686e-05,
+      "loss": 1.8695,
+      "step": 195
+    },
+    {
+      "epoch": 0.020359405837748,
+      "grad_norm": 0.3354018032550812,
+      "learning_rate": 9.989775975307272e-05,
+      "loss": 1.7589,
+      "step": 196
+    },
+    {
+      "epoch": 0.020463280357328348,
+      "grad_norm": 0.4081863462924957,
+      "learning_rate": 9.989671418409636e-05,
+      "loss": 2.1527,
+      "step": 197
+    },
+    {
+      "epoch": 0.020567154876908696,
+      "grad_norm": 0.33487629890441895,
+      "learning_rate": 9.989566330150914e-05,
+      "loss": 1.7132,
+      "step": 198
+    },
+    {
+      "epoch": 0.02067102939648904,
+      "grad_norm": 0.3445211350917816,
+      "learning_rate": 9.989460710542301e-05,
+      "loss": 1.8913,
+      "step": 199
+    },
+    {
+      "epoch": 0.020774903916069388,
+      "grad_norm": 0.40155649185180664,
+      "learning_rate": 9.98935455959504e-05,
+      "loss": 2.191,
+      "step": 200
+    },
+    {
+      "epoch": 0.020878778435649736,
+      "grad_norm": 0.35195931792259216,
+      "learning_rate": 9.989247877320436e-05,
+      "loss": 1.9199,
+      "step": 201
+    },
+    {
+      "epoch": 0.020982652955230083,
+      "grad_norm": 0.34090521931648254,
+      "learning_rate": 9.989140663729852e-05,
+      "loss": 1.831,
+      "step": 202
+    },
+    {
+      "epoch": 0.021086527474810428,
+      "grad_norm": 0.3321789801120758,
+      "learning_rate": 9.989032918834704e-05,
+      "loss": 1.8922,
+      "step": 203
+    },
+    {
+      "epoch": 0.021190401994390776,
+      "grad_norm": 0.321913480758667,
+      "learning_rate": 9.988924642646466e-05,
+      "loss": 1.8271,
+      "step": 204
+    },
+    {
+      "epoch": 0.021294276513971123,
+      "grad_norm": 0.39516186714172363,
+      "learning_rate": 9.988815835176668e-05,
+      "loss": 1.9391,
+      "step": 205
+    },
+    {
+      "epoch": 0.02139815103355147,
+      "grad_norm": 0.35352519154548645,
+      "learning_rate": 9.9887064964369e-05,
+      "loss": 1.6495,
+      "step": 206
+    },
+    {
+      "epoch": 0.021502025553131816,
+      "grad_norm": 0.3435503840446472,
+      "learning_rate": 9.988596626438801e-05,
+      "loss": 1.8272,
+      "step": 207
+    },
+    {
+      "epoch": 0.021605900072712163,
+      "grad_norm": 0.35792356729507446,
+      "learning_rate": 9.988486225194075e-05,
+      "loss": 1.8506,
+      "step": 208
+    },
+    {
+      "epoch": 0.02170977459229251,
+      "grad_norm": 0.32657817006111145,
+      "learning_rate": 9.988375292714478e-05,
+      "loss": 1.8219,
+      "step": 209
+    },
+    {
+      "epoch": 0.02181364911187286,
+      "grad_norm": 0.35596343874931335,
+      "learning_rate": 9.988263829011822e-05,
+      "loss": 1.8984,
+      "step": 210
+    },
+    {
+      "epoch": 0.021917523631453203,
+      "grad_norm": 0.35671266913414,
+      "learning_rate": 9.988151834097979e-05,
+      "loss": 1.8276,
+      "step": 211
+    },
+    {
+      "epoch": 0.02202139815103355,
+      "grad_norm": 0.3640732765197754,
+      "learning_rate": 9.988039307984874e-05,
+      "loss": 1.8979,
+      "step": 212
+    },
+    {
+      "epoch": 0.0221252726706139,
+      "grad_norm": 0.32779642939567566,
+      "learning_rate": 9.987926250684491e-05,
+      "loss": 1.8392,
+      "step": 213
+    },
+    {
+      "epoch": 0.022229147190194247,
+      "grad_norm": 0.3471597731113434,
+      "learning_rate": 9.98781266220887e-05,
+      "loss": 1.8287,
+      "step": 214
+    },
+    {
+      "epoch": 0.02233302170977459,
+      "grad_norm": 0.3553629219532013,
+      "learning_rate": 9.987698542570107e-05,
+      "loss": 1.8612,
+      "step": 215
+    },
+    {
+      "epoch": 0.02243689622935494,
+      "grad_norm": 0.3513118028640747,
+      "learning_rate": 9.987583891780355e-05,
+      "loss": 1.9711,
+      "step": 216
+    },
+    {
+      "epoch": 0.022540770748935287,
+      "grad_norm": 0.3648374378681183,
+      "learning_rate": 9.987468709851823e-05,
+      "loss": 1.7615,
+      "step": 217
+    },
+    {
+      "epoch": 0.022644645268515635,
+      "grad_norm": 0.3236382007598877,
+      "learning_rate": 9.987352996796777e-05,
+      "loss": 1.694,
+      "step": 218
+    },
+    {
+      "epoch": 0.02274851978809598,
+      "grad_norm": 0.3441942632198334,
+      "learning_rate": 9.98723675262754e-05,
+      "loss": 1.7834,
+      "step": 219
+    },
+    {
+      "epoch": 0.022852394307676327,
+      "grad_norm": 0.3617098331451416,
+      "learning_rate": 9.987119977356491e-05,
+      "loss": 1.8788,
+      "step": 220
+    },
+    {
+      "epoch": 0.022956268827256675,
+      "grad_norm": 0.3559406101703644,
+      "learning_rate": 9.987002670996064e-05,
+      "loss": 1.9663,
+      "step": 221
+    },
+    {
+      "epoch": 0.023060143346837023,
+      "grad_norm": 0.3278948664665222,
+      "learning_rate": 9.986884833558754e-05,
+      "loss": 1.822,
+      "step": 222
+    },
+    {
+      "epoch": 0.023164017866417367,
+      "grad_norm": 0.3330172896385193,
+      "learning_rate": 9.986766465057108e-05,
+      "loss": 1.8147,
+      "step": 223
+    },
+    {
+      "epoch": 0.023267892385997715,
+      "grad_norm": 0.3398604094982147,
+      "learning_rate": 9.986647565503731e-05,
+      "loss": 1.7854,
+      "step": 224
+    },
+    {
+      "epoch": 0.023371766905578063,
+      "grad_norm": 0.2998906970024109,
+      "learning_rate": 9.986528134911288e-05,
+      "loss": 1.6783,
+      "step": 225
+    },
+    {
+      "epoch": 0.023475641425158407,
+      "grad_norm": 0.3525906801223755,
+      "learning_rate": 9.986408173292492e-05,
+      "loss": 1.9996,
+      "step": 226
+    },
+    {
+      "epoch": 0.023579515944738755,
+      "grad_norm": 0.34057801961898804,
+      "learning_rate": 9.986287680660123e-05,
+      "loss": 1.8241,
+      "step": 227
+    },
+    {
+      "epoch": 0.023683390464319103,
+      "grad_norm": 0.3602697551250458,
+      "learning_rate": 9.98616665702701e-05,
+      "loss": 1.9336,
+      "step": 228
+    },
+    {
+      "epoch": 0.02378726498389945,
+      "grad_norm": 0.31595826148986816,
+      "learning_rate": 9.986045102406042e-05,
+      "loss": 1.7719,
+      "step": 229
+    },
+    {
+      "epoch": 0.023891139503479795,
+      "grad_norm": 0.34162575006484985,
+      "learning_rate": 9.985923016810163e-05,
+      "loss": 1.8026,
+      "step": 230
+    },
+    {
+      "epoch": 0.023995014023060143,
+      "grad_norm": 0.3397705852985382,
+      "learning_rate": 9.985800400252374e-05,
+      "loss": 1.9196,
+      "step": 231
+    },
+    {
+      "epoch": 0.02409888854264049,
+      "grad_norm": 0.3344639539718628,
+      "learning_rate": 9.985677252745733e-05,
+      "loss": 1.8512,
+      "step": 232
+    },
+    {
+      "epoch": 0.024202763062220838,
+      "grad_norm": 0.44500732421875,
+      "learning_rate": 9.985553574303354e-05,
+      "loss": 1.8336,
+      "step": 233
+    },
+    {
+      "epoch": 0.024306637581801183,
+      "grad_norm": 0.3289027512073517,
+      "learning_rate": 9.98542936493841e-05,
+      "loss": 1.6927,
+      "step": 234
+    },
+    {
+      "epoch": 0.02441051210138153,
+      "grad_norm": 0.3510255217552185,
+      "learning_rate": 9.985304624664125e-05,
+      "loss": 1.8582,
+      "step": 235
+    },
+    {
+      "epoch": 0.024514386620961878,
+      "grad_norm": 0.3442867696285248,
+      "learning_rate": 9.985179353493785e-05,
+      "loss": 1.9107,
+      "step": 236
+    },
+    {
+      "epoch": 0.024618261140542226,
+      "grad_norm": 0.3105999529361725,
+      "learning_rate": 9.98505355144073e-05,
+      "loss": 1.6351,
+      "step": 237
+    },
+    {
+      "epoch": 0.02472213566012257,
+      "grad_norm": 0.3260045647621155,
+      "learning_rate": 9.984927218518356e-05,
+      "loss": 1.7564,
+      "step": 238
+    },
+    {
+      "epoch": 0.024826010179702918,
+      "grad_norm": 0.32900920510292053,
+      "learning_rate": 9.984800354740117e-05,
+      "loss": 1.7372,
+      "step": 239
+    },
+    {
+      "epoch": 0.024929884699283266,
+      "grad_norm": 0.3402916193008423,
+      "learning_rate": 9.984672960119523e-05,
+      "loss": 1.7131,
+      "step": 240
+    },
+    {
+      "epoch": 0.025033759218863614,
+      "grad_norm": 0.35345762968063354,
+      "learning_rate": 9.984545034670142e-05,
+      "loss": 1.7636,
+      "step": 241
+    },
+    {
+      "epoch": 0.025137633738443958,
+      "grad_norm": 0.3537079393863678,
+      "learning_rate": 9.984416578405596e-05,
+      "loss": 1.9847,
+      "step": 242
+    },
+    {
+      "epoch": 0.025241508258024306,
+      "grad_norm": 0.3712044358253479,
+      "learning_rate": 9.984287591339562e-05,
+      "loss": 2.0124,
+      "step": 243
+    },
+    {
+      "epoch": 0.025345382777604654,
+      "grad_norm": 0.3348478376865387,
+      "learning_rate": 9.98415807348578e-05,
+      "loss": 1.8527,
+      "step": 244
+    },
+    {
+      "epoch": 0.025449257297185002,
+      "grad_norm": 0.3452380299568176,
+      "learning_rate": 9.984028024858041e-05,
+      "loss": 1.8608,
+      "step": 245
+    },
+    {
+      "epoch": 0.025553131816765346,
+      "grad_norm": 0.38733747601509094,
+      "learning_rate": 9.983897445470194e-05,
+      "loss": 2.029,
+      "step": 246
+    },
+    {
+      "epoch": 0.025657006336345694,
+      "grad_norm": 0.3828756809234619,
+      "learning_rate": 9.983766335336144e-05,
+      "loss": 1.9682,
+      "step": 247
+    },
+    {
+      "epoch": 0.025760880855926042,
+      "grad_norm": 0.38744086027145386,
+      "learning_rate": 9.983634694469855e-05,
+      "loss": 1.6414,
+      "step": 248
+    },
+    {
+      "epoch": 0.02586475537550639,
+      "grad_norm": 0.3316837251186371,
+      "learning_rate": 9.983502522885347e-05,
+      "loss": 1.7982,
+      "step": 249
+    },
+    {
+      "epoch": 0.025968629895086734,
+      "grad_norm": 0.3396342694759369,
+      "learning_rate": 9.983369820596691e-05,
+      "loss": 1.8246,
+      "step": 250
+    },
+    {
+      "epoch": 0.026072504414667082,
+      "grad_norm": 0.3051803708076477,
+      "learning_rate": 9.98323658761802e-05,
+      "loss": 1.733,
+      "step": 251
+    },
+    {
+      "epoch": 0.02617637893424743,
+      "grad_norm": 0.31522175669670105,
+      "learning_rate": 9.983102823963524e-05,
+      "loss": 1.6995,
+      "step": 252
+    },
+    {
+      "epoch": 0.026280253453827777,
+      "grad_norm": 0.33415964245796204,
+      "learning_rate": 9.982968529647447e-05,
+      "loss": 1.5215,
+      "step": 253
+    },
+    {
+      "epoch": 0.026384127973408122,
+      "grad_norm": 0.35238194465637207,
+      "learning_rate": 9.982833704684091e-05,
+      "loss": 1.8707,
+      "step": 254
+    },
+    {
+      "epoch": 0.02648800249298847,
+      "grad_norm": 0.3153392970561981,
+      "learning_rate": 9.982698349087812e-05,
+      "loss": 1.7123,
+      "step": 255
+    },
+    {
+      "epoch": 0.026591877012568817,
+      "grad_norm": 0.3397294282913208,
+      "learning_rate": 9.982562462873026e-05,
+      "loss": 1.7335,
+      "step": 256
+    },
+    {
+      "epoch": 0.026695751532149165,
+      "grad_norm": 0.3764454126358032,
+      "learning_rate": 9.982426046054204e-05,
+      "loss": 1.9431,
+      "step": 257
+    },
+    {
+      "epoch": 0.02679962605172951,
+      "grad_norm": 0.3482568860054016,
+      "learning_rate": 9.982289098645872e-05,
+      "loss": 1.8024,
+      "step": 258
+    },
+    {
+      "epoch": 0.026903500571309857,
+      "grad_norm": 0.3208499550819397,
+      "learning_rate": 9.982151620662612e-05,
+      "loss": 1.8129,
+      "step": 259
+    },
+    {
+      "epoch": 0.027007375090890205,
+      "grad_norm": 0.3308181166648865,
+      "learning_rate": 9.98201361211907e-05,
+      "loss": 1.8653,
+      "step": 260
+    },
+    {
+      "epoch": 0.027111249610470553,
+      "grad_norm": 0.3466727137565613,
+      "learning_rate": 9.981875073029938e-05,
+      "loss": 1.7716,
+      "step": 261
+    },
+    {
+      "epoch": 0.027215124130050897,
+      "grad_norm": 0.33778107166290283,
+      "learning_rate": 9.981736003409971e-05,
+      "loss": 1.8481,
+      "step": 262
+    },
+    {
+      "epoch": 0.027318998649631245,
+      "grad_norm": 0.367851197719574,
+      "learning_rate": 9.981596403273978e-05,
+      "loss": 2.0175,
+      "step": 263
+    },
+    {
+      "epoch": 0.027422873169211593,
+      "grad_norm": 0.3659020662307739,
+      "learning_rate": 9.981456272636826e-05,
+      "loss": 1.6928,
+      "step": 264
+    },
+    {
+      "epoch": 0.02752674768879194,
+      "grad_norm": 0.35648471117019653,
+      "learning_rate": 9.981315611513438e-05,
+      "loss": 1.9024,
+      "step": 265
+    },
+    {
+      "epoch": 0.027630622208372285,
+      "grad_norm": 0.3503887355327606,
+      "learning_rate": 9.981174419918795e-05,
+      "loss": 1.8439,
+      "step": 266
+    },
+    {
+      "epoch": 0.027734496727952633,
+      "grad_norm": 0.37520813941955566,
+      "learning_rate": 9.981032697867929e-05,
+      "loss": 2.0282,
+      "step": 267
+    },
+    {
+      "epoch": 0.02783837124753298,
+      "grad_norm": 0.3300826847553253,
+      "learning_rate": 9.980890445375934e-05,
+      "loss": 1.7793,
+      "step": 268
+    },
+    {
+      "epoch": 0.02794224576711333,
+      "grad_norm": 0.32417240738868713,
+      "learning_rate": 9.980747662457961e-05,
+      "loss": 1.8332,
+      "step": 269
+    },
+    {
+      "epoch": 0.028046120286693673,
+      "grad_norm": 0.32837119698524475,
+      "learning_rate": 9.980604349129211e-05,
+      "loss": 1.6546,
+      "step": 270
+    },
+    {
+      "epoch": 0.02814999480627402,
+      "grad_norm": 0.36148953437805176,
+      "learning_rate": 9.980460505404949e-05,
+      "loss": 1.9148,
+      "step": 271
+    },
+    {
+      "epoch": 0.02825386932585437,
+      "grad_norm": 0.3625042736530304,
+      "learning_rate": 9.980316131300493e-05,
+      "loss": 1.8963,
+      "step": 272
+    },
+    {
+      "epoch": 0.028357743845434717,
+      "grad_norm": 0.322068452835083,
+      "learning_rate": 9.980171226831216e-05,
+      "loss": 1.7182,
+      "step": 273
+    },
+    {
+      "epoch": 0.02846161836501506,
+      "grad_norm": 0.3443426787853241,
+      "learning_rate": 9.980025792012551e-05,
+      "loss": 1.7612,
+      "step": 274
+    },
+    {
+      "epoch": 0.02856549288459541,
+      "grad_norm": 0.3243624269962311,
+      "learning_rate": 9.979879826859983e-05,
+      "loss": 1.8107,
+      "step": 275
+    },
+    {
+      "epoch": 0.028669367404175757,
+      "grad_norm": 0.34362664818763733,
+      "learning_rate": 9.97973333138906e-05,
+      "loss": 1.8116,
+      "step": 276
+    },
+    {
+      "epoch": 0.028773241923756104,
+      "grad_norm": 0.3423418402671814,
+      "learning_rate": 9.97958630561538e-05,
+      "loss": 1.968,
+      "step": 277
+    },
+    {
+      "epoch": 0.02887711644333645,
+      "grad_norm": 0.3265102207660675,
+      "learning_rate": 9.9794387495546e-05,
+      "loss": 1.7281,
+      "step": 278
+    },
+    {
+      "epoch": 0.028980990962916797,
+      "grad_norm": 0.34078219532966614,
+      "learning_rate": 9.979290663222434e-05,
+      "loss": 1.8249,
+      "step": 279
+    },
+    {
+      "epoch": 0.029084865482497144,
+      "grad_norm": 0.34047967195510864,
+      "learning_rate": 9.979142046634653e-05,
+      "loss": 1.8084,
+      "step": 280
+    },
+    {
+      "epoch": 0.02918874000207749,
+      "grad_norm": 0.36841025948524475,
+      "learning_rate": 9.978992899807084e-05,
+      "loss": 1.9606,
+      "step": 281
+    },
+    {
+      "epoch": 0.029292614521657837,
+      "grad_norm": 0.32618024945259094,
+      "learning_rate": 9.978843222755607e-05,
+      "loss": 1.8753,
+      "step": 282
+    },
+    {
+      "epoch": 0.029396489041238184,
+      "grad_norm": 0.33189335465431213,
+      "learning_rate": 9.978693015496165e-05,
+      "loss": 1.948,
+      "step": 283
+    },
+    {
+      "epoch": 0.029500363560818532,
+      "grad_norm": 0.32388558983802795,
+      "learning_rate": 9.978542278044751e-05,
+      "loss": 1.8796,
+      "step": 284
+    },
+    {
+      "epoch": 0.029604238080398877,
+      "grad_norm": 0.34063830971717834,
+      "learning_rate": 9.978391010417418e-05,
+      "loss": 1.8221,
+      "step": 285
+    },
+    {
+      "epoch": 0.029708112599979224,
+      "grad_norm": 0.32267820835113525,
+      "learning_rate": 9.978239212630277e-05,
+      "loss": 1.7903,
+      "step": 286
+    },
+    {
+      "epoch": 0.029811987119559572,
+      "grad_norm": 0.35702529549598694,
+      "learning_rate": 9.978086884699492e-05,
+      "loss": 1.9183,
+      "step": 287
+    },
+    {
+      "epoch": 0.02991586163913992,
+      "grad_norm": 0.36471986770629883,
+      "learning_rate": 9.977934026641282e-05,
+      "loss": 1.8877,
+      "step": 288
+    },
+    {
+      "epoch": 0.030019736158720264,
+      "grad_norm": 0.3516945540904999,
+      "learning_rate": 9.977780638471928e-05,
+      "loss": 1.7437,
+      "step": 289
+    },
+    {
+      "epoch": 0.030123610678300612,
+      "grad_norm": 0.37206095457077026,
+      "learning_rate": 9.977626720207764e-05,
+      "loss": 1.794,
+      "step": 290
+    },
+    {
+      "epoch": 0.03022748519788096,
+      "grad_norm": 0.3756014108657837,
+      "learning_rate": 9.977472271865182e-05,
+      "loss": 1.8116,
+      "step": 291
+    },
+    {
+      "epoch": 0.030331359717461308,
+      "grad_norm": 0.3230532705783844,
+      "learning_rate": 9.977317293460631e-05,
+      "loss": 1.7871,
+      "step": 292
+    },
+    {
+      "epoch": 0.030435234237041652,
+      "grad_norm": 0.3606550693511963,
+      "learning_rate": 9.97716178501061e-05,
+      "loss": 1.8669,
+      "step": 293
+    },
+    {
+      "epoch": 0.030539108756622,
+      "grad_norm": 0.34765157103538513,
+      "learning_rate": 9.977005746531682e-05,
+      "loss": 1.8061,
+      "step": 294
+    },
+    {
+      "epoch": 0.030642983276202348,
+      "grad_norm": 0.3419150114059448,
+      "learning_rate": 9.976849178040466e-05,
+      "loss": 1.8189,
+      "step": 295
+    },
+    {
+      "epoch": 0.030746857795782696,
+      "grad_norm": 0.33086076378822327,
+      "learning_rate": 9.976692079553633e-05,
+      "loss": 1.7308,
+      "step": 296
+    },
+    {
+      "epoch": 0.03085073231536304,
+      "grad_norm": 0.3839011490345001,
+      "learning_rate": 9.976534451087913e-05,
+      "loss": 1.881,
+      "step": 297
+    },
+    {
+      "epoch": 0.030954606834943388,
+      "grad_norm": 0.3362378478050232,
+      "learning_rate": 9.976376292660091e-05,
+      "loss": 2.0111,
+      "step": 298
+    },
+    {
+      "epoch": 0.031058481354523736,
+      "grad_norm": 0.3804178535938263,
+      "learning_rate": 9.976217604287013e-05,
+      "loss": 2.1103,
+      "step": 299
+    },
+    {
+      "epoch": 0.031162355874104083,
+      "grad_norm": 0.38724952936172485,
+      "learning_rate": 9.976058385985575e-05,
+      "loss": 1.8858,
+      "step": 300
+    },
+    {
+      "epoch": 0.03126623039368443,
+      "grad_norm": 0.35871621966362,
+      "learning_rate": 9.975898637772734e-05,
+      "loss": 1.7076,
+      "step": 301
+    },
+    {
+      "epoch": 0.03137010491326478,
+      "grad_norm": 0.32210198044776917,
+      "learning_rate": 9.975738359665501e-05,
+      "loss": 1.8058,
+      "step": 302
+    },
+    {
+      "epoch": 0.03147397943284512,
+      "grad_norm": 0.34519657492637634,
+      "learning_rate": 9.975577551680946e-05,
+      "loss": 1.8827,
+      "step": 303
+    },
+    {
+      "epoch": 0.03157785395242547,
+      "grad_norm": 0.3406042456626892,
+      "learning_rate": 9.975416213836193e-05,
+      "loss": 1.801,
+      "step": 304
+    },
+    {
+      "epoch": 0.031681728472005816,
+      "grad_norm": 0.33171379566192627,
+      "learning_rate": 9.975254346148422e-05,
+      "loss": 1.7826,
+      "step": 305
+    },
+    {
+      "epoch": 0.031785602991586163,
+      "grad_norm": 0.31799814105033875,
+      "learning_rate": 9.975091948634871e-05,
+      "loss": 1.7969,
+      "step": 306
+    },
+    {
+      "epoch": 0.03188947751116651,
+      "grad_norm": 0.33427590131759644,
+      "learning_rate": 9.974929021312836e-05,
+      "loss": 1.9499,
+      "step": 307
+    },
+    {
+      "epoch": 0.03199335203074686,
+      "grad_norm": 0.336112916469574,
+      "learning_rate": 9.974765564199665e-05,
+      "loss": 1.9003,
+      "step": 308
+    },
+    {
+      "epoch": 0.03209722655032721,
+      "grad_norm": 0.33370164036750793,
+      "learning_rate": 9.974601577312768e-05,
+      "loss": 1.8738,
+      "step": 309
+    },
+    {
+      "epoch": 0.032201101069907555,
+      "grad_norm": 0.3412269949913025,
+      "learning_rate": 9.974437060669603e-05,
+      "loss": 1.8817,
+      "step": 310
+    },
+    {
+      "epoch": 0.032304975589487896,
+      "grad_norm": 0.34361353516578674,
+      "learning_rate": 9.974272014287697e-05,
+      "loss": 1.8391,
+      "step": 311
+    },
+    {
+      "epoch": 0.032408850109068243,
+      "grad_norm": 0.3552245795726776,
+      "learning_rate": 9.97410643818462e-05,
+      "loss": 1.8881,
+      "step": 312
+    },
+    {
+      "epoch": 0.03251272462864859,
+      "grad_norm": 0.3323882222175598,
+      "learning_rate": 9.973940332378007e-05,
+      "loss": 1.8975,
+      "step": 313
+    },
+    {
+      "epoch": 0.03261659914822894,
+      "grad_norm": 0.34055736660957336,
+      "learning_rate": 9.973773696885547e-05,
+      "loss": 1.7162,
+      "step": 314
+    },
+    {
+      "epoch": 0.03272047366780929,
+      "grad_norm": 0.3609575927257538,
+      "learning_rate": 9.973606531724985e-05,
+      "loss": 1.9902,
+      "step": 315
+    },
+    {
+      "epoch": 0.032824348187389635,
+      "grad_norm": 0.3580215573310852,
+      "learning_rate": 9.973438836914124e-05,
+      "loss": 1.9183,
+      "step": 316
+    },
+    {
+      "epoch": 0.03292822270696998,
+      "grad_norm": 0.34166282415390015,
+      "learning_rate": 9.973270612470822e-05,
+      "loss": 1.8884,
+      "step": 317
+    },
+    {
+      "epoch": 0.03303209722655033,
+      "grad_norm": 0.3344467878341675,
+      "learning_rate": 9.97310185841299e-05,
+      "loss": 1.9219,
+      "step": 318
+    },
+    {
+      "epoch": 0.03313597174613067,
+      "grad_norm": 0.3929234743118286,
+      "learning_rate": 9.972932574758604e-05,
+      "loss": 1.9071,
+      "step": 319
+    },
+    {
+      "epoch": 0.03323984626571102,
+      "grad_norm": 0.3376573324203491,
+      "learning_rate": 9.972762761525689e-05,
+      "loss": 1.7418,
+      "step": 320
+    },
+    {
+      "epoch": 0.03334372078529137,
+      "grad_norm": 0.33105456829071045,
+      "learning_rate": 9.972592418732327e-05,
+      "loss": 1.7941,
+      "step": 321
+    },
+    {
+      "epoch": 0.033447595304871715,
+      "grad_norm": 0.3593199551105499,
+      "learning_rate": 9.972421546396662e-05,
+      "loss": 1.8983,
+      "step": 322
+    },
+    {
+      "epoch": 0.03355146982445206,
+      "grad_norm": 0.37045425176620483,
+      "learning_rate": 9.972250144536888e-05,
+      "loss": 1.947,
+      "step": 323
+    },
+    {
+      "epoch": 0.03365534434403241,
+      "grad_norm": 0.3378489911556244,
+      "learning_rate": 9.972078213171259e-05,
+      "loss": 1.7318,
+      "step": 324
+    },
+    {
+      "epoch": 0.03375921886361276,
+      "grad_norm": 0.354125440120697,
+      "learning_rate": 9.971905752318084e-05,
+      "loss": 1.6396,
+      "step": 325
+    },
+    {
+      "epoch": 0.033863093383193106,
+      "grad_norm": 0.36457526683807373,
+      "learning_rate": 9.971732761995728e-05,
+      "loss": 1.8063,
+      "step": 326
+    },
+    {
+      "epoch": 0.03396696790277345,
+      "grad_norm": 0.33944860100746155,
+      "learning_rate": 9.971559242222615e-05,
+      "loss": 1.7624,
+      "step": 327
+    },
+    {
+      "epoch": 0.034070842422353795,
+      "grad_norm": 0.327573299407959,
+      "learning_rate": 9.971385193017221e-05,
+      "loss": 1.7963,
+      "step": 328
+    },
+    {
+      "epoch": 0.03417471694193414,
+      "grad_norm": 0.3168799877166748,
+      "learning_rate": 9.971210614398084e-05,
+      "loss": 1.7648,
+      "step": 329
+    },
+    {
+      "epoch": 0.03427859146151449,
+      "grad_norm": 0.3881419003009796,
+      "learning_rate": 9.971035506383792e-05,
+      "loss": 1.8365,
+      "step": 330
+    },
+    {
+      "epoch": 0.03438246598109484,
+      "grad_norm": 0.3255634903907776,
+      "learning_rate": 9.970859868992995e-05,
+      "loss": 1.9249,
+      "step": 331
+    },
+    {
+      "epoch": 0.034486340500675186,
+      "grad_norm": 0.3285115361213684,
+      "learning_rate": 9.970683702244395e-05,
+      "loss": 1.881,
+      "step": 332
+    },
+    {
+      "epoch": 0.034590215020255534,
+      "grad_norm": 0.34438276290893555,
+      "learning_rate": 9.970507006156755e-05,
+      "loss": 1.8085,
+      "step": 333
+    },
+    {
+      "epoch": 0.034694089539835875,
+      "grad_norm": 0.347380667924881,
+      "learning_rate": 9.970329780748888e-05,
+      "loss": 1.8762,
+      "step": 334
+    },
+    {
+      "epoch": 0.03479796405941622,
+      "grad_norm": 0.32174697518348694,
+      "learning_rate": 9.970152026039672e-05,
+      "loss": 1.7262,
+      "step": 335
+    },
+    {
+      "epoch": 0.03490183857899657,
+      "grad_norm": 0.3503490686416626,
+      "learning_rate": 9.969973742048032e-05,
+      "loss": 1.7248,
+      "step": 336
+    },
+    {
+      "epoch": 0.03500571309857692,
+      "grad_norm": 0.327619343996048,
+      "learning_rate": 9.969794928792957e-05,
+      "loss": 1.7031,
+      "step": 337
+    },
+    {
+      "epoch": 0.035109587618157266,
+      "grad_norm": 0.4252755343914032,
+      "learning_rate": 9.969615586293488e-05,
+      "loss": 2.0297,
+      "step": 338
+    },
+    {
+      "epoch": 0.035213462137737614,
+      "grad_norm": 0.36586689949035645,
+      "learning_rate": 9.969435714568722e-05,
+      "loss": 1.9272,
+      "step": 339
+    },
+    {
+      "epoch": 0.03531733665731796,
+      "grad_norm": 0.3425196409225464,
+      "learning_rate": 9.969255313637818e-05,
+      "loss": 1.8647,
+      "step": 340
+    },
+    {
+      "epoch": 0.03542121117689831,
+      "grad_norm": 0.3617844581604004,
+      "learning_rate": 9.969074383519983e-05,
+      "loss": 1.8622,
+      "step": 341
+    },
+    {
+      "epoch": 0.03552508569647865,
+      "grad_norm": 0.33944642543792725,
+      "learning_rate": 9.968892924234487e-05,
+      "loss": 1.7988,
+      "step": 342
+    },
+    {
+      "epoch": 0.035628960216059,
+      "grad_norm": 0.3269334137439728,
+      "learning_rate": 9.968710935800652e-05,
+      "loss": 1.8039,
+      "step": 343
+    },
+    {
+      "epoch": 0.035732834735639346,
+      "grad_norm": 0.3931077718734741,
+      "learning_rate": 9.968528418237862e-05,
+      "loss": 2.0847,
+      "step": 344
+    },
+    {
+      "epoch": 0.035836709255219694,
+      "grad_norm": 0.3752608299255371,
+      "learning_rate": 9.96834537156555e-05,
+      "loss": 2.0207,
+      "step": 345
+    },
+    {
+      "epoch": 0.03594058377480004,
+      "grad_norm": 0.33372077345848083,
+      "learning_rate": 9.96816179580321e-05,
+      "loss": 1.9464,
+      "step": 346
+    },
+    {
+      "epoch": 0.03604445829438039,
+      "grad_norm": 0.36559945344924927,
+      "learning_rate": 9.967977690970393e-05,
+      "loss": 1.9318,
+      "step": 347
+    },
+    {
+      "epoch": 0.03614833281396074,
+      "grad_norm": 0.3152011036872864,
+      "learning_rate": 9.967793057086706e-05,
+      "loss": 1.8081,
+      "step": 348
+    },
+    {
+      "epoch": 0.036252207333541085,
+      "grad_norm": 0.32508155703544617,
+      "learning_rate": 9.967607894171804e-05,
+      "loss": 1.7424,
+      "step": 349
+    },
+    {
+      "epoch": 0.036356081853121426,
+      "grad_norm": 0.3422446846961975,
+      "learning_rate": 9.967422202245413e-05,
+      "loss": 1.9266,
+      "step": 350
+    },
+    {
+      "epoch": 0.036459956372701774,
+      "grad_norm": 0.3278721570968628,
+      "learning_rate": 9.967235981327304e-05,
+      "loss": 1.8344,
+      "step": 351
+    },
+    {
+      "epoch": 0.03656383089228212,
+      "grad_norm": 0.32580530643463135,
+      "learning_rate": 9.967049231437309e-05,
+      "loss": 1.7158,
+      "step": 352
+    },
+    {
+      "epoch": 0.03666770541186247,
+      "grad_norm": 0.38664954900741577,
+      "learning_rate": 9.966861952595316e-05,
+      "loss": 1.9844,
+      "step": 353
+    },
+    {
+      "epoch": 0.03677157993144282,
+      "grad_norm": 0.40230998396873474,
+      "learning_rate": 9.966674144821266e-05,
+      "loss": 2.0037,
+      "step": 354
+    },
+    {
+      "epoch": 0.036875454451023165,
+      "grad_norm": 0.33027613162994385,
+      "learning_rate": 9.966485808135165e-05,
+      "loss": 1.7922,
+      "step": 355
+    },
+    {
+      "epoch": 0.03697932897060351,
+      "grad_norm": 0.342254102230072,
+      "learning_rate": 9.966296942557062e-05,
+      "loss": 1.8094,
+      "step": 356
+    },
+    {
+      "epoch": 0.03708320349018386,
+      "grad_norm": 0.3351558446884155,
+      "learning_rate": 9.966107548107074e-05,
+      "loss": 1.803,
+      "step": 357
+    },
+    {
+      "epoch": 0.0371870780097642,
+      "grad_norm": 0.34574928879737854,
+      "learning_rate": 9.96591762480537e-05,
+      "loss": 1.9305,
+      "step": 358
+    },
+    {
+      "epoch": 0.03729095252934455,
+      "grad_norm": 0.34656211733818054,
+      "learning_rate": 9.965727172672174e-05,
+      "loss": 1.781,
+      "step": 359
+    },
+    {
+      "epoch": 0.0373948270489249,
+      "grad_norm": 0.35084661841392517,
+      "learning_rate": 9.965536191727769e-05,
+      "loss": 1.886,
+      "step": 360
+    },
+    {
+      "epoch": 0.037498701568505245,
+      "grad_norm": 0.34373828768730164,
+      "learning_rate": 9.965344681992491e-05,
+      "loss": 1.8182,
+      "step": 361
+    },
+    {
+      "epoch": 0.03760257608808559,
+      "grad_norm": 0.33524051308631897,
+      "learning_rate": 9.965152643486738e-05,
+      "loss": 1.8175,
+      "step": 362
+    },
+    {
+      "epoch": 0.03770645060766594,
+      "grad_norm": 0.3301374912261963,
+      "learning_rate": 9.964960076230955e-05,
+      "loss": 1.7128,
+      "step": 363
+    },
+    {
+      "epoch": 0.03781032512724629,
+      "grad_norm": 0.35539349913597107,
+      "learning_rate": 9.964766980245653e-05,
+      "loss": 1.8029,
+      "step": 364
+    },
+    {
+      "epoch": 0.037914199646826637,
+      "grad_norm": 0.3248385190963745,
+      "learning_rate": 9.964573355551394e-05,
+      "loss": 1.6429,
+      "step": 365
+    },
+    {
+      "epoch": 0.03801807416640698,
+      "grad_norm": 0.34955573081970215,
+      "learning_rate": 9.964379202168799e-05,
+      "loss": 1.9152,
+      "step": 366
+    },
+    {
+      "epoch": 0.038121948685987325,
+      "grad_norm": 0.32366281747817993,
+      "learning_rate": 9.964184520118542e-05,
+      "loss": 1.7536,
+      "step": 367
+    },
+    {
+      "epoch": 0.03822582320556767,
+      "grad_norm": 0.3177925646305084,
+      "learning_rate": 9.963989309421356e-05,
+      "loss": 1.7626,
+      "step": 368
+    },
+    {
+      "epoch": 0.03832969772514802,
+      "grad_norm": 0.35649746656417847,
+      "learning_rate": 9.96379357009803e-05,
+      "loss": 1.822,
+      "step": 369
+    },
+    {
+      "epoch": 0.03843357224472837,
+      "grad_norm": 0.3334449827671051,
+      "learning_rate": 9.963597302169406e-05,
+      "loss": 1.7122,
+      "step": 370
+    },
+    {
+      "epoch": 0.038537446764308717,
+      "grad_norm": 0.35319963097572327,
+      "learning_rate": 9.963400505656388e-05,
+      "loss": 1.8249,
+      "step": 371
+    },
+    {
+      "epoch": 0.038641321283889064,
+      "grad_norm": 0.34933820366859436,
+      "learning_rate": 9.963203180579932e-05,
+      "loss": 2.1094,
+      "step": 372
+    },
+    {
+      "epoch": 0.03874519580346941,
+      "grad_norm": 0.35011550784111023,
+      "learning_rate": 9.963005326961052e-05,
+      "loss": 1.8247,
+      "step": 373
+    },
+    {
+      "epoch": 0.03884907032304975,
+      "grad_norm": 0.4046299159526825,
+      "learning_rate": 9.962806944820817e-05,
+      "loss": 2.1071,
+      "step": 374
+    },
+    {
+      "epoch": 0.0389529448426301,
+      "grad_norm": 0.3381158113479614,
+      "learning_rate": 9.962608034180353e-05,
+      "loss": 1.8242,
+      "step": 375
+    },
+    {
+      "epoch": 0.03905681936221045,
+      "grad_norm": 0.35168975591659546,
+      "learning_rate": 9.962408595060845e-05,
+      "loss": 2.0001,
+      "step": 376
+    },
+    {
+      "epoch": 0.039160693881790797,
+      "grad_norm": 0.35159188508987427,
+      "learning_rate": 9.962208627483529e-05,
+      "loss": 1.8534,
+      "step": 377
+    },
+    {
+      "epoch": 0.039264568401371144,
+      "grad_norm": 0.31816786527633667,
+      "learning_rate": 9.962008131469703e-05,
+      "loss": 1.6624,
+      "step": 378
+    },
+    {
+      "epoch": 0.03936844292095149,
+      "grad_norm": 0.33059608936309814,
+      "learning_rate": 9.961807107040712e-05,
+      "loss": 1.7612,
+      "step": 379
+    },
+    {
+      "epoch": 0.03947231744053184,
+      "grad_norm": 0.33961811661720276,
+      "learning_rate": 9.96160555421797e-05,
+      "loss": 1.8432,
+      "step": 380
+    },
+    {
+      "epoch": 0.03957619196011219,
+      "grad_norm": 0.3444601595401764,
+      "learning_rate": 9.961403473022939e-05,
+      "loss": 1.8842,
+      "step": 381
+    },
+    {
+      "epoch": 0.03968006647969253,
+      "grad_norm": 0.3596336841583252,
+      "learning_rate": 9.961200863477139e-05,
+      "loss": 2.0091,
+      "step": 382
+    },
+    {
+      "epoch": 0.039783940999272877,
+      "grad_norm": 0.35697799921035767,
+      "learning_rate": 9.960997725602144e-05,
+      "loss": 1.8124,
+      "step": 383
+    },
+    {
+      "epoch": 0.039887815518853224,
+      "grad_norm": 0.33248475193977356,
+      "learning_rate": 9.96079405941959e-05,
+      "loss": 1.9146,
+      "step": 384
+    },
+    {
+      "epoch": 0.03999169003843357,
+      "grad_norm": 0.3686772882938385,
+      "learning_rate": 9.960589864951162e-05,
+      "loss": 1.8714,
+      "step": 385
+    },
+    {
+      "epoch": 0.04009556455801392,
+      "grad_norm": 0.36085087060928345,
+      "learning_rate": 9.960385142218609e-05,
+      "loss": 1.8268,
+      "step": 386
+    },
+    {
+      "epoch": 0.04019943907759427,
+      "grad_norm": 0.32262122631073,
+      "learning_rate": 9.960179891243731e-05,
+      "loss": 1.6027,
+      "step": 387
+    },
+    {
+      "epoch": 0.040303313597174616,
+      "grad_norm": 0.361555278301239,
+      "learning_rate": 9.959974112048386e-05,
+      "loss": 1.9221,
+      "step": 388
+    },
+    {
+      "epoch": 0.040407188116754963,
+      "grad_norm": 0.3558778166770935,
+      "learning_rate": 9.959767804654487e-05,
+      "loss": 1.8234,
+      "step": 389
+    },
+    {
+      "epoch": 0.040511062636335304,
+      "grad_norm": 0.3629186749458313,
+      "learning_rate": 9.959560969084003e-05,
+      "loss": 1.9483,
+      "step": 390
+    },
+    {
+      "epoch": 0.04061493715591565,
+      "grad_norm": 0.37084469199180603,
+      "learning_rate": 9.959353605358964e-05,
+      "loss": 1.7024,
+      "step": 391
+    },
+    {
+      "epoch": 0.040718811675496,
+      "grad_norm": 0.3552490770816803,
+      "learning_rate": 9.95914571350145e-05,
+      "loss": 1.884,
+      "step": 392
+    },
+    {
+      "epoch": 0.04082268619507635,
+      "grad_norm": 0.34589436650276184,
+      "learning_rate": 9.958937293533599e-05,
+      "loss": 1.8548,
+      "step": 393
+    },
+    {
+      "epoch": 0.040926560714656696,
+      "grad_norm": 0.3737829327583313,
+      "learning_rate": 9.958728345477608e-05,
+      "loss": 2.0568,
+      "step": 394
+    },
+    {
+      "epoch": 0.041030435234237043,
+      "grad_norm": 0.35843873023986816,
+      "learning_rate": 9.958518869355728e-05,
+      "loss": 1.7908,
+      "step": 395
+    },
+    {
+      "epoch": 0.04113430975381739,
+      "grad_norm": 0.6880629062652588,
+      "learning_rate": 9.958308865190267e-05,
+      "loss": 1.8225,
+      "step": 396
+    },
+    {
+      "epoch": 0.04123818427339773,
+      "grad_norm": 0.3438432812690735,
+      "learning_rate": 9.958098333003588e-05,
+      "loss": 1.7719,
+      "step": 397
+    },
+    {
+      "epoch": 0.04134205879297808,
+      "grad_norm": 0.3586975634098053,
+      "learning_rate": 9.957887272818112e-05,
+      "loss": 1.9513,
+      "step": 398
+    },
+    {
+      "epoch": 0.04144593331255843,
+      "grad_norm": 0.3421470522880554,
+      "learning_rate": 9.957675684656312e-05,
+      "loss": 1.8509,
+      "step": 399
+    },
+    {
+      "epoch": 0.041549807832138776,
+      "grad_norm": 0.3313000202178955,
+      "learning_rate": 9.957463568540725e-05,
+      "loss": 1.8173,
+      "step": 400
+    },
+    {
+      "epoch": 0.041653682351719123,
+      "grad_norm": 0.3544430732727051,
+      "learning_rate": 9.957250924493938e-05,
+      "loss": 1.8059,
+      "step": 401
+    },
+    {
+      "epoch": 0.04175755687129947,
+      "grad_norm": 0.36002522706985474,
+      "learning_rate": 9.957037752538592e-05,
+      "loss": 1.9456,
+      "step": 402
+    },
+    {
+      "epoch": 0.04186143139087982,
+      "grad_norm": 0.3440776765346527,
+      "learning_rate": 9.956824052697395e-05,
+      "loss": 1.7831,
+      "step": 403
+    },
+    {
+      "epoch": 0.04196530591046017,
+      "grad_norm": 0.36796486377716064,
+      "learning_rate": 9.956609824993099e-05,
+      "loss": 1.791,
+      "step": 404
+    },
+    {
+      "epoch": 0.04206918043004051,
+      "grad_norm": 0.3968099355697632,
+      "learning_rate": 9.956395069448523e-05,
+      "loss": 1.8728,
+      "step": 405
+    },
+    {
+      "epoch": 0.042173054949620856,
+      "grad_norm": 0.3467409610748291,
+      "learning_rate": 9.95617978608653e-05,
+      "loss": 1.8004,
+      "step": 406
+    },
+    {
+      "epoch": 0.042276929469201203,
+      "grad_norm": 0.3945915102958679,
+      "learning_rate": 9.955963974930052e-05,
+      "loss": 1.7891,
+      "step": 407
+    },
+    {
+      "epoch": 0.04238080398878155,
+      "grad_norm": 0.3563166856765747,
+      "learning_rate": 9.955747636002068e-05,
+      "loss": 1.9755,
+      "step": 408
+    },
+    {
+      "epoch": 0.0424846785083619,
+      "grad_norm": 0.4015011191368103,
+      "learning_rate": 9.955530769325616e-05,
+      "loss": 1.6478,
+      "step": 409
+    },
+    {
+      "epoch": 0.04258855302794225,
+      "grad_norm": 0.36491334438323975,
+      "learning_rate": 9.955313374923791e-05,
+      "loss": 1.9403,
+      "step": 410
+    },
+    {
+      "epoch": 0.042692427547522595,
+      "grad_norm": 0.3891184628009796,
+      "learning_rate": 9.955095452819747e-05,
+      "loss": 2.0413,
+      "step": 411
+    },
+    {
+      "epoch": 0.04279630206710294,
+      "grad_norm": 0.3426514267921448,
+      "learning_rate": 9.954877003036687e-05,
+      "loss": 1.6253,
+      "step": 412
+    },
+    {
+      "epoch": 0.042900176586683283,
+      "grad_norm": 0.48841214179992676,
+      "learning_rate": 9.954658025597876e-05,
+      "loss": 2.1789,
+      "step": 413
+    },
+    {
+      "epoch": 0.04300405110626363,
+      "grad_norm": 0.3180578351020813,
+      "learning_rate": 9.954438520526635e-05,
+      "loss": 1.4562,
+      "step": 414
+    },
+    {
+      "epoch": 0.04310792562584398,
+      "grad_norm": 0.3487285077571869,
+      "learning_rate": 9.954218487846334e-05,
+      "loss": 1.7904,
+      "step": 415
+    },
+    {
+      "epoch": 0.04321180014542433,
+      "grad_norm": 0.3434458374977112,
+      "learning_rate": 9.953997927580411e-05,
+      "loss": 1.8548,
+      "step": 416
+    },
+    {
+      "epoch": 0.043315674665004675,
+      "grad_norm": 0.3450503647327423,
+      "learning_rate": 9.953776839752351e-05,
+      "loss": 1.7331,
+      "step": 417
+    },
+    {
+      "epoch": 0.04341954918458502,
+      "grad_norm": 0.33946654200553894,
+      "learning_rate": 9.953555224385698e-05,
+      "loss": 1.883,
+      "step": 418
+    },
+    {
+      "epoch": 0.04352342370416537,
+      "grad_norm": 0.37217575311660767,
+      "learning_rate": 9.953333081504052e-05,
+      "loss": 1.6892,
+      "step": 419
+    },
+    {
+      "epoch": 0.04362729822374572,
+      "grad_norm": 0.3341975510120392,
+      "learning_rate": 9.953110411131072e-05,
+      "loss": 1.734,
+      "step": 420
+    },
+    {
+      "epoch": 0.04373117274332606,
+      "grad_norm": 0.3347923159599304,
+      "learning_rate": 9.952887213290469e-05,
+      "loss": 1.8173,
+      "step": 421
+    },
+    {
+      "epoch": 0.04383504726290641,
+      "grad_norm": 0.35578978061676025,
+      "learning_rate": 9.952663488006011e-05,
+      "loss": 1.9877,
+      "step": 422
+    },
+    {
+      "epoch": 0.043938921782486755,
+      "grad_norm": 0.3420504331588745,
+      "learning_rate": 9.952439235301522e-05,
+      "loss": 1.8042,
+      "step": 423
+    },
+    {
+      "epoch": 0.0440427963020671,
+      "grad_norm": 0.3436269760131836,
+      "learning_rate": 9.952214455200887e-05,
+      "loss": 1.7833,
+      "step": 424
+    },
+    {
+      "epoch": 0.04414667082164745,
+      "grad_norm": 0.3772909641265869,
+      "learning_rate": 9.951989147728043e-05,
+      "loss": 1.9778,
+      "step": 425
+    },
+    {
+      "epoch": 0.0442505453412278,
+      "grad_norm": 0.3602845370769501,
+      "learning_rate": 9.951763312906979e-05,
+      "loss": 1.8447,
+      "step": 426
+    },
+    {
+      "epoch": 0.044354419860808146,
+      "grad_norm": 0.3500402271747589,
+      "learning_rate": 9.951536950761748e-05,
+      "loss": 1.6772,
+      "step": 427
+    },
+    {
+      "epoch": 0.044458294380388494,
+      "grad_norm": 0.3483547866344452,
+      "learning_rate": 9.951310061316455e-05,
+      "loss": 1.8328,
+      "step": 428
+    },
+    {
+      "epoch": 0.044562168899968835,
+      "grad_norm": 0.3545311391353607,
+      "learning_rate": 9.951082644595264e-05,
+      "loss": 1.8814,
+      "step": 429
+    },
+    {
+      "epoch": 0.04466604341954918,
+      "grad_norm": 0.33376967906951904,
+      "learning_rate": 9.95085470062239e-05,
+      "loss": 1.8108,
+      "step": 430
+    },
+    {
+      "epoch": 0.04476991793912953,
+      "grad_norm": 0.330287903547287,
+      "learning_rate": 9.950626229422107e-05,
+      "loss": 1.7423,
+      "step": 431
+    },
+    {
+      "epoch": 0.04487379245870988,
+      "grad_norm": 0.36695408821105957,
+      "learning_rate": 9.950397231018748e-05,
+      "loss": 1.7873,
+      "step": 432
+    },
+    {
+      "epoch": 0.044977666978290226,
+      "grad_norm": 0.3538840711116791,
+      "learning_rate": 9.950167705436699e-05,
+      "loss": 1.9949,
+      "step": 433
+    },
+    {
+      "epoch": 0.045081541497870574,
+      "grad_norm": 0.31848835945129395,
+      "learning_rate": 9.949937652700402e-05,
+      "loss": 1.5736,
+      "step": 434
+    },
+    {
+      "epoch": 0.04518541601745092,
+      "grad_norm": 0.3445279598236084,
+      "learning_rate": 9.949707072834356e-05,
+      "loss": 1.7662,
+      "step": 435
+    },
+    {
+      "epoch": 0.04528929053703127,
+      "grad_norm": 0.37137672305107117,
+      "learning_rate": 9.949475965863116e-05,
+      "loss": 1.8209,
+      "step": 436
+    },
+    {
+      "epoch": 0.04539316505661161,
+      "grad_norm": 0.43532824516296387,
+      "learning_rate": 9.949244331811293e-05,
+      "loss": 2.1643,
+      "step": 437
+    },
+    {
+      "epoch": 0.04549703957619196,
+      "grad_norm": 0.35221028327941895,
+      "learning_rate": 9.949012170703556e-05,
+      "loss": 1.8557,
+      "step": 438
+    },
+    {
+      "epoch": 0.045600914095772306,
+      "grad_norm": 0.34177568554878235,
+      "learning_rate": 9.948779482564624e-05,
+      "loss": 1.7119,
+      "step": 439
+    },
+    {
+      "epoch": 0.045704788615352654,
+      "grad_norm": 0.3455469012260437,
+      "learning_rate": 9.948546267419278e-05,
+      "loss": 1.8411,
+      "step": 440
+    },
+    {
+      "epoch": 0.045808663134933,
+      "grad_norm": 0.34957823157310486,
+      "learning_rate": 9.948312525292358e-05,
+      "loss": 1.7674,
+      "step": 441
+    },
+    {
+      "epoch": 0.04591253765451335,
+      "grad_norm": 0.3494671583175659,
+      "learning_rate": 9.948078256208751e-05,
+      "loss": 1.9521,
+      "step": 442
+    },
+    {
+      "epoch": 0.0460164121740937,
+      "grad_norm": 0.3724217116832733,
+      "learning_rate": 9.947843460193407e-05,
+      "loss": 1.9358,
+      "step": 443
+    },
+    {
+      "epoch": 0.046120286693674045,
+      "grad_norm": 0.34524059295654297,
+      "learning_rate": 9.947608137271328e-05,
+      "loss": 1.8426,
+      "step": 444
+    },
+    {
+      "epoch": 0.046224161213254386,
+      "grad_norm": 0.3344314396381378,
+      "learning_rate": 9.947372287467576e-05,
+      "loss": 1.7263,
+      "step": 445
+    },
+    {
+      "epoch": 0.046328035732834734,
+      "grad_norm": 0.3640967309474945,
+      "learning_rate": 9.947135910807265e-05,
+      "loss": 1.9716,
+      "step": 446
+    },
+    {
+      "epoch": 0.04643191025241508,
+      "grad_norm": 0.38016462326049805,
+      "learning_rate": 9.946899007315569e-05,
+      "loss": 1.7952,
+      "step": 447
+    },
+    {
+      "epoch": 0.04653578477199543,
+      "grad_norm": 0.35454219579696655,
+      "learning_rate": 9.946661577017717e-05,
+      "loss": 1.8804,
+      "step": 448
+    },
+    {
+      "epoch": 0.04663965929157578,
+      "grad_norm": 0.3755532503128052,
+      "learning_rate": 9.946423619938992e-05,
+      "loss": 2.0071,
+      "step": 449
+    },
+    {
+      "epoch": 0.046743533811156125,
+      "grad_norm": 0.3422885537147522,
+      "learning_rate": 9.946185136104735e-05,
+      "loss": 1.6811,
+      "step": 450
+    },
+    {
+      "epoch": 0.04684740833073647,
+      "grad_norm": 0.33953240513801575,
+      "learning_rate": 9.945946125540343e-05,
+      "loss": 1.7127,
+      "step": 451
+    },
+    {
+      "epoch": 0.046951282850316814,
+      "grad_norm": 0.313754677772522,
+      "learning_rate": 9.945706588271269e-05,
+      "loss": 1.7116,
+      "step": 452
+    },
+    {
+      "epoch": 0.04705515736989716,
+      "grad_norm": 0.3217809200286865,
+      "learning_rate": 9.94546652432302e-05,
+      "loss": 1.7475,
+      "step": 453
+    },
+    {
+      "epoch": 0.04715903188947751,
+      "grad_norm": 0.35068783164024353,
+      "learning_rate": 9.945225933721163e-05,
+      "loss": 1.7588,
+      "step": 454
+    },
+    {
+      "epoch": 0.04726290640905786,
+      "grad_norm": 0.32471030950546265,
+      "learning_rate": 9.944984816491318e-05,
+      "loss": 1.6391,
+      "step": 455
+    },
+    {
+      "epoch": 0.047366780928638205,
+      "grad_norm": 0.3416842222213745,
+      "learning_rate": 9.944743172659164e-05,
+      "loss": 1.7689,
+      "step": 456
+    },
+    {
+      "epoch": 0.04747065544821855,
+      "grad_norm": 0.3535037934780121,
+      "learning_rate": 9.94450100225043e-05,
+      "loss": 1.6911,
+      "step": 457
+    },
+    {
+      "epoch": 0.0475745299677989,
+      "grad_norm": 0.3632698953151703,
+      "learning_rate": 9.944258305290908e-05,
+      "loss": 1.949,
+      "step": 458
+    },
+    {
+      "epoch": 0.04767840448737925,
+      "grad_norm": 0.38424357771873474,
+      "learning_rate": 9.944015081806445e-05,
+      "loss": 1.9012,
+      "step": 459
+    },
+    {
+      "epoch": 0.04778227900695959,
+      "grad_norm": 0.3942817747592926,
+      "learning_rate": 9.94377133182294e-05,
+      "loss": 2.0238,
+      "step": 460
+    },
+    {
+      "epoch": 0.04788615352653994,
+      "grad_norm": 0.3337697684764862,
+      "learning_rate": 9.943527055366351e-05,
+      "loss": 1.7753,
+      "step": 461
+    },
+    {
+      "epoch": 0.047990028046120285,
+      "grad_norm": 0.3492131233215332,
+      "learning_rate": 9.94328225246269e-05,
+      "loss": 1.7137,
+      "step": 462
+    },
+    {
+      "epoch": 0.04809390256570063,
+      "grad_norm": 0.3351963758468628,
+      "learning_rate": 9.943036923138032e-05,
+      "loss": 1.7599,
+      "step": 463
+    },
+    {
+      "epoch": 0.04819777708528098,
+      "grad_norm": 0.33879536390304565,
+      "learning_rate": 9.942791067418496e-05,
+      "loss": 1.7868,
+      "step": 464
+    },
+    {
+      "epoch": 0.04830165160486133,
+      "grad_norm": 0.3504945933818817,
+      "learning_rate": 9.942544685330267e-05,
+      "loss": 1.93,
+      "step": 465
+    },
+    {
+      "epoch": 0.048405526124441677,
+      "grad_norm": 0.3818790912628174,
+      "learning_rate": 9.942297776899583e-05,
+      "loss": 1.9025,
+      "step": 466
+    },
+    {
+      "epoch": 0.048509400644022024,
+      "grad_norm": 0.346236914396286,
+      "learning_rate": 9.942050342152736e-05,
+      "loss": 1.7295,
+      "step": 467
+    },
+    {
+      "epoch": 0.048613275163602365,
+      "grad_norm": 0.39538463950157166,
+      "learning_rate": 9.941802381116078e-05,
+      "loss": 1.5522,
+      "step": 468
+    },
+    {
+      "epoch": 0.04871714968318271,
+      "grad_norm": 0.35648438334465027,
+      "learning_rate": 9.941553893816014e-05,
+      "loss": 1.7374,
+      "step": 469
+    },
+    {
+      "epoch": 0.04882102420276306,
+      "grad_norm": 0.3466714024543762,
+      "learning_rate": 9.941304880279005e-05,
+      "loss": 1.7781,
+      "step": 470
+    },
+    {
+      "epoch": 0.04892489872234341,
+      "grad_norm": 0.38975635170936584,
+      "learning_rate": 9.941055340531571e-05,
+      "loss": 1.9754,
+      "step": 471
+    },
+    {
+      "epoch": 0.049028773241923757,
+      "grad_norm": 0.3584454655647278,
+      "learning_rate": 9.940805274600285e-05,
+      "loss": 1.8484,
+      "step": 472
+    },
+    {
+      "epoch": 0.049132647761504104,
+      "grad_norm": 0.33715423941612244,
+      "learning_rate": 9.940554682511775e-05,
+      "loss": 1.9061,
+      "step": 473
+    },
+    {
+      "epoch": 0.04923652228108445,
+      "grad_norm": 0.35068365931510925,
+      "learning_rate": 9.940303564292731e-05,
+      "loss": 1.759,
+      "step": 474
+    },
+    {
+      "epoch": 0.0493403968006648,
+      "grad_norm": 0.37917405366897583,
+      "learning_rate": 9.940051919969893e-05,
+      "loss": 1.8638,
+      "step": 475
+    },
+    {
+      "epoch": 0.04944427132024514,
+      "grad_norm": 0.38259661197662354,
+      "learning_rate": 9.939799749570059e-05,
+      "loss": 1.9465,
+      "step": 476
+    },
+    {
+      "epoch": 0.04954814583982549,
+      "grad_norm": 0.334435373544693,
+      "learning_rate": 9.939547053120084e-05,
+      "loss": 1.8029,
+      "step": 477
+    },
+    {
+      "epoch": 0.049652020359405837,
+      "grad_norm": 0.3447658121585846,
+      "learning_rate": 9.939293830646878e-05,
+      "loss": 1.6021,
+      "step": 478
+    },
+    {
+      "epoch": 0.049755894878986184,
+      "grad_norm": 0.3410322964191437,
+      "learning_rate": 9.939040082177407e-05,
+      "loss": 1.9107,
+      "step": 479
+    },
+    {
+      "epoch": 0.04985976939856653,
+      "grad_norm": 0.3771478831768036,
+      "learning_rate": 9.938785807738693e-05,
+      "loss": 2.0531,
+      "step": 480
+    },
+    {
+      "epoch": 0.04996364391814688,
+      "grad_norm": 0.421427845954895,
+      "learning_rate": 9.938531007357813e-05,
+      "loss": 1.9953,
+      "step": 481
+    },
+    {
+      "epoch": 0.05006751843772723,
+      "grad_norm": 0.347931444644928,
+      "learning_rate": 9.938275681061903e-05,
+      "loss": 1.75,
+      "step": 482
+    },
+    {
+      "epoch": 0.050171392957307576,
+      "grad_norm": 0.38480523228645325,
+      "learning_rate": 9.938019828878155e-05,
+      "loss": 1.9639,
+      "step": 483
+    },
+    {
+      "epoch": 0.050275267476887917,
+      "grad_norm": 0.3436914384365082,
+      "learning_rate": 9.937763450833813e-05,
+      "loss": 1.8305,
+      "step": 484
+    },
+    {
+      "epoch": 0.050379141996468264,
+      "grad_norm": 0.3643799126148224,
+      "learning_rate": 9.93750654695618e-05,
+      "loss": 1.8667,
+      "step": 485
+    },
+    {
+      "epoch": 0.05048301651604861,
+      "grad_norm": 0.3727877140045166,
+      "learning_rate": 9.937249117272612e-05,
+      "loss": 1.8097,
+      "step": 486
+    },
+    {
+      "epoch": 0.05058689103562896,
+      "grad_norm": 0.34742361307144165,
+      "learning_rate": 9.936991161810525e-05,
+      "loss": 1.872,
+      "step": 487
+    },
+    {
+      "epoch": 0.05069076555520931,
+      "grad_norm": 0.34192922711372375,
+      "learning_rate": 9.93673268059739e-05,
+      "loss": 1.777,
+      "step": 488
+    },
+    {
+      "epoch": 0.050794640074789656,
+      "grad_norm": 0.34614014625549316,
+      "learning_rate": 9.936473673660733e-05,
+      "loss": 1.8364,
+      "step": 489
+    },
+    {
+      "epoch": 0.050898514594370003,
+      "grad_norm": 0.3394782841205597,
+      "learning_rate": 9.936214141028135e-05,
+      "loss": 1.7858,
+      "step": 490
+    },
+    {
+      "epoch": 0.05100238911395035,
+      "grad_norm": 0.3589742183685303,
+      "learning_rate": 9.935954082727235e-05,
+      "loss": 2.1136,
+      "step": 491
+    },
+    {
+      "epoch": 0.05110626363353069,
+      "grad_norm": 0.3754216134548187,
+      "learning_rate": 9.935693498785728e-05,
+      "loss": 1.8873,
+      "step": 492
+    },
+    {
+      "epoch": 0.05121013815311104,
+      "grad_norm": 0.3502977192401886,
+      "learning_rate": 9.93543238923136e-05,
+      "loss": 1.6858,
+      "step": 493
+    },
+    {
+      "epoch": 0.05131401267269139,
+      "grad_norm": 0.3583030104637146,
+      "learning_rate": 9.935170754091945e-05,
+      "loss": 1.8528,
+      "step": 494
+    },
+    {
+      "epoch": 0.051417887192271736,
+      "grad_norm": 0.3400440216064453,
+      "learning_rate": 9.934908593395338e-05,
+      "loss": 1.7657,
+      "step": 495
+    },
+    {
+      "epoch": 0.051521761711852083,
+      "grad_norm": 0.31973132491111755,
+      "learning_rate": 9.934645907169459e-05,
+      "loss": 1.7152,
+      "step": 496
+    },
+    {
+      "epoch": 0.05162563623143243,
+      "grad_norm": 0.33886149525642395,
+      "learning_rate": 9.934382695442284e-05,
+      "loss": 1.6483,
+      "step": 497
+    },
+    {
+      "epoch": 0.05172951075101278,
+      "grad_norm": 0.3560135066509247,
+      "learning_rate": 9.934118958241842e-05,
+      "loss": 1.8188,
+      "step": 498
+    },
+    {
+      "epoch": 0.05183338527059313,
+      "grad_norm": 0.32066768407821655,
+      "learning_rate": 9.933854695596216e-05,
+      "loss": 1.6657,
+      "step": 499
+    },
+    {
+      "epoch": 0.05193725979017347,
+      "grad_norm": 0.3544224798679352,
+      "learning_rate": 9.933589907533554e-05,
+      "loss": 1.7714,
+      "step": 500
+    },
+    {
+      "epoch": 0.052041134309753816,
+      "grad_norm": 0.3491422235965729,
+      "learning_rate": 9.933324594082048e-05,
+      "loss": 1.9087,
+      "step": 501
+    },
+    {
+      "epoch": 0.052145008829334163,
+      "grad_norm": 0.32746124267578125,
+      "learning_rate": 9.933058755269955e-05,
+      "loss": 1.8386,
+      "step": 502
+    },
+    {
+      "epoch": 0.05224888334891451,
+      "grad_norm": 0.32933396100997925,
+      "learning_rate": 9.932792391125583e-05,
+      "loss": 1.7782,
+      "step": 503
+    },
+    {
+      "epoch": 0.05235275786849486,
+      "grad_norm": 0.35885295271873474,
+      "learning_rate": 9.932525501677299e-05,
+      "loss": 1.8822,
+      "step": 504
+    },
+    {
+      "epoch": 0.05245663238807521,
+      "grad_norm": 0.3308489918708801,
+      "learning_rate": 9.932258086953525e-05,
+      "loss": 1.8777,
+      "step": 505
+    },
+    {
+      "epoch": 0.052560506907655555,
+      "grad_norm": 0.32857421040534973,
+      "learning_rate": 9.931990146982736e-05,
+      "loss": 1.6693,
+      "step": 506
+    },
+    {
+      "epoch": 0.052664381427235896,
+      "grad_norm": 0.37959519028663635,
+      "learning_rate": 9.931721681793468e-05,
+      "loss": 1.6766,
+      "step": 507
+    },
+    {
+      "epoch": 0.052768255946816243,
+      "grad_norm": 0.39978745579719543,
+      "learning_rate": 9.931452691414311e-05,
+      "loss": 1.9316,
+      "step": 508
+    },
+    {
+      "epoch": 0.05287213046639659,
+      "grad_norm": 0.36127495765686035,
+      "learning_rate": 9.93118317587391e-05,
+      "loss": 1.9157,
+      "step": 509
+    },
+    {
+      "epoch": 0.05297600498597694,
+      "grad_norm": 0.3422827422618866,
+      "learning_rate": 9.930913135200963e-05,
+      "loss": 1.718,
+      "step": 510
+    },
+    {
+      "epoch": 0.05307987950555729,
+      "grad_norm": 0.3710615932941437,
+      "learning_rate": 9.930642569424231e-05,
+      "loss": 1.7702,
+      "step": 511
+    },
+    {
+      "epoch": 0.053183754025137635,
+      "grad_norm": 0.3597875237464905,
+      "learning_rate": 9.930371478572526e-05,
+      "loss": 1.6665,
+      "step": 512
+    },
+    {
+      "epoch": 0.05328762854471798,
+      "grad_norm": 0.3505166172981262,
+      "learning_rate": 9.930099862674716e-05,
+      "loss": 1.7484,
+      "step": 513
+    },
+    {
+      "epoch": 0.05339150306429833,
+      "grad_norm": 0.343287855386734,
+      "learning_rate": 9.929827721759728e-05,
+      "loss": 1.9197,
+      "step": 514
+    },
+    {
+      "epoch": 0.05349537758387867,
+      "grad_norm": 0.35222482681274414,
+      "learning_rate": 9.92955505585654e-05,
+      "loss": 1.8467,
+      "step": 515
+    },
+    {
+      "epoch": 0.05359925210345902,
+      "grad_norm": 0.39000675082206726,
+      "learning_rate": 9.929281864994193e-05,
+      "loss": 1.8709,
+      "step": 516
+    },
+    {
+      "epoch": 0.05370312662303937,
+      "grad_norm": 0.3652053773403168,
+      "learning_rate": 9.929008149201774e-05,
+      "loss": 1.9486,
+      "step": 517
+    },
+    {
+      "epoch": 0.053807001142619715,
+      "grad_norm": 0.3311000168323517,
+      "learning_rate": 9.928733908508438e-05,
+      "loss": 1.7749,
+      "step": 518
+    },
+    {
+      "epoch": 0.05391087566220006,
+      "grad_norm": 0.3284410536289215,
+      "learning_rate": 9.928459142943385e-05,
+      "loss": 1.8424,
+      "step": 519
+    },
+    {
+      "epoch": 0.05401475018178041,
+      "grad_norm": 0.3507472574710846,
+      "learning_rate": 9.928183852535877e-05,
+      "loss": 1.8719,
+      "step": 520
+    },
+    {
+      "epoch": 0.05411862470136076,
+      "grad_norm": 0.35009628534317017,
+      "learning_rate": 9.92790803731523e-05,
+      "loss": 1.8434,
+      "step": 521
+    },
+    {
+      "epoch": 0.054222499220941106,
+      "grad_norm": 0.39859047532081604,
+      "learning_rate": 9.927631697310816e-05,
+      "loss": 1.9473,
+      "step": 522
+    },
+    {
+      "epoch": 0.05432637374052145,
+      "grad_norm": 0.369045615196228,
+      "learning_rate": 9.927354832552063e-05,
+      "loss": 1.9063,
+      "step": 523
+    },
+    {
+      "epoch": 0.054430248260101795,
+      "grad_norm": 0.3442009389400482,
+      "learning_rate": 9.927077443068455e-05,
+      "loss": 1.7318,
+      "step": 524
+    },
+    {
+      "epoch": 0.05453412277968214,
+      "grad_norm": 0.3606685996055603,
+      "learning_rate": 9.926799528889534e-05,
+      "loss": 1.8991,
+      "step": 525
+    },
+    {
+      "epoch": 0.05463799729926249,
+      "grad_norm": 0.36547592282295227,
+      "learning_rate": 9.926521090044891e-05,
+      "loss": 2.1391,
+      "step": 526
+    },
+    {
+      "epoch": 0.05474187181884284,
+      "grad_norm": 0.334673136472702,
+      "learning_rate": 9.926242126564182e-05,
+      "loss": 1.8072,
+      "step": 527
+    },
+    {
+      "epoch": 0.054845746338423186,
+      "grad_norm": 0.3335951864719391,
+      "learning_rate": 9.925962638477113e-05,
+      "loss": 1.7599,
+      "step": 528
+    },
+    {
+      "epoch": 0.054949620858003534,
+      "grad_norm": 0.33183959126472473,
+      "learning_rate": 9.925682625813446e-05,
+      "loss": 1.7878,
+      "step": 529
+    },
+    {
+      "epoch": 0.05505349537758388,
+      "grad_norm": 0.33561477065086365,
+      "learning_rate": 9.925402088603e-05,
+      "loss": 1.7764,
+      "step": 530
+    },
+    {
+      "epoch": 0.05515736989716422,
+      "grad_norm": 0.33029311895370483,
+      "learning_rate": 9.925121026875654e-05,
+      "loss": 1.7976,
+      "step": 531
+    },
+    {
+      "epoch": 0.05526124441674457,
+      "grad_norm": 0.3405066430568695,
+      "learning_rate": 9.924839440661334e-05,
+      "loss": 1.8439,
+      "step": 532
+    },
+    {
+      "epoch": 0.05536511893632492,
+      "grad_norm": 0.3600308299064636,
+      "learning_rate": 9.92455732999003e-05,
+      "loss": 1.9871,
+      "step": 533
+    },
+    {
+      "epoch": 0.055468993455905266,
+      "grad_norm": 0.4009020924568176,
+      "learning_rate": 9.924274694891782e-05,
+      "loss": 1.7941,
+      "step": 534
+    },
+    {
+      "epoch": 0.055572867975485614,
+      "grad_norm": 0.3483946919441223,
+      "learning_rate": 9.92399153539669e-05,
+      "loss": 1.7015,
+      "step": 535
+    },
+    {
+      "epoch": 0.05567674249506596,
+      "grad_norm": 0.37152403593063354,
+      "learning_rate": 9.923707851534909e-05,
+      "loss": 1.791,
+      "step": 536
+    },
+    {
+      "epoch": 0.05578061701464631,
+      "grad_norm": 0.3480570912361145,
+      "learning_rate": 9.923423643336648e-05,
+      "loss": 1.7576,
+      "step": 537
+    },
+    {
+      "epoch": 0.05588449153422666,
+      "grad_norm": 0.38511136174201965,
+      "learning_rate": 9.923138910832172e-05,
+      "loss": 1.963,
+      "step": 538
+    },
+    {
+      "epoch": 0.055988366053807,
+      "grad_norm": 0.31676754355430603,
+      "learning_rate": 9.922853654051806e-05,
+      "loss": 1.6796,
+      "step": 539
+    },
+    {
+      "epoch": 0.056092240573387346,
+      "grad_norm": 0.325538694858551,
+      "learning_rate": 9.922567873025924e-05,
+      "loss": 1.7261,
+      "step": 540
+    },
+    {
+      "epoch": 0.056196115092967694,
+      "grad_norm": 0.3761240839958191,
+      "learning_rate": 9.922281567784961e-05,
+      "loss": 1.825,
+      "step": 541
+    },
+    {
+      "epoch": 0.05629998961254804,
+      "grad_norm": 0.33543258905410767,
+      "learning_rate": 9.921994738359409e-05,
+      "loss": 1.791,
+      "step": 542
+    },
+    {
+      "epoch": 0.05640386413212839,
+      "grad_norm": 0.34951722621917725,
+      "learning_rate": 9.921707384779807e-05,
+      "loss": 1.8141,
+      "step": 543
+    },
+    {
+      "epoch": 0.05650773865170874,
+      "grad_norm": 0.34788161516189575,
+      "learning_rate": 9.92141950707676e-05,
+      "loss": 1.7599,
+      "step": 544
+    },
+    {
+      "epoch": 0.056611613171289085,
+      "grad_norm": 0.32082879543304443,
+      "learning_rate": 9.921131105280925e-05,
+      "loss": 1.5647,
+      "step": 545
+    },
+    {
+      "epoch": 0.05671548769086943,
+      "grad_norm": 0.37312445044517517,
+      "learning_rate": 9.920842179423013e-05,
+      "loss": 1.9926,
+      "step": 546
+    },
+    {
+      "epoch": 0.056819362210449774,
+      "grad_norm": 0.32151979207992554,
+      "learning_rate": 9.920552729533792e-05,
+      "loss": 1.6564,
+      "step": 547
+    },
+    {
+      "epoch": 0.05692323673003012,
+      "grad_norm": 0.3291560709476471,
+      "learning_rate": 9.92026275564409e-05,
+      "loss": 1.8871,
+      "step": 548
+    },
+    {
+      "epoch": 0.05702711124961047,
+      "grad_norm": 0.370978444814682,
+      "learning_rate": 9.919972257784783e-05,
+      "loss": 2.0246,
+      "step": 549
+    },
+    {
+      "epoch": 0.05713098576919082,
+      "grad_norm": 0.3654458820819855,
+      "learning_rate": 9.919681235986806e-05,
+      "loss": 1.8523,
+      "step": 550
+    },
+    {
+      "epoch": 0.057234860288771165,
+      "grad_norm": 0.3720274567604065,
+      "learning_rate": 9.919389690281154e-05,
+      "loss": 1.7427,
+      "step": 551
+    },
+    {
+      "epoch": 0.05733873480835151,
+      "grad_norm": 0.3372805714607239,
+      "learning_rate": 9.919097620698872e-05,
+      "loss": 1.8491,
+      "step": 552
+    },
+    {
+      "epoch": 0.05744260932793186,
+      "grad_norm": 0.3331336975097656,
+      "learning_rate": 9.918805027271064e-05,
+      "loss": 1.7309,
+      "step": 553
+    },
+    {
+      "epoch": 0.05754648384751221,
+      "grad_norm": 0.3529910743236542,
+      "learning_rate": 9.918511910028888e-05,
+      "loss": 1.8213,
+      "step": 554
+    },
+    {
+      "epoch": 0.05765035836709255,
+      "grad_norm": 0.3342648446559906,
+      "learning_rate": 9.918218269003561e-05,
+      "loss": 1.6017,
+      "step": 555
+    },
+    {
+      "epoch": 0.0577542328866729,
+      "grad_norm": 0.41469427943229675,
+      "learning_rate": 9.917924104226351e-05,
+      "loss": 1.7434,
+      "step": 556
+    },
+    {
+      "epoch": 0.057858107406253245,
+      "grad_norm": 0.3674336373806,
+      "learning_rate": 9.917629415728584e-05,
+      "loss": 1.8997,
+      "step": 557
+    },
+    {
+      "epoch": 0.05796198192583359,
+      "grad_norm": 0.3625181317329407,
+      "learning_rate": 9.917334203541645e-05,
+      "loss": 1.8736,
+      "step": 558
+    },
+    {
+      "epoch": 0.05806585644541394,
+      "grad_norm": 0.35162436962127686,
+      "learning_rate": 9.917038467696968e-05,
+      "loss": 1.8951,
+      "step": 559
+    },
+    {
+      "epoch": 0.05816973096499429,
+      "grad_norm": 0.35170653462409973,
+      "learning_rate": 9.916742208226051e-05,
+      "loss": 1.9565,
+      "step": 560
+    },
+    {
+      "epoch": 0.058273605484574637,
+      "grad_norm": 0.3626965284347534,
+      "learning_rate": 9.91644542516044e-05,
+      "loss": 1.9673,
+      "step": 561
+    },
+    {
+      "epoch": 0.05837748000415498,
+      "grad_norm": 0.36217907071113586,
+      "learning_rate": 9.91614811853174e-05,
+      "loss": 2.0416,
+      "step": 562
+    },
+    {
+      "epoch": 0.058481354523735325,
+      "grad_norm": 0.3251532018184662,
+      "learning_rate": 9.915850288371616e-05,
+      "loss": 1.8444,
+      "step": 563
+    },
+    {
+      "epoch": 0.05858522904331567,
+      "grad_norm": 0.3693266212940216,
+      "learning_rate": 9.915551934711777e-05,
+      "loss": 1.7398,
+      "step": 564
+    },
+    {
+      "epoch": 0.05868910356289602,
+      "grad_norm": 0.34907597303390503,
+      "learning_rate": 9.915253057584003e-05,
+      "loss": 1.5016,
+      "step": 565
+    },
+    {
+      "epoch": 0.05879297808247637,
+      "grad_norm": 0.3413955867290497,
+      "learning_rate": 9.914953657020118e-05,
+      "loss": 1.7225,
+      "step": 566
+    },
+    {
+      "epoch": 0.058896852602056717,
+      "grad_norm": 0.34406933188438416,
+      "learning_rate": 9.914653733052006e-05,
+      "loss": 1.7444,
+      "step": 567
+    },
+    {
+      "epoch": 0.059000727121637064,
+      "grad_norm": 0.36242082715034485,
+      "learning_rate": 9.914353285711607e-05,
+      "loss": 1.8678,
+      "step": 568
+    },
+    {
+      "epoch": 0.05910460164121741,
+      "grad_norm": 0.359131395816803,
+      "learning_rate": 9.914052315030917e-05,
+      "loss": 1.9416,
+      "step": 569
+    },
+    {
+      "epoch": 0.05920847616079775,
+      "grad_norm": 0.35352620482444763,
+      "learning_rate": 9.913750821041987e-05,
+      "loss": 1.8087,
+      "step": 570
+    },
+    {
+      "epoch": 0.0593123506803781,
+      "grad_norm": 0.3557063043117523,
+      "learning_rate": 9.913448803776922e-05,
+      "loss": 1.9329,
+      "step": 571
+    },
+    {
+      "epoch": 0.05941622519995845,
+      "grad_norm": 0.3462260067462921,
+      "learning_rate": 9.913146263267887e-05,
+      "loss": 1.882,
+      "step": 572
+    },
+    {
+      "epoch": 0.059520099719538797,
+      "grad_norm": 0.32889804244041443,
+      "learning_rate": 9.912843199547099e-05,
+      "loss": 1.6253,
+      "step": 573
+    },
+    {
+      "epoch": 0.059623974239119144,
+      "grad_norm": 0.3543769419193268,
+      "learning_rate": 9.912539612646832e-05,
+      "loss": 1.7809,
+      "step": 574
+    },
+    {
+      "epoch": 0.05972784875869949,
+      "grad_norm": 0.38579022884368896,
+      "learning_rate": 9.912235502599414e-05,
+      "loss": 2.0883,
+      "step": 575
+    },
+    {
+      "epoch": 0.05983172327827984,
+      "grad_norm": 0.3288674056529999,
+      "learning_rate": 9.911930869437233e-05,
+      "loss": 1.7313,
+      "step": 576
+    },
+    {
+      "epoch": 0.05993559779786019,
+      "grad_norm": 0.35373395681381226,
+      "learning_rate": 9.911625713192729e-05,
+      "loss": 1.906,
+      "step": 577
+    },
+    {
+      "epoch": 0.06003947231744053,
+      "grad_norm": 0.33762502670288086,
+      "learning_rate": 9.911320033898398e-05,
+      "loss": 1.8296,
+      "step": 578
+    },
+    {
+      "epoch": 0.060143346837020877,
+      "grad_norm": 0.3767317831516266,
+      "learning_rate": 9.911013831586795e-05,
+      "loss": 1.8199,
+      "step": 579
+    },
+    {
+      "epoch": 0.060247221356601224,
+      "grad_norm": 0.38876909017562866,
+      "learning_rate": 9.910707106290525e-05,
+      "loss": 1.6965,
+      "step": 580
+    },
+    {
+      "epoch": 0.06035109587618157,
+      "grad_norm": 0.34621986746788025,
+      "learning_rate": 9.910399858042252e-05,
+      "loss": 1.7974,
+      "step": 581
+    },
+    {
+      "epoch": 0.06045497039576192,
+      "grad_norm": 0.3526759743690491,
+      "learning_rate": 9.910092086874699e-05,
+      "loss": 1.7587,
+      "step": 582
+    },
+    {
+      "epoch": 0.06055884491534227,
+      "grad_norm": 0.3554968237876892,
+      "learning_rate": 9.909783792820637e-05,
+      "loss": 1.7185,
+      "step": 583
+    },
+    {
+      "epoch": 0.060662719434922616,
+      "grad_norm": 0.34225958585739136,
+      "learning_rate": 9.909474975912899e-05,
+      "loss": 1.9277,
+      "step": 584
+    },
+    {
+      "epoch": 0.060766593954502963,
+      "grad_norm": 0.35381531715393066,
+      "learning_rate": 9.909165636184372e-05,
+      "loss": 1.8624,
+      "step": 585
+    },
+    {
+      "epoch": 0.060870468474083304,
+      "grad_norm": 0.3494366407394409,
+      "learning_rate": 9.908855773667997e-05,
+      "loss": 1.8565,
+      "step": 586
+    },
+    {
+      "epoch": 0.06097434299366365,
+      "grad_norm": 0.358967661857605,
+      "learning_rate": 9.908545388396775e-05,
+      "loss": 1.8135,
+      "step": 587
+    },
+    {
+      "epoch": 0.061078217513244,
+      "grad_norm": 0.32239392399787903,
+      "learning_rate": 9.908234480403755e-05,
+      "loss": 1.7439,
+      "step": 588
+    },
+    {
+      "epoch": 0.06118209203282435,
+      "grad_norm": 0.33216288685798645,
+      "learning_rate": 9.90792304972205e-05,
+      "loss": 1.8637,
+      "step": 589
+    },
+    {
+      "epoch": 0.061285966552404696,
+      "grad_norm": 0.3340390622615814,
+      "learning_rate": 9.907611096384823e-05,
+      "loss": 1.6959,
+      "step": 590
+    },
+    {
+      "epoch": 0.061389841071985043,
+      "grad_norm": 0.3495895564556122,
+      "learning_rate": 9.907298620425294e-05,
+      "loss": 1.7969,
+      "step": 591
+    },
+    {
+      "epoch": 0.06149371559156539,
+      "grad_norm": 0.365520715713501,
+      "learning_rate": 9.90698562187674e-05,
+      "loss": 1.8457,
+      "step": 592
+    },
+    {
+      "epoch": 0.06159759011114574,
+      "grad_norm": 0.33886629343032837,
+      "learning_rate": 9.906672100772496e-05,
+      "loss": 1.6876,
+      "step": 593
+    },
+    {
+      "epoch": 0.06170146463072608,
+      "grad_norm": 0.3432716131210327,
+      "learning_rate": 9.906358057145945e-05,
+      "loss": 1.7644,
+      "step": 594
+    },
+    {
+      "epoch": 0.06180533915030643,
+      "grad_norm": 0.35947611927986145,
+      "learning_rate": 9.906043491030532e-05,
+      "loss": 1.9074,
+      "step": 595
+    },
+    {
+      "epoch": 0.061909213669886776,
+      "grad_norm": 0.3474954664707184,
+      "learning_rate": 9.905728402459755e-05,
+      "loss": 1.834,
+      "step": 596
+    },
+    {
+      "epoch": 0.062013088189467123,
+      "grad_norm": 0.32925835251808167,
+      "learning_rate": 9.905412791467171e-05,
+      "loss": 1.7292,
+      "step": 597
+    },
+    {
+      "epoch": 0.06211696270904747,
+      "grad_norm": 0.37957310676574707,
+      "learning_rate": 9.905096658086387e-05,
+      "loss": 2.0142,
+      "step": 598
+    },
+    {
+      "epoch": 0.06222083722862782,
+      "grad_norm": 0.3465014398097992,
+      "learning_rate": 9.904780002351071e-05,
+      "loss": 1.852,
+      "step": 599
+    },
+    {
+      "epoch": 0.06232471174820817,
+      "grad_norm": 0.364199697971344,
+      "learning_rate": 9.904462824294944e-05,
+      "loss": 1.7035,
+      "step": 600
+    },
+    {
+      "epoch": 0.062428586267788515,
+      "grad_norm": 0.33271247148513794,
+      "learning_rate": 9.904145123951781e-05,
+      "loss": 1.7046,
+      "step": 601
+    },
+    {
+      "epoch": 0.06253246078736886,
+      "grad_norm": 0.38924211263656616,
+      "learning_rate": 9.903826901355417e-05,
+      "loss": 1.8107,
+      "step": 602
+    },
+    {
+      "epoch": 0.0626363353069492,
+      "grad_norm": 0.39860212802886963,
+      "learning_rate": 9.90350815653974e-05,
+      "loss": 1.8613,
+      "step": 603
+    },
+    {
+      "epoch": 0.06274020982652956,
+      "grad_norm": 0.335452139377594,
+      "learning_rate": 9.903188889538692e-05,
+      "loss": 1.6605,
+      "step": 604
+    },
+    {
+      "epoch": 0.0628440843461099,
+      "grad_norm": 0.3518449664115906,
+      "learning_rate": 9.902869100386273e-05,
+      "loss": 1.8248,
+      "step": 605
+    },
+    {
+      "epoch": 0.06294795886569024,
+      "grad_norm": 0.36134716868400574,
+      "learning_rate": 9.90254878911654e-05,
+      "loss": 1.8456,
+      "step": 606
+    },
+    {
+      "epoch": 0.0630518333852706,
+      "grad_norm": 0.3638603687286377,
+      "learning_rate": 9.902227955763602e-05,
+      "loss": 1.6454,
+      "step": 607
+    },
+    {
+      "epoch": 0.06315570790485094,
+      "grad_norm": 0.3790292739868164,
+      "learning_rate": 9.901906600361625e-05,
+      "loss": 1.8036,
+      "step": 608
+    },
+    {
+      "epoch": 0.06325958242443129,
+      "grad_norm": 0.3539649546146393,
+      "learning_rate": 9.901584722944833e-05,
+      "loss": 1.8153,
+      "step": 609
+    },
+    {
+      "epoch": 0.06336345694401163,
+      "grad_norm": 0.33538633584976196,
+      "learning_rate": 9.9012623235475e-05,
+      "loss": 1.7256,
+      "step": 610
+    },
+    {
+      "epoch": 0.06346733146359199,
+      "grad_norm": 0.3604874610900879,
+      "learning_rate": 9.900939402203962e-05,
+      "loss": 2.0305,
+      "step": 611
+    },
+    {
+      "epoch": 0.06357120598317233,
+      "grad_norm": 0.34128594398498535,
+      "learning_rate": 9.900615958948606e-05,
+      "loss": 1.7801,
+      "step": 612
+    },
+    {
+      "epoch": 0.06367508050275267,
+      "grad_norm": 0.3321555554866791,
+      "learning_rate": 9.900291993815877e-05,
+      "loss": 1.6963,
+      "step": 613
+    },
+    {
+      "epoch": 0.06377895502233302,
+      "grad_norm": 0.32713523507118225,
+      "learning_rate": 9.899967506840273e-05,
+      "loss": 1.8211,
+      "step": 614
+    },
+    {
+      "epoch": 0.06388282954191336,
+      "grad_norm": 0.34873101115226746,
+      "learning_rate": 9.899642498056353e-05,
+      "loss": 1.8328,
+      "step": 615
+    },
+    {
+      "epoch": 0.06398670406149372,
+      "grad_norm": 0.36176908016204834,
+      "learning_rate": 9.899316967498724e-05,
+      "loss": 1.9433,
+      "step": 616
+    },
+    {
+      "epoch": 0.06409057858107406,
+      "grad_norm": 0.35845959186553955,
+      "learning_rate": 9.898990915202055e-05,
+      "loss": 1.8383,
+      "step": 617
+    },
+    {
+      "epoch": 0.06419445310065441,
+      "grad_norm": 0.3352903425693512,
+      "learning_rate": 9.898664341201067e-05,
+      "loss": 1.8279,
+      "step": 618
+    },
+    {
+      "epoch": 0.06429832762023475,
+      "grad_norm": 0.3244091272354126,
+      "learning_rate": 9.898337245530536e-05,
+      "loss": 1.6484,
+      "step": 619
+    },
+    {
+      "epoch": 0.06440220213981511,
+      "grad_norm": 0.39862748980522156,
+      "learning_rate": 9.8980096282253e-05,
+      "loss": 1.9679,
+      "step": 620
+    },
+    {
+      "epoch": 0.06450607665939545,
+      "grad_norm": 0.34560245275497437,
+      "learning_rate": 9.897681489320242e-05,
+      "loss": 1.8297,
+      "step": 621
+    },
+    {
+      "epoch": 0.06460995117897579,
+      "grad_norm": 0.37943777441978455,
+      "learning_rate": 9.897352828850309e-05,
+      "loss": 1.9845,
+      "step": 622
+    },
+    {
+      "epoch": 0.06471382569855615,
+      "grad_norm": 0.3497175872325897,
+      "learning_rate": 9.897023646850502e-05,
+      "loss": 1.7576,
+      "step": 623
+    },
+    {
+      "epoch": 0.06481770021813649,
+      "grad_norm": 0.3344300091266632,
+      "learning_rate": 9.896693943355874e-05,
+      "loss": 1.7502,
+      "step": 624
+    },
+    {
+      "epoch": 0.06492157473771684,
+      "grad_norm": 0.3299304246902466,
+      "learning_rate": 9.896363718401536e-05,
+      "loss": 1.5323,
+      "step": 625
+    },
+    {
+      "epoch": 0.06502544925729718,
+      "grad_norm": 0.3126195967197418,
+      "learning_rate": 9.896032972022656e-05,
+      "loss": 1.5328,
+      "step": 626
+    },
+    {
+      "epoch": 0.06512932377687754,
+      "grad_norm": 0.33860689401626587,
+      "learning_rate": 9.895701704254455e-05,
+      "loss": 1.7393,
+      "step": 627
+    },
+    {
+      "epoch": 0.06523319829645788,
+      "grad_norm": 0.332501083612442,
+      "learning_rate": 9.89536991513221e-05,
+      "loss": 1.6363,
+      "step": 628
+    },
+    {
+      "epoch": 0.06533707281603822,
+      "grad_norm": 0.32221221923828125,
+      "learning_rate": 9.895037604691256e-05,
+      "loss": 1.6374,
+      "step": 629
+    },
+    {
+      "epoch": 0.06544094733561857,
+      "grad_norm": 0.381583034992218,
+      "learning_rate": 9.894704772966977e-05,
+      "loss": 1.773,
+      "step": 630
+    },
+    {
+      "epoch": 0.06554482185519891,
+      "grad_norm": 0.36580127477645874,
+      "learning_rate": 9.894371419994821e-05,
+      "loss": 1.7417,
+      "step": 631
+    },
+    {
+      "epoch": 0.06564869637477927,
+      "grad_norm": 0.3486615717411041,
+      "learning_rate": 9.894037545810288e-05,
+      "loss": 1.8542,
+      "step": 632
+    },
+    {
+      "epoch": 0.06575257089435961,
+      "grad_norm": 0.3709062337875366,
+      "learning_rate": 9.89370315044893e-05,
+      "loss": 1.917,
+      "step": 633
+    },
+    {
+      "epoch": 0.06585644541393997,
+      "grad_norm": 0.37561196088790894,
+      "learning_rate": 9.893368233946359e-05,
+      "loss": 1.9253,
+      "step": 634
+    },
+    {
+      "epoch": 0.0659603199335203,
+      "grad_norm": 0.32539495825767517,
+      "learning_rate": 9.89303279633824e-05,
+      "loss": 1.6948,
+      "step": 635
+    },
+    {
+      "epoch": 0.06606419445310066,
+      "grad_norm": 0.32918450236320496,
+      "learning_rate": 9.892696837660295e-05,
+      "loss": 1.8504,
+      "step": 636
+    },
+    {
+      "epoch": 0.066168068972681,
+      "grad_norm": 0.32979822158813477,
+      "learning_rate": 9.892360357948303e-05,
+      "loss": 1.7728,
+      "step": 637
+    },
+    {
+      "epoch": 0.06627194349226134,
+      "grad_norm": 0.39485815167427063,
+      "learning_rate": 9.892023357238094e-05,
+      "loss": 1.8216,
+      "step": 638
+    },
+    {
+      "epoch": 0.0663758180118417,
+      "grad_norm": 0.3529796302318573,
+      "learning_rate": 9.891685835565557e-05,
+      "loss": 1.725,
+      "step": 639
+    },
+    {
+      "epoch": 0.06647969253142204,
+      "grad_norm": 0.34411367774009705,
+      "learning_rate": 9.891347792966635e-05,
+      "loss": 1.7718,
+      "step": 640
+    },
+    {
+      "epoch": 0.06658356705100239,
+      "grad_norm": 0.38490092754364014,
+      "learning_rate": 9.891009229477326e-05,
+      "loss": 1.9705,
+      "step": 641
+    },
+    {
+      "epoch": 0.06668744157058273,
+      "grad_norm": 0.3569811284542084,
+      "learning_rate": 9.890670145133687e-05,
+      "loss": 2.0534,
+      "step": 642
+    },
+    {
+      "epoch": 0.06679131609016309,
+      "grad_norm": 0.3439379334449768,
+      "learning_rate": 9.890330539971825e-05,
+      "loss": 1.9179,
+      "step": 643
+    },
+    {
+      "epoch": 0.06689519060974343,
+      "grad_norm": 0.3299177587032318,
+      "learning_rate": 9.889990414027907e-05,
+      "loss": 1.8077,
+      "step": 644
+    },
+    {
+      "epoch": 0.06699906512932377,
+      "grad_norm": 0.3420833647251129,
+      "learning_rate": 9.889649767338153e-05,
+      "loss": 1.6712,
+      "step": 645
+    },
+    {
+      "epoch": 0.06710293964890413,
+      "grad_norm": 0.3195805549621582,
+      "learning_rate": 9.88930859993884e-05,
+      "loss": 1.629,
+      "step": 646
+    },
+    {
+      "epoch": 0.06720681416848447,
+      "grad_norm": 0.3861030638217926,
+      "learning_rate": 9.888966911866299e-05,
+      "loss": 2.02,
+      "step": 647
+    },
+    {
+      "epoch": 0.06731068868806482,
+      "grad_norm": 0.3432799279689789,
+      "learning_rate": 9.888624703156918e-05,
+      "loss": 1.6907,
+      "step": 648
+    },
+    {
+      "epoch": 0.06741456320764516,
+      "grad_norm": 0.33698904514312744,
+      "learning_rate": 9.888281973847138e-05,
+      "loss": 1.6967,
+      "step": 649
+    },
+    {
+      "epoch": 0.06751843772722552,
+      "grad_norm": 0.3321373462677002,
+      "learning_rate": 9.887938723973458e-05,
+      "loss": 1.7385,
+      "step": 650
+    },
+    {
+      "epoch": 0.06762231224680586,
+      "grad_norm": 0.3467056155204773,
+      "learning_rate": 9.887594953572432e-05,
+      "loss": 1.6206,
+      "step": 651
+    },
+    {
+      "epoch": 0.06772618676638621,
+      "grad_norm": 0.3018084764480591,
+      "learning_rate": 9.887250662680667e-05,
+      "loss": 1.6185,
+      "step": 652
+    },
+    {
+      "epoch": 0.06783006128596655,
+      "grad_norm": 0.37331244349479675,
+      "learning_rate": 9.886905851334829e-05,
+      "loss": 1.8004,
+      "step": 653
+    },
+    {
+      "epoch": 0.0679339358055469,
+      "grad_norm": 0.36824825406074524,
+      "learning_rate": 9.886560519571637e-05,
+      "loss": 1.7545,
+      "step": 654
+    },
+    {
+      "epoch": 0.06803781032512725,
+      "grad_norm": 0.33622682094573975,
+      "learning_rate": 9.886214667427867e-05,
+      "loss": 1.7991,
+      "step": 655
+    },
+    {
+      "epoch": 0.06814168484470759,
+      "grad_norm": 0.3627051115036011,
+      "learning_rate": 9.885868294940349e-05,
+      "loss": 1.7288,
+      "step": 656
+    },
+    {
+      "epoch": 0.06824555936428794,
+      "grad_norm": 0.3434629440307617,
+      "learning_rate": 9.885521402145967e-05,
+      "loss": 1.7928,
+      "step": 657
+    },
+    {
+      "epoch": 0.06834943388386829,
+      "grad_norm": 0.3145413100719452,
+      "learning_rate": 9.885173989081664e-05,
+      "loss": 1.6569,
+      "step": 658
+    },
+    {
+      "epoch": 0.06845330840344864,
+      "grad_norm": 0.3603506088256836,
+      "learning_rate": 9.88482605578444e-05,
+      "loss": 1.9213,
+      "step": 659
+    },
+    {
+      "epoch": 0.06855718292302898,
+      "grad_norm": 0.33825626969337463,
+      "learning_rate": 9.884477602291343e-05,
+      "loss": 1.6129,
+      "step": 660
+    },
+    {
+      "epoch": 0.06866105744260932,
+      "grad_norm": 0.35486510396003723,
+      "learning_rate": 9.88412862863948e-05,
+      "loss": 1.8986,
+      "step": 661
+    },
+    {
+      "epoch": 0.06876493196218968,
+      "grad_norm": 0.3892877399921417,
+      "learning_rate": 9.883779134866016e-05,
+      "loss": 2.0618,
+      "step": 662
+    },
+    {
+      "epoch": 0.06886880648177002,
+      "grad_norm": 0.3327191472053528,
+      "learning_rate": 9.883429121008169e-05,
+      "loss": 1.6847,
+      "step": 663
+    },
+    {
+      "epoch": 0.06897268100135037,
+      "grad_norm": 0.32258662581443787,
+      "learning_rate": 9.883078587103215e-05,
+      "loss": 1.7325,
+      "step": 664
+    },
+    {
+      "epoch": 0.06907655552093071,
+      "grad_norm": 0.3689654767513275,
+      "learning_rate": 9.882727533188481e-05,
+      "loss": 2.0511,
+      "step": 665
+    },
+    {
+      "epoch": 0.06918043004051107,
+      "grad_norm": 0.3166467547416687,
+      "learning_rate": 9.882375959301349e-05,
+      "loss": 1.7457,
+      "step": 666
+    },
+    {
+      "epoch": 0.06928430456009141,
+      "grad_norm": 0.3648569583892822,
+      "learning_rate": 9.882023865479263e-05,
+      "loss": 1.9451,
+      "step": 667
+    },
+    {
+      "epoch": 0.06938817907967175,
+      "grad_norm": 0.35240545868873596,
+      "learning_rate": 9.881671251759717e-05,
+      "loss": 1.9018,
+      "step": 668
+    },
+    {
+      "epoch": 0.0694920535992521,
+      "grad_norm": 0.3445391356945038,
+      "learning_rate": 9.881318118180258e-05,
+      "loss": 1.9383,
+      "step": 669
+    },
+    {
+      "epoch": 0.06959592811883245,
+      "grad_norm": 0.33014625310897827,
+      "learning_rate": 9.880964464778499e-05,
+      "loss": 1.6511,
+      "step": 670
+    },
+    {
+      "epoch": 0.0696998026384128,
+      "grad_norm": 0.3588254451751709,
+      "learning_rate": 9.880610291592096e-05,
+      "loss": 1.7165,
+      "step": 671
+    },
+    {
+      "epoch": 0.06980367715799314,
+      "grad_norm": 0.3672218322753906,
+      "learning_rate": 9.880255598658767e-05,
+      "loss": 1.6612,
+      "step": 672
+    },
+    {
+      "epoch": 0.0699075516775735,
+      "grad_norm": 0.3593609035015106,
+      "learning_rate": 9.879900386016284e-05,
+      "loss": 1.8946,
+      "step": 673
+    },
+    {
+      "epoch": 0.07001142619715384,
+      "grad_norm": 0.45219045877456665,
+      "learning_rate": 9.879544653702476e-05,
+      "loss": 1.8398,
+      "step": 674
+    },
+    {
+      "epoch": 0.07011530071673419,
+      "grad_norm": 0.3960374891757965,
+      "learning_rate": 9.879188401755222e-05,
+      "loss": 1.928,
+      "step": 675
+    },
+    {
+      "epoch": 0.07021917523631453,
+      "grad_norm": 0.38661623001098633,
+      "learning_rate": 9.878831630212465e-05,
+      "loss": 1.9605,
+      "step": 676
+    },
+    {
+      "epoch": 0.07032304975589487,
+      "grad_norm": 0.34862831234931946,
+      "learning_rate": 9.878474339112193e-05,
+      "loss": 1.7933,
+      "step": 677
+    },
+    {
+      "epoch": 0.07042692427547523,
+      "grad_norm": 0.3596080243587494,
+      "learning_rate": 9.878116528492458e-05,
+      "loss": 1.7487,
+      "step": 678
+    },
+    {
+      "epoch": 0.07053079879505557,
+      "grad_norm": 0.3401545584201813,
+      "learning_rate": 9.877758198391365e-05,
+      "loss": 1.8613,
+      "step": 679
+    },
+    {
+      "epoch": 0.07063467331463592,
+      "grad_norm": 0.3352885842323303,
+      "learning_rate": 9.87739934884707e-05,
+      "loss": 1.6381,
+      "step": 680
+    },
+    {
+      "epoch": 0.07073854783421626,
+      "grad_norm": 0.36678406596183777,
+      "learning_rate": 9.87703997989779e-05,
+      "loss": 1.944,
+      "step": 681
+    },
+    {
+      "epoch": 0.07084242235379662,
+      "grad_norm": 0.3366321623325348,
+      "learning_rate": 9.876680091581796e-05,
+      "loss": 1.7308,
+      "step": 682
+    },
+    {
+      "epoch": 0.07094629687337696,
+      "grad_norm": 0.37314730882644653,
+      "learning_rate": 9.876319683937412e-05,
+      "loss": 1.8055,
+      "step": 683
+    },
+    {
+      "epoch": 0.0710501713929573,
+      "grad_norm": 0.3612748384475708,
+      "learning_rate": 9.875958757003017e-05,
+      "loss": 1.7043,
+      "step": 684
+    },
+    {
+      "epoch": 0.07115404591253766,
+      "grad_norm": 0.37414270639419556,
+      "learning_rate": 9.875597310817049e-05,
+      "loss": 1.8259,
+      "step": 685
+    },
+    {
+      "epoch": 0.071257920432118,
+      "grad_norm": 0.3539772033691406,
+      "learning_rate": 9.875235345417997e-05,
+      "loss": 1.7541,
+      "step": 686
+    },
+    {
+      "epoch": 0.07136179495169835,
+      "grad_norm": 0.35429245233535767,
+      "learning_rate": 9.874872860844411e-05,
+      "loss": 1.5966,
+      "step": 687
+    },
+    {
+      "epoch": 0.07146566947127869,
+      "grad_norm": 0.32633036375045776,
+      "learning_rate": 9.87450985713489e-05,
+      "loss": 1.7182,
+      "step": 688
+    },
+    {
+      "epoch": 0.07156954399085905,
+      "grad_norm": 0.316303014755249,
+      "learning_rate": 9.874146334328092e-05,
+      "loss": 1.553,
+      "step": 689
+    },
+    {
+      "epoch": 0.07167341851043939,
+      "grad_norm": 0.39488279819488525,
+      "learning_rate": 9.873782292462727e-05,
+      "loss": 1.9359,
+      "step": 690
+    },
+    {
+      "epoch": 0.07177729303001974,
+      "grad_norm": 0.3820980489253998,
+      "learning_rate": 9.873417731577568e-05,
+      "loss": 1.9939,
+      "step": 691
+    },
+    {
+      "epoch": 0.07188116754960008,
+      "grad_norm": 0.4182794392108917,
+      "learning_rate": 9.873052651711434e-05,
+      "loss": 1.8192,
+      "step": 692
+    },
+    {
+      "epoch": 0.07198504206918042,
+      "grad_norm": 0.3251241147518158,
+      "learning_rate": 9.872687052903203e-05,
+      "loss": 1.6569,
+      "step": 693
+    },
+    {
+      "epoch": 0.07208891658876078,
+      "grad_norm": 0.33113572001457214,
+      "learning_rate": 9.87232093519181e-05,
+      "loss": 1.743,
+      "step": 694
+    },
+    {
+      "epoch": 0.07219279110834112,
+      "grad_norm": 0.3508453369140625,
+      "learning_rate": 9.871954298616242e-05,
+      "loss": 1.7822,
+      "step": 695
+    },
+    {
+      "epoch": 0.07229666562792147,
+      "grad_norm": 0.36456069350242615,
+      "learning_rate": 9.871587143215545e-05,
+      "loss": 1.7646,
+      "step": 696
+    },
+    {
+      "epoch": 0.07240054014750182,
+      "grad_norm": 0.3450940251350403,
+      "learning_rate": 9.871219469028816e-05,
+      "loss": 1.798,
+      "step": 697
+    },
+    {
+      "epoch": 0.07250441466708217,
+      "grad_norm": 0.3370673358440399,
+      "learning_rate": 9.870851276095212e-05,
+      "loss": 1.757,
+      "step": 698
+    },
+    {
+      "epoch": 0.07260828918666251,
+      "grad_norm": 0.34746289253234863,
+      "learning_rate": 9.87048256445394e-05,
+      "loss": 1.6328,
+      "step": 699
+    },
+    {
+      "epoch": 0.07271216370624285,
+      "grad_norm": 0.3589167296886444,
+      "learning_rate": 9.870113334144267e-05,
+      "loss": 1.7512,
+      "step": 700
+    },
+    {
+      "epoch": 0.07281603822582321,
+      "grad_norm": 0.4009071886539459,
+      "learning_rate": 9.869743585205511e-05,
+      "loss": 1.8032,
+      "step": 701
+    },
+    {
+      "epoch": 0.07291991274540355,
+      "grad_norm": 0.38780277967453003,
+      "learning_rate": 9.869373317677049e-05,
+      "loss": 1.8526,
+      "step": 702
+    },
+    {
+      "epoch": 0.0730237872649839,
+      "grad_norm": 0.37040579319000244,
+      "learning_rate": 9.869002531598312e-05,
+      "loss": 1.8994,
+      "step": 703
+    },
+    {
+      "epoch": 0.07312766178456424,
+      "grad_norm": 0.3585062325000763,
+      "learning_rate": 9.868631227008785e-05,
+      "loss": 1.853,
+      "step": 704
+    },
+    {
+      "epoch": 0.0732315363041446,
+      "grad_norm": 0.3617033362388611,
+      "learning_rate": 9.868259403948008e-05,
+      "loss": 1.9527,
+      "step": 705
+    },
+    {
+      "epoch": 0.07333541082372494,
+      "grad_norm": 0.35170087218284607,
+      "learning_rate": 9.867887062455579e-05,
+      "loss": 1.7766,
+      "step": 706
+    },
+    {
+      "epoch": 0.0734392853433053,
+      "grad_norm": 0.34852275252342224,
+      "learning_rate": 9.867514202571149e-05,
+      "loss": 1.5639,
+      "step": 707
+    },
+    {
+      "epoch": 0.07354315986288563,
+      "grad_norm": 0.3750884532928467,
+      "learning_rate": 9.867140824334424e-05,
+      "loss": 1.9035,
+      "step": 708
+    },
+    {
+      "epoch": 0.07364703438246598,
+      "grad_norm": 0.37743544578552246,
+      "learning_rate": 9.866766927785167e-05,
+      "loss": 1.8442,
+      "step": 709
+    },
+    {
+      "epoch": 0.07375090890204633,
+      "grad_norm": 0.3599265217781067,
+      "learning_rate": 9.866392512963195e-05,
+      "loss": 1.6519,
+      "step": 710
+    },
+    {
+      "epoch": 0.07385478342162667,
+      "grad_norm": 0.35439547896385193,
+      "learning_rate": 9.86601757990838e-05,
+      "loss": 1.8225,
+      "step": 711
+    },
+    {
+      "epoch": 0.07395865794120703,
+      "grad_norm": 0.35483318567276,
+      "learning_rate": 9.865642128660647e-05,
+      "loss": 1.778,
+      "step": 712
+    },
+    {
+      "epoch": 0.07406253246078737,
+      "grad_norm": 0.39381521940231323,
+      "learning_rate": 9.865266159259981e-05,
+      "loss": 1.9915,
+      "step": 713
+    },
+    {
+      "epoch": 0.07416640698036772,
+      "grad_norm": 0.4099736213684082,
+      "learning_rate": 9.864889671746421e-05,
+      "loss": 2.0117,
+      "step": 714
+    },
+    {
+      "epoch": 0.07427028149994806,
+      "grad_norm": 0.3895813822746277,
+      "learning_rate": 9.864512666160058e-05,
+      "loss": 2.2284,
+      "step": 715
+    },
+    {
+      "epoch": 0.0743741560195284,
+      "grad_norm": 0.3480944037437439,
+      "learning_rate": 9.86413514254104e-05,
+      "loss": 1.7284,
+      "step": 716
+    },
+    {
+      "epoch": 0.07447803053910876,
+      "grad_norm": 0.3395494818687439,
+      "learning_rate": 9.863757100929573e-05,
+      "loss": 1.7128,
+      "step": 717
+    },
+    {
+      "epoch": 0.0745819050586891,
+      "grad_norm": 0.4045158624649048,
+      "learning_rate": 9.863378541365912e-05,
+      "loss": 1.8116,
+      "step": 718
+    },
+    {
+      "epoch": 0.07468577957826945,
+      "grad_norm": 0.34248921275138855,
+      "learning_rate": 9.862999463890372e-05,
+      "loss": 1.9717,
+      "step": 719
+    },
+    {
+      "epoch": 0.0747896540978498,
+      "grad_norm": 0.33910906314849854,
+      "learning_rate": 9.862619868543322e-05,
+      "loss": 1.886,
+      "step": 720
+    },
+    {
+      "epoch": 0.07489352861743015,
+      "grad_norm": 0.33199310302734375,
+      "learning_rate": 9.862239755365186e-05,
+      "loss": 1.7202,
+      "step": 721
+    },
+    {
+      "epoch": 0.07499740313701049,
+      "grad_norm": 0.39172807335853577,
+      "learning_rate": 9.861859124396444e-05,
+      "loss": 1.9074,
+      "step": 722
+    },
+    {
+      "epoch": 0.07510127765659085,
+      "grad_norm": 0.38816559314727783,
+      "learning_rate": 9.86147797567763e-05,
+      "loss": 1.963,
+      "step": 723
+    },
+    {
+      "epoch": 0.07520515217617119,
+      "grad_norm": 0.39771151542663574,
+      "learning_rate": 9.86109630924933e-05,
+      "loss": 2.119,
+      "step": 724
+    },
+    {
+      "epoch": 0.07530902669575153,
+      "grad_norm": 0.34102803468704224,
+      "learning_rate": 9.860714125152191e-05,
+      "loss": 1.8536,
+      "step": 725
+    },
+    {
+      "epoch": 0.07541290121533188,
+      "grad_norm": 0.35469919443130493,
+      "learning_rate": 9.860331423426914e-05,
+      "loss": 1.7498,
+      "step": 726
+    },
+    {
+      "epoch": 0.07551677573491222,
+      "grad_norm": 0.3550052046775818,
+      "learning_rate": 9.859948204114253e-05,
+      "loss": 1.9569,
+      "step": 727
+    },
+    {
+      "epoch": 0.07562065025449258,
+      "grad_norm": 0.3915771245956421,
+      "learning_rate": 9.859564467255015e-05,
+      "loss": 2.0232,
+      "step": 728
+    },
+    {
+      "epoch": 0.07572452477407292,
+      "grad_norm": 0.326684832572937,
+      "learning_rate": 9.859180212890069e-05,
+      "loss": 1.6597,
+      "step": 729
+    },
+    {
+      "epoch": 0.07582839929365327,
+      "grad_norm": 0.33311307430267334,
+      "learning_rate": 9.858795441060333e-05,
+      "loss": 1.8534,
+      "step": 730
+    },
+    {
+      "epoch": 0.07593227381323361,
+      "grad_norm": 0.3527379631996155,
+      "learning_rate": 9.858410151806783e-05,
+      "loss": 1.8047,
+      "step": 731
+    },
+    {
+      "epoch": 0.07603614833281395,
+      "grad_norm": 0.3848365545272827,
+      "learning_rate": 9.85802434517045e-05,
+      "loss": 1.7222,
+      "step": 732
+    },
+    {
+      "epoch": 0.07614002285239431,
+      "grad_norm": 0.3726034462451935,
+      "learning_rate": 9.857638021192417e-05,
+      "loss": 1.9158,
+      "step": 733
+    },
+    {
+      "epoch": 0.07624389737197465,
+      "grad_norm": 0.4738229513168335,
+      "learning_rate": 9.857251179913824e-05,
+      "loss": 2.1136,
+      "step": 734
+    },
+    {
+      "epoch": 0.076347771891555,
+      "grad_norm": 0.3687533438205719,
+      "learning_rate": 9.856863821375872e-05,
+      "loss": 1.802,
+      "step": 735
+    },
+    {
+      "epoch": 0.07645164641113535,
+      "grad_norm": 0.35550445318222046,
+      "learning_rate": 9.856475945619806e-05,
+      "loss": 1.8454,
+      "step": 736
+    },
+    {
+      "epoch": 0.0765555209307157,
+      "grad_norm": 0.3472285866737366,
+      "learning_rate": 9.856087552686932e-05,
+      "loss": 1.7785,
+      "step": 737
+    },
+    {
+      "epoch": 0.07665939545029604,
+      "grad_norm": 0.3866637051105499,
+      "learning_rate": 9.855698642618616e-05,
+      "loss": 1.9353,
+      "step": 738
+    },
+    {
+      "epoch": 0.07676326996987638,
+      "grad_norm": 0.3777254819869995,
+      "learning_rate": 9.855309215456268e-05,
+      "loss": 1.9781,
+      "step": 739
+    },
+    {
+      "epoch": 0.07686714448945674,
+      "grad_norm": 0.413335919380188,
+      "learning_rate": 9.854919271241363e-05,
+      "loss": 1.917,
+      "step": 740
+    },
+    {
+      "epoch": 0.07697101900903708,
+      "grad_norm": 0.3310924470424652,
+      "learning_rate": 9.854528810015425e-05,
+      "loss": 1.6168,
+      "step": 741
+    },
+    {
+      "epoch": 0.07707489352861743,
+      "grad_norm": 0.34101495146751404,
+      "learning_rate": 9.854137831820035e-05,
+      "loss": 1.7935,
+      "step": 742
+    },
+    {
+      "epoch": 0.07717876804819777,
+      "grad_norm": 0.34622523188591003,
+      "learning_rate": 9.85374633669683e-05,
+      "loss": 1.7695,
+      "step": 743
+    },
+    {
+      "epoch": 0.07728264256777813,
+      "grad_norm": 0.39640340209007263,
+      "learning_rate": 9.8533543246875e-05,
+      "loss": 1.8163,
+      "step": 744
+    },
+    {
+      "epoch": 0.07738651708735847,
+      "grad_norm": 0.35431715846061707,
+      "learning_rate": 9.852961795833793e-05,
+      "loss": 1.7706,
+      "step": 745
+    },
+    {
+      "epoch": 0.07749039160693882,
+      "grad_norm": 0.37833094596862793,
+      "learning_rate": 9.852568750177508e-05,
+      "loss": 1.6099,
+      "step": 746
+    },
+    {
+      "epoch": 0.07759426612651917,
+      "grad_norm": 0.3214152753353119,
+      "learning_rate": 9.852175187760504e-05,
+      "loss": 1.5333,
+      "step": 747
+    },
+    {
+      "epoch": 0.0776981406460995,
+      "grad_norm": 0.3399621546268463,
+      "learning_rate": 9.851781108624692e-05,
+      "loss": 1.7676,
+      "step": 748
+    },
+    {
+      "epoch": 0.07780201516567986,
+      "grad_norm": 0.3282534182071686,
+      "learning_rate": 9.851386512812036e-05,
+      "loss": 1.8172,
+      "step": 749
+    },
+    {
+      "epoch": 0.0779058896852602,
+      "grad_norm": 0.37723109126091003,
+      "learning_rate": 9.850991400364557e-05,
+      "loss": 1.8644,
+      "step": 750
+    },
+    {
+      "epoch": 0.07800976420484056,
+      "grad_norm": 0.36292698979377747,
+      "learning_rate": 9.850595771324335e-05,
+      "loss": 1.8879,
+      "step": 751
+    },
+    {
+      "epoch": 0.0781136387244209,
+      "grad_norm": 0.3753519654273987,
+      "learning_rate": 9.850199625733501e-05,
+      "loss": 1.9628,
+      "step": 752
+    },
+    {
+      "epoch": 0.07821751324400125,
+      "grad_norm": 0.33497151732444763,
+      "learning_rate": 9.849802963634238e-05,
+      "loss": 1.8383,
+      "step": 753
+    },
+    {
+      "epoch": 0.07832138776358159,
+      "grad_norm": 0.3661734163761139,
+      "learning_rate": 9.849405785068789e-05,
+      "loss": 1.903,
+      "step": 754
+    },
+    {
+      "epoch": 0.07842526228316193,
+      "grad_norm": 0.33177104592323303,
+      "learning_rate": 9.84900809007945e-05,
+      "loss": 1.7703,
+      "step": 755
+    },
+    {
+      "epoch": 0.07852913680274229,
+      "grad_norm": 0.41701647639274597,
+      "learning_rate": 9.848609878708575e-05,
+      "loss": 2.017,
+      "step": 756
+    },
+    {
+      "epoch": 0.07863301132232263,
+      "grad_norm": 0.3774707317352295,
+      "learning_rate": 9.848211150998568e-05,
+      "loss": 1.8528,
+      "step": 757
+    },
+    {
+      "epoch": 0.07873688584190298,
+      "grad_norm": 0.3493809700012207,
+      "learning_rate": 9.84781190699189e-05,
+      "loss": 1.8787,
+      "step": 758
+    },
+    {
+      "epoch": 0.07884076036148333,
+      "grad_norm": 0.4020559787750244,
+      "learning_rate": 9.847412146731058e-05,
+      "loss": 2.0583,
+      "step": 759
+    },
+    {
+      "epoch": 0.07894463488106368,
+      "grad_norm": 0.37172991037368774,
+      "learning_rate": 9.847011870258644e-05,
+      "loss": 1.9294,
+      "step": 760
+    },
+    {
+      "epoch": 0.07904850940064402,
+      "grad_norm": 0.3182731568813324,
+      "learning_rate": 9.846611077617274e-05,
+      "loss": 1.7817,
+      "step": 761
+    },
+    {
+      "epoch": 0.07915238392022438,
+      "grad_norm": 0.4014347791671753,
+      "learning_rate": 9.84620976884963e-05,
+      "loss": 1.8936,
+      "step": 762
+    },
+    {
+      "epoch": 0.07925625843980472,
+      "grad_norm": 0.317242294549942,
+      "learning_rate": 9.845807943998448e-05,
+      "loss": 1.6904,
+      "step": 763
+    },
+    {
+      "epoch": 0.07936013295938506,
+      "grad_norm": 0.3695957064628601,
+      "learning_rate": 9.845405603106518e-05,
+      "loss": 1.9056,
+      "step": 764
+    },
+    {
+      "epoch": 0.07946400747896541,
+      "grad_norm": 0.3249827027320862,
+      "learning_rate": 9.845002746216687e-05,
+      "loss": 1.8097,
+      "step": 765
+    },
+    {
+      "epoch": 0.07956788199854575,
+      "grad_norm": 0.3424146771430969,
+      "learning_rate": 9.844599373371855e-05,
+      "loss": 1.6326,
+      "step": 766
+    },
+    {
+      "epoch": 0.07967175651812611,
+      "grad_norm": 0.35552307963371277,
+      "learning_rate": 9.84419548461498e-05,
+      "loss": 1.9049,
+      "step": 767
+    },
+    {
+      "epoch": 0.07977563103770645,
+      "grad_norm": 0.3346835970878601,
+      "learning_rate": 9.843791079989071e-05,
+      "loss": 1.7215,
+      "step": 768
+    },
+    {
+      "epoch": 0.0798795055572868,
+      "grad_norm": 0.339160680770874,
+      "learning_rate": 9.843386159537196e-05,
+      "loss": 1.6931,
+      "step": 769
+    },
+    {
+      "epoch": 0.07998338007686714,
+      "grad_norm": 0.3619834780693054,
+      "learning_rate": 9.842980723302476e-05,
+      "loss": 1.7751,
+      "step": 770
+    },
+    {
+      "epoch": 0.08008725459644749,
+      "grad_norm": 0.37088942527770996,
+      "learning_rate": 9.842574771328084e-05,
+      "loss": 1.7259,
+      "step": 771
+    },
+    {
+      "epoch": 0.08019112911602784,
+      "grad_norm": 0.3355197012424469,
+      "learning_rate": 9.842168303657253e-05,
+      "loss": 1.8242,
+      "step": 772
+    },
+    {
+      "epoch": 0.08029500363560818,
+      "grad_norm": 0.37746527791023254,
+      "learning_rate": 9.841761320333266e-05,
+      "loss": 1.7493,
+      "step": 773
+    },
+    {
+      "epoch": 0.08039887815518854,
+      "grad_norm": 0.3684033751487732,
+      "learning_rate": 9.841353821399469e-05,
+      "loss": 1.7955,
+      "step": 774
+    },
+    {
+      "epoch": 0.08050275267476888,
+      "grad_norm": 0.3551725447177887,
+      "learning_rate": 9.840945806899253e-05,
+      "loss": 1.9765,
+      "step": 775
+    },
+    {
+      "epoch": 0.08060662719434923,
+      "grad_norm": 0.32390061020851135,
+      "learning_rate": 9.840537276876069e-05,
+      "loss": 1.6581,
+      "step": 776
+    },
+    {
+      "epoch": 0.08071050171392957,
+      "grad_norm": 0.3525629937648773,
+      "learning_rate": 9.840128231373422e-05,
+      "loss": 1.8436,
+      "step": 777
+    },
+    {
+      "epoch": 0.08081437623350993,
+      "grad_norm": 0.3583666682243347,
+      "learning_rate": 9.839718670434873e-05,
+      "loss": 1.8948,
+      "step": 778
+    },
+    {
+      "epoch": 0.08091825075309027,
+      "grad_norm": 0.3378392457962036,
+      "learning_rate": 9.839308594104038e-05,
+      "loss": 1.7955,
+      "step": 779
+    },
+    {
+      "epoch": 0.08102212527267061,
+      "grad_norm": 0.3844630718231201,
+      "learning_rate": 9.838898002424585e-05,
+      "loss": 1.987,
+      "step": 780
+    },
+    {
+      "epoch": 0.08112599979225096,
+      "grad_norm": 0.41022759675979614,
+      "learning_rate": 9.83848689544024e-05,
+      "loss": 1.959,
+      "step": 781
+    },
+    {
+      "epoch": 0.0812298743118313,
+      "grad_norm": 0.3520353436470032,
+      "learning_rate": 9.83807527319478e-05,
+      "loss": 1.8185,
+      "step": 782
+    },
+    {
+      "epoch": 0.08133374883141166,
+      "grad_norm": 0.36484450101852417,
+      "learning_rate": 9.837663135732045e-05,
+      "loss": 1.8079,
+      "step": 783
+    },
+    {
+      "epoch": 0.081437623350992,
+      "grad_norm": 0.34820452332496643,
+      "learning_rate": 9.83725048309592e-05,
+      "loss": 1.7167,
+      "step": 784
+    },
+    {
+      "epoch": 0.08154149787057235,
+      "grad_norm": 0.35253769159317017,
+      "learning_rate": 9.83683731533035e-05,
+      "loss": 1.8118,
+      "step": 785
+    },
+    {
+      "epoch": 0.0816453723901527,
+      "grad_norm": 0.3355175256729126,
+      "learning_rate": 9.836423632479335e-05,
+      "loss": 1.7767,
+      "step": 786
+    },
+    {
+      "epoch": 0.08174924690973304,
+      "grad_norm": 0.3633790612220764,
+      "learning_rate": 9.836009434586929e-05,
+      "loss": 1.9734,
+      "step": 787
+    },
+    {
+      "epoch": 0.08185312142931339,
+      "grad_norm": 0.36171820759773254,
+      "learning_rate": 9.835594721697241e-05,
+      "loss": 1.8049,
+      "step": 788
+    },
+    {
+      "epoch": 0.08195699594889373,
+      "grad_norm": 0.329885333776474,
+      "learning_rate": 9.835179493854433e-05,
+      "loss": 1.8358,
+      "step": 789
+    },
+    {
+      "epoch": 0.08206087046847409,
+      "grad_norm": 0.33959150314331055,
+      "learning_rate": 9.834763751102726e-05,
+      "loss": 1.6816,
+      "step": 790
+    },
+    {
+      "epoch": 0.08216474498805443,
+      "grad_norm": 0.3971327245235443,
+      "learning_rate": 9.834347493486391e-05,
+      "loss": 1.8448,
+      "step": 791
+    },
+    {
+      "epoch": 0.08226861950763478,
+      "grad_norm": 0.3603891134262085,
+      "learning_rate": 9.833930721049758e-05,
+      "loss": 1.9219,
+      "step": 792
+    },
+    {
+      "epoch": 0.08237249402721512,
+      "grad_norm": 0.3858663737773895,
+      "learning_rate": 9.833513433837209e-05,
+      "loss": 2.0203,
+      "step": 793
+    },
+    {
+      "epoch": 0.08247636854679546,
+      "grad_norm": 0.36965909600257874,
+      "learning_rate": 9.833095631893183e-05,
+      "loss": 1.5018,
+      "step": 794
+    },
+    {
+      "epoch": 0.08258024306637582,
+      "grad_norm": 0.3826972544193268,
+      "learning_rate": 9.832677315262171e-05,
+      "loss": 1.9229,
+      "step": 795
+    },
+    {
+      "epoch": 0.08268411758595616,
+      "grad_norm": 0.3521627187728882,
+      "learning_rate": 9.832258483988721e-05,
+      "loss": 1.8676,
+      "step": 796
+    },
+    {
+      "epoch": 0.08278799210553651,
+      "grad_norm": 0.38269051909446716,
+      "learning_rate": 9.831839138117435e-05,
+      "loss": 1.8478,
+      "step": 797
+    },
+    {
+      "epoch": 0.08289186662511686,
+      "grad_norm": 0.33963099122047424,
+      "learning_rate": 9.831419277692972e-05,
+      "loss": 1.7087,
+      "step": 798
+    },
+    {
+      "epoch": 0.08299574114469721,
+      "grad_norm": 0.3614671528339386,
+      "learning_rate": 9.830998902760043e-05,
+      "loss": 1.8661,
+      "step": 799
+    },
+    {
+      "epoch": 0.08309961566427755,
+      "grad_norm": 0.3493926227092743,
+      "learning_rate": 9.830578013363411e-05,
+      "loss": 1.8653,
+      "step": 800
+    },
+    {
+      "epoch": 0.0832034901838579,
+      "grad_norm": 0.3648037314414978,
+      "learning_rate": 9.830156609547905e-05,
+      "loss": 1.8252,
+      "step": 801
+    },
+    {
+      "epoch": 0.08330736470343825,
+      "grad_norm": 0.35134628415107727,
+      "learning_rate": 9.829734691358393e-05,
+      "loss": 1.8055,
+      "step": 802
+    },
+    {
+      "epoch": 0.08341123922301859,
+      "grad_norm": 0.3299328088760376,
+      "learning_rate": 9.829312258839811e-05,
+      "loss": 1.7022,
+      "step": 803
+    },
+    {
+      "epoch": 0.08351511374259894,
+      "grad_norm": 0.39070454239845276,
+      "learning_rate": 9.828889312037143e-05,
+      "loss": 2.0613,
+      "step": 804
+    },
+    {
+      "epoch": 0.08361898826217928,
+      "grad_norm": 0.3565296232700348,
+      "learning_rate": 9.828465850995433e-05,
+      "loss": 1.973,
+      "step": 805
+    },
+    {
+      "epoch": 0.08372286278175964,
+      "grad_norm": 0.35673460364341736,
+      "learning_rate": 9.82804187575977e-05,
+      "loss": 1.8557,
+      "step": 806
+    },
+    {
+      "epoch": 0.08382673730133998,
+      "grad_norm": 0.34928351640701294,
+      "learning_rate": 9.827617386375307e-05,
+      "loss": 1.9153,
+      "step": 807
+    },
+    {
+      "epoch": 0.08393061182092033,
+      "grad_norm": 0.35845404863357544,
+      "learning_rate": 9.827192382887251e-05,
+      "loss": 1.8209,
+      "step": 808
+    },
+    {
+      "epoch": 0.08403448634050067,
+      "grad_norm": 0.3234979510307312,
+      "learning_rate": 9.82676686534086e-05,
+      "loss": 1.6632,
+      "step": 809
+    },
+    {
+      "epoch": 0.08413836086008102,
+      "grad_norm": 0.3533652722835541,
+      "learning_rate": 9.826340833781447e-05,
+      "loss": 1.7217,
+      "step": 810
+    },
+    {
+      "epoch": 0.08424223537966137,
+      "grad_norm": 0.3984024226665497,
+      "learning_rate": 9.825914288254383e-05,
+      "loss": 1.9889,
+      "step": 811
+    },
+    {
+      "epoch": 0.08434610989924171,
+      "grad_norm": 0.3283498287200928,
+      "learning_rate": 9.82548722880509e-05,
+      "loss": 1.7551,
+      "step": 812
+    },
+    {
+      "epoch": 0.08444998441882207,
+      "grad_norm": 0.3530416190624237,
+      "learning_rate": 9.825059655479047e-05,
+      "loss": 1.5998,
+      "step": 813
+    },
+    {
+      "epoch": 0.08455385893840241,
+      "grad_norm": 0.377096951007843,
+      "learning_rate": 9.82463156832179e-05,
+      "loss": 1.93,
+      "step": 814
+    },
+    {
+      "epoch": 0.08465773345798276,
+      "grad_norm": 0.3375512361526489,
+      "learning_rate": 9.824202967378903e-05,
+      "loss": 1.7851,
+      "step": 815
+    },
+    {
+      "epoch": 0.0847616079775631,
+      "grad_norm": 0.3516164720058441,
+      "learning_rate": 9.82377385269603e-05,
+      "loss": 1.7856,
+      "step": 816
+    },
+    {
+      "epoch": 0.08486548249714346,
+      "grad_norm": 0.34516972303390503,
+      "learning_rate": 9.823344224318869e-05,
+      "loss": 1.7032,
+      "step": 817
+    },
+    {
+      "epoch": 0.0849693570167238,
+      "grad_norm": 0.33883360028266907,
+      "learning_rate": 9.822914082293171e-05,
+      "loss": 1.8525,
+      "step": 818
+    },
+    {
+      "epoch": 0.08507323153630414,
+      "grad_norm": 0.3980461657047272,
+      "learning_rate": 9.822483426664744e-05,
+      "loss": 1.8626,
+      "step": 819
+    },
+    {
+      "epoch": 0.0851771060558845,
+      "grad_norm": 0.36254748702049255,
+      "learning_rate": 9.822052257479448e-05,
+      "loss": 1.7556,
+      "step": 820
+    },
+    {
+      "epoch": 0.08528098057546483,
+      "grad_norm": 0.33397209644317627,
+      "learning_rate": 9.821620574783202e-05,
+      "loss": 1.713,
+      "step": 821
+    },
+    {
+      "epoch": 0.08538485509504519,
+      "grad_norm": 0.3705761432647705,
+      "learning_rate": 9.821188378621973e-05,
+      "loss": 2.0036,
+      "step": 822
+    },
+    {
+      "epoch": 0.08548872961462553,
+      "grad_norm": 0.40211713314056396,
+      "learning_rate": 9.820755669041787e-05,
+      "loss": 1.8456,
+      "step": 823
+    },
+    {
+      "epoch": 0.08559260413420589,
+      "grad_norm": 0.3149651288986206,
+      "learning_rate": 9.820322446088729e-05,
+      "loss": 1.6277,
+      "step": 824
+    },
+    {
+      "epoch": 0.08569647865378623,
+      "grad_norm": 0.36002975702285767,
+      "learning_rate": 9.819888709808929e-05,
+      "loss": 1.8758,
+      "step": 825
+    },
+    {
+      "epoch": 0.08580035317336657,
+      "grad_norm": 0.40512141585350037,
+      "learning_rate": 9.819454460248578e-05,
+      "loss": 1.8693,
+      "step": 826
+    },
+    {
+      "epoch": 0.08590422769294692,
+      "grad_norm": 0.33460405468940735,
+      "learning_rate": 9.81901969745392e-05,
+      "loss": 1.8883,
+      "step": 827
+    },
+    {
+      "epoch": 0.08600810221252726,
+      "grad_norm": 0.3666040301322937,
+      "learning_rate": 9.818584421471254e-05,
+      "loss": 1.7379,
+      "step": 828
+    },
+    {
+      "epoch": 0.08611197673210762,
+      "grad_norm": 0.3242300748825073,
+      "learning_rate": 9.818148632346934e-05,
+      "loss": 1.67,
+      "step": 829
+    },
+    {
+      "epoch": 0.08621585125168796,
+      "grad_norm": 0.37964361906051636,
+      "learning_rate": 9.817712330127367e-05,
+      "loss": 1.9572,
+      "step": 830
+    },
+    {
+      "epoch": 0.08631972577126831,
+      "grad_norm": 0.3320227265357971,
+      "learning_rate": 9.817275514859017e-05,
+      "loss": 1.7383,
+      "step": 831
+    },
+    {
+      "epoch": 0.08642360029084865,
+      "grad_norm": 0.3565314710140228,
+      "learning_rate": 9.8168381865884e-05,
+      "loss": 1.9483,
+      "step": 832
+    },
+    {
+      "epoch": 0.08652747481042901,
+      "grad_norm": 0.35897499322891235,
+      "learning_rate": 9.81640034536209e-05,
+      "loss": 1.7132,
+      "step": 833
+    },
+    {
+      "epoch": 0.08663134933000935,
+      "grad_norm": 0.37124866247177124,
+      "learning_rate": 9.815961991226711e-05,
+      "loss": 2.0216,
+      "step": 834
+    },
+    {
+      "epoch": 0.08673522384958969,
+      "grad_norm": 0.36309394240379333,
+      "learning_rate": 9.815523124228949e-05,
+      "loss": 1.7324,
+      "step": 835
+    },
+    {
+      "epoch": 0.08683909836917005,
+      "grad_norm": 0.35258007049560547,
+      "learning_rate": 9.815083744415534e-05,
+      "loss": 1.7634,
+      "step": 836
+    },
+    {
+      "epoch": 0.08694297288875039,
+      "grad_norm": 0.335984468460083,
+      "learning_rate": 9.81464385183326e-05,
+      "loss": 1.4646,
+      "step": 837
+    },
+    {
+      "epoch": 0.08704684740833074,
+      "grad_norm": 0.31050655245780945,
+      "learning_rate": 9.814203446528971e-05,
+      "loss": 1.3127,
+      "step": 838
+    },
+    {
+      "epoch": 0.08715072192791108,
+      "grad_norm": 0.3857136368751526,
+      "learning_rate": 9.813762528549568e-05,
+      "loss": 1.9647,
+      "step": 839
+    },
+    {
+      "epoch": 0.08725459644749144,
+      "grad_norm": 0.35393643379211426,
+      "learning_rate": 9.813321097942005e-05,
+      "loss": 1.7981,
+      "step": 840
+    },
+    {
+      "epoch": 0.08735847096707178,
+      "grad_norm": 0.3497065305709839,
+      "learning_rate": 9.812879154753289e-05,
+      "loss": 1.8632,
+      "step": 841
+    },
+    {
+      "epoch": 0.08746234548665212,
+      "grad_norm": 0.34144020080566406,
+      "learning_rate": 9.812436699030486e-05,
+      "loss": 1.653,
+      "step": 842
+    },
+    {
+      "epoch": 0.08756622000623247,
+      "grad_norm": 0.375434935092926,
+      "learning_rate": 9.811993730820712e-05,
+      "loss": 1.798,
+      "step": 843
+    },
+    {
+      "epoch": 0.08767009452581281,
+      "grad_norm": 0.33210819959640503,
+      "learning_rate": 9.811550250171142e-05,
+      "loss": 1.7409,
+      "step": 844
+    },
+    {
+      "epoch": 0.08777396904539317,
+      "grad_norm": 0.3796077072620392,
+      "learning_rate": 9.811106257129001e-05,
+      "loss": 1.8968,
+      "step": 845
+    },
+    {
+      "epoch": 0.08787784356497351,
+      "grad_norm": 0.3507767915725708,
+      "learning_rate": 9.810661751741573e-05,
+      "loss": 1.7981,
+      "step": 846
+    },
+    {
+      "epoch": 0.08798171808455386,
+      "grad_norm": 0.3411443829536438,
+      "learning_rate": 9.810216734056192e-05,
+      "loss": 1.5857,
+      "step": 847
+    },
+    {
+      "epoch": 0.0880855926041342,
+      "grad_norm": 0.3522975444793701,
+      "learning_rate": 9.809771204120248e-05,
+      "loss": 1.7766,
+      "step": 848
+    },
+    {
+      "epoch": 0.08818946712371455,
+      "grad_norm": 0.35262516140937805,
+      "learning_rate": 9.809325161981191e-05,
+      "loss": 1.6296,
+      "step": 849
+    },
+    {
+      "epoch": 0.0882933416432949,
+      "grad_norm": 0.36089739203453064,
+      "learning_rate": 9.808878607686518e-05,
+      "loss": 1.8433,
+      "step": 850
+    },
+    {
+      "epoch": 0.08839721616287524,
+      "grad_norm": 0.3271440863609314,
+      "learning_rate": 9.808431541283783e-05,
+      "loss": 1.8521,
+      "step": 851
+    },
+    {
+      "epoch": 0.0885010906824556,
+      "grad_norm": 0.39304596185684204,
+      "learning_rate": 9.807983962820597e-05,
+      "loss": 2.0822,
+      "step": 852
+    },
+    {
+      "epoch": 0.08860496520203594,
+      "grad_norm": 0.3730714023113251,
+      "learning_rate": 9.807535872344622e-05,
+      "loss": 1.8393,
+      "step": 853
+    },
+    {
+      "epoch": 0.08870883972161629,
+      "grad_norm": 0.37659719586372375,
+      "learning_rate": 9.807087269903577e-05,
+      "loss": 1.921,
+      "step": 854
+    },
+    {
+      "epoch": 0.08881271424119663,
+      "grad_norm": 0.3870750367641449,
+      "learning_rate": 9.806638155545236e-05,
+      "loss": 2.0497,
+      "step": 855
+    },
+    {
+      "epoch": 0.08891658876077699,
+      "grad_norm": 0.3502962291240692,
+      "learning_rate": 9.806188529317423e-05,
+      "loss": 1.849,
+      "step": 856
+    },
+    {
+      "epoch": 0.08902046328035733,
+      "grad_norm": 0.3573035001754761,
+      "learning_rate": 9.805738391268021e-05,
+      "loss": 1.6662,
+      "step": 857
+    },
+    {
+      "epoch": 0.08912433779993767,
+      "grad_norm": 0.46759751439094543,
+      "learning_rate": 9.805287741444968e-05,
+      "loss": 1.5148,
+      "step": 858
+    },
+    {
+      "epoch": 0.08922821231951802,
+      "grad_norm": 0.3513263463973999,
+      "learning_rate": 9.804836579896253e-05,
+      "loss": 1.9624,
+      "step": 859
+    },
+    {
+      "epoch": 0.08933208683909837,
+      "grad_norm": 0.3362573981285095,
+      "learning_rate": 9.80438490666992e-05,
+      "loss": 1.7757,
+      "step": 860
+    },
+    {
+      "epoch": 0.08943596135867872,
+      "grad_norm": 0.3206601142883301,
+      "learning_rate": 9.803932721814071e-05,
+      "loss": 1.692,
+      "step": 861
+    },
+    {
+      "epoch": 0.08953983587825906,
+      "grad_norm": 0.3470083773136139,
+      "learning_rate": 9.803480025376858e-05,
+      "loss": 1.8445,
+      "step": 862
+    },
+    {
+      "epoch": 0.08964371039783942,
+      "grad_norm": 0.3480392098426819,
+      "learning_rate": 9.803026817406493e-05,
+      "loss": 1.8338,
+      "step": 863
+    },
+    {
+      "epoch": 0.08974758491741976,
+      "grad_norm": 0.3229011297225952,
+      "learning_rate": 9.802573097951235e-05,
+      "loss": 1.7572,
+      "step": 864
+    },
+    {
+      "epoch": 0.0898514594370001,
+      "grad_norm": 0.3733616769313812,
+      "learning_rate": 9.802118867059407e-05,
+      "loss": 1.8358,
+      "step": 865
+    },
+    {
+      "epoch": 0.08995533395658045,
+      "grad_norm": 0.35253623127937317,
+      "learning_rate": 9.801664124779373e-05,
+      "loss": 1.8347,
+      "step": 866
+    },
+    {
+      "epoch": 0.09005920847616079,
+      "grad_norm": 0.3676695227622986,
+      "learning_rate": 9.801208871159567e-05,
+      "loss": 1.8158,
+      "step": 867
+    },
+    {
+      "epoch": 0.09016308299574115,
+      "grad_norm": 0.3626701235771179,
+      "learning_rate": 9.800753106248468e-05,
+      "loss": 1.9947,
+      "step": 868
+    },
+    {
+      "epoch": 0.09026695751532149,
+      "grad_norm": 0.3680635392665863,
+      "learning_rate": 9.800296830094609e-05,
+      "loss": 1.7856,
+      "step": 869
+    },
+    {
+      "epoch": 0.09037083203490184,
+      "grad_norm": 0.34749433398246765,
+      "learning_rate": 9.79984004274658e-05,
+      "loss": 1.7665,
+      "step": 870
+    },
+    {
+      "epoch": 0.09047470655448218,
+      "grad_norm": 0.3646162152290344,
+      "learning_rate": 9.799382744253029e-05,
+      "loss": 1.715,
+      "step": 871
+    },
+    {
+      "epoch": 0.09057858107406254,
+      "grad_norm": 0.37328121066093445,
+      "learning_rate": 9.79892493466265e-05,
+      "loss": 1.7945,
+      "step": 872
+    },
+    {
+      "epoch": 0.09068245559364288,
+      "grad_norm": 0.35278576612472534,
+      "learning_rate": 9.7984666140242e-05,
+      "loss": 1.7249,
+      "step": 873
+    },
+    {
+      "epoch": 0.09078633011322322,
+      "grad_norm": 0.33178964257240295,
+      "learning_rate": 9.798007782386485e-05,
+      "loss": 1.6934,
+      "step": 874
+    },
+    {
+      "epoch": 0.09089020463280358,
+      "grad_norm": 0.3599662780761719,
+      "learning_rate": 9.797548439798368e-05,
+      "loss": 1.9192,
+      "step": 875
+    },
+    {
+      "epoch": 0.09099407915238392,
+      "grad_norm": 0.37787213921546936,
+      "learning_rate": 9.797088586308762e-05,
+      "loss": 1.9863,
+      "step": 876
+    },
+    {
+      "epoch": 0.09109795367196427,
+      "grad_norm": 0.31273946166038513,
+      "learning_rate": 9.796628221966642e-05,
+      "loss": 1.6311,
+      "step": 877
+    },
+    {
+      "epoch": 0.09120182819154461,
+      "grad_norm": 0.33196964859962463,
+      "learning_rate": 9.796167346821032e-05,
+      "loss": 1.6965,
+      "step": 878
+    },
+    {
+      "epoch": 0.09130570271112497,
+      "grad_norm": 0.3330535590648651,
+      "learning_rate": 9.79570596092101e-05,
+      "loss": 1.745,
+      "step": 879
+    },
+    {
+      "epoch": 0.09140957723070531,
+      "grad_norm": 0.3526756465435028,
+      "learning_rate": 9.795244064315712e-05,
+      "loss": 1.886,
+      "step": 880
+    },
+    {
+      "epoch": 0.09151345175028565,
+      "grad_norm": 0.3859866261482239,
+      "learning_rate": 9.794781657054324e-05,
+      "loss": 1.8935,
+      "step": 881
+    },
+    {
+      "epoch": 0.091617326269866,
+      "grad_norm": 0.3710609972476959,
+      "learning_rate": 9.794318739186092e-05,
+      "loss": 1.9279,
+      "step": 882
+    },
+    {
+      "epoch": 0.09172120078944634,
+      "grad_norm": 0.37800589203834534,
+      "learning_rate": 9.79385531076031e-05,
+      "loss": 1.9028,
+      "step": 883
+    },
+    {
+      "epoch": 0.0918250753090267,
+      "grad_norm": 0.36296647787094116,
+      "learning_rate": 9.793391371826332e-05,
+      "loss": 1.9369,
+      "step": 884
+    },
+    {
+      "epoch": 0.09192894982860704,
+      "grad_norm": 0.3620465397834778,
+      "learning_rate": 9.792926922433563e-05,
+      "loss": 1.9072,
+      "step": 885
+    },
+    {
+      "epoch": 0.0920328243481874,
+      "grad_norm": 0.35136178135871887,
+      "learning_rate": 9.792461962631462e-05,
+      "loss": 1.6753,
+      "step": 886
+    },
+    {
+      "epoch": 0.09213669886776774,
+      "grad_norm": 0.33308306336402893,
+      "learning_rate": 9.791996492469544e-05,
+      "loss": 1.8044,
+      "step": 887
+    },
+    {
+      "epoch": 0.09224057338734809,
+      "grad_norm": 0.36434537172317505,
+      "learning_rate": 9.79153051199738e-05,
+      "loss": 1.6905,
+      "step": 888
+    },
+    {
+      "epoch": 0.09234444790692843,
+      "grad_norm": 0.3682979345321655,
+      "learning_rate": 9.791064021264591e-05,
+      "loss": 1.9002,
+      "step": 889
+    },
+    {
+      "epoch": 0.09244832242650877,
+      "grad_norm": 0.39494451880455017,
+      "learning_rate": 9.790597020320856e-05,
+      "loss": 1.8316,
+      "step": 890
+    },
+    {
+      "epoch": 0.09255219694608913,
+      "grad_norm": 0.3634776473045349,
+      "learning_rate": 9.790129509215907e-05,
+      "loss": 1.7874,
+      "step": 891
+    },
+    {
+      "epoch": 0.09265607146566947,
+      "grad_norm": 0.49316370487213135,
+      "learning_rate": 9.789661487999528e-05,
+      "loss": 1.8666,
+      "step": 892
+    },
+    {
+      "epoch": 0.09275994598524982,
+      "grad_norm": 0.4129713475704193,
+      "learning_rate": 9.789192956721563e-05,
+      "loss": 2.0314,
+      "step": 893
+    },
+    {
+      "epoch": 0.09286382050483016,
+      "grad_norm": 0.33426743745803833,
+      "learning_rate": 9.788723915431905e-05,
+      "loss": 1.6938,
+      "step": 894
+    },
+    {
+      "epoch": 0.09296769502441052,
+      "grad_norm": 0.3889116048812866,
+      "learning_rate": 9.788254364180503e-05,
+      "loss": 1.9863,
+      "step": 895
+    },
+    {
+      "epoch": 0.09307156954399086,
+      "grad_norm": 0.37265413999557495,
+      "learning_rate": 9.78778430301736e-05,
+      "loss": 1.7899,
+      "step": 896
+    },
+    {
+      "epoch": 0.0931754440635712,
+      "grad_norm": 0.4090532958507538,
+      "learning_rate": 9.787313731992537e-05,
+      "loss": 1.8823,
+      "step": 897
+    },
+    {
+      "epoch": 0.09327931858315155,
+      "grad_norm": 0.3560778796672821,
+      "learning_rate": 9.786842651156143e-05,
+      "loss": 1.8849,
+      "step": 898
+    },
+    {
+      "epoch": 0.0933831931027319,
+      "grad_norm": 0.43143633008003235,
+      "learning_rate": 9.786371060558346e-05,
+      "loss": 1.7419,
+      "step": 899
+    },
+    {
+      "epoch": 0.09348706762231225,
+      "grad_norm": 0.4012468755245209,
+      "learning_rate": 9.785898960249364e-05,
+      "loss": 1.9191,
+      "step": 900
+    },
+    {
+      "epoch": 0.09359094214189259,
+      "grad_norm": 0.38962993025779724,
+      "learning_rate": 9.785426350279477e-05,
+      "loss": 1.8023,
+      "step": 901
+    },
+    {
+      "epoch": 0.09369481666147295,
+      "grad_norm": 0.38585248589515686,
+      "learning_rate": 9.78495323069901e-05,
+      "loss": 1.9383,
+      "step": 902
+    },
+    {
+      "epoch": 0.09379869118105329,
+      "grad_norm": 0.347837895154953,
+      "learning_rate": 9.784479601558348e-05,
+      "loss": 1.8187,
+      "step": 903
+    },
+    {
+      "epoch": 0.09390256570063363,
+      "grad_norm": 0.32775819301605225,
+      "learning_rate": 9.78400546290793e-05,
+      "loss": 1.5762,
+      "step": 904
+    },
+    {
+      "epoch": 0.09400644022021398,
+      "grad_norm": 0.3671472668647766,
+      "learning_rate": 9.783530814798245e-05,
+      "loss": 1.8772,
+      "step": 905
+    },
+    {
+      "epoch": 0.09411031473979432,
+      "grad_norm": 0.32809990644454956,
+      "learning_rate": 9.783055657279843e-05,
+      "loss": 1.7432,
+      "step": 906
+    },
+    {
+      "epoch": 0.09421418925937468,
+      "grad_norm": 0.33797743916511536,
+      "learning_rate": 9.782579990403321e-05,
+      "loss": 1.7051,
+      "step": 907
+    },
+    {
+      "epoch": 0.09431806377895502,
+      "grad_norm": 0.3659290373325348,
+      "learning_rate": 9.782103814219337e-05,
+      "loss": 1.7051,
+      "step": 908
+    },
+    {
+      "epoch": 0.09442193829853537,
+      "grad_norm": 0.4114789068698883,
+      "learning_rate": 9.781627128778598e-05,
+      "loss": 2.0238,
+      "step": 909
+    },
+    {
+      "epoch": 0.09452581281811571,
+      "grad_norm": 0.3247332274913788,
+      "learning_rate": 9.781149934131868e-05,
+      "loss": 1.6112,
+      "step": 910
+    },
+    {
+      "epoch": 0.09462968733769607,
+      "grad_norm": 0.3638227880001068,
+      "learning_rate": 9.780672230329964e-05,
+      "loss": 1.7238,
+      "step": 911
+    },
+    {
+      "epoch": 0.09473356185727641,
+      "grad_norm": 0.35563889145851135,
+      "learning_rate": 9.780194017423758e-05,
+      "loss": 1.8491,
+      "step": 912
+    },
+    {
+      "epoch": 0.09483743637685675,
+      "grad_norm": 0.38154855370521545,
+      "learning_rate": 9.779715295464178e-05,
+      "loss": 1.666,
+      "step": 913
+    },
+    {
+      "epoch": 0.0949413108964371,
+      "grad_norm": 0.3501157760620117,
+      "learning_rate": 9.779236064502201e-05,
+      "loss": 1.8345,
+      "step": 914
+    },
+    {
+      "epoch": 0.09504518541601745,
+      "grad_norm": 0.3432106375694275,
+      "learning_rate": 9.778756324588863e-05,
+      "loss": 1.7373,
+      "step": 915
+    },
+    {
+      "epoch": 0.0951490599355978,
+      "grad_norm": 0.39171236753463745,
+      "learning_rate": 9.778276075775251e-05,
+      "loss": 1.9764,
+      "step": 916
+    },
+    {
+      "epoch": 0.09525293445517814,
+      "grad_norm": 0.35228273272514343,
+      "learning_rate": 9.77779531811251e-05,
+      "loss": 1.8864,
+      "step": 917
+    },
+    {
+      "epoch": 0.0953568089747585,
+      "grad_norm": 0.3456995487213135,
+      "learning_rate": 9.777314051651836e-05,
+      "loss": 1.8929,
+      "step": 918
+    },
+    {
+      "epoch": 0.09546068349433884,
+      "grad_norm": 0.39113515615463257,
+      "learning_rate": 9.776832276444479e-05,
+      "loss": 1.8455,
+      "step": 919
+    },
+    {
+      "epoch": 0.09556455801391918,
+      "grad_norm": 0.35368141531944275,
+      "learning_rate": 9.776349992541746e-05,
+      "loss": 1.7881,
+      "step": 920
+    },
+    {
+      "epoch": 0.09566843253349953,
+      "grad_norm": 0.3645842671394348,
+      "learning_rate": 9.775867199994994e-05,
+      "loss": 1.9618,
+      "step": 921
+    },
+    {
+      "epoch": 0.09577230705307987,
+      "grad_norm": 0.4213750958442688,
+      "learning_rate": 9.775383898855641e-05,
+      "loss": 1.9835,
+      "step": 922
+    },
+    {
+      "epoch": 0.09587618157266023,
+      "grad_norm": 0.3259928226470947,
+      "learning_rate": 9.77490008917515e-05,
+      "loss": 1.6711,
+      "step": 923
+    },
+    {
+      "epoch": 0.09598005609224057,
+      "grad_norm": 0.33279410004615784,
+      "learning_rate": 9.774415771005048e-05,
+      "loss": 1.7854,
+      "step": 924
+    },
+    {
+      "epoch": 0.09608393061182093,
+      "grad_norm": 0.3503129482269287,
+      "learning_rate": 9.773930944396905e-05,
+      "loss": 1.8105,
+      "step": 925
+    },
+    {
+      "epoch": 0.09618780513140127,
+      "grad_norm": 0.3861035704612732,
+      "learning_rate": 9.773445609402356e-05,
+      "loss": 1.8127,
+      "step": 926
+    },
+    {
+      "epoch": 0.09629167965098162,
+      "grad_norm": 0.4022195339202881,
+      "learning_rate": 9.772959766073084e-05,
+      "loss": 1.8921,
+      "step": 927
+    },
+    {
+      "epoch": 0.09639555417056196,
+      "grad_norm": 0.34826424717903137,
+      "learning_rate": 9.772473414460828e-05,
+      "loss": 1.8411,
+      "step": 928
+    },
+    {
+      "epoch": 0.0964994286901423,
+      "grad_norm": 0.3265725374221802,
+      "learning_rate": 9.77198655461738e-05,
+      "loss": 1.6894,
+      "step": 929
+    },
+    {
+      "epoch": 0.09660330320972266,
+      "grad_norm": 0.3737447261810303,
+      "learning_rate": 9.771499186594586e-05,
+      "loss": 1.8592,
+      "step": 930
+    },
+    {
+      "epoch": 0.096707177729303,
+      "grad_norm": 0.3337661623954773,
+      "learning_rate": 9.771011310444348e-05,
+      "loss": 1.6941,
+      "step": 931
+    },
+    {
+      "epoch": 0.09681105224888335,
+      "grad_norm": 0.3704545199871063,
+      "learning_rate": 9.770522926218622e-05,
+      "loss": 1.7519,
+      "step": 932
+    },
+    {
+      "epoch": 0.0969149267684637,
+      "grad_norm": 0.35930871963500977,
+      "learning_rate": 9.770034033969413e-05,
+      "loss": 1.9341,
+      "step": 933
+    },
+    {
+      "epoch": 0.09701880128804405,
+      "grad_norm": 0.38769471645355225,
+      "learning_rate": 9.769544633748789e-05,
+      "loss": 1.8773,
+      "step": 934
+    },
+    {
+      "epoch": 0.09712267580762439,
+      "grad_norm": 0.37353989481925964,
+      "learning_rate": 9.769054725608867e-05,
+      "loss": 1.9544,
+      "step": 935
+    },
+    {
+      "epoch": 0.09722655032720473,
+      "grad_norm": 0.39026400446891785,
+      "learning_rate": 9.768564309601813e-05,
+      "loss": 2.0178,
+      "step": 936
+    },
+    {
+      "epoch": 0.09733042484678509,
+      "grad_norm": 0.37478938698768616,
+      "learning_rate": 9.768073385779859e-05,
+      "loss": 1.6323,
+      "step": 937
+    },
+    {
+      "epoch": 0.09743429936636543,
+      "grad_norm": 0.3543946146965027,
+      "learning_rate": 9.76758195419528e-05,
+      "loss": 1.8568,
+      "step": 938
+    },
+    {
+      "epoch": 0.09753817388594578,
+      "grad_norm": 0.3243783712387085,
+      "learning_rate": 9.767090014900413e-05,
+      "loss": 1.6446,
+      "step": 939
+    },
+    {
+      "epoch": 0.09764204840552612,
+      "grad_norm": 0.390176922082901,
+      "learning_rate": 9.766597567947644e-05,
+      "loss": 1.7432,
+      "step": 940
+    },
+    {
+      "epoch": 0.09774592292510648,
+      "grad_norm": 0.40544116497039795,
+      "learning_rate": 9.766104613389413e-05,
+      "loss": 1.8066,
+      "step": 941
+    },
+    {
+      "epoch": 0.09784979744468682,
+      "grad_norm": 0.4119703471660614,
+      "learning_rate": 9.76561115127822e-05,
+      "loss": 1.8682,
+      "step": 942
+    },
+    {
+      "epoch": 0.09795367196426717,
+      "grad_norm": 0.41198158264160156,
+      "learning_rate": 9.765117181666612e-05,
+      "loss": 1.8556,
+      "step": 943
+    },
+    {
+      "epoch": 0.09805754648384751,
+      "grad_norm": 0.3242831528186798,
+      "learning_rate": 9.764622704607193e-05,
+      "loss": 1.8249,
+      "step": 944
+    },
+    {
+      "epoch": 0.09816142100342785,
+      "grad_norm": 0.336739718914032,
+      "learning_rate": 9.764127720152623e-05,
+      "loss": 1.7983,
+      "step": 945
+    },
+    {
+      "epoch": 0.09826529552300821,
+      "grad_norm": 0.43454012274742126,
+      "learning_rate": 9.763632228355611e-05,
+      "loss": 1.8482,
+      "step": 946
+    },
+    {
+      "epoch": 0.09836917004258855,
+      "grad_norm": 0.39079445600509644,
+      "learning_rate": 9.763136229268924e-05,
+      "loss": 1.9735,
+      "step": 947
+    },
+    {
+      "epoch": 0.0984730445621689,
+      "grad_norm": 0.3446981906890869,
+      "learning_rate": 9.762639722945384e-05,
+      "loss": 1.8289,
+      "step": 948
+    },
+    {
+      "epoch": 0.09857691908174925,
+      "grad_norm": 0.3308192193508148,
+      "learning_rate": 9.762142709437863e-05,
+      "loss": 1.6953,
+      "step": 949
+    },
+    {
+      "epoch": 0.0986807936013296,
+      "grad_norm": 0.40982964634895325,
+      "learning_rate": 9.76164518879929e-05,
+      "loss": 1.9994,
+      "step": 950
+    },
+    {
+      "epoch": 0.09878466812090994,
+      "grad_norm": 0.3556596040725708,
+      "learning_rate": 9.761147161082644e-05,
+      "loss": 1.7591,
+      "step": 951
+    },
+    {
+      "epoch": 0.09888854264049028,
+      "grad_norm": 0.361565500497818,
+      "learning_rate": 9.760648626340965e-05,
+      "loss": 1.8883,
+      "step": 952
+    },
+    {
+      "epoch": 0.09899241716007064,
+      "grad_norm": 0.36020350456237793,
+      "learning_rate": 9.760149584627343e-05,
+      "loss": 1.7859,
+      "step": 953
+    },
+    {
+      "epoch": 0.09909629167965098,
+      "grad_norm": 0.3831625282764435,
+      "learning_rate": 9.759650035994921e-05,
+      "loss": 1.9948,
+      "step": 954
+    },
+    {
+      "epoch": 0.09920016619923133,
+      "grad_norm": 0.34753647446632385,
+      "learning_rate": 9.759149980496896e-05,
+      "loss": 1.8239,
+      "step": 955
+    },
+    {
+      "epoch": 0.09930404071881167,
+      "grad_norm": 0.3743743896484375,
+      "learning_rate": 9.758649418186521e-05,
+      "loss": 1.8389,
+      "step": 956
+    },
+    {
+      "epoch": 0.09940791523839203,
+      "grad_norm": 0.38041260838508606,
+      "learning_rate": 9.758148349117101e-05,
+      "loss": 1.9304,
+      "step": 957
+    },
+    {
+      "epoch": 0.09951178975797237,
+      "grad_norm": 0.4073218107223511,
+      "learning_rate": 9.757646773341998e-05,
+      "loss": 1.9184,
+      "step": 958
+    },
+    {
+      "epoch": 0.09961566427755271,
+      "grad_norm": 0.3602345585823059,
+      "learning_rate": 9.757144690914624e-05,
+      "loss": 1.8935,
+      "step": 959
+    },
+    {
+      "epoch": 0.09971953879713306,
+      "grad_norm": 0.3824581801891327,
+      "learning_rate": 9.756642101888448e-05,
+      "loss": 1.8969,
+      "step": 960
+    },
+    {
+      "epoch": 0.0998234133167134,
+      "grad_norm": 0.3472011387348175,
+      "learning_rate": 9.756139006316991e-05,
+      "loss": 1.7702,
+      "step": 961
+    },
+    {
+      "epoch": 0.09992728783629376,
+      "grad_norm": 0.3461735248565674,
+      "learning_rate": 9.755635404253829e-05,
+      "loss": 1.6524,
+      "step": 962
+    },
+    {
+      "epoch": 0.1000311623558741,
+      "grad_norm": 0.34832313656806946,
+      "learning_rate": 9.755131295752593e-05,
+      "loss": 1.701,
+      "step": 963
+    },
+    {
+      "epoch": 0.10013503687545446,
+      "grad_norm": 0.3486674726009369,
+      "learning_rate": 9.754626680866964e-05,
+      "loss": 1.8465,
+      "step": 964
+    },
+    {
+      "epoch": 0.1002389113950348,
+      "grad_norm": 0.3982466161251068,
+      "learning_rate": 9.754121559650681e-05,
+      "loss": 2.0038,
+      "step": 965
+    },
+    {
+      "epoch": 0.10034278591461515,
+      "grad_norm": 0.3699825406074524,
+      "learning_rate": 9.753615932157536e-05,
+      "loss": 2.1116,
+      "step": 966
+    },
+    {
+      "epoch": 0.10044666043419549,
+      "grad_norm": 0.37160253524780273,
+      "learning_rate": 9.753109798441375e-05,
+      "loss": 1.9158,
+      "step": 967
+    },
+    {
+      "epoch": 0.10055053495377583,
+      "grad_norm": 0.34480687975883484,
+      "learning_rate": 9.752603158556095e-05,
+      "loss": 1.7587,
+      "step": 968
+    },
+    {
+      "epoch": 0.10065440947335619,
+      "grad_norm": 0.34491807222366333,
+      "learning_rate": 9.752096012555652e-05,
+      "loss": 1.8125,
+      "step": 969
+    },
+    {
+      "epoch": 0.10075828399293653,
+      "grad_norm": 0.36737021803855896,
+      "learning_rate": 9.751588360494048e-05,
+      "loss": 1.9693,
+      "step": 970
+    },
+    {
+      "epoch": 0.10086215851251688,
+      "grad_norm": 0.372712641954422,
+      "learning_rate": 9.75108020242535e-05,
+      "loss": 1.7953,
+      "step": 971
+    },
+    {
+      "epoch": 0.10096603303209722,
+      "grad_norm": 0.3740496039390564,
+      "learning_rate": 9.75057153840367e-05,
+      "loss": 1.7753,
+      "step": 972
+    },
+    {
+      "epoch": 0.10106990755167758,
+      "grad_norm": 0.46863672137260437,
+      "learning_rate": 9.750062368483176e-05,
+      "loss": 2.1136,
+      "step": 973
+    },
+    {
+      "epoch": 0.10117378207125792,
+      "grad_norm": 0.3760620355606079,
+      "learning_rate": 9.749552692718093e-05,
+      "loss": 1.7655,
+      "step": 974
+    },
+    {
+      "epoch": 0.10127765659083826,
+      "grad_norm": 0.35611656308174133,
+      "learning_rate": 9.749042511162696e-05,
+      "loss": 1.8326,
+      "step": 975
+    },
+    {
+      "epoch": 0.10138153111041862,
+      "grad_norm": 0.33176377415657043,
+      "learning_rate": 9.748531823871314e-05,
+      "loss": 1.7562,
+      "step": 976
+    },
+    {
+      "epoch": 0.10148540562999896,
+      "grad_norm": 0.33755457401275635,
+      "learning_rate": 9.748020630898333e-05,
+      "loss": 1.7919,
+      "step": 977
+    },
+    {
+      "epoch": 0.10158928014957931,
+      "grad_norm": 0.40104183554649353,
+      "learning_rate": 9.747508932298192e-05,
+      "loss": 1.7349,
+      "step": 978
+    },
+    {
+      "epoch": 0.10169315466915965,
+      "grad_norm": 0.33952096104621887,
+      "learning_rate": 9.746996728125382e-05,
+      "loss": 1.6836,
+      "step": 979
+    },
+    {
+      "epoch": 0.10179702918874001,
+      "grad_norm": 0.3854829668998718,
+      "learning_rate": 9.746484018434447e-05,
+      "loss": 1.6913,
+      "step": 980
+    },
+    {
+      "epoch": 0.10190090370832035,
+      "grad_norm": 0.44178688526153564,
+      "learning_rate": 9.745970803279989e-05,
+      "loss": 2.2579,
+      "step": 981
+    },
+    {
+      "epoch": 0.1020047782279007,
+      "grad_norm": 0.35316595435142517,
+      "learning_rate": 9.74545708271666e-05,
+      "loss": 1.7616,
+      "step": 982
+    },
+    {
+      "epoch": 0.10210865274748104,
+      "grad_norm": 0.36709627509117126,
+      "learning_rate": 9.744942856799169e-05,
+      "loss": 1.6341,
+      "step": 983
+    },
+    {
+      "epoch": 0.10221252726706138,
+      "grad_norm": 0.4684566557407379,
+      "learning_rate": 9.744428125582275e-05,
+      "loss": 2.0114,
+      "step": 984
+    },
+    {
+      "epoch": 0.10231640178664174,
+      "grad_norm": 0.34113651514053345,
+      "learning_rate": 9.743912889120791e-05,
+      "loss": 1.7651,
+      "step": 985
+    },
+    {
+      "epoch": 0.10242027630622208,
+      "grad_norm": 0.4054688513278961,
+      "learning_rate": 9.743397147469592e-05,
+      "loss": 2.0589,
+      "step": 986
+    },
+    {
+      "epoch": 0.10252415082580243,
+      "grad_norm": 0.36325010657310486,
+      "learning_rate": 9.742880900683596e-05,
+      "loss": 1.8003,
+      "step": 987
+    },
+    {
+      "epoch": 0.10262802534538278,
+      "grad_norm": 0.3484811782836914,
+      "learning_rate": 9.742364148817777e-05,
+      "loss": 1.8546,
+      "step": 988
+    },
+    {
+      "epoch": 0.10273189986496313,
+      "grad_norm": 0.3415622115135193,
+      "learning_rate": 9.74184689192717e-05,
+      "loss": 1.6446,
+      "step": 989
+    },
+    {
+      "epoch": 0.10283577438454347,
+      "grad_norm": 0.377069354057312,
+      "learning_rate": 9.741329130066859e-05,
+      "loss": 2.0174,
+      "step": 990
+    },
+    {
+      "epoch": 0.10293964890412381,
+      "grad_norm": 0.37598928809165955,
+      "learning_rate": 9.740810863291976e-05,
+      "loss": 1.9646,
+      "step": 991
+    },
+    {
+      "epoch": 0.10304352342370417,
+      "grad_norm": 0.34742510318756104,
+      "learning_rate": 9.740292091657718e-05,
+      "loss": 1.7112,
+      "step": 992
+    },
+    {
+      "epoch": 0.10314739794328451,
+      "grad_norm": 0.35279062390327454,
+      "learning_rate": 9.739772815219327e-05,
+      "loss": 1.736,
+      "step": 993
+    },
+    {
+      "epoch": 0.10325127246286486,
+      "grad_norm": 0.33563071489334106,
+      "learning_rate": 9.739253034032103e-05,
+      "loss": 1.7482,
+      "step": 994
+    },
+    {
+      "epoch": 0.1033551469824452,
+      "grad_norm": 0.4013693332672119,
+      "learning_rate": 9.738732748151397e-05,
+      "loss": 1.9931,
+      "step": 995
+    },
+    {
+      "epoch": 0.10345902150202556,
+      "grad_norm": 0.3738352954387665,
+      "learning_rate": 9.738211957632619e-05,
+      "loss": 1.7324,
+      "step": 996
+    },
+    {
+      "epoch": 0.1035628960216059,
+      "grad_norm": 0.34439367055892944,
+      "learning_rate": 9.737690662531227e-05,
+      "loss": 1.7277,
+      "step": 997
+    },
+    {
+      "epoch": 0.10366677054118625,
+      "grad_norm": 0.34663471579551697,
+      "learning_rate": 9.737168862902735e-05,
+      "loss": 1.8904,
+      "step": 998
+    },
+    {
+      "epoch": 0.1037706450607666,
+      "grad_norm": 0.3536369800567627,
+      "learning_rate": 9.736646558802709e-05,
+      "loss": 1.898,
+      "step": 999
+    },
+    {
+      "epoch": 0.10387451958034694,
+      "grad_norm": 0.37141430377960205,
+      "learning_rate": 9.736123750286772e-05,
+      "loss": 1.9826,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10397839409992729,
+      "grad_norm": 0.33750221133232117,
+      "learning_rate": 9.7356004374106e-05,
+      "loss": 1.6245,
+      "step": 1001
+    },
+    {
+      "epoch": 0.10408226861950763,
+      "grad_norm": 0.39704445004463196,
+      "learning_rate": 9.735076620229919e-05,
+      "loss": 1.9813,
+      "step": 1002
+    },
+    {
+      "epoch": 0.10418614313908799,
+      "grad_norm": 0.3396252691745758,
+      "learning_rate": 9.734552298800513e-05,
+      "loss": 1.7119,
+      "step": 1003
+    },
+    {
+      "epoch": 0.10429001765866833,
+      "grad_norm": 0.3506071865558624,
+      "learning_rate": 9.734027473178219e-05,
+      "loss": 1.8563,
+      "step": 1004
+    },
+    {
+      "epoch": 0.10439389217824868,
+      "grad_norm": 0.3697708249092102,
+      "learning_rate": 9.733502143418925e-05,
+      "loss": 1.5552,
+      "step": 1005
+    },
+    {
+      "epoch": 0.10449776669782902,
+      "grad_norm": 0.34835517406463623,
+      "learning_rate": 9.732976309578577e-05,
+      "loss": 1.6061,
+      "step": 1006
+    },
+    {
+      "epoch": 0.10460164121740936,
+      "grad_norm": 0.4172792434692383,
+      "learning_rate": 9.73244997171317e-05,
+      "loss": 2.0223,
+      "step": 1007
+    },
+    {
+      "epoch": 0.10470551573698972,
+      "grad_norm": 0.37893515825271606,
+      "learning_rate": 9.731923129878754e-05,
+      "loss": 1.8512,
+      "step": 1008
+    },
+    {
+      "epoch": 0.10480939025657006,
+      "grad_norm": 0.3517685830593109,
+      "learning_rate": 9.731395784131437e-05,
+      "loss": 1.7601,
+      "step": 1009
+    },
+    {
+      "epoch": 0.10491326477615041,
+      "grad_norm": 0.3533296585083008,
+      "learning_rate": 9.730867934527374e-05,
+      "loss": 1.7264,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10501713929573075,
+      "grad_norm": 0.3656322956085205,
+      "learning_rate": 9.73033958112278e-05,
+      "loss": 1.8542,
+      "step": 1011
+    },
+    {
+      "epoch": 0.10512101381531111,
+      "grad_norm": 0.36052483320236206,
+      "learning_rate": 9.729810723973917e-05,
+      "loss": 1.7967,
+      "step": 1012
+    },
+    {
+      "epoch": 0.10522488833489145,
+      "grad_norm": 0.3687455356121063,
+      "learning_rate": 9.729281363137106e-05,
+      "loss": 1.9559,
+      "step": 1013
+    },
+    {
+      "epoch": 0.10532876285447179,
+      "grad_norm": 0.3520287275314331,
+      "learning_rate": 9.72875149866872e-05,
+      "loss": 1.8265,
+      "step": 1014
+    },
+    {
+      "epoch": 0.10543263737405215,
+      "grad_norm": 0.36388278007507324,
+      "learning_rate": 9.728221130625186e-05,
+      "loss": 1.7892,
+      "step": 1015
+    },
+    {
+      "epoch": 0.10553651189363249,
+      "grad_norm": 0.3430229127407074,
+      "learning_rate": 9.727690259062982e-05,
+      "loss": 1.7436,
+      "step": 1016
+    },
+    {
+      "epoch": 0.10564038641321284,
+      "grad_norm": 0.40209057927131653,
+      "learning_rate": 9.727158884038643e-05,
+      "loss": 2.0679,
+      "step": 1017
+    },
+    {
+      "epoch": 0.10574426093279318,
+      "grad_norm": 0.35472702980041504,
+      "learning_rate": 9.726627005608756e-05,
+      "loss": 1.8322,
+      "step": 1018
+    },
+    {
+      "epoch": 0.10584813545237354,
+      "grad_norm": 0.3761141896247864,
+      "learning_rate": 9.726094623829962e-05,
+      "loss": 1.9986,
+      "step": 1019
+    },
+    {
+      "epoch": 0.10595200997195388,
+      "grad_norm": 0.37328705191612244,
+      "learning_rate": 9.725561738758956e-05,
+      "loss": 1.8012,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10605588449153423,
+      "grad_norm": 0.3950223922729492,
+      "learning_rate": 9.725028350452485e-05,
+      "loss": 1.859,
+      "step": 1021
+    },
+    {
+      "epoch": 0.10615975901111457,
+      "grad_norm": 0.37431395053863525,
+      "learning_rate": 9.724494458967352e-05,
+      "loss": 1.6539,
+      "step": 1022
+    },
+    {
+      "epoch": 0.10626363353069491,
+      "grad_norm": 0.3411618173122406,
+      "learning_rate": 9.723960064360412e-05,
+      "loss": 1.6884,
+      "step": 1023
+    },
+    {
+      "epoch": 0.10636750805027527,
+      "grad_norm": 0.3407668173313141,
+      "learning_rate": 9.72342516668857e-05,
+      "loss": 1.7554,
+      "step": 1024
+    },
+    {
+      "epoch": 0.10647138256985561,
+      "grad_norm": 0.3708615005016327,
+      "learning_rate": 9.722889766008796e-05,
+      "loss": 1.7656,
+      "step": 1025
+    },
+    {
+      "epoch": 0.10657525708943597,
+      "grad_norm": 0.3609970510005951,
+      "learning_rate": 9.722353862378101e-05,
+      "loss": 1.8226,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1066791316090163,
+      "grad_norm": 0.341370552778244,
+      "learning_rate": 9.721817455853554e-05,
+      "loss": 1.6735,
+      "step": 1027
+    },
+    {
+      "epoch": 0.10678300612859666,
+      "grad_norm": 0.3461858630180359,
+      "learning_rate": 9.72128054649228e-05,
+      "loss": 1.6466,
+      "step": 1028
+    },
+    {
+      "epoch": 0.106886880648177,
+      "grad_norm": 0.3534831404685974,
+      "learning_rate": 9.720743134351457e-05,
+      "loss": 1.8349,
+      "step": 1029
+    },
+    {
+      "epoch": 0.10699075516775734,
+      "grad_norm": 0.33681240677833557,
+      "learning_rate": 9.720205219488312e-05,
+      "loss": 1.7535,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1070946296873377,
+      "grad_norm": 0.3691225051879883,
+      "learning_rate": 9.71966680196013e-05,
+      "loss": 1.7835,
+      "step": 1031
+    },
+    {
+      "epoch": 0.10719850420691804,
+      "grad_norm": 0.3454303443431854,
+      "learning_rate": 9.719127881824249e-05,
+      "loss": 1.7698,
+      "step": 1032
+    },
+    {
+      "epoch": 0.10730237872649839,
+      "grad_norm": 0.3353855311870575,
+      "learning_rate": 9.718588459138059e-05,
+      "loss": 1.8369,
+      "step": 1033
+    },
+    {
+      "epoch": 0.10740625324607873,
+      "grad_norm": 0.3854605257511139,
+      "learning_rate": 9.718048533959004e-05,
+      "loss": 1.8615,
+      "step": 1034
+    },
+    {
+      "epoch": 0.10751012776565909,
+      "grad_norm": 0.34844595193862915,
+      "learning_rate": 9.717508106344582e-05,
+      "loss": 1.9681,
+      "step": 1035
+    },
+    {
+      "epoch": 0.10761400228523943,
+      "grad_norm": 0.3415117561817169,
+      "learning_rate": 9.716967176352345e-05,
+      "loss": 1.8015,
+      "step": 1036
+    },
+    {
+      "epoch": 0.10771787680481978,
+      "grad_norm": 0.3829508125782013,
+      "learning_rate": 9.716425744039898e-05,
+      "loss": 1.9817,
+      "step": 1037
+    },
+    {
+      "epoch": 0.10782175132440013,
+      "grad_norm": 0.3588506877422333,
+      "learning_rate": 9.715883809464899e-05,
+      "loss": 1.7871,
+      "step": 1038
+    },
+    {
+      "epoch": 0.10792562584398047,
+      "grad_norm": 0.36895090341567993,
+      "learning_rate": 9.715341372685059e-05,
+      "loss": 1.9029,
+      "step": 1039
+    },
+    {
+      "epoch": 0.10802950036356082,
+      "grad_norm": 0.4375429153442383,
+      "learning_rate": 9.714798433758144e-05,
+      "loss": 1.6931,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10813337488314116,
+      "grad_norm": 0.34080272912979126,
+      "learning_rate": 9.714254992741971e-05,
+      "loss": 1.7422,
+      "step": 1041
+    },
+    {
+      "epoch": 0.10823724940272152,
+      "grad_norm": 0.35768958926200867,
+      "learning_rate": 9.713711049694416e-05,
+      "loss": 1.7005,
+      "step": 1042
+    },
+    {
+      "epoch": 0.10834112392230186,
+      "grad_norm": 0.35327664017677307,
+      "learning_rate": 9.713166604673403e-05,
+      "loss": 1.7248,
+      "step": 1043
+    },
+    {
+      "epoch": 0.10844499844188221,
+      "grad_norm": 0.33711856603622437,
+      "learning_rate": 9.71262165773691e-05,
+      "loss": 1.719,
+      "step": 1044
+    },
+    {
+      "epoch": 0.10854887296146255,
+      "grad_norm": 0.35835373401641846,
+      "learning_rate": 9.712076208942968e-05,
+      "loss": 1.7452,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1086527474810429,
+      "grad_norm": 0.3395202159881592,
+      "learning_rate": 9.711530258349668e-05,
+      "loss": 1.7313,
+      "step": 1046
+    },
+    {
+      "epoch": 0.10875662200062325,
+      "grad_norm": 0.3273974061012268,
+      "learning_rate": 9.710983806015148e-05,
+      "loss": 1.7674,
+      "step": 1047
+    },
+    {
+      "epoch": 0.10886049652020359,
+      "grad_norm": 0.35871952772140503,
+      "learning_rate": 9.710436851997598e-05,
+      "loss": 1.964,
+      "step": 1048
+    },
+    {
+      "epoch": 0.10896437103978394,
+      "grad_norm": 0.3521716296672821,
+      "learning_rate": 9.709889396355265e-05,
+      "loss": 1.8366,
+      "step": 1049
+    },
+    {
+      "epoch": 0.10906824555936429,
+      "grad_norm": 0.37341251969337463,
+      "learning_rate": 9.709341439146452e-05,
+      "loss": 1.8713,
+      "step": 1050
+    },
+    {
+      "epoch": 0.10917212007894464,
+      "grad_norm": 0.37906524538993835,
+      "learning_rate": 9.708792980429508e-05,
+      "loss": 1.988,
+      "step": 1051
+    },
+    {
+      "epoch": 0.10927599459852498,
+      "grad_norm": 0.3761735260486603,
+      "learning_rate": 9.708244020262842e-05,
+      "loss": 1.9337,
+      "step": 1052
+    },
+    {
+      "epoch": 0.10937986911810534,
+      "grad_norm": 0.33354657888412476,
+      "learning_rate": 9.707694558704914e-05,
+      "loss": 1.8026,
+      "step": 1053
+    },
+    {
+      "epoch": 0.10948374363768568,
+      "grad_norm": 0.3411722779273987,
+      "learning_rate": 9.707144595814237e-05,
+      "loss": 1.6728,
+      "step": 1054
+    },
+    {
+      "epoch": 0.10958761815726602,
+      "grad_norm": 0.3556554913520813,
+      "learning_rate": 9.706594131649378e-05,
+      "loss": 1.8825,
+      "step": 1055
+    },
+    {
+      "epoch": 0.10969149267684637,
+      "grad_norm": 0.3727862536907196,
+      "learning_rate": 9.706043166268957e-05,
+      "loss": 1.8012,
+      "step": 1056
+    },
+    {
+      "epoch": 0.10979536719642671,
+      "grad_norm": 0.3387303650379181,
+      "learning_rate": 9.705491699731646e-05,
+      "loss": 1.6768,
+      "step": 1057
+    },
+    {
+      "epoch": 0.10989924171600707,
+      "grad_norm": 0.36352434754371643,
+      "learning_rate": 9.704939732096175e-05,
+      "loss": 1.8888,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11000311623558741,
+      "grad_norm": 0.3399582803249359,
+      "learning_rate": 9.704387263421322e-05,
+      "loss": 1.6474,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11010699075516776,
+      "grad_norm": 0.41029804944992065,
+      "learning_rate": 9.70383429376592e-05,
+      "loss": 1.9323,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1102108652747481,
+      "grad_norm": 0.40055909752845764,
+      "learning_rate": 9.703280823188858e-05,
+      "loss": 1.8642,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11031473979432845,
+      "grad_norm": 0.365138441324234,
+      "learning_rate": 9.702726851749074e-05,
+      "loss": 1.7669,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1104186143139088,
+      "grad_norm": 0.3488384187221527,
+      "learning_rate": 9.702172379505565e-05,
+      "loss": 1.6532,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11052248883348914,
+      "grad_norm": 0.34194380044937134,
+      "learning_rate": 9.701617406517374e-05,
+      "loss": 1.6235,
+      "step": 1064
+    },
+    {
+      "epoch": 0.1106263633530695,
+      "grad_norm": 0.3830929696559906,
+      "learning_rate": 9.701061932843605e-05,
+      "loss": 1.9023,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11073023787264984,
+      "grad_norm": 0.3511880934238434,
+      "learning_rate": 9.700505958543409e-05,
+      "loss": 1.6955,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11083411239223019,
+      "grad_norm": 0.35780179500579834,
+      "learning_rate": 9.699949483675994e-05,
+      "loss": 1.7417,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11093798691181053,
+      "grad_norm": 0.3555978536605835,
+      "learning_rate": 9.69939250830062e-05,
+      "loss": 1.7226,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11104186143139087,
+      "grad_norm": 0.34007731080055237,
+      "learning_rate": 9.698835032476598e-05,
+      "loss": 1.7571,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11114573595097123,
+      "grad_norm": 0.34321436285972595,
+      "learning_rate": 9.6982770562633e-05,
+      "loss": 1.8685,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11124961047055157,
+      "grad_norm": 0.32458704710006714,
+      "learning_rate": 9.697718579720144e-05,
+      "loss": 1.6949,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11135348499013192,
+      "grad_norm": 0.37475547194480896,
+      "learning_rate": 9.697159602906603e-05,
+      "loss": 1.9906,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11145735950971226,
+      "grad_norm": 0.38911253213882446,
+      "learning_rate": 9.696600125882202e-05,
+      "loss": 1.9341,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11156123402929262,
+      "grad_norm": 0.3872954249382019,
+      "learning_rate": 9.696040148706523e-05,
+      "loss": 1.929,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11166510854887296,
+      "grad_norm": 0.3806956708431244,
+      "learning_rate": 9.695479671439198e-05,
+      "loss": 1.8981,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11176898306845331,
+      "grad_norm": 0.3513946235179901,
+      "learning_rate": 9.694918694139915e-05,
+      "loss": 1.689,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11187285758803366,
+      "grad_norm": 0.3258873522281647,
+      "learning_rate": 9.694357216868413e-05,
+      "loss": 1.6753,
+      "step": 1077
+    },
+    {
+      "epoch": 0.111976732107614,
+      "grad_norm": 0.36498063802719116,
+      "learning_rate": 9.693795239684484e-05,
+      "loss": 1.6901,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11208060662719435,
+      "grad_norm": 0.34291768074035645,
+      "learning_rate": 9.693232762647976e-05,
+      "loss": 1.6372,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11218448114677469,
+      "grad_norm": 0.38301602005958557,
+      "learning_rate": 9.692669785818786e-05,
+      "loss": 1.8587,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11228835566635505,
+      "grad_norm": 0.3735436201095581,
+      "learning_rate": 9.69210630925687e-05,
+      "loss": 1.9017,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11239223018593539,
+      "grad_norm": 0.34948012232780457,
+      "learning_rate": 9.69154233302223e-05,
+      "loss": 1.7744,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11249610470551574,
+      "grad_norm": 0.30937257409095764,
+      "learning_rate": 9.690977857174927e-05,
+      "loss": 1.6596,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11259997922509608,
+      "grad_norm": 0.3430674970149994,
+      "learning_rate": 9.690412881775076e-05,
+      "loss": 1.8651,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11270385374467642,
+      "grad_norm": 0.4052720069885254,
+      "learning_rate": 9.689847406882838e-05,
+      "loss": 1.9878,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11280772826425678,
+      "grad_norm": 0.3367058038711548,
+      "learning_rate": 9.689281432558432e-05,
+      "loss": 1.9459,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11291160278383712,
+      "grad_norm": 0.32296720147132874,
+      "learning_rate": 9.688714958862133e-05,
+      "loss": 1.7604,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11301547730341747,
+      "grad_norm": 0.36404433846473694,
+      "learning_rate": 9.688147985854264e-05,
+      "loss": 1.8102,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11311935182299782,
+      "grad_norm": 0.4039054811000824,
+      "learning_rate": 9.687580513595202e-05,
+      "loss": 1.8611,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11322322634257817,
+      "grad_norm": 0.4218135476112366,
+      "learning_rate": 9.687012542145382e-05,
+      "loss": 1.962,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11332710086215851,
+      "grad_norm": 0.36868998408317566,
+      "learning_rate": 9.686444071565285e-05,
+      "loss": 1.7224,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11343097538173887,
+      "grad_norm": 0.3493179380893707,
+      "learning_rate": 9.68587510191545e-05,
+      "loss": 1.715,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11353484990131921,
+      "grad_norm": 0.3461728096008301,
+      "learning_rate": 9.685305633256469e-05,
+      "loss": 1.6688,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11363872442089955,
+      "grad_norm": 0.35678285360336304,
+      "learning_rate": 9.684735665648983e-05,
+      "loss": 1.851,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1137425989404799,
+      "grad_norm": 0.3611486852169037,
+      "learning_rate": 9.684165199153692e-05,
+      "loss": 1.8039,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11384647346006024,
+      "grad_norm": 0.3748013377189636,
+      "learning_rate": 9.683594233831345e-05,
+      "loss": 1.8317,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1139503479796406,
+      "grad_norm": 0.39046037197113037,
+      "learning_rate": 9.683022769742746e-05,
+      "loss": 1.9756,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11405422249922094,
+      "grad_norm": 0.32861414551734924,
+      "learning_rate": 9.68245080694875e-05,
+      "loss": 1.6867,
+      "step": 1098
+    },
+    {
+      "epoch": 0.1141580970188013,
+      "grad_norm": 0.4004691541194916,
+      "learning_rate": 9.68187834551027e-05,
+      "loss": 1.9397,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11426197153838163,
+      "grad_norm": 0.33561286330223083,
+      "learning_rate": 9.681305385488264e-05,
+      "loss": 1.7001,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11436584605796198,
+      "grad_norm": 0.3517715036869049,
+      "learning_rate": 9.680731926943752e-05,
+      "loss": 1.774,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11446972057754233,
+      "grad_norm": 0.37641438841819763,
+      "learning_rate": 9.680157969937799e-05,
+      "loss": 1.7585,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11457359509712267,
+      "grad_norm": 0.36919966340065,
+      "learning_rate": 9.67958351453153e-05,
+      "loss": 1.8175,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11467746961670303,
+      "grad_norm": 0.3542039096355438,
+      "learning_rate": 9.679008560786117e-05,
+      "loss": 1.7941,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11478134413628337,
+      "grad_norm": 0.3409385681152344,
+      "learning_rate": 9.678433108762791e-05,
+      "loss": 1.7144,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11488521865586372,
+      "grad_norm": 0.3563219904899597,
+      "learning_rate": 9.677857158522833e-05,
+      "loss": 1.9284,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11498909317544406,
+      "grad_norm": 0.3560921251773834,
+      "learning_rate": 9.677280710127575e-05,
+      "loss": 1.5077,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11509296769502442,
+      "grad_norm": 0.43264439702033997,
+      "learning_rate": 9.676703763638406e-05,
+      "loss": 2.0279,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11519684221460476,
+      "grad_norm": 0.36117497086524963,
+      "learning_rate": 9.676126319116766e-05,
+      "loss": 1.9437,
+      "step": 1109
+    },
+    {
+      "epoch": 0.1153007167341851,
+      "grad_norm": 0.3572199046611786,
+      "learning_rate": 9.675548376624148e-05,
+      "loss": 1.8091,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11540459125376545,
+      "grad_norm": 0.35779669880867004,
+      "learning_rate": 9.674969936222099e-05,
+      "loss": 1.7403,
+      "step": 1111
+    },
+    {
+      "epoch": 0.1155084657733458,
+      "grad_norm": 0.36604002118110657,
+      "learning_rate": 9.674390997972218e-05,
+      "loss": 1.9467,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11561234029292615,
+      "grad_norm": 0.38619542121887207,
+      "learning_rate": 9.673811561936157e-05,
+      "loss": 1.8208,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11571621481250649,
+      "grad_norm": 0.33228105306625366,
+      "learning_rate": 9.673231628175622e-05,
+      "loss": 1.7336,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11582008933208685,
+      "grad_norm": 0.3648672103881836,
+      "learning_rate": 9.672651196752371e-05,
+      "loss": 1.9191,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11592396385166719,
+      "grad_norm": 0.33783960342407227,
+      "learning_rate": 9.672070267728215e-05,
+      "loss": 1.7849,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11602783837124753,
+      "grad_norm": 0.4384811818599701,
+      "learning_rate": 9.67148884116502e-05,
+      "loss": 2.076,
+      "step": 1117
+    },
+    {
+      "epoch": 0.11613171289082788,
+      "grad_norm": 0.3607281446456909,
+      "learning_rate": 9.670906917124703e-05,
+      "loss": 1.7839,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11623558741040822,
+      "grad_norm": 0.3579871356487274,
+      "learning_rate": 9.670324495669231e-05,
+      "loss": 1.7778,
+      "step": 1119
+    },
+    {
+      "epoch": 0.11633946192998858,
+      "grad_norm": 0.3565009832382202,
+      "learning_rate": 9.669741576860634e-05,
+      "loss": 1.8653,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11644333644956892,
+      "grad_norm": 0.36752045154571533,
+      "learning_rate": 9.669158160760982e-05,
+      "loss": 1.9041,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11654721096914927,
+      "grad_norm": 0.32525351643562317,
+      "learning_rate": 9.668574247432407e-05,
+      "loss": 1.5888,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11665108548872961,
+      "grad_norm": 0.3645164370536804,
+      "learning_rate": 9.667989836937092e-05,
+      "loss": 1.8802,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11675496000830995,
+      "grad_norm": 0.34458303451538086,
+      "learning_rate": 9.66740492933727e-05,
+      "loss": 1.8201,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11685883452789031,
+      "grad_norm": 0.33625590801239014,
+      "learning_rate": 9.66681952469523e-05,
+      "loss": 1.758,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11696270904747065,
+      "grad_norm": 0.3612918257713318,
+      "learning_rate": 9.666233623073315e-05,
+      "loss": 1.8567,
+      "step": 1126
+    },
+    {
+      "epoch": 0.117066583567051,
+      "grad_norm": 0.3521936535835266,
+      "learning_rate": 9.665647224533915e-05,
+      "loss": 1.8514,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11717045808663135,
+      "grad_norm": 0.42195868492126465,
+      "learning_rate": 9.665060329139479e-05,
+      "loss": 1.9686,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1172743326062117,
+      "grad_norm": 0.35592785477638245,
+      "learning_rate": 9.66447293695251e-05,
+      "loss": 1.6642,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11737820712579204,
+      "grad_norm": 0.3389096260070801,
+      "learning_rate": 9.663885048035555e-05,
+      "loss": 1.6711,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1174820816453724,
+      "grad_norm": 0.3547491729259491,
+      "learning_rate": 9.663296662451221e-05,
+      "loss": 1.7615,
+      "step": 1131
+    },
+    {
+      "epoch": 0.11758595616495274,
+      "grad_norm": 0.33912357687950134,
+      "learning_rate": 9.66270778026217e-05,
+      "loss": 1.6615,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11768983068453308,
+      "grad_norm": 0.37360331416130066,
+      "learning_rate": 9.662118401531108e-05,
+      "loss": 1.8592,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11779370520411343,
+      "grad_norm": 0.3422006666660309,
+      "learning_rate": 9.661528526320804e-05,
+      "loss": 1.8253,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11789757972369377,
+      "grad_norm": 0.36613208055496216,
+      "learning_rate": 9.660938154694075e-05,
+      "loss": 1.592,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11800145424327413,
+      "grad_norm": 0.38055816292762756,
+      "learning_rate": 9.660347286713787e-05,
+      "loss": 1.6502,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11810532876285447,
+      "grad_norm": 0.3511364459991455,
+      "learning_rate": 9.659755922442866e-05,
+      "loss": 1.6945,
+      "step": 1137
+    },
+    {
+      "epoch": 0.11820920328243482,
+      "grad_norm": 0.3394559621810913,
+      "learning_rate": 9.659164061944285e-05,
+      "loss": 1.7614,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11831307780201517,
+      "grad_norm": 0.38711434602737427,
+      "learning_rate": 9.658571705281076e-05,
+      "loss": 1.8105,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1184169523215955,
+      "grad_norm": 0.3627341389656067,
+      "learning_rate": 9.657978852516319e-05,
+      "loss": 1.6527,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11852082684117586,
+      "grad_norm": 0.3539731800556183,
+      "learning_rate": 9.657385503713146e-05,
+      "loss": 1.8393,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1186247013607562,
+      "grad_norm": 0.3663812279701233,
+      "learning_rate": 9.656791658934747e-05,
+      "loss": 1.7833,
+      "step": 1142
+    },
+    {
+      "epoch": 0.11872857588033656,
+      "grad_norm": 0.3631526231765747,
+      "learning_rate": 9.656197318244359e-05,
+      "loss": 1.9837,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1188324503999169,
+      "grad_norm": 0.382820725440979,
+      "learning_rate": 9.655602481705279e-05,
+      "loss": 1.8836,
+      "step": 1144
+    },
+    {
+      "epoch": 0.11893632491949725,
+      "grad_norm": 0.33744722604751587,
+      "learning_rate": 9.655007149380849e-05,
+      "loss": 1.7552,
+      "step": 1145
+    },
+    {
+      "epoch": 0.11904019943907759,
+      "grad_norm": 0.34012195467948914,
+      "learning_rate": 9.654411321334468e-05,
+      "loss": 1.6319,
+      "step": 1146
+    },
+    {
+      "epoch": 0.11914407395865795,
+      "grad_norm": 0.34437379240989685,
+      "learning_rate": 9.653814997629587e-05,
+      "loss": 1.6788,
+      "step": 1147
+    },
+    {
+      "epoch": 0.11924794847823829,
+      "grad_norm": 0.3765188753604889,
+      "learning_rate": 9.653218178329709e-05,
+      "loss": 1.8694,
+      "step": 1148
+    },
+    {
+      "epoch": 0.11935182299781863,
+      "grad_norm": 0.341296911239624,
+      "learning_rate": 9.652620863498393e-05,
+      "loss": 1.7641,
+      "step": 1149
+    },
+    {
+      "epoch": 0.11945569751739898,
+      "grad_norm": 0.37878480553627014,
+      "learning_rate": 9.652023053199247e-05,
+      "loss": 1.8416,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11955957203697933,
+      "grad_norm": 0.39128556847572327,
+      "learning_rate": 9.651424747495931e-05,
+      "loss": 1.9727,
+      "step": 1151
+    },
+    {
+      "epoch": 0.11966344655655968,
+      "grad_norm": 0.32593607902526855,
+      "learning_rate": 9.650825946452164e-05,
+      "loss": 1.6429,
+      "step": 1152
+    },
+    {
+      "epoch": 0.11976732107614002,
+      "grad_norm": 0.365972101688385,
+      "learning_rate": 9.65022665013171e-05,
+      "loss": 1.7789,
+      "step": 1153
+    },
+    {
+      "epoch": 0.11987119559572038,
+      "grad_norm": 0.3774997889995575,
+      "learning_rate": 9.64962685859839e-05,
+      "loss": 1.8482,
+      "step": 1154
+    },
+    {
+      "epoch": 0.11997507011530072,
+      "grad_norm": 0.36030101776123047,
+      "learning_rate": 9.64902657191608e-05,
+      "loss": 1.621,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12007894463488106,
+      "grad_norm": 0.40642428398132324,
+      "learning_rate": 9.648425790148704e-05,
+      "loss": 1.869,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12018281915446141,
+      "grad_norm": 0.374668151140213,
+      "learning_rate": 9.64782451336024e-05,
+      "loss": 1.8155,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12028669367404175,
+      "grad_norm": 0.33325710892677307,
+      "learning_rate": 9.64722274161472e-05,
+      "loss": 1.6419,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12039056819362211,
+      "grad_norm": 0.36537379026412964,
+      "learning_rate": 9.646620474976226e-05,
+      "loss": 1.8969,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12049444271320245,
+      "grad_norm": 0.3427730202674866,
+      "learning_rate": 9.6460177135089e-05,
+      "loss": 1.7814,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1205983172327828,
+      "grad_norm": 0.39888283610343933,
+      "learning_rate": 9.645414457276925e-05,
+      "loss": 1.8611,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12070219175236314,
+      "grad_norm": 0.35970908403396606,
+      "learning_rate": 9.644810706344547e-05,
+      "loss": 1.838,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1208060662719435,
+      "grad_norm": 0.35236963629722595,
+      "learning_rate": 9.644206460776058e-05,
+      "loss": 1.7295,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12090994079152384,
+      "grad_norm": 0.3539031445980072,
+      "learning_rate": 9.643601720635809e-05,
+      "loss": 1.8078,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12101381531110418,
+      "grad_norm": 0.3607688546180725,
+      "learning_rate": 9.642996485988198e-05,
+      "loss": 1.8958,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12111768983068454,
+      "grad_norm": 0.363936185836792,
+      "learning_rate": 9.642390756897677e-05,
+      "loss": 1.7418,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12122156435026488,
+      "grad_norm": 0.3548314571380615,
+      "learning_rate": 9.641784533428754e-05,
+      "loss": 1.6745,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12132543886984523,
+      "grad_norm": 0.43917304277420044,
+      "learning_rate": 9.641177815645984e-05,
+      "loss": 2.0477,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12142931338942557,
+      "grad_norm": 0.36484494805336,
+      "learning_rate": 9.64057060361398e-05,
+      "loss": 1.8097,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12153318790900593,
+      "grad_norm": 0.41941696405410767,
+      "learning_rate": 9.639962897397404e-05,
+      "loss": 1.921,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12163706242858627,
+      "grad_norm": 0.3769179582595825,
+      "learning_rate": 9.639354697060974e-05,
+      "loss": 1.9793,
+      "step": 1171
+    },
+    {
+      "epoch": 0.12174093694816661,
+      "grad_norm": 0.3361814618110657,
+      "learning_rate": 9.638746002669455e-05,
+      "loss": 1.685,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12184481146774696,
+      "grad_norm": 0.3244837522506714,
+      "learning_rate": 9.638136814287672e-05,
+      "loss": 1.6935,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1219486859873273,
+      "grad_norm": 0.3598596751689911,
+      "learning_rate": 9.637527131980496e-05,
+      "loss": 1.6975,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12205256050690766,
+      "grad_norm": 0.396697461605072,
+      "learning_rate": 9.636916955812856e-05,
+      "loss": 1.982,
+      "step": 1175
+    },
+    {
+      "epoch": 0.122156435026488,
+      "grad_norm": 0.37619683146476746,
+      "learning_rate": 9.636306285849728e-05,
+      "loss": 1.8531,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12226030954606835,
+      "grad_norm": 0.3777252733707428,
+      "learning_rate": 9.635695122156145e-05,
+      "loss": 1.9038,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1223641840656487,
+      "grad_norm": 0.39798396825790405,
+      "learning_rate": 9.635083464797192e-05,
+      "loss": 2.0153,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12246805858522904,
+      "grad_norm": 0.3506300449371338,
+      "learning_rate": 9.634471313838005e-05,
+      "loss": 1.743,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12257193310480939,
+      "grad_norm": 0.3415655493736267,
+      "learning_rate": 9.633858669343774e-05,
+      "loss": 1.7878,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12267580762438973,
+      "grad_norm": 0.39789021015167236,
+      "learning_rate": 9.633245531379739e-05,
+      "loss": 1.9123,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12277968214397009,
+      "grad_norm": 0.37617623805999756,
+      "learning_rate": 9.632631900011197e-05,
+      "loss": 1.8383,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12288355666355043,
+      "grad_norm": 0.37131214141845703,
+      "learning_rate": 9.632017775303492e-05,
+      "loss": 1.817,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12298743118313078,
+      "grad_norm": 0.4080236256122589,
+      "learning_rate": 9.631403157322026e-05,
+      "loss": 1.9845,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12309130570271112,
+      "grad_norm": 0.37740015983581543,
+      "learning_rate": 9.630788046132251e-05,
+      "loss": 1.904,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12319518022229148,
+      "grad_norm": 0.33329272270202637,
+      "learning_rate": 9.63017244179967e-05,
+      "loss": 1.7073,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12329905474187182,
+      "grad_norm": 0.3489798307418823,
+      "learning_rate": 9.629556344389839e-05,
+      "loss": 1.7045,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12340292926145216,
+      "grad_norm": 0.3395266532897949,
+      "learning_rate": 9.628939753968371e-05,
+      "loss": 1.7442,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12350680378103251,
+      "grad_norm": 0.3406863808631897,
+      "learning_rate": 9.628322670600926e-05,
+      "loss": 1.7656,
+      "step": 1189
+    },
+    {
+      "epoch": 0.12361067830061286,
+      "grad_norm": 0.35596734285354614,
+      "learning_rate": 9.627705094353219e-05,
+      "loss": 2.0144,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12371455282019321,
+      "grad_norm": 0.3764468729496002,
+      "learning_rate": 9.627087025291015e-05,
+      "loss": 1.8333,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12381842733977355,
+      "grad_norm": 0.3464232087135315,
+      "learning_rate": 9.626468463480138e-05,
+      "loss": 1.5966,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1239223018593539,
+      "grad_norm": 0.35695046186447144,
+      "learning_rate": 9.625849408986456e-05,
+      "loss": 1.7578,
+      "step": 1193
+    },
+    {
+      "epoch": 0.12402617637893425,
+      "grad_norm": 0.3387443721294403,
+      "learning_rate": 9.625229861875897e-05,
+      "loss": 1.6883,
+      "step": 1194
+    },
+    {
+      "epoch": 0.12413005089851459,
+      "grad_norm": 0.3431154191493988,
+      "learning_rate": 9.624609822214434e-05,
+      "loss": 1.534,
+      "step": 1195
+    },
+    {
+      "epoch": 0.12423392541809494,
+      "grad_norm": 0.3574540317058563,
+      "learning_rate": 9.623989290068099e-05,
+      "loss": 1.804,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12433779993767528,
+      "grad_norm": 0.35088303685188293,
+      "learning_rate": 9.623368265502973e-05,
+      "loss": 1.8282,
+      "step": 1197
+    },
+    {
+      "epoch": 0.12444167445725564,
+      "grad_norm": 0.35268113017082214,
+      "learning_rate": 9.62274674858519e-05,
+      "loss": 1.748,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12454554897683598,
+      "grad_norm": 0.4166131019592285,
+      "learning_rate": 9.622124739380936e-05,
+      "loss": 1.9086,
+      "step": 1199
+    },
+    {
+      "epoch": 0.12464942349641633,
+      "grad_norm": 0.38566648960113525,
+      "learning_rate": 9.621502237956452e-05,
+      "loss": 1.7812,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12475329801599667,
+      "grad_norm": 0.35631853342056274,
+      "learning_rate": 9.62087924437803e-05,
+      "loss": 1.6562,
+      "step": 1201
+    },
+    {
+      "epoch": 0.12485717253557703,
+      "grad_norm": 0.34664246439933777,
+      "learning_rate": 9.620255758712012e-05,
+      "loss": 1.743,
+      "step": 1202
+    },
+    {
+      "epoch": 0.12496104705515737,
+      "grad_norm": 0.4030589759349823,
+      "learning_rate": 9.619631781024797e-05,
+      "loss": 1.7807,
+      "step": 1203
+    },
+    {
+      "epoch": 0.12506492157473773,
+      "grad_norm": 0.3608097732067108,
+      "learning_rate": 9.619007311382828e-05,
+      "loss": 1.8727,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12516879609431805,
+      "grad_norm": 0.3390495777130127,
+      "learning_rate": 9.618382349852612e-05,
+      "loss": 1.6474,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1252726706138984,
+      "grad_norm": 0.34509339928627014,
+      "learning_rate": 9.617756896500701e-05,
+      "loss": 1.7651,
+      "step": 1206
+    },
+    {
+      "epoch": 0.12537654513347876,
+      "grad_norm": 0.37745997309684753,
+      "learning_rate": 9.6171309513937e-05,
+      "loss": 1.6898,
+      "step": 1207
+    },
+    {
+      "epoch": 0.12548041965305912,
+      "grad_norm": 0.33759596943855286,
+      "learning_rate": 9.616504514598267e-05,
+      "loss": 1.5825,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12558429417263944,
+      "grad_norm": 0.3818475008010864,
+      "learning_rate": 9.615877586181114e-05,
+      "loss": 1.7711,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1256881686922198,
+      "grad_norm": 0.3622841536998749,
+      "learning_rate": 9.615250166209003e-05,
+      "loss": 1.8772,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12579204321180015,
+      "grad_norm": 0.38078466057777405,
+      "learning_rate": 9.614622254748749e-05,
+      "loss": 1.6718,
+      "step": 1211
+    },
+    {
+      "epoch": 0.12589591773138048,
+      "grad_norm": 0.3571094572544098,
+      "learning_rate": 9.613993851867221e-05,
+      "loss": 1.7133,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12599979225096083,
+      "grad_norm": 0.3744667172431946,
+      "learning_rate": 9.613364957631339e-05,
+      "loss": 1.8524,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1261036667705412,
+      "grad_norm": 0.3752327561378479,
+      "learning_rate": 9.612735572108073e-05,
+      "loss": 1.7747,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12620754129012154,
+      "grad_norm": 0.3768329620361328,
+      "learning_rate": 9.61210569536445e-05,
+      "loss": 1.7503,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12631141580970187,
+      "grad_norm": 0.35730409622192383,
+      "learning_rate": 9.611475327467547e-05,
+      "loss": 1.8134,
+      "step": 1216
+    },
+    {
+      "epoch": 0.12641529032928223,
+      "grad_norm": 0.38782814145088196,
+      "learning_rate": 9.610844468484492e-05,
+      "loss": 1.9043,
+      "step": 1217
+    },
+    {
+      "epoch": 0.12651916484886258,
+      "grad_norm": 0.35042253136634827,
+      "learning_rate": 9.610213118482468e-05,
+      "loss": 1.6922,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1266230393684429,
+      "grad_norm": 0.3477722108364105,
+      "learning_rate": 9.609581277528708e-05,
+      "loss": 1.8249,
+      "step": 1219
+    },
+    {
+      "epoch": 0.12672691388802326,
+      "grad_norm": 0.358598530292511,
+      "learning_rate": 9.608948945690496e-05,
+      "loss": 1.6571,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12683078840760362,
+      "grad_norm": 0.34587663412094116,
+      "learning_rate": 9.608316123035172e-05,
+      "loss": 1.8308,
+      "step": 1221
+    },
+    {
+      "epoch": 0.12693466292718397,
+      "grad_norm": 0.3662189543247223,
+      "learning_rate": 9.60768280963013e-05,
+      "loss": 1.8263,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1270385374467643,
+      "grad_norm": 0.4229857325553894,
+      "learning_rate": 9.60704900554281e-05,
+      "loss": 1.8465,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12714241196634465,
+      "grad_norm": 0.3771613538265228,
+      "learning_rate": 9.606414710840706e-05,
+      "loss": 1.9523,
+      "step": 1224
+    },
+    {
+      "epoch": 0.127246286485925,
+      "grad_norm": 0.37686896324157715,
+      "learning_rate": 9.605779925591366e-05,
+      "loss": 1.8436,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12735016100550534,
+      "grad_norm": 0.3408556580543518,
+      "learning_rate": 9.605144649862391e-05,
+      "loss": 1.8095,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1274540355250857,
+      "grad_norm": 0.3675098717212677,
+      "learning_rate": 9.604508883721432e-05,
+      "loss": 1.9452,
+      "step": 1227
+    },
+    {
+      "epoch": 0.12755791004466605,
+      "grad_norm": 0.3680543899536133,
+      "learning_rate": 9.603872627236194e-05,
+      "loss": 1.8987,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1276617845642464,
+      "grad_norm": 0.3140665888786316,
+      "learning_rate": 9.603235880474433e-05,
+      "loss": 1.6107,
+      "step": 1229
+    },
+    {
+      "epoch": 0.12776565908382673,
+      "grad_norm": 0.35767602920532227,
+      "learning_rate": 9.602598643503957e-05,
+      "loss": 1.5594,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12786953360340708,
+      "grad_norm": 0.40325549244880676,
+      "learning_rate": 9.601960916392627e-05,
+      "loss": 2.082,
+      "step": 1231
+    },
+    {
+      "epoch": 0.12797340812298744,
+      "grad_norm": 0.34380728006362915,
+      "learning_rate": 9.601322699208357e-05,
+      "loss": 1.7435,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1280772826425678,
+      "grad_norm": 0.3566792607307434,
+      "learning_rate": 9.60068399201911e-05,
+      "loss": 1.7594,
+      "step": 1233
+    },
+    {
+      "epoch": 0.12818115716214812,
+      "grad_norm": 0.34519535303115845,
+      "learning_rate": 9.600044794892905e-05,
+      "loss": 1.8354,
+      "step": 1234
+    },
+    {
+      "epoch": 0.12828503168172847,
+      "grad_norm": 0.36046940088272095,
+      "learning_rate": 9.599405107897811e-05,
+      "loss": 1.7715,
+      "step": 1235
+    },
+    {
+      "epoch": 0.12838890620130883,
+      "grad_norm": 0.33479416370391846,
+      "learning_rate": 9.598764931101949e-05,
+      "loss": 1.6034,
+      "step": 1236
+    },
+    {
+      "epoch": 0.12849278072088915,
+      "grad_norm": 0.3153260052204132,
+      "learning_rate": 9.598124264573495e-05,
+      "loss": 1.6561,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1285966552404695,
+      "grad_norm": 0.3522169888019562,
+      "learning_rate": 9.597483108380674e-05,
+      "loss": 1.8149,
+      "step": 1238
+    },
+    {
+      "epoch": 0.12870052976004986,
+      "grad_norm": 0.39202046394348145,
+      "learning_rate": 9.596841462591761e-05,
+      "loss": 1.8963,
+      "step": 1239
+    },
+    {
+      "epoch": 0.12880440427963022,
+      "grad_norm": 0.3598194718360901,
+      "learning_rate": 9.596199327275091e-05,
+      "loss": 1.7161,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12890827879921055,
+      "grad_norm": 0.36012735962867737,
+      "learning_rate": 9.595556702499045e-05,
+      "loss": 1.86,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1290121533187909,
+      "grad_norm": 0.32633063197135925,
+      "learning_rate": 9.594913588332055e-05,
+      "loss": 1.5897,
+      "step": 1242
+    },
+    {
+      "epoch": 0.12911602783837126,
+      "grad_norm": 0.35175660252571106,
+      "learning_rate": 9.594269984842612e-05,
+      "loss": 1.8549,
+      "step": 1243
+    },
+    {
+      "epoch": 0.12921990235795158,
+      "grad_norm": 0.3716254234313965,
+      "learning_rate": 9.59362589209925e-05,
+      "loss": 1.8379,
+      "step": 1244
+    },
+    {
+      "epoch": 0.12932377687753194,
+      "grad_norm": 0.3607979118824005,
+      "learning_rate": 9.592981310170563e-05,
+      "loss": 1.8592,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1294276513971123,
+      "grad_norm": 0.3418334424495697,
+      "learning_rate": 9.592336239125193e-05,
+      "loss": 1.7502,
+      "step": 1246
+    },
+    {
+      "epoch": 0.12953152591669265,
+      "grad_norm": 0.3638349771499634,
+      "learning_rate": 9.591690679031835e-05,
+      "loss": 1.8981,
+      "step": 1247
+    },
+    {
+      "epoch": 0.12963540043627297,
+      "grad_norm": 0.36420029401779175,
+      "learning_rate": 9.591044629959236e-05,
+      "loss": 1.913,
+      "step": 1248
+    },
+    {
+      "epoch": 0.12973927495585333,
+      "grad_norm": 0.3521292805671692,
+      "learning_rate": 9.590398091976195e-05,
+      "loss": 1.6886,
+      "step": 1249
+    },
+    {
+      "epoch": 0.12984314947543368,
+      "grad_norm": 0.3583468198776245,
+      "learning_rate": 9.589751065151567e-05,
+      "loss": 1.7328,
+      "step": 1250
+    },
+    {
+      "epoch": 0.129947023995014,
+      "grad_norm": 0.3620645999908447,
+      "learning_rate": 9.589103549554248e-05,
+      "loss": 1.9107,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13005089851459437,
+      "grad_norm": 0.361101895570755,
+      "learning_rate": 9.5884555452532e-05,
+      "loss": 1.7693,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13015477303417472,
+      "grad_norm": 0.3591129183769226,
+      "learning_rate": 9.587807052317426e-05,
+      "loss": 1.696,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13025864755375507,
+      "grad_norm": 0.3596014678478241,
+      "learning_rate": 9.587158070815988e-05,
+      "loss": 1.9454,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1303625220733354,
+      "grad_norm": 0.35805267095565796,
+      "learning_rate": 9.586508600817997e-05,
+      "loss": 1.8159,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13046639659291576,
+      "grad_norm": 0.3794294595718384,
+      "learning_rate": 9.585858642392616e-05,
+      "loss": 1.7915,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1305702711124961,
+      "grad_norm": 0.350418359041214,
+      "learning_rate": 9.585208195609059e-05,
+      "loss": 1.754,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13067414563207644,
+      "grad_norm": 0.37485983967781067,
+      "learning_rate": 9.584557260536597e-05,
+      "loss": 1.8592,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1307780201516568,
+      "grad_norm": 0.38537418842315674,
+      "learning_rate": 9.583905837244547e-05,
+      "loss": 1.814,
+      "step": 1259
+    },
+    {
+      "epoch": 0.13088189467123715,
+      "grad_norm": 0.3657982349395752,
+      "learning_rate": 9.583253925802283e-05,
+      "loss": 1.7968,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1309857691908175,
+      "grad_norm": 0.4190506935119629,
+      "learning_rate": 9.582601526279225e-05,
+      "loss": 2.0191,
+      "step": 1261
+    },
+    {
+      "epoch": 0.13108964371039783,
+      "grad_norm": 0.3682914674282074,
+      "learning_rate": 9.581948638744849e-05,
+      "loss": 1.7923,
+      "step": 1262
+    },
+    {
+      "epoch": 0.13119351822997818,
+      "grad_norm": 0.33630824089050293,
+      "learning_rate": 9.581295263268684e-05,
+      "loss": 1.6561,
+      "step": 1263
+    },
+    {
+      "epoch": 0.13129739274955854,
+      "grad_norm": 0.37690606713294983,
+      "learning_rate": 9.580641399920311e-05,
+      "loss": 1.8836,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1314012672691389,
+      "grad_norm": 0.3465617895126343,
+      "learning_rate": 9.579987048769356e-05,
+      "loss": 1.7498,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13150514178871922,
+      "grad_norm": 0.3308243155479431,
+      "learning_rate": 9.579332209885508e-05,
+      "loss": 1.6476,
+      "step": 1266
+    },
+    {
+      "epoch": 0.13160901630829958,
+      "grad_norm": 0.38376688957214355,
+      "learning_rate": 9.5786768833385e-05,
+      "loss": 1.7795,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13171289082787993,
+      "grad_norm": 0.35049548745155334,
+      "learning_rate": 9.578021069198118e-05,
+      "loss": 1.718,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13181676534746026,
+      "grad_norm": 0.34886178374290466,
+      "learning_rate": 9.577364767534202e-05,
+      "loss": 1.7214,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1319206398670406,
+      "grad_norm": 0.36722105741500854,
+      "learning_rate": 9.576707978416644e-05,
+      "loss": 1.6402,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13202451438662097,
+      "grad_norm": 0.4114188849925995,
+      "learning_rate": 9.576050701915385e-05,
+      "loss": 1.8282,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13212838890620132,
+      "grad_norm": 0.40800419449806213,
+      "learning_rate": 9.575392938100422e-05,
+      "loss": 2.0049,
+      "step": 1272
+    },
+    {
+      "epoch": 0.13223226342578165,
+      "grad_norm": 0.3313359320163727,
+      "learning_rate": 9.5747346870418e-05,
+      "loss": 1.678,
+      "step": 1273
+    },
+    {
+      "epoch": 0.132336137945362,
+      "grad_norm": 0.33527714014053345,
+      "learning_rate": 9.574075948809618e-05,
+      "loss": 1.7132,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13244001246494236,
+      "grad_norm": 0.37509602308273315,
+      "learning_rate": 9.573416723474026e-05,
+      "loss": 1.7306,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13254388698452269,
+      "grad_norm": 0.41604387760162354,
+      "learning_rate": 9.572757011105227e-05,
+      "loss": 1.8486,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13264776150410304,
+      "grad_norm": 0.3609203100204468,
+      "learning_rate": 9.572096811773477e-05,
+      "loss": 1.7497,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1327516360236834,
+      "grad_norm": 0.346780389547348,
+      "learning_rate": 9.571436125549078e-05,
+      "loss": 1.6892,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13285551054326375,
+      "grad_norm": 0.37450289726257324,
+      "learning_rate": 9.57077495250239e-05,
+      "loss": 1.77,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13295938506284408,
+      "grad_norm": 0.36386212706565857,
+      "learning_rate": 9.570113292703826e-05,
+      "loss": 1.8936,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13306325958242443,
+      "grad_norm": 0.381849080324173,
+      "learning_rate": 9.569451146223843e-05,
+      "loss": 1.7727,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13316713410200479,
+      "grad_norm": 0.3608989715576172,
+      "learning_rate": 9.568788513132955e-05,
+      "loss": 1.8763,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1332710086215851,
+      "grad_norm": 0.37927472591400146,
+      "learning_rate": 9.568125393501728e-05,
+      "loss": 1.864,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13337488314116547,
+      "grad_norm": 0.37271663546562195,
+      "learning_rate": 9.567461787400781e-05,
+      "loss": 1.7884,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13347875766074582,
+      "grad_norm": 0.34601160883903503,
+      "learning_rate": 9.566797694900779e-05,
+      "loss": 1.7951,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13358263218032618,
+      "grad_norm": 0.32820773124694824,
+      "learning_rate": 9.566133116072447e-05,
+      "loss": 1.602,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1336865066999065,
+      "grad_norm": 0.36603811383247375,
+      "learning_rate": 9.565468050986556e-05,
+      "loss": 1.7799,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13379038121948686,
+      "grad_norm": 0.33986133337020874,
+      "learning_rate": 9.564802499713928e-05,
+      "loss": 1.7055,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13389425573906721,
+      "grad_norm": 0.37927207350730896,
+      "learning_rate": 9.564136462325442e-05,
+      "loss": 1.8803,
+      "step": 1289
+    },
+    {
+      "epoch": 0.13399813025864754,
+      "grad_norm": 0.3654598593711853,
+      "learning_rate": 9.563469938892022e-05,
+      "loss": 1.8929,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1341020047782279,
+      "grad_norm": 0.38735446333885193,
+      "learning_rate": 9.562802929484651e-05,
+      "loss": 1.6864,
+      "step": 1291
+    },
+    {
+      "epoch": 0.13420587929780825,
+      "grad_norm": 0.3329310417175293,
+      "learning_rate": 9.56213543417436e-05,
+      "loss": 1.7724,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1343097538173886,
+      "grad_norm": 0.37568461894989014,
+      "learning_rate": 9.561467453032231e-05,
+      "loss": 1.794,
+      "step": 1293
+    },
+    {
+      "epoch": 0.13441362833696893,
+      "grad_norm": 0.3253065049648285,
+      "learning_rate": 9.560798986129399e-05,
+      "loss": 1.7469,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1345175028565493,
+      "grad_norm": 0.32807695865631104,
+      "learning_rate": 9.56013003353705e-05,
+      "loss": 1.7181,
+      "step": 1295
+    },
+    {
+      "epoch": 0.13462137737612964,
+      "grad_norm": 0.34811633825302124,
+      "learning_rate": 9.559460595326424e-05,
+      "loss": 1.6823,
+      "step": 1296
+    },
+    {
+      "epoch": 0.13472525189570997,
+      "grad_norm": 0.3585864007472992,
+      "learning_rate": 9.558790671568809e-05,
+      "loss": 1.8669,
+      "step": 1297
+    },
+    {
+      "epoch": 0.13482912641529032,
+      "grad_norm": 0.3390277624130249,
+      "learning_rate": 9.558120262335546e-05,
+      "loss": 1.6597,
+      "step": 1298
+    },
+    {
+      "epoch": 0.13493300093487068,
+      "grad_norm": 0.3469536006450653,
+      "learning_rate": 9.557449367698032e-05,
+      "loss": 1.8582,
+      "step": 1299
+    },
+    {
+      "epoch": 0.13503687545445103,
+      "grad_norm": 0.35477346181869507,
+      "learning_rate": 9.55677798772771e-05,
+      "loss": 1.8483,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13514074997403136,
+      "grad_norm": 0.3245786726474762,
+      "learning_rate": 9.556106122496076e-05,
+      "loss": 1.6194,
+      "step": 1301
+    },
+    {
+      "epoch": 0.13524462449361171,
+      "grad_norm": 0.37269967794418335,
+      "learning_rate": 9.555433772074678e-05,
+      "loss": 1.8113,
+      "step": 1302
+    },
+    {
+      "epoch": 0.13534849901319207,
+      "grad_norm": 0.41088709235191345,
+      "learning_rate": 9.554760936535116e-05,
+      "loss": 1.9416,
+      "step": 1303
+    },
+    {
+      "epoch": 0.13545237353277242,
+      "grad_norm": 0.35153427720069885,
+      "learning_rate": 9.554087615949046e-05,
+      "loss": 1.6847,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13555624805235275,
+      "grad_norm": 0.38244035840034485,
+      "learning_rate": 9.553413810388165e-05,
+      "loss": 1.8235,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1356601225719331,
+      "grad_norm": 0.33529362082481384,
+      "learning_rate": 9.552739519924235e-05,
+      "loss": 1.6378,
+      "step": 1306
+    },
+    {
+      "epoch": 0.13576399709151346,
+      "grad_norm": 0.335685133934021,
+      "learning_rate": 9.552064744629054e-05,
+      "loss": 1.8144,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1358678716110938,
+      "grad_norm": 0.33851975202560425,
+      "learning_rate": 9.551389484574489e-05,
+      "loss": 1.6885,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13597174613067414,
+      "grad_norm": 0.4007343649864197,
+      "learning_rate": 9.550713739832444e-05,
+      "loss": 2.0096,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1360756206502545,
+      "grad_norm": 0.3581026494503021,
+      "learning_rate": 9.550037510474883e-05,
+      "loss": 1.7002,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13617949516983485,
+      "grad_norm": 0.3608642518520355,
+      "learning_rate": 9.549360796573821e-05,
+      "loss": 1.7883,
+      "step": 1311
+    },
+    {
+      "epoch": 0.13628336968941518,
+      "grad_norm": 0.39638829231262207,
+      "learning_rate": 9.548683598201317e-05,
+      "loss": 1.7193,
+      "step": 1312
+    },
+    {
+      "epoch": 0.13638724420899553,
+      "grad_norm": 0.3727129399776459,
+      "learning_rate": 9.548005915429495e-05,
+      "loss": 1.8664,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1364911187285759,
+      "grad_norm": 0.34268417954444885,
+      "learning_rate": 9.547327748330514e-05,
+      "loss": 1.7827,
+      "step": 1314
+    },
+    {
+      "epoch": 0.13659499324815622,
+      "grad_norm": 0.3418048918247223,
+      "learning_rate": 9.5466490969766e-05,
+      "loss": 1.7287,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13669886776773657,
+      "grad_norm": 0.3453529477119446,
+      "learning_rate": 9.545969961440025e-05,
+      "loss": 1.7565,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13680274228731693,
+      "grad_norm": 0.35973402857780457,
+      "learning_rate": 9.545290341793107e-05,
+      "loss": 1.8727,
+      "step": 1317
+    },
+    {
+      "epoch": 0.13690661680689728,
+      "grad_norm": 0.3567400872707367,
+      "learning_rate": 9.544610238108223e-05,
+      "loss": 1.8201,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1370104913264776,
+      "grad_norm": 0.37093833088874817,
+      "learning_rate": 9.543929650457796e-05,
+      "loss": 1.8134,
+      "step": 1319
+    },
+    {
+      "epoch": 0.13711436584605796,
+      "grad_norm": 0.3393130600452423,
+      "learning_rate": 9.543248578914307e-05,
+      "loss": 1.6867,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13721824036563832,
+      "grad_norm": 0.3978770673274994,
+      "learning_rate": 9.542567023550283e-05,
+      "loss": 1.9418,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13732211488521864,
+      "grad_norm": 0.33980122208595276,
+      "learning_rate": 9.541884984438304e-05,
+      "loss": 1.6986,
+      "step": 1322
+    },
+    {
+      "epoch": 0.137425989404799,
+      "grad_norm": 0.3643237054347992,
+      "learning_rate": 9.541202461651002e-05,
+      "loss": 1.6953,
+      "step": 1323
+    },
+    {
+      "epoch": 0.13752986392437935,
+      "grad_norm": 0.3416014313697815,
+      "learning_rate": 9.54051945526106e-05,
+      "loss": 1.6347,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1376337384439597,
+      "grad_norm": 0.42336559295654297,
+      "learning_rate": 9.539835965341215e-05,
+      "loss": 2.0321,
+      "step": 1325
+    },
+    {
+      "epoch": 0.13773761296354003,
+      "grad_norm": 0.3578283488750458,
+      "learning_rate": 9.539151991964253e-05,
+      "loss": 1.7978,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1378414874831204,
+      "grad_norm": 0.39390215277671814,
+      "learning_rate": 9.538467535203007e-05,
+      "loss": 2.0113,
+      "step": 1327
+    },
+    {
+      "epoch": 0.13794536200270074,
+      "grad_norm": 0.34902918338775635,
+      "learning_rate": 9.537782595130371e-05,
+      "loss": 1.7936,
+      "step": 1328
+    },
+    {
+      "epoch": 0.13804923652228107,
+      "grad_norm": 0.3545171916484833,
+      "learning_rate": 9.537097171819287e-05,
+      "loss": 1.8352,
+      "step": 1329
+    },
+    {
+      "epoch": 0.13815311104186143,
+      "grad_norm": 0.41688135266304016,
+      "learning_rate": 9.536411265342742e-05,
+      "loss": 1.9648,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13825698556144178,
+      "grad_norm": 0.34818556904792786,
+      "learning_rate": 9.535724875773784e-05,
+      "loss": 1.6804,
+      "step": 1331
+    },
+    {
+      "epoch": 0.13836086008102214,
+      "grad_norm": 0.3562292754650116,
+      "learning_rate": 9.535038003185507e-05,
+      "loss": 1.697,
+      "step": 1332
+    },
+    {
+      "epoch": 0.13846473460060246,
+      "grad_norm": 0.3660946786403656,
+      "learning_rate": 9.534350647651056e-05,
+      "loss": 1.8382,
+      "step": 1333
+    },
+    {
+      "epoch": 0.13856860912018282,
+      "grad_norm": 0.38286975026130676,
+      "learning_rate": 9.533662809243632e-05,
+      "loss": 1.8386,
+      "step": 1334
+    },
+    {
+      "epoch": 0.13867248363976317,
+      "grad_norm": 0.3726559281349182,
+      "learning_rate": 9.532974488036481e-05,
+      "loss": 1.701,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1387763581593435,
+      "grad_norm": 0.37573710083961487,
+      "learning_rate": 9.532285684102907e-05,
+      "loss": 1.9953,
+      "step": 1336
+    },
+    {
+      "epoch": 0.13888023267892385,
+      "grad_norm": 0.36941513419151306,
+      "learning_rate": 9.53159639751626e-05,
+      "loss": 1.7335,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1389841071985042,
+      "grad_norm": 0.3815058767795563,
+      "learning_rate": 9.530906628349945e-05,
+      "loss": 1.929,
+      "step": 1338
+    },
+    {
+      "epoch": 0.13908798171808456,
+      "grad_norm": 0.3447597026824951,
+      "learning_rate": 9.530216376677417e-05,
+      "loss": 1.7699,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1391918562376649,
+      "grad_norm": 0.3957836627960205,
+      "learning_rate": 9.529525642572182e-05,
+      "loss": 1.8501,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13929573075724525,
+      "grad_norm": 0.44103363156318665,
+      "learning_rate": 9.528834426107795e-05,
+      "loss": 1.9139,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1393996052768256,
+      "grad_norm": 0.3422475755214691,
+      "learning_rate": 9.52814272735787e-05,
+      "loss": 1.7613,
+      "step": 1342
+    },
+    {
+      "epoch": 0.13950347979640595,
+      "grad_norm": 0.3645455837249756,
+      "learning_rate": 9.527450546396067e-05,
+      "loss": 1.7897,
+      "step": 1343
+    },
+    {
+      "epoch": 0.13960735431598628,
+      "grad_norm": 0.3628111779689789,
+      "learning_rate": 9.526757883296094e-05,
+      "loss": 1.782,
+      "step": 1344
+    },
+    {
+      "epoch": 0.13971122883556664,
+      "grad_norm": 0.3505362868309021,
+      "learning_rate": 9.526064738131717e-05,
+      "loss": 1.7047,
+      "step": 1345
+    },
+    {
+      "epoch": 0.139815103355147,
+      "grad_norm": 0.36256182193756104,
+      "learning_rate": 9.52537111097675e-05,
+      "loss": 1.8108,
+      "step": 1346
+    },
+    {
+      "epoch": 0.13991897787472732,
+      "grad_norm": 0.34468889236450195,
+      "learning_rate": 9.524677001905059e-05,
+      "loss": 1.7674,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14002285239430767,
+      "grad_norm": 0.36335834860801697,
+      "learning_rate": 9.52398241099056e-05,
+      "loss": 1.7181,
+      "step": 1348
+    },
+    {
+      "epoch": 0.14012672691388803,
+      "grad_norm": 0.3589484393596649,
+      "learning_rate": 9.523287338307224e-05,
+      "loss": 1.8431,
+      "step": 1349
+    },
+    {
+      "epoch": 0.14023060143346838,
+      "grad_norm": 0.3601551949977875,
+      "learning_rate": 9.522591783929069e-05,
+      "loss": 1.6761,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1403344759530487,
+      "grad_norm": 0.34505337476730347,
+      "learning_rate": 9.521895747930168e-05,
+      "loss": 1.7923,
+      "step": 1351
+    },
+    {
+      "epoch": 0.14043835047262906,
+      "grad_norm": 0.3898836374282837,
+      "learning_rate": 9.521199230384641e-05,
+      "loss": 1.9424,
+      "step": 1352
+    },
+    {
+      "epoch": 0.14054222499220942,
+      "grad_norm": 0.4122181534767151,
+      "learning_rate": 9.520502231366661e-05,
+      "loss": 1.7709,
+      "step": 1353
+    },
+    {
+      "epoch": 0.14064609951178975,
+      "grad_norm": 0.36563369631767273,
+      "learning_rate": 9.519804750950457e-05,
+      "loss": 1.8535,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1407499740313701,
+      "grad_norm": 0.3425121009349823,
+      "learning_rate": 9.519106789210301e-05,
+      "loss": 1.6418,
+      "step": 1355
+    },
+    {
+      "epoch": 0.14085384855095046,
+      "grad_norm": 0.36973652243614197,
+      "learning_rate": 9.518408346220525e-05,
+      "loss": 1.7506,
+      "step": 1356
+    },
+    {
+      "epoch": 0.1409577230705308,
+      "grad_norm": 0.36737698316574097,
+      "learning_rate": 9.517709422055502e-05,
+      "loss": 1.756,
+      "step": 1357
+    },
+    {
+      "epoch": 0.14106159759011114,
+      "grad_norm": 0.36824169754981995,
+      "learning_rate": 9.517010016789667e-05,
+      "loss": 1.7849,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1411654721096915,
+      "grad_norm": 0.44827938079833984,
+      "learning_rate": 9.516310130497498e-05,
+      "loss": 2.191,
+      "step": 1359
+    },
+    {
+      "epoch": 0.14126934662927185,
+      "grad_norm": 0.33879658579826355,
+      "learning_rate": 9.51560976325353e-05,
+      "loss": 1.6468,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14137322114885217,
+      "grad_norm": 0.3727847933769226,
+      "learning_rate": 9.514908915132346e-05,
+      "loss": 1.5898,
+      "step": 1361
+    },
+    {
+      "epoch": 0.14147709566843253,
+      "grad_norm": 0.34985223412513733,
+      "learning_rate": 9.514207586208578e-05,
+      "loss": 1.5243,
+      "step": 1362
+    },
+    {
+      "epoch": 0.14158097018801288,
+      "grad_norm": 0.39218172430992126,
+      "learning_rate": 9.513505776556915e-05,
+      "loss": 1.8003,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14168484470759324,
+      "grad_norm": 0.3964877426624298,
+      "learning_rate": 9.512803486252093e-05,
+      "loss": 1.7633,
+      "step": 1364
+    },
+    {
+      "epoch": 0.14178871922717357,
+      "grad_norm": 0.3854914903640747,
+      "learning_rate": 9.512100715368902e-05,
+      "loss": 1.8179,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14189259374675392,
+      "grad_norm": 0.35278111696243286,
+      "learning_rate": 9.511397463982181e-05,
+      "loss": 1.8105,
+      "step": 1366
+    },
+    {
+      "epoch": 0.14199646826633427,
+      "grad_norm": 0.45402637124061584,
+      "learning_rate": 9.510693732166819e-05,
+      "loss": 1.9222,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1421003427859146,
+      "grad_norm": 0.3488807678222656,
+      "learning_rate": 9.50998951999776e-05,
+      "loss": 1.6662,
+      "step": 1368
+    },
+    {
+      "epoch": 0.14220421730549496,
+      "grad_norm": 0.3388180136680603,
+      "learning_rate": 9.509284827549996e-05,
+      "loss": 1.6714,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1423080918250753,
+      "grad_norm": 0.36008358001708984,
+      "learning_rate": 9.508579654898571e-05,
+      "loss": 1.8606,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14241196634465567,
+      "grad_norm": 0.37350037693977356,
+      "learning_rate": 9.507874002118582e-05,
+      "loss": 1.8999,
+      "step": 1371
+    },
+    {
+      "epoch": 0.142515840864236,
+      "grad_norm": 0.3648621737957001,
+      "learning_rate": 9.507167869285174e-05,
+      "loss": 1.9342,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14261971538381635,
+      "grad_norm": 0.3702144920825958,
+      "learning_rate": 9.506461256473545e-05,
+      "loss": 1.8062,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1427235899033967,
+      "grad_norm": 0.37488213181495667,
+      "learning_rate": 9.505754163758946e-05,
+      "loss": 1.8751,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14282746442297706,
+      "grad_norm": 0.37895387411117554,
+      "learning_rate": 9.505046591216674e-05,
+      "loss": 1.6707,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14293133894255738,
+      "grad_norm": 0.3528745174407959,
+      "learning_rate": 9.504338538922078e-05,
+      "loss": 1.8434,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14303521346213774,
+      "grad_norm": 0.3752725422382355,
+      "learning_rate": 9.503630006950566e-05,
+      "loss": 1.9281,
+      "step": 1377
+    },
+    {
+      "epoch": 0.1431390879817181,
+      "grad_norm": 0.38882383704185486,
+      "learning_rate": 9.502920995377587e-05,
+      "loss": 1.6748,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14324296250129842,
+      "grad_norm": 0.3682365417480469,
+      "learning_rate": 9.502211504278644e-05,
+      "loss": 1.767,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14334683702087878,
+      "grad_norm": 0.339139848947525,
+      "learning_rate": 9.501501533729297e-05,
+      "loss": 1.5963,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14345071154045913,
+      "grad_norm": 0.3614344298839569,
+      "learning_rate": 9.50079108380515e-05,
+      "loss": 1.7736,
+      "step": 1381
+    },
+    {
+      "epoch": 0.14355458606003949,
+      "grad_norm": 0.33216235041618347,
+      "learning_rate": 9.500080154581859e-05,
+      "loss": 1.8277,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1436584605796198,
+      "grad_norm": 0.3475283086299896,
+      "learning_rate": 9.499368746135133e-05,
+      "loss": 1.7244,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14376233509920017,
+      "grad_norm": 0.35839635133743286,
+      "learning_rate": 9.498656858540731e-05,
+      "loss": 1.8232,
+      "step": 1384
+    },
+    {
+      "epoch": 0.14386620961878052,
+      "grad_norm": 0.37137824296951294,
+      "learning_rate": 9.497944491874467e-05,
+      "loss": 1.7515,
+      "step": 1385
+    },
+    {
+      "epoch": 0.14397008413836085,
+      "grad_norm": 0.35471752285957336,
+      "learning_rate": 9.4972316462122e-05,
+      "loss": 1.8021,
+      "step": 1386
+    },
+    {
+      "epoch": 0.1440739586579412,
+      "grad_norm": 0.3838224411010742,
+      "learning_rate": 9.49651832162984e-05,
+      "loss": 1.8813,
+      "step": 1387
+    },
+    {
+      "epoch": 0.14417783317752156,
+      "grad_norm": 0.32974106073379517,
+      "learning_rate": 9.495804518203355e-05,
+      "loss": 1.7155,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1442817076971019,
+      "grad_norm": 0.35001546144485474,
+      "learning_rate": 9.495090236008756e-05,
+      "loss": 1.8771,
+      "step": 1389
+    },
+    {
+      "epoch": 0.14438558221668224,
+      "grad_norm": 0.37329480051994324,
+      "learning_rate": 9.49437547512211e-05,
+      "loss": 1.894,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1444894567362626,
+      "grad_norm": 0.35633528232574463,
+      "learning_rate": 9.493660235619535e-05,
+      "loss": 1.7673,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14459333125584295,
+      "grad_norm": 0.37964895367622375,
+      "learning_rate": 9.492944517577196e-05,
+      "loss": 2.0935,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14469720577542328,
+      "grad_norm": 0.33758020401000977,
+      "learning_rate": 9.492228321071311e-05,
+      "loss": 1.6482,
+      "step": 1393
+    },
+    {
+      "epoch": 0.14480108029500363,
+      "grad_norm": 0.39080965518951416,
+      "learning_rate": 9.491511646178151e-05,
+      "loss": 1.6859,
+      "step": 1394
+    },
+    {
+      "epoch": 0.14490495481458399,
+      "grad_norm": 0.36754679679870605,
+      "learning_rate": 9.490794492974038e-05,
+      "loss": 1.6268,
+      "step": 1395
+    },
+    {
+      "epoch": 0.14500882933416434,
+      "grad_norm": 0.33904823660850525,
+      "learning_rate": 9.490076861535339e-05,
+      "loss": 1.8217,
+      "step": 1396
+    },
+    {
+      "epoch": 0.14511270385374467,
+      "grad_norm": 0.36078742146492004,
+      "learning_rate": 9.489358751938477e-05,
+      "loss": 1.7979,
+      "step": 1397
+    },
+    {
+      "epoch": 0.14521657837332502,
+      "grad_norm": 0.37665989995002747,
+      "learning_rate": 9.488640164259929e-05,
+      "loss": 1.7041,
+      "step": 1398
+    },
+    {
+      "epoch": 0.14532045289290538,
+      "grad_norm": 0.3702797591686249,
+      "learning_rate": 9.487921098576216e-05,
+      "loss": 1.8746,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1454243274124857,
+      "grad_norm": 0.3909272253513336,
+      "learning_rate": 9.487201554963911e-05,
+      "loss": 1.9115,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14552820193206606,
+      "grad_norm": 0.33725446462631226,
+      "learning_rate": 9.486481533499642e-05,
+      "loss": 1.6462,
+      "step": 1401
+    },
+    {
+      "epoch": 0.14563207645164641,
+      "grad_norm": 0.3404362201690674,
+      "learning_rate": 9.485761034260087e-05,
+      "loss": 1.5963,
+      "step": 1402
+    },
+    {
+      "epoch": 0.14573595097122677,
+      "grad_norm": 0.3509582281112671,
+      "learning_rate": 9.485040057321972e-05,
+      "loss": 1.7931,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1458398254908071,
+      "grad_norm": 0.3783257007598877,
+      "learning_rate": 9.484318602762077e-05,
+      "loss": 2.0324,
+      "step": 1404
+    },
+    {
+      "epoch": 0.14594370001038745,
+      "grad_norm": 0.36144283413887024,
+      "learning_rate": 9.483596670657228e-05,
+      "loss": 1.5949,
+      "step": 1405
+    },
+    {
+      "epoch": 0.1460475745299678,
+      "grad_norm": 0.33135420083999634,
+      "learning_rate": 9.482874261084307e-05,
+      "loss": 1.6486,
+      "step": 1406
+    },
+    {
+      "epoch": 0.14615144904954813,
+      "grad_norm": 0.35105010867118835,
+      "learning_rate": 9.482151374120244e-05,
+      "loss": 1.7674,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1462553235691285,
+      "grad_norm": 0.3350829780101776,
+      "learning_rate": 9.481428009842024e-05,
+      "loss": 1.6668,
+      "step": 1408
+    },
+    {
+      "epoch": 0.14635919808870884,
+      "grad_norm": 0.3917206823825836,
+      "learning_rate": 9.480704168326676e-05,
+      "loss": 1.7319,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1464630726082892,
+      "grad_norm": 0.35552677512168884,
+      "learning_rate": 9.479979849651286e-05,
+      "loss": 1.7501,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14656694712786952,
+      "grad_norm": 0.3432137072086334,
+      "learning_rate": 9.479255053892986e-05,
+      "loss": 1.7739,
+      "step": 1411
+    },
+    {
+      "epoch": 0.14667082164744988,
+      "grad_norm": 0.37724825739860535,
+      "learning_rate": 9.478529781128962e-05,
+      "loss": 1.8712,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14677469616703023,
+      "grad_norm": 0.37483900785446167,
+      "learning_rate": 9.477804031436451e-05,
+      "loss": 1.7262,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1468785706866106,
+      "grad_norm": 0.35302111506462097,
+      "learning_rate": 9.47707780489274e-05,
+      "loss": 1.68,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14698244520619091,
+      "grad_norm": 0.3883351683616638,
+      "learning_rate": 9.476351101575163e-05,
+      "loss": 1.7192,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14708631972577127,
+      "grad_norm": 0.3823285400867462,
+      "learning_rate": 9.475623921561112e-05,
+      "loss": 1.8355,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14719019424535162,
+      "grad_norm": 0.3662102520465851,
+      "learning_rate": 9.474896264928024e-05,
+      "loss": 1.5081,
+      "step": 1417
+    },
+    {
+      "epoch": 0.14729406876493195,
+      "grad_norm": 0.35046282410621643,
+      "learning_rate": 9.47416813175339e-05,
+      "loss": 1.7609,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1473979432845123,
+      "grad_norm": 0.37491992115974426,
+      "learning_rate": 9.47343952211475e-05,
+      "loss": 1.9254,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14750181780409266,
+      "grad_norm": 0.38457345962524414,
+      "learning_rate": 9.472710436089693e-05,
+      "loss": 1.8976,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14760569232367302,
+      "grad_norm": 0.39856263995170593,
+      "learning_rate": 9.471980873755865e-05,
+      "loss": 1.989,
+      "step": 1421
+    },
+    {
+      "epoch": 0.14770956684325334,
+      "grad_norm": 0.3745686113834381,
+      "learning_rate": 9.471250835190957e-05,
+      "loss": 1.7134,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1478134413628337,
+      "grad_norm": 0.360574871301651,
+      "learning_rate": 9.470520320472712e-05,
+      "loss": 1.722,
+      "step": 1423
+    },
+    {
+      "epoch": 0.14791731588241405,
+      "grad_norm": 0.35429859161376953,
+      "learning_rate": 9.469789329678922e-05,
+      "loss": 1.5907,
+      "step": 1424
+    },
+    {
+      "epoch": 0.14802119040199438,
+      "grad_norm": 0.3642517030239105,
+      "learning_rate": 9.469057862887436e-05,
+      "loss": 1.7298,
+      "step": 1425
+    },
+    {
+      "epoch": 0.14812506492157473,
+      "grad_norm": 0.4171449840068817,
+      "learning_rate": 9.468325920176148e-05,
+      "loss": 1.9391,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1482289394411551,
+      "grad_norm": 0.3378536105155945,
+      "learning_rate": 9.467593501623003e-05,
+      "loss": 1.7072,
+      "step": 1427
+    },
+    {
+      "epoch": 0.14833281396073544,
+      "grad_norm": 0.37104731798171997,
+      "learning_rate": 9.466860607305998e-05,
+      "loss": 1.7929,
+      "step": 1428
+    },
+    {
+      "epoch": 0.14843668848031577,
+      "grad_norm": 0.38163742423057556,
+      "learning_rate": 9.466127237303182e-05,
+      "loss": 1.8288,
+      "step": 1429
+    },
+    {
+      "epoch": 0.14854056299989613,
+      "grad_norm": 0.3787825107574463,
+      "learning_rate": 9.465393391692653e-05,
+      "loss": 1.7313,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14864443751947648,
+      "grad_norm": 0.3504464626312256,
+      "learning_rate": 9.464659070552558e-05,
+      "loss": 1.7905,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1487483120390568,
+      "grad_norm": 0.3874281346797943,
+      "learning_rate": 9.463924273961099e-05,
+      "loss": 1.8856,
+      "step": 1432
+    },
+    {
+      "epoch": 0.14885218655863716,
+      "grad_norm": 0.35603591799736023,
+      "learning_rate": 9.463189001996525e-05,
+      "loss": 1.7671,
+      "step": 1433
+    },
+    {
+      "epoch": 0.14895606107821752,
+      "grad_norm": 0.37857985496520996,
+      "learning_rate": 9.462453254737133e-05,
+      "loss": 1.7791,
+      "step": 1434
+    },
+    {
+      "epoch": 0.14905993559779787,
+      "grad_norm": 0.3580288290977478,
+      "learning_rate": 9.461717032261282e-05,
+      "loss": 1.6404,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1491638101173782,
+      "grad_norm": 0.38064590096473694,
+      "learning_rate": 9.460980334647367e-05,
+      "loss": 1.8078,
+      "step": 1436
+    },
+    {
+      "epoch": 0.14926768463695855,
+      "grad_norm": 0.3878273069858551,
+      "learning_rate": 9.460243161973846e-05,
+      "loss": 1.6382,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1493715591565389,
+      "grad_norm": 0.3887486457824707,
+      "learning_rate": 9.459505514319216e-05,
+      "loss": 1.8884,
+      "step": 1438
+    },
+    {
+      "epoch": 0.14947543367611923,
+      "grad_norm": 0.42063435912132263,
+      "learning_rate": 9.458767391762038e-05,
+      "loss": 1.9866,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1495793081956996,
+      "grad_norm": 0.39019012451171875,
+      "learning_rate": 9.45802879438091e-05,
+      "loss": 1.8562,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14968318271527994,
+      "grad_norm": 0.36793825030326843,
+      "learning_rate": 9.457289722254489e-05,
+      "loss": 1.8974,
+      "step": 1441
+    },
+    {
+      "epoch": 0.1497870572348603,
+      "grad_norm": 0.37522581219673157,
+      "learning_rate": 9.456550175461483e-05,
+      "loss": 1.878,
+      "step": 1442
+    },
+    {
+      "epoch": 0.14989093175444063,
+      "grad_norm": 0.3458879590034485,
+      "learning_rate": 9.455810154080644e-05,
+      "loss": 1.7597,
+      "step": 1443
+    },
+    {
+      "epoch": 0.14999480627402098,
+      "grad_norm": 0.3603193759918213,
+      "learning_rate": 9.45506965819078e-05,
+      "loss": 1.7119,
+      "step": 1444
+    },
+    {
+      "epoch": 0.15009868079360134,
+      "grad_norm": 0.35782983899116516,
+      "learning_rate": 9.45432868787075e-05,
+      "loss": 1.7925,
+      "step": 1445
+    },
+    {
+      "epoch": 0.1502025553131817,
+      "grad_norm": 0.3702535331249237,
+      "learning_rate": 9.453587243199458e-05,
+      "loss": 1.8274,
+      "step": 1446
+    },
+    {
+      "epoch": 0.15030642983276202,
+      "grad_norm": 0.37376201152801514,
+      "learning_rate": 9.452845324255865e-05,
+      "loss": 1.6782,
+      "step": 1447
+    },
+    {
+      "epoch": 0.15041030435234237,
+      "grad_norm": 0.3699168264865875,
+      "learning_rate": 9.452102931118979e-05,
+      "loss": 1.8381,
+      "step": 1448
+    },
+    {
+      "epoch": 0.15051417887192273,
+      "grad_norm": 0.36657705903053284,
+      "learning_rate": 9.451360063867857e-05,
+      "loss": 1.8949,
+      "step": 1449
+    },
+    {
+      "epoch": 0.15061805339150305,
+      "grad_norm": 0.35056763887405396,
+      "learning_rate": 9.450616722581611e-05,
+      "loss": 1.7087,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1507219279110834,
+      "grad_norm": 0.33786895871162415,
+      "learning_rate": 9.449872907339401e-05,
+      "loss": 1.7222,
+      "step": 1451
+    },
+    {
+      "epoch": 0.15082580243066376,
+      "grad_norm": 0.3312149941921234,
+      "learning_rate": 9.449128618220436e-05,
+      "loss": 1.7871,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15092967695024412,
+      "grad_norm": 0.3575022518634796,
+      "learning_rate": 9.448383855303977e-05,
+      "loss": 1.7017,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15103355146982445,
+      "grad_norm": 0.35011738538742065,
+      "learning_rate": 9.447638618669336e-05,
+      "loss": 1.9229,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1511374259894048,
+      "grad_norm": 0.37051519751548767,
+      "learning_rate": 9.446892908395877e-05,
+      "loss": 1.8912,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15124130050898515,
+      "grad_norm": 0.3656577169895172,
+      "learning_rate": 9.446146724563009e-05,
+      "loss": 1.8665,
+      "step": 1456
+    },
+    {
+      "epoch": 0.15134517502856548,
+      "grad_norm": 0.34047558903694153,
+      "learning_rate": 9.445400067250196e-05,
+      "loss": 1.6188,
+      "step": 1457
+    },
+    {
+      "epoch": 0.15144904954814584,
+      "grad_norm": 0.3526557385921478,
+      "learning_rate": 9.44465293653695e-05,
+      "loss": 1.7484,
+      "step": 1458
+    },
+    {
+      "epoch": 0.1515529240677262,
+      "grad_norm": 0.3618306815624237,
+      "learning_rate": 9.443905332502837e-05,
+      "loss": 1.6957,
+      "step": 1459
+    },
+    {
+      "epoch": 0.15165679858730655,
+      "grad_norm": 0.39453551173210144,
+      "learning_rate": 9.443157255227472e-05,
+      "loss": 1.866,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15176067310688687,
+      "grad_norm": 0.3841235637664795,
+      "learning_rate": 9.442408704790516e-05,
+      "loss": 1.8885,
+      "step": 1461
+    },
+    {
+      "epoch": 0.15186454762646723,
+      "grad_norm": 0.3559393286705017,
+      "learning_rate": 9.441659681271684e-05,
+      "loss": 1.5994,
+      "step": 1462
+    },
+    {
+      "epoch": 0.15196842214604758,
+      "grad_norm": 0.3648277521133423,
+      "learning_rate": 9.440910184750743e-05,
+      "loss": 1.741,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1520722966656279,
+      "grad_norm": 0.35485556721687317,
+      "learning_rate": 9.440160215307508e-05,
+      "loss": 1.7756,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15217617118520826,
+      "grad_norm": 0.35386818647384644,
+      "learning_rate": 9.439409773021844e-05,
+      "loss": 1.877,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15228004570478862,
+      "grad_norm": 0.3678518235683441,
+      "learning_rate": 9.438658857973668e-05,
+      "loss": 1.7585,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15238392022436897,
+      "grad_norm": 0.36395224928855896,
+      "learning_rate": 9.437907470242946e-05,
+      "loss": 1.7805,
+      "step": 1467
+    },
+    {
+      "epoch": 0.1524877947439493,
+      "grad_norm": 0.3579014539718628,
+      "learning_rate": 9.437155609909696e-05,
+      "loss": 1.7271,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15259166926352966,
+      "grad_norm": 0.3654552102088928,
+      "learning_rate": 9.436403277053984e-05,
+      "loss": 1.858,
+      "step": 1469
+    },
+    {
+      "epoch": 0.15269554378311,
+      "grad_norm": 0.3869788646697998,
+      "learning_rate": 9.435650471755931e-05,
+      "loss": 1.8507,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15279941830269034,
+      "grad_norm": 0.3339064121246338,
+      "learning_rate": 9.4348971940957e-05,
+      "loss": 1.7086,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1529032928222707,
+      "grad_norm": 0.3778363764286041,
+      "learning_rate": 9.434143444153511e-05,
+      "loss": 1.936,
+      "step": 1472
+    },
+    {
+      "epoch": 0.15300716734185105,
+      "grad_norm": 0.4051001965999603,
+      "learning_rate": 9.433389222009633e-05,
+      "loss": 1.7736,
+      "step": 1473
+    },
+    {
+      "epoch": 0.1531110418614314,
+      "grad_norm": 0.35207393765449524,
+      "learning_rate": 9.432634527744386e-05,
+      "loss": 1.7176,
+      "step": 1474
+    },
+    {
+      "epoch": 0.15321491638101173,
+      "grad_norm": 0.3851469159126282,
+      "learning_rate": 9.431879361438137e-05,
+      "loss": 1.9448,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15331879090059208,
+      "grad_norm": 0.35818904638290405,
+      "learning_rate": 9.431123723171305e-05,
+      "loss": 1.7773,
+      "step": 1476
+    },
+    {
+      "epoch": 0.15342266542017244,
+      "grad_norm": 0.37601613998413086,
+      "learning_rate": 9.430367613024361e-05,
+      "loss": 1.9316,
+      "step": 1477
+    },
+    {
+      "epoch": 0.15352653993975277,
+      "grad_norm": 0.3351070284843445,
+      "learning_rate": 9.429611031077825e-05,
+      "loss": 1.5443,
+      "step": 1478
+    },
+    {
+      "epoch": 0.15363041445933312,
+      "grad_norm": 0.4057861566543579,
+      "learning_rate": 9.428853977412266e-05,
+      "loss": 1.8619,
+      "step": 1479
+    },
+    {
+      "epoch": 0.15373428897891347,
+      "grad_norm": 0.3662700057029724,
+      "learning_rate": 9.428096452108305e-05,
+      "loss": 1.7752,
+      "step": 1480
+    },
+    {
+      "epoch": 0.15383816349849383,
+      "grad_norm": 0.38205060362815857,
+      "learning_rate": 9.427338455246612e-05,
+      "loss": 1.7397,
+      "step": 1481
+    },
+    {
+      "epoch": 0.15394203801807416,
+      "grad_norm": 0.3742862641811371,
+      "learning_rate": 9.426579986907909e-05,
+      "loss": 1.7385,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1540459125376545,
+      "grad_norm": 0.37825533747673035,
+      "learning_rate": 9.425821047172965e-05,
+      "loss": 1.8288,
+      "step": 1483
+    },
+    {
+      "epoch": 0.15414978705723487,
+      "grad_norm": 0.42222508788108826,
+      "learning_rate": 9.425061636122603e-05,
+      "loss": 1.9567,
+      "step": 1484
+    },
+    {
+      "epoch": 0.15425366157681522,
+      "grad_norm": 0.3459799885749817,
+      "learning_rate": 9.424301753837692e-05,
+      "loss": 1.7383,
+      "step": 1485
+    },
+    {
+      "epoch": 0.15435753609639555,
+      "grad_norm": 0.38764312863349915,
+      "learning_rate": 9.423541400399157e-05,
+      "loss": 1.8508,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1544614106159759,
+      "grad_norm": 0.3998440206050873,
+      "learning_rate": 9.422780575887967e-05,
+      "loss": 1.9329,
+      "step": 1487
+    },
+    {
+      "epoch": 0.15456528513555626,
+      "grad_norm": 0.365713894367218,
+      "learning_rate": 9.422019280385145e-05,
+      "loss": 1.9418,
+      "step": 1488
+    },
+    {
+      "epoch": 0.15466915965513658,
+      "grad_norm": 0.3910145163536072,
+      "learning_rate": 9.421257513971762e-05,
+      "loss": 1.8998,
+      "step": 1489
+    },
+    {
+      "epoch": 0.15477303417471694,
+      "grad_norm": 0.3563111424446106,
+      "learning_rate": 9.420495276728943e-05,
+      "loss": 1.8042,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1548769086942973,
+      "grad_norm": 0.3810329735279083,
+      "learning_rate": 9.419732568737856e-05,
+      "loss": 1.9044,
+      "step": 1491
+    },
+    {
+      "epoch": 0.15498078321387765,
+      "grad_norm": 0.4031761884689331,
+      "learning_rate": 9.418969390079726e-05,
+      "loss": 2.098,
+      "step": 1492
+    },
+    {
+      "epoch": 0.15508465773345798,
+      "grad_norm": 0.3679083287715912,
+      "learning_rate": 9.418205740835825e-05,
+      "loss": 1.7396,
+      "step": 1493
+    },
+    {
+      "epoch": 0.15518853225303833,
+      "grad_norm": 0.35544419288635254,
+      "learning_rate": 9.417441621087476e-05,
+      "loss": 1.7114,
+      "step": 1494
+    },
+    {
+      "epoch": 0.15529240677261869,
+      "grad_norm": 0.36036983132362366,
+      "learning_rate": 9.416677030916053e-05,
+      "loss": 1.828,
+      "step": 1495
+    },
+    {
+      "epoch": 0.155396281292199,
+      "grad_norm": 0.34864529967308044,
+      "learning_rate": 9.415911970402976e-05,
+      "loss": 1.7821,
+      "step": 1496
+    },
+    {
+      "epoch": 0.15550015581177937,
+      "grad_norm": 0.4014929533004761,
+      "learning_rate": 9.41514643962972e-05,
+      "loss": 1.8746,
+      "step": 1497
+    },
+    {
+      "epoch": 0.15560403033135972,
+      "grad_norm": 0.39249807596206665,
+      "learning_rate": 9.414380438677808e-05,
+      "loss": 1.9818,
+      "step": 1498
+    },
+    {
+      "epoch": 0.15570790485094008,
+      "grad_norm": 0.33457159996032715,
+      "learning_rate": 9.413613967628814e-05,
+      "loss": 1.7178,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1558117793705204,
+      "grad_norm": 0.326725572347641,
+      "learning_rate": 9.412847026564357e-05,
+      "loss": 1.5971,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15591565389010076,
+      "grad_norm": 0.36268436908721924,
+      "learning_rate": 9.412079615566116e-05,
+      "loss": 1.8022,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1560195284096811,
+      "grad_norm": 0.3470126986503601,
+      "learning_rate": 9.411311734715811e-05,
+      "loss": 1.8473,
+      "step": 1502
+    },
+    {
+      "epoch": 0.15612340292926144,
+      "grad_norm": 0.3726639747619629,
+      "learning_rate": 9.410543384095214e-05,
+      "loss": 1.7718,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1562272774488418,
+      "grad_norm": 0.37733760476112366,
+      "learning_rate": 9.409774563786152e-05,
+      "loss": 1.8614,
+      "step": 1504
+    },
+    {
+      "epoch": 0.15633115196842215,
+      "grad_norm": 0.4124131202697754,
+      "learning_rate": 9.409005273870495e-05,
+      "loss": 2.1187,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1564350264880025,
+      "grad_norm": 0.3965660631656647,
+      "learning_rate": 9.40823551443017e-05,
+      "loss": 1.7244,
+      "step": 1506
+    },
+    {
+      "epoch": 0.15653890100758283,
+      "grad_norm": 0.390770822763443,
+      "learning_rate": 9.407465285547147e-05,
+      "loss": 1.8046,
+      "step": 1507
+    },
+    {
+      "epoch": 0.15664277552716319,
+      "grad_norm": 0.369103342294693,
+      "learning_rate": 9.406694587303451e-05,
+      "loss": 1.7574,
+      "step": 1508
+    },
+    {
+      "epoch": 0.15674665004674354,
+      "grad_norm": 0.38057541847229004,
+      "learning_rate": 9.405923419781154e-05,
+      "loss": 1.8895,
+      "step": 1509
+    },
+    {
+      "epoch": 0.15685052456632387,
+      "grad_norm": 0.3767298460006714,
+      "learning_rate": 9.405151783062381e-05,
+      "loss": 1.6415,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15695439908590422,
+      "grad_norm": 0.35092034935951233,
+      "learning_rate": 9.404379677229305e-05,
+      "loss": 1.7707,
+      "step": 1511
+    },
+    {
+      "epoch": 0.15705827360548458,
+      "grad_norm": 0.3567756712436676,
+      "learning_rate": 9.403607102364148e-05,
+      "loss": 1.7814,
+      "step": 1512
+    },
+    {
+      "epoch": 0.15716214812506493,
+      "grad_norm": 0.38232964277267456,
+      "learning_rate": 9.402834058549184e-05,
+      "loss": 2.0017,
+      "step": 1513
+    },
+    {
+      "epoch": 0.15726602264464526,
+      "grad_norm": 0.35852310061454773,
+      "learning_rate": 9.402060545866737e-05,
+      "loss": 1.8552,
+      "step": 1514
+    },
+    {
+      "epoch": 0.15736989716422561,
+      "grad_norm": 0.3555811047554016,
+      "learning_rate": 9.401286564399178e-05,
+      "loss": 1.8447,
+      "step": 1515
+    },
+    {
+      "epoch": 0.15747377168380597,
+      "grad_norm": 0.33671462535858154,
+      "learning_rate": 9.400512114228935e-05,
+      "loss": 1.6783,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1575776462033863,
+      "grad_norm": 0.3423287570476532,
+      "learning_rate": 9.399737195438473e-05,
+      "loss": 1.7901,
+      "step": 1517
+    },
+    {
+      "epoch": 0.15768152072296665,
+      "grad_norm": 0.3938925564289093,
+      "learning_rate": 9.398961808110321e-05,
+      "loss": 1.6647,
+      "step": 1518
+    },
+    {
+      "epoch": 0.157785395242547,
+      "grad_norm": 0.366240918636322,
+      "learning_rate": 9.398185952327048e-05,
+      "loss": 1.7797,
+      "step": 1519
+    },
+    {
+      "epoch": 0.15788926976212736,
+      "grad_norm": 0.3558516204357147,
+      "learning_rate": 9.397409628171279e-05,
+      "loss": 1.6302,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1579931442817077,
+      "grad_norm": 0.3891289234161377,
+      "learning_rate": 9.396632835725687e-05,
+      "loss": 1.9296,
+      "step": 1521
+    },
+    {
+      "epoch": 0.15809701880128804,
+      "grad_norm": 0.4402526319026947,
+      "learning_rate": 9.395855575072991e-05,
+      "loss": 1.9748,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1582008933208684,
+      "grad_norm": 0.3928702473640442,
+      "learning_rate": 9.395077846295967e-05,
+      "loss": 1.9662,
+      "step": 1523
+    },
+    {
+      "epoch": 0.15830476784044875,
+      "grad_norm": 0.36685439944267273,
+      "learning_rate": 9.394299649477434e-05,
+      "loss": 1.7051,
+      "step": 1524
+    },
+    {
+      "epoch": 0.15840864236002908,
+      "grad_norm": 0.36427903175354004,
+      "learning_rate": 9.393520984700266e-05,
+      "loss": 1.8365,
+      "step": 1525
+    },
+    {
+      "epoch": 0.15851251687960943,
+      "grad_norm": 0.3728634715080261,
+      "learning_rate": 9.392741852047385e-05,
+      "loss": 1.8812,
+      "step": 1526
+    },
+    {
+      "epoch": 0.1586163913991898,
+      "grad_norm": 0.3916126489639282,
+      "learning_rate": 9.391962251601758e-05,
+      "loss": 1.8,
+      "step": 1527
+    },
+    {
+      "epoch": 0.15872026591877011,
+      "grad_norm": 0.3563532531261444,
+      "learning_rate": 9.391182183446413e-05,
+      "loss": 1.5527,
+      "step": 1528
+    },
+    {
+      "epoch": 0.15882414043835047,
+      "grad_norm": 0.3756893277168274,
+      "learning_rate": 9.390401647664417e-05,
+      "loss": 1.9147,
+      "step": 1529
+    },
+    {
+      "epoch": 0.15892801495793082,
+      "grad_norm": 0.35163572430610657,
+      "learning_rate": 9.389620644338892e-05,
+      "loss": 1.7478,
+      "step": 1530
+    },
+    {
+      "epoch": 0.15903188947751118,
+      "grad_norm": 0.3400569260120392,
+      "learning_rate": 9.388839173553007e-05,
+      "loss": 1.5696,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1591357639970915,
+      "grad_norm": 0.3381504416465759,
+      "learning_rate": 9.388057235389986e-05,
+      "loss": 1.6998,
+      "step": 1532
+    },
+    {
+      "epoch": 0.15923963851667186,
+      "grad_norm": 0.35516080260276794,
+      "learning_rate": 9.3872748299331e-05,
+      "loss": 1.8103,
+      "step": 1533
+    },
+    {
+      "epoch": 0.15934351303625222,
+      "grad_norm": 0.3808714747428894,
+      "learning_rate": 9.386491957265662e-05,
+      "loss": 1.7707,
+      "step": 1534
+    },
+    {
+      "epoch": 0.15944738755583254,
+      "grad_norm": 0.3629775941371918,
+      "learning_rate": 9.38570861747105e-05,
+      "loss": 1.8373,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1595512620754129,
+      "grad_norm": 0.340930700302124,
+      "learning_rate": 9.384924810632679e-05,
+      "loss": 1.7374,
+      "step": 1536
+    },
+    {
+      "epoch": 0.15965513659499325,
+      "grad_norm": 0.40548157691955566,
+      "learning_rate": 9.384140536834018e-05,
+      "loss": 2.0285,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1597590111145736,
+      "grad_norm": 0.39659062027931213,
+      "learning_rate": 9.383355796158588e-05,
+      "loss": 1.8455,
+      "step": 1538
+    },
+    {
+      "epoch": 0.15986288563415393,
+      "grad_norm": 0.35021933913230896,
+      "learning_rate": 9.382570588689958e-05,
+      "loss": 1.8248,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1599667601537343,
+      "grad_norm": 0.39401671290397644,
+      "learning_rate": 9.381784914511746e-05,
+      "loss": 2.1518,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16007063467331464,
+      "grad_norm": 0.3451925814151764,
+      "learning_rate": 9.38099877370762e-05,
+      "loss": 1.7234,
+      "step": 1541
+    },
+    {
+      "epoch": 0.16017450919289497,
+      "grad_norm": 0.3629453778266907,
+      "learning_rate": 9.380212166361298e-05,
+      "loss": 1.7744,
+      "step": 1542
+    },
+    {
+      "epoch": 0.16027838371247533,
+      "grad_norm": 0.4119042456150055,
+      "learning_rate": 9.379425092556545e-05,
+      "loss": 2.0752,
+      "step": 1543
+    },
+    {
+      "epoch": 0.16038225823205568,
+      "grad_norm": 0.347231388092041,
+      "learning_rate": 9.378637552377181e-05,
+      "loss": 1.7362,
+      "step": 1544
+    },
+    {
+      "epoch": 0.16048613275163603,
+      "grad_norm": 0.4104901850223541,
+      "learning_rate": 9.377849545907074e-05,
+      "loss": 1.9631,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16059000727121636,
+      "grad_norm": 0.396040141582489,
+      "learning_rate": 9.377061073230138e-05,
+      "loss": 1.955,
+      "step": 1546
+    },
+    {
+      "epoch": 0.16069388179079672,
+      "grad_norm": 0.3335644602775574,
+      "learning_rate": 9.37627213443034e-05,
+      "loss": 1.5817,
+      "step": 1547
+    },
+    {
+      "epoch": 0.16079775631037707,
+      "grad_norm": 0.3619517385959625,
+      "learning_rate": 9.375482729591696e-05,
+      "loss": 1.7563,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1609016308299574,
+      "grad_norm": 0.35229092836380005,
+      "learning_rate": 9.37469285879827e-05,
+      "loss": 1.7638,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16100550534953775,
+      "grad_norm": 0.36668309569358826,
+      "learning_rate": 9.373902522134181e-05,
+      "loss": 1.6821,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1611093798691181,
+      "grad_norm": 0.38943883776664734,
+      "learning_rate": 9.37311171968359e-05,
+      "loss": 1.6423,
+      "step": 1551
+    },
+    {
+      "epoch": 0.16121325438869846,
+      "grad_norm": 0.35452452301979065,
+      "learning_rate": 9.372320451530711e-05,
+      "loss": 1.7089,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1613171289082788,
+      "grad_norm": 0.39926689863204956,
+      "learning_rate": 9.371528717759811e-05,
+      "loss": 1.801,
+      "step": 1553
+    },
+    {
+      "epoch": 0.16142100342785914,
+      "grad_norm": 0.37590524554252625,
+      "learning_rate": 9.370736518455203e-05,
+      "loss": 1.7258,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1615248779474395,
+      "grad_norm": 0.35222113132476807,
+      "learning_rate": 9.369943853701246e-05,
+      "loss": 1.8522,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16162875246701985,
+      "grad_norm": 0.35563012957572937,
+      "learning_rate": 9.369150723582358e-05,
+      "loss": 1.6075,
+      "step": 1556
+    },
+    {
+      "epoch": 0.16173262698660018,
+      "grad_norm": 0.39180493354797363,
+      "learning_rate": 9.368357128182998e-05,
+      "loss": 1.7797,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16183650150618054,
+      "grad_norm": 0.3635503053665161,
+      "learning_rate": 9.367563067587677e-05,
+      "loss": 1.8579,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1619403760257609,
+      "grad_norm": 0.3636340796947479,
+      "learning_rate": 9.366768541880958e-05,
+      "loss": 1.708,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16204425054534122,
+      "grad_norm": 0.39017453789711,
+      "learning_rate": 9.365973551147453e-05,
+      "loss": 1.8155,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16214812506492157,
+      "grad_norm": 0.3406469225883484,
+      "learning_rate": 9.365178095471819e-05,
+      "loss": 1.7573,
+      "step": 1561
+    },
+    {
+      "epoch": 0.16225199958450193,
+      "grad_norm": 0.36579304933547974,
+      "learning_rate": 9.364382174938768e-05,
+      "loss": 1.898,
+      "step": 1562
+    },
+    {
+      "epoch": 0.16235587410408228,
+      "grad_norm": 0.3369811773300171,
+      "learning_rate": 9.363585789633057e-05,
+      "loss": 1.6292,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1624597486236626,
+      "grad_norm": 0.3355672061443329,
+      "learning_rate": 9.3627889396395e-05,
+      "loss": 1.6484,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16256362314324296,
+      "grad_norm": 0.37393924593925476,
+      "learning_rate": 9.36199162504295e-05,
+      "loss": 1.7915,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16266749766282332,
+      "grad_norm": 0.37588125467300415,
+      "learning_rate": 9.361193845928318e-05,
+      "loss": 1.6923,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16277137218240365,
+      "grad_norm": 0.4559624493122101,
+      "learning_rate": 9.360395602380559e-05,
+      "loss": 1.8842,
+      "step": 1567
+    },
+    {
+      "epoch": 0.162875246701984,
+      "grad_norm": 0.39234891533851624,
+      "learning_rate": 9.359596894484682e-05,
+      "loss": 1.9783,
+      "step": 1568
+    },
+    {
+      "epoch": 0.16297912122156435,
+      "grad_norm": 0.36912286281585693,
+      "learning_rate": 9.35879772232574e-05,
+      "loss": 1.8153,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1630829957411447,
+      "grad_norm": 0.3693285584449768,
+      "learning_rate": 9.357998085988842e-05,
+      "loss": 1.5263,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16318687026072504,
+      "grad_norm": 0.38860028982162476,
+      "learning_rate": 9.35719798555914e-05,
+      "loss": 1.6806,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1632907447803054,
+      "grad_norm": 0.401967853307724,
+      "learning_rate": 9.356397421121842e-05,
+      "loss": 1.855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.16339461929988575,
+      "grad_norm": 0.35193613171577454,
+      "learning_rate": 9.355596392762197e-05,
+      "loss": 1.8909,
+      "step": 1573
+    },
+    {
+      "epoch": 0.16349849381946607,
+      "grad_norm": 0.38218873739242554,
+      "learning_rate": 9.354794900565515e-05,
+      "loss": 1.874,
+      "step": 1574
+    },
+    {
+      "epoch": 0.16360236833904643,
+      "grad_norm": 0.3593958020210266,
+      "learning_rate": 9.353992944617142e-05,
+      "loss": 1.6768,
+      "step": 1575
+    },
+    {
+      "epoch": 0.16370624285862678,
+      "grad_norm": 0.3737293481826782,
+      "learning_rate": 9.353190525002484e-05,
+      "loss": 1.6279,
+      "step": 1576
+    },
+    {
+      "epoch": 0.16381011737820714,
+      "grad_norm": 0.38832589983940125,
+      "learning_rate": 9.35238764180699e-05,
+      "loss": 1.8541,
+      "step": 1577
+    },
+    {
+      "epoch": 0.16391399189778746,
+      "grad_norm": 0.37959024310112,
+      "learning_rate": 9.351584295116165e-05,
+      "loss": 1.5282,
+      "step": 1578
+    },
+    {
+      "epoch": 0.16401786641736782,
+      "grad_norm": 0.37193652987480164,
+      "learning_rate": 9.350780485015553e-05,
+      "loss": 1.6896,
+      "step": 1579
+    },
+    {
+      "epoch": 0.16412174093694817,
+      "grad_norm": 0.35193783044815063,
+      "learning_rate": 9.349976211590758e-05,
+      "loss": 1.57,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1642256154565285,
+      "grad_norm": 0.3922156095504761,
+      "learning_rate": 9.349171474927427e-05,
+      "loss": 1.5819,
+      "step": 1581
+    },
+    {
+      "epoch": 0.16432948997610886,
+      "grad_norm": 0.36705270409584045,
+      "learning_rate": 9.34836627511126e-05,
+      "loss": 1.6357,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1644333644956892,
+      "grad_norm": 0.36877331137657166,
+      "learning_rate": 9.347560612228003e-05,
+      "loss": 1.7312,
+      "step": 1583
+    },
+    {
+      "epoch": 0.16453723901526957,
+      "grad_norm": 0.3872382342815399,
+      "learning_rate": 9.346754486363452e-05,
+      "loss": 1.8537,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1646411135348499,
+      "grad_norm": 0.3447991609573364,
+      "learning_rate": 9.345947897603455e-05,
+      "loss": 1.6428,
+      "step": 1585
+    },
+    {
+      "epoch": 0.16474498805443025,
+      "grad_norm": 0.3650977909564972,
+      "learning_rate": 9.345140846033906e-05,
+      "loss": 1.8029,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1648488625740106,
+      "grad_norm": 0.35949772596359253,
+      "learning_rate": 9.344333331740751e-05,
+      "loss": 1.6589,
+      "step": 1587
+    },
+    {
+      "epoch": 0.16495273709359093,
+      "grad_norm": 0.38417744636535645,
+      "learning_rate": 9.343525354809984e-05,
+      "loss": 1.7891,
+      "step": 1588
+    },
+    {
+      "epoch": 0.16505661161317128,
+      "grad_norm": 0.36684268712997437,
+      "learning_rate": 9.342716915327647e-05,
+      "loss": 1.7708,
+      "step": 1589
+    },
+    {
+      "epoch": 0.16516048613275164,
+      "grad_norm": 0.3678865134716034,
+      "learning_rate": 9.341908013379831e-05,
+      "loss": 1.7155,
+      "step": 1590
+    },
+    {
+      "epoch": 0.165264360652332,
+      "grad_norm": 0.38696667551994324,
+      "learning_rate": 9.341098649052681e-05,
+      "loss": 1.8784,
+      "step": 1591
+    },
+    {
+      "epoch": 0.16536823517191232,
+      "grad_norm": 0.3689199388027191,
+      "learning_rate": 9.340288822432388e-05,
+      "loss": 1.874,
+      "step": 1592
+    },
+    {
+      "epoch": 0.16547210969149267,
+      "grad_norm": 0.34134095907211304,
+      "learning_rate": 9.339478533605191e-05,
+      "loss": 1.5365,
+      "step": 1593
+    },
+    {
+      "epoch": 0.16557598421107303,
+      "grad_norm": 0.3707195818424225,
+      "learning_rate": 9.338667782657378e-05,
+      "loss": 1.7984,
+      "step": 1594
+    },
+    {
+      "epoch": 0.16567985873065338,
+      "grad_norm": 0.3389596939086914,
+      "learning_rate": 9.337856569675288e-05,
+      "loss": 1.6244,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1657837332502337,
+      "grad_norm": 0.3761754035949707,
+      "learning_rate": 9.337044894745313e-05,
+      "loss": 1.8774,
+      "step": 1596
+    },
+    {
+      "epoch": 0.16588760776981407,
+      "grad_norm": 0.37581124901771545,
+      "learning_rate": 9.336232757953884e-05,
+      "loss": 1.7503,
+      "step": 1597
+    },
+    {
+      "epoch": 0.16599148228939442,
+      "grad_norm": 0.37551531195640564,
+      "learning_rate": 9.335420159387491e-05,
+      "loss": 1.8433,
+      "step": 1598
+    },
+    {
+      "epoch": 0.16609535680897475,
+      "grad_norm": 0.3547391891479492,
+      "learning_rate": 9.33460709913267e-05,
+      "loss": 1.8019,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1661992313285551,
+      "grad_norm": 0.40177521109580994,
+      "learning_rate": 9.333793577276004e-05,
+      "loss": 1.9248,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16630310584813546,
+      "grad_norm": 0.345114141702652,
+      "learning_rate": 9.332979593904125e-05,
+      "loss": 1.674,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1664069803677158,
+      "grad_norm": 0.36117351055145264,
+      "learning_rate": 9.33216514910372e-05,
+      "loss": 1.9506,
+      "step": 1602
+    },
+    {
+      "epoch": 0.16651085488729614,
+      "grad_norm": 0.3475273847579956,
+      "learning_rate": 9.331350242961518e-05,
+      "loss": 1.716,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1666147294068765,
+      "grad_norm": 0.3611377477645874,
+      "learning_rate": 9.3305348755643e-05,
+      "loss": 1.7274,
+      "step": 1604
+    },
+    {
+      "epoch": 0.16671860392645685,
+      "grad_norm": 0.40187951922416687,
+      "learning_rate": 9.329719046998898e-05,
+      "loss": 1.7231,
+      "step": 1605
+    },
+    {
+      "epoch": 0.16682247844603718,
+      "grad_norm": 0.3646019399166107,
+      "learning_rate": 9.328902757352191e-05,
+      "loss": 1.6772,
+      "step": 1606
+    },
+    {
+      "epoch": 0.16692635296561753,
+      "grad_norm": 0.34303680062294006,
+      "learning_rate": 9.328086006711108e-05,
+      "loss": 1.6853,
+      "step": 1607
+    },
+    {
+      "epoch": 0.16703022748519789,
+      "grad_norm": 0.36761873960494995,
+      "learning_rate": 9.327268795162622e-05,
+      "loss": 1.7584,
+      "step": 1608
+    },
+    {
+      "epoch": 0.16713410200477824,
+      "grad_norm": 0.36471322178840637,
+      "learning_rate": 9.326451122793766e-05,
+      "loss": 1.7157,
+      "step": 1609
+    },
+    {
+      "epoch": 0.16723797652435857,
+      "grad_norm": 0.385139524936676,
+      "learning_rate": 9.325632989691613e-05,
+      "loss": 1.7911,
+      "step": 1610
+    },
+    {
+      "epoch": 0.16734185104393892,
+      "grad_norm": 0.3613028824329376,
+      "learning_rate": 9.324814395943289e-05,
+      "loss": 1.7849,
+      "step": 1611
+    },
+    {
+      "epoch": 0.16744572556351928,
+      "grad_norm": 0.3828355669975281,
+      "learning_rate": 9.323995341635963e-05,
+      "loss": 1.9662,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1675496000830996,
+      "grad_norm": 0.3685465157032013,
+      "learning_rate": 9.323175826856865e-05,
+      "loss": 1.8091,
+      "step": 1613
+    },
+    {
+      "epoch": 0.16765347460267996,
+      "grad_norm": 0.36837613582611084,
+      "learning_rate": 9.322355851693263e-05,
+      "loss": 1.716,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1677573491222603,
+      "grad_norm": 0.4158804416656494,
+      "learning_rate": 9.321535416232476e-05,
+      "loss": 1.8553,
+      "step": 1615
+    },
+    {
+      "epoch": 0.16786122364184067,
+      "grad_norm": 0.35498541593551636,
+      "learning_rate": 9.32071452056188e-05,
+      "loss": 1.6873,
+      "step": 1616
+    },
+    {
+      "epoch": 0.167965098161421,
+      "grad_norm": 0.37819525599479675,
+      "learning_rate": 9.319893164768888e-05,
+      "loss": 1.7064,
+      "step": 1617
+    },
+    {
+      "epoch": 0.16806897268100135,
+      "grad_norm": 0.3379732668399811,
+      "learning_rate": 9.31907134894097e-05,
+      "loss": 1.6289,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1681728472005817,
+      "grad_norm": 0.38348469138145447,
+      "learning_rate": 9.318249073165646e-05,
+      "loss": 1.7028,
+      "step": 1619
+    },
+    {
+      "epoch": 0.16827672172016203,
+      "grad_norm": 0.3996196985244751,
+      "learning_rate": 9.317426337530477e-05,
+      "loss": 1.7427,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16838059623974239,
+      "grad_norm": 0.3799591064453125,
+      "learning_rate": 9.316603142123082e-05,
+      "loss": 1.6961,
+      "step": 1621
+    },
+    {
+      "epoch": 0.16848447075932274,
+      "grad_norm": 0.36466121673583984,
+      "learning_rate": 9.315779487031122e-05,
+      "loss": 1.8285,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1685883452789031,
+      "grad_norm": 0.35838162899017334,
+      "learning_rate": 9.314955372342311e-05,
+      "loss": 1.643,
+      "step": 1623
+    },
+    {
+      "epoch": 0.16869221979848342,
+      "grad_norm": 0.37461066246032715,
+      "learning_rate": 9.314130798144412e-05,
+      "loss": 1.8822,
+      "step": 1624
+    },
+    {
+      "epoch": 0.16879609431806378,
+      "grad_norm": 0.4094388484954834,
+      "learning_rate": 9.313305764525232e-05,
+      "loss": 1.9001,
+      "step": 1625
+    },
+    {
+      "epoch": 0.16889996883764413,
+      "grad_norm": 0.382744163274765,
+      "learning_rate": 9.312480271572634e-05,
+      "loss": 1.8858,
+      "step": 1626
+    },
+    {
+      "epoch": 0.16900384335722446,
+      "grad_norm": 0.3661576807498932,
+      "learning_rate": 9.311654319374528e-05,
+      "loss": 1.8909,
+      "step": 1627
+    },
+    {
+      "epoch": 0.16910771787680481,
+      "grad_norm": 0.38933804631233215,
+      "learning_rate": 9.310827908018866e-05,
+      "loss": 1.8039,
+      "step": 1628
+    },
+    {
+      "epoch": 0.16921159239638517,
+      "grad_norm": 0.3421494662761688,
+      "learning_rate": 9.31000103759366e-05,
+      "loss": 1.6478,
+      "step": 1629
+    },
+    {
+      "epoch": 0.16931546691596552,
+      "grad_norm": 0.3714037239551544,
+      "learning_rate": 9.309173708186963e-05,
+      "loss": 1.7259,
+      "step": 1630
+    },
+    {
+      "epoch": 0.16941934143554585,
+      "grad_norm": 0.4035167694091797,
+      "learning_rate": 9.308345919886877e-05,
+      "loss": 1.8386,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1695232159551262,
+      "grad_norm": 0.37035608291625977,
+      "learning_rate": 9.307517672781557e-05,
+      "loss": 1.8391,
+      "step": 1632
+    },
+    {
+      "epoch": 0.16962709047470656,
+      "grad_norm": 0.37822678685188293,
+      "learning_rate": 9.306688966959205e-05,
+      "loss": 1.8277,
+      "step": 1633
+    },
+    {
+      "epoch": 0.16973096499428691,
+      "grad_norm": 0.3702380955219269,
+      "learning_rate": 9.305859802508073e-05,
+      "loss": 1.7996,
+      "step": 1634
+    },
+    {
+      "epoch": 0.16983483951386724,
+      "grad_norm": 0.37005430459976196,
+      "learning_rate": 9.305030179516458e-05,
+      "loss": 1.7441,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1699387140334476,
+      "grad_norm": 0.38137298822402954,
+      "learning_rate": 9.304200098072709e-05,
+      "loss": 1.9145,
+      "step": 1636
+    },
+    {
+      "epoch": 0.17004258855302795,
+      "grad_norm": 0.39384132623672485,
+      "learning_rate": 9.303369558265224e-05,
+      "loss": 1.7381,
+      "step": 1637
+    },
+    {
+      "epoch": 0.17014646307260828,
+      "grad_norm": 0.34569051861763,
+      "learning_rate": 9.30253856018245e-05,
+      "loss": 1.6391,
+      "step": 1638
+    },
+    {
+      "epoch": 0.17025033759218863,
+      "grad_norm": 0.38778260350227356,
+      "learning_rate": 9.301707103912878e-05,
+      "loss": 1.8337,
+      "step": 1639
+    },
+    {
+      "epoch": 0.170354212111769,
+      "grad_norm": 0.390543133020401,
+      "learning_rate": 9.300875189545055e-05,
+      "loss": 1.9798,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17045808663134934,
+      "grad_norm": 0.3621176779270172,
+      "learning_rate": 9.300042817167574e-05,
+      "loss": 1.6744,
+      "step": 1641
+    },
+    {
+      "epoch": 0.17056196115092967,
+      "grad_norm": 0.43153461813926697,
+      "learning_rate": 9.299209986869073e-05,
+      "loss": 2.0351,
+      "step": 1642
+    },
+    {
+      "epoch": 0.17066583567051002,
+      "grad_norm": 0.39063888788223267,
+      "learning_rate": 9.298376698738242e-05,
+      "loss": 1.8298,
+      "step": 1643
+    },
+    {
+      "epoch": 0.17076971019009038,
+      "grad_norm": 0.36155179142951965,
+      "learning_rate": 9.297542952863824e-05,
+      "loss": 1.7124,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1708735847096707,
+      "grad_norm": 0.4032329022884369,
+      "learning_rate": 9.296708749334603e-05,
+      "loss": 2.0271,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17097745922925106,
+      "grad_norm": 0.39317673444747925,
+      "learning_rate": 9.295874088239416e-05,
+      "loss": 1.8808,
+      "step": 1646
+    },
+    {
+      "epoch": 0.17108133374883142,
+      "grad_norm": 0.36430609226226807,
+      "learning_rate": 9.295038969667145e-05,
+      "loss": 1.8879,
+      "step": 1647
+    },
+    {
+      "epoch": 0.17118520826841177,
+      "grad_norm": 0.3385809361934662,
+      "learning_rate": 9.29420339370673e-05,
+      "loss": 1.644,
+      "step": 1648
+    },
+    {
+      "epoch": 0.1712890827879921,
+      "grad_norm": 0.3685392737388611,
+      "learning_rate": 9.293367360447148e-05,
+      "loss": 1.907,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17139295730757245,
+      "grad_norm": 0.36233484745025635,
+      "learning_rate": 9.292530869977432e-05,
+      "loss": 1.6838,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1714968318271528,
+      "grad_norm": 0.4106137752532959,
+      "learning_rate": 9.29169392238666e-05,
+      "loss": 1.8516,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17160070634673313,
+      "grad_norm": 0.4084801971912384,
+      "learning_rate": 9.290856517763965e-05,
+      "loss": 1.8696,
+      "step": 1652
+    },
+    {
+      "epoch": 0.1717045808663135,
+      "grad_norm": 0.3670850694179535,
+      "learning_rate": 9.290018656198517e-05,
+      "loss": 1.7186,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17180845538589384,
+      "grad_norm": 0.362164705991745,
+      "learning_rate": 9.289180337779547e-05,
+      "loss": 1.6732,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1719123299054742,
+      "grad_norm": 0.3619086742401123,
+      "learning_rate": 9.288341562596328e-05,
+      "loss": 1.6082,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17201620442505453,
+      "grad_norm": 0.3776550889015198,
+      "learning_rate": 9.287502330738183e-05,
+      "loss": 1.7092,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17212007894463488,
+      "grad_norm": 0.3793870210647583,
+      "learning_rate": 9.286662642294484e-05,
+      "loss": 1.6828,
+      "step": 1657
+    },
+    {
+      "epoch": 0.17222395346421523,
+      "grad_norm": 0.3709464967250824,
+      "learning_rate": 9.28582249735465e-05,
+      "loss": 1.7037,
+      "step": 1658
+    },
+    {
+      "epoch": 0.17232782798379556,
+      "grad_norm": 0.35695111751556396,
+      "learning_rate": 9.284981896008152e-05,
+      "loss": 1.7627,
+      "step": 1659
+    },
+    {
+      "epoch": 0.17243170250337592,
+      "grad_norm": 0.3538592457771301,
+      "learning_rate": 9.284140838344505e-05,
+      "loss": 1.8424,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17253557702295627,
+      "grad_norm": 0.3740057647228241,
+      "learning_rate": 9.283299324453276e-05,
+      "loss": 1.9412,
+      "step": 1661
+    },
+    {
+      "epoch": 0.17263945154253663,
+      "grad_norm": 0.3617721199989319,
+      "learning_rate": 9.28245735442408e-05,
+      "loss": 1.6443,
+      "step": 1662
+    },
+    {
+      "epoch": 0.17274332606211695,
+      "grad_norm": 0.40431293845176697,
+      "learning_rate": 9.281614928346582e-05,
+      "loss": 1.9303,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1728472005816973,
+      "grad_norm": 0.3580422103404999,
+      "learning_rate": 9.280772046310491e-05,
+      "loss": 1.5823,
+      "step": 1664
+    },
+    {
+      "epoch": 0.17295107510127766,
+      "grad_norm": 0.38240423798561096,
+      "learning_rate": 9.279928708405568e-05,
+      "loss": 1.6381,
+      "step": 1665
+    },
+    {
+      "epoch": 0.17305494962085802,
+      "grad_norm": 0.3335290253162384,
+      "learning_rate": 9.279084914721623e-05,
+      "loss": 1.686,
+      "step": 1666
+    },
+    {
+      "epoch": 0.17315882414043834,
+      "grad_norm": 0.3627927899360657,
+      "learning_rate": 9.278240665348512e-05,
+      "loss": 1.8426,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1732626986600187,
+      "grad_norm": 0.4823165833950043,
+      "learning_rate": 9.277395960376143e-05,
+      "loss": 2.0619,
+      "step": 1668
+    },
+    {
+      "epoch": 0.17336657317959905,
+      "grad_norm": 0.36185434460639954,
+      "learning_rate": 9.276550799894468e-05,
+      "loss": 1.7277,
+      "step": 1669
+    },
+    {
+      "epoch": 0.17347044769917938,
+      "grad_norm": 0.36980822682380676,
+      "learning_rate": 9.275705183993491e-05,
+      "loss": 1.7649,
+      "step": 1670
+    },
+    {
+      "epoch": 0.17357432221875974,
+      "grad_norm": 0.3434619605541229,
+      "learning_rate": 9.274859112763264e-05,
+      "loss": 1.702,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1736781967383401,
+      "grad_norm": 0.3595278859138489,
+      "learning_rate": 9.274012586293887e-05,
+      "loss": 1.7485,
+      "step": 1672
+    },
+    {
+      "epoch": 0.17378207125792045,
+      "grad_norm": 0.4083694815635681,
+      "learning_rate": 9.273165604675509e-05,
+      "loss": 1.7587,
+      "step": 1673
+    },
+    {
+      "epoch": 0.17388594577750077,
+      "grad_norm": 0.3653002679347992,
+      "learning_rate": 9.272318167998325e-05,
+      "loss": 1.7234,
+      "step": 1674
+    },
+    {
+      "epoch": 0.17398982029708113,
+      "grad_norm": 0.40430814027786255,
+      "learning_rate": 9.27147027635258e-05,
+      "loss": 1.8682,
+      "step": 1675
+    },
+    {
+      "epoch": 0.17409369481666148,
+      "grad_norm": 0.4132486879825592,
+      "learning_rate": 9.270621929828571e-05,
+      "loss": 1.9371,
+      "step": 1676
+    },
+    {
+      "epoch": 0.1741975693362418,
+      "grad_norm": 0.3685455322265625,
+      "learning_rate": 9.269773128516638e-05,
+      "loss": 1.8035,
+      "step": 1677
+    },
+    {
+      "epoch": 0.17430144385582216,
+      "grad_norm": 0.35275524854660034,
+      "learning_rate": 9.268923872507174e-05,
+      "loss": 1.7536,
+      "step": 1678
+    },
+    {
+      "epoch": 0.17440531837540252,
+      "grad_norm": 0.364773154258728,
+      "learning_rate": 9.268074161890614e-05,
+      "loss": 1.8075,
+      "step": 1679
+    },
+    {
+      "epoch": 0.17450919289498287,
+      "grad_norm": 0.40323731303215027,
+      "learning_rate": 9.267223996757451e-05,
+      "loss": 1.8028,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1746130674145632,
+      "grad_norm": 0.34777867794036865,
+      "learning_rate": 9.266373377198216e-05,
+      "loss": 1.7143,
+      "step": 1681
+    },
+    {
+      "epoch": 0.17471694193414355,
+      "grad_norm": 0.39449429512023926,
+      "learning_rate": 9.265522303303494e-05,
+      "loss": 1.8556,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1748208164537239,
+      "grad_norm": 0.3480819761753082,
+      "learning_rate": 9.264670775163922e-05,
+      "loss": 1.6931,
+      "step": 1683
+    },
+    {
+      "epoch": 0.17492469097330424,
+      "grad_norm": 0.3808116316795349,
+      "learning_rate": 9.263818792870176e-05,
+      "loss": 1.6403,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1750285654928846,
+      "grad_norm": 0.3749628961086273,
+      "learning_rate": 9.262966356512987e-05,
+      "loss": 1.6206,
+      "step": 1685
+    },
+    {
+      "epoch": 0.17513244001246495,
+      "grad_norm": 0.3475877642631531,
+      "learning_rate": 9.262113466183136e-05,
+      "loss": 1.6314,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1752363145320453,
+      "grad_norm": 0.3954698443412781,
+      "learning_rate": 9.261260121971445e-05,
+      "loss": 1.7696,
+      "step": 1687
+    },
+    {
+      "epoch": 0.17534018905162563,
+      "grad_norm": 0.36054256558418274,
+      "learning_rate": 9.260406323968791e-05,
+      "loss": 1.734,
+      "step": 1688
+    },
+    {
+      "epoch": 0.17544406357120598,
+      "grad_norm": 0.3659268617630005,
+      "learning_rate": 9.259552072266096e-05,
+      "loss": 1.6301,
+      "step": 1689
+    },
+    {
+      "epoch": 0.17554793809078634,
+      "grad_norm": 0.3359396457672119,
+      "learning_rate": 9.258697366954329e-05,
+      "loss": 1.6708,
+      "step": 1690
+    },
+    {
+      "epoch": 0.17565181261036666,
+      "grad_norm": 0.3496209979057312,
+      "learning_rate": 9.257842208124514e-05,
+      "loss": 1.7327,
+      "step": 1691
+    },
+    {
+      "epoch": 0.17575568712994702,
+      "grad_norm": 0.4164491295814514,
+      "learning_rate": 9.256986595867715e-05,
+      "loss": 1.7489,
+      "step": 1692
+    },
+    {
+      "epoch": 0.17585956164952737,
+      "grad_norm": 0.4587213695049286,
+      "learning_rate": 9.256130530275048e-05,
+      "loss": 2.1299,
+      "step": 1693
+    },
+    {
+      "epoch": 0.17596343616910773,
+      "grad_norm": 0.40228530764579773,
+      "learning_rate": 9.255274011437683e-05,
+      "loss": 1.8252,
+      "step": 1694
+    },
+    {
+      "epoch": 0.17606731068868806,
+      "grad_norm": 0.3676128685474396,
+      "learning_rate": 9.254417039446824e-05,
+      "loss": 1.6666,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1761711852082684,
+      "grad_norm": 0.3423883318901062,
+      "learning_rate": 9.253559614393739e-05,
+      "loss": 1.7213,
+      "step": 1696
+    },
+    {
+      "epoch": 0.17627505972784877,
+      "grad_norm": 0.3809538781642914,
+      "learning_rate": 9.252701736369732e-05,
+      "loss": 1.8218,
+      "step": 1697
+    },
+    {
+      "epoch": 0.1763789342474291,
+      "grad_norm": 0.351162314414978,
+      "learning_rate": 9.251843405466162e-05,
+      "loss": 1.7478,
+      "step": 1698
+    },
+    {
+      "epoch": 0.17648280876700945,
+      "grad_norm": 0.3520200550556183,
+      "learning_rate": 9.250984621774437e-05,
+      "loss": 1.7541,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1765866832865898,
+      "grad_norm": 0.3424931764602661,
+      "learning_rate": 9.250125385386007e-05,
+      "loss": 1.6592,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17669055780617016,
+      "grad_norm": 0.3925075829029083,
+      "learning_rate": 9.249265696392377e-05,
+      "loss": 1.784,
+      "step": 1701
+    },
+    {
+      "epoch": 0.17679443232575048,
+      "grad_norm": 0.3799852132797241,
+      "learning_rate": 9.248405554885095e-05,
+      "loss": 1.7904,
+      "step": 1702
+    },
+    {
+      "epoch": 0.17689830684533084,
+      "grad_norm": 0.37215399742126465,
+      "learning_rate": 9.24754496095576e-05,
+      "loss": 1.8846,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1770021813649112,
+      "grad_norm": 0.37840306758880615,
+      "learning_rate": 9.246683914696019e-05,
+      "loss": 1.7425,
+      "step": 1704
+    },
+    {
+      "epoch": 0.17710605588449155,
+      "grad_norm": 0.362456351518631,
+      "learning_rate": 9.245822416197567e-05,
+      "loss": 1.8047,
+      "step": 1705
+    },
+    {
+      "epoch": 0.17720993040407187,
+      "grad_norm": 0.3728417456150055,
+      "learning_rate": 9.244960465552148e-05,
+      "loss": 1.8033,
+      "step": 1706
+    },
+    {
+      "epoch": 0.17731380492365223,
+      "grad_norm": 0.3797183334827423,
+      "learning_rate": 9.244098062851549e-05,
+      "loss": 1.9075,
+      "step": 1707
+    },
+    {
+      "epoch": 0.17741767944323258,
+      "grad_norm": 0.38207125663757324,
+      "learning_rate": 9.243235208187612e-05,
+      "loss": 1.8054,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1775215539628129,
+      "grad_norm": 0.3418576419353485,
+      "learning_rate": 9.242371901652223e-05,
+      "loss": 1.8205,
+      "step": 1709
+    },
+    {
+      "epoch": 0.17762542848239327,
+      "grad_norm": 0.4341420829296112,
+      "learning_rate": 9.241508143337319e-05,
+      "loss": 2.1321,
+      "step": 1710
+    },
+    {
+      "epoch": 0.17772930300197362,
+      "grad_norm": 0.38019636273384094,
+      "learning_rate": 9.240643933334884e-05,
+      "loss": 1.6551,
+      "step": 1711
+    },
+    {
+      "epoch": 0.17783317752155398,
+      "grad_norm": 0.3897992968559265,
+      "learning_rate": 9.239779271736947e-05,
+      "loss": 1.8644,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1779370520411343,
+      "grad_norm": 0.3626013398170471,
+      "learning_rate": 9.238914158635589e-05,
+      "loss": 1.6417,
+      "step": 1713
+    },
+    {
+      "epoch": 0.17804092656071466,
+      "grad_norm": 0.38448286056518555,
+      "learning_rate": 9.238048594122939e-05,
+      "loss": 1.7762,
+      "step": 1714
+    },
+    {
+      "epoch": 0.178144801080295,
+      "grad_norm": 0.3795235753059387,
+      "learning_rate": 9.237182578291171e-05,
+      "loss": 1.7897,
+      "step": 1715
+    },
+    {
+      "epoch": 0.17824867559987534,
+      "grad_norm": 0.3505716323852539,
+      "learning_rate": 9.23631611123251e-05,
+      "loss": 1.8132,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1783525501194557,
+      "grad_norm": 0.36175671219825745,
+      "learning_rate": 9.235449193039227e-05,
+      "loss": 1.6795,
+      "step": 1717
+    },
+    {
+      "epoch": 0.17845642463903605,
+      "grad_norm": 0.3952975571155548,
+      "learning_rate": 9.234581823803645e-05,
+      "loss": 1.7109,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1785602991586164,
+      "grad_norm": 0.3580237925052643,
+      "learning_rate": 9.233714003618127e-05,
+      "loss": 1.8293,
+      "step": 1719
+    },
+    {
+      "epoch": 0.17866417367819673,
+      "grad_norm": 0.33214688301086426,
+      "learning_rate": 9.232845732575093e-05,
+      "loss": 1.6718,
+      "step": 1720
+    },
+    {
+      "epoch": 0.17876804819777709,
+      "grad_norm": 0.359581857919693,
+      "learning_rate": 9.231977010767005e-05,
+      "loss": 1.8014,
+      "step": 1721
+    },
+    {
+      "epoch": 0.17887192271735744,
+      "grad_norm": 0.3323659300804138,
+      "learning_rate": 9.231107838286378e-05,
+      "loss": 1.6053,
+      "step": 1722
+    },
+    {
+      "epoch": 0.17897579723693777,
+      "grad_norm": 0.3609966039657593,
+      "learning_rate": 9.230238215225769e-05,
+      "loss": 1.9612,
+      "step": 1723
+    },
+    {
+      "epoch": 0.17907967175651812,
+      "grad_norm": 0.33431074023246765,
+      "learning_rate": 9.229368141677788e-05,
+      "loss": 1.6002,
+      "step": 1724
+    },
+    {
+      "epoch": 0.17918354627609848,
+      "grad_norm": 0.37472623586654663,
+      "learning_rate": 9.228497617735088e-05,
+      "loss": 1.7805,
+      "step": 1725
+    },
+    {
+      "epoch": 0.17928742079567883,
+      "grad_norm": 0.3726039528846741,
+      "learning_rate": 9.227626643490377e-05,
+      "loss": 1.8624,
+      "step": 1726
+    },
+    {
+      "epoch": 0.17939129531525916,
+      "grad_norm": 0.4065108895301819,
+      "learning_rate": 9.226755219036405e-05,
+      "loss": 1.7586,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1794951698348395,
+      "grad_norm": 0.35552752017974854,
+      "learning_rate": 9.225883344465972e-05,
+      "loss": 1.6,
+      "step": 1728
+    },
+    {
+      "epoch": 0.17959904435441987,
+      "grad_norm": 0.3508833944797516,
+      "learning_rate": 9.225011019871927e-05,
+      "loss": 1.7917,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1797029188740002,
+      "grad_norm": 0.374521404504776,
+      "learning_rate": 9.224138245347163e-05,
+      "loss": 1.7524,
+      "step": 1730
+    },
+    {
+      "epoch": 0.17980679339358055,
+      "grad_norm": 0.36706939339637756,
+      "learning_rate": 9.223265020984626e-05,
+      "loss": 1.7125,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1799106679131609,
+      "grad_norm": 0.3556000590324402,
+      "learning_rate": 9.222391346877308e-05,
+      "loss": 1.6935,
+      "step": 1732
+    },
+    {
+      "epoch": 0.18001454243274126,
+      "grad_norm": 0.3797661066055298,
+      "learning_rate": 9.221517223118245e-05,
+      "loss": 1.7725,
+      "step": 1733
+    },
+    {
+      "epoch": 0.18011841695232159,
+      "grad_norm": 0.3610544502735138,
+      "learning_rate": 9.22064264980053e-05,
+      "loss": 1.8192,
+      "step": 1734
+    },
+    {
+      "epoch": 0.18022229147190194,
+      "grad_norm": 0.46052882075309753,
+      "learning_rate": 9.219767627017293e-05,
+      "loss": 2.0761,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1803261659914823,
+      "grad_norm": 0.4026206135749817,
+      "learning_rate": 9.218892154861721e-05,
+      "loss": 1.978,
+      "step": 1736
+    },
+    {
+      "epoch": 0.18043004051106262,
+      "grad_norm": 0.37845221161842346,
+      "learning_rate": 9.218016233427042e-05,
+      "loss": 1.7811,
+      "step": 1737
+    },
+    {
+      "epoch": 0.18053391503064298,
+      "grad_norm": 0.4019536077976227,
+      "learning_rate": 9.217139862806534e-05,
+      "loss": 1.9864,
+      "step": 1738
+    },
+    {
+      "epoch": 0.18063778955022333,
+      "grad_norm": 0.36031562089920044,
+      "learning_rate": 9.21626304309353e-05,
+      "loss": 1.6712,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1807416640698037,
+      "grad_norm": 0.37847834825515747,
+      "learning_rate": 9.215385774381395e-05,
+      "loss": 1.7967,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18084553858938401,
+      "grad_norm": 0.38624006509780884,
+      "learning_rate": 9.21450805676356e-05,
+      "loss": 1.8833,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18094941310896437,
+      "grad_norm": 0.34793728590011597,
+      "learning_rate": 9.213629890333489e-05,
+      "loss": 1.8027,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18105328762854472,
+      "grad_norm": 0.3407808542251587,
+      "learning_rate": 9.212751275184704e-05,
+      "loss": 1.658,
+      "step": 1743
+    },
+    {
+      "epoch": 0.18115716214812508,
+      "grad_norm": 0.3690899908542633,
+      "learning_rate": 9.211872211410769e-05,
+      "loss": 1.6251,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1812610366677054,
+      "grad_norm": 0.3415517210960388,
+      "learning_rate": 9.210992699105296e-05,
+      "loss": 1.5998,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18136491118728576,
+      "grad_norm": 0.36001309752464294,
+      "learning_rate": 9.210112738361949e-05,
+      "loss": 1.8308,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18146878570686611,
+      "grad_norm": 0.3720461130142212,
+      "learning_rate": 9.209232329274433e-05,
+      "loss": 1.7902,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18157266022644644,
+      "grad_norm": 0.42275193333625793,
+      "learning_rate": 9.208351471936509e-05,
+      "loss": 1.9225,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1816765347460268,
+      "grad_norm": 0.36535343527793884,
+      "learning_rate": 9.20747016644198e-05,
+      "loss": 1.7431,
+      "step": 1749
+    },
+    {
+      "epoch": 0.18178040926560715,
+      "grad_norm": 0.3501549959182739,
+      "learning_rate": 9.206588412884698e-05,
+      "loss": 1.6417,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1818842837851875,
+      "grad_norm": 0.34967172145843506,
+      "learning_rate": 9.205706211358563e-05,
+      "loss": 1.7612,
+      "step": 1751
+    },
+    {
+      "epoch": 0.18198815830476783,
+      "grad_norm": 0.3451713025569916,
+      "learning_rate": 9.204823561957521e-05,
+      "loss": 1.6711,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1820920328243482,
+      "grad_norm": 0.37724578380584717,
+      "learning_rate": 9.203940464775568e-05,
+      "loss": 1.9513,
+      "step": 1753
+    },
+    {
+      "epoch": 0.18219590734392854,
+      "grad_norm": 0.34096843004226685,
+      "learning_rate": 9.203056919906748e-05,
+      "loss": 1.6691,
+      "step": 1754
+    },
+    {
+      "epoch": 0.18229978186350887,
+      "grad_norm": 0.3424280881881714,
+      "learning_rate": 9.202172927445152e-05,
+      "loss": 1.6556,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18240365638308922,
+      "grad_norm": 0.3700230121612549,
+      "learning_rate": 9.201288487484916e-05,
+      "loss": 1.7761,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18250753090266958,
+      "grad_norm": 0.4259064197540283,
+      "learning_rate": 9.200403600120229e-05,
+      "loss": 2.1197,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18261140542224993,
+      "grad_norm": 0.40020573139190674,
+      "learning_rate": 9.199518265445321e-05,
+      "loss": 1.9191,
+      "step": 1758
+    },
+    {
+      "epoch": 0.18271527994183026,
+      "grad_norm": 0.36124834418296814,
+      "learning_rate": 9.198632483554478e-05,
+      "loss": 1.8001,
+      "step": 1759
+    },
+    {
+      "epoch": 0.18281915446141062,
+      "grad_norm": 0.4086042642593384,
+      "learning_rate": 9.197746254542025e-05,
+      "loss": 1.8518,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18292302898099097,
+      "grad_norm": 0.37972450256347656,
+      "learning_rate": 9.196859578502338e-05,
+      "loss": 1.952,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1830269035005713,
+      "grad_norm": 0.4148435890674591,
+      "learning_rate": 9.195972455529844e-05,
+      "loss": 1.8547,
+      "step": 1762
+    },
+    {
+      "epoch": 0.18313077802015165,
+      "grad_norm": 0.3922263979911804,
+      "learning_rate": 9.195084885719014e-05,
+      "loss": 1.9248,
+      "step": 1763
+    },
+    {
+      "epoch": 0.183234652539732,
+      "grad_norm": 0.3774312138557434,
+      "learning_rate": 9.194196869164366e-05,
+      "loss": 1.8411,
+      "step": 1764
+    },
+    {
+      "epoch": 0.18333852705931236,
+      "grad_norm": 0.35800519585609436,
+      "learning_rate": 9.193308405960466e-05,
+      "loss": 1.6903,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1834424015788927,
+      "grad_norm": 0.40143853425979614,
+      "learning_rate": 9.192419496201932e-05,
+      "loss": 1.7171,
+      "step": 1766
+    },
+    {
+      "epoch": 0.18354627609847304,
+      "grad_norm": 0.3324778974056244,
+      "learning_rate": 9.191530139983423e-05,
+      "loss": 1.6295,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1836501506180534,
+      "grad_norm": 0.3377951979637146,
+      "learning_rate": 9.190640337399647e-05,
+      "loss": 1.5843,
+      "step": 1768
+    },
+    {
+      "epoch": 0.18375402513763373,
+      "grad_norm": 0.39436647295951843,
+      "learning_rate": 9.189750088545367e-05,
+      "loss": 1.957,
+      "step": 1769
+    },
+    {
+      "epoch": 0.18385789965721408,
+      "grad_norm": 0.42431554198265076,
+      "learning_rate": 9.188859393515382e-05,
+      "loss": 1.9529,
+      "step": 1770
+    },
+    {
+      "epoch": 0.18396177417679443,
+      "grad_norm": 0.3626527786254883,
+      "learning_rate": 9.187968252404547e-05,
+      "loss": 1.7235,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1840656486963748,
+      "grad_norm": 0.42632752656936646,
+      "learning_rate": 9.187076665307758e-05,
+      "loss": 2.1363,
+      "step": 1772
+    },
+    {
+      "epoch": 0.18416952321595512,
+      "grad_norm": 0.41993650794029236,
+      "learning_rate": 9.186184632319968e-05,
+      "loss": 1.7975,
+      "step": 1773
+    },
+    {
+      "epoch": 0.18427339773553547,
+      "grad_norm": 0.37287190556526184,
+      "learning_rate": 9.185292153536165e-05,
+      "loss": 1.8601,
+      "step": 1774
+    },
+    {
+      "epoch": 0.18437727225511583,
+      "grad_norm": 0.38827162981033325,
+      "learning_rate": 9.184399229051397e-05,
+      "loss": 1.8401,
+      "step": 1775
+    },
+    {
+      "epoch": 0.18448114677469618,
+      "grad_norm": 0.3717474937438965,
+      "learning_rate": 9.183505858960748e-05,
+      "loss": 1.7523,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1845850212942765,
+      "grad_norm": 0.3732149600982666,
+      "learning_rate": 9.18261204335936e-05,
+      "loss": 1.7327,
+      "step": 1777
+    },
+    {
+      "epoch": 0.18468889581385686,
+      "grad_norm": 0.3960385024547577,
+      "learning_rate": 9.181717782342414e-05,
+      "loss": 1.9453,
+      "step": 1778
+    },
+    {
+      "epoch": 0.18479277033343722,
+      "grad_norm": 0.38847383856773376,
+      "learning_rate": 9.180823076005143e-05,
+      "loss": 1.5841,
+      "step": 1779
+    },
+    {
+      "epoch": 0.18489664485301754,
+      "grad_norm": 0.37122729420661926,
+      "learning_rate": 9.179927924442826e-05,
+      "loss": 1.8535,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1850005193725979,
+      "grad_norm": 0.4589596688747406,
+      "learning_rate": 9.17903232775079e-05,
+      "loss": 2.1207,
+      "step": 1781
+    },
+    {
+      "epoch": 0.18510439389217825,
+      "grad_norm": 0.4033050537109375,
+      "learning_rate": 9.178136286024408e-05,
+      "loss": 1.9405,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1852082684117586,
+      "grad_norm": 0.37945422530174255,
+      "learning_rate": 9.177239799359102e-05,
+      "loss": 1.8607,
+      "step": 1783
+    },
+    {
+      "epoch": 0.18531214293133894,
+      "grad_norm": 0.44369664788246155,
+      "learning_rate": 9.176342867850343e-05,
+      "loss": 1.9337,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1854160174509193,
+      "grad_norm": 0.3960365951061249,
+      "learning_rate": 9.175445491593644e-05,
+      "loss": 1.8079,
+      "step": 1785
+    },
+    {
+      "epoch": 0.18551989197049965,
+      "grad_norm": 0.34727954864501953,
+      "learning_rate": 9.174547670684569e-05,
+      "loss": 1.5856,
+      "step": 1786
+    },
+    {
+      "epoch": 0.18562376649007997,
+      "grad_norm": 0.3571012616157532,
+      "learning_rate": 9.173649405218732e-05,
+      "loss": 1.6828,
+      "step": 1787
+    },
+    {
+      "epoch": 0.18572764100966033,
+      "grad_norm": 0.35931140184402466,
+      "learning_rate": 9.172750695291787e-05,
+      "loss": 1.5845,
+      "step": 1788
+    },
+    {
+      "epoch": 0.18583151552924068,
+      "grad_norm": 0.36305472254753113,
+      "learning_rate": 9.171851540999442e-05,
+      "loss": 1.7407,
+      "step": 1789
+    },
+    {
+      "epoch": 0.18593539004882104,
+      "grad_norm": 0.377492219209671,
+      "learning_rate": 9.17095194243745e-05,
+      "loss": 1.8655,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18603926456840136,
+      "grad_norm": 0.3644843101501465,
+      "learning_rate": 9.17005189970161e-05,
+      "loss": 1.7179,
+      "step": 1791
+    },
+    {
+      "epoch": 0.18614313908798172,
+      "grad_norm": 0.3360661268234253,
+      "learning_rate": 9.169151412887774e-05,
+      "loss": 1.6426,
+      "step": 1792
+    },
+    {
+      "epoch": 0.18624701360756207,
+      "grad_norm": 0.38643765449523926,
+      "learning_rate": 9.16825048209183e-05,
+      "loss": 1.7468,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1863508881271424,
+      "grad_norm": 0.34640881419181824,
+      "learning_rate": 9.167349107409721e-05,
+      "loss": 1.5946,
+      "step": 1794
+    },
+    {
+      "epoch": 0.18645476264672275,
+      "grad_norm": 0.33807694911956787,
+      "learning_rate": 9.166447288937444e-05,
+      "loss": 1.6169,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1865586371663031,
+      "grad_norm": 0.4004231095314026,
+      "learning_rate": 9.165545026771026e-05,
+      "loss": 1.7109,
+      "step": 1796
+    },
+    {
+      "epoch": 0.18666251168588346,
+      "grad_norm": 0.3662926256656647,
+      "learning_rate": 9.164642321006555e-05,
+      "loss": 1.7274,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1867663862054638,
+      "grad_norm": 0.37025517225265503,
+      "learning_rate": 9.163739171740167e-05,
+      "loss": 1.8982,
+      "step": 1798
+    },
+    {
+      "epoch": 0.18687026072504415,
+      "grad_norm": 0.4032931327819824,
+      "learning_rate": 9.162835579068032e-05,
+      "loss": 1.8586,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1869741352446245,
+      "grad_norm": 0.36516958475112915,
+      "learning_rate": 9.161931543086379e-05,
+      "loss": 1.8012,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18707800976420483,
+      "grad_norm": 0.35469850897789,
+      "learning_rate": 9.161027063891481e-05,
+      "loss": 1.6804,
+      "step": 1801
+    },
+    {
+      "epoch": 0.18718188428378518,
+      "grad_norm": 0.4029258191585541,
+      "learning_rate": 9.160122141579659e-05,
+      "loss": 2.0161,
+      "step": 1802
+    },
+    {
+      "epoch": 0.18728575880336554,
+      "grad_norm": 0.3713259696960449,
+      "learning_rate": 9.159216776247279e-05,
+      "loss": 1.7743,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1873896333229459,
+      "grad_norm": 0.3715076744556427,
+      "learning_rate": 9.158310967990755e-05,
+      "loss": 1.8117,
+      "step": 1804
+    },
+    {
+      "epoch": 0.18749350784252622,
+      "grad_norm": 0.43088141083717346,
+      "learning_rate": 9.15740471690655e-05,
+      "loss": 1.8559,
+      "step": 1805
+    },
+    {
+      "epoch": 0.18759738236210657,
+      "grad_norm": 0.38683950901031494,
+      "learning_rate": 9.156498023091172e-05,
+      "loss": 1.8888,
+      "step": 1806
+    },
+    {
+      "epoch": 0.18770125688168693,
+      "grad_norm": 0.3979988694190979,
+      "learning_rate": 9.155590886641174e-05,
+      "loss": 1.8934,
+      "step": 1807
+    },
+    {
+      "epoch": 0.18780513140126726,
+      "grad_norm": 0.39728519320487976,
+      "learning_rate": 9.154683307653162e-05,
+      "loss": 1.8035,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1879090059208476,
+      "grad_norm": 0.42086923122406006,
+      "learning_rate": 9.153775286223787e-05,
+      "loss": 1.9612,
+      "step": 1809
+    },
+    {
+      "epoch": 0.18801288044042797,
+      "grad_norm": 0.40852025151252747,
+      "learning_rate": 9.152866822449743e-05,
+      "loss": 1.6764,
+      "step": 1810
+    },
+    {
+      "epoch": 0.18811675496000832,
+      "grad_norm": 0.3808266222476959,
+      "learning_rate": 9.151957916427778e-05,
+      "loss": 1.7292,
+      "step": 1811
+    },
+    {
+      "epoch": 0.18822062947958865,
+      "grad_norm": 0.4031354784965515,
+      "learning_rate": 9.151048568254682e-05,
+      "loss": 1.9939,
+      "step": 1812
+    },
+    {
+      "epoch": 0.188324503999169,
+      "grad_norm": 0.3865586817264557,
+      "learning_rate": 9.15013877802729e-05,
+      "loss": 1.7508,
+      "step": 1813
+    },
+    {
+      "epoch": 0.18842837851874936,
+      "grad_norm": 0.35779932141304016,
+      "learning_rate": 9.149228545842492e-05,
+      "loss": 1.766,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1885322530383297,
+      "grad_norm": 0.35968711972236633,
+      "learning_rate": 9.148317871797219e-05,
+      "loss": 1.8,
+      "step": 1815
+    },
+    {
+      "epoch": 0.18863612755791004,
+      "grad_norm": 0.44821977615356445,
+      "learning_rate": 9.14740675598845e-05,
+      "loss": 1.8142,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1887400020774904,
+      "grad_norm": 0.433288037776947,
+      "learning_rate": 9.146495198513214e-05,
+      "loss": 1.9191,
+      "step": 1817
+    },
+    {
+      "epoch": 0.18884387659707075,
+      "grad_norm": 0.39990267157554626,
+      "learning_rate": 9.145583199468582e-05,
+      "loss": 1.8575,
+      "step": 1818
+    },
+    {
+      "epoch": 0.18894775111665107,
+      "grad_norm": 0.37523946166038513,
+      "learning_rate": 9.144670758951677e-05,
+      "loss": 1.799,
+      "step": 1819
+    },
+    {
+      "epoch": 0.18905162563623143,
+      "grad_norm": 0.3855798542499542,
+      "learning_rate": 9.143757877059665e-05,
+      "loss": 1.8317,
+      "step": 1820
+    },
+    {
+      "epoch": 0.18915550015581178,
+      "grad_norm": 0.35304591059684753,
+      "learning_rate": 9.142844553889762e-05,
+      "loss": 1.5719,
+      "step": 1821
+    },
+    {
+      "epoch": 0.18925937467539214,
+      "grad_norm": 0.3850051760673523,
+      "learning_rate": 9.141930789539229e-05,
+      "loss": 1.7996,
+      "step": 1822
+    },
+    {
+      "epoch": 0.18936324919497247,
+      "grad_norm": 0.3510708212852478,
+      "learning_rate": 9.141016584105376e-05,
+      "loss": 1.5699,
+      "step": 1823
+    },
+    {
+      "epoch": 0.18946712371455282,
+      "grad_norm": 0.3610669672489166,
+      "learning_rate": 9.140101937685558e-05,
+      "loss": 1.7303,
+      "step": 1824
+    },
+    {
+      "epoch": 0.18957099823413318,
+      "grad_norm": 0.34818708896636963,
+      "learning_rate": 9.139186850377177e-05,
+      "loss": 1.7005,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1896748727537135,
+      "grad_norm": 0.35581493377685547,
+      "learning_rate": 9.138271322277685e-05,
+      "loss": 1.6708,
+      "step": 1826
+    },
+    {
+      "epoch": 0.18977874727329386,
+      "grad_norm": 0.35102319717407227,
+      "learning_rate": 9.137355353484573e-05,
+      "loss": 1.8142,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1898826217928742,
+      "grad_norm": 0.3373032212257385,
+      "learning_rate": 9.136438944095392e-05,
+      "loss": 1.741,
+      "step": 1828
+    },
+    {
+      "epoch": 0.18998649631245457,
+      "grad_norm": 0.37855619192123413,
+      "learning_rate": 9.135522094207729e-05,
+      "loss": 1.7995,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1900903708320349,
+      "grad_norm": 0.35194653272628784,
+      "learning_rate": 9.13460480391922e-05,
+      "loss": 1.5989,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19019424535161525,
+      "grad_norm": 0.3600010871887207,
+      "learning_rate": 9.13368707332755e-05,
+      "loss": 1.8601,
+      "step": 1831
+    },
+    {
+      "epoch": 0.1902981198711956,
+      "grad_norm": 0.37306833267211914,
+      "learning_rate": 9.13276890253045e-05,
+      "loss": 1.86,
+      "step": 1832
+    },
+    {
+      "epoch": 0.19040199439077593,
+      "grad_norm": 0.355836421251297,
+      "learning_rate": 9.131850291625702e-05,
+      "loss": 1.6658,
+      "step": 1833
+    },
+    {
+      "epoch": 0.19050586891035629,
+      "grad_norm": 0.4272918999195099,
+      "learning_rate": 9.130931240711126e-05,
+      "loss": 1.9726,
+      "step": 1834
+    },
+    {
+      "epoch": 0.19060974342993664,
+      "grad_norm": 0.4324601888656616,
+      "learning_rate": 9.130011749884595e-05,
+      "loss": 1.851,
+      "step": 1835
+    },
+    {
+      "epoch": 0.190713617949517,
+      "grad_norm": 0.36520737409591675,
+      "learning_rate": 9.129091819244028e-05,
+      "loss": 1.8169,
+      "step": 1836
+    },
+    {
+      "epoch": 0.19081749246909732,
+      "grad_norm": 0.4533885419368744,
+      "learning_rate": 9.12817144888739e-05,
+      "loss": 2.0705,
+      "step": 1837
+    },
+    {
+      "epoch": 0.19092136698867768,
+      "grad_norm": 0.3979325592517853,
+      "learning_rate": 9.127250638912695e-05,
+      "loss": 1.7838,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19102524150825803,
+      "grad_norm": 0.3549365997314453,
+      "learning_rate": 9.126329389418e-05,
+      "loss": 1.8239,
+      "step": 1839
+    },
+    {
+      "epoch": 0.19112911602783836,
+      "grad_norm": 0.3781158924102783,
+      "learning_rate": 9.125407700501412e-05,
+      "loss": 1.6151,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1912329905474187,
+      "grad_norm": 0.388790100812912,
+      "learning_rate": 9.124485572261082e-05,
+      "loss": 1.9243,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19133686506699907,
+      "grad_norm": 0.4158736765384674,
+      "learning_rate": 9.123563004795213e-05,
+      "loss": 1.9735,
+      "step": 1842
+    },
+    {
+      "epoch": 0.19144073958657942,
+      "grad_norm": 0.39373651146888733,
+      "learning_rate": 9.122639998202047e-05,
+      "loss": 1.8774,
+      "step": 1843
+    },
+    {
+      "epoch": 0.19154461410615975,
+      "grad_norm": 0.3896943926811218,
+      "learning_rate": 9.12171655257988e-05,
+      "loss": 1.8337,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1916484886257401,
+      "grad_norm": 0.3980332016944885,
+      "learning_rate": 9.120792668027049e-05,
+      "loss": 1.8081,
+      "step": 1845
+    },
+    {
+      "epoch": 0.19175236314532046,
+      "grad_norm": 0.4258505403995514,
+      "learning_rate": 9.119868344641944e-05,
+      "loss": 1.8096,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19185623766490079,
+      "grad_norm": 0.3683486580848694,
+      "learning_rate": 9.118943582522996e-05,
+      "loss": 1.826,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19196011218448114,
+      "grad_norm": 0.3608863949775696,
+      "learning_rate": 9.118018381768687e-05,
+      "loss": 1.6942,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1920639867040615,
+      "grad_norm": 0.35737764835357666,
+      "learning_rate": 9.11709274247754e-05,
+      "loss": 1.8741,
+      "step": 1849
+    },
+    {
+      "epoch": 0.19216786122364185,
+      "grad_norm": 0.36617088317871094,
+      "learning_rate": 9.116166664748131e-05,
+      "loss": 1.7608,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19227173574322218,
+      "grad_norm": 0.40373408794403076,
+      "learning_rate": 9.11524014867908e-05,
+      "loss": 1.715,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19237561026280253,
+      "grad_norm": 0.36378052830696106,
+      "learning_rate": 9.114313194369053e-05,
+      "loss": 1.5889,
+      "step": 1852
+    },
+    {
+      "epoch": 0.1924794847823829,
+      "grad_norm": 0.37960201501846313,
+      "learning_rate": 9.113385801916765e-05,
+      "loss": 1.6605,
+      "step": 1853
+    },
+    {
+      "epoch": 0.19258335930196324,
+      "grad_norm": 0.3682219982147217,
+      "learning_rate": 9.112457971420973e-05,
+      "loss": 1.8351,
+      "step": 1854
+    },
+    {
+      "epoch": 0.19268723382154357,
+      "grad_norm": 0.38188329339027405,
+      "learning_rate": 9.111529702980486e-05,
+      "loss": 1.8845,
+      "step": 1855
+    },
+    {
+      "epoch": 0.19279110834112392,
+      "grad_norm": 0.39885640144348145,
+      "learning_rate": 9.110600996694157e-05,
+      "loss": 1.8629,
+      "step": 1856
+    },
+    {
+      "epoch": 0.19289498286070428,
+      "grad_norm": 0.36927422881126404,
+      "learning_rate": 9.109671852660885e-05,
+      "loss": 1.718,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1929988573802846,
+      "grad_norm": 0.35329553484916687,
+      "learning_rate": 9.108742270979618e-05,
+      "loss": 1.6209,
+      "step": 1858
+    },
+    {
+      "epoch": 0.19310273189986496,
+      "grad_norm": 0.40359756350517273,
+      "learning_rate": 9.107812251749349e-05,
+      "loss": 1.936,
+      "step": 1859
+    },
+    {
+      "epoch": 0.19320660641944531,
+      "grad_norm": 0.3696520924568176,
+      "learning_rate": 9.106881795069116e-05,
+      "loss": 1.8329,
+      "step": 1860
+    },
+    {
+      "epoch": 0.19331048093902567,
+      "grad_norm": 0.38084790110588074,
+      "learning_rate": 9.105950901038008e-05,
+      "loss": 1.8072,
+      "step": 1861
+    },
+    {
+      "epoch": 0.193414355458606,
+      "grad_norm": 0.4722822606563568,
+      "learning_rate": 9.105019569755157e-05,
+      "loss": 1.8516,
+      "step": 1862
+    },
+    {
+      "epoch": 0.19351822997818635,
+      "grad_norm": 0.3589233160018921,
+      "learning_rate": 9.104087801319741e-05,
+      "loss": 1.7609,
+      "step": 1863
+    },
+    {
+      "epoch": 0.1936221044977667,
+      "grad_norm": 0.4372257590293884,
+      "learning_rate": 9.103155595830988e-05,
+      "loss": 1.9742,
+      "step": 1864
+    },
+    {
+      "epoch": 0.19372597901734703,
+      "grad_norm": 0.3735795021057129,
+      "learning_rate": 9.10222295338817e-05,
+      "loss": 1.71,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1938298535369274,
+      "grad_norm": 0.35734692215919495,
+      "learning_rate": 9.101289874090605e-05,
+      "loss": 1.6526,
+      "step": 1866
+    },
+    {
+      "epoch": 0.19393372805650774,
+      "grad_norm": 0.36955031752586365,
+      "learning_rate": 9.10035635803766e-05,
+      "loss": 1.787,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1940376025760881,
+      "grad_norm": 0.3762964904308319,
+      "learning_rate": 9.099422405328748e-05,
+      "loss": 1.6645,
+      "step": 1868
+    },
+    {
+      "epoch": 0.19414147709566842,
+      "grad_norm": 0.40508878231048584,
+      "learning_rate": 9.098488016063325e-05,
+      "loss": 1.8264,
+      "step": 1869
+    },
+    {
+      "epoch": 0.19424535161524878,
+      "grad_norm": 0.43582627177238464,
+      "learning_rate": 9.097553190340899e-05,
+      "loss": 2.0092,
+      "step": 1870
+    },
+    {
+      "epoch": 0.19434922613482913,
+      "grad_norm": 0.379056453704834,
+      "learning_rate": 9.09661792826102e-05,
+      "loss": 1.7522,
+      "step": 1871
+    },
+    {
+      "epoch": 0.19445310065440946,
+      "grad_norm": 0.35009434819221497,
+      "learning_rate": 9.095682229923287e-05,
+      "loss": 1.7053,
+      "step": 1872
+    },
+    {
+      "epoch": 0.19455697517398982,
+      "grad_norm": 0.3844042420387268,
+      "learning_rate": 9.094746095427342e-05,
+      "loss": 1.7458,
+      "step": 1873
+    },
+    {
+      "epoch": 0.19466084969357017,
+      "grad_norm": 0.3690871000289917,
+      "learning_rate": 9.09380952487288e-05,
+      "loss": 1.7563,
+      "step": 1874
+    },
+    {
+      "epoch": 0.19476472421315053,
+      "grad_norm": 0.3871258497238159,
+      "learning_rate": 9.092872518359637e-05,
+      "loss": 1.9391,
+      "step": 1875
+    },
+    {
+      "epoch": 0.19486859873273085,
+      "grad_norm": 0.3885674774646759,
+      "learning_rate": 9.091935075987395e-05,
+      "loss": 1.7193,
+      "step": 1876
+    },
+    {
+      "epoch": 0.1949724732523112,
+      "grad_norm": 0.37015703320503235,
+      "learning_rate": 9.090997197855986e-05,
+      "loss": 1.7996,
+      "step": 1877
+    },
+    {
+      "epoch": 0.19507634777189156,
+      "grad_norm": 0.3506444990634918,
+      "learning_rate": 9.090058884065287e-05,
+      "loss": 1.5394,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1951802222914719,
+      "grad_norm": 0.350041002035141,
+      "learning_rate": 9.089120134715222e-05,
+      "loss": 1.8134,
+      "step": 1879
+    },
+    {
+      "epoch": 0.19528409681105224,
+      "grad_norm": 0.35820406675338745,
+      "learning_rate": 9.088180949905756e-05,
+      "loss": 1.7417,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1953879713306326,
+      "grad_norm": 0.37615320086479187,
+      "learning_rate": 9.087241329736909e-05,
+      "loss": 1.7162,
+      "step": 1881
+    },
+    {
+      "epoch": 0.19549184585021295,
+      "grad_norm": 0.35000884532928467,
+      "learning_rate": 9.086301274308742e-05,
+      "loss": 1.6961,
+      "step": 1882
+    },
+    {
+      "epoch": 0.19559572036979328,
+      "grad_norm": 0.3935372233390808,
+      "learning_rate": 9.085360783721363e-05,
+      "loss": 1.6883,
+      "step": 1883
+    },
+    {
+      "epoch": 0.19569959488937363,
+      "grad_norm": 0.37884512543678284,
+      "learning_rate": 9.084419858074928e-05,
+      "loss": 1.756,
+      "step": 1884
+    },
+    {
+      "epoch": 0.195803469408954,
+      "grad_norm": 0.33219221234321594,
+      "learning_rate": 9.08347849746964e-05,
+      "loss": 1.5695,
+      "step": 1885
+    },
+    {
+      "epoch": 0.19590734392853434,
+      "grad_norm": 0.3523041903972626,
+      "learning_rate": 9.082536702005741e-05,
+      "loss": 1.7402,
+      "step": 1886
+    },
+    {
+      "epoch": 0.19601121844811467,
+      "grad_norm": 0.38876470923423767,
+      "learning_rate": 9.08159447178353e-05,
+      "loss": 1.6901,
+      "step": 1887
+    },
+    {
+      "epoch": 0.19611509296769503,
+      "grad_norm": 0.35667043924331665,
+      "learning_rate": 9.080651806903346e-05,
+      "loss": 1.6679,
+      "step": 1888
+    },
+    {
+      "epoch": 0.19621896748727538,
+      "grad_norm": 0.3476751446723938,
+      "learning_rate": 9.079708707465575e-05,
+      "loss": 1.6656,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1963228420068557,
+      "grad_norm": 0.3749213218688965,
+      "learning_rate": 9.078765173570648e-05,
+      "loss": 1.6932,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19642671652643606,
+      "grad_norm": 0.38287267088890076,
+      "learning_rate": 9.077821205319045e-05,
+      "loss": 1.8927,
+      "step": 1891
+    },
+    {
+      "epoch": 0.19653059104601642,
+      "grad_norm": 0.3563637137413025,
+      "learning_rate": 9.076876802811293e-05,
+      "loss": 1.7309,
+      "step": 1892
+    },
+    {
+      "epoch": 0.19663446556559677,
+      "grad_norm": 0.4049491882324219,
+      "learning_rate": 9.075931966147964e-05,
+      "loss": 1.8704,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1967383400851771,
+      "grad_norm": 0.38649019598960876,
+      "learning_rate": 9.074986695429673e-05,
+      "loss": 1.7995,
+      "step": 1894
+    },
+    {
+      "epoch": 0.19684221460475745,
+      "grad_norm": 0.38346853852272034,
+      "learning_rate": 9.074040990757085e-05,
+      "loss": 1.7893,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1969460891243378,
+      "grad_norm": 0.39437806606292725,
+      "learning_rate": 9.07309485223091e-05,
+      "loss": 1.8473,
+      "step": 1896
+    },
+    {
+      "epoch": 0.19704996364391814,
+      "grad_norm": 0.3689126670360565,
+      "learning_rate": 9.072148279951905e-05,
+      "loss": 1.6973,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1971538381634985,
+      "grad_norm": 0.37634965777397156,
+      "learning_rate": 9.07120127402087e-05,
+      "loss": 1.5333,
+      "step": 1898
+    },
+    {
+      "epoch": 0.19725771268307885,
+      "grad_norm": 0.365343302488327,
+      "learning_rate": 9.070253834538658e-05,
+      "loss": 1.7896,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1973615872026592,
+      "grad_norm": 0.3354664444923401,
+      "learning_rate": 9.069305961606162e-05,
+      "loss": 1.5792,
+      "step": 1900
+    },
+    {
+      "epoch": 0.19746546172223953,
+      "grad_norm": 0.4631555676460266,
+      "learning_rate": 9.068357655324322e-05,
+      "loss": 1.9998,
+      "step": 1901
+    },
+    {
+      "epoch": 0.19756933624181988,
+      "grad_norm": 0.3880319893360138,
+      "learning_rate": 9.067408915794126e-05,
+      "loss": 1.6883,
+      "step": 1902
+    },
+    {
+      "epoch": 0.19767321076140024,
+      "grad_norm": 0.3678785562515259,
+      "learning_rate": 9.066459743116609e-05,
+      "loss": 1.7639,
+      "step": 1903
+    },
+    {
+      "epoch": 0.19777708528098056,
+      "grad_norm": 0.39961332082748413,
+      "learning_rate": 9.065510137392848e-05,
+      "loss": 1.8451,
+      "step": 1904
+    },
+    {
+      "epoch": 0.19788095980056092,
+      "grad_norm": 0.4052257239818573,
+      "learning_rate": 9.06456009872397e-05,
+      "loss": 1.7716,
+      "step": 1905
+    },
+    {
+      "epoch": 0.19798483432014127,
+      "grad_norm": 0.3587489128112793,
+      "learning_rate": 9.063609627211145e-05,
+      "loss": 1.7665,
+      "step": 1906
+    },
+    {
+      "epoch": 0.19808870883972163,
+      "grad_norm": 0.36075523495674133,
+      "learning_rate": 9.062658722955593e-05,
+      "loss": 1.7998,
+      "step": 1907
+    },
+    {
+      "epoch": 0.19819258335930195,
+      "grad_norm": 0.3680761754512787,
+      "learning_rate": 9.061707386058577e-05,
+      "loss": 1.7033,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1982964578788823,
+      "grad_norm": 0.37647727131843567,
+      "learning_rate": 9.060755616621406e-05,
+      "loss": 1.8304,
+      "step": 1909
+    },
+    {
+      "epoch": 0.19840033239846266,
+      "grad_norm": 0.3684910237789154,
+      "learning_rate": 9.059803414745438e-05,
+      "loss": 1.6814,
+      "step": 1910
+    },
+    {
+      "epoch": 0.198504206918043,
+      "grad_norm": 0.3947627544403076,
+      "learning_rate": 9.058850780532074e-05,
+      "loss": 1.8224,
+      "step": 1911
+    },
+    {
+      "epoch": 0.19860808143762335,
+      "grad_norm": 0.38467562198638916,
+      "learning_rate": 9.057897714082761e-05,
+      "loss": 1.6889,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1987119559572037,
+      "grad_norm": 0.38776230812072754,
+      "learning_rate": 9.056944215498994e-05,
+      "loss": 1.5652,
+      "step": 1913
+    },
+    {
+      "epoch": 0.19881583047678406,
+      "grad_norm": 0.36061033606529236,
+      "learning_rate": 9.055990284882317e-05,
+      "loss": 1.7888,
+      "step": 1914
+    },
+    {
+      "epoch": 0.19891970499636438,
+      "grad_norm": 0.37597185373306274,
+      "learning_rate": 9.05503592233431e-05,
+      "loss": 1.8526,
+      "step": 1915
+    },
+    {
+      "epoch": 0.19902357951594474,
+      "grad_norm": 0.39023831486701965,
+      "learning_rate": 9.054081127956607e-05,
+      "loss": 1.7178,
+      "step": 1916
+    },
+    {
+      "epoch": 0.1991274540355251,
+      "grad_norm": 0.37832608819007874,
+      "learning_rate": 9.053125901850888e-05,
+      "loss": 1.7645,
+      "step": 1917
+    },
+    {
+      "epoch": 0.19923132855510542,
+      "grad_norm": 0.3865016996860504,
+      "learning_rate": 9.052170244118876e-05,
+      "loss": 1.9984,
+      "step": 1918
+    },
+    {
+      "epoch": 0.19933520307468577,
+      "grad_norm": 0.42755410075187683,
+      "learning_rate": 9.05121415486234e-05,
+      "loss": 1.7952,
+      "step": 1919
+    },
+    {
+      "epoch": 0.19943907759426613,
+      "grad_norm": 0.3337387442588806,
+      "learning_rate": 9.050257634183099e-05,
+      "loss": 1.5986,
+      "step": 1920
+    },
+    {
+      "epoch": 0.19954295211384648,
+      "grad_norm": 0.35701167583465576,
+      "learning_rate": 9.049300682183012e-05,
+      "loss": 1.5109,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1996468266334268,
+      "grad_norm": 0.3579074442386627,
+      "learning_rate": 9.048343298963988e-05,
+      "loss": 1.8023,
+      "step": 1922
+    },
+    {
+      "epoch": 0.19975070115300717,
+      "grad_norm": 0.3522278666496277,
+      "learning_rate": 9.047385484627982e-05,
+      "loss": 1.8454,
+      "step": 1923
+    },
+    {
+      "epoch": 0.19985457567258752,
+      "grad_norm": 0.35523468255996704,
+      "learning_rate": 9.046427239276991e-05,
+      "loss": 1.7483,
+      "step": 1924
+    },
+    {
+      "epoch": 0.19995845019216787,
+      "grad_norm": 0.32691970467567444,
+      "learning_rate": 9.045468563013064e-05,
+      "loss": 1.6541,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2000623247117482,
+      "grad_norm": 0.3599613606929779,
+      "learning_rate": 9.04450945593829e-05,
+      "loss": 1.7776,
+      "step": 1926
+    },
+    {
+      "epoch": 0.20016619923132856,
+      "grad_norm": 0.3572538495063782,
+      "learning_rate": 9.043549918154808e-05,
+      "loss": 1.7373,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2002700737509089,
+      "grad_norm": 0.3507987856864929,
+      "learning_rate": 9.0425899497648e-05,
+      "loss": 1.7154,
+      "step": 1928
+    },
+    {
+      "epoch": 0.20037394827048924,
+      "grad_norm": 0.3993472456932068,
+      "learning_rate": 9.041629550870496e-05,
+      "loss": 1.8059,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2004778227900696,
+      "grad_norm": 0.34402957558631897,
+      "learning_rate": 9.040668721574171e-05,
+      "loss": 1.7075,
+      "step": 1930
+    },
+    {
+      "epoch": 0.20058169730964995,
+      "grad_norm": 0.3825515806674957,
+      "learning_rate": 9.039707461978146e-05,
+      "loss": 1.5313,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2006855718292303,
+      "grad_norm": 0.3753582537174225,
+      "learning_rate": 9.038745772184786e-05,
+      "loss": 1.792,
+      "step": 1932
+    },
+    {
+      "epoch": 0.20078944634881063,
+      "grad_norm": 0.39727622270584106,
+      "learning_rate": 9.037783652296505e-05,
+      "loss": 1.7264,
+      "step": 1933
+    },
+    {
+      "epoch": 0.20089332086839098,
+      "grad_norm": 0.3653308153152466,
+      "learning_rate": 9.036821102415762e-05,
+      "loss": 1.6462,
+      "step": 1934
+    },
+    {
+      "epoch": 0.20099719538797134,
+      "grad_norm": 0.3842056095600128,
+      "learning_rate": 9.03585812264506e-05,
+      "loss": 1.8057,
+      "step": 1935
+    },
+    {
+      "epoch": 0.20110106990755167,
+      "grad_norm": 0.3687531054019928,
+      "learning_rate": 9.034894713086947e-05,
+      "loss": 1.6832,
+      "step": 1936
+    },
+    {
+      "epoch": 0.20120494442713202,
+      "grad_norm": 0.4252033829689026,
+      "learning_rate": 9.033930873844023e-05,
+      "loss": 1.8803,
+      "step": 1937
+    },
+    {
+      "epoch": 0.20130881894671238,
+      "grad_norm": 0.3911295235157013,
+      "learning_rate": 9.032966605018924e-05,
+      "loss": 1.8326,
+      "step": 1938
+    },
+    {
+      "epoch": 0.20141269346629273,
+      "grad_norm": 0.3815619647502899,
+      "learning_rate": 9.032001906714343e-05,
+      "loss": 1.8654,
+      "step": 1939
+    },
+    {
+      "epoch": 0.20151656798587306,
+      "grad_norm": 0.42655375599861145,
+      "learning_rate": 9.031036779033006e-05,
+      "loss": 1.8325,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2016204425054534,
+      "grad_norm": 0.3972933292388916,
+      "learning_rate": 9.030071222077698e-05,
+      "loss": 1.5501,
+      "step": 1941
+    },
+    {
+      "epoch": 0.20172431702503377,
+      "grad_norm": 0.3478373885154724,
+      "learning_rate": 9.02910523595124e-05,
+      "loss": 1.5961,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2018281915446141,
+      "grad_norm": 0.36317178606987,
+      "learning_rate": 9.028138820756504e-05,
+      "loss": 1.7352,
+      "step": 1943
+    },
+    {
+      "epoch": 0.20193206606419445,
+      "grad_norm": 0.36241620779037476,
+      "learning_rate": 9.027171976596402e-05,
+      "loss": 1.7462,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2020359405837748,
+      "grad_norm": 0.3782672584056854,
+      "learning_rate": 9.026204703573899e-05,
+      "loss": 1.7317,
+      "step": 1945
+    },
+    {
+      "epoch": 0.20213981510335516,
+      "grad_norm": 0.3928276002407074,
+      "learning_rate": 9.025237001791999e-05,
+      "loss": 1.7494,
+      "step": 1946
+    },
+    {
+      "epoch": 0.20224368962293549,
+      "grad_norm": 0.38330700993537903,
+      "learning_rate": 9.024268871353758e-05,
+      "loss": 1.6797,
+      "step": 1947
+    },
+    {
+      "epoch": 0.20234756414251584,
+      "grad_norm": 0.38679707050323486,
+      "learning_rate": 9.023300312362273e-05,
+      "loss": 1.7813,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2024514386620962,
+      "grad_norm": 0.41488325595855713,
+      "learning_rate": 9.022331324920687e-05,
+      "loss": 1.8702,
+      "step": 1949
+    },
+    {
+      "epoch": 0.20255531318167652,
+      "grad_norm": 0.39658311009407043,
+      "learning_rate": 9.02136190913219e-05,
+      "loss": 1.7838,
+      "step": 1950
+    },
+    {
+      "epoch": 0.20265918770125688,
+      "grad_norm": 0.39057308435440063,
+      "learning_rate": 9.020392065100018e-05,
+      "loss": 1.814,
+      "step": 1951
+    },
+    {
+      "epoch": 0.20276306222083723,
+      "grad_norm": 0.374887079000473,
+      "learning_rate": 9.019421792927452e-05,
+      "loss": 1.7414,
+      "step": 1952
+    },
+    {
+      "epoch": 0.20286693674041759,
+      "grad_norm": 0.3583182990550995,
+      "learning_rate": 9.018451092717816e-05,
+      "loss": 1.6376,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2029708112599979,
+      "grad_norm": 0.35195374488830566,
+      "learning_rate": 9.017479964574485e-05,
+      "loss": 1.6618,
+      "step": 1954
+    },
+    {
+      "epoch": 0.20307468577957827,
+      "grad_norm": 0.3840351104736328,
+      "learning_rate": 9.016508408600875e-05,
+      "loss": 1.6952,
+      "step": 1955
+    },
+    {
+      "epoch": 0.20317856029915862,
+      "grad_norm": 0.37217509746551514,
+      "learning_rate": 9.015536424900449e-05,
+      "loss": 1.6792,
+      "step": 1956
+    },
+    {
+      "epoch": 0.20328243481873895,
+      "grad_norm": 0.39523565769195557,
+      "learning_rate": 9.014564013576716e-05,
+      "loss": 1.7578,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2033863093383193,
+      "grad_norm": 0.38107234239578247,
+      "learning_rate": 9.013591174733231e-05,
+      "loss": 1.8505,
+      "step": 1958
+    },
+    {
+      "epoch": 0.20349018385789966,
+      "grad_norm": 0.3982243835926056,
+      "learning_rate": 9.012617908473593e-05,
+      "loss": 1.9007,
+      "step": 1959
+    },
+    {
+      "epoch": 0.20359405837748001,
+      "grad_norm": 0.3710818290710449,
+      "learning_rate": 9.011644214901447e-05,
+      "loss": 1.7955,
+      "step": 1960
+    },
+    {
+      "epoch": 0.20369793289706034,
+      "grad_norm": 0.37171775102615356,
+      "learning_rate": 9.010670094120485e-05,
+      "loss": 1.8637,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2038018074166407,
+      "grad_norm": 0.3907235562801361,
+      "learning_rate": 9.009695546234443e-05,
+      "loss": 1.7001,
+      "step": 1962
+    },
+    {
+      "epoch": 0.20390568193622105,
+      "grad_norm": 0.34530210494995117,
+      "learning_rate": 9.0087205713471e-05,
+      "loss": 1.6479,
+      "step": 1963
+    },
+    {
+      "epoch": 0.2040095564558014,
+      "grad_norm": 0.37163665890693665,
+      "learning_rate": 9.007745169562285e-05,
+      "loss": 1.7622,
+      "step": 1964
+    },
+    {
+      "epoch": 0.20411343097538173,
+      "grad_norm": 0.36681413650512695,
+      "learning_rate": 9.006769340983873e-05,
+      "loss": 1.9061,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2042173054949621,
+      "grad_norm": 0.370410680770874,
+      "learning_rate": 9.00579308571578e-05,
+      "loss": 1.823,
+      "step": 1966
+    },
+    {
+      "epoch": 0.20432118001454244,
+      "grad_norm": 0.43219515681266785,
+      "learning_rate": 9.004816403861969e-05,
+      "loss": 1.9272,
+      "step": 1967
+    },
+    {
+      "epoch": 0.20442505453412277,
+      "grad_norm": 0.3857935667037964,
+      "learning_rate": 9.00383929552645e-05,
+      "loss": 1.7746,
+      "step": 1968
+    },
+    {
+      "epoch": 0.20452892905370312,
+      "grad_norm": 0.37894824147224426,
+      "learning_rate": 9.002861760813278e-05,
+      "loss": 1.7197,
+      "step": 1969
+    },
+    {
+      "epoch": 0.20463280357328348,
+      "grad_norm": 0.35982224345207214,
+      "learning_rate": 9.001883799826551e-05,
+      "loss": 1.7175,
+      "step": 1970
+    },
+    {
+      "epoch": 0.20473667809286383,
+      "grad_norm": 0.3776380717754364,
+      "learning_rate": 9.000905412670415e-05,
+      "loss": 1.7624,
+      "step": 1971
+    },
+    {
+      "epoch": 0.20484055261244416,
+      "grad_norm": 0.38274866342544556,
+      "learning_rate": 8.999926599449062e-05,
+      "loss": 1.7201,
+      "step": 1972
+    },
+    {
+      "epoch": 0.20494442713202451,
+      "grad_norm": 0.3876325488090515,
+      "learning_rate": 8.998947360266726e-05,
+      "loss": 1.8462,
+      "step": 1973
+    },
+    {
+      "epoch": 0.20504830165160487,
+      "grad_norm": 0.3528856039047241,
+      "learning_rate": 8.99796769522769e-05,
+      "loss": 1.7983,
+      "step": 1974
+    },
+    {
+      "epoch": 0.2051521761711852,
+      "grad_norm": 0.3806709945201874,
+      "learning_rate": 8.996987604436279e-05,
+      "loss": 2.0234,
+      "step": 1975
+    },
+    {
+      "epoch": 0.20525605069076555,
+      "grad_norm": 0.3677120506763458,
+      "learning_rate": 8.996007087996866e-05,
+      "loss": 1.7771,
+      "step": 1976
+    },
+    {
+      "epoch": 0.2053599252103459,
+      "grad_norm": 0.3892456591129303,
+      "learning_rate": 8.995026146013867e-05,
+      "loss": 1.6017,
+      "step": 1977
+    },
+    {
+      "epoch": 0.20546379972992626,
+      "grad_norm": 0.4247187077999115,
+      "learning_rate": 8.994044778591749e-05,
+      "loss": 1.7722,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2055676742495066,
+      "grad_norm": 0.3515567183494568,
+      "learning_rate": 8.993062985835013e-05,
+      "loss": 1.4704,
+      "step": 1979
+    },
+    {
+      "epoch": 0.20567154876908694,
+      "grad_norm": 0.3741825520992279,
+      "learning_rate": 8.99208076784822e-05,
+      "loss": 1.6708,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2057754232886673,
+      "grad_norm": 0.37131038308143616,
+      "learning_rate": 8.991098124735961e-05,
+      "loss": 1.7259,
+      "step": 1981
+    },
+    {
+      "epoch": 0.20587929780824762,
+      "grad_norm": 0.3691762387752533,
+      "learning_rate": 8.990115056602885e-05,
+      "loss": 1.8153,
+      "step": 1982
+    },
+    {
+      "epoch": 0.20598317232782798,
+      "grad_norm": 0.3380625545978546,
+      "learning_rate": 8.989131563553679e-05,
+      "loss": 1.7063,
+      "step": 1983
+    },
+    {
+      "epoch": 0.20608704684740833,
+      "grad_norm": 0.4201214909553528,
+      "learning_rate": 8.98814764569308e-05,
+      "loss": 1.8642,
+      "step": 1984
+    },
+    {
+      "epoch": 0.2061909213669887,
+      "grad_norm": 0.3742324709892273,
+      "learning_rate": 8.987163303125863e-05,
+      "loss": 1.6623,
+      "step": 1985
+    },
+    {
+      "epoch": 0.20629479588656902,
+      "grad_norm": 0.3564354181289673,
+      "learning_rate": 8.986178535956856e-05,
+      "loss": 1.7477,
+      "step": 1986
+    },
+    {
+      "epoch": 0.20639867040614937,
+      "grad_norm": 0.41312122344970703,
+      "learning_rate": 8.985193344290929e-05,
+      "loss": 1.901,
+      "step": 1987
+    },
+    {
+      "epoch": 0.20650254492572973,
+      "grad_norm": 0.4109034538269043,
+      "learning_rate": 8.984207728232995e-05,
+      "loss": 1.7853,
+      "step": 1988
+    },
+    {
+      "epoch": 0.20660641944531005,
+      "grad_norm": 0.383986234664917,
+      "learning_rate": 8.983221687888017e-05,
+      "loss": 1.8616,
+      "step": 1989
+    },
+    {
+      "epoch": 0.2067102939648904,
+      "grad_norm": 0.38250911235809326,
+      "learning_rate": 8.982235223360999e-05,
+      "loss": 1.7691,
+      "step": 1990
+    },
+    {
+      "epoch": 0.20681416848447076,
+      "grad_norm": 0.35506096482276917,
+      "learning_rate": 8.981248334756994e-05,
+      "loss": 1.8013,
+      "step": 1991
+    },
+    {
+      "epoch": 0.20691804300405112,
+      "grad_norm": 0.33986783027648926,
+      "learning_rate": 8.980261022181095e-05,
+      "loss": 1.624,
+      "step": 1992
+    },
+    {
+      "epoch": 0.20702191752363144,
+      "grad_norm": 0.35081538558006287,
+      "learning_rate": 8.979273285738445e-05,
+      "loss": 1.6776,
+      "step": 1993
+    },
+    {
+      "epoch": 0.2071257920432118,
+      "grad_norm": 0.3631971478462219,
+      "learning_rate": 8.978285125534229e-05,
+      "loss": 1.5715,
+      "step": 1994
+    },
+    {
+      "epoch": 0.20722966656279215,
+      "grad_norm": 0.38734301924705505,
+      "learning_rate": 8.977296541673678e-05,
+      "loss": 1.781,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2073335410823725,
+      "grad_norm": 0.40530499815940857,
+      "learning_rate": 8.976307534262072e-05,
+      "loss": 1.8291,
+      "step": 1996
+    },
+    {
+      "epoch": 0.20743741560195283,
+      "grad_norm": 0.4037350118160248,
+      "learning_rate": 8.975318103404728e-05,
+      "loss": 1.9573,
+      "step": 1997
+    },
+    {
+      "epoch": 0.2075412901215332,
+      "grad_norm": 0.3434680700302124,
+      "learning_rate": 8.974328249207015e-05,
+      "loss": 1.7654,
+      "step": 1998
+    },
+    {
+      "epoch": 0.20764516464111354,
+      "grad_norm": 0.38100457191467285,
+      "learning_rate": 8.973337971774344e-05,
+      "loss": 1.7762,
+      "step": 1999
+    },
+    {
+      "epoch": 0.20774903916069387,
+      "grad_norm": 0.3673883080482483,
+      "learning_rate": 8.972347271212173e-05,
+      "loss": 1.7563,
+      "step": 2000
+    },
+    {
+      "epoch": 0.20785291368027423,
+      "grad_norm": 0.3593236207962036,
+      "learning_rate": 8.971356147626004e-05,
+      "loss": 1.6374,
+      "step": 2001
+    },
+    {
+      "epoch": 0.20795678819985458,
+      "grad_norm": 0.4013366401195526,
+      "learning_rate": 8.970364601121382e-05,
+      "loss": 1.7564,
+      "step": 2002
+    },
+    {
+      "epoch": 0.20806066271943494,
+      "grad_norm": 0.36467140913009644,
+      "learning_rate": 8.9693726318039e-05,
+      "loss": 1.7618,
+      "step": 2003
+    },
+    {
+      "epoch": 0.20816453723901526,
+      "grad_norm": 0.36317500472068787,
+      "learning_rate": 8.968380239779194e-05,
+      "loss": 1.6911,
+      "step": 2004
+    },
+    {
+      "epoch": 0.20826841175859562,
+      "grad_norm": 0.3494802713394165,
+      "learning_rate": 8.96738742515295e-05,
+      "loss": 1.7087,
+      "step": 2005
+    },
+    {
+      "epoch": 0.20837228627817597,
+      "grad_norm": 0.3602224290370941,
+      "learning_rate": 8.96639418803089e-05,
+      "loss": 1.698,
+      "step": 2006
+    },
+    {
+      "epoch": 0.2084761607977563,
+      "grad_norm": 0.3831143379211426,
+      "learning_rate": 8.965400528518787e-05,
+      "loss": 1.7733,
+      "step": 2007
+    },
+    {
+      "epoch": 0.20858003531733665,
+      "grad_norm": 0.3766723871231079,
+      "learning_rate": 8.964406446722459e-05,
+      "loss": 1.7092,
+      "step": 2008
+    },
+    {
+      "epoch": 0.208683909836917,
+      "grad_norm": 0.38016197085380554,
+      "learning_rate": 8.96341194274777e-05,
+      "loss": 1.6285,
+      "step": 2009
+    },
+    {
+      "epoch": 0.20878778435649736,
+      "grad_norm": 0.3981560170650482,
+      "learning_rate": 8.962417016700624e-05,
+      "loss": 1.7449,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2088916588760777,
+      "grad_norm": 0.3760923445224762,
+      "learning_rate": 8.961421668686974e-05,
+      "loss": 1.7459,
+      "step": 2011
+    },
+    {
+      "epoch": 0.20899553339565805,
+      "grad_norm": 0.3543214797973633,
+      "learning_rate": 8.960425898812816e-05,
+      "loss": 1.7329,
+      "step": 2012
+    },
+    {
+      "epoch": 0.2090994079152384,
+      "grad_norm": 0.3547232747077942,
+      "learning_rate": 8.959429707184192e-05,
+      "loss": 1.7992,
+      "step": 2013
+    },
+    {
+      "epoch": 0.20920328243481873,
+      "grad_norm": 0.43568694591522217,
+      "learning_rate": 8.95843309390719e-05,
+      "loss": 2.0488,
+      "step": 2014
+    },
+    {
+      "epoch": 0.20930715695439908,
+      "grad_norm": 0.4254542291164398,
+      "learning_rate": 8.95743605908794e-05,
+      "loss": 1.796,
+      "step": 2015
+    },
+    {
+      "epoch": 0.20941103147397944,
+      "grad_norm": 0.3894093632698059,
+      "learning_rate": 8.956438602832619e-05,
+      "loss": 1.8641,
+      "step": 2016
+    },
+    {
+      "epoch": 0.2095149059935598,
+      "grad_norm": 0.383073091506958,
+      "learning_rate": 8.955440725247448e-05,
+      "loss": 1.7898,
+      "step": 2017
+    },
+    {
+      "epoch": 0.20961878051314012,
+      "grad_norm": 0.37630245089530945,
+      "learning_rate": 8.954442426438694e-05,
+      "loss": 1.738,
+      "step": 2018
+    },
+    {
+      "epoch": 0.20972265503272047,
+      "grad_norm": 0.3684820234775543,
+      "learning_rate": 8.953443706512667e-05,
+      "loss": 1.8819,
+      "step": 2019
+    },
+    {
+      "epoch": 0.20982652955230083,
+      "grad_norm": 0.36796849966049194,
+      "learning_rate": 8.952444565575723e-05,
+      "loss": 1.7886,
+      "step": 2020
+    },
+    {
+      "epoch": 0.20993040407188115,
+      "grad_norm": 0.35004639625549316,
+      "learning_rate": 8.951445003734263e-05,
+      "loss": 1.7872,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2100342785914615,
+      "grad_norm": 0.3540632724761963,
+      "learning_rate": 8.950445021094733e-05,
+      "loss": 1.6917,
+      "step": 2022
+    },
+    {
+      "epoch": 0.21013815311104186,
+      "grad_norm": 0.3964502215385437,
+      "learning_rate": 8.949444617763623e-05,
+      "loss": 1.7487,
+      "step": 2023
+    },
+    {
+      "epoch": 0.21024202763062222,
+      "grad_norm": 0.3707832098007202,
+      "learning_rate": 8.948443793847467e-05,
+      "loss": 1.8752,
+      "step": 2024
+    },
+    {
+      "epoch": 0.21034590215020255,
+      "grad_norm": 0.3899782598018646,
+      "learning_rate": 8.947442549452846e-05,
+      "loss": 1.7619,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2104497766697829,
+      "grad_norm": 0.349178671836853,
+      "learning_rate": 8.946440884686387e-05,
+      "loss": 1.5852,
+      "step": 2026
+    },
+    {
+      "epoch": 0.21055365118936326,
+      "grad_norm": 0.3928735554218292,
+      "learning_rate": 8.945438799654756e-05,
+      "loss": 1.8503,
+      "step": 2027
+    },
+    {
+      "epoch": 0.21065752570894358,
+      "grad_norm": 0.3914186954498291,
+      "learning_rate": 8.944436294464669e-05,
+      "loss": 1.6764,
+      "step": 2028
+    },
+    {
+      "epoch": 0.21076140022852394,
+      "grad_norm": 0.370089054107666,
+      "learning_rate": 8.943433369222882e-05,
+      "loss": 1.6127,
+      "step": 2029
+    },
+    {
+      "epoch": 0.2108652747481043,
+      "grad_norm": 0.35939261317253113,
+      "learning_rate": 8.942430024036203e-05,
+      "loss": 1.7161,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21096914926768465,
+      "grad_norm": 0.36686834692955017,
+      "learning_rate": 8.941426259011478e-05,
+      "loss": 1.8648,
+      "step": 2031
+    },
+    {
+      "epoch": 0.21107302378726497,
+      "grad_norm": 0.3749295473098755,
+      "learning_rate": 8.940422074255602e-05,
+      "loss": 1.7433,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21117689830684533,
+      "grad_norm": 0.36779385805130005,
+      "learning_rate": 8.93941746987551e-05,
+      "loss": 1.778,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21128077282642568,
+      "grad_norm": 0.3840864896774292,
+      "learning_rate": 8.938412445978186e-05,
+      "loss": 1.7633,
+      "step": 2034
+    },
+    {
+      "epoch": 0.21138464734600604,
+      "grad_norm": 0.3512420356273651,
+      "learning_rate": 8.937407002670659e-05,
+      "loss": 1.6726,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21148852186558637,
+      "grad_norm": 0.3997120261192322,
+      "learning_rate": 8.936401140059998e-05,
+      "loss": 1.7913,
+      "step": 2036
+    },
+    {
+      "epoch": 0.21159239638516672,
+      "grad_norm": 0.4093436300754547,
+      "learning_rate": 8.935394858253321e-05,
+      "loss": 1.7161,
+      "step": 2037
+    },
+    {
+      "epoch": 0.21169627090474707,
+      "grad_norm": 0.3791636824607849,
+      "learning_rate": 8.934388157357788e-05,
+      "loss": 1.9097,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2118001454243274,
+      "grad_norm": 0.39315351843833923,
+      "learning_rate": 8.933381037480606e-05,
+      "loss": 1.9521,
+      "step": 2039
+    },
+    {
+      "epoch": 0.21190401994390776,
+      "grad_norm": 0.3954058289527893,
+      "learning_rate": 8.932373498729025e-05,
+      "loss": 1.7008,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2120078944634881,
+      "grad_norm": 0.44486141204833984,
+      "learning_rate": 8.931365541210342e-05,
+      "loss": 1.8255,
+      "step": 2041
+    },
+    {
+      "epoch": 0.21211176898306847,
+      "grad_norm": 0.4147118330001831,
+      "learning_rate": 8.930357165031893e-05,
+      "loss": 1.9377,
+      "step": 2042
+    },
+    {
+      "epoch": 0.2122156435026488,
+      "grad_norm": 0.4043649733066559,
+      "learning_rate": 8.929348370301063e-05,
+      "loss": 1.8809,
+      "step": 2043
+    },
+    {
+      "epoch": 0.21231951802222915,
+      "grad_norm": 0.37084200978279114,
+      "learning_rate": 8.928339157125283e-05,
+      "loss": 1.7093,
+      "step": 2044
+    },
+    {
+      "epoch": 0.2124233925418095,
+      "grad_norm": 0.33814066648483276,
+      "learning_rate": 8.927329525612025e-05,
+      "loss": 1.7248,
+      "step": 2045
+    },
+    {
+      "epoch": 0.21252726706138983,
+      "grad_norm": 0.3607522249221802,
+      "learning_rate": 8.926319475868807e-05,
+      "loss": 1.6424,
+      "step": 2046
+    },
+    {
+      "epoch": 0.21263114158097018,
+      "grad_norm": 0.35810667276382446,
+      "learning_rate": 8.925309008003192e-05,
+      "loss": 1.6919,
+      "step": 2047
+    },
+    {
+      "epoch": 0.21273501610055054,
+      "grad_norm": 0.3557438850402832,
+      "learning_rate": 8.924298122122783e-05,
+      "loss": 1.8568,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2128388906201309,
+      "grad_norm": 0.37123459577560425,
+      "learning_rate": 8.923286818335237e-05,
+      "loss": 1.8214,
+      "step": 2049
+    },
+    {
+      "epoch": 0.21294276513971122,
+      "grad_norm": 0.42854633927345276,
+      "learning_rate": 8.922275096748247e-05,
+      "loss": 2.1206,
+      "step": 2050
+    },
+    {
+      "epoch": 0.21304663965929158,
+      "grad_norm": 0.4002090394496918,
+      "learning_rate": 8.921262957469554e-05,
+      "loss": 1.8579,
+      "step": 2051
+    },
+    {
+      "epoch": 0.21315051417887193,
+      "grad_norm": 0.34181690216064453,
+      "learning_rate": 8.920250400606944e-05,
+      "loss": 1.7962,
+      "step": 2052
+    },
+    {
+      "epoch": 0.21325438869845226,
+      "grad_norm": 0.35831955075263977,
+      "learning_rate": 8.919237426268246e-05,
+      "loss": 1.6304,
+      "step": 2053
+    },
+    {
+      "epoch": 0.2133582632180326,
+      "grad_norm": 0.4297682046890259,
+      "learning_rate": 8.918224034561333e-05,
+      "loss": 1.9758,
+      "step": 2054
+    },
+    {
+      "epoch": 0.21346213773761297,
+      "grad_norm": 0.3815910518169403,
+      "learning_rate": 8.917210225594122e-05,
+      "loss": 1.7476,
+      "step": 2055
+    },
+    {
+      "epoch": 0.21356601225719332,
+      "grad_norm": 0.3674294650554657,
+      "learning_rate": 8.916195999474579e-05,
+      "loss": 1.7741,
+      "step": 2056
+    },
+    {
+      "epoch": 0.21366988677677365,
+      "grad_norm": 0.3721073567867279,
+      "learning_rate": 8.915181356310709e-05,
+      "loss": 1.7451,
+      "step": 2057
+    },
+    {
+      "epoch": 0.213773761296354,
+      "grad_norm": 0.379859983921051,
+      "learning_rate": 8.914166296210564e-05,
+      "loss": 1.9665,
+      "step": 2058
+    },
+    {
+      "epoch": 0.21387763581593436,
+      "grad_norm": 0.3658583462238312,
+      "learning_rate": 8.913150819282242e-05,
+      "loss": 1.6796,
+      "step": 2059
+    },
+    {
+      "epoch": 0.21398151033551469,
+      "grad_norm": 0.3702607750892639,
+      "learning_rate": 8.912134925633878e-05,
+      "loss": 1.7107,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21408538485509504,
+      "grad_norm": 0.37530508637428284,
+      "learning_rate": 8.911118615373661e-05,
+      "loss": 1.6896,
+      "step": 2061
+    },
+    {
+      "epoch": 0.2141892593746754,
+      "grad_norm": 0.3804415464401245,
+      "learning_rate": 8.910101888609821e-05,
+      "loss": 1.8049,
+      "step": 2062
+    },
+    {
+      "epoch": 0.21429313389425575,
+      "grad_norm": 0.36921432614326477,
+      "learning_rate": 8.909084745450628e-05,
+      "loss": 1.7652,
+      "step": 2063
+    },
+    {
+      "epoch": 0.21439700841383608,
+      "grad_norm": 0.38994041085243225,
+      "learning_rate": 8.908067186004405e-05,
+      "loss": 1.701,
+      "step": 2064
+    },
+    {
+      "epoch": 0.21450088293341643,
+      "grad_norm": 0.3705255389213562,
+      "learning_rate": 8.907049210379508e-05,
+      "loss": 1.781,
+      "step": 2065
+    },
+    {
+      "epoch": 0.21460475745299679,
+      "grad_norm": 0.3754466772079468,
+      "learning_rate": 8.906030818684348e-05,
+      "loss": 1.6878,
+      "step": 2066
+    },
+    {
+      "epoch": 0.2147086319725771,
+      "grad_norm": 0.3889169991016388,
+      "learning_rate": 8.905012011027372e-05,
+      "loss": 1.6964,
+      "step": 2067
+    },
+    {
+      "epoch": 0.21481250649215747,
+      "grad_norm": 0.3674185872077942,
+      "learning_rate": 8.903992787517078e-05,
+      "loss": 1.2466,
+      "step": 2068
+    },
+    {
+      "epoch": 0.21491638101173782,
+      "grad_norm": 0.37006455659866333,
+      "learning_rate": 8.902973148262004e-05,
+      "loss": 1.7607,
+      "step": 2069
+    },
+    {
+      "epoch": 0.21502025553131818,
+      "grad_norm": 0.3947128355503082,
+      "learning_rate": 8.901953093370734e-05,
+      "loss": 1.8514,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2151241300508985,
+      "grad_norm": 0.36148831248283386,
+      "learning_rate": 8.900932622951897e-05,
+      "loss": 1.5947,
+      "step": 2071
+    },
+    {
+      "epoch": 0.21522800457047886,
+      "grad_norm": 0.34942540526390076,
+      "learning_rate": 8.899911737114163e-05,
+      "loss": 1.7449,
+      "step": 2072
+    },
+    {
+      "epoch": 0.21533187909005921,
+      "grad_norm": 0.3522939383983612,
+      "learning_rate": 8.898890435966251e-05,
+      "loss": 1.7266,
+      "step": 2073
+    },
+    {
+      "epoch": 0.21543575360963957,
+      "grad_norm": 0.4223746955394745,
+      "learning_rate": 8.897868719616919e-05,
+      "loss": 1.8562,
+      "step": 2074
+    },
+    {
+      "epoch": 0.2155396281292199,
+      "grad_norm": 0.3850705921649933,
+      "learning_rate": 8.896846588174973e-05,
+      "loss": 1.8528,
+      "step": 2075
+    },
+    {
+      "epoch": 0.21564350264880025,
+      "grad_norm": 0.391237735748291,
+      "learning_rate": 8.89582404174926e-05,
+      "loss": 1.6268,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2157473771683806,
+      "grad_norm": 0.40240150690078735,
+      "learning_rate": 8.89480108044868e-05,
+      "loss": 1.7256,
+      "step": 2077
+    },
+    {
+      "epoch": 0.21585125168796093,
+      "grad_norm": 0.400774210691452,
+      "learning_rate": 8.893777704382163e-05,
+      "loss": 1.7654,
+      "step": 2078
+    },
+    {
+      "epoch": 0.2159551262075413,
+      "grad_norm": 0.4116402268409729,
+      "learning_rate": 8.892753913658691e-05,
+      "loss": 1.5765,
+      "step": 2079
+    },
+    {
+      "epoch": 0.21605900072712164,
+      "grad_norm": 0.3588711619377136,
+      "learning_rate": 8.891729708387294e-05,
+      "loss": 1.7399,
+      "step": 2080
+    },
+    {
+      "epoch": 0.216162875246702,
+      "grad_norm": 0.40262454748153687,
+      "learning_rate": 8.890705088677039e-05,
+      "loss": 1.8868,
+      "step": 2081
+    },
+    {
+      "epoch": 0.21626674976628232,
+      "grad_norm": 0.38147544860839844,
+      "learning_rate": 8.889680054637042e-05,
+      "loss": 1.907,
+      "step": 2082
+    },
+    {
+      "epoch": 0.21637062428586268,
+      "grad_norm": 0.40591171383857727,
+      "learning_rate": 8.888654606376459e-05,
+      "loss": 1.9046,
+      "step": 2083
+    },
+    {
+      "epoch": 0.21647449880544303,
+      "grad_norm": 0.3788270652294159,
+      "learning_rate": 8.887628744004493e-05,
+      "loss": 1.7253,
+      "step": 2084
+    },
+    {
+      "epoch": 0.21657837332502336,
+      "grad_norm": 0.38902533054351807,
+      "learning_rate": 8.88660246763039e-05,
+      "loss": 1.7522,
+      "step": 2085
+    },
+    {
+      "epoch": 0.21668224784460371,
+      "grad_norm": 0.3673681318759918,
+      "learning_rate": 8.885575777363442e-05,
+      "loss": 1.7231,
+      "step": 2086
+    },
+    {
+      "epoch": 0.21678612236418407,
+      "grad_norm": 0.33182385563850403,
+      "learning_rate": 8.884548673312981e-05,
+      "loss": 1.6142,
+      "step": 2087
+    },
+    {
+      "epoch": 0.21688999688376442,
+      "grad_norm": 0.38326361775398254,
+      "learning_rate": 8.883521155588388e-05,
+      "loss": 1.8635,
+      "step": 2088
+    },
+    {
+      "epoch": 0.21699387140334475,
+      "grad_norm": 0.38930484652519226,
+      "learning_rate": 8.882493224299084e-05,
+      "loss": 1.6453,
+      "step": 2089
+    },
+    {
+      "epoch": 0.2170977459229251,
+      "grad_norm": 0.3819931447505951,
+      "learning_rate": 8.881464879554536e-05,
+      "loss": 1.821,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21720162044250546,
+      "grad_norm": 0.3670012354850769,
+      "learning_rate": 8.880436121464255e-05,
+      "loss": 1.8536,
+      "step": 2091
+    },
+    {
+      "epoch": 0.2173054949620858,
+      "grad_norm": 0.3817991614341736,
+      "learning_rate": 8.879406950137796e-05,
+      "loss": 1.7896,
+      "step": 2092
+    },
+    {
+      "epoch": 0.21740936948166614,
+      "grad_norm": 0.44066059589385986,
+      "learning_rate": 8.878377365684758e-05,
+      "loss": 1.7763,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2175132440012465,
+      "grad_norm": 0.362020343542099,
+      "learning_rate": 8.877347368214783e-05,
+      "loss": 1.7594,
+      "step": 2094
+    },
+    {
+      "epoch": 0.21761711852082685,
+      "grad_norm": 0.385455459356308,
+      "learning_rate": 8.876316957837556e-05,
+      "loss": 1.9007,
+      "step": 2095
+    },
+    {
+      "epoch": 0.21772099304040718,
+      "grad_norm": 0.3932596743106842,
+      "learning_rate": 8.87528613466281e-05,
+      "loss": 1.6164,
+      "step": 2096
+    },
+    {
+      "epoch": 0.21782486755998753,
+      "grad_norm": 0.3642088770866394,
+      "learning_rate": 8.874254898800321e-05,
+      "loss": 1.626,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2179287420795679,
+      "grad_norm": 0.4024796783924103,
+      "learning_rate": 8.873223250359903e-05,
+      "loss": 1.8004,
+      "step": 2098
+    },
+    {
+      "epoch": 0.21803261659914822,
+      "grad_norm": 0.3582911491394043,
+      "learning_rate": 8.872191189451422e-05,
+      "loss": 1.7169,
+      "step": 2099
+    },
+    {
+      "epoch": 0.21813649111872857,
+      "grad_norm": 0.40865403413772583,
+      "learning_rate": 8.871158716184784e-05,
+      "loss": 1.6472,
+      "step": 2100
+    },
+    {
+      "epoch": 0.21824036563830893,
+      "grad_norm": 0.39730289578437805,
+      "learning_rate": 8.87012583066994e-05,
+      "loss": 1.6177,
+      "step": 2101
+    },
+    {
+      "epoch": 0.21834424015788928,
+      "grad_norm": 0.3871819078922272,
+      "learning_rate": 8.869092533016882e-05,
+      "loss": 1.6924,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2184481146774696,
+      "grad_norm": 0.377941757440567,
+      "learning_rate": 8.86805882333565e-05,
+      "loss": 1.8299,
+      "step": 2103
+    },
+    {
+      "epoch": 0.21855198919704996,
+      "grad_norm": 0.397905558347702,
+      "learning_rate": 8.867024701736325e-05,
+      "loss": 1.8411,
+      "step": 2104
+    },
+    {
+      "epoch": 0.21865586371663032,
+      "grad_norm": 0.3686088025569916,
+      "learning_rate": 8.865990168329031e-05,
+      "loss": 1.6618,
+      "step": 2105
+    },
+    {
+      "epoch": 0.21875973823621067,
+      "grad_norm": 0.394415020942688,
+      "learning_rate": 8.864955223223943e-05,
+      "loss": 1.8495,
+      "step": 2106
+    },
+    {
+      "epoch": 0.218863612755791,
+      "grad_norm": 0.3614397943019867,
+      "learning_rate": 8.863919866531269e-05,
+      "loss": 1.7565,
+      "step": 2107
+    },
+    {
+      "epoch": 0.21896748727537135,
+      "grad_norm": 0.3643103837966919,
+      "learning_rate": 8.862884098361271e-05,
+      "loss": 1.7943,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2190713617949517,
+      "grad_norm": 0.37451231479644775,
+      "learning_rate": 8.861847918824248e-05,
+      "loss": 1.5722,
+      "step": 2109
+    },
+    {
+      "epoch": 0.21917523631453203,
+      "grad_norm": 0.40897324681282043,
+      "learning_rate": 8.860811328030545e-05,
+      "loss": 1.8011,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2192791108341124,
+      "grad_norm": 0.38252517580986023,
+      "learning_rate": 8.85977432609055e-05,
+      "loss": 1.7692,
+      "step": 2111
+    },
+    {
+      "epoch": 0.21938298535369274,
+      "grad_norm": 0.3585284352302551,
+      "learning_rate": 8.858736913114698e-05,
+      "loss": 1.6858,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2194868598732731,
+      "grad_norm": 0.40061917901039124,
+      "learning_rate": 8.857699089213462e-05,
+      "loss": 1.5681,
+      "step": 2113
+    },
+    {
+      "epoch": 0.21959073439285343,
+      "grad_norm": 0.38758501410484314,
+      "learning_rate": 8.856660854497367e-05,
+      "loss": 1.7354,
+      "step": 2114
+    },
+    {
+      "epoch": 0.21969460891243378,
+      "grad_norm": 0.37637653946876526,
+      "learning_rate": 8.855622209076972e-05,
+      "loss": 1.6739,
+      "step": 2115
+    },
+    {
+      "epoch": 0.21979848343201414,
+      "grad_norm": 0.4087028205394745,
+      "learning_rate": 8.854583153062887e-05,
+      "loss": 2.0526,
+      "step": 2116
+    },
+    {
+      "epoch": 0.21990235795159446,
+      "grad_norm": 0.3950178623199463,
+      "learning_rate": 8.853543686565765e-05,
+      "loss": 1.8204,
+      "step": 2117
+    },
+    {
+      "epoch": 0.22000623247117482,
+      "grad_norm": 0.3853364586830139,
+      "learning_rate": 8.852503809696297e-05,
+      "loss": 1.7236,
+      "step": 2118
+    },
+    {
+      "epoch": 0.22011010699075517,
+      "grad_norm": 0.36730098724365234,
+      "learning_rate": 8.851463522565226e-05,
+      "loss": 1.7896,
+      "step": 2119
+    },
+    {
+      "epoch": 0.22021398151033553,
+      "grad_norm": 0.3799102008342743,
+      "learning_rate": 8.85042282528333e-05,
+      "loss": 1.8652,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22031785602991585,
+      "grad_norm": 0.37538158893585205,
+      "learning_rate": 8.84938171796144e-05,
+      "loss": 1.8323,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2204217305494962,
+      "grad_norm": 0.3702995777130127,
+      "learning_rate": 8.848340200710421e-05,
+      "loss": 1.6752,
+      "step": 2122
+    },
+    {
+      "epoch": 0.22052560506907656,
+      "grad_norm": 0.3841915726661682,
+      "learning_rate": 8.847298273641192e-05,
+      "loss": 1.8893,
+      "step": 2123
+    },
+    {
+      "epoch": 0.2206294795886569,
+      "grad_norm": 0.3636316955089569,
+      "learning_rate": 8.846255936864703e-05,
+      "loss": 1.6729,
+      "step": 2124
+    },
+    {
+      "epoch": 0.22073335410823725,
+      "grad_norm": 0.3771231770515442,
+      "learning_rate": 8.845213190491962e-05,
+      "loss": 1.6048,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2208372286278176,
+      "grad_norm": 0.4207725524902344,
+      "learning_rate": 8.844170034634007e-05,
+      "loss": 1.9401,
+      "step": 2126
+    },
+    {
+      "epoch": 0.22094110314739795,
+      "grad_norm": 0.42666593194007874,
+      "learning_rate": 8.843126469401931e-05,
+      "loss": 1.9608,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22104497766697828,
+      "grad_norm": 0.3714624345302582,
+      "learning_rate": 8.842082494906863e-05,
+      "loss": 1.7066,
+      "step": 2128
+    },
+    {
+      "epoch": 0.22114885218655864,
+      "grad_norm": 0.41800081729888916,
+      "learning_rate": 8.841038111259979e-05,
+      "loss": 1.8553,
+      "step": 2129
+    },
+    {
+      "epoch": 0.221252726706139,
+      "grad_norm": 0.34756773710250854,
+      "learning_rate": 8.839993318572497e-05,
+      "loss": 1.6992,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22135660122571932,
+      "grad_norm": 0.3820962905883789,
+      "learning_rate": 8.83894811695568e-05,
+      "loss": 1.9582,
+      "step": 2131
+    },
+    {
+      "epoch": 0.22146047574529967,
+      "grad_norm": 0.3578221797943115,
+      "learning_rate": 8.837902506520835e-05,
+      "loss": 1.7399,
+      "step": 2132
+    },
+    {
+      "epoch": 0.22156435026488003,
+      "grad_norm": 0.3778248727321625,
+      "learning_rate": 8.836856487379306e-05,
+      "loss": 1.6818,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22166822478446038,
+      "grad_norm": 0.36357828974723816,
+      "learning_rate": 8.835810059642493e-05,
+      "loss": 1.6495,
+      "step": 2134
+    },
+    {
+      "epoch": 0.2217720993040407,
+      "grad_norm": 0.384027898311615,
+      "learning_rate": 8.834763223421828e-05,
+      "loss": 1.7112,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22187597382362106,
+      "grad_norm": 0.3461948037147522,
+      "learning_rate": 8.833715978828793e-05,
+      "loss": 1.713,
+      "step": 2136
+    },
+    {
+      "epoch": 0.22197984834320142,
+      "grad_norm": 0.38593146204948425,
+      "learning_rate": 8.83266832597491e-05,
+      "loss": 1.6994,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22208372286278175,
+      "grad_norm": 0.346113383769989,
+      "learning_rate": 8.831620264971745e-05,
+      "loss": 1.6051,
+      "step": 2138
+    },
+    {
+      "epoch": 0.2221875973823621,
+      "grad_norm": 0.3933405876159668,
+      "learning_rate": 8.830571795930911e-05,
+      "loss": 1.678,
+      "step": 2139
+    },
+    {
+      "epoch": 0.22229147190194246,
+      "grad_norm": 0.34719663858413696,
+      "learning_rate": 8.829522918964058e-05,
+      "loss": 1.6881,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2223953464215228,
+      "grad_norm": 0.3760213553905487,
+      "learning_rate": 8.828473634182888e-05,
+      "loss": 1.8456,
+      "step": 2141
+    },
+    {
+      "epoch": 0.22249922094110314,
+      "grad_norm": 0.3988967537879944,
+      "learning_rate": 8.827423941699137e-05,
+      "loss": 1.8562,
+      "step": 2142
+    },
+    {
+      "epoch": 0.2226030954606835,
+      "grad_norm": 0.40549615025520325,
+      "learning_rate": 8.826373841624593e-05,
+      "loss": 1.7444,
+      "step": 2143
+    },
+    {
+      "epoch": 0.22270696998026385,
+      "grad_norm": 0.37362509965896606,
+      "learning_rate": 8.825323334071083e-05,
+      "loss": 1.7901,
+      "step": 2144
+    },
+    {
+      "epoch": 0.2228108444998442,
+      "grad_norm": 0.34235090017318726,
+      "learning_rate": 8.824272419150475e-05,
+      "loss": 1.7461,
+      "step": 2145
+    },
+    {
+      "epoch": 0.22291471901942453,
+      "grad_norm": 0.3918958306312561,
+      "learning_rate": 8.823221096974684e-05,
+      "loss": 1.782,
+      "step": 2146
+    },
+    {
+      "epoch": 0.22301859353900488,
+      "grad_norm": 0.37165430188179016,
+      "learning_rate": 8.822169367655669e-05,
+      "loss": 1.803,
+      "step": 2147
+    },
+    {
+      "epoch": 0.22312246805858524,
+      "grad_norm": 0.3669700026512146,
+      "learning_rate": 8.821117231305431e-05,
+      "loss": 1.8741,
+      "step": 2148
+    },
+    {
+      "epoch": 0.22322634257816557,
+      "grad_norm": 0.3532780408859253,
+      "learning_rate": 8.820064688036014e-05,
+      "loss": 1.7279,
+      "step": 2149
+    },
+    {
+      "epoch": 0.22333021709774592,
+      "grad_norm": 0.3830491006374359,
+      "learning_rate": 8.819011737959504e-05,
+      "loss": 1.8127,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22343409161732627,
+      "grad_norm": 0.36550477147102356,
+      "learning_rate": 8.817958381188033e-05,
+      "loss": 1.7985,
+      "step": 2151
+    },
+    {
+      "epoch": 0.22353796613690663,
+      "grad_norm": 0.3734526038169861,
+      "learning_rate": 8.816904617833778e-05,
+      "loss": 1.7096,
+      "step": 2152
+    },
+    {
+      "epoch": 0.22364184065648696,
+      "grad_norm": 0.36251428723335266,
+      "learning_rate": 8.815850448008953e-05,
+      "loss": 1.7936,
+      "step": 2153
+    },
+    {
+      "epoch": 0.2237457151760673,
+      "grad_norm": 0.3801165223121643,
+      "learning_rate": 8.81479587182582e-05,
+      "loss": 1.7445,
+      "step": 2154
+    },
+    {
+      "epoch": 0.22384958969564767,
+      "grad_norm": 0.38031628727912903,
+      "learning_rate": 8.813740889396682e-05,
+      "loss": 1.8105,
+      "step": 2155
+    },
+    {
+      "epoch": 0.223953464215228,
+      "grad_norm": 0.42004433274269104,
+      "learning_rate": 8.81268550083389e-05,
+      "loss": 1.7913,
+      "step": 2156
+    },
+    {
+      "epoch": 0.22405733873480835,
+      "grad_norm": 0.38428744673728943,
+      "learning_rate": 8.81162970624983e-05,
+      "loss": 1.805,
+      "step": 2157
+    },
+    {
+      "epoch": 0.2241612132543887,
+      "grad_norm": 0.4019632935523987,
+      "learning_rate": 8.81057350575694e-05,
+      "loss": 1.6997,
+      "step": 2158
+    },
+    {
+      "epoch": 0.22426508777396906,
+      "grad_norm": 0.3467335104942322,
+      "learning_rate": 8.809516899467694e-05,
+      "loss": 1.5612,
+      "step": 2159
+    },
+    {
+      "epoch": 0.22436896229354938,
+      "grad_norm": 0.36768898367881775,
+      "learning_rate": 8.808459887494615e-05,
+      "loss": 1.6928,
+      "step": 2160
+    },
+    {
+      "epoch": 0.22447283681312974,
+      "grad_norm": 0.41053932905197144,
+      "learning_rate": 8.807402469950263e-05,
+      "loss": 1.7605,
+      "step": 2161
+    },
+    {
+      "epoch": 0.2245767113327101,
+      "grad_norm": 0.36271244287490845,
+      "learning_rate": 8.806344646947249e-05,
+      "loss": 1.7773,
+      "step": 2162
+    },
+    {
+      "epoch": 0.22468058585229042,
+      "grad_norm": 0.38008955121040344,
+      "learning_rate": 8.80528641859822e-05,
+      "loss": 1.8473,
+      "step": 2163
+    },
+    {
+      "epoch": 0.22478446037187078,
+      "grad_norm": 0.382253497838974,
+      "learning_rate": 8.804227785015869e-05,
+      "loss": 1.686,
+      "step": 2164
+    },
+    {
+      "epoch": 0.22488833489145113,
+      "grad_norm": 0.3781398832798004,
+      "learning_rate": 8.803168746312934e-05,
+      "loss": 1.7894,
+      "step": 2165
+    },
+    {
+      "epoch": 0.22499220941103149,
+      "grad_norm": 0.39482182264328003,
+      "learning_rate": 8.802109302602193e-05,
+      "loss": 1.8895,
+      "step": 2166
+    },
+    {
+      "epoch": 0.2250960839306118,
+      "grad_norm": 0.40680569410324097,
+      "learning_rate": 8.801049453996468e-05,
+      "loss": 2.0375,
+      "step": 2167
+    },
+    {
+      "epoch": 0.22519995845019217,
+      "grad_norm": 0.3483993113040924,
+      "learning_rate": 8.799989200608627e-05,
+      "loss": 1.6489,
+      "step": 2168
+    },
+    {
+      "epoch": 0.22530383296977252,
+      "grad_norm": 0.3627791702747345,
+      "learning_rate": 8.798928542551576e-05,
+      "loss": 1.8188,
+      "step": 2169
+    },
+    {
+      "epoch": 0.22540770748935285,
+      "grad_norm": 0.3764524459838867,
+      "learning_rate": 8.797867479938269e-05,
+      "loss": 1.9716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2255115820089332,
+      "grad_norm": 0.3761730492115021,
+      "learning_rate": 8.796806012881699e-05,
+      "loss": 1.857,
+      "step": 2171
+    },
+    {
+      "epoch": 0.22561545652851356,
+      "grad_norm": 0.3673245906829834,
+      "learning_rate": 8.795744141494905e-05,
+      "loss": 1.6663,
+      "step": 2172
+    },
+    {
+      "epoch": 0.2257193310480939,
+      "grad_norm": 0.3899873197078705,
+      "learning_rate": 8.794681865890968e-05,
+      "loss": 1.7631,
+      "step": 2173
+    },
+    {
+      "epoch": 0.22582320556767424,
+      "grad_norm": 0.36863207817077637,
+      "learning_rate": 8.793619186183011e-05,
+      "loss": 1.8269,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2259270800872546,
+      "grad_norm": 0.36547142267227173,
+      "learning_rate": 8.792556102484204e-05,
+      "loss": 1.6963,
+      "step": 2175
+    },
+    {
+      "epoch": 0.22603095460683495,
+      "grad_norm": 0.34882575273513794,
+      "learning_rate": 8.791492614907754e-05,
+      "loss": 1.6336,
+      "step": 2176
+    },
+    {
+      "epoch": 0.22613482912641528,
+      "grad_norm": 0.38748350739479065,
+      "learning_rate": 8.790428723566915e-05,
+      "loss": 1.8569,
+      "step": 2177
+    },
+    {
+      "epoch": 0.22623870364599563,
+      "grad_norm": 0.362943559885025,
+      "learning_rate": 8.789364428574984e-05,
+      "loss": 1.7161,
+      "step": 2178
+    },
+    {
+      "epoch": 0.22634257816557599,
+      "grad_norm": 0.37946629524230957,
+      "learning_rate": 8.788299730045299e-05,
+      "loss": 1.8211,
+      "step": 2179
+    },
+    {
+      "epoch": 0.22644645268515634,
+      "grad_norm": 0.3608231842517853,
+      "learning_rate": 8.787234628091243e-05,
+      "loss": 1.7704,
+      "step": 2180
+    },
+    {
+      "epoch": 0.22655032720473667,
+      "grad_norm": 0.38280248641967773,
+      "learning_rate": 8.786169122826242e-05,
+      "loss": 1.7376,
+      "step": 2181
+    },
+    {
+      "epoch": 0.22665420172431702,
+      "grad_norm": 0.36834195256233215,
+      "learning_rate": 8.785103214363761e-05,
+      "loss": 1.6503,
+      "step": 2182
+    },
+    {
+      "epoch": 0.22675807624389738,
+      "grad_norm": 0.3779491186141968,
+      "learning_rate": 8.784036902817312e-05,
+      "loss": 1.7421,
+      "step": 2183
+    },
+    {
+      "epoch": 0.22686195076347773,
+      "grad_norm": 0.40714192390441895,
+      "learning_rate": 8.78297018830045e-05,
+      "loss": 1.7828,
+      "step": 2184
+    },
+    {
+      "epoch": 0.22696582528305806,
+      "grad_norm": 0.36981451511383057,
+      "learning_rate": 8.781903070926774e-05,
+      "loss": 1.7687,
+      "step": 2185
+    },
+    {
+      "epoch": 0.22706969980263841,
+      "grad_norm": 0.37374863028526306,
+      "learning_rate": 8.780835550809918e-05,
+      "loss": 1.7469,
+      "step": 2186
+    },
+    {
+      "epoch": 0.22717357432221877,
+      "grad_norm": 0.3481188416481018,
+      "learning_rate": 8.779767628063567e-05,
+      "loss": 1.5268,
+      "step": 2187
+    },
+    {
+      "epoch": 0.2272774488417991,
+      "grad_norm": 0.352603942155838,
+      "learning_rate": 8.77869930280145e-05,
+      "loss": 1.5966,
+      "step": 2188
+    },
+    {
+      "epoch": 0.22738132336137945,
+      "grad_norm": 0.38702160120010376,
+      "learning_rate": 8.77763057513733e-05,
+      "loss": 1.7469,
+      "step": 2189
+    },
+    {
+      "epoch": 0.2274851978809598,
+      "grad_norm": 0.37263423204421997,
+      "learning_rate": 8.77656144518502e-05,
+      "loss": 1.6986,
+      "step": 2190
+    },
+    {
+      "epoch": 0.22758907240054016,
+      "grad_norm": 0.3658749759197235,
+      "learning_rate": 8.775491913058376e-05,
+      "loss": 1.7425,
+      "step": 2191
+    },
+    {
+      "epoch": 0.2276929469201205,
+      "grad_norm": 0.38823366165161133,
+      "learning_rate": 8.774421978871292e-05,
+      "loss": 1.7665,
+      "step": 2192
+    },
+    {
+      "epoch": 0.22779682143970084,
+      "grad_norm": 0.35970309376716614,
+      "learning_rate": 8.773351642737708e-05,
+      "loss": 1.6914,
+      "step": 2193
+    },
+    {
+      "epoch": 0.2279006959592812,
+      "grad_norm": 0.3693389892578125,
+      "learning_rate": 8.772280904771608e-05,
+      "loss": 1.8031,
+      "step": 2194
+    },
+    {
+      "epoch": 0.22800457047886152,
+      "grad_norm": 0.34769538044929504,
+      "learning_rate": 8.771209765087018e-05,
+      "loss": 1.7248,
+      "step": 2195
+    },
+    {
+      "epoch": 0.22810844499844188,
+      "grad_norm": 0.37290412187576294,
+      "learning_rate": 8.770138223798003e-05,
+      "loss": 1.7994,
+      "step": 2196
+    },
+    {
+      "epoch": 0.22821231951802223,
+      "grad_norm": 0.37960392236709595,
+      "learning_rate": 8.769066281018676e-05,
+      "loss": 1.8668,
+      "step": 2197
+    },
+    {
+      "epoch": 0.2283161940376026,
+      "grad_norm": 0.3551698625087738,
+      "learning_rate": 8.76799393686319e-05,
+      "loss": 1.7928,
+      "step": 2198
+    },
+    {
+      "epoch": 0.22842006855718291,
+      "grad_norm": 0.3406142294406891,
+      "learning_rate": 8.766921191445742e-05,
+      "loss": 1.6981,
+      "step": 2199
+    },
+    {
+      "epoch": 0.22852394307676327,
+      "grad_norm": 0.410206139087677,
+      "learning_rate": 8.765848044880569e-05,
+      "loss": 1.634,
+      "step": 2200
+    },
+    {
+      "epoch": 0.22862781759634362,
+      "grad_norm": 0.42800506949424744,
+      "learning_rate": 8.764774497281954e-05,
+      "loss": 1.7564,
+      "step": 2201
+    },
+    {
+      "epoch": 0.22873169211592395,
+      "grad_norm": 0.39173564314842224,
+      "learning_rate": 8.763700548764223e-05,
+      "loss": 1.8504,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2288355666355043,
+      "grad_norm": 0.35888996720314026,
+      "learning_rate": 8.76262619944174e-05,
+      "loss": 1.5756,
+      "step": 2203
+    },
+    {
+      "epoch": 0.22893944115508466,
+      "grad_norm": 0.4852285087108612,
+      "learning_rate": 8.761551449428918e-05,
+      "loss": 1.9804,
+      "step": 2204
+    },
+    {
+      "epoch": 0.22904331567466502,
+      "grad_norm": 0.4047568440437317,
+      "learning_rate": 8.760476298840206e-05,
+      "loss": 1.81,
+      "step": 2205
+    },
+    {
+      "epoch": 0.22914719019424534,
+      "grad_norm": 0.3965972065925598,
+      "learning_rate": 8.7594007477901e-05,
+      "loss": 1.8159,
+      "step": 2206
+    },
+    {
+      "epoch": 0.2292510647138257,
+      "grad_norm": 0.4031708538532257,
+      "learning_rate": 8.758324796393142e-05,
+      "loss": 1.8237,
+      "step": 2207
+    },
+    {
+      "epoch": 0.22935493923340605,
+      "grad_norm": 0.3491591215133667,
+      "learning_rate": 8.757248444763908e-05,
+      "loss": 1.6837,
+      "step": 2208
+    },
+    {
+      "epoch": 0.22945881375298638,
+      "grad_norm": 0.36505648493766785,
+      "learning_rate": 8.756171693017022e-05,
+      "loss": 1.6408,
+      "step": 2209
+    },
+    {
+      "epoch": 0.22956268827256673,
+      "grad_norm": 0.394775927066803,
+      "learning_rate": 8.755094541267149e-05,
+      "loss": 1.8816,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2296665627921471,
+      "grad_norm": 0.4164559543132782,
+      "learning_rate": 8.754016989629e-05,
+      "loss": 1.9372,
+      "step": 2211
+    },
+    {
+      "epoch": 0.22977043731172744,
+      "grad_norm": 0.3899494707584381,
+      "learning_rate": 8.752939038217322e-05,
+      "loss": 1.7356,
+      "step": 2212
+    },
+    {
+      "epoch": 0.22987431183130777,
+      "grad_norm": 0.36787334084510803,
+      "learning_rate": 8.751860687146912e-05,
+      "loss": 1.7804,
+      "step": 2213
+    },
+    {
+      "epoch": 0.22997818635088813,
+      "grad_norm": 0.3581325113773346,
+      "learning_rate": 8.750781936532604e-05,
+      "loss": 1.825,
+      "step": 2214
+    },
+    {
+      "epoch": 0.23008206087046848,
+      "grad_norm": 0.3675483465194702,
+      "learning_rate": 8.749702786489277e-05,
+      "loss": 1.7208,
+      "step": 2215
+    },
+    {
+      "epoch": 0.23018593539004883,
+      "grad_norm": 0.38968801498413086,
+      "learning_rate": 8.748623237131853e-05,
+      "loss": 1.8447,
+      "step": 2216
+    },
+    {
+      "epoch": 0.23028980990962916,
+      "grad_norm": 0.3647545576095581,
+      "learning_rate": 8.747543288575293e-05,
+      "loss": 1.6169,
+      "step": 2217
+    },
+    {
+      "epoch": 0.23039368442920952,
+      "grad_norm": 0.3703446388244629,
+      "learning_rate": 8.746462940934605e-05,
+      "loss": 1.6316,
+      "step": 2218
+    },
+    {
+      "epoch": 0.23049755894878987,
+      "grad_norm": 0.3884913921356201,
+      "learning_rate": 8.745382194324839e-05,
+      "loss": 1.7359,
+      "step": 2219
+    },
+    {
+      "epoch": 0.2306014334683702,
+      "grad_norm": 0.39596831798553467,
+      "learning_rate": 8.744301048861083e-05,
+      "loss": 1.7425,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23070530798795055,
+      "grad_norm": 0.41258224844932556,
+      "learning_rate": 8.743219504658472e-05,
+      "loss": 1.8832,
+      "step": 2221
+    },
+    {
+      "epoch": 0.2308091825075309,
+      "grad_norm": 0.3803333342075348,
+      "learning_rate": 8.742137561832182e-05,
+      "loss": 1.8591,
+      "step": 2222
+    },
+    {
+      "epoch": 0.23091305702711126,
+      "grad_norm": 0.3818683326244354,
+      "learning_rate": 8.741055220497431e-05,
+      "loss": 1.9393,
+      "step": 2223
+    },
+    {
+      "epoch": 0.2310169315466916,
+      "grad_norm": 0.3759542405605316,
+      "learning_rate": 8.739972480769481e-05,
+      "loss": 1.8439,
+      "step": 2224
+    },
+    {
+      "epoch": 0.23112080606627194,
+      "grad_norm": 0.3411445617675781,
+      "learning_rate": 8.738889342763635e-05,
+      "loss": 1.7198,
+      "step": 2225
+    },
+    {
+      "epoch": 0.2312246805858523,
+      "grad_norm": 0.3545193076133728,
+      "learning_rate": 8.737805806595239e-05,
+      "loss": 1.6465,
+      "step": 2226
+    },
+    {
+      "epoch": 0.23132855510543263,
+      "grad_norm": 0.3674876093864441,
+      "learning_rate": 8.73672187237968e-05,
+      "loss": 1.7859,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23143242962501298,
+      "grad_norm": 0.3644886612892151,
+      "learning_rate": 8.735637540232389e-05,
+      "loss": 1.7008,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23153630414459334,
+      "grad_norm": 0.3750406503677368,
+      "learning_rate": 8.734552810268838e-05,
+      "loss": 1.6847,
+      "step": 2229
+    },
+    {
+      "epoch": 0.2316401786641737,
+      "grad_norm": 0.3923105299472809,
+      "learning_rate": 8.733467682604545e-05,
+      "loss": 1.7455,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23174405318375402,
+      "grad_norm": 0.3961966931819916,
+      "learning_rate": 8.732382157355066e-05,
+      "loss": 1.6726,
+      "step": 2231
+    },
+    {
+      "epoch": 0.23184792770333437,
+      "grad_norm": 0.4062844514846802,
+      "learning_rate": 8.731296234636e-05,
+      "loss": 1.9202,
+      "step": 2232
+    },
+    {
+      "epoch": 0.23195180222291473,
+      "grad_norm": 0.4073842167854309,
+      "learning_rate": 8.73020991456299e-05,
+      "loss": 1.9388,
+      "step": 2233
+    },
+    {
+      "epoch": 0.23205567674249505,
+      "grad_norm": 0.38866791129112244,
+      "learning_rate": 8.72912319725172e-05,
+      "loss": 1.6893,
+      "step": 2234
+    },
+    {
+      "epoch": 0.2321595512620754,
+      "grad_norm": 0.3830127716064453,
+      "learning_rate": 8.72803608281792e-05,
+      "loss": 1.812,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23226342578165576,
+      "grad_norm": 0.4291350245475769,
+      "learning_rate": 8.726948571377356e-05,
+      "loss": 1.8426,
+      "step": 2236
+    },
+    {
+      "epoch": 0.23236730030123612,
+      "grad_norm": 0.36664703488349915,
+      "learning_rate": 8.72586066304584e-05,
+      "loss": 1.6714,
+      "step": 2237
+    },
+    {
+      "epoch": 0.23247117482081645,
+      "grad_norm": 0.34283211827278137,
+      "learning_rate": 8.724772357939229e-05,
+      "loss": 1.7595,
+      "step": 2238
+    },
+    {
+      "epoch": 0.2325750493403968,
+      "grad_norm": 0.3679027557373047,
+      "learning_rate": 8.723683656173413e-05,
+      "loss": 1.6493,
+      "step": 2239
+    },
+    {
+      "epoch": 0.23267892385997715,
+      "grad_norm": 0.3953079879283905,
+      "learning_rate": 8.722594557864335e-05,
+      "loss": 1.7885,
+      "step": 2240
+    },
+    {
+      "epoch": 0.23278279837955748,
+      "grad_norm": 0.37095022201538086,
+      "learning_rate": 8.721505063127972e-05,
+      "loss": 1.6947,
+      "step": 2241
+    },
+    {
+      "epoch": 0.23288667289913784,
+      "grad_norm": 0.38997137546539307,
+      "learning_rate": 8.72041517208035e-05,
+      "loss": 1.8158,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2329905474187182,
+      "grad_norm": 0.36092621088027954,
+      "learning_rate": 8.719324884837531e-05,
+      "loss": 1.7928,
+      "step": 2243
+    },
+    {
+      "epoch": 0.23309442193829855,
+      "grad_norm": 0.3828462064266205,
+      "learning_rate": 8.718234201515627e-05,
+      "loss": 1.7782,
+      "step": 2244
+    },
+    {
+      "epoch": 0.23319829645787887,
+      "grad_norm": 0.35145506262779236,
+      "learning_rate": 8.717143122230782e-05,
+      "loss": 1.6527,
+      "step": 2245
+    },
+    {
+      "epoch": 0.23330217097745923,
+      "grad_norm": 0.3912096321582794,
+      "learning_rate": 8.71605164709919e-05,
+      "loss": 1.8666,
+      "step": 2246
+    },
+    {
+      "epoch": 0.23340604549703958,
+      "grad_norm": 0.3545803725719452,
+      "learning_rate": 8.714959776237083e-05,
+      "loss": 1.7043,
+      "step": 2247
+    },
+    {
+      "epoch": 0.2335099200166199,
+      "grad_norm": 0.3926384449005127,
+      "learning_rate": 8.713867509760738e-05,
+      "loss": 2.007,
+      "step": 2248
+    },
+    {
+      "epoch": 0.23361379453620026,
+      "grad_norm": 0.3713902533054352,
+      "learning_rate": 8.712774847786471e-05,
+      "loss": 1.7151,
+      "step": 2249
+    },
+    {
+      "epoch": 0.23371766905578062,
+      "grad_norm": 0.39191189408302307,
+      "learning_rate": 8.711681790430645e-05,
+      "loss": 1.8144,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23382154357536097,
+      "grad_norm": 0.40152493119239807,
+      "learning_rate": 8.710588337809662e-05,
+      "loss": 1.9487,
+      "step": 2251
+    },
+    {
+      "epoch": 0.2339254180949413,
+      "grad_norm": 0.3732980489730835,
+      "learning_rate": 8.709494490039963e-05,
+      "loss": 1.6245,
+      "step": 2252
+    },
+    {
+      "epoch": 0.23402929261452166,
+      "grad_norm": 0.3951932191848755,
+      "learning_rate": 8.708400247238035e-05,
+      "loss": 1.8078,
+      "step": 2253
+    },
+    {
+      "epoch": 0.234133167134102,
+      "grad_norm": 0.3723011910915375,
+      "learning_rate": 8.707305609520408e-05,
+      "loss": 1.7107,
+      "step": 2254
+    },
+    {
+      "epoch": 0.23423704165368237,
+      "grad_norm": 0.3788382411003113,
+      "learning_rate": 8.706210577003653e-05,
+      "loss": 1.6238,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2343409161732627,
+      "grad_norm": 0.3792799115180969,
+      "learning_rate": 8.705115149804381e-05,
+      "loss": 1.7069,
+      "step": 2256
+    },
+    {
+      "epoch": 0.23444479069284305,
+      "grad_norm": 0.3782130181789398,
+      "learning_rate": 8.704019328039244e-05,
+      "loss": 1.6198,
+      "step": 2257
+    },
+    {
+      "epoch": 0.2345486652124234,
+      "grad_norm": 0.3798910975456238,
+      "learning_rate": 8.702923111824943e-05,
+      "loss": 1.8176,
+      "step": 2258
+    },
+    {
+      "epoch": 0.23465253973200373,
+      "grad_norm": 0.38398462533950806,
+      "learning_rate": 8.701826501278216e-05,
+      "loss": 1.7607,
+      "step": 2259
+    },
+    {
+      "epoch": 0.23475641425158408,
+      "grad_norm": 0.3783421814441681,
+      "learning_rate": 8.70072949651584e-05,
+      "loss": 1.765,
+      "step": 2260
+    },
+    {
+      "epoch": 0.23486028877116444,
+      "grad_norm": 0.34123924374580383,
+      "learning_rate": 8.69963209765464e-05,
+      "loss": 1.6595,
+      "step": 2261
+    },
+    {
+      "epoch": 0.2349641632907448,
+      "grad_norm": 0.33665284514427185,
+      "learning_rate": 8.698534304811478e-05,
+      "loss": 1.4557,
+      "step": 2262
+    },
+    {
+      "epoch": 0.23506803781032512,
+      "grad_norm": 0.37269553542137146,
+      "learning_rate": 8.697436118103264e-05,
+      "loss": 1.8497,
+      "step": 2263
+    },
+    {
+      "epoch": 0.23517191232990547,
+      "grad_norm": 0.38616353273391724,
+      "learning_rate": 8.696337537646944e-05,
+      "loss": 1.7943,
+      "step": 2264
+    },
+    {
+      "epoch": 0.23527578684948583,
+      "grad_norm": 0.4167322814464569,
+      "learning_rate": 8.695238563559509e-05,
+      "loss": 1.9579,
+      "step": 2265
+    },
+    {
+      "epoch": 0.23537966136906616,
+      "grad_norm": 0.38582777976989746,
+      "learning_rate": 8.694139195957991e-05,
+      "loss": 1.7782,
+      "step": 2266
+    },
+    {
+      "epoch": 0.2354835358886465,
+      "grad_norm": 0.3635927438735962,
+      "learning_rate": 8.693039434959464e-05,
+      "loss": 1.7276,
+      "step": 2267
+    },
+    {
+      "epoch": 0.23558741040822687,
+      "grad_norm": 0.42506465315818787,
+      "learning_rate": 8.691939280681045e-05,
+      "loss": 1.818,
+      "step": 2268
+    },
+    {
+      "epoch": 0.23569128492780722,
+      "grad_norm": 0.38665080070495605,
+      "learning_rate": 8.69083873323989e-05,
+      "loss": 1.7603,
+      "step": 2269
+    },
+    {
+      "epoch": 0.23579515944738755,
+      "grad_norm": 0.3853725790977478,
+      "learning_rate": 8.689737792753198e-05,
+      "loss": 1.6039,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2358990339669679,
+      "grad_norm": 0.37736237049102783,
+      "learning_rate": 8.688636459338215e-05,
+      "loss": 1.8416,
+      "step": 2271
+    },
+    {
+      "epoch": 0.23600290848654826,
+      "grad_norm": 0.4537546634674072,
+      "learning_rate": 8.68753473311222e-05,
+      "loss": 1.6987,
+      "step": 2272
+    },
+    {
+      "epoch": 0.23610678300612858,
+      "grad_norm": 0.3903726041316986,
+      "learning_rate": 8.686432614192538e-05,
+      "loss": 1.9472,
+      "step": 2273
+    },
+    {
+      "epoch": 0.23621065752570894,
+      "grad_norm": 0.41961470246315,
+      "learning_rate": 8.68533010269654e-05,
+      "loss": 1.8414,
+      "step": 2274
+    },
+    {
+      "epoch": 0.2363145320452893,
+      "grad_norm": 0.36322250962257385,
+      "learning_rate": 8.684227198741633e-05,
+      "loss": 1.5978,
+      "step": 2275
+    },
+    {
+      "epoch": 0.23641840656486965,
+      "grad_norm": 0.4004223644733429,
+      "learning_rate": 8.683123902445267e-05,
+      "loss": 1.7981,
+      "step": 2276
+    },
+    {
+      "epoch": 0.23652228108444998,
+      "grad_norm": 0.38842740654945374,
+      "learning_rate": 8.682020213924935e-05,
+      "loss": 1.7098,
+      "step": 2277
+    },
+    {
+      "epoch": 0.23662615560403033,
+      "grad_norm": 0.3592800199985504,
+      "learning_rate": 8.680916133298171e-05,
+      "loss": 1.661,
+      "step": 2278
+    },
+    {
+      "epoch": 0.23673003012361069,
+      "grad_norm": 0.357117623090744,
+      "learning_rate": 8.67981166068255e-05,
+      "loss": 1.5888,
+      "step": 2279
+    },
+    {
+      "epoch": 0.236833904643191,
+      "grad_norm": 0.3849972188472748,
+      "learning_rate": 8.678706796195692e-05,
+      "loss": 1.7733,
+      "step": 2280
+    },
+    {
+      "epoch": 0.23693777916277137,
+      "grad_norm": 0.36867091059684753,
+      "learning_rate": 8.677601539955256e-05,
+      "loss": 1.8096,
+      "step": 2281
+    },
+    {
+      "epoch": 0.23704165368235172,
+      "grad_norm": 0.3629715144634247,
+      "learning_rate": 8.676495892078941e-05,
+      "loss": 1.6919,
+      "step": 2282
+    },
+    {
+      "epoch": 0.23714552820193208,
+      "grad_norm": 0.4076031446456909,
+      "learning_rate": 8.675389852684492e-05,
+      "loss": 1.9165,
+      "step": 2283
+    },
+    {
+      "epoch": 0.2372494027215124,
+      "grad_norm": 0.37175339460372925,
+      "learning_rate": 8.674283421889691e-05,
+      "loss": 1.6767,
+      "step": 2284
+    },
+    {
+      "epoch": 0.23735327724109276,
+      "grad_norm": 0.3934227526187897,
+      "learning_rate": 8.673176599812368e-05,
+      "loss": 1.7891,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2374571517606731,
+      "grad_norm": 0.4014788269996643,
+      "learning_rate": 8.672069386570389e-05,
+      "loss": 1.9999,
+      "step": 2286
+    },
+    {
+      "epoch": 0.23756102628025344,
+      "grad_norm": 0.35122254490852356,
+      "learning_rate": 8.670961782281664e-05,
+      "loss": 1.8249,
+      "step": 2287
+    },
+    {
+      "epoch": 0.2376649007998338,
+      "grad_norm": 0.36434510350227356,
+      "learning_rate": 8.669853787064142e-05,
+      "loss": 1.8532,
+      "step": 2288
+    },
+    {
+      "epoch": 0.23776877531941415,
+      "grad_norm": 0.3808654546737671,
+      "learning_rate": 8.668745401035818e-05,
+      "loss": 1.7474,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2378726498389945,
+      "grad_norm": 0.3531155288219452,
+      "learning_rate": 8.667636624314725e-05,
+      "loss": 1.8941,
+      "step": 2290
+    },
+    {
+      "epoch": 0.23797652435857483,
+      "grad_norm": 0.472049742937088,
+      "learning_rate": 8.666527457018943e-05,
+      "loss": 1.985,
+      "step": 2291
+    },
+    {
+      "epoch": 0.23808039887815519,
+      "grad_norm": 0.39368936419487,
+      "learning_rate": 8.665417899266586e-05,
+      "loss": 1.6889,
+      "step": 2292
+    },
+    {
+      "epoch": 0.23818427339773554,
+      "grad_norm": 0.3707980811595917,
+      "learning_rate": 8.664307951175814e-05,
+      "loss": 1.5088,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2382881479173159,
+      "grad_norm": 0.3548358380794525,
+      "learning_rate": 8.663197612864827e-05,
+      "loss": 1.7288,
+      "step": 2294
+    },
+    {
+      "epoch": 0.23839202243689622,
+      "grad_norm": 0.47809383273124695,
+      "learning_rate": 8.662086884451869e-05,
+      "loss": 1.8472,
+      "step": 2295
+    },
+    {
+      "epoch": 0.23849589695647658,
+      "grad_norm": 0.4353974163532257,
+      "learning_rate": 8.660975766055224e-05,
+      "loss": 1.8897,
+      "step": 2296
+    },
+    {
+      "epoch": 0.23859977147605693,
+      "grad_norm": 0.4466110169887543,
+      "learning_rate": 8.659864257793215e-05,
+      "loss": 1.9391,
+      "step": 2297
+    },
+    {
+      "epoch": 0.23870364599563726,
+      "grad_norm": 0.39143791794776917,
+      "learning_rate": 8.65875235978421e-05,
+      "loss": 1.7464,
+      "step": 2298
+    },
+    {
+      "epoch": 0.23880752051521761,
+      "grad_norm": 0.38728442788124084,
+      "learning_rate": 8.65764007214662e-05,
+      "loss": 1.7393,
+      "step": 2299
+    },
+    {
+      "epoch": 0.23891139503479797,
+      "grad_norm": 0.3674345910549164,
+      "learning_rate": 8.656527394998892e-05,
+      "loss": 1.6336,
+      "step": 2300
+    },
+    {
+      "epoch": 0.23901526955437832,
+      "grad_norm": 0.38445186614990234,
+      "learning_rate": 8.655414328459519e-05,
+      "loss": 1.8982,
+      "step": 2301
+    },
+    {
+      "epoch": 0.23911914407395865,
+      "grad_norm": 0.38056597113609314,
+      "learning_rate": 8.654300872647033e-05,
+      "loss": 1.6851,
+      "step": 2302
+    },
+    {
+      "epoch": 0.239223018593539,
+      "grad_norm": 0.4110845625400543,
+      "learning_rate": 8.65318702768001e-05,
+      "loss": 1.8767,
+      "step": 2303
+    },
+    {
+      "epoch": 0.23932689311311936,
+      "grad_norm": 0.38014286756515503,
+      "learning_rate": 8.652072793677061e-05,
+      "loss": 1.725,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2394307676326997,
+      "grad_norm": 0.39622190594673157,
+      "learning_rate": 8.650958170756852e-05,
+      "loss": 1.9115,
+      "step": 2305
+    },
+    {
+      "epoch": 0.23953464215228004,
+      "grad_norm": 0.36890217661857605,
+      "learning_rate": 8.649843159038071e-05,
+      "loss": 1.698,
+      "step": 2306
+    },
+    {
+      "epoch": 0.2396385166718604,
+      "grad_norm": 0.38787323236465454,
+      "learning_rate": 8.648727758639467e-05,
+      "loss": 1.8998,
+      "step": 2307
+    },
+    {
+      "epoch": 0.23974239119144075,
+      "grad_norm": 0.38866347074508667,
+      "learning_rate": 8.647611969679816e-05,
+      "loss": 1.7095,
+      "step": 2308
+    },
+    {
+      "epoch": 0.23984626571102108,
+      "grad_norm": 0.39878830313682556,
+      "learning_rate": 8.646495792277943e-05,
+      "loss": 1.9387,
+      "step": 2309
+    },
+    {
+      "epoch": 0.23995014023060143,
+      "grad_norm": 0.3923484981060028,
+      "learning_rate": 8.645379226552712e-05,
+      "loss": 1.8302,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2400540147501818,
+      "grad_norm": 0.3727096617221832,
+      "learning_rate": 8.644262272623029e-05,
+      "loss": 1.6717,
+      "step": 2311
+    },
+    {
+      "epoch": 0.24015788926976211,
+      "grad_norm": 0.3535787761211395,
+      "learning_rate": 8.64314493060784e-05,
+      "loss": 1.6446,
+      "step": 2312
+    },
+    {
+      "epoch": 0.24026176378934247,
+      "grad_norm": 0.3555033504962921,
+      "learning_rate": 8.642027200626135e-05,
+      "loss": 1.7145,
+      "step": 2313
+    },
+    {
+      "epoch": 0.24036563830892282,
+      "grad_norm": 0.3748113512992859,
+      "learning_rate": 8.640909082796939e-05,
+      "loss": 1.7144,
+      "step": 2314
+    },
+    {
+      "epoch": 0.24046951282850318,
+      "grad_norm": 0.37921851873397827,
+      "learning_rate": 8.639790577239328e-05,
+      "loss": 1.8516,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2405733873480835,
+      "grad_norm": 0.36275529861450195,
+      "learning_rate": 8.638671684072412e-05,
+      "loss": 1.6144,
+      "step": 2316
+    },
+    {
+      "epoch": 0.24067726186766386,
+      "grad_norm": 0.35582250356674194,
+      "learning_rate": 8.637552403415343e-05,
+      "loss": 1.7601,
+      "step": 2317
+    },
+    {
+      "epoch": 0.24078113638724422,
+      "grad_norm": 0.3954647183418274,
+      "learning_rate": 8.636432735387319e-05,
+      "loss": 1.798,
+      "step": 2318
+    },
+    {
+      "epoch": 0.24088501090682454,
+      "grad_norm": 0.36659133434295654,
+      "learning_rate": 8.635312680107572e-05,
+      "loss": 1.551,
+      "step": 2319
+    },
+    {
+      "epoch": 0.2409888854264049,
+      "grad_norm": 0.35685068368911743,
+      "learning_rate": 8.634192237695382e-05,
+      "loss": 1.6926,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24109275994598525,
+      "grad_norm": 0.35937169194221497,
+      "learning_rate": 8.633071408270065e-05,
+      "loss": 1.604,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2411966344655656,
+      "grad_norm": 0.44768527150154114,
+      "learning_rate": 8.631950191950983e-05,
+      "loss": 2.0036,
+      "step": 2322
+    },
+    {
+      "epoch": 0.24130050898514593,
+      "grad_norm": 0.37177640199661255,
+      "learning_rate": 8.630828588857537e-05,
+      "loss": 1.7145,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2414043835047263,
+      "grad_norm": 0.38756901025772095,
+      "learning_rate": 8.629706599109169e-05,
+      "loss": 1.7844,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24150825802430664,
+      "grad_norm": 0.37525761127471924,
+      "learning_rate": 8.628584222825357e-05,
+      "loss": 1.586,
+      "step": 2325
+    },
+    {
+      "epoch": 0.241612132543887,
+      "grad_norm": 0.4448417127132416,
+      "learning_rate": 8.627461460125632e-05,
+      "loss": 1.6915,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24171600706346733,
+      "grad_norm": 0.39078620076179504,
+      "learning_rate": 8.626338311129557e-05,
+      "loss": 1.7567,
+      "step": 2327
+    },
+    {
+      "epoch": 0.24181988158304768,
+      "grad_norm": 0.38780951499938965,
+      "learning_rate": 8.625214775956737e-05,
+      "loss": 1.6969,
+      "step": 2328
+    },
+    {
+      "epoch": 0.24192375610262803,
+      "grad_norm": 0.3700442314147949,
+      "learning_rate": 8.624090854726822e-05,
+      "loss": 1.6198,
+      "step": 2329
+    },
+    {
+      "epoch": 0.24202763062220836,
+      "grad_norm": 0.37854552268981934,
+      "learning_rate": 8.622966547559499e-05,
+      "loss": 1.7932,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24213150514178872,
+      "grad_norm": 0.49389779567718506,
+      "learning_rate": 8.621841854574501e-05,
+      "loss": 1.8561,
+      "step": 2331
+    },
+    {
+      "epoch": 0.24223537966136907,
+      "grad_norm": 0.35576608777046204,
+      "learning_rate": 8.620716775891595e-05,
+      "loss": 1.5224,
+      "step": 2332
+    },
+    {
+      "epoch": 0.24233925418094943,
+      "grad_norm": 0.37670719623565674,
+      "learning_rate": 8.619591311630595e-05,
+      "loss": 1.8392,
+      "step": 2333
+    },
+    {
+      "epoch": 0.24244312870052975,
+      "grad_norm": 0.3652971386909485,
+      "learning_rate": 8.618465461911355e-05,
+      "loss": 1.7624,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2425470032201101,
+      "grad_norm": 0.3692745268344879,
+      "learning_rate": 8.617339226853768e-05,
+      "loss": 1.6901,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24265087773969046,
+      "grad_norm": 0.36297428607940674,
+      "learning_rate": 8.61621260657777e-05,
+      "loss": 1.8461,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2427547522592708,
+      "grad_norm": 0.37688395380973816,
+      "learning_rate": 8.615085601203337e-05,
+      "loss": 1.6562,
+      "step": 2337
+    },
+    {
+      "epoch": 0.24285862677885114,
+      "grad_norm": 0.3952527642250061,
+      "learning_rate": 8.613958210850485e-05,
+      "loss": 1.7862,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2429625012984315,
+      "grad_norm": 0.35479936003685,
+      "learning_rate": 8.612830435639275e-05,
+      "loss": 1.9233,
+      "step": 2339
+    },
+    {
+      "epoch": 0.24306637581801185,
+      "grad_norm": 0.40649887919425964,
+      "learning_rate": 8.611702275689805e-05,
+      "loss": 1.8637,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24317025033759218,
+      "grad_norm": 0.36132219433784485,
+      "learning_rate": 8.610573731122214e-05,
+      "loss": 1.7277,
+      "step": 2341
+    },
+    {
+      "epoch": 0.24327412485717254,
+      "grad_norm": 0.35456719994544983,
+      "learning_rate": 8.609444802056686e-05,
+      "loss": 1.7141,
+      "step": 2342
+    },
+    {
+      "epoch": 0.2433779993767529,
+      "grad_norm": 0.38762810826301575,
+      "learning_rate": 8.608315488613439e-05,
+      "loss": 1.7315,
+      "step": 2343
+    },
+    {
+      "epoch": 0.24348187389633322,
+      "grad_norm": 0.3507632911205292,
+      "learning_rate": 8.607185790912739e-05,
+      "loss": 1.712,
+      "step": 2344
+    },
+    {
+      "epoch": 0.24358574841591357,
+      "grad_norm": 0.39255252480506897,
+      "learning_rate": 8.60605570907489e-05,
+      "loss": 1.7046,
+      "step": 2345
+    },
+    {
+      "epoch": 0.24368962293549393,
+      "grad_norm": 0.35092976689338684,
+      "learning_rate": 8.604925243220235e-05,
+      "loss": 1.8009,
+      "step": 2346
+    },
+    {
+      "epoch": 0.24379349745507428,
+      "grad_norm": 0.37961629033088684,
+      "learning_rate": 8.603794393469162e-05,
+      "loss": 1.8602,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2438973719746546,
+      "grad_norm": 0.38758939504623413,
+      "learning_rate": 8.602663159942098e-05,
+      "loss": 1.8524,
+      "step": 2348
+    },
+    {
+      "epoch": 0.24400124649423496,
+      "grad_norm": 0.3866504430770874,
+      "learning_rate": 8.601531542759506e-05,
+      "loss": 1.722,
+      "step": 2349
+    },
+    {
+      "epoch": 0.24410512101381532,
+      "grad_norm": 0.3876989483833313,
+      "learning_rate": 8.600399542041901e-05,
+      "loss": 1.7846,
+      "step": 2350
+    },
+    {
+      "epoch": 0.24420899553339565,
+      "grad_norm": 0.3649909198284149,
+      "learning_rate": 8.599267157909827e-05,
+      "loss": 1.7255,
+      "step": 2351
+    },
+    {
+      "epoch": 0.244312870052976,
+      "grad_norm": 0.3867836892604828,
+      "learning_rate": 8.598134390483879e-05,
+      "loss": 1.8129,
+      "step": 2352
+    },
+    {
+      "epoch": 0.24441674457255635,
+      "grad_norm": 0.40322503447532654,
+      "learning_rate": 8.59700123988468e-05,
+      "loss": 2.0753,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2445206190921367,
+      "grad_norm": 0.37695807218551636,
+      "learning_rate": 8.595867706232911e-05,
+      "loss": 1.8298,
+      "step": 2354
+    },
+    {
+      "epoch": 0.24462449361171704,
+      "grad_norm": 0.3737477958202362,
+      "learning_rate": 8.594733789649279e-05,
+      "loss": 1.7315,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2447283681312974,
+      "grad_norm": 0.3679026663303375,
+      "learning_rate": 8.593599490254538e-05,
+      "loss": 1.6989,
+      "step": 2356
+    },
+    {
+      "epoch": 0.24483224265087775,
+      "grad_norm": 0.3665264844894409,
+      "learning_rate": 8.592464808169482e-05,
+      "loss": 1.7326,
+      "step": 2357
+    },
+    {
+      "epoch": 0.24493611717045807,
+      "grad_norm": 0.4071952700614929,
+      "learning_rate": 8.591329743514947e-05,
+      "loss": 1.774,
+      "step": 2358
+    },
+    {
+      "epoch": 0.24503999169003843,
+      "grad_norm": 0.4158788025379181,
+      "learning_rate": 8.590194296411806e-05,
+      "loss": 2.0193,
+      "step": 2359
+    },
+    {
+      "epoch": 0.24514386620961878,
+      "grad_norm": 0.3840784430503845,
+      "learning_rate": 8.589058466980979e-05,
+      "loss": 1.6699,
+      "step": 2360
+    },
+    {
+      "epoch": 0.24524774072919914,
+      "grad_norm": 0.36834099888801575,
+      "learning_rate": 8.587922255343421e-05,
+      "loss": 1.6759,
+      "step": 2361
+    },
+    {
+      "epoch": 0.24535161524877946,
+      "grad_norm": 0.35662734508514404,
+      "learning_rate": 8.58678566162013e-05,
+      "loss": 1.4792,
+      "step": 2362
+    },
+    {
+      "epoch": 0.24545548976835982,
+      "grad_norm": 0.369302362203598,
+      "learning_rate": 8.585648685932141e-05,
+      "loss": 1.7362,
+      "step": 2363
+    },
+    {
+      "epoch": 0.24555936428794017,
+      "grad_norm": 0.38440340757369995,
+      "learning_rate": 8.58451132840054e-05,
+      "loss": 1.8942,
+      "step": 2364
+    },
+    {
+      "epoch": 0.24566323880752053,
+      "grad_norm": 0.3850240707397461,
+      "learning_rate": 8.58337358914644e-05,
+      "loss": 1.8449,
+      "step": 2365
+    },
+    {
+      "epoch": 0.24576711332710086,
+      "grad_norm": 0.393264502286911,
+      "learning_rate": 8.582235468291007e-05,
+      "loss": 1.88,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2458709878466812,
+      "grad_norm": 0.40909549593925476,
+      "learning_rate": 8.581096965955436e-05,
+      "loss": 1.8228,
+      "step": 2367
+    },
+    {
+      "epoch": 0.24597486236626157,
+      "grad_norm": 0.36390751600265503,
+      "learning_rate": 8.579958082260973e-05,
+      "loss": 1.7457,
+      "step": 2368
+    },
+    {
+      "epoch": 0.2460787368858419,
+      "grad_norm": 0.3853476047515869,
+      "learning_rate": 8.5788188173289e-05,
+      "loss": 1.8158,
+      "step": 2369
+    },
+    {
+      "epoch": 0.24618261140542225,
+      "grad_norm": 0.37678855657577515,
+      "learning_rate": 8.577679171280537e-05,
+      "loss": 1.8368,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2462864859250026,
+      "grad_norm": 0.3758280277252197,
+      "learning_rate": 8.57653914423725e-05,
+      "loss": 1.6665,
+      "step": 2371
+    },
+    {
+      "epoch": 0.24639036044458296,
+      "grad_norm": 0.38738977909088135,
+      "learning_rate": 8.575398736320442e-05,
+      "loss": 1.8557,
+      "step": 2372
+    },
+    {
+      "epoch": 0.24649423496416328,
+      "grad_norm": 0.35604584217071533,
+      "learning_rate": 8.574257947651558e-05,
+      "loss": 1.6039,
+      "step": 2373
+    },
+    {
+      "epoch": 0.24659810948374364,
+      "grad_norm": 0.3709351420402527,
+      "learning_rate": 8.573116778352084e-05,
+      "loss": 1.85,
+      "step": 2374
+    },
+    {
+      "epoch": 0.246701984003324,
+      "grad_norm": 0.37650248408317566,
+      "learning_rate": 8.571975228543543e-05,
+      "loss": 1.8249,
+      "step": 2375
+    },
+    {
+      "epoch": 0.24680585852290432,
+      "grad_norm": 0.3389437794685364,
+      "learning_rate": 8.570833298347502e-05,
+      "loss": 1.7355,
+      "step": 2376
+    },
+    {
+      "epoch": 0.24690973304248467,
+      "grad_norm": 0.3376055061817169,
+      "learning_rate": 8.56969098788557e-05,
+      "loss": 1.5235,
+      "step": 2377
+    },
+    {
+      "epoch": 0.24701360756206503,
+      "grad_norm": 0.3974694013595581,
+      "learning_rate": 8.568548297279392e-05,
+      "loss": 1.8553,
+      "step": 2378
+    },
+    {
+      "epoch": 0.24711748208164538,
+      "grad_norm": 0.35596492886543274,
+      "learning_rate": 8.567405226650656e-05,
+      "loss": 1.4544,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2472213566012257,
+      "grad_norm": 0.37058472633361816,
+      "learning_rate": 8.56626177612109e-05,
+      "loss": 1.7564,
+      "step": 2380
+    },
+    {
+      "epoch": 0.24732523112080607,
+      "grad_norm": 0.3887064456939697,
+      "learning_rate": 8.565117945812463e-05,
+      "loss": 1.8172,
+      "step": 2381
+    },
+    {
+      "epoch": 0.24742910564038642,
+      "grad_norm": 0.3546907901763916,
+      "learning_rate": 8.563973735846583e-05,
+      "loss": 1.6763,
+      "step": 2382
+    },
+    {
+      "epoch": 0.24753298015996675,
+      "grad_norm": 0.38635915517807007,
+      "learning_rate": 8.562829146345301e-05,
+      "loss": 1.7017,
+      "step": 2383
+    },
+    {
+      "epoch": 0.2476368546795471,
+      "grad_norm": 0.3599991500377655,
+      "learning_rate": 8.561684177430507e-05,
+      "loss": 1.6087,
+      "step": 2384
+    },
+    {
+      "epoch": 0.24774072919912746,
+      "grad_norm": 0.3766249120235443,
+      "learning_rate": 8.560538829224129e-05,
+      "loss": 1.6795,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2478446037187078,
+      "grad_norm": 0.38669300079345703,
+      "learning_rate": 8.559393101848139e-05,
+      "loss": 1.7984,
+      "step": 2386
+    },
+    {
+      "epoch": 0.24794847823828814,
+      "grad_norm": 0.35774049162864685,
+      "learning_rate": 8.558246995424548e-05,
+      "loss": 1.5587,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2480523527578685,
+      "grad_norm": 0.3456031382083893,
+      "learning_rate": 8.557100510075406e-05,
+      "loss": 1.6269,
+      "step": 2388
+    },
+    {
+      "epoch": 0.24815622727744885,
+      "grad_norm": 0.42325082421302795,
+      "learning_rate": 8.555953645922809e-05,
+      "loss": 2.0851,
+      "step": 2389
+    },
+    {
+      "epoch": 0.24826010179702918,
+      "grad_norm": 0.32227158546447754,
+      "learning_rate": 8.554806403088884e-05,
+      "loss": 1.5321,
+      "step": 2390
+    },
+    {
+      "epoch": 0.24836397631660953,
+      "grad_norm": 0.35737884044647217,
+      "learning_rate": 8.553658781695807e-05,
+      "loss": 1.7002,
+      "step": 2391
+    },
+    {
+      "epoch": 0.24846785083618989,
+      "grad_norm": 0.3779980540275574,
+      "learning_rate": 8.552510781865788e-05,
+      "loss": 1.74,
+      "step": 2392
+    },
+    {
+      "epoch": 0.24857172535577024,
+      "grad_norm": 0.3488198518753052,
+      "learning_rate": 8.551362403721084e-05,
+      "loss": 1.7332,
+      "step": 2393
+    },
+    {
+      "epoch": 0.24867559987535057,
+      "grad_norm": 0.3871583342552185,
+      "learning_rate": 8.550213647383982e-05,
+      "loss": 1.8173,
+      "step": 2394
+    },
+    {
+      "epoch": 0.24877947439493092,
+      "grad_norm": 0.3559076488018036,
+      "learning_rate": 8.549064512976822e-05,
+      "loss": 1.824,
+      "step": 2395
+    },
+    {
+      "epoch": 0.24888334891451128,
+      "grad_norm": 0.36714842915534973,
+      "learning_rate": 8.547915000621974e-05,
+      "loss": 1.7691,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2489872234340916,
+      "grad_norm": 0.35242655873298645,
+      "learning_rate": 8.546765110441855e-05,
+      "loss": 1.7228,
+      "step": 2397
+    },
+    {
+      "epoch": 0.24909109795367196,
+      "grad_norm": 0.3446803689002991,
+      "learning_rate": 8.545614842558915e-05,
+      "loss": 1.613,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2491949724732523,
+      "grad_norm": 0.36509135365486145,
+      "learning_rate": 8.544464197095651e-05,
+      "loss": 1.6443,
+      "step": 2399
+    },
+    {
+      "epoch": 0.24929884699283267,
+      "grad_norm": 0.3790837228298187,
+      "learning_rate": 8.5433131741746e-05,
+      "loss": 1.7364,
+      "step": 2400
+    },
+    {
+      "epoch": 0.249402721512413,
+      "grad_norm": 0.4487643837928772,
+      "learning_rate": 8.542161773918334e-05,
+      "loss": 1.8304,
+      "step": 2401
+    },
+    {
+      "epoch": 0.24950659603199335,
+      "grad_norm": 0.3896706700325012,
+      "learning_rate": 8.54100999644947e-05,
+      "loss": 1.7448,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2496104705515737,
+      "grad_norm": 0.382664293050766,
+      "learning_rate": 8.53985784189066e-05,
+      "loss": 1.7384,
+      "step": 2403
+    },
+    {
+      "epoch": 0.24971434507115406,
+      "grad_norm": 0.3803345561027527,
+      "learning_rate": 8.538705310364603e-05,
+      "loss": 1.8425,
+      "step": 2404
+    },
+    {
+      "epoch": 0.24981821959073439,
+      "grad_norm": 0.37627753615379333,
+      "learning_rate": 8.537552401994034e-05,
+      "loss": 1.7596,
+      "step": 2405
+    },
+    {
+      "epoch": 0.24992209411031474,
+      "grad_norm": 0.3746028542518616,
+      "learning_rate": 8.536399116901728e-05,
+      "loss": 1.8014,
+      "step": 2406
+    },
+    {
+      "epoch": 0.25002596862989507,
+      "grad_norm": 0.3502449095249176,
+      "learning_rate": 8.535245455210501e-05,
+      "loss": 1.5849,
+      "step": 2407
+    },
+    {
+      "epoch": 0.25012984314947545,
+      "grad_norm": 0.40077903866767883,
+      "learning_rate": 8.534091417043208e-05,
+      "loss": 1.9111,
+      "step": 2408
+    },
+    {
+      "epoch": 0.2502337176690558,
+      "grad_norm": 0.3578510284423828,
+      "learning_rate": 8.532937002522747e-05,
+      "loss": 1.6662,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2503375921886361,
+      "grad_norm": 0.38965925574302673,
+      "learning_rate": 8.531782211772052e-05,
+      "loss": 1.5775,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2504414667082165,
+      "grad_norm": 0.35679998993873596,
+      "learning_rate": 8.530627044914101e-05,
+      "loss": 1.796,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2505453412277968,
+      "grad_norm": 0.3703193962574005,
+      "learning_rate": 8.52947150207191e-05,
+      "loss": 1.8143,
+      "step": 2412
+    },
+    {
+      "epoch": 0.25064921574737714,
+      "grad_norm": 0.3648841977119446,
+      "learning_rate": 8.528315583368536e-05,
+      "loss": 1.7078,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2507530902669575,
+      "grad_norm": 0.38748499751091003,
+      "learning_rate": 8.527159288927074e-05,
+      "loss": 1.876,
+      "step": 2414
+    },
+    {
+      "epoch": 0.25085696478653785,
+      "grad_norm": 0.37417733669281006,
+      "learning_rate": 8.526002618870659e-05,
+      "loss": 1.646,
+      "step": 2415
+    },
+    {
+      "epoch": 0.25096083930611823,
+      "grad_norm": 0.3735947012901306,
+      "learning_rate": 8.52484557332247e-05,
+      "loss": 1.8018,
+      "step": 2416
+    },
+    {
+      "epoch": 0.25106471382569856,
+      "grad_norm": 0.36857619881629944,
+      "learning_rate": 8.523688152405722e-05,
+      "loss": 1.6909,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2511685883452789,
+      "grad_norm": 0.41932663321495056,
+      "learning_rate": 8.522530356243671e-05,
+      "loss": 1.6112,
+      "step": 2418
+    },
+    {
+      "epoch": 0.25127246286485927,
+      "grad_norm": 0.4023738503456116,
+      "learning_rate": 8.521372184959615e-05,
+      "loss": 1.9221,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2513763373844396,
+      "grad_norm": 0.35625362396240234,
+      "learning_rate": 8.520213638676889e-05,
+      "loss": 1.6999,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2514802119040199,
+      "grad_norm": 0.36791056394577026,
+      "learning_rate": 8.519054717518868e-05,
+      "loss": 1.8009,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2515840864236003,
+      "grad_norm": 0.3679400086402893,
+      "learning_rate": 8.517895421608972e-05,
+      "loss": 1.6935,
+      "step": 2422
+    },
+    {
+      "epoch": 0.25168796094318063,
+      "grad_norm": 0.36350300908088684,
+      "learning_rate": 8.516735751070652e-05,
+      "loss": 1.7611,
+      "step": 2423
+    },
+    {
+      "epoch": 0.25179183546276096,
+      "grad_norm": 0.36680689454078674,
+      "learning_rate": 8.515575706027406e-05,
+      "loss": 1.7451,
+      "step": 2424
+    },
+    {
+      "epoch": 0.25189570998234134,
+      "grad_norm": 0.3586255609989166,
+      "learning_rate": 8.514415286602771e-05,
+      "loss": 1.6652,
+      "step": 2425
+    },
+    {
+      "epoch": 0.25199958450192167,
+      "grad_norm": 0.380088210105896,
+      "learning_rate": 8.51325449292032e-05,
+      "loss": 1.7082,
+      "step": 2426
+    },
+    {
+      "epoch": 0.25210345902150205,
+      "grad_norm": 0.3451653718948364,
+      "learning_rate": 8.512093325103671e-05,
+      "loss": 1.6654,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2522073335410824,
+      "grad_norm": 0.42833617329597473,
+      "learning_rate": 8.510931783276477e-05,
+      "loss": 1.9137,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2523112080606627,
+      "grad_norm": 0.42482277750968933,
+      "learning_rate": 8.509769867562436e-05,
+      "loss": 2.0342,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2524150825802431,
+      "grad_norm": 0.3890083134174347,
+      "learning_rate": 8.508607578085282e-05,
+      "loss": 1.7393,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2525189570998234,
+      "grad_norm": 0.37014102935791016,
+      "learning_rate": 8.507444914968788e-05,
+      "loss": 1.8187,
+      "step": 2431
+    },
+    {
+      "epoch": 0.25262283161940374,
+      "grad_norm": 0.370259553194046,
+      "learning_rate": 8.506281878336768e-05,
+      "loss": 1.733,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2527267061389841,
+      "grad_norm": 0.37391120195388794,
+      "learning_rate": 8.50511846831308e-05,
+      "loss": 1.691,
+      "step": 2433
+    },
+    {
+      "epoch": 0.25283058065856445,
+      "grad_norm": 0.3874087631702423,
+      "learning_rate": 8.503954685021616e-05,
+      "loss": 1.7962,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2529344551781448,
+      "grad_norm": 0.3736341893672943,
+      "learning_rate": 8.50279052858631e-05,
+      "loss": 1.8226,
+      "step": 2435
+    },
+    {
+      "epoch": 0.25303832969772516,
+      "grad_norm": 0.37461403012275696,
+      "learning_rate": 8.501625999131134e-05,
+      "loss": 1.8345,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2531422042173055,
+      "grad_norm": 0.39281967282295227,
+      "learning_rate": 8.500461096780105e-05,
+      "loss": 1.6779,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2532460787368858,
+      "grad_norm": 0.4174926280975342,
+      "learning_rate": 8.499295821657273e-05,
+      "loss": 1.8124,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2533499532564662,
+      "grad_norm": 0.4330326318740845,
+      "learning_rate": 8.498130173886731e-05,
+      "loss": 1.9252,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2534538277760465,
+      "grad_norm": 0.357661634683609,
+      "learning_rate": 8.496964153592613e-05,
+      "loss": 1.7423,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2535577022956269,
+      "grad_norm": 0.3977978527545929,
+      "learning_rate": 8.495797760899088e-05,
+      "loss": 1.4438,
+      "step": 2441
+    },
+    {
+      "epoch": 0.25366157681520723,
+      "grad_norm": 0.4523005783557892,
+      "learning_rate": 8.494630995930372e-05,
+      "loss": 1.5389,
+      "step": 2442
+    },
+    {
+      "epoch": 0.25376545133478756,
+      "grad_norm": 0.3693365156650543,
+      "learning_rate": 8.493463858810713e-05,
+      "loss": 1.8215,
+      "step": 2443
+    },
+    {
+      "epoch": 0.25386932585436794,
+      "grad_norm": 0.40159082412719727,
+      "learning_rate": 8.492296349664401e-05,
+      "loss": 1.8385,
+      "step": 2444
+    },
+    {
+      "epoch": 0.25397320037394827,
+      "grad_norm": 0.3591073751449585,
+      "learning_rate": 8.491128468615772e-05,
+      "loss": 1.642,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2540770748935286,
+      "grad_norm": 0.3740187883377075,
+      "learning_rate": 8.489960215789192e-05,
+      "loss": 1.8311,
+      "step": 2446
+    },
+    {
+      "epoch": 0.254180949413109,
+      "grad_norm": 0.38675010204315186,
+      "learning_rate": 8.488791591309072e-05,
+      "loss": 1.7373,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2542848239326893,
+      "grad_norm": 0.34369611740112305,
+      "learning_rate": 8.48762259529986e-05,
+      "loss": 1.7054,
+      "step": 2448
+    },
+    {
+      "epoch": 0.25438869845226963,
+      "grad_norm": 0.3643989861011505,
+      "learning_rate": 8.486453227886049e-05,
+      "loss": 1.6371,
+      "step": 2449
+    },
+    {
+      "epoch": 0.25449257297185,
+      "grad_norm": 0.38395020365715027,
+      "learning_rate": 8.485283489192162e-05,
+      "loss": 1.5869,
+      "step": 2450
+    },
+    {
+      "epoch": 0.25459644749143034,
+      "grad_norm": 0.365596741437912,
+      "learning_rate": 8.48411337934277e-05,
+      "loss": 1.5886,
+      "step": 2451
+    },
+    {
+      "epoch": 0.25470032201101067,
+      "grad_norm": 0.3677592873573303,
+      "learning_rate": 8.48294289846248e-05,
+      "loss": 1.8044,
+      "step": 2452
+    },
+    {
+      "epoch": 0.25480419653059105,
+      "grad_norm": 0.39587604999542236,
+      "learning_rate": 8.48177204667594e-05,
+      "loss": 1.8413,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2549080710501714,
+      "grad_norm": 0.3834674656391144,
+      "learning_rate": 8.480600824107837e-05,
+      "loss": 1.6651,
+      "step": 2454
+    },
+    {
+      "epoch": 0.25501194556975176,
+      "grad_norm": 0.38077157735824585,
+      "learning_rate": 8.479429230882893e-05,
+      "loss": 1.558,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2551158200893321,
+      "grad_norm": 0.4156430959701538,
+      "learning_rate": 8.478257267125878e-05,
+      "loss": 1.7901,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2552196946089124,
+      "grad_norm": 0.3814376890659332,
+      "learning_rate": 8.477084932961596e-05,
+      "loss": 1.6559,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2553235691284928,
+      "grad_norm": 0.3800637722015381,
+      "learning_rate": 8.475912228514889e-05,
+      "loss": 1.7842,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2554274436480731,
+      "grad_norm": 0.36574795842170715,
+      "learning_rate": 8.474739153910646e-05,
+      "loss": 1.6281,
+      "step": 2459
+    },
+    {
+      "epoch": 0.25553131816765345,
+      "grad_norm": 0.3729517161846161,
+      "learning_rate": 8.473565709273785e-05,
+      "loss": 1.8031,
+      "step": 2460
+    },
+    {
+      "epoch": 0.25563519268723384,
+      "grad_norm": 0.4632837176322937,
+      "learning_rate": 8.47239189472927e-05,
+      "loss": 1.9434,
+      "step": 2461
+    },
+    {
+      "epoch": 0.25573906720681416,
+      "grad_norm": 0.3544566333293915,
+      "learning_rate": 8.471217710402106e-05,
+      "loss": 1.8198,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2558429417263945,
+      "grad_norm": 0.3790525794029236,
+      "learning_rate": 8.470043156417333e-05,
+      "loss": 1.8017,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2559468162459749,
+      "grad_norm": 0.3937493562698364,
+      "learning_rate": 8.46886823290003e-05,
+      "loss": 1.7512,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2560506907655552,
+      "grad_norm": 0.4141665995121002,
+      "learning_rate": 8.467692939975316e-05,
+      "loss": 1.9163,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2561545652851356,
+      "grad_norm": 0.4007442891597748,
+      "learning_rate": 8.466517277768355e-05,
+      "loss": 1.9303,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2562584398047159,
+      "grad_norm": 0.36880823969841003,
+      "learning_rate": 8.465341246404345e-05,
+      "loss": 1.6991,
+      "step": 2467
+    },
+    {
+      "epoch": 0.25636231432429624,
+      "grad_norm": 0.3562661409378052,
+      "learning_rate": 8.464164846008522e-05,
+      "loss": 1.7796,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2564661888438766,
+      "grad_norm": 0.3477732241153717,
+      "learning_rate": 8.462988076706164e-05,
+      "loss": 1.7226,
+      "step": 2469
+    },
+    {
+      "epoch": 0.25657006336345695,
+      "grad_norm": 0.35839709639549255,
+      "learning_rate": 8.461810938622589e-05,
+      "loss": 1.7105,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2566739378830373,
+      "grad_norm": 0.39441442489624023,
+      "learning_rate": 8.460633431883151e-05,
+      "loss": 1.6964,
+      "step": 2471
+    },
+    {
+      "epoch": 0.25677781240261766,
+      "grad_norm": 0.41418394446372986,
+      "learning_rate": 8.459455556613247e-05,
+      "loss": 1.7889,
+      "step": 2472
+    },
+    {
+      "epoch": 0.256881686922198,
+      "grad_norm": 0.3982704281806946,
+      "learning_rate": 8.458277312938312e-05,
+      "loss": 2.066,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2569855614417783,
+      "grad_norm": 0.38881179690361023,
+      "learning_rate": 8.457098700983818e-05,
+      "loss": 1.8605,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2570894359613587,
+      "grad_norm": 0.39533254504203796,
+      "learning_rate": 8.455919720875279e-05,
+      "loss": 1.8728,
+      "step": 2475
+    },
+    {
+      "epoch": 0.257193310480939,
+      "grad_norm": 0.4153515100479126,
+      "learning_rate": 8.454740372738246e-05,
+      "loss": 1.6178,
+      "step": 2476
+    },
+    {
+      "epoch": 0.25729718500051935,
+      "grad_norm": 0.35797804594039917,
+      "learning_rate": 8.453560656698311e-05,
+      "loss": 1.5962,
+      "step": 2477
+    },
+    {
+      "epoch": 0.25740105952009973,
+      "grad_norm": 0.3884628415107727,
+      "learning_rate": 8.452380572881107e-05,
+      "loss": 1.6725,
+      "step": 2478
+    },
+    {
+      "epoch": 0.25750493403968006,
+      "grad_norm": 0.4121970534324646,
+      "learning_rate": 8.451200121412299e-05,
+      "loss": 1.731,
+      "step": 2479
+    },
+    {
+      "epoch": 0.25760880855926044,
+      "grad_norm": 0.3405035734176636,
+      "learning_rate": 8.4500193024176e-05,
+      "loss": 1.5419,
+      "step": 2480
+    },
+    {
+      "epoch": 0.25771268307884077,
+      "grad_norm": 0.38876980543136597,
+      "learning_rate": 8.448838116022758e-05,
+      "loss": 1.6488,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2578165575984211,
+      "grad_norm": 0.38169679045677185,
+      "learning_rate": 8.447656562353557e-05,
+      "loss": 1.8325,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2579204321180015,
+      "grad_norm": 0.36781740188598633,
+      "learning_rate": 8.446474641535824e-05,
+      "loss": 1.5927,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2580243066375818,
+      "grad_norm": 0.39887794852256775,
+      "learning_rate": 8.445292353695427e-05,
+      "loss": 1.7471,
+      "step": 2484
+    },
+    {
+      "epoch": 0.25812818115716213,
+      "grad_norm": 0.381283164024353,
+      "learning_rate": 8.444109698958267e-05,
+      "loss": 1.7668,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2582320556767425,
+      "grad_norm": 0.38031065464019775,
+      "learning_rate": 8.44292667745029e-05,
+      "loss": 1.7104,
+      "step": 2486
+    },
+    {
+      "epoch": 0.25833593019632284,
+      "grad_norm": 0.36943575739860535,
+      "learning_rate": 8.441743289297476e-05,
+      "loss": 1.8359,
+      "step": 2487
+    },
+    {
+      "epoch": 0.25843980471590317,
+      "grad_norm": 0.3790445625782013,
+      "learning_rate": 8.440559534625851e-05,
+      "loss": 1.6393,
+      "step": 2488
+    },
+    {
+      "epoch": 0.25854367923548355,
+      "grad_norm": 0.3714222013950348,
+      "learning_rate": 8.439375413561472e-05,
+      "loss": 1.6612,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2586475537550639,
+      "grad_norm": 0.38887256383895874,
+      "learning_rate": 8.43819092623044e-05,
+      "loss": 1.7044,
+      "step": 2490
+    },
+    {
+      "epoch": 0.25875142827464426,
+      "grad_norm": 0.3530022203922272,
+      "learning_rate": 8.437006072758891e-05,
+      "loss": 1.608,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2588553027942246,
+      "grad_norm": 0.38229474425315857,
+      "learning_rate": 8.435820853273007e-05,
+      "loss": 1.7361,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2589591773138049,
+      "grad_norm": 0.35574570298194885,
+      "learning_rate": 8.434635267899002e-05,
+      "loss": 1.799,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2590630518333853,
+      "grad_norm": 0.40232568979263306,
+      "learning_rate": 8.433449316763133e-05,
+      "loss": 1.8073,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2591669263529656,
+      "grad_norm": 0.42104196548461914,
+      "learning_rate": 8.432262999991694e-05,
+      "loss": 2.0099,
+      "step": 2495
+    },
+    {
+      "epoch": 0.25927080087254595,
+      "grad_norm": 0.37814000248908997,
+      "learning_rate": 8.431076317711017e-05,
+      "loss": 1.6813,
+      "step": 2496
+    },
+    {
+      "epoch": 0.25937467539212633,
+      "grad_norm": 0.38011592626571655,
+      "learning_rate": 8.429889270047475e-05,
+      "loss": 1.6921,
+      "step": 2497
+    },
+    {
+      "epoch": 0.25947854991170666,
+      "grad_norm": 0.3853015601634979,
+      "learning_rate": 8.428701857127481e-05,
+      "loss": 1.8239,
+      "step": 2498
+    },
+    {
+      "epoch": 0.259582424431287,
+      "grad_norm": 0.351633757352829,
+      "learning_rate": 8.427514079077485e-05,
+      "loss": 1.6058,
+      "step": 2499
+    },
+    {
+      "epoch": 0.25968629895086737,
+      "grad_norm": 0.4057175815105438,
+      "learning_rate": 8.426325936023974e-05,
+      "loss": 1.8754,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2597901734704477,
+      "grad_norm": 0.3578610420227051,
+      "learning_rate": 8.425137428093477e-05,
+      "loss": 1.7336,
+      "step": 2501
+    },
+    {
+      "epoch": 0.259894047990028,
+      "grad_norm": 0.3505629301071167,
+      "learning_rate": 8.423948555412562e-05,
+      "loss": 1.7356,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2599979225096084,
+      "grad_norm": 0.3745979368686676,
+      "learning_rate": 8.422759318107832e-05,
+      "loss": 1.8169,
+      "step": 2503
+    },
+    {
+      "epoch": 0.26010179702918873,
+      "grad_norm": 0.37145760655403137,
+      "learning_rate": 8.421569716305934e-05,
+      "loss": 1.904,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2602056715487691,
+      "grad_norm": 0.3965916931629181,
+      "learning_rate": 8.42037975013355e-05,
+      "loss": 1.809,
+      "step": 2505
+    },
+    {
+      "epoch": 0.26030954606834944,
+      "grad_norm": 0.39221835136413574,
+      "learning_rate": 8.419189419717404e-05,
+      "loss": 1.7811,
+      "step": 2506
+    },
+    {
+      "epoch": 0.26041342058792977,
+      "grad_norm": 0.3546196222305298,
+      "learning_rate": 8.417998725184254e-05,
+      "loss": 1.7371,
+      "step": 2507
+    },
+    {
+      "epoch": 0.26051729510751015,
+      "grad_norm": 0.3630322217941284,
+      "learning_rate": 8.416807666660901e-05,
+      "loss": 1.719,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2606211696270905,
+      "grad_norm": 0.34440553188323975,
+      "learning_rate": 8.415616244274185e-05,
+      "loss": 1.6887,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2607250441466708,
+      "grad_norm": 0.39099493622779846,
+      "learning_rate": 8.414424458150983e-05,
+      "loss": 1.7395,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2608289186662512,
+      "grad_norm": 0.35405251383781433,
+      "learning_rate": 8.413232308418206e-05,
+      "loss": 1.6363,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2609327931858315,
+      "grad_norm": 0.41946062445640564,
+      "learning_rate": 8.412039795202816e-05,
+      "loss": 1.91,
+      "step": 2512
+    },
+    {
+      "epoch": 0.26103666770541184,
+      "grad_norm": 0.4249524474143982,
+      "learning_rate": 8.4108469186318e-05,
+      "loss": 1.7341,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2611405422249922,
+      "grad_norm": 0.3654036223888397,
+      "learning_rate": 8.409653678832194e-05,
+      "loss": 1.6744,
+      "step": 2514
+    },
+    {
+      "epoch": 0.26124441674457255,
+      "grad_norm": 0.35637834668159485,
+      "learning_rate": 8.408460075931068e-05,
+      "loss": 1.6597,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2613482912641529,
+      "grad_norm": 0.39124104380607605,
+      "learning_rate": 8.407266110055531e-05,
+      "loss": 1.6995,
+      "step": 2516
+    },
+    {
+      "epoch": 0.26145216578373326,
+      "grad_norm": 0.3535708487033844,
+      "learning_rate": 8.40607178133273e-05,
+      "loss": 1.657,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2615560403033136,
+      "grad_norm": 0.4048800766468048,
+      "learning_rate": 8.404877089889853e-05,
+      "loss": 1.7071,
+      "step": 2518
+    },
+    {
+      "epoch": 0.26165991482289397,
+      "grad_norm": 0.3928733170032501,
+      "learning_rate": 8.403682035854125e-05,
+      "loss": 1.5713,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2617637893424743,
+      "grad_norm": 0.3696751892566681,
+      "learning_rate": 8.40248661935281e-05,
+      "loss": 1.6968,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2618676638620546,
+      "grad_norm": 0.35033291578292847,
+      "learning_rate": 8.40129084051321e-05,
+      "loss": 1.5176,
+      "step": 2521
+    },
+    {
+      "epoch": 0.261971538381635,
+      "grad_norm": 0.38887977600097656,
+      "learning_rate": 8.400094699462667e-05,
+      "loss": 1.84,
+      "step": 2522
+    },
+    {
+      "epoch": 0.26207541290121533,
+      "grad_norm": 0.382174015045166,
+      "learning_rate": 8.398898196328561e-05,
+      "loss": 1.818,
+      "step": 2523
+    },
+    {
+      "epoch": 0.26217928742079566,
+      "grad_norm": 0.3753756880760193,
+      "learning_rate": 8.39770133123831e-05,
+      "loss": 1.764,
+      "step": 2524
+    },
+    {
+      "epoch": 0.26228316194037604,
+      "grad_norm": 0.37997132539749146,
+      "learning_rate": 8.396504104319366e-05,
+      "loss": 1.8256,
+      "step": 2525
+    },
+    {
+      "epoch": 0.26238703645995637,
+      "grad_norm": 0.36985552310943604,
+      "learning_rate": 8.395306515699234e-05,
+      "loss": 1.5408,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2624909109795367,
+      "grad_norm": 0.36994847655296326,
+      "learning_rate": 8.394108565505441e-05,
+      "loss": 1.6385,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2625947854991171,
+      "grad_norm": 0.3789272904396057,
+      "learning_rate": 8.392910253865557e-05,
+      "loss": 1.6302,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2626986600186974,
+      "grad_norm": 0.39001405239105225,
+      "learning_rate": 8.391711580907202e-05,
+      "loss": 1.8292,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2628025345382778,
+      "grad_norm": 0.3820188343524933,
+      "learning_rate": 8.390512546758016e-05,
+      "loss": 1.8284,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2629064090578581,
+      "grad_norm": 0.43149465322494507,
+      "learning_rate": 8.389313151545694e-05,
+      "loss": 1.8026,
+      "step": 2531
+    },
+    {
+      "epoch": 0.26301028357743844,
+      "grad_norm": 0.4724147319793701,
+      "learning_rate": 8.388113395397957e-05,
+      "loss": 1.7194,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2631141580970188,
+      "grad_norm": 0.3973971903324127,
+      "learning_rate": 8.386913278442571e-05,
+      "loss": 1.703,
+      "step": 2533
+    },
+    {
+      "epoch": 0.26321803261659915,
+      "grad_norm": 0.38330528140068054,
+      "learning_rate": 8.385712800807343e-05,
+      "loss": 1.8659,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2633219071361795,
+      "grad_norm": 0.37077972292900085,
+      "learning_rate": 8.38451196262011e-05,
+      "loss": 1.8752,
+      "step": 2535
+    },
+    {
+      "epoch": 0.26342578165575986,
+      "grad_norm": 0.36941617727279663,
+      "learning_rate": 8.383310764008751e-05,
+      "loss": 1.765,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2635296561753402,
+      "grad_norm": 0.3706973195075989,
+      "learning_rate": 8.382109205101188e-05,
+      "loss": 1.6668,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2636335306949205,
+      "grad_norm": 0.37633809447288513,
+      "learning_rate": 8.380907286025377e-05,
+      "loss": 1.6682,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2637374052145009,
+      "grad_norm": 0.39212766289711,
+      "learning_rate": 8.37970500690931e-05,
+      "loss": 1.6501,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2638412797340812,
+      "grad_norm": 0.3803769052028656,
+      "learning_rate": 8.378502367881025e-05,
+      "loss": 1.6854,
+      "step": 2540
+    },
+    {
+      "epoch": 0.26394515425366155,
+      "grad_norm": 0.40050208568573,
+      "learning_rate": 8.37729936906859e-05,
+      "loss": 1.8676,
+      "step": 2541
+    },
+    {
+      "epoch": 0.26404902877324193,
+      "grad_norm": 0.39779967069625854,
+      "learning_rate": 8.376096010600116e-05,
+      "loss": 1.7777,
+      "step": 2542
+    },
+    {
+      "epoch": 0.26415290329282226,
+      "grad_norm": 0.4135241210460663,
+      "learning_rate": 8.374892292603751e-05,
+      "loss": 1.7512,
+      "step": 2543
+    },
+    {
+      "epoch": 0.26425677781240264,
+      "grad_norm": 0.36392831802368164,
+      "learning_rate": 8.373688215207682e-05,
+      "loss": 1.7389,
+      "step": 2544
+    },
+    {
+      "epoch": 0.26436065233198297,
+      "grad_norm": 0.40198683738708496,
+      "learning_rate": 8.372483778540134e-05,
+      "loss": 1.9784,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2644645268515633,
+      "grad_norm": 0.35998931527137756,
+      "learning_rate": 8.371278982729371e-05,
+      "loss": 1.8175,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2645684013711437,
+      "grad_norm": 0.4466514587402344,
+      "learning_rate": 8.370073827903693e-05,
+      "loss": 1.8014,
+      "step": 2547
+    },
+    {
+      "epoch": 0.264672275890724,
+      "grad_norm": 0.39324405789375305,
+      "learning_rate": 8.368868314191439e-05,
+      "loss": 1.8389,
+      "step": 2548
+    },
+    {
+      "epoch": 0.26477615041030433,
+      "grad_norm": 0.3979739248752594,
+      "learning_rate": 8.367662441720989e-05,
+      "loss": 1.7975,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2648800249298847,
+      "grad_norm": 0.3831605315208435,
+      "learning_rate": 8.366456210620757e-05,
+      "loss": 1.7531,
+      "step": 2550
+    },
+    {
+      "epoch": 0.26498389944946504,
+      "grad_norm": 0.3608654737472534,
+      "learning_rate": 8.365249621019197e-05,
+      "loss": 1.6416,
+      "step": 2551
+    },
+    {
+      "epoch": 0.26508777396904537,
+      "grad_norm": 0.37154313921928406,
+      "learning_rate": 8.364042673044803e-05,
+      "loss": 1.775,
+      "step": 2552
+    },
+    {
+      "epoch": 0.26519164848862575,
+      "grad_norm": 0.36865541338920593,
+      "learning_rate": 8.362835366826105e-05,
+      "loss": 1.6302,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2652955230082061,
+      "grad_norm": 0.40405401587486267,
+      "learning_rate": 8.361627702491673e-05,
+      "loss": 1.8325,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2653993975277864,
+      "grad_norm": 0.3631226122379303,
+      "learning_rate": 8.360419680170111e-05,
+      "loss": 1.6433,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2655032720473668,
+      "grad_norm": 0.3895721435546875,
+      "learning_rate": 8.359211299990064e-05,
+      "loss": 1.5261,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2656071465669471,
+      "grad_norm": 0.34275680780410767,
+      "learning_rate": 8.358002562080219e-05,
+      "loss": 1.7036,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2657110210865275,
+      "grad_norm": 0.4039117991924286,
+      "learning_rate": 8.356793466569293e-05,
+      "loss": 1.801,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2658148956061078,
+      "grad_norm": 0.444367915391922,
+      "learning_rate": 8.355584013586047e-05,
+      "loss": 2.0989,
+      "step": 2559
+    },
+    {
+      "epoch": 0.26591877012568815,
+      "grad_norm": 0.3847164809703827,
+      "learning_rate": 8.354374203259278e-05,
+      "loss": 1.8125,
+      "step": 2560
+    },
+    {
+      "epoch": 0.26602264464526854,
+      "grad_norm": 0.36114174127578735,
+      "learning_rate": 8.353164035717822e-05,
+      "loss": 1.6731,
+      "step": 2561
+    },
+    {
+      "epoch": 0.26612651916484886,
+      "grad_norm": 0.37462207674980164,
+      "learning_rate": 8.351953511090551e-05,
+      "loss": 1.7634,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2662303936844292,
+      "grad_norm": 0.38300392031669617,
+      "learning_rate": 8.350742629506378e-05,
+      "loss": 1.8874,
+      "step": 2563
+    },
+    {
+      "epoch": 0.26633426820400957,
+      "grad_norm": 0.37371304631233215,
+      "learning_rate": 8.349531391094251e-05,
+      "loss": 1.7228,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2664381427235899,
+      "grad_norm": 0.33728665113449097,
+      "learning_rate": 8.34831979598316e-05,
+      "loss": 1.6236,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2665420172431702,
+      "grad_norm": 0.38045376539230347,
+      "learning_rate": 8.347107844302129e-05,
+      "loss": 1.8109,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2666458917627506,
+      "grad_norm": 0.36660969257354736,
+      "learning_rate": 8.345895536180219e-05,
+      "loss": 1.7463,
+      "step": 2567
+    },
+    {
+      "epoch": 0.26674976628233094,
+      "grad_norm": 0.3986735939979553,
+      "learning_rate": 8.344682871746534e-05,
+      "loss": 1.8362,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2668536408019113,
+      "grad_norm": 0.41932711005210876,
+      "learning_rate": 8.343469851130212e-05,
+      "loss": 1.845,
+      "step": 2569
+    },
+    {
+      "epoch": 0.26695751532149165,
+      "grad_norm": 0.40920954942703247,
+      "learning_rate": 8.342256474460431e-05,
+      "loss": 1.6763,
+      "step": 2570
+    },
+    {
+      "epoch": 0.26706138984107197,
+      "grad_norm": 0.3709944784641266,
+      "learning_rate": 8.341042741866408e-05,
+      "loss": 1.8428,
+      "step": 2571
+    },
+    {
+      "epoch": 0.26716526436065235,
+      "grad_norm": 0.3826795220375061,
+      "learning_rate": 8.339828653477391e-05,
+      "loss": 1.5346,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2672691388802327,
+      "grad_norm": 0.37877601385116577,
+      "learning_rate": 8.338614209422677e-05,
+      "loss": 1.7763,
+      "step": 2573
+    },
+    {
+      "epoch": 0.267373013399813,
+      "grad_norm": 0.4136562645435333,
+      "learning_rate": 8.33739940983159e-05,
+      "loss": 1.9153,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2674768879193934,
+      "grad_norm": 0.3574436604976654,
+      "learning_rate": 8.336184254833499e-05,
+      "loss": 1.4252,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2675807624389737,
+      "grad_norm": 0.37612104415893555,
+      "learning_rate": 8.33496874455781e-05,
+      "loss": 1.8131,
+      "step": 2576
+    },
+    {
+      "epoch": 0.26768463695855405,
+      "grad_norm": 0.4038298726081848,
+      "learning_rate": 8.33375287913396e-05,
+      "loss": 1.769,
+      "step": 2577
+    },
+    {
+      "epoch": 0.26778851147813443,
+      "grad_norm": 0.3684050440788269,
+      "learning_rate": 8.332536658691435e-05,
+      "loss": 1.806,
+      "step": 2578
+    },
+    {
+      "epoch": 0.26789238599771475,
+      "grad_norm": 0.39676815271377563,
+      "learning_rate": 8.331320083359751e-05,
+      "loss": 1.8308,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2679962605172951,
+      "grad_norm": 0.37601250410079956,
+      "learning_rate": 8.330103153268462e-05,
+      "loss": 1.6708,
+      "step": 2580
+    },
+    {
+      "epoch": 0.26810013503687546,
+      "grad_norm": 0.35616087913513184,
+      "learning_rate": 8.328885868547164e-05,
+      "loss": 1.6567,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2682040095564558,
+      "grad_norm": 0.4200495183467865,
+      "learning_rate": 8.327668229325487e-05,
+      "loss": 1.9518,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2683078840760362,
+      "grad_norm": 0.4047677516937256,
+      "learning_rate": 8.326450235733098e-05,
+      "loss": 1.8429,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2684117585956165,
+      "grad_norm": 0.35730311274528503,
+      "learning_rate": 8.32523188789971e-05,
+      "loss": 1.5773,
+      "step": 2584
+    },
+    {
+      "epoch": 0.26851563311519683,
+      "grad_norm": 0.38395121693611145,
+      "learning_rate": 8.324013185955062e-05,
+      "loss": 1.685,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2686195076347772,
+      "grad_norm": 0.3953067362308502,
+      "learning_rate": 8.322794130028938e-05,
+      "loss": 1.8974,
+      "step": 2586
+    },
+    {
+      "epoch": 0.26872338215435754,
+      "grad_norm": 0.3572498559951782,
+      "learning_rate": 8.321574720251158e-05,
+      "loss": 1.6507,
+      "step": 2587
+    },
+    {
+      "epoch": 0.26882725667393786,
+      "grad_norm": 0.38150545954704285,
+      "learning_rate": 8.32035495675158e-05,
+      "loss": 1.7104,
+      "step": 2588
+    },
+    {
+      "epoch": 0.26893113119351825,
+      "grad_norm": 0.36428940296173096,
+      "learning_rate": 8.3191348396601e-05,
+      "loss": 1.6052,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2690350057130986,
+      "grad_norm": 0.4125150144100189,
+      "learning_rate": 8.317914369106648e-05,
+      "loss": 1.9357,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2691388802326789,
+      "grad_norm": 0.3636937439441681,
+      "learning_rate": 8.316693545221196e-05,
+      "loss": 1.7948,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2692427547522593,
+      "grad_norm": 0.40792933106422424,
+      "learning_rate": 8.315472368133754e-05,
+      "loss": 1.6696,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2693466292718396,
+      "grad_norm": 0.40981337428092957,
+      "learning_rate": 8.314250837974364e-05,
+      "loss": 1.8025,
+      "step": 2593
+    },
+    {
+      "epoch": 0.26945050379141994,
+      "grad_norm": 0.3620801270008087,
+      "learning_rate": 8.313028954873114e-05,
+      "loss": 1.725,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2695543783110003,
+      "grad_norm": 0.35369589924812317,
+      "learning_rate": 8.311806718960124e-05,
+      "loss": 1.5303,
+      "step": 2595
+    },
+    {
+      "epoch": 0.26965825283058065,
+      "grad_norm": 0.38566333055496216,
+      "learning_rate": 8.310584130365551e-05,
+      "loss": 1.9351,
+      "step": 2596
+    },
+    {
+      "epoch": 0.26976212735016103,
+      "grad_norm": 0.3703290820121765,
+      "learning_rate": 8.309361189219589e-05,
+      "loss": 1.6705,
+      "step": 2597
+    },
+    {
+      "epoch": 0.26986600186974136,
+      "grad_norm": 0.3818873167037964,
+      "learning_rate": 8.308137895652477e-05,
+      "loss": 1.7879,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2699698763893217,
+      "grad_norm": 0.3748812675476074,
+      "learning_rate": 8.306914249794483e-05,
+      "loss": 1.7799,
+      "step": 2599
+    },
+    {
+      "epoch": 0.27007375090890207,
+      "grad_norm": 0.3769199252128601,
+      "learning_rate": 8.305690251775915e-05,
+      "loss": 1.6268,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2701776254284824,
+      "grad_norm": 0.3615618944168091,
+      "learning_rate": 8.304465901727123e-05,
+      "loss": 1.7004,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2702814999480627,
+      "grad_norm": 0.36809107661247253,
+      "learning_rate": 8.303241199778486e-05,
+      "loss": 1.7739,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2703853744676431,
+      "grad_norm": 0.37990957498550415,
+      "learning_rate": 8.30201614606043e-05,
+      "loss": 1.5496,
+      "step": 2603
+    },
+    {
+      "epoch": 0.27048924898722343,
+      "grad_norm": 0.3890332579612732,
+      "learning_rate": 8.300790740703409e-05,
+      "loss": 1.7503,
+      "step": 2604
+    },
+    {
+      "epoch": 0.27059312350680376,
+      "grad_norm": 0.3829896152019501,
+      "learning_rate": 8.299564983837922e-05,
+      "loss": 1.7856,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27069699802638414,
+      "grad_norm": 0.36139553785324097,
+      "learning_rate": 8.298338875594501e-05,
+      "loss": 1.7488,
+      "step": 2606
+    },
+    {
+      "epoch": 0.27080087254596447,
+      "grad_norm": 0.3622409403324127,
+      "learning_rate": 8.29711241610372e-05,
+      "loss": 1.6934,
+      "step": 2607
+    },
+    {
+      "epoch": 0.27090474706554485,
+      "grad_norm": 0.37896421551704407,
+      "learning_rate": 8.295885605496184e-05,
+      "loss": 1.785,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2710086215851252,
+      "grad_norm": 0.36020195484161377,
+      "learning_rate": 8.294658443902539e-05,
+      "loss": 1.7482,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2711124961047055,
+      "grad_norm": 0.3811848759651184,
+      "learning_rate": 8.29343093145347e-05,
+      "loss": 1.7077,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2712163706242859,
+      "grad_norm": 0.3960614502429962,
+      "learning_rate": 8.292203068279695e-05,
+      "loss": 1.8574,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2713202451438662,
+      "grad_norm": 0.4237765371799469,
+      "learning_rate": 8.290974854511974e-05,
+      "loss": 1.7988,
+      "step": 2612
+    },
+    {
+      "epoch": 0.27142411966344654,
+      "grad_norm": 0.3800017237663269,
+      "learning_rate": 8.289746290281104e-05,
+      "loss": 1.6488,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2715279941830269,
+      "grad_norm": 0.38529491424560547,
+      "learning_rate": 8.288517375717913e-05,
+      "loss": 1.8813,
+      "step": 2614
+    },
+    {
+      "epoch": 0.27163186870260725,
+      "grad_norm": 0.3625027537345886,
+      "learning_rate": 8.287288110953274e-05,
+      "loss": 1.4855,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2717357432221876,
+      "grad_norm": 0.4243226647377014,
+      "learning_rate": 8.286058496118091e-05,
+      "loss": 1.921,
+      "step": 2616
+    },
+    {
+      "epoch": 0.27183961774176796,
+      "grad_norm": 0.36872926354408264,
+      "learning_rate": 8.284828531343312e-05,
+      "loss": 1.6206,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2719434922613483,
+      "grad_norm": 0.40654125809669495,
+      "learning_rate": 8.283598216759915e-05,
+      "loss": 1.8668,
+      "step": 2618
+    },
+    {
+      "epoch": 0.2720473667809286,
+      "grad_norm": 0.3534424602985382,
+      "learning_rate": 8.282367552498925e-05,
+      "loss": 1.6809,
+      "step": 2619
+    },
+    {
+      "epoch": 0.272151241300509,
+      "grad_norm": 0.41357290744781494,
+      "learning_rate": 8.28113653869139e-05,
+      "loss": 1.922,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2722551158200893,
+      "grad_norm": 0.41436415910720825,
+      "learning_rate": 8.27990517546841e-05,
+      "loss": 1.8944,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2723589903396697,
+      "grad_norm": 0.3990755081176758,
+      "learning_rate": 8.278673462961112e-05,
+      "loss": 1.8936,
+      "step": 2622
+    },
+    {
+      "epoch": 0.27246286485925003,
+      "grad_norm": 0.3946811258792877,
+      "learning_rate": 8.277441401300665e-05,
+      "loss": 1.9192,
+      "step": 2623
+    },
+    {
+      "epoch": 0.27256673937883036,
+      "grad_norm": 0.37876197695732117,
+      "learning_rate": 8.276208990618274e-05,
+      "loss": 1.6761,
+      "step": 2624
+    },
+    {
+      "epoch": 0.27267061389841074,
+      "grad_norm": 0.4105803072452545,
+      "learning_rate": 8.27497623104518e-05,
+      "loss": 1.7843,
+      "step": 2625
+    },
+    {
+      "epoch": 0.27277448841799107,
+      "grad_norm": 0.37790194153785706,
+      "learning_rate": 8.273743122712664e-05,
+      "loss": 1.5891,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2728783629375714,
+      "grad_norm": 0.43674150109291077,
+      "learning_rate": 8.272509665752041e-05,
+      "loss": 1.793,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2729822374571518,
+      "grad_norm": 0.45887747406959534,
+      "learning_rate": 8.271275860294667e-05,
+      "loss": 1.9712,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2730861119767321,
+      "grad_norm": 0.3662078082561493,
+      "learning_rate": 8.27004170647193e-05,
+      "loss": 1.6459,
+      "step": 2629
+    },
+    {
+      "epoch": 0.27318998649631243,
+      "grad_norm": 0.364083468914032,
+      "learning_rate": 8.268807204415258e-05,
+      "loss": 1.6445,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2732938610158928,
+      "grad_norm": 0.3766544759273529,
+      "learning_rate": 8.267572354256117e-05,
+      "loss": 1.6357,
+      "step": 2631
+    },
+    {
+      "epoch": 0.27339773553547314,
+      "grad_norm": 0.38634899258613586,
+      "learning_rate": 8.266337156126008e-05,
+      "loss": 1.7254,
+      "step": 2632
+    },
+    {
+      "epoch": 0.27350161005505347,
+      "grad_norm": 0.391665518283844,
+      "learning_rate": 8.26510161015647e-05,
+      "loss": 1.7073,
+      "step": 2633
+    },
+    {
+      "epoch": 0.27360548457463385,
+      "grad_norm": 0.5298207998275757,
+      "learning_rate": 8.263865716479078e-05,
+      "loss": 1.8689,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2737093590942142,
+      "grad_norm": 0.4072588086128235,
+      "learning_rate": 8.262629475225448e-05,
+      "loss": 1.8721,
+      "step": 2635
+    },
+    {
+      "epoch": 0.27381323361379456,
+      "grad_norm": 0.3996543288230896,
+      "learning_rate": 8.261392886527229e-05,
+      "loss": 1.7335,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2739171081333749,
+      "grad_norm": 0.414880633354187,
+      "learning_rate": 8.260155950516106e-05,
+      "loss": 1.8539,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2740209826529552,
+      "grad_norm": 0.36667972803115845,
+      "learning_rate": 8.258918667323806e-05,
+      "loss": 1.6214,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2741248571725356,
+      "grad_norm": 0.39150235056877136,
+      "learning_rate": 8.257681037082085e-05,
+      "loss": 1.7361,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2742287316921159,
+      "grad_norm": 0.38801634311676025,
+      "learning_rate": 8.256443059922749e-05,
+      "loss": 1.8081,
+      "step": 2640
+    },
+    {
+      "epoch": 0.27433260621169625,
+      "grad_norm": 0.42313215136528015,
+      "learning_rate": 8.255204735977626e-05,
+      "loss": 1.7,
+      "step": 2641
+    },
+    {
+      "epoch": 0.27443648073127663,
+      "grad_norm": 0.36863112449645996,
+      "learning_rate": 8.25396606537859e-05,
+      "loss": 1.5137,
+      "step": 2642
+    },
+    {
+      "epoch": 0.27454035525085696,
+      "grad_norm": 0.39456894993782043,
+      "learning_rate": 8.252727048257551e-05,
+      "loss": 1.7127,
+      "step": 2643
+    },
+    {
+      "epoch": 0.2746442297704373,
+      "grad_norm": 0.3853185474872589,
+      "learning_rate": 8.251487684746453e-05,
+      "loss": 1.8356,
+      "step": 2644
+    },
+    {
+      "epoch": 0.27474810429001767,
+      "grad_norm": 0.4087388813495636,
+      "learning_rate": 8.250247974977279e-05,
+      "loss": 1.6803,
+      "step": 2645
+    },
+    {
+      "epoch": 0.274851978809598,
+      "grad_norm": 0.3977982997894287,
+      "learning_rate": 8.249007919082048e-05,
+      "loss": 1.7868,
+      "step": 2646
+    },
+    {
+      "epoch": 0.2749558533291784,
+      "grad_norm": 0.3721441626548767,
+      "learning_rate": 8.247767517192818e-05,
+      "loss": 1.8521,
+      "step": 2647
+    },
+    {
+      "epoch": 0.2750597278487587,
+      "grad_norm": 0.3743959963321686,
+      "learning_rate": 8.246526769441681e-05,
+      "loss": 1.7785,
+      "step": 2648
+    },
+    {
+      "epoch": 0.27516360236833903,
+      "grad_norm": 0.39412543177604675,
+      "learning_rate": 8.245285675960767e-05,
+      "loss": 1.8763,
+      "step": 2649
+    },
+    {
+      "epoch": 0.2752674768879194,
+      "grad_norm": 0.3404049575328827,
+      "learning_rate": 8.244044236882242e-05,
+      "loss": 1.4576,
+      "step": 2650
+    },
+    {
+      "epoch": 0.27537135140749974,
+      "grad_norm": 0.3886807858943939,
+      "learning_rate": 8.24280245233831e-05,
+      "loss": 1.7468,
+      "step": 2651
+    },
+    {
+      "epoch": 0.27547522592708007,
+      "grad_norm": 0.3996857702732086,
+      "learning_rate": 8.241560322461212e-05,
+      "loss": 1.92,
+      "step": 2652
+    },
+    {
+      "epoch": 0.27557910044666045,
+      "grad_norm": 0.3869626224040985,
+      "learning_rate": 8.240317847383226e-05,
+      "loss": 1.6042,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2756829749662408,
+      "grad_norm": 0.4047796130180359,
+      "learning_rate": 8.239075027236663e-05,
+      "loss": 1.8099,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2757868494858211,
+      "grad_norm": 0.44502919912338257,
+      "learning_rate": 8.237831862153877e-05,
+      "loss": 1.7363,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2758907240054015,
+      "grad_norm": 0.388954222202301,
+      "learning_rate": 8.236588352267252e-05,
+      "loss": 1.8691,
+      "step": 2656
+    },
+    {
+      "epoch": 0.2759945985249818,
+      "grad_norm": 0.39591512084007263,
+      "learning_rate": 8.235344497709214e-05,
+      "loss": 1.8538,
+      "step": 2657
+    },
+    {
+      "epoch": 0.27609847304456214,
+      "grad_norm": 0.38203978538513184,
+      "learning_rate": 8.234100298612226e-05,
+      "loss": 1.28,
+      "step": 2658
+    },
+    {
+      "epoch": 0.2762023475641425,
+      "grad_norm": 0.43226009607315063,
+      "learning_rate": 8.23285575510878e-05,
+      "loss": 1.9326,
+      "step": 2659
+    },
+    {
+      "epoch": 0.27630622208372285,
+      "grad_norm": 0.37048643827438354,
+      "learning_rate": 8.231610867331415e-05,
+      "loss": 1.7221,
+      "step": 2660
+    },
+    {
+      "epoch": 0.27641009660330323,
+      "grad_norm": 0.41580793261528015,
+      "learning_rate": 8.2303656354127e-05,
+      "loss": 1.9031,
+      "step": 2661
+    },
+    {
+      "epoch": 0.27651397112288356,
+      "grad_norm": 0.3827223777770996,
+      "learning_rate": 8.229120059485243e-05,
+      "loss": 1.6825,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2766178456424639,
+      "grad_norm": 0.37424013018608093,
+      "learning_rate": 8.227874139681688e-05,
+      "loss": 1.7172,
+      "step": 2663
+    },
+    {
+      "epoch": 0.27672172016204427,
+      "grad_norm": 0.43709954619407654,
+      "learning_rate": 8.226627876134714e-05,
+      "loss": 1.8136,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2768255946816246,
+      "grad_norm": 0.35949236154556274,
+      "learning_rate": 8.22538126897704e-05,
+      "loss": 1.7346,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2769294692012049,
+      "grad_norm": 0.39692017436027527,
+      "learning_rate": 8.224134318341421e-05,
+      "loss": 1.925,
+      "step": 2666
+    },
+    {
+      "epoch": 0.2770333437207853,
+      "grad_norm": 0.3917725384235382,
+      "learning_rate": 8.222887024360646e-05,
+      "loss": 1.7681,
+      "step": 2667
+    },
+    {
+      "epoch": 0.27713721824036563,
+      "grad_norm": 0.4082455635070801,
+      "learning_rate": 8.221639387167543e-05,
+      "loss": 1.8118,
+      "step": 2668
+    },
+    {
+      "epoch": 0.27724109275994596,
+      "grad_norm": 0.36559662222862244,
+      "learning_rate": 8.220391406894973e-05,
+      "loss": 1.8121,
+      "step": 2669
+    },
+    {
+      "epoch": 0.27734496727952634,
+      "grad_norm": 0.372790664434433,
+      "learning_rate": 8.21914308367584e-05,
+      "loss": 1.7102,
+      "step": 2670
+    },
+    {
+      "epoch": 0.27744884179910667,
+      "grad_norm": 0.37069472670555115,
+      "learning_rate": 8.217894417643078e-05,
+      "loss": 1.6396,
+      "step": 2671
+    },
+    {
+      "epoch": 0.277552716318687,
+      "grad_norm": 0.38368502259254456,
+      "learning_rate": 8.216645408929661e-05,
+      "loss": 1.8278,
+      "step": 2672
+    },
+    {
+      "epoch": 0.2776565908382674,
+      "grad_norm": 0.4187244474887848,
+      "learning_rate": 8.215396057668598e-05,
+      "loss": 1.7567,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2777604653578477,
+      "grad_norm": 0.37204065918922424,
+      "learning_rate": 8.214146363992939e-05,
+      "loss": 1.7282,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2778643398774281,
+      "grad_norm": 0.37890782952308655,
+      "learning_rate": 8.21289632803576e-05,
+      "loss": 1.6675,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2779682143970084,
+      "grad_norm": 0.37642747163772583,
+      "learning_rate": 8.211645949930187e-05,
+      "loss": 1.6311,
+      "step": 2676
+    },
+    {
+      "epoch": 0.27807208891658874,
+      "grad_norm": 0.3810623288154602,
+      "learning_rate": 8.210395229809369e-05,
+      "loss": 1.7166,
+      "step": 2677
+    },
+    {
+      "epoch": 0.2781759634361691,
+      "grad_norm": 0.3708658516407013,
+      "learning_rate": 8.209144167806502e-05,
+      "loss": 1.7207,
+      "step": 2678
+    },
+    {
+      "epoch": 0.27827983795574945,
+      "grad_norm": 0.3930984139442444,
+      "learning_rate": 8.207892764054814e-05,
+      "loss": 1.7495,
+      "step": 2679
+    },
+    {
+      "epoch": 0.2783837124753298,
+      "grad_norm": 0.3427664339542389,
+      "learning_rate": 8.20664101868757e-05,
+      "loss": 1.6362,
+      "step": 2680
+    },
+    {
+      "epoch": 0.27848758699491016,
+      "grad_norm": 0.3461865484714508,
+      "learning_rate": 8.205388931838068e-05,
+      "loss": 1.594,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2785914615144905,
+      "grad_norm": 0.43719297647476196,
+      "learning_rate": 8.20413650363965e-05,
+      "loss": 1.9431,
+      "step": 2682
+    },
+    {
+      "epoch": 0.2786953360340708,
+      "grad_norm": 0.3867358863353729,
+      "learning_rate": 8.202883734225686e-05,
+      "loss": 1.7749,
+      "step": 2683
+    },
+    {
+      "epoch": 0.2787992105536512,
+      "grad_norm": 0.4017643928527832,
+      "learning_rate": 8.201630623729588e-05,
+      "loss": 1.8466,
+      "step": 2684
+    },
+    {
+      "epoch": 0.2789030850732315,
+      "grad_norm": 0.4061945974826813,
+      "learning_rate": 8.200377172284803e-05,
+      "loss": 2.0036,
+      "step": 2685
+    },
+    {
+      "epoch": 0.2790069595928119,
+      "grad_norm": 0.36883074045181274,
+      "learning_rate": 8.199123380024812e-05,
+      "loss": 1.6884,
+      "step": 2686
+    },
+    {
+      "epoch": 0.27911083411239224,
+      "grad_norm": 0.37533366680145264,
+      "learning_rate": 8.197869247083136e-05,
+      "loss": 1.6512,
+      "step": 2687
+    },
+    {
+      "epoch": 0.27921470863197256,
+      "grad_norm": 0.3572383522987366,
+      "learning_rate": 8.196614773593329e-05,
+      "loss": 1.6059,
+      "step": 2688
+    },
+    {
+      "epoch": 0.27931858315155295,
+      "grad_norm": 0.37231144309043884,
+      "learning_rate": 8.195359959688985e-05,
+      "loss": 1.6252,
+      "step": 2689
+    },
+    {
+      "epoch": 0.2794224576711333,
+      "grad_norm": 0.3484655022621155,
+      "learning_rate": 8.194104805503728e-05,
+      "loss": 1.7011,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2795263321907136,
+      "grad_norm": 0.3684258460998535,
+      "learning_rate": 8.192849311171225e-05,
+      "loss": 1.6864,
+      "step": 2691
+    },
+    {
+      "epoch": 0.279630206710294,
+      "grad_norm": 0.34496328234672546,
+      "learning_rate": 8.191593476825174e-05,
+      "loss": 1.5545,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2797340812298743,
+      "grad_norm": 0.3760117292404175,
+      "learning_rate": 8.190337302599314e-05,
+      "loss": 1.7473,
+      "step": 2693
+    },
+    {
+      "epoch": 0.27983795574945464,
+      "grad_norm": 0.3527992367744446,
+      "learning_rate": 8.189080788627419e-05,
+      "loss": 1.6099,
+      "step": 2694
+    },
+    {
+      "epoch": 0.279941830269035,
+      "grad_norm": 0.37164172530174255,
+      "learning_rate": 8.187823935043292e-05,
+      "loss": 1.6274,
+      "step": 2695
+    },
+    {
+      "epoch": 0.28004570478861535,
+      "grad_norm": 0.37998443841934204,
+      "learning_rate": 8.186566741980786e-05,
+      "loss": 1.8226,
+      "step": 2696
+    },
+    {
+      "epoch": 0.2801495793081957,
+      "grad_norm": 0.3952192962169647,
+      "learning_rate": 8.185309209573775e-05,
+      "loss": 1.6987,
+      "step": 2697
+    },
+    {
+      "epoch": 0.28025345382777606,
+      "grad_norm": 0.3859109878540039,
+      "learning_rate": 8.18405133795618e-05,
+      "loss": 1.7248,
+      "step": 2698
+    },
+    {
+      "epoch": 0.2803573283473564,
+      "grad_norm": 0.3739028573036194,
+      "learning_rate": 8.182793127261955e-05,
+      "loss": 1.8529,
+      "step": 2699
+    },
+    {
+      "epoch": 0.28046120286693677,
+      "grad_norm": 0.44212570786476135,
+      "learning_rate": 8.181534577625088e-05,
+      "loss": 1.8915,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2805650773865171,
+      "grad_norm": 0.3574367165565491,
+      "learning_rate": 8.180275689179607e-05,
+      "loss": 1.7146,
+      "step": 2701
+    },
+    {
+      "epoch": 0.2806689519060974,
+      "grad_norm": 0.36751818656921387,
+      "learning_rate": 8.17901646205957e-05,
+      "loss": 1.5943,
+      "step": 2702
+    },
+    {
+      "epoch": 0.2807728264256778,
+      "grad_norm": 0.3904435634613037,
+      "learning_rate": 8.177756896399078e-05,
+      "loss": 1.7849,
+      "step": 2703
+    },
+    {
+      "epoch": 0.28087670094525813,
+      "grad_norm": 0.38835206627845764,
+      "learning_rate": 8.176496992332265e-05,
+      "loss": 1.7837,
+      "step": 2704
+    },
+    {
+      "epoch": 0.28098057546483846,
+      "grad_norm": 0.46619629859924316,
+      "learning_rate": 8.175236749993298e-05,
+      "loss": 1.9918,
+      "step": 2705
+    },
+    {
+      "epoch": 0.28108444998441884,
+      "grad_norm": 0.412544846534729,
+      "learning_rate": 8.173976169516386e-05,
+      "loss": 1.8577,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28118832450399917,
+      "grad_norm": 0.36168932914733887,
+      "learning_rate": 8.17271525103577e-05,
+      "loss": 1.6675,
+      "step": 2707
+    },
+    {
+      "epoch": 0.2812921990235795,
+      "grad_norm": 0.4263473451137543,
+      "learning_rate": 8.171453994685728e-05,
+      "loss": 1.8988,
+      "step": 2708
+    },
+    {
+      "epoch": 0.2813960735431599,
+      "grad_norm": 0.369133323431015,
+      "learning_rate": 8.170192400600574e-05,
+      "loss": 1.6811,
+      "step": 2709
+    },
+    {
+      "epoch": 0.2814999480627402,
+      "grad_norm": 0.39230549335479736,
+      "learning_rate": 8.168930468914658e-05,
+      "loss": 1.5888,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2816038225823206,
+      "grad_norm": 0.3522338569164276,
+      "learning_rate": 8.167668199762364e-05,
+      "loss": 1.5904,
+      "step": 2711
+    },
+    {
+      "epoch": 0.2817076971019009,
+      "grad_norm": 0.3948661983013153,
+      "learning_rate": 8.166405593278116e-05,
+      "loss": 1.7703,
+      "step": 2712
+    },
+    {
+      "epoch": 0.28181157162148124,
+      "grad_norm": 0.35214731097221375,
+      "learning_rate": 8.165142649596372e-05,
+      "loss": 1.6956,
+      "step": 2713
+    },
+    {
+      "epoch": 0.2819154461410616,
+      "grad_norm": 0.4061570167541504,
+      "learning_rate": 8.163879368851623e-05,
+      "loss": 1.9759,
+      "step": 2714
+    },
+    {
+      "epoch": 0.28201932066064195,
+      "grad_norm": 0.3767627477645874,
+      "learning_rate": 8.162615751178405e-05,
+      "loss": 1.6924,
+      "step": 2715
+    },
+    {
+      "epoch": 0.2821231951802223,
+      "grad_norm": 0.37810075283050537,
+      "learning_rate": 8.161351796711274e-05,
+      "loss": 1.8406,
+      "step": 2716
+    },
+    {
+      "epoch": 0.28222706969980266,
+      "grad_norm": 0.37295323610305786,
+      "learning_rate": 8.16008750558484e-05,
+      "loss": 1.7651,
+      "step": 2717
+    },
+    {
+      "epoch": 0.282330944219383,
+      "grad_norm": 0.3852209448814392,
+      "learning_rate": 8.158822877933733e-05,
+      "loss": 1.7538,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2824348187389633,
+      "grad_norm": 0.36513835191726685,
+      "learning_rate": 8.15755791389263e-05,
+      "loss": 1.7825,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2825386932585437,
+      "grad_norm": 0.3820558190345764,
+      "learning_rate": 8.15629261359624e-05,
+      "loss": 1.737,
+      "step": 2720
+    },
+    {
+      "epoch": 0.282642567778124,
+      "grad_norm": 0.3873055875301361,
+      "learning_rate": 8.155026977179307e-05,
+      "loss": 1.774,
+      "step": 2721
+    },
+    {
+      "epoch": 0.28274644229770435,
+      "grad_norm": 0.3791974186897278,
+      "learning_rate": 8.15376100477661e-05,
+      "loss": 1.7034,
+      "step": 2722
+    },
+    {
+      "epoch": 0.28285031681728473,
+      "grad_norm": 0.35163450241088867,
+      "learning_rate": 8.152494696522965e-05,
+      "loss": 1.5747,
+      "step": 2723
+    },
+    {
+      "epoch": 0.28295419133686506,
+      "grad_norm": 0.390493780374527,
+      "learning_rate": 8.151228052553228e-05,
+      "loss": 1.7802,
+      "step": 2724
+    },
+    {
+      "epoch": 0.28305806585644544,
+      "grad_norm": 0.3626655340194702,
+      "learning_rate": 8.149961073002283e-05,
+      "loss": 1.7353,
+      "step": 2725
+    },
+    {
+      "epoch": 0.28316194037602577,
+      "grad_norm": 0.40224525332450867,
+      "learning_rate": 8.148693758005054e-05,
+      "loss": 1.8108,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2832658148956061,
+      "grad_norm": 0.3655959367752075,
+      "learning_rate": 8.1474261076965e-05,
+      "loss": 1.7521,
+      "step": 2727
+    },
+    {
+      "epoch": 0.2833696894151865,
+      "grad_norm": 0.3598358929157257,
+      "learning_rate": 8.146158122211618e-05,
+      "loss": 1.6139,
+      "step": 2728
+    },
+    {
+      "epoch": 0.2834735639347668,
+      "grad_norm": 0.4097810983657837,
+      "learning_rate": 8.144889801685436e-05,
+      "loss": 1.9751,
+      "step": 2729
+    },
+    {
+      "epoch": 0.28357743845434713,
+      "grad_norm": 0.3773553967475891,
+      "learning_rate": 8.143621146253022e-05,
+      "loss": 1.7618,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2836813129739275,
+      "grad_norm": 0.3876679837703705,
+      "learning_rate": 8.142352156049475e-05,
+      "loss": 1.7756,
+      "step": 2731
+    },
+    {
+      "epoch": 0.28378518749350784,
+      "grad_norm": 0.3985688090324402,
+      "learning_rate": 8.141082831209938e-05,
+      "loss": 1.8868,
+      "step": 2732
+    },
+    {
+      "epoch": 0.28388906201308817,
+      "grad_norm": 0.40376704931259155,
+      "learning_rate": 8.139813171869579e-05,
+      "loss": 1.9038,
+      "step": 2733
+    },
+    {
+      "epoch": 0.28399293653266855,
+      "grad_norm": 0.333291620016098,
+      "learning_rate": 8.138543178163607e-05,
+      "loss": 1.5568,
+      "step": 2734
+    },
+    {
+      "epoch": 0.2840968110522489,
+      "grad_norm": 0.39361095428466797,
+      "learning_rate": 8.137272850227271e-05,
+      "loss": 1.6809,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2842006855718292,
+      "grad_norm": 0.39014434814453125,
+      "learning_rate": 8.136002188195846e-05,
+      "loss": 1.7479,
+      "step": 2736
+    },
+    {
+      "epoch": 0.2843045600914096,
+      "grad_norm": 0.4338228404521942,
+      "learning_rate": 8.13473119220465e-05,
+      "loss": 1.8751,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2844084346109899,
+      "grad_norm": 0.44359880685806274,
+      "learning_rate": 8.133459862389036e-05,
+      "loss": 1.8787,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2845123091305703,
+      "grad_norm": 0.42322838306427,
+      "learning_rate": 8.132188198884386e-05,
+      "loss": 1.827,
+      "step": 2739
+    },
+    {
+      "epoch": 0.2846161836501506,
+      "grad_norm": 0.363816499710083,
+      "learning_rate": 8.130916201826126e-05,
+      "loss": 1.5635,
+      "step": 2740
+    },
+    {
+      "epoch": 0.28472005816973095,
+      "grad_norm": 0.39882132411003113,
+      "learning_rate": 8.129643871349712e-05,
+      "loss": 2.0049,
+      "step": 2741
+    },
+    {
+      "epoch": 0.28482393268931133,
+      "grad_norm": 0.3964424729347229,
+      "learning_rate": 8.128371207590639e-05,
+      "loss": 1.6739,
+      "step": 2742
+    },
+    {
+      "epoch": 0.28492780720889166,
+      "grad_norm": 0.42000818252563477,
+      "learning_rate": 8.127098210684434e-05,
+      "loss": 1.9091,
+      "step": 2743
+    },
+    {
+      "epoch": 0.285031681728472,
+      "grad_norm": 0.3911057412624359,
+      "learning_rate": 8.125824880766661e-05,
+      "loss": 1.7084,
+      "step": 2744
+    },
+    {
+      "epoch": 0.28513555624805237,
+      "grad_norm": 0.37378305196762085,
+      "learning_rate": 8.124551217972922e-05,
+      "loss": 1.7252,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2852394307676327,
+      "grad_norm": 0.40639790892601013,
+      "learning_rate": 8.12327722243885e-05,
+      "loss": 1.8441,
+      "step": 2746
+    },
+    {
+      "epoch": 0.285343305287213,
+      "grad_norm": 0.3873542249202728,
+      "learning_rate": 8.122002894300117e-05,
+      "loss": 1.506,
+      "step": 2747
+    },
+    {
+      "epoch": 0.2854471798067934,
+      "grad_norm": 0.34335261583328247,
+      "learning_rate": 8.120728233692428e-05,
+      "loss": 1.6598,
+      "step": 2748
+    },
+    {
+      "epoch": 0.28555105432637373,
+      "grad_norm": 0.4100406765937805,
+      "learning_rate": 8.119453240751525e-05,
+      "loss": 1.6315,
+      "step": 2749
+    },
+    {
+      "epoch": 0.2856549288459541,
+      "grad_norm": 0.34580525755882263,
+      "learning_rate": 8.118177915613182e-05,
+      "loss": 1.562,
+      "step": 2750
+    },
+    {
+      "epoch": 0.28575880336553444,
+      "grad_norm": 0.39648303389549255,
+      "learning_rate": 8.116902258413216e-05,
+      "loss": 1.9186,
+      "step": 2751
+    },
+    {
+      "epoch": 0.28586267788511477,
+      "grad_norm": 0.3890984356403351,
+      "learning_rate": 8.115626269287473e-05,
+      "loss": 1.862,
+      "step": 2752
+    },
+    {
+      "epoch": 0.28596655240469515,
+      "grad_norm": 0.3671417236328125,
+      "learning_rate": 8.114349948371833e-05,
+      "loss": 1.7655,
+      "step": 2753
+    },
+    {
+      "epoch": 0.2860704269242755,
+      "grad_norm": 0.3553464412689209,
+      "learning_rate": 8.113073295802219e-05,
+      "loss": 1.5148,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2861743014438558,
+      "grad_norm": 0.3629017472267151,
+      "learning_rate": 8.111796311714579e-05,
+      "loss": 1.5992,
+      "step": 2755
+    },
+    {
+      "epoch": 0.2862781759634362,
+      "grad_norm": 0.4407372772693634,
+      "learning_rate": 8.110518996244907e-05,
+      "loss": 1.7642,
+      "step": 2756
+    },
+    {
+      "epoch": 0.2863820504830165,
+      "grad_norm": 0.42658326029777527,
+      "learning_rate": 8.109241349529223e-05,
+      "loss": 1.9091,
+      "step": 2757
+    },
+    {
+      "epoch": 0.28648592500259684,
+      "grad_norm": 0.38717007637023926,
+      "learning_rate": 8.107963371703589e-05,
+      "loss": 1.7881,
+      "step": 2758
+    },
+    {
+      "epoch": 0.2865897995221772,
+      "grad_norm": 0.39365988969802856,
+      "learning_rate": 8.1066850629041e-05,
+      "loss": 1.6742,
+      "step": 2759
+    },
+    {
+      "epoch": 0.28669367404175755,
+      "grad_norm": 0.3701498806476593,
+      "learning_rate": 8.105406423266884e-05,
+      "loss": 1.6718,
+      "step": 2760
+    },
+    {
+      "epoch": 0.2867975485613379,
+      "grad_norm": 0.3934311866760254,
+      "learning_rate": 8.104127452928107e-05,
+      "loss": 1.7174,
+      "step": 2761
+    },
+    {
+      "epoch": 0.28690142308091826,
+      "grad_norm": 0.3584936261177063,
+      "learning_rate": 8.10284815202397e-05,
+      "loss": 1.7269,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2870052976004986,
+      "grad_norm": 0.4173574447631836,
+      "learning_rate": 8.101568520690708e-05,
+      "loss": 1.7251,
+      "step": 2763
+    },
+    {
+      "epoch": 0.28710917212007897,
+      "grad_norm": 0.3903685212135315,
+      "learning_rate": 8.10028855906459e-05,
+      "loss": 1.7643,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2872130466396593,
+      "grad_norm": 0.37448278069496155,
+      "learning_rate": 8.099008267281924e-05,
+      "loss": 1.7648,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2873169211592396,
+      "grad_norm": 0.3669279217720032,
+      "learning_rate": 8.097727645479053e-05,
+      "loss": 1.7988,
+      "step": 2766
+    },
+    {
+      "epoch": 0.28742079567882,
+      "grad_norm": 0.4136315584182739,
+      "learning_rate": 8.096446693792347e-05,
+      "loss": 1.857,
+      "step": 2767
+    },
+    {
+      "epoch": 0.28752467019840033,
+      "grad_norm": 0.33900344371795654,
+      "learning_rate": 8.095165412358225e-05,
+      "loss": 1.5582,
+      "step": 2768
+    },
+    {
+      "epoch": 0.28762854471798066,
+      "grad_norm": 0.36668214201927185,
+      "learning_rate": 8.093883801313129e-05,
+      "loss": 1.6839,
+      "step": 2769
+    },
+    {
+      "epoch": 0.28773241923756104,
+      "grad_norm": 0.4014098048210144,
+      "learning_rate": 8.09260186079354e-05,
+      "loss": 1.7254,
+      "step": 2770
+    },
+    {
+      "epoch": 0.28783629375714137,
+      "grad_norm": 0.44136208295822144,
+      "learning_rate": 8.091319590935978e-05,
+      "loss": 1.6938,
+      "step": 2771
+    },
+    {
+      "epoch": 0.2879401682767217,
+      "grad_norm": 0.42486312985420227,
+      "learning_rate": 8.090036991876992e-05,
+      "loss": 1.7066,
+      "step": 2772
+    },
+    {
+      "epoch": 0.2880440427963021,
+      "grad_norm": 0.4575170576572418,
+      "learning_rate": 8.08875406375317e-05,
+      "loss": 2.0641,
+      "step": 2773
+    },
+    {
+      "epoch": 0.2881479173158824,
+      "grad_norm": 0.4238605499267578,
+      "learning_rate": 8.087470806701131e-05,
+      "loss": 1.9339,
+      "step": 2774
+    },
+    {
+      "epoch": 0.28825179183546273,
+      "grad_norm": 0.37625259160995483,
+      "learning_rate": 8.086187220857537e-05,
+      "loss": 1.6855,
+      "step": 2775
+    },
+    {
+      "epoch": 0.2883556663550431,
+      "grad_norm": 0.3760613799095154,
+      "learning_rate": 8.084903306359078e-05,
+      "loss": 1.786,
+      "step": 2776
+    },
+    {
+      "epoch": 0.28845954087462344,
+      "grad_norm": 0.4117826521396637,
+      "learning_rate": 8.083619063342477e-05,
+      "loss": 1.796,
+      "step": 2777
+    },
+    {
+      "epoch": 0.2885634153942038,
+      "grad_norm": 0.3975922465324402,
+      "learning_rate": 8.0823344919445e-05,
+      "loss": 1.8309,
+      "step": 2778
+    },
+    {
+      "epoch": 0.28866728991378415,
+      "grad_norm": 0.39822137355804443,
+      "learning_rate": 8.081049592301944e-05,
+      "loss": 1.8686,
+      "step": 2779
+    },
+    {
+      "epoch": 0.2887711644333645,
+      "grad_norm": 0.3700173795223236,
+      "learning_rate": 8.079764364551638e-05,
+      "loss": 1.7419,
+      "step": 2780
+    },
+    {
+      "epoch": 0.28887503895294486,
+      "grad_norm": 0.3475184440612793,
+      "learning_rate": 8.07847880883045e-05,
+      "loss": 1.478,
+      "step": 2781
+    },
+    {
+      "epoch": 0.2889789134725252,
+      "grad_norm": 0.37445908784866333,
+      "learning_rate": 8.077192925275283e-05,
+      "loss": 1.8183,
+      "step": 2782
+    },
+    {
+      "epoch": 0.2890827879921055,
+      "grad_norm": 0.3727985918521881,
+      "learning_rate": 8.075906714023073e-05,
+      "loss": 1.7757,
+      "step": 2783
+    },
+    {
+      "epoch": 0.2891866625116859,
+      "grad_norm": 0.36759766936302185,
+      "learning_rate": 8.07462017521079e-05,
+      "loss": 1.6917,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2892905370312662,
+      "grad_norm": 0.39550402760505676,
+      "learning_rate": 8.073333308975442e-05,
+      "loss": 1.7787,
+      "step": 2785
+    },
+    {
+      "epoch": 0.28939441155084655,
+      "grad_norm": 0.3845430612564087,
+      "learning_rate": 8.072046115454069e-05,
+      "loss": 1.9255,
+      "step": 2786
+    },
+    {
+      "epoch": 0.28949828607042694,
+      "grad_norm": 0.39458414912223816,
+      "learning_rate": 8.070758594783748e-05,
+      "loss": 1.6905,
+      "step": 2787
+    },
+    {
+      "epoch": 0.28960216059000726,
+      "grad_norm": 0.38234516978263855,
+      "learning_rate": 8.069470747101588e-05,
+      "loss": 1.784,
+      "step": 2788
+    },
+    {
+      "epoch": 0.28970603510958765,
+      "grad_norm": 0.4639817774295807,
+      "learning_rate": 8.068182572544736e-05,
+      "loss": 1.7567,
+      "step": 2789
+    },
+    {
+      "epoch": 0.28980990962916797,
+      "grad_norm": 0.43574076890945435,
+      "learning_rate": 8.066894071250375e-05,
+      "loss": 1.8726,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2899137841487483,
+      "grad_norm": 0.37440717220306396,
+      "learning_rate": 8.065605243355717e-05,
+      "loss": 1.7629,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2900176586683287,
+      "grad_norm": 0.4101695716381073,
+      "learning_rate": 8.064316088998012e-05,
+      "loss": 1.7499,
+      "step": 2792
+    },
+    {
+      "epoch": 0.290121533187909,
+      "grad_norm": 0.3932081162929535,
+      "learning_rate": 8.063026608314547e-05,
+      "loss": 1.7148,
+      "step": 2793
+    },
+    {
+      "epoch": 0.29022540770748934,
+      "grad_norm": 0.37535983324050903,
+      "learning_rate": 8.061736801442643e-05,
+      "loss": 1.8031,
+      "step": 2794
+    },
+    {
+      "epoch": 0.2903292822270697,
+      "grad_norm": 0.3890056610107422,
+      "learning_rate": 8.06044666851965e-05,
+      "loss": 1.718,
+      "step": 2795
+    },
+    {
+      "epoch": 0.29043315674665005,
+      "grad_norm": 0.44528940320014954,
+      "learning_rate": 8.059156209682959e-05,
+      "loss": 1.7655,
+      "step": 2796
+    },
+    {
+      "epoch": 0.29053703126623037,
+      "grad_norm": 0.3969075679779053,
+      "learning_rate": 8.057865425069995e-05,
+      "loss": 1.7541,
+      "step": 2797
+    },
+    {
+      "epoch": 0.29064090578581075,
+      "grad_norm": 0.37456732988357544,
+      "learning_rate": 8.056574314818215e-05,
+      "loss": 1.747,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2907447803053911,
+      "grad_norm": 0.3638935387134552,
+      "learning_rate": 8.055282879065114e-05,
+      "loss": 1.6764,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2908486548249714,
+      "grad_norm": 0.3996514081954956,
+      "learning_rate": 8.053991117948216e-05,
+      "loss": 1.7869,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2909525293445518,
+      "grad_norm": 0.3886312246322632,
+      "learning_rate": 8.052699031605087e-05,
+      "loss": 1.8845,
+      "step": 2801
+    },
+    {
+      "epoch": 0.2910564038641321,
+      "grad_norm": 0.37657225131988525,
+      "learning_rate": 8.051406620173323e-05,
+      "loss": 1.6482,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2911602783837125,
+      "grad_norm": 0.3608708083629608,
+      "learning_rate": 8.050113883790556e-05,
+      "loss": 1.6466,
+      "step": 2803
+    },
+    {
+      "epoch": 0.29126415290329283,
+      "grad_norm": 0.41230741143226624,
+      "learning_rate": 8.048820822594452e-05,
+      "loss": 1.8663,
+      "step": 2804
+    },
+    {
+      "epoch": 0.29136802742287315,
+      "grad_norm": 0.3793978989124298,
+      "learning_rate": 8.047527436722713e-05,
+      "loss": 1.7933,
+      "step": 2805
+    },
+    {
+      "epoch": 0.29147190194245354,
+      "grad_norm": 0.35737892985343933,
+      "learning_rate": 8.046233726313073e-05,
+      "loss": 1.6788,
+      "step": 2806
+    },
+    {
+      "epoch": 0.29157577646203386,
+      "grad_norm": 0.4132479429244995,
+      "learning_rate": 8.044939691503304e-05,
+      "loss": 1.7392,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2916796509816142,
+      "grad_norm": 0.4083154499530792,
+      "learning_rate": 8.043645332431206e-05,
+      "loss": 1.6476,
+      "step": 2808
+    },
+    {
+      "epoch": 0.2917835255011946,
+      "grad_norm": 0.38581231236457825,
+      "learning_rate": 8.042350649234625e-05,
+      "loss": 1.7906,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2918874000207749,
+      "grad_norm": 0.4001144766807556,
+      "learning_rate": 8.041055642051429e-05,
+      "loss": 1.5869,
+      "step": 2810
+    },
+    {
+      "epoch": 0.29199127454035523,
+      "grad_norm": 0.3624385893344879,
+      "learning_rate": 8.039760311019529e-05,
+      "loss": 1.7447,
+      "step": 2811
+    },
+    {
+      "epoch": 0.2920951490599356,
+      "grad_norm": 0.3716486394405365,
+      "learning_rate": 8.038464656276867e-05,
+      "loss": 1.8424,
+      "step": 2812
+    },
+    {
+      "epoch": 0.29219902357951594,
+      "grad_norm": 0.39141982793807983,
+      "learning_rate": 8.03716867796142e-05,
+      "loss": 1.6962,
+      "step": 2813
+    },
+    {
+      "epoch": 0.29230289809909626,
+      "grad_norm": 0.3740139901638031,
+      "learning_rate": 8.035872376211199e-05,
+      "loss": 1.6874,
+      "step": 2814
+    },
+    {
+      "epoch": 0.29240677261867665,
+      "grad_norm": 0.464403361082077,
+      "learning_rate": 8.03457575116425e-05,
+      "loss": 1.8638,
+      "step": 2815
+    },
+    {
+      "epoch": 0.292510647138257,
+      "grad_norm": 0.3929448425769806,
+      "learning_rate": 8.033278802958656e-05,
+      "loss": 1.808,
+      "step": 2816
+    },
+    {
+      "epoch": 0.29261452165783736,
+      "grad_norm": 0.4005575180053711,
+      "learning_rate": 8.031981531732529e-05,
+      "loss": 1.6983,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2927183961774177,
+      "grad_norm": 0.3648793697357178,
+      "learning_rate": 8.03068393762402e-05,
+      "loss": 1.7543,
+      "step": 2818
+    },
+    {
+      "epoch": 0.292822270696998,
+      "grad_norm": 0.3590324819087982,
+      "learning_rate": 8.02938602077131e-05,
+      "loss": 1.8414,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2929261452165784,
+      "grad_norm": 0.3629642426967621,
+      "learning_rate": 8.02808778131262e-05,
+      "loss": 1.8325,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2930300197361587,
+      "grad_norm": 0.42311742901802063,
+      "learning_rate": 8.026789219386201e-05,
+      "loss": 1.7774,
+      "step": 2821
+    },
+    {
+      "epoch": 0.29313389425573905,
+      "grad_norm": 0.37744253873825073,
+      "learning_rate": 8.025490335130341e-05,
+      "loss": 1.7133,
+      "step": 2822
+    },
+    {
+      "epoch": 0.29323776877531943,
+      "grad_norm": 0.3763290047645569,
+      "learning_rate": 8.02419112868336e-05,
+      "loss": 1.7046,
+      "step": 2823
+    },
+    {
+      "epoch": 0.29334164329489976,
+      "grad_norm": 0.3736268877983093,
+      "learning_rate": 8.022891600183613e-05,
+      "loss": 1.6936,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2934455178144801,
+      "grad_norm": 0.3851562440395355,
+      "learning_rate": 8.021591749769489e-05,
+      "loss": 1.7645,
+      "step": 2825
+    },
+    {
+      "epoch": 0.29354939233406047,
+      "grad_norm": 0.4103233814239502,
+      "learning_rate": 8.020291577579413e-05,
+      "loss": 1.8886,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2936532668536408,
+      "grad_norm": 0.38178038597106934,
+      "learning_rate": 8.018991083751846e-05,
+      "loss": 1.7208,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2937571413732212,
+      "grad_norm": 0.36348870396614075,
+      "learning_rate": 8.017690268425276e-05,
+      "loss": 1.6585,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2938610158928015,
+      "grad_norm": 0.3771371841430664,
+      "learning_rate": 8.016389131738229e-05,
+      "loss": 1.7227,
+      "step": 2829
+    },
+    {
+      "epoch": 0.29396489041238183,
+      "grad_norm": 0.38507938385009766,
+      "learning_rate": 8.015087673829271e-05,
+      "loss": 1.8273,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2940687649319622,
+      "grad_norm": 0.3727617561817169,
+      "learning_rate": 8.013785894836993e-05,
+      "loss": 1.6725,
+      "step": 2831
+    },
+    {
+      "epoch": 0.29417263945154254,
+      "grad_norm": 0.37300458550453186,
+      "learning_rate": 8.012483794900026e-05,
+      "loss": 1.729,
+      "step": 2832
+    },
+    {
+      "epoch": 0.29427651397112287,
+      "grad_norm": 0.4091015160083771,
+      "learning_rate": 8.011181374157034e-05,
+      "loss": 1.6771,
+      "step": 2833
+    },
+    {
+      "epoch": 0.29438038849070325,
+      "grad_norm": 0.4204038083553314,
+      "learning_rate": 8.009878632746712e-05,
+      "loss": 1.8316,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2944842630102836,
+      "grad_norm": 0.4067396819591522,
+      "learning_rate": 8.008575570807795e-05,
+      "loss": 1.7071,
+      "step": 2835
+    },
+    {
+      "epoch": 0.2945881375298639,
+      "grad_norm": 0.4235256314277649,
+      "learning_rate": 8.007272188479049e-05,
+      "loss": 1.8557,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2946920120494443,
+      "grad_norm": 0.3819893002510071,
+      "learning_rate": 8.005968485899269e-05,
+      "loss": 1.7695,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2947958865690246,
+      "grad_norm": 0.3644767999649048,
+      "learning_rate": 8.004664463207294e-05,
+      "loss": 1.6902,
+      "step": 2838
+    },
+    {
+      "epoch": 0.29489976108860494,
+      "grad_norm": 0.36920779943466187,
+      "learning_rate": 8.003360120541992e-05,
+      "loss": 1.7109,
+      "step": 2839
+    },
+    {
+      "epoch": 0.2950036356081853,
+      "grad_norm": 0.3806985318660736,
+      "learning_rate": 8.002055458042264e-05,
+      "loss": 1.8152,
+      "step": 2840
+    },
+    {
+      "epoch": 0.29510751012776565,
+      "grad_norm": 0.4298724830150604,
+      "learning_rate": 8.000750475847045e-05,
+      "loss": 1.8906,
+      "step": 2841
+    },
+    {
+      "epoch": 0.29521138464734603,
+      "grad_norm": 0.3639010787010193,
+      "learning_rate": 7.999445174095308e-05,
+      "loss": 1.6325,
+      "step": 2842
+    },
+    {
+      "epoch": 0.29531525916692636,
+      "grad_norm": 0.36708855628967285,
+      "learning_rate": 7.998139552926057e-05,
+      "loss": 1.6156,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2954191336865067,
+      "grad_norm": 0.3640385866165161,
+      "learning_rate": 7.996833612478331e-05,
+      "loss": 1.71,
+      "step": 2844
+    },
+    {
+      "epoch": 0.29552300820608707,
+      "grad_norm": 0.44802719354629517,
+      "learning_rate": 7.9955273528912e-05,
+      "loss": 1.7186,
+      "step": 2845
+    },
+    {
+      "epoch": 0.2956268827256674,
+      "grad_norm": 0.37511417269706726,
+      "learning_rate": 7.994220774303771e-05,
+      "loss": 1.5034,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2957307572452477,
+      "grad_norm": 0.36259564757347107,
+      "learning_rate": 7.992913876855188e-05,
+      "loss": 1.6271,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2958346317648281,
+      "grad_norm": 0.3824392855167389,
+      "learning_rate": 7.99160666068462e-05,
+      "loss": 1.7693,
+      "step": 2848
+    },
+    {
+      "epoch": 0.29593850628440843,
+      "grad_norm": 0.41231241822242737,
+      "learning_rate": 7.99029912593128e-05,
+      "loss": 1.5047,
+      "step": 2849
+    },
+    {
+      "epoch": 0.29604238080398876,
+      "grad_norm": 0.3836499750614166,
+      "learning_rate": 7.988991272734407e-05,
+      "loss": 1.6688,
+      "step": 2850
+    },
+    {
+      "epoch": 0.29614625532356914,
+      "grad_norm": 0.4020180106163025,
+      "learning_rate": 7.987683101233279e-05,
+      "loss": 1.7687,
+      "step": 2851
+    },
+    {
+      "epoch": 0.29625012984314947,
+      "grad_norm": 0.3831713795661926,
+      "learning_rate": 7.986374611567207e-05,
+      "loss": 1.7141,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2963540043627298,
+      "grad_norm": 0.44648897647857666,
+      "learning_rate": 7.985065803875532e-05,
+      "loss": 1.9484,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2964578788823102,
+      "grad_norm": 0.36257603764533997,
+      "learning_rate": 7.983756678297633e-05,
+      "loss": 1.6849,
+      "step": 2854
+    },
+    {
+      "epoch": 0.2965617534018905,
+      "grad_norm": 0.40492746233940125,
+      "learning_rate": 7.982447234972922e-05,
+      "loss": 1.7369,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2966656279214709,
+      "grad_norm": 0.37346115708351135,
+      "learning_rate": 7.981137474040845e-05,
+      "loss": 1.7769,
+      "step": 2856
+    },
+    {
+      "epoch": 0.2967695024410512,
+      "grad_norm": 0.3808712959289551,
+      "learning_rate": 7.979827395640883e-05,
+      "loss": 1.6387,
+      "step": 2857
+    },
+    {
+      "epoch": 0.29687337696063154,
+      "grad_norm": 0.3708186149597168,
+      "learning_rate": 7.978516999912544e-05,
+      "loss": 1.7448,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2969772514802119,
+      "grad_norm": 0.3663538992404938,
+      "learning_rate": 7.977206286995379e-05,
+      "loss": 1.7705,
+      "step": 2859
+    },
+    {
+      "epoch": 0.29708112599979225,
+      "grad_norm": 0.41986000537872314,
+      "learning_rate": 7.975895257028965e-05,
+      "loss": 1.9017,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2971850005193726,
+      "grad_norm": 0.36581292748451233,
+      "learning_rate": 7.974583910152922e-05,
+      "loss": 1.5924,
+      "step": 2861
+    },
+    {
+      "epoch": 0.29728887503895296,
+      "grad_norm": 0.38697493076324463,
+      "learning_rate": 7.973272246506893e-05,
+      "loss": 1.7093,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2973927495585333,
+      "grad_norm": 0.38015836477279663,
+      "learning_rate": 7.971960266230565e-05,
+      "loss": 1.7662,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2974966240781136,
+      "grad_norm": 0.37382447719573975,
+      "learning_rate": 7.97064796946365e-05,
+      "loss": 1.6753,
+      "step": 2864
+    },
+    {
+      "epoch": 0.297600498597694,
+      "grad_norm": 0.42534875869750977,
+      "learning_rate": 7.969335356345894e-05,
+      "loss": 2.0455,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2977043731172743,
+      "grad_norm": 0.39144277572631836,
+      "learning_rate": 7.968022427017088e-05,
+      "loss": 1.4737,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2978082476368547,
+      "grad_norm": 0.3727049231529236,
+      "learning_rate": 7.966709181617045e-05,
+      "loss": 1.8034,
+      "step": 2867
+    },
+    {
+      "epoch": 0.29791212215643503,
+      "grad_norm": 0.3644360601902008,
+      "learning_rate": 7.965395620285616e-05,
+      "loss": 1.7262,
+      "step": 2868
+    },
+    {
+      "epoch": 0.29801599667601536,
+      "grad_norm": 0.40929368138313293,
+      "learning_rate": 7.964081743162684e-05,
+      "loss": 1.8834,
+      "step": 2869
+    },
+    {
+      "epoch": 0.29811987119559574,
+      "grad_norm": 0.40274330973625183,
+      "learning_rate": 7.962767550388166e-05,
+      "loss": 1.6383,
+      "step": 2870
+    },
+    {
+      "epoch": 0.29822374571517607,
+      "grad_norm": 0.38310062885284424,
+      "learning_rate": 7.961453042102014e-05,
+      "loss": 1.6704,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2983276202347564,
+      "grad_norm": 0.3942394554615021,
+      "learning_rate": 7.960138218444215e-05,
+      "loss": 1.8623,
+      "step": 2872
+    },
+    {
+      "epoch": 0.2984314947543368,
+      "grad_norm": 0.3523310124874115,
+      "learning_rate": 7.958823079554785e-05,
+      "loss": 1.7932,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2985353692739171,
+      "grad_norm": 0.36860281229019165,
+      "learning_rate": 7.957507625573776e-05,
+      "loss": 1.7049,
+      "step": 2874
+    },
+    {
+      "epoch": 0.29863924379349743,
+      "grad_norm": 0.37940865755081177,
+      "learning_rate": 7.956191856641276e-05,
+      "loss": 1.682,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2987431183130778,
+      "grad_norm": 0.4322430193424225,
+      "learning_rate": 7.9548757728974e-05,
+      "loss": 1.9822,
+      "step": 2876
+    },
+    {
+      "epoch": 0.29884699283265814,
+      "grad_norm": 0.39275041222572327,
+      "learning_rate": 7.953559374482305e-05,
+      "loss": 1.8279,
+      "step": 2877
+    },
+    {
+      "epoch": 0.29895086735223847,
+      "grad_norm": 0.3673188388347626,
+      "learning_rate": 7.952242661536175e-05,
+      "loss": 1.6634,
+      "step": 2878
+    },
+    {
+      "epoch": 0.29905474187181885,
+      "grad_norm": 0.39935851097106934,
+      "learning_rate": 7.950925634199228e-05,
+      "loss": 1.7705,
+      "step": 2879
+    },
+    {
+      "epoch": 0.2991586163913992,
+      "grad_norm": 0.4122634530067444,
+      "learning_rate": 7.949608292611721e-05,
+      "loss": 1.7795,
+      "step": 2880
+    },
+    {
+      "epoch": 0.29926249091097956,
+      "grad_norm": 0.40487560629844666,
+      "learning_rate": 7.948290636913939e-05,
+      "loss": 1.9119,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2993663654305599,
+      "grad_norm": 0.3445805013179779,
+      "learning_rate": 7.946972667246201e-05,
+      "loss": 1.5662,
+      "step": 2882
+    },
+    {
+      "epoch": 0.2994702399501402,
+      "grad_norm": 0.3658605217933655,
+      "learning_rate": 7.945654383748861e-05,
+      "loss": 1.5546,
+      "step": 2883
+    },
+    {
+      "epoch": 0.2995741144697206,
+      "grad_norm": 0.3820507824420929,
+      "learning_rate": 7.944335786562307e-05,
+      "loss": 1.5994,
+      "step": 2884
+    },
+    {
+      "epoch": 0.2996779889893009,
+      "grad_norm": 0.4138948321342468,
+      "learning_rate": 7.943016875826957e-05,
+      "loss": 1.829,
+      "step": 2885
+    },
+    {
+      "epoch": 0.29978186350888125,
+      "grad_norm": 0.35458049178123474,
+      "learning_rate": 7.941697651683267e-05,
+      "loss": 1.6141,
+      "step": 2886
+    },
+    {
+      "epoch": 0.29988573802846163,
+      "grad_norm": 0.4207065999507904,
+      "learning_rate": 7.940378114271723e-05,
+      "loss": 1.837,
+      "step": 2887
+    },
+    {
+      "epoch": 0.29998961254804196,
+      "grad_norm": 0.35922253131866455,
+      "learning_rate": 7.939058263732846e-05,
+      "loss": 1.5883,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3000934870676223,
+      "grad_norm": 0.38362789154052734,
+      "learning_rate": 7.93773810020719e-05,
+      "loss": 1.7448,
+      "step": 2889
+    },
+    {
+      "epoch": 0.30019736158720267,
+      "grad_norm": 0.39529237151145935,
+      "learning_rate": 7.93641762383534e-05,
+      "loss": 1.7747,
+      "step": 2890
+    },
+    {
+      "epoch": 0.300301236106783,
+      "grad_norm": 0.37079519033432007,
+      "learning_rate": 7.93509683475792e-05,
+      "loss": 1.6808,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3004051106263634,
+      "grad_norm": 0.37962087988853455,
+      "learning_rate": 7.93377573311558e-05,
+      "loss": 1.6966,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3005089851459437,
+      "grad_norm": 0.36086609959602356,
+      "learning_rate": 7.932454319049008e-05,
+      "loss": 1.6517,
+      "step": 2893
+    },
+    {
+      "epoch": 0.30061285966552403,
+      "grad_norm": 0.3676837980747223,
+      "learning_rate": 7.931132592698927e-05,
+      "loss": 1.5688,
+      "step": 2894
+    },
+    {
+      "epoch": 0.3007167341851044,
+      "grad_norm": 0.38151493668556213,
+      "learning_rate": 7.929810554206088e-05,
+      "loss": 1.7309,
+      "step": 2895
+    },
+    {
+      "epoch": 0.30082060870468474,
+      "grad_norm": 0.38594383001327515,
+      "learning_rate": 7.928488203711279e-05,
+      "loss": 1.7323,
+      "step": 2896
+    },
+    {
+      "epoch": 0.30092448322426507,
+      "grad_norm": 0.3639463186264038,
+      "learning_rate": 7.927165541355319e-05,
+      "loss": 1.6411,
+      "step": 2897
+    },
+    {
+      "epoch": 0.30102835774384545,
+      "grad_norm": 0.36965441703796387,
+      "learning_rate": 7.92584256727906e-05,
+      "loss": 1.6233,
+      "step": 2898
+    },
+    {
+      "epoch": 0.3011322322634258,
+      "grad_norm": 0.40190815925598145,
+      "learning_rate": 7.924519281623393e-05,
+      "loss": 1.8847,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3012361067830061,
+      "grad_norm": 0.3761938214302063,
+      "learning_rate": 7.923195684529232e-05,
+      "loss": 1.6344,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3013399813025865,
+      "grad_norm": 0.39748868346214294,
+      "learning_rate": 7.921871776137533e-05,
+      "loss": 1.9062,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3014438558221668,
+      "grad_norm": 0.3873717188835144,
+      "learning_rate": 7.920547556589282e-05,
+      "loss": 1.8971,
+      "step": 2902
+    },
+    {
+      "epoch": 0.30154773034174714,
+      "grad_norm": 0.39114367961883545,
+      "learning_rate": 7.919223026025498e-05,
+      "loss": 1.7381,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3016516048613275,
+      "grad_norm": 0.38005557656288147,
+      "learning_rate": 7.917898184587231e-05,
+      "loss": 1.8914,
+      "step": 2904
+    },
+    {
+      "epoch": 0.30175547938090785,
+      "grad_norm": 0.3953699469566345,
+      "learning_rate": 7.916573032415569e-05,
+      "loss": 1.8567,
+      "step": 2905
+    },
+    {
+      "epoch": 0.30185935390048824,
+      "grad_norm": 0.3641657829284668,
+      "learning_rate": 7.915247569651627e-05,
+      "loss": 1.5998,
+      "step": 2906
+    },
+    {
+      "epoch": 0.30196322842006856,
+      "grad_norm": 0.35653895139694214,
+      "learning_rate": 7.913921796436561e-05,
+      "loss": 1.5639,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3020671029396489,
+      "grad_norm": 0.3951781690120697,
+      "learning_rate": 7.912595712911552e-05,
+      "loss": 1.7428,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3021709774592293,
+      "grad_norm": 0.47132858633995056,
+      "learning_rate": 7.91126931921782e-05,
+      "loss": 1.7877,
+      "step": 2909
+    },
+    {
+      "epoch": 0.3022748519788096,
+      "grad_norm": 0.4093819260597229,
+      "learning_rate": 7.909942615496613e-05,
+      "loss": 1.8667,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3023787264983899,
+      "grad_norm": 0.43610236048698425,
+      "learning_rate": 7.908615601889218e-05,
+      "loss": 1.6627,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3024826010179703,
+      "grad_norm": 0.42912378907203674,
+      "learning_rate": 7.907288278536947e-05,
+      "loss": 1.6026,
+      "step": 2912
+    },
+    {
+      "epoch": 0.30258647553755064,
+      "grad_norm": 0.4064681828022003,
+      "learning_rate": 7.905960645581152e-05,
+      "loss": 1.7998,
+      "step": 2913
+    },
+    {
+      "epoch": 0.30269035005713096,
+      "grad_norm": 0.373544842004776,
+      "learning_rate": 7.904632703163218e-05,
+      "loss": 1.6253,
+      "step": 2914
+    },
+    {
+      "epoch": 0.30279422457671135,
+      "grad_norm": 0.376240074634552,
+      "learning_rate": 7.903304451424556e-05,
+      "loss": 1.7304,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3028980990962917,
+      "grad_norm": 0.3998112678527832,
+      "learning_rate": 7.901975890506616e-05,
+      "loss": 1.7403,
+      "step": 2916
+    },
+    {
+      "epoch": 0.303001973615872,
+      "grad_norm": 0.38144102692604065,
+      "learning_rate": 7.900647020550882e-05,
+      "loss": 1.7706,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3031058481354524,
+      "grad_norm": 0.370951771736145,
+      "learning_rate": 7.899317841698864e-05,
+      "loss": 1.7351,
+      "step": 2918
+    },
+    {
+      "epoch": 0.3032097226550327,
+      "grad_norm": 0.41188183426856995,
+      "learning_rate": 7.897988354092113e-05,
+      "loss": 1.8509,
+      "step": 2919
+    },
+    {
+      "epoch": 0.3033135971746131,
+      "grad_norm": 0.4149079918861389,
+      "learning_rate": 7.896658557872207e-05,
+      "loss": 1.6496,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3034174716941934,
+      "grad_norm": 0.38840773701667786,
+      "learning_rate": 7.89532845318076e-05,
+      "loss": 1.4942,
+      "step": 2921
+    },
+    {
+      "epoch": 0.30352134621377375,
+      "grad_norm": 0.4095969498157501,
+      "learning_rate": 7.893998040159418e-05,
+      "loss": 1.5979,
+      "step": 2922
+    },
+    {
+      "epoch": 0.30362522073335413,
+      "grad_norm": 0.3632447123527527,
+      "learning_rate": 7.892667318949855e-05,
+      "loss": 1.5617,
+      "step": 2923
+    },
+    {
+      "epoch": 0.30372909525293446,
+      "grad_norm": 0.34822526574134827,
+      "learning_rate": 7.891336289693789e-05,
+      "loss": 1.6198,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3038329697725148,
+      "grad_norm": 0.3828188180923462,
+      "learning_rate": 7.890004952532962e-05,
+      "loss": 1.7494,
+      "step": 2925
+    },
+    {
+      "epoch": 0.30393684429209517,
+      "grad_norm": 0.37072688341140747,
+      "learning_rate": 7.888673307609149e-05,
+      "loss": 1.5492,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3040407188116755,
+      "grad_norm": 0.3795013427734375,
+      "learning_rate": 7.887341355064162e-05,
+      "loss": 1.6222,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3041445933312558,
+      "grad_norm": 0.40049639344215393,
+      "learning_rate": 7.886009095039843e-05,
+      "loss": 1.8504,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3042484678508362,
+      "grad_norm": 0.4613747000694275,
+      "learning_rate": 7.884676527678068e-05,
+      "loss": 2.1015,
+      "step": 2929
+    },
+    {
+      "epoch": 0.30435234237041653,
+      "grad_norm": 0.3635263442993164,
+      "learning_rate": 7.883343653120743e-05,
+      "loss": 1.6364,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3044562168899969,
+      "grad_norm": 0.4116704761981964,
+      "learning_rate": 7.882010471509809e-05,
+      "loss": 1.5827,
+      "step": 2931
+    },
+    {
+      "epoch": 0.30456009140957724,
+      "grad_norm": 0.41490453481674194,
+      "learning_rate": 7.88067698298724e-05,
+      "loss": 1.9797,
+      "step": 2932
+    },
+    {
+      "epoch": 0.30466396592915757,
+      "grad_norm": 0.3712293207645416,
+      "learning_rate": 7.879343187695045e-05,
+      "loss": 1.7037,
+      "step": 2933
+    },
+    {
+      "epoch": 0.30476784044873795,
+      "grad_norm": 0.37144172191619873,
+      "learning_rate": 7.878009085775257e-05,
+      "loss": 1.7411,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3048717149683183,
+      "grad_norm": 0.38958555459976196,
+      "learning_rate": 7.876674677369954e-05,
+      "loss": 1.7444,
+      "step": 2935
+    },
+    {
+      "epoch": 0.3049755894878986,
+      "grad_norm": 0.3905205726623535,
+      "learning_rate": 7.875339962621235e-05,
+      "loss": 1.8388,
+      "step": 2936
+    },
+    {
+      "epoch": 0.305079464007479,
+      "grad_norm": 0.4097367525100708,
+      "learning_rate": 7.874004941671239e-05,
+      "loss": 1.7877,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3051833385270593,
+      "grad_norm": 0.39029520750045776,
+      "learning_rate": 7.872669614662135e-05,
+      "loss": 1.7491,
+      "step": 2938
+    },
+    {
+      "epoch": 0.30528721304663964,
+      "grad_norm": 0.3893304467201233,
+      "learning_rate": 7.871333981736124e-05,
+      "loss": 1.7892,
+      "step": 2939
+    },
+    {
+      "epoch": 0.30539108756622,
+      "grad_norm": 0.39329490065574646,
+      "learning_rate": 7.869998043035442e-05,
+      "loss": 1.7247,
+      "step": 2940
+    },
+    {
+      "epoch": 0.30549496208580035,
+      "grad_norm": 0.4152490496635437,
+      "learning_rate": 7.868661798702355e-05,
+      "loss": 1.767,
+      "step": 2941
+    },
+    {
+      "epoch": 0.3055988366053807,
+      "grad_norm": 0.40060338377952576,
+      "learning_rate": 7.867325248879163e-05,
+      "loss": 1.6133,
+      "step": 2942
+    },
+    {
+      "epoch": 0.30570271112496106,
+      "grad_norm": 0.422410786151886,
+      "learning_rate": 7.865988393708197e-05,
+      "loss": 1.6876,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3058065856445414,
+      "grad_norm": 0.4146443009376526,
+      "learning_rate": 7.864651233331823e-05,
+      "loss": 1.7866,
+      "step": 2944
+    },
+    {
+      "epoch": 0.30591046016412177,
+      "grad_norm": 0.363182008266449,
+      "learning_rate": 7.863313767892438e-05,
+      "loss": 1.5605,
+      "step": 2945
+    },
+    {
+      "epoch": 0.3060143346837021,
+      "grad_norm": 0.4126920998096466,
+      "learning_rate": 7.861975997532471e-05,
+      "loss": 1.8965,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3061182092032824,
+      "grad_norm": 0.4038070738315582,
+      "learning_rate": 7.860637922394387e-05,
+      "loss": 1.8668,
+      "step": 2947
+    },
+    {
+      "epoch": 0.3062220837228628,
+      "grad_norm": 0.3751528859138489,
+      "learning_rate": 7.859299542620675e-05,
+      "loss": 1.8258,
+      "step": 2948
+    },
+    {
+      "epoch": 0.30632595824244313,
+      "grad_norm": 0.422244668006897,
+      "learning_rate": 7.857960858353866e-05,
+      "loss": 1.8309,
+      "step": 2949
+    },
+    {
+      "epoch": 0.30642983276202346,
+      "grad_norm": 0.37908682227134705,
+      "learning_rate": 7.85662186973652e-05,
+      "loss": 1.7936,
+      "step": 2950
+    },
+    {
+      "epoch": 0.30653370728160384,
+      "grad_norm": 0.36694470047950745,
+      "learning_rate": 7.855282576911224e-05,
+      "loss": 1.7259,
+      "step": 2951
+    },
+    {
+      "epoch": 0.30663758180118417,
+      "grad_norm": 0.3961583077907562,
+      "learning_rate": 7.85394298002061e-05,
+      "loss": 1.7949,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3067414563207645,
+      "grad_norm": 0.3951317071914673,
+      "learning_rate": 7.852603079207328e-05,
+      "loss": 1.6679,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3068453308403449,
+      "grad_norm": 0.3763517737388611,
+      "learning_rate": 7.851262874614069e-05,
+      "loss": 1.6965,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3069492053599252,
+      "grad_norm": 0.3869110941886902,
+      "learning_rate": 7.849922366383555e-05,
+      "loss": 1.6044,
+      "step": 2955
+    },
+    {
+      "epoch": 0.30705307987950553,
+      "grad_norm": 0.4839021563529968,
+      "learning_rate": 7.848581554658539e-05,
+      "loss": 2.1448,
+      "step": 2956
+    },
+    {
+      "epoch": 0.3071569543990859,
+      "grad_norm": 0.4335784912109375,
+      "learning_rate": 7.847240439581806e-05,
+      "loss": 1.7972,
+      "step": 2957
+    },
+    {
+      "epoch": 0.30726082891866624,
+      "grad_norm": 0.5460455417633057,
+      "learning_rate": 7.845899021296178e-05,
+      "loss": 1.732,
+      "step": 2958
+    },
+    {
+      "epoch": 0.3073647034382466,
+      "grad_norm": 0.39001351594924927,
+      "learning_rate": 7.844557299944501e-05,
+      "loss": 1.7266,
+      "step": 2959
+    },
+    {
+      "epoch": 0.30746857795782695,
+      "grad_norm": 0.39050352573394775,
+      "learning_rate": 7.843215275669663e-05,
+      "loss": 1.7042,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3075724524774073,
+      "grad_norm": 0.3831944465637207,
+      "learning_rate": 7.841872948614573e-05,
+      "loss": 1.7086,
+      "step": 2961
+    },
+    {
+      "epoch": 0.30767632699698766,
+      "grad_norm": 0.4240471422672272,
+      "learning_rate": 7.840530318922181e-05,
+      "loss": 1.7215,
+      "step": 2962
+    },
+    {
+      "epoch": 0.307780201516568,
+      "grad_norm": 0.3831408619880676,
+      "learning_rate": 7.839187386735469e-05,
+      "loss": 1.7995,
+      "step": 2963
+    },
+    {
+      "epoch": 0.3078840760361483,
+      "grad_norm": 0.43340522050857544,
+      "learning_rate": 7.837844152197447e-05,
+      "loss": 1.7028,
+      "step": 2964
+    },
+    {
+      "epoch": 0.3079879505557287,
+      "grad_norm": 0.3613108992576599,
+      "learning_rate": 7.836500615451156e-05,
+      "loss": 1.4274,
+      "step": 2965
+    },
+    {
+      "epoch": 0.308091825075309,
+      "grad_norm": 0.4255940914154053,
+      "learning_rate": 7.835156776639679e-05,
+      "loss": 1.5747,
+      "step": 2966
+    },
+    {
+      "epoch": 0.30819569959488935,
+      "grad_norm": 0.3656274974346161,
+      "learning_rate": 7.833812635906118e-05,
+      "loss": 1.6061,
+      "step": 2967
+    },
+    {
+      "epoch": 0.30829957411446973,
+      "grad_norm": 0.4550713002681732,
+      "learning_rate": 7.832468193393616e-05,
+      "loss": 1.8704,
+      "step": 2968
+    },
+    {
+      "epoch": 0.30840344863405006,
+      "grad_norm": 0.40488240122795105,
+      "learning_rate": 7.831123449245345e-05,
+      "loss": 1.7132,
+      "step": 2969
+    },
+    {
+      "epoch": 0.30850732315363044,
+      "grad_norm": 0.3986703157424927,
+      "learning_rate": 7.829778403604513e-05,
+      "loss": 1.8511,
+      "step": 2970
+    },
+    {
+      "epoch": 0.30861119767321077,
+      "grad_norm": 0.41377493739128113,
+      "learning_rate": 7.828433056614351e-05,
+      "loss": 1.8369,
+      "step": 2971
+    },
+    {
+      "epoch": 0.3087150721927911,
+      "grad_norm": 0.3639225959777832,
+      "learning_rate": 7.827087408418132e-05,
+      "loss": 1.7907,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3088189467123715,
+      "grad_norm": 0.3887461721897125,
+      "learning_rate": 7.825741459159157e-05,
+      "loss": 1.7639,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3089228212319518,
+      "grad_norm": 0.38962462544441223,
+      "learning_rate": 7.824395208980758e-05,
+      "loss": 1.7871,
+      "step": 2974
+    },
+    {
+      "epoch": 0.30902669575153213,
+      "grad_norm": 0.43991488218307495,
+      "learning_rate": 7.8230486580263e-05,
+      "loss": 1.988,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3091305702711125,
+      "grad_norm": 0.4023156464099884,
+      "learning_rate": 7.821701806439179e-05,
+      "loss": 1.812,
+      "step": 2976
+    },
+    {
+      "epoch": 0.30923444479069284,
+      "grad_norm": 0.37170884013175964,
+      "learning_rate": 7.820354654362828e-05,
+      "loss": 1.7231,
+      "step": 2977
+    },
+    {
+      "epoch": 0.30933831931027317,
+      "grad_norm": 0.38935086131095886,
+      "learning_rate": 7.819007201940706e-05,
+      "loss": 1.7493,
+      "step": 2978
+    },
+    {
+      "epoch": 0.30944219382985355,
+      "grad_norm": 0.3613695204257965,
+      "learning_rate": 7.817659449316305e-05,
+      "loss": 1.621,
+      "step": 2979
+    },
+    {
+      "epoch": 0.3095460683494339,
+      "grad_norm": 0.3804933726787567,
+      "learning_rate": 7.816311396633152e-05,
+      "loss": 1.7622,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3096499428690142,
+      "grad_norm": 0.3554391860961914,
+      "learning_rate": 7.814963044034802e-05,
+      "loss": 1.615,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3097538173885946,
+      "grad_norm": 0.42907533049583435,
+      "learning_rate": 7.813614391664847e-05,
+      "loss": 1.9264,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3098576919081749,
+      "grad_norm": 0.46726834774017334,
+      "learning_rate": 7.812265439666906e-05,
+      "loss": 1.6321,
+      "step": 2983
+    },
+    {
+      "epoch": 0.3099615664277553,
+      "grad_norm": 0.43149927258491516,
+      "learning_rate": 7.81091618818463e-05,
+      "loss": 1.902,
+      "step": 2984
+    },
+    {
+      "epoch": 0.3100654409473356,
+      "grad_norm": 0.42148879170417786,
+      "learning_rate": 7.80956663736171e-05,
+      "loss": 1.8288,
+      "step": 2985
+    },
+    {
+      "epoch": 0.31016931546691595,
+      "grad_norm": 0.39359238743782043,
+      "learning_rate": 7.808216787341855e-05,
+      "loss": 1.7853,
+      "step": 2986
+    },
+    {
+      "epoch": 0.31027318998649633,
+      "grad_norm": 0.37277668714523315,
+      "learning_rate": 7.806866638268818e-05,
+      "loss": 1.702,
+      "step": 2987
+    },
+    {
+      "epoch": 0.31037706450607666,
+      "grad_norm": 0.3854857087135315,
+      "learning_rate": 7.80551619028638e-05,
+      "loss": 1.7776,
+      "step": 2988
+    },
+    {
+      "epoch": 0.310480939025657,
+      "grad_norm": 0.3613017201423645,
+      "learning_rate": 7.80416544353835e-05,
+      "loss": 1.6451,
+      "step": 2989
+    },
+    {
+      "epoch": 0.31058481354523737,
+      "grad_norm": 0.3885599970817566,
+      "learning_rate": 7.802814398168572e-05,
+      "loss": 1.8862,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3106886880648177,
+      "grad_norm": 0.3823319375514984,
+      "learning_rate": 7.801463054320927e-05,
+      "loss": 1.7783,
+      "step": 2991
+    },
+    {
+      "epoch": 0.310792562584398,
+      "grad_norm": 0.43862950801849365,
+      "learning_rate": 7.800111412139318e-05,
+      "loss": 1.654,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3108964371039784,
+      "grad_norm": 0.4039106070995331,
+      "learning_rate": 7.798759471767684e-05,
+      "loss": 1.7425,
+      "step": 2993
+    },
+    {
+      "epoch": 0.31100031162355873,
+      "grad_norm": 0.3942197263240814,
+      "learning_rate": 7.797407233349998e-05,
+      "loss": 1.7184,
+      "step": 2994
+    },
+    {
+      "epoch": 0.31110418614313906,
+      "grad_norm": 0.35280346870422363,
+      "learning_rate": 7.796054697030262e-05,
+      "loss": 1.5123,
+      "step": 2995
+    },
+    {
+      "epoch": 0.31120806066271944,
+      "grad_norm": 0.38291507959365845,
+      "learning_rate": 7.794701862952512e-05,
+      "loss": 1.7482,
+      "step": 2996
+    },
+    {
+      "epoch": 0.31131193518229977,
+      "grad_norm": 0.4096364378929138,
+      "learning_rate": 7.793348731260813e-05,
+      "loss": 1.8343,
+      "step": 2997
+    },
+    {
+      "epoch": 0.31141580970188015,
+      "grad_norm": 0.40099769830703735,
+      "learning_rate": 7.79199530209926e-05,
+      "loss": 1.655,
+      "step": 2998
+    },
+    {
+      "epoch": 0.3115196842214605,
+      "grad_norm": 0.3833656907081604,
+      "learning_rate": 7.79064157561199e-05,
+      "loss": 1.7954,
+      "step": 2999
+    },
+    {
+      "epoch": 0.3116235587410408,
+      "grad_norm": 0.3824611008167267,
+      "learning_rate": 7.789287551943158e-05,
+      "loss": 1.5854,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3117274332606212,
+      "grad_norm": 0.3726818561553955,
+      "learning_rate": 7.787933231236958e-05,
+      "loss": 1.6898,
+      "step": 3001
+    },
+    {
+      "epoch": 0.3118313077802015,
+      "grad_norm": 0.37149757146835327,
+      "learning_rate": 7.786578613637617e-05,
+      "loss": 1.6463,
+      "step": 3002
+    },
+    {
+      "epoch": 0.31193518229978184,
+      "grad_norm": 0.42311716079711914,
+      "learning_rate": 7.78522369928939e-05,
+      "loss": 1.951,
+      "step": 3003
+    },
+    {
+      "epoch": 0.3120390568193622,
+      "grad_norm": 0.47437840700149536,
+      "learning_rate": 7.783868488336563e-05,
+      "loss": 1.6597,
+      "step": 3004
+    },
+    {
+      "epoch": 0.31214293133894255,
+      "grad_norm": 0.3761157989501953,
+      "learning_rate": 7.782512980923459e-05,
+      "loss": 1.8036,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3122468058585229,
+      "grad_norm": 0.3912384510040283,
+      "learning_rate": 7.781157177194426e-05,
+      "loss": 1.6382,
+      "step": 3006
+    },
+    {
+      "epoch": 0.31235068037810326,
+      "grad_norm": 0.3848210871219635,
+      "learning_rate": 7.779801077293847e-05,
+      "loss": 1.7521,
+      "step": 3007
+    },
+    {
+      "epoch": 0.3124545548976836,
+      "grad_norm": 0.3805733323097229,
+      "learning_rate": 7.778444681366137e-05,
+      "loss": 1.7019,
+      "step": 3008
+    },
+    {
+      "epoch": 0.31255842941726397,
+      "grad_norm": 0.38343513011932373,
+      "learning_rate": 7.777087989555741e-05,
+      "loss": 1.7686,
+      "step": 3009
+    },
+    {
+      "epoch": 0.3126623039368443,
+      "grad_norm": 0.37739554047584534,
+      "learning_rate": 7.775731002007138e-05,
+      "loss": 1.5242,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3127661784564246,
+      "grad_norm": 0.37050339579582214,
+      "learning_rate": 7.774373718864833e-05,
+      "loss": 1.771,
+      "step": 3011
+    },
+    {
+      "epoch": 0.312870052976005,
+      "grad_norm": 0.39899301528930664,
+      "learning_rate": 7.773016140273368e-05,
+      "loss": 1.8089,
+      "step": 3012
+    },
+    {
+      "epoch": 0.31297392749558534,
+      "grad_norm": 0.3865603506565094,
+      "learning_rate": 7.771658266377315e-05,
+      "loss": 1.6336,
+      "step": 3013
+    },
+    {
+      "epoch": 0.31307780201516566,
+      "grad_norm": 0.3612881302833557,
+      "learning_rate": 7.770300097321276e-05,
+      "loss": 1.4883,
+      "step": 3014
+    },
+    {
+      "epoch": 0.31318167653474605,
+      "grad_norm": 0.422632098197937,
+      "learning_rate": 7.768941633249884e-05,
+      "loss": 1.8948,
+      "step": 3015
+    },
+    {
+      "epoch": 0.31328555105432637,
+      "grad_norm": 0.3684242069721222,
+      "learning_rate": 7.76758287430781e-05,
+      "loss": 1.702,
+      "step": 3016
+    },
+    {
+      "epoch": 0.3133894255739067,
+      "grad_norm": 0.36709776520729065,
+      "learning_rate": 7.766223820639747e-05,
+      "loss": 1.687,
+      "step": 3017
+    },
+    {
+      "epoch": 0.3134933000934871,
+      "grad_norm": 0.39332762360572815,
+      "learning_rate": 7.764864472390423e-05,
+      "loss": 1.7333,
+      "step": 3018
+    },
+    {
+      "epoch": 0.3135971746130674,
+      "grad_norm": 0.3547021448612213,
+      "learning_rate": 7.7635048297046e-05,
+      "loss": 1.6883,
+      "step": 3019
+    },
+    {
+      "epoch": 0.31370104913264774,
+      "grad_norm": 0.393284410238266,
+      "learning_rate": 7.762144892727069e-05,
+      "loss": 1.9121,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3138049236522281,
+      "grad_norm": 0.4200425148010254,
+      "learning_rate": 7.760784661602653e-05,
+      "loss": 1.8525,
+      "step": 3021
+    },
+    {
+      "epoch": 0.31390879817180845,
+      "grad_norm": 0.3398437798023224,
+      "learning_rate": 7.759424136476206e-05,
+      "loss": 1.529,
+      "step": 3022
+    },
+    {
+      "epoch": 0.31401267269138883,
+      "grad_norm": 0.3940331041812897,
+      "learning_rate": 7.75806331749261e-05,
+      "loss": 1.8153,
+      "step": 3023
+    },
+    {
+      "epoch": 0.31411654721096915,
+      "grad_norm": 0.37523049116134644,
+      "learning_rate": 7.756702204796786e-05,
+      "loss": 1.6807,
+      "step": 3024
+    },
+    {
+      "epoch": 0.3142204217305495,
+      "grad_norm": 0.3940492868423462,
+      "learning_rate": 7.75534079853368e-05,
+      "loss": 1.8526,
+      "step": 3025
+    },
+    {
+      "epoch": 0.31432429625012986,
+      "grad_norm": 0.40810003876686096,
+      "learning_rate": 7.753979098848272e-05,
+      "loss": 1.6186,
+      "step": 3026
+    },
+    {
+      "epoch": 0.3144281707697102,
+      "grad_norm": 0.3908647894859314,
+      "learning_rate": 7.752617105885574e-05,
+      "loss": 1.6896,
+      "step": 3027
+    },
+    {
+      "epoch": 0.3145320452892905,
+      "grad_norm": 0.35582828521728516,
+      "learning_rate": 7.751254819790623e-05,
+      "loss": 1.719,
+      "step": 3028
+    },
+    {
+      "epoch": 0.3146359198088709,
+      "grad_norm": 0.3710813820362091,
+      "learning_rate": 7.749892240708494e-05,
+      "loss": 1.7155,
+      "step": 3029
+    },
+    {
+      "epoch": 0.31473979432845123,
+      "grad_norm": 0.4214909076690674,
+      "learning_rate": 7.748529368784292e-05,
+      "loss": 1.8792,
+      "step": 3030
+    },
+    {
+      "epoch": 0.31484366884803155,
+      "grad_norm": 0.4322544038295746,
+      "learning_rate": 7.747166204163152e-05,
+      "loss": 1.8348,
+      "step": 3031
+    },
+    {
+      "epoch": 0.31494754336761194,
+      "grad_norm": 0.38530245423316956,
+      "learning_rate": 7.745802746990239e-05,
+      "loss": 1.7037,
+      "step": 3032
+    },
+    {
+      "epoch": 0.31505141788719226,
+      "grad_norm": 0.3923969268798828,
+      "learning_rate": 7.744438997410752e-05,
+      "loss": 1.6523,
+      "step": 3033
+    },
+    {
+      "epoch": 0.3151552924067726,
+      "grad_norm": 0.3855164349079132,
+      "learning_rate": 7.74307495556992e-05,
+      "loss": 1.705,
+      "step": 3034
+    },
+    {
+      "epoch": 0.315259166926353,
+      "grad_norm": 0.3986038267612457,
+      "learning_rate": 7.741710621613002e-05,
+      "loss": 1.9339,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3153630414459333,
+      "grad_norm": 0.42609813809394836,
+      "learning_rate": 7.740345995685287e-05,
+      "loss": 1.8712,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3154669159655137,
+      "grad_norm": 0.4008728563785553,
+      "learning_rate": 7.738981077932099e-05,
+      "loss": 1.8456,
+      "step": 3037
+    },
+    {
+      "epoch": 0.315570790485094,
+      "grad_norm": 0.4060449004173279,
+      "learning_rate": 7.737615868498793e-05,
+      "loss": 1.7126,
+      "step": 3038
+    },
+    {
+      "epoch": 0.31567466500467434,
+      "grad_norm": 0.36810052394866943,
+      "learning_rate": 7.736250367530751e-05,
+      "loss": 1.6292,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3157785395242547,
+      "grad_norm": 0.39102548360824585,
+      "learning_rate": 7.734884575173387e-05,
+      "loss": 1.7703,
+      "step": 3040
+    },
+    {
+      "epoch": 0.31588241404383505,
+      "grad_norm": 0.4772125780582428,
+      "learning_rate": 7.73351849157215e-05,
+      "loss": 1.7719,
+      "step": 3041
+    },
+    {
+      "epoch": 0.3159862885634154,
+      "grad_norm": 0.40599432587623596,
+      "learning_rate": 7.732152116872515e-05,
+      "loss": 1.6538,
+      "step": 3042
+    },
+    {
+      "epoch": 0.31609016308299576,
+      "grad_norm": 0.4301255941390991,
+      "learning_rate": 7.730785451219991e-05,
+      "loss": 1.755,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3161940376025761,
+      "grad_norm": 0.386177122592926,
+      "learning_rate": 7.729418494760119e-05,
+      "loss": 1.5637,
+      "step": 3044
+    },
+    {
+      "epoch": 0.3162979121221564,
+      "grad_norm": 0.3636477589607239,
+      "learning_rate": 7.728051247638468e-05,
+      "loss": 1.5991,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3164017866417368,
+      "grad_norm": 0.38973626494407654,
+      "learning_rate": 7.726683710000637e-05,
+      "loss": 1.7331,
+      "step": 3046
+    },
+    {
+      "epoch": 0.3165056611613171,
+      "grad_norm": 0.42155221104621887,
+      "learning_rate": 7.72531588199226e-05,
+      "loss": 1.8179,
+      "step": 3047
+    },
+    {
+      "epoch": 0.3166095356808975,
+      "grad_norm": 0.4188525080680847,
+      "learning_rate": 7.723947763759e-05,
+      "loss": 1.7464,
+      "step": 3048
+    },
+    {
+      "epoch": 0.31671341020047783,
+      "grad_norm": 0.38648781180381775,
+      "learning_rate": 7.722579355446551e-05,
+      "loss": 1.5823,
+      "step": 3049
+    },
+    {
+      "epoch": 0.31681728472005816,
+      "grad_norm": 0.4101918339729309,
+      "learning_rate": 7.721210657200637e-05,
+      "loss": 1.7697,
+      "step": 3050
+    },
+    {
+      "epoch": 0.31692115923963854,
+      "grad_norm": 0.366359680891037,
+      "learning_rate": 7.719841669167014e-05,
+      "loss": 1.6861,
+      "step": 3051
+    },
+    {
+      "epoch": 0.31702503375921887,
+      "grad_norm": 0.4024447202682495,
+      "learning_rate": 7.71847239149147e-05,
+      "loss": 1.9079,
+      "step": 3052
+    },
+    {
+      "epoch": 0.3171289082787992,
+      "grad_norm": 0.3973731994628906,
+      "learning_rate": 7.717102824319822e-05,
+      "loss": 1.8736,
+      "step": 3053
+    },
+    {
+      "epoch": 0.3172327827983796,
+      "grad_norm": 0.42864885926246643,
+      "learning_rate": 7.715732967797915e-05,
+      "loss": 1.7298,
+      "step": 3054
+    },
+    {
+      "epoch": 0.3173366573179599,
+      "grad_norm": 0.37840789556503296,
+      "learning_rate": 7.714362822071632e-05,
+      "loss": 1.8096,
+      "step": 3055
+    },
+    {
+      "epoch": 0.31744053183754023,
+      "grad_norm": 0.3534778654575348,
+      "learning_rate": 7.712992387286882e-05,
+      "loss": 1.5215,
+      "step": 3056
+    },
+    {
+      "epoch": 0.3175444063571206,
+      "grad_norm": 0.37285417318344116,
+      "learning_rate": 7.711621663589602e-05,
+      "loss": 1.6497,
+      "step": 3057
+    },
+    {
+      "epoch": 0.31764828087670094,
+      "grad_norm": 0.35734882950782776,
+      "learning_rate": 7.71025065112577e-05,
+      "loss": 1.6053,
+      "step": 3058
+    },
+    {
+      "epoch": 0.31775215539628127,
+      "grad_norm": 0.38647857308387756,
+      "learning_rate": 7.708879350041383e-05,
+      "loss": 1.6968,
+      "step": 3059
+    },
+    {
+      "epoch": 0.31785602991586165,
+      "grad_norm": 0.4510471522808075,
+      "learning_rate": 7.707507760482474e-05,
+      "loss": 1.6815,
+      "step": 3060
+    },
+    {
+      "epoch": 0.317959904435442,
+      "grad_norm": 0.4056614935398102,
+      "learning_rate": 7.706135882595108e-05,
+      "loss": 1.6113,
+      "step": 3061
+    },
+    {
+      "epoch": 0.31806377895502236,
+      "grad_norm": 0.4500080645084381,
+      "learning_rate": 7.70476371652538e-05,
+      "loss": 1.9964,
+      "step": 3062
+    },
+    {
+      "epoch": 0.3181676534746027,
+      "grad_norm": 0.3800092935562134,
+      "learning_rate": 7.703391262419413e-05,
+      "loss": 1.67,
+      "step": 3063
+    },
+    {
+      "epoch": 0.318271527994183,
+      "grad_norm": 0.4021591544151306,
+      "learning_rate": 7.702018520423367e-05,
+      "loss": 1.7399,
+      "step": 3064
+    },
+    {
+      "epoch": 0.3183754025137634,
+      "grad_norm": 0.42205873131752014,
+      "learning_rate": 7.700645490683421e-05,
+      "loss": 1.8155,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3184792770333437,
+      "grad_norm": 0.3718779385089874,
+      "learning_rate": 7.699272173345799e-05,
+      "loss": 1.7583,
+      "step": 3066
+    },
+    {
+      "epoch": 0.31858315155292405,
+      "grad_norm": 0.3647688031196594,
+      "learning_rate": 7.697898568556745e-05,
+      "loss": 1.5891,
+      "step": 3067
+    },
+    {
+      "epoch": 0.31868702607250443,
+      "grad_norm": 0.3905600309371948,
+      "learning_rate": 7.696524676462538e-05,
+      "loss": 1.7358,
+      "step": 3068
+    },
+    {
+      "epoch": 0.31879090059208476,
+      "grad_norm": 0.36496713757514954,
+      "learning_rate": 7.695150497209485e-05,
+      "loss": 1.7218,
+      "step": 3069
+    },
+    {
+      "epoch": 0.3188947751116651,
+      "grad_norm": 0.4626975357532501,
+      "learning_rate": 7.69377603094393e-05,
+      "loss": 1.8209,
+      "step": 3070
+    },
+    {
+      "epoch": 0.31899864963124547,
+      "grad_norm": 0.4146973490715027,
+      "learning_rate": 7.692401277812236e-05,
+      "loss": 1.6095,
+      "step": 3071
+    },
+    {
+      "epoch": 0.3191025241508258,
+      "grad_norm": 0.3558516800403595,
+      "learning_rate": 7.69102623796081e-05,
+      "loss": 1.5576,
+      "step": 3072
+    },
+    {
+      "epoch": 0.3192063986704061,
+      "grad_norm": 0.4161911606788635,
+      "learning_rate": 7.689650911536081e-05,
+      "loss": 1.656,
+      "step": 3073
+    },
+    {
+      "epoch": 0.3193102731899865,
+      "grad_norm": 0.40754130482673645,
+      "learning_rate": 7.688275298684509e-05,
+      "loss": 1.7435,
+      "step": 3074
+    },
+    {
+      "epoch": 0.31941414770956683,
+      "grad_norm": 0.36731454730033875,
+      "learning_rate": 7.686899399552587e-05,
+      "loss": 1.6421,
+      "step": 3075
+    },
+    {
+      "epoch": 0.3195180222291472,
+      "grad_norm": 0.4020611047744751,
+      "learning_rate": 7.685523214286839e-05,
+      "loss": 1.7208,
+      "step": 3076
+    },
+    {
+      "epoch": 0.31962189674872754,
+      "grad_norm": 0.48006507754325867,
+      "learning_rate": 7.684146743033815e-05,
+      "loss": 1.7648,
+      "step": 3077
+    },
+    {
+      "epoch": 0.31972577126830787,
+      "grad_norm": 0.39849093556404114,
+      "learning_rate": 7.6827699859401e-05,
+      "loss": 1.751,
+      "step": 3078
+    },
+    {
+      "epoch": 0.31982964578788825,
+      "grad_norm": 0.3924444615840912,
+      "learning_rate": 7.681392943152308e-05,
+      "loss": 1.7878,
+      "step": 3079
+    },
+    {
+      "epoch": 0.3199335203074686,
+      "grad_norm": 0.4005417823791504,
+      "learning_rate": 7.680015614817083e-05,
+      "loss": 1.7752,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3200373948270489,
+      "grad_norm": 0.39548689126968384,
+      "learning_rate": 7.678638001081102e-05,
+      "loss": 1.7377,
+      "step": 3081
+    },
+    {
+      "epoch": 0.3201412693466293,
+      "grad_norm": 0.4022079408168793,
+      "learning_rate": 7.677260102091066e-05,
+      "loss": 1.7732,
+      "step": 3082
+    },
+    {
+      "epoch": 0.3202451438662096,
+      "grad_norm": 0.4020305573940277,
+      "learning_rate": 7.675881917993713e-05,
+      "loss": 1.6768,
+      "step": 3083
+    },
+    {
+      "epoch": 0.32034901838578994,
+      "grad_norm": 0.374347448348999,
+      "learning_rate": 7.674503448935808e-05,
+      "loss": 1.59,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3204528929053703,
+      "grad_norm": 0.3810427784919739,
+      "learning_rate": 7.673124695064148e-05,
+      "loss": 1.3873,
+      "step": 3085
+    },
+    {
+      "epoch": 0.32055676742495065,
+      "grad_norm": 0.43509384989738464,
+      "learning_rate": 7.671745656525559e-05,
+      "loss": 1.5037,
+      "step": 3086
+    },
+    {
+      "epoch": 0.32066064194453103,
+      "grad_norm": 0.38209038972854614,
+      "learning_rate": 7.670366333466899e-05,
+      "loss": 1.6842,
+      "step": 3087
+    },
+    {
+      "epoch": 0.32076451646411136,
+      "grad_norm": 0.4374238848686218,
+      "learning_rate": 7.668986726035054e-05,
+      "loss": 1.858,
+      "step": 3088
+    },
+    {
+      "epoch": 0.3208683909836917,
+      "grad_norm": 0.4977831244468689,
+      "learning_rate": 7.66760683437694e-05,
+      "loss": 1.9369,
+      "step": 3089
+    },
+    {
+      "epoch": 0.32097226550327207,
+      "grad_norm": 0.37174275517463684,
+      "learning_rate": 7.666226658639507e-05,
+      "loss": 1.6107,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3210761400228524,
+      "grad_norm": 0.3945876657962799,
+      "learning_rate": 7.664846198969733e-05,
+      "loss": 1.8063,
+      "step": 3091
+    },
+    {
+      "epoch": 0.3211800145424327,
+      "grad_norm": 0.4560282230377197,
+      "learning_rate": 7.663465455514625e-05,
+      "loss": 2.2025,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3212838890620131,
+      "grad_norm": 0.3639695346355438,
+      "learning_rate": 7.662084428421221e-05,
+      "loss": 1.6021,
+      "step": 3093
+    },
+    {
+      "epoch": 0.32138776358159343,
+      "grad_norm": 0.36369383335113525,
+      "learning_rate": 7.660703117836591e-05,
+      "loss": 1.6781,
+      "step": 3094
+    },
+    {
+      "epoch": 0.32149163810117376,
+      "grad_norm": 0.3726049065589905,
+      "learning_rate": 7.659321523907834e-05,
+      "loss": 1.8033,
+      "step": 3095
+    },
+    {
+      "epoch": 0.32159551262075414,
+      "grad_norm": 0.3972417116165161,
+      "learning_rate": 7.657939646782077e-05,
+      "loss": 1.6971,
+      "step": 3096
+    },
+    {
+      "epoch": 0.32169938714033447,
+      "grad_norm": 0.4059099555015564,
+      "learning_rate": 7.656557486606482e-05,
+      "loss": 1.7625,
+      "step": 3097
+    },
+    {
+      "epoch": 0.3218032616599148,
+      "grad_norm": 0.3809730112552643,
+      "learning_rate": 7.655175043528235e-05,
+      "loss": 1.79,
+      "step": 3098
+    },
+    {
+      "epoch": 0.3219071361794952,
+      "grad_norm": 0.37229931354522705,
+      "learning_rate": 7.653792317694556e-05,
+      "loss": 1.7281,
+      "step": 3099
+    },
+    {
+      "epoch": 0.3220110106990755,
+      "grad_norm": 0.40562358498573303,
+      "learning_rate": 7.652409309252699e-05,
+      "loss": 1.7644,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3221148852186559,
+      "grad_norm": 0.3879106044769287,
+      "learning_rate": 7.651026018349938e-05,
+      "loss": 1.815,
+      "step": 3101
+    },
+    {
+      "epoch": 0.3222187597382362,
+      "grad_norm": 0.39501863718032837,
+      "learning_rate": 7.649642445133585e-05,
+      "loss": 1.8004,
+      "step": 3102
+    },
+    {
+      "epoch": 0.32232263425781654,
+      "grad_norm": 0.42894792556762695,
+      "learning_rate": 7.64825858975098e-05,
+      "loss": 1.8058,
+      "step": 3103
+    },
+    {
+      "epoch": 0.3224265087773969,
+      "grad_norm": 0.4245396852493286,
+      "learning_rate": 7.646874452349491e-05,
+      "loss": 1.7229,
+      "step": 3104
+    },
+    {
+      "epoch": 0.32253038329697725,
+      "grad_norm": 0.3750874996185303,
+      "learning_rate": 7.64549003307652e-05,
+      "loss": 1.6437,
+      "step": 3105
+    },
+    {
+      "epoch": 0.3226342578165576,
+      "grad_norm": 0.44294023513793945,
+      "learning_rate": 7.644105332079497e-05,
+      "loss": 1.8526,
+      "step": 3106
+    },
+    {
+      "epoch": 0.32273813233613796,
+      "grad_norm": 0.4349030554294586,
+      "learning_rate": 7.642720349505881e-05,
+      "loss": 1.8628,
+      "step": 3107
+    },
+    {
+      "epoch": 0.3228420068557183,
+      "grad_norm": 0.4126608967781067,
+      "learning_rate": 7.641335085503161e-05,
+      "loss": 1.7873,
+      "step": 3108
+    },
+    {
+      "epoch": 0.3229458813752986,
+      "grad_norm": 0.3600351810455322,
+      "learning_rate": 7.639949540218858e-05,
+      "loss": 1.7672,
+      "step": 3109
+    },
+    {
+      "epoch": 0.323049755894879,
+      "grad_norm": 0.38123077154159546,
+      "learning_rate": 7.63856371380052e-05,
+      "loss": 1.7417,
+      "step": 3110
+    },
+    {
+      "epoch": 0.3231536304144593,
+      "grad_norm": 0.41736799478530884,
+      "learning_rate": 7.63717760639573e-05,
+      "loss": 1.855,
+      "step": 3111
+    },
+    {
+      "epoch": 0.3232575049340397,
+      "grad_norm": 0.42737138271331787,
+      "learning_rate": 7.635791218152094e-05,
+      "loss": 1.766,
+      "step": 3112
+    },
+    {
+      "epoch": 0.32336137945362003,
+      "grad_norm": 0.37469685077667236,
+      "learning_rate": 7.634404549217254e-05,
+      "loss": 1.6696,
+      "step": 3113
+    },
+    {
+      "epoch": 0.32346525397320036,
+      "grad_norm": 0.4223494231700897,
+      "learning_rate": 7.633017599738879e-05,
+      "loss": 1.7205,
+      "step": 3114
+    },
+    {
+      "epoch": 0.32356912849278074,
+      "grad_norm": 0.39290282130241394,
+      "learning_rate": 7.631630369864668e-05,
+      "loss": 1.6764,
+      "step": 3115
+    },
+    {
+      "epoch": 0.32367300301236107,
+      "grad_norm": 0.38165774941444397,
+      "learning_rate": 7.63024285974235e-05,
+      "loss": 1.633,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3237768775319414,
+      "grad_norm": 0.46332311630249023,
+      "learning_rate": 7.628855069519685e-05,
+      "loss": 1.9059,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3238807520515218,
+      "grad_norm": 0.38814160227775574,
+      "learning_rate": 7.627466999344459e-05,
+      "loss": 1.6475,
+      "step": 3118
+    },
+    {
+      "epoch": 0.3239846265711021,
+      "grad_norm": 0.44505298137664795,
+      "learning_rate": 7.626078649364494e-05,
+      "loss": 1.8901,
+      "step": 3119
+    },
+    {
+      "epoch": 0.32408850109068243,
+      "grad_norm": 0.4025139808654785,
+      "learning_rate": 7.624690019727636e-05,
+      "loss": 1.62,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3241923756102628,
+      "grad_norm": 0.37177082896232605,
+      "learning_rate": 7.623301110581764e-05,
+      "loss": 1.7613,
+      "step": 3121
+    },
+    {
+      "epoch": 0.32429625012984314,
+      "grad_norm": 0.38354530930519104,
+      "learning_rate": 7.621911922074787e-05,
+      "loss": 1.6685,
+      "step": 3122
+    },
+    {
+      "epoch": 0.32440012464942347,
+      "grad_norm": 0.404633492231369,
+      "learning_rate": 7.620522454354644e-05,
+      "loss": 1.8067,
+      "step": 3123
+    },
+    {
+      "epoch": 0.32450399916900385,
+      "grad_norm": 0.3900499641895294,
+      "learning_rate": 7.619132707569299e-05,
+      "loss": 1.8046,
+      "step": 3124
+    },
+    {
+      "epoch": 0.3246078736885842,
+      "grad_norm": 0.4414837062358856,
+      "learning_rate": 7.617742681866749e-05,
+      "loss": 1.9807,
+      "step": 3125
+    },
+    {
+      "epoch": 0.32471174820816456,
+      "grad_norm": 0.3875233232975006,
+      "learning_rate": 7.616352377395025e-05,
+      "loss": 1.7041,
+      "step": 3126
+    },
+    {
+      "epoch": 0.3248156227277449,
+      "grad_norm": 0.4394824206829071,
+      "learning_rate": 7.614961794302178e-05,
+      "loss": 1.7103,
+      "step": 3127
+    },
+    {
+      "epoch": 0.3249194972473252,
+      "grad_norm": 0.38259413838386536,
+      "learning_rate": 7.6135709327363e-05,
+      "loss": 1.6377,
+      "step": 3128
+    },
+    {
+      "epoch": 0.3250233717669056,
+      "grad_norm": 0.4169695973396301,
+      "learning_rate": 7.612179792845504e-05,
+      "loss": 1.7518,
+      "step": 3129
+    },
+    {
+      "epoch": 0.3251272462864859,
+      "grad_norm": 0.4079383909702301,
+      "learning_rate": 7.610788374777935e-05,
+      "loss": 1.6768,
+      "step": 3130
+    },
+    {
+      "epoch": 0.32523112080606625,
+      "grad_norm": 0.4074368476867676,
+      "learning_rate": 7.609396678681771e-05,
+      "loss": 1.8424,
+      "step": 3131
+    },
+    {
+      "epoch": 0.32533499532564664,
+      "grad_norm": 0.41162872314453125,
+      "learning_rate": 7.608004704705212e-05,
+      "loss": 1.8505,
+      "step": 3132
+    },
+    {
+      "epoch": 0.32543886984522696,
+      "grad_norm": 0.41227778792381287,
+      "learning_rate": 7.606612452996495e-05,
+      "loss": 1.6893,
+      "step": 3133
+    },
+    {
+      "epoch": 0.3255427443648073,
+      "grad_norm": 0.3721327781677246,
+      "learning_rate": 7.605219923703886e-05,
+      "loss": 1.7895,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3256466188843877,
+      "grad_norm": 0.3553870916366577,
+      "learning_rate": 7.603827116975677e-05,
+      "loss": 1.5966,
+      "step": 3135
+    },
+    {
+      "epoch": 0.325750493403968,
+      "grad_norm": 0.35715457797050476,
+      "learning_rate": 7.602434032960189e-05,
+      "loss": 1.7413,
+      "step": 3136
+    },
+    {
+      "epoch": 0.3258543679235483,
+      "grad_norm": 0.39415115118026733,
+      "learning_rate": 7.601040671805779e-05,
+      "loss": 1.3362,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3259582424431287,
+      "grad_norm": 0.3815273344516754,
+      "learning_rate": 7.599647033660824e-05,
+      "loss": 1.6745,
+      "step": 3138
+    },
+    {
+      "epoch": 0.32606211696270904,
+      "grad_norm": 0.3810808062553406,
+      "learning_rate": 7.598253118673738e-05,
+      "loss": 1.6381,
+      "step": 3139
+    },
+    {
+      "epoch": 0.3261659914822894,
+      "grad_norm": 0.43945252895355225,
+      "learning_rate": 7.596858926992962e-05,
+      "loss": 1.8387,
+      "step": 3140
+    },
+    {
+      "epoch": 0.32626986600186975,
+      "grad_norm": 0.38324853777885437,
+      "learning_rate": 7.595464458766968e-05,
+      "loss": 1.6446,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3263737405214501,
+      "grad_norm": 0.4006134569644928,
+      "learning_rate": 7.594069714144252e-05,
+      "loss": 1.4691,
+      "step": 3142
+    },
+    {
+      "epoch": 0.32647761504103046,
+      "grad_norm": 0.4041348993778229,
+      "learning_rate": 7.592674693273348e-05,
+      "loss": 1.8108,
+      "step": 3143
+    },
+    {
+      "epoch": 0.3265814895606108,
+      "grad_norm": 0.380634605884552,
+      "learning_rate": 7.591279396302812e-05,
+      "loss": 1.7062,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3266853640801911,
+      "grad_norm": 0.3991457223892212,
+      "learning_rate": 7.589883823381234e-05,
+      "loss": 1.6658,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3267892385997715,
+      "grad_norm": 0.4109683334827423,
+      "learning_rate": 7.58848797465723e-05,
+      "loss": 1.7899,
+      "step": 3146
+    },
+    {
+      "epoch": 0.3268931131193518,
+      "grad_norm": 0.38975322246551514,
+      "learning_rate": 7.587091850279447e-05,
+      "loss": 1.7506,
+      "step": 3147
+    },
+    {
+      "epoch": 0.32699698763893215,
+      "grad_norm": 0.3871016800403595,
+      "learning_rate": 7.585695450396564e-05,
+      "loss": 1.7103,
+      "step": 3148
+    },
+    {
+      "epoch": 0.32710086215851253,
+      "grad_norm": 0.36475732922554016,
+      "learning_rate": 7.584298775157282e-05,
+      "loss": 1.5979,
+      "step": 3149
+    },
+    {
+      "epoch": 0.32720473667809286,
+      "grad_norm": 0.3877626359462738,
+      "learning_rate": 7.58290182471034e-05,
+      "loss": 1.8509,
+      "step": 3150
+    },
+    {
+      "epoch": 0.32730861119767324,
+      "grad_norm": 0.36147022247314453,
+      "learning_rate": 7.581504599204503e-05,
+      "loss": 1.686,
+      "step": 3151
+    },
+    {
+      "epoch": 0.32741248571725357,
+      "grad_norm": 0.38483577966690063,
+      "learning_rate": 7.58010709878856e-05,
+      "loss": 1.8448,
+      "step": 3152
+    },
+    {
+      "epoch": 0.3275163602368339,
+      "grad_norm": 0.415414035320282,
+      "learning_rate": 7.578709323611338e-05,
+      "loss": 1.7782,
+      "step": 3153
+    },
+    {
+      "epoch": 0.3276202347564143,
+      "grad_norm": 0.3928332030773163,
+      "learning_rate": 7.577311273821687e-05,
+      "loss": 1.7276,
+      "step": 3154
+    },
+    {
+      "epoch": 0.3277241092759946,
+      "grad_norm": 0.36280107498168945,
+      "learning_rate": 7.575912949568489e-05,
+      "loss": 1.7094,
+      "step": 3155
+    },
+    {
+      "epoch": 0.32782798379557493,
+      "grad_norm": 0.36135435104370117,
+      "learning_rate": 7.574514351000657e-05,
+      "loss": 1.6252,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3279318583151553,
+      "grad_norm": 0.38758692145347595,
+      "learning_rate": 7.573115478267126e-05,
+      "loss": 1.7715,
+      "step": 3157
+    },
+    {
+      "epoch": 0.32803573283473564,
+      "grad_norm": 0.39279675483703613,
+      "learning_rate": 7.571716331516869e-05,
+      "loss": 1.71,
+      "step": 3158
+    },
+    {
+      "epoch": 0.32813960735431597,
+      "grad_norm": 0.36881011724472046,
+      "learning_rate": 7.570316910898882e-05,
+      "loss": 1.6717,
+      "step": 3159
+    },
+    {
+      "epoch": 0.32824348187389635,
+      "grad_norm": 0.3746923506259918,
+      "learning_rate": 7.568917216562193e-05,
+      "loss": 1.6144,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3283473563934767,
+      "grad_norm": 0.4102190434932709,
+      "learning_rate": 7.56751724865586e-05,
+      "loss": 1.7229,
+      "step": 3161
+    },
+    {
+      "epoch": 0.328451230913057,
+      "grad_norm": 0.39784857630729675,
+      "learning_rate": 7.566117007328967e-05,
+      "loss": 1.6128,
+      "step": 3162
+    },
+    {
+      "epoch": 0.3285551054326374,
+      "grad_norm": 0.37368080019950867,
+      "learning_rate": 7.564716492730628e-05,
+      "loss": 1.8158,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3286589799522177,
+      "grad_norm": 0.4041743576526642,
+      "learning_rate": 7.56331570500999e-05,
+      "loss": 1.5258,
+      "step": 3164
+    },
+    {
+      "epoch": 0.3287628544717981,
+      "grad_norm": 0.4261610507965088,
+      "learning_rate": 7.561914644316221e-05,
+      "loss": 1.8395,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3288667289913784,
+      "grad_norm": 0.3845587372779846,
+      "learning_rate": 7.560513310798529e-05,
+      "loss": 1.7382,
+      "step": 3166
+    },
+    {
+      "epoch": 0.32897060351095875,
+      "grad_norm": 0.37753623723983765,
+      "learning_rate": 7.559111704606142e-05,
+      "loss": 1.7647,
+      "step": 3167
+    },
+    {
+      "epoch": 0.32907447803053913,
+      "grad_norm": 0.3923317492008209,
+      "learning_rate": 7.557709825888319e-05,
+      "loss": 1.6671,
+      "step": 3168
+    },
+    {
+      "epoch": 0.32917835255011946,
+      "grad_norm": 0.4080863893032074,
+      "learning_rate": 7.556307674794351e-05,
+      "loss": 1.8564,
+      "step": 3169
+    },
+    {
+      "epoch": 0.3292822270696998,
+      "grad_norm": 0.38211774826049805,
+      "learning_rate": 7.554905251473557e-05,
+      "loss": 1.8405,
+      "step": 3170
+    },
+    {
+      "epoch": 0.32938610158928017,
+      "grad_norm": 0.3729119896888733,
+      "learning_rate": 7.55350255607528e-05,
+      "loss": 1.6871,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3294899761088605,
+      "grad_norm": 0.36250361800193787,
+      "learning_rate": 7.5520995887489e-05,
+      "loss": 1.6192,
+      "step": 3172
+    },
+    {
+      "epoch": 0.3295938506284408,
+      "grad_norm": 0.4741998314857483,
+      "learning_rate": 7.550696349643822e-05,
+      "loss": 1.9924,
+      "step": 3173
+    },
+    {
+      "epoch": 0.3296977251480212,
+      "grad_norm": 0.39271846413612366,
+      "learning_rate": 7.549292838909478e-05,
+      "loss": 1.7255,
+      "step": 3174
+    },
+    {
+      "epoch": 0.32980159966760153,
+      "grad_norm": 0.4134219288825989,
+      "learning_rate": 7.547889056695332e-05,
+      "loss": 1.7508,
+      "step": 3175
+    },
+    {
+      "epoch": 0.32990547418718186,
+      "grad_norm": 0.377145379781723,
+      "learning_rate": 7.546485003150876e-05,
+      "loss": 1.7239,
+      "step": 3176
+    },
+    {
+      "epoch": 0.33000934870676224,
+      "grad_norm": 0.37137743830680847,
+      "learning_rate": 7.54508067842563e-05,
+      "loss": 1.6987,
+      "step": 3177
+    },
+    {
+      "epoch": 0.33011322322634257,
+      "grad_norm": 0.36599913239479065,
+      "learning_rate": 7.543676082669146e-05,
+      "loss": 1.675,
+      "step": 3178
+    },
+    {
+      "epoch": 0.33021709774592295,
+      "grad_norm": 0.3733290731906891,
+      "learning_rate": 7.542271216030998e-05,
+      "loss": 1.6841,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3303209722655033,
+      "grad_norm": 0.3710775375366211,
+      "learning_rate": 7.540866078660797e-05,
+      "loss": 1.7604,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3304248467850836,
+      "grad_norm": 0.4162429869174957,
+      "learning_rate": 7.539460670708178e-05,
+      "loss": 1.7021,
+      "step": 3181
+    },
+    {
+      "epoch": 0.330528721304664,
+      "grad_norm": 0.4121178090572357,
+      "learning_rate": 7.538054992322806e-05,
+      "loss": 1.885,
+      "step": 3182
+    },
+    {
+      "epoch": 0.3306325958242443,
+      "grad_norm": 0.3530847728252411,
+      "learning_rate": 7.536649043654376e-05,
+      "loss": 1.3871,
+      "step": 3183
+    },
+    {
+      "epoch": 0.33073647034382464,
+      "grad_norm": 0.42063409090042114,
+      "learning_rate": 7.535242824852608e-05,
+      "loss": 1.7323,
+      "step": 3184
+    },
+    {
+      "epoch": 0.330840344863405,
+      "grad_norm": 0.4144652485847473,
+      "learning_rate": 7.533836336067257e-05,
+      "loss": 1.7173,
+      "step": 3185
+    },
+    {
+      "epoch": 0.33094421938298535,
+      "grad_norm": 0.3715633749961853,
+      "learning_rate": 7.5324295774481e-05,
+      "loss": 1.6793,
+      "step": 3186
+    },
+    {
+      "epoch": 0.3310480939025657,
+      "grad_norm": 0.40452414751052856,
+      "learning_rate": 7.531022549144946e-05,
+      "loss": 1.7443,
+      "step": 3187
+    },
+    {
+      "epoch": 0.33115196842214606,
+      "grad_norm": 0.3777833878993988,
+      "learning_rate": 7.529615251307631e-05,
+      "loss": 1.675,
+      "step": 3188
+    },
+    {
+      "epoch": 0.3312558429417264,
+      "grad_norm": 0.3736540973186493,
+      "learning_rate": 7.528207684086027e-05,
+      "loss": 1.7026,
+      "step": 3189
+    },
+    {
+      "epoch": 0.33135971746130677,
+      "grad_norm": 0.3794771134853363,
+      "learning_rate": 7.526799847630024e-05,
+      "loss": 1.6583,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3314635919808871,
+      "grad_norm": 0.38491761684417725,
+      "learning_rate": 7.525391742089547e-05,
+      "loss": 1.6339,
+      "step": 3191
+    },
+    {
+      "epoch": 0.3315674665004674,
+      "grad_norm": 0.3660680055618286,
+      "learning_rate": 7.52398336761455e-05,
+      "loss": 1.7295,
+      "step": 3192
+    },
+    {
+      "epoch": 0.3316713410200478,
+      "grad_norm": 0.4049983024597168,
+      "learning_rate": 7.522574724355009e-05,
+      "loss": 1.8415,
+      "step": 3193
+    },
+    {
+      "epoch": 0.33177521553962813,
+      "grad_norm": 0.4031364917755127,
+      "learning_rate": 7.52116581246094e-05,
+      "loss": 1.8444,
+      "step": 3194
+    },
+    {
+      "epoch": 0.33187909005920846,
+      "grad_norm": 0.3920493721961975,
+      "learning_rate": 7.519756632082376e-05,
+      "loss": 1.9211,
+      "step": 3195
+    },
+    {
+      "epoch": 0.33198296457878884,
+      "grad_norm": 0.40728676319122314,
+      "learning_rate": 7.518347183369385e-05,
+      "loss": 1.5216,
+      "step": 3196
+    },
+    {
+      "epoch": 0.33208683909836917,
+      "grad_norm": 0.3954370617866516,
+      "learning_rate": 7.516937466472065e-05,
+      "loss": 1.6549,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3321907136179495,
+      "grad_norm": 0.3614353537559509,
+      "learning_rate": 7.515527481540536e-05,
+      "loss": 1.6972,
+      "step": 3198
+    },
+    {
+      "epoch": 0.3322945881375299,
+      "grad_norm": 0.35628658533096313,
+      "learning_rate": 7.514117228724953e-05,
+      "loss": 1.6622,
+      "step": 3199
+    },
+    {
+      "epoch": 0.3323984626571102,
+      "grad_norm": 0.3558284342288971,
+      "learning_rate": 7.512706708175496e-05,
+      "loss": 1.5866,
+      "step": 3200
+    },
+    {
+      "epoch": 0.33250233717669053,
+      "grad_norm": 0.3872045576572418,
+      "learning_rate": 7.511295920042374e-05,
+      "loss": 1.7721,
+      "step": 3201
+    },
+    {
+      "epoch": 0.3326062116962709,
+      "grad_norm": 0.3917664587497711,
+      "learning_rate": 7.509884864475825e-05,
+      "loss": 1.7278,
+      "step": 3202
+    },
+    {
+      "epoch": 0.33271008621585124,
+      "grad_norm": 0.3571791648864746,
+      "learning_rate": 7.508473541626115e-05,
+      "loss": 1.7481,
+      "step": 3203
+    },
+    {
+      "epoch": 0.3328139607354316,
+      "grad_norm": 0.3802284002304077,
+      "learning_rate": 7.507061951643541e-05,
+      "loss": 1.7085,
+      "step": 3204
+    },
+    {
+      "epoch": 0.33291783525501195,
+      "grad_norm": 0.41221603751182556,
+      "learning_rate": 7.505650094678425e-05,
+      "loss": 1.8142,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3330217097745923,
+      "grad_norm": 0.41970524191856384,
+      "learning_rate": 7.504237970881118e-05,
+      "loss": 1.7213,
+      "step": 3206
+    },
+    {
+      "epoch": 0.33312558429417266,
+      "grad_norm": 0.38147467374801636,
+      "learning_rate": 7.502825580402002e-05,
+      "loss": 1.637,
+      "step": 3207
+    },
+    {
+      "epoch": 0.333229458813753,
+      "grad_norm": 0.430268257856369,
+      "learning_rate": 7.501412923391483e-05,
+      "loss": 1.7966,
+      "step": 3208
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.422599732875824,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.7097,
+      "step": 3209
+    },
+    {
+      "epoch": 0.3334372078529137,
+      "grad_norm": 0.3947455585002899,
+      "learning_rate": 7.498586810378019e-05,
+      "loss": 1.6064,
+      "step": 3210
+    },
+    {
+      "epoch": 0.333541082372494,
+      "grad_norm": 0.43581533432006836,
+      "learning_rate": 7.49717335467603e-05,
+      "loss": 1.8244,
+      "step": 3211
+    },
+    {
+      "epoch": 0.33364495689207435,
+      "grad_norm": 0.42619240283966064,
+      "learning_rate": 7.49575963304456e-05,
+      "loss": 1.8339,
+      "step": 3212
+    },
+    {
+      "epoch": 0.33374883141165473,
+      "grad_norm": 0.4488205015659332,
+      "learning_rate": 7.494345645634156e-05,
+      "loss": 1.6953,
+      "step": 3213
+    },
+    {
+      "epoch": 0.33385270593123506,
+      "grad_norm": 0.35742267966270447,
+      "learning_rate": 7.492931392595397e-05,
+      "loss": 1.7246,
+      "step": 3214
+    },
+    {
+      "epoch": 0.3339565804508154,
+      "grad_norm": 0.37364768981933594,
+      "learning_rate": 7.49151687407889e-05,
+      "loss": 1.6528,
+      "step": 3215
+    },
+    {
+      "epoch": 0.33406045497039577,
+      "grad_norm": 0.3801439702510834,
+      "learning_rate": 7.49010209023527e-05,
+      "loss": 1.5913,
+      "step": 3216
+    },
+    {
+      "epoch": 0.3341643294899761,
+      "grad_norm": 0.37618300318717957,
+      "learning_rate": 7.488687041215202e-05,
+      "loss": 1.609,
+      "step": 3217
+    },
+    {
+      "epoch": 0.3342682040095565,
+      "grad_norm": 0.37247294187545776,
+      "learning_rate": 7.487271727169377e-05,
+      "loss": 1.6473,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3343720785291368,
+      "grad_norm": 0.3996807336807251,
+      "learning_rate": 7.485856148248513e-05,
+      "loss": 1.7744,
+      "step": 3219
+    },
+    {
+      "epoch": 0.33447595304871713,
+      "grad_norm": 0.3898894488811493,
+      "learning_rate": 7.484440304603362e-05,
+      "loss": 1.8776,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3345798275682975,
+      "grad_norm": 0.4372991621494293,
+      "learning_rate": 7.483024196384696e-05,
+      "loss": 1.8293,
+      "step": 3221
+    },
+    {
+      "epoch": 0.33468370208787784,
+      "grad_norm": 0.4678659439086914,
+      "learning_rate": 7.481607823743321e-05,
+      "loss": 1.8113,
+      "step": 3222
+    },
+    {
+      "epoch": 0.33478757660745817,
+      "grad_norm": 0.38630014657974243,
+      "learning_rate": 7.480191186830072e-05,
+      "loss": 1.6579,
+      "step": 3223
+    },
+    {
+      "epoch": 0.33489145112703855,
+      "grad_norm": 0.4188450276851654,
+      "learning_rate": 7.478774285795805e-05,
+      "loss": 1.9302,
+      "step": 3224
+    },
+    {
+      "epoch": 0.3349953256466189,
+      "grad_norm": 0.3890570402145386,
+      "learning_rate": 7.477357120791415e-05,
+      "loss": 1.8298,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3350992001661992,
+      "grad_norm": 0.4579260051250458,
+      "learning_rate": 7.475939691967814e-05,
+      "loss": 1.673,
+      "step": 3226
+    },
+    {
+      "epoch": 0.3352030746857796,
+      "grad_norm": 0.39062821865081787,
+      "learning_rate": 7.47452199947595e-05,
+      "loss": 1.7051,
+      "step": 3227
+    },
+    {
+      "epoch": 0.3353069492053599,
+      "grad_norm": 0.37739062309265137,
+      "learning_rate": 7.473104043466794e-05,
+      "loss": 1.6151,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3354108237249403,
+      "grad_norm": 0.3767510950565338,
+      "learning_rate": 7.471685824091348e-05,
+      "loss": 1.7629,
+      "step": 3229
+    },
+    {
+      "epoch": 0.3355146982445206,
+      "grad_norm": 0.41610831022262573,
+      "learning_rate": 7.470267341500642e-05,
+      "loss": 1.7987,
+      "step": 3230
+    },
+    {
+      "epoch": 0.33561857276410095,
+      "grad_norm": 0.3895459771156311,
+      "learning_rate": 7.468848595845733e-05,
+      "loss": 1.5922,
+      "step": 3231
+    },
+    {
+      "epoch": 0.33572244728368134,
+      "grad_norm": 0.3886757493019104,
+      "learning_rate": 7.467429587277705e-05,
+      "loss": 1.6534,
+      "step": 3232
+    },
+    {
+      "epoch": 0.33582632180326166,
+      "grad_norm": 0.37435784935951233,
+      "learning_rate": 7.466010315947676e-05,
+      "loss": 1.7319,
+      "step": 3233
+    },
+    {
+      "epoch": 0.335930196322842,
+      "grad_norm": 0.4168533682823181,
+      "learning_rate": 7.464590782006782e-05,
+      "loss": 1.8126,
+      "step": 3234
+    },
+    {
+      "epoch": 0.33603407084242237,
+      "grad_norm": 0.3834715187549591,
+      "learning_rate": 7.463170985606194e-05,
+      "loss": 1.7701,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3361379453620027,
+      "grad_norm": 0.44510364532470703,
+      "learning_rate": 7.46175092689711e-05,
+      "loss": 1.8166,
+      "step": 3236
+    },
+    {
+      "epoch": 0.336241819881583,
+      "grad_norm": 0.41382500529289246,
+      "learning_rate": 7.460330606030754e-05,
+      "loss": 1.9408,
+      "step": 3237
+    },
+    {
+      "epoch": 0.3363456944011634,
+      "grad_norm": 0.3592391014099121,
+      "learning_rate": 7.45891002315838e-05,
+      "loss": 1.6185,
+      "step": 3238
+    },
+    {
+      "epoch": 0.33644956892074374,
+      "grad_norm": 0.36257967352867126,
+      "learning_rate": 7.457489178431269e-05,
+      "loss": 1.5973,
+      "step": 3239
+    },
+    {
+      "epoch": 0.33655344344032406,
+      "grad_norm": 0.368685245513916,
+      "learning_rate": 7.45606807200073e-05,
+      "loss": 1.5255,
+      "step": 3240
+    },
+    {
+      "epoch": 0.33665731795990445,
+      "grad_norm": 0.3916224241256714,
+      "learning_rate": 7.4546467040181e-05,
+      "loss": 1.6482,
+      "step": 3241
+    },
+    {
+      "epoch": 0.33676119247948477,
+      "grad_norm": 0.36512747406959534,
+      "learning_rate": 7.453225074634742e-05,
+      "loss": 1.5935,
+      "step": 3242
+    },
+    {
+      "epoch": 0.33686506699906515,
+      "grad_norm": 0.40340378880500793,
+      "learning_rate": 7.451803184002048e-05,
+      "loss": 1.8217,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3369689415186455,
+      "grad_norm": 0.4013363718986511,
+      "learning_rate": 7.450381032271442e-05,
+      "loss": 1.698,
+      "step": 3244
+    },
+    {
+      "epoch": 0.3370728160382258,
+      "grad_norm": 0.4022335708141327,
+      "learning_rate": 7.44895861959437e-05,
+      "loss": 1.7154,
+      "step": 3245
+    },
+    {
+      "epoch": 0.3371766905578062,
+      "grad_norm": 0.4084652364253998,
+      "learning_rate": 7.447535946122304e-05,
+      "loss": 1.6902,
+      "step": 3246
+    },
+    {
+      "epoch": 0.3372805650773865,
+      "grad_norm": 0.41394492983818054,
+      "learning_rate": 7.446113012006756e-05,
+      "loss": 1.6136,
+      "step": 3247
+    },
+    {
+      "epoch": 0.33738443959696685,
+      "grad_norm": 0.3942053020000458,
+      "learning_rate": 7.444689817399251e-05,
+      "loss": 1.7797,
+      "step": 3248
+    },
+    {
+      "epoch": 0.33748831411654723,
+      "grad_norm": 0.41565534472465515,
+      "learning_rate": 7.443266362451349e-05,
+      "loss": 1.7013,
+      "step": 3249
+    },
+    {
+      "epoch": 0.33759218863612755,
+      "grad_norm": 0.3893018960952759,
+      "learning_rate": 7.441842647314638e-05,
+      "loss": 1.7065,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3376960631557079,
+      "grad_norm": 0.4059962034225464,
+      "learning_rate": 7.440418672140733e-05,
+      "loss": 1.7509,
+      "step": 3251
+    },
+    {
+      "epoch": 0.33779993767528826,
+      "grad_norm": 0.4266146719455719,
+      "learning_rate": 7.438994437081272e-05,
+      "loss": 1.7629,
+      "step": 3252
+    },
+    {
+      "epoch": 0.3379038121948686,
+      "grad_norm": 0.37736377120018005,
+      "learning_rate": 7.437569942287932e-05,
+      "loss": 1.8509,
+      "step": 3253
+    },
+    {
+      "epoch": 0.3380076867144489,
+      "grad_norm": 0.3954285681247711,
+      "learning_rate": 7.436145187912406e-05,
+      "loss": 1.6512,
+      "step": 3254
+    },
+    {
+      "epoch": 0.3381115612340293,
+      "grad_norm": 0.3633100092411041,
+      "learning_rate": 7.43472017410642e-05,
+      "loss": 1.5277,
+      "step": 3255
+    },
+    {
+      "epoch": 0.33821543575360963,
+      "grad_norm": 0.3525071144104004,
+      "learning_rate": 7.433294901021727e-05,
+      "loss": 1.5944,
+      "step": 3256
+    },
+    {
+      "epoch": 0.33831931027319,
+      "grad_norm": 0.4135834574699402,
+      "learning_rate": 7.431869368810106e-05,
+      "loss": 1.6653,
+      "step": 3257
+    },
+    {
+      "epoch": 0.33842318479277034,
+      "grad_norm": 0.40316396951675415,
+      "learning_rate": 7.430443577623369e-05,
+      "loss": 1.7842,
+      "step": 3258
+    },
+    {
+      "epoch": 0.33852705931235066,
+      "grad_norm": 0.3851187825202942,
+      "learning_rate": 7.429017527613347e-05,
+      "loss": 1.721,
+      "step": 3259
+    },
+    {
+      "epoch": 0.33863093383193105,
+      "grad_norm": 0.4158236086368561,
+      "learning_rate": 7.427591218931906e-05,
+      "loss": 1.9518,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3387348083515114,
+      "grad_norm": 0.3790571689605713,
+      "learning_rate": 7.426164651730937e-05,
+      "loss": 1.5893,
+      "step": 3261
+    },
+    {
+      "epoch": 0.3388386828710917,
+      "grad_norm": 0.37710022926330566,
+      "learning_rate": 7.424737826162358e-05,
+      "loss": 1.6549,
+      "step": 3262
+    },
+    {
+      "epoch": 0.3389425573906721,
+      "grad_norm": 0.3882453441619873,
+      "learning_rate": 7.423310742378113e-05,
+      "loss": 1.6472,
+      "step": 3263
+    },
+    {
+      "epoch": 0.3390464319102524,
+      "grad_norm": 0.39669013023376465,
+      "learning_rate": 7.421883400530176e-05,
+      "loss": 1.7416,
+      "step": 3264
+    },
+    {
+      "epoch": 0.33915030642983274,
+      "grad_norm": 0.41558176279067993,
+      "learning_rate": 7.42045580077055e-05,
+      "loss": 1.6492,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3392541809494131,
+      "grad_norm": 0.4130480885505676,
+      "learning_rate": 7.41902794325126e-05,
+      "loss": 1.7518,
+      "step": 3266
+    },
+    {
+      "epoch": 0.33935805546899345,
+      "grad_norm": 0.398885577917099,
+      "learning_rate": 7.417599828124363e-05,
+      "loss": 1.6747,
+      "step": 3267
+    },
+    {
+      "epoch": 0.33946192998857383,
+      "grad_norm": 0.39864227175712585,
+      "learning_rate": 7.416171455541943e-05,
+      "loss": 1.7727,
+      "step": 3268
+    },
+    {
+      "epoch": 0.33956580450815416,
+      "grad_norm": 0.4171668589115143,
+      "learning_rate": 7.414742825656111e-05,
+      "loss": 1.6616,
+      "step": 3269
+    },
+    {
+      "epoch": 0.3396696790277345,
+      "grad_norm": 0.379443883895874,
+      "learning_rate": 7.413313938619003e-05,
+      "loss": 1.8302,
+      "step": 3270
+    },
+    {
+      "epoch": 0.33977355354731487,
+      "grad_norm": 0.3963342607021332,
+      "learning_rate": 7.411884794582784e-05,
+      "loss": 1.7629,
+      "step": 3271
+    },
+    {
+      "epoch": 0.3398774280668952,
+      "grad_norm": 0.3690603971481323,
+      "learning_rate": 7.410455393699647e-05,
+      "loss": 1.6549,
+      "step": 3272
+    },
+    {
+      "epoch": 0.3399813025864755,
+      "grad_norm": 0.36229610443115234,
+      "learning_rate": 7.409025736121816e-05,
+      "loss": 1.6949,
+      "step": 3273
+    },
+    {
+      "epoch": 0.3400851771060559,
+      "grad_norm": 0.3683745265007019,
+      "learning_rate": 7.407595822001532e-05,
+      "loss": 1.8114,
+      "step": 3274
+    },
+    {
+      "epoch": 0.34018905162563623,
+      "grad_norm": 0.3948541283607483,
+      "learning_rate": 7.406165651491075e-05,
+      "loss": 1.6883,
+      "step": 3275
+    },
+    {
+      "epoch": 0.34029292614521656,
+      "grad_norm": 0.4150826036930084,
+      "learning_rate": 7.404735224742745e-05,
+      "loss": 1.7254,
+      "step": 3276
+    },
+    {
+      "epoch": 0.34039680066479694,
+      "grad_norm": 0.4049845337867737,
+      "learning_rate": 7.40330454190887e-05,
+      "loss": 1.6418,
+      "step": 3277
+    },
+    {
+      "epoch": 0.34050067518437727,
+      "grad_norm": 0.37728235125541687,
+      "learning_rate": 7.401873603141809e-05,
+      "loss": 1.8138,
+      "step": 3278
+    },
+    {
+      "epoch": 0.3406045497039576,
+      "grad_norm": 0.4334777891635895,
+      "learning_rate": 7.400442408593944e-05,
+      "loss": 1.9727,
+      "step": 3279
+    },
+    {
+      "epoch": 0.340708424223538,
+      "grad_norm": 0.395805299282074,
+      "learning_rate": 7.399010958417684e-05,
+      "loss": 1.7511,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3408122987431183,
+      "grad_norm": 0.3635464310646057,
+      "learning_rate": 7.397579252765475e-05,
+      "loss": 1.7119,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3409161732626987,
+      "grad_norm": 0.3630931079387665,
+      "learning_rate": 7.396147291789775e-05,
+      "loss": 1.7253,
+      "step": 3282
+    },
+    {
+      "epoch": 0.341020047782279,
+      "grad_norm": 0.37721481919288635,
+      "learning_rate": 7.394715075643078e-05,
+      "loss": 1.7831,
+      "step": 3283
+    },
+    {
+      "epoch": 0.34112392230185934,
+      "grad_norm": 0.41346338391304016,
+      "learning_rate": 7.393282604477907e-05,
+      "loss": 1.8239,
+      "step": 3284
+    },
+    {
+      "epoch": 0.3412277968214397,
+      "grad_norm": 0.3926272392272949,
+      "learning_rate": 7.391849878446805e-05,
+      "loss": 1.8158,
+      "step": 3285
+    },
+    {
+      "epoch": 0.34133167134102005,
+      "grad_norm": 0.37988725304603577,
+      "learning_rate": 7.39041689770235e-05,
+      "loss": 1.6549,
+      "step": 3286
+    },
+    {
+      "epoch": 0.3414355458606004,
+      "grad_norm": 0.46685877442359924,
+      "learning_rate": 7.388983662397139e-05,
+      "loss": 1.8314,
+      "step": 3287
+    },
+    {
+      "epoch": 0.34153942038018076,
+      "grad_norm": 0.38952314853668213,
+      "learning_rate": 7.387550172683803e-05,
+      "loss": 1.9213,
+      "step": 3288
+    },
+    {
+      "epoch": 0.3416432948997611,
+      "grad_norm": 0.3613092005252838,
+      "learning_rate": 7.386116428715e-05,
+      "loss": 1.7343,
+      "step": 3289
+    },
+    {
+      "epoch": 0.3417471694193414,
+      "grad_norm": 0.3758079707622528,
+      "learning_rate": 7.384682430643409e-05,
+      "loss": 1.6233,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3418510439389218,
+      "grad_norm": 0.36782777309417725,
+      "learning_rate": 7.383248178621739e-05,
+      "loss": 1.7206,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3419549184585021,
+      "grad_norm": 0.36162370443344116,
+      "learning_rate": 7.381813672802729e-05,
+      "loss": 1.6387,
+      "step": 3292
+    },
+    {
+      "epoch": 0.34205879297808245,
+      "grad_norm": 0.3870641887187958,
+      "learning_rate": 7.380378913339142e-05,
+      "loss": 1.7021,
+      "step": 3293
+    },
+    {
+      "epoch": 0.34216266749766283,
+      "grad_norm": 0.366452157497406,
+      "learning_rate": 7.378943900383766e-05,
+      "loss": 1.6213,
+      "step": 3294
+    },
+    {
+      "epoch": 0.34226654201724316,
+      "grad_norm": 0.40911996364593506,
+      "learning_rate": 7.377508634089422e-05,
+      "loss": 1.8177,
+      "step": 3295
+    },
+    {
+      "epoch": 0.34237041653682354,
+      "grad_norm": 0.396062433719635,
+      "learning_rate": 7.376073114608955e-05,
+      "loss": 1.7282,
+      "step": 3296
+    },
+    {
+      "epoch": 0.34247429105640387,
+      "grad_norm": 0.4013812839984894,
+      "learning_rate": 7.374637342095236e-05,
+      "loss": 1.7302,
+      "step": 3297
+    },
+    {
+      "epoch": 0.3425781655759842,
+      "grad_norm": 0.37226590514183044,
+      "learning_rate": 7.373201316701162e-05,
+      "loss": 1.6312,
+      "step": 3298
+    },
+    {
+      "epoch": 0.3426820400955646,
+      "grad_norm": 0.39679157733917236,
+      "learning_rate": 7.371765038579658e-05,
+      "loss": 1.6591,
+      "step": 3299
+    },
+    {
+      "epoch": 0.3427859146151449,
+      "grad_norm": 0.4248996078968048,
+      "learning_rate": 7.37032850788368e-05,
+      "loss": 1.8358,
+      "step": 3300
+    },
+    {
+      "epoch": 0.34288978913472523,
+      "grad_norm": 0.37451761960983276,
+      "learning_rate": 7.368891724766204e-05,
+      "loss": 1.7225,
+      "step": 3301
+    },
+    {
+      "epoch": 0.3429936636543056,
+      "grad_norm": 0.38723859190940857,
+      "learning_rate": 7.367454689380238e-05,
+      "loss": 1.5813,
+      "step": 3302
+    },
+    {
+      "epoch": 0.34309753817388594,
+      "grad_norm": 0.4239073395729065,
+      "learning_rate": 7.366017401878813e-05,
+      "loss": 1.8608,
+      "step": 3303
+    },
+    {
+      "epoch": 0.34320141269346627,
+      "grad_norm": 0.39057475328445435,
+      "learning_rate": 7.364579862414993e-05,
+      "loss": 1.7135,
+      "step": 3304
+    },
+    {
+      "epoch": 0.34330528721304665,
+      "grad_norm": 0.4171145558357239,
+      "learning_rate": 7.36314207114186e-05,
+      "loss": 1.8721,
+      "step": 3305
+    },
+    {
+      "epoch": 0.343409161732627,
+      "grad_norm": 0.4564177095890045,
+      "learning_rate": 7.36170402821253e-05,
+      "loss": 1.6559,
+      "step": 3306
+    },
+    {
+      "epoch": 0.34351303625220736,
+      "grad_norm": 0.4414724111557007,
+      "learning_rate": 7.360265733780142e-05,
+      "loss": 1.7733,
+      "step": 3307
+    },
+    {
+      "epoch": 0.3436169107717877,
+      "grad_norm": 0.35561293363571167,
+      "learning_rate": 7.358827187997867e-05,
+      "loss": 1.5218,
+      "step": 3308
+    },
+    {
+      "epoch": 0.343720785291368,
+      "grad_norm": 0.41624075174331665,
+      "learning_rate": 7.357388391018893e-05,
+      "loss": 1.6683,
+      "step": 3309
+    },
+    {
+      "epoch": 0.3438246598109484,
+      "grad_norm": 0.38379234075546265,
+      "learning_rate": 7.355949342996447e-05,
+      "loss": 1.7642,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3439285343305287,
+      "grad_norm": 0.40274274349212646,
+      "learning_rate": 7.35451004408377e-05,
+      "loss": 1.7529,
+      "step": 3311
+    },
+    {
+      "epoch": 0.34403240885010905,
+      "grad_norm": 0.4023614525794983,
+      "learning_rate": 7.353070494434141e-05,
+      "loss": 1.77,
+      "step": 3312
+    },
+    {
+      "epoch": 0.34413628336968943,
+      "grad_norm": 0.41878741979599,
+      "learning_rate": 7.351630694200857e-05,
+      "loss": 1.7528,
+      "step": 3313
+    },
+    {
+      "epoch": 0.34424015788926976,
+      "grad_norm": 0.3700146973133087,
+      "learning_rate": 7.35019064353725e-05,
+      "loss": 1.677,
+      "step": 3314
+    },
+    {
+      "epoch": 0.3443440324088501,
+      "grad_norm": 0.44311147928237915,
+      "learning_rate": 7.348750342596671e-05,
+      "loss": 1.9235,
+      "step": 3315
+    },
+    {
+      "epoch": 0.34444790692843047,
+      "grad_norm": 0.4150424599647522,
+      "learning_rate": 7.3473097915325e-05,
+      "loss": 1.6732,
+      "step": 3316
+    },
+    {
+      "epoch": 0.3445517814480108,
+      "grad_norm": 0.3866238594055176,
+      "learning_rate": 7.345868990498146e-05,
+      "loss": 1.6598,
+      "step": 3317
+    },
+    {
+      "epoch": 0.3446556559675911,
+      "grad_norm": 0.42701780796051025,
+      "learning_rate": 7.344427939647045e-05,
+      "loss": 1.7695,
+      "step": 3318
+    },
+    {
+      "epoch": 0.3447595304871715,
+      "grad_norm": 0.3794342577457428,
+      "learning_rate": 7.342986639132654e-05,
+      "loss": 1.6552,
+      "step": 3319
+    },
+    {
+      "epoch": 0.34486340500675183,
+      "grad_norm": 0.4094541668891907,
+      "learning_rate": 7.341545089108464e-05,
+      "loss": 1.9607,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3449672795263322,
+      "grad_norm": 0.4172968566417694,
+      "learning_rate": 7.340103289727983e-05,
+      "loss": 1.7933,
+      "step": 3321
+    },
+    {
+      "epoch": 0.34507115404591254,
+      "grad_norm": 0.402832955121994,
+      "learning_rate": 7.338661241144758e-05,
+      "loss": 1.8795,
+      "step": 3322
+    },
+    {
+      "epoch": 0.34517502856549287,
+      "grad_norm": 0.38485434651374817,
+      "learning_rate": 7.337218943512353e-05,
+      "loss": 1.764,
+      "step": 3323
+    },
+    {
+      "epoch": 0.34527890308507325,
+      "grad_norm": 0.4096323847770691,
+      "learning_rate": 7.335776396984361e-05,
+      "loss": 1.814,
+      "step": 3324
+    },
+    {
+      "epoch": 0.3453827776046536,
+      "grad_norm": 0.347913920879364,
+      "learning_rate": 7.334333601714403e-05,
+      "loss": 1.5926,
+      "step": 3325
+    },
+    {
+      "epoch": 0.3454866521242339,
+      "grad_norm": 0.404429167509079,
+      "learning_rate": 7.332890557856125e-05,
+      "loss": 1.8246,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3455905266438143,
+      "grad_norm": 0.3735859990119934,
+      "learning_rate": 7.331447265563197e-05,
+      "loss": 1.5425,
+      "step": 3327
+    },
+    {
+      "epoch": 0.3456944011633946,
+      "grad_norm": 0.35603833198547363,
+      "learning_rate": 7.330003724989324e-05,
+      "loss": 1.5314,
+      "step": 3328
+    },
+    {
+      "epoch": 0.34579827568297494,
+      "grad_norm": 0.3887482285499573,
+      "learning_rate": 7.328559936288228e-05,
+      "loss": 1.688,
+      "step": 3329
+    },
+    {
+      "epoch": 0.3459021502025553,
+      "grad_norm": 0.4250495731830597,
+      "learning_rate": 7.327115899613662e-05,
+      "loss": 1.8245,
+      "step": 3330
+    },
+    {
+      "epoch": 0.34600602472213565,
+      "grad_norm": 0.36763063073158264,
+      "learning_rate": 7.325671615119407e-05,
+      "loss": 1.4173,
+      "step": 3331
+    },
+    {
+      "epoch": 0.34610989924171603,
+      "grad_norm": 0.3886317312717438,
+      "learning_rate": 7.324227082959264e-05,
+      "loss": 1.7182,
+      "step": 3332
+    },
+    {
+      "epoch": 0.34621377376129636,
+      "grad_norm": 0.4716704487800598,
+      "learning_rate": 7.322782303287067e-05,
+      "loss": 1.6838,
+      "step": 3333
+    },
+    {
+      "epoch": 0.3463176482808767,
+      "grad_norm": 0.3932298719882965,
+      "learning_rate": 7.321337276256674e-05,
+      "loss": 1.8532,
+      "step": 3334
+    },
+    {
+      "epoch": 0.34642152280045707,
+      "grad_norm": 0.4240279793739319,
+      "learning_rate": 7.319892002021966e-05,
+      "loss": 1.7919,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3465253973200374,
+      "grad_norm": 0.41877415776252747,
+      "learning_rate": 7.318446480736857e-05,
+      "loss": 1.6985,
+      "step": 3336
+    },
+    {
+      "epoch": 0.3466292718396177,
+      "grad_norm": 0.45144563913345337,
+      "learning_rate": 7.317000712555283e-05,
+      "loss": 1.9947,
+      "step": 3337
+    },
+    {
+      "epoch": 0.3467331463591981,
+      "grad_norm": 0.3652666509151459,
+      "learning_rate": 7.315554697631205e-05,
+      "loss": 1.7206,
+      "step": 3338
+    },
+    {
+      "epoch": 0.34683702087877843,
+      "grad_norm": 0.3598730266094208,
+      "learning_rate": 7.314108436118614e-05,
+      "loss": 1.6871,
+      "step": 3339
+    },
+    {
+      "epoch": 0.34694089539835876,
+      "grad_norm": 0.3652264177799225,
+      "learning_rate": 7.312661928171525e-05,
+      "loss": 1.5859,
+      "step": 3340
+    },
+    {
+      "epoch": 0.34704476991793914,
+      "grad_norm": 0.36526259779930115,
+      "learning_rate": 7.311215173943982e-05,
+      "loss": 1.5721,
+      "step": 3341
+    },
+    {
+      "epoch": 0.34714864443751947,
+      "grad_norm": 0.3666329085826874,
+      "learning_rate": 7.309768173590049e-05,
+      "loss": 1.6762,
+      "step": 3342
+    },
+    {
+      "epoch": 0.3472525189570998,
+      "grad_norm": 0.39144429564476013,
+      "learning_rate": 7.308320927263823e-05,
+      "loss": 1.6709,
+      "step": 3343
+    },
+    {
+      "epoch": 0.3473563934766802,
+      "grad_norm": 0.41937124729156494,
+      "learning_rate": 7.306873435119423e-05,
+      "loss": 1.7338,
+      "step": 3344
+    },
+    {
+      "epoch": 0.3474602679962605,
+      "grad_norm": 0.38105306029319763,
+      "learning_rate": 7.305425697311e-05,
+      "loss": 1.5917,
+      "step": 3345
+    },
+    {
+      "epoch": 0.3475641425158409,
+      "grad_norm": 0.3741595149040222,
+      "learning_rate": 7.30397771399272e-05,
+      "loss": 1.6511,
+      "step": 3346
+    },
+    {
+      "epoch": 0.3476680170354212,
+      "grad_norm": 0.399387389421463,
+      "learning_rate": 7.302529485318786e-05,
+      "loss": 1.6948,
+      "step": 3347
+    },
+    {
+      "epoch": 0.34777189155500154,
+      "grad_norm": 0.3723578155040741,
+      "learning_rate": 7.301081011443423e-05,
+      "loss": 1.6319,
+      "step": 3348
+    },
+    {
+      "epoch": 0.3478757660745819,
+      "grad_norm": 0.37226778268814087,
+      "learning_rate": 7.29963229252088e-05,
+      "loss": 1.7856,
+      "step": 3349
+    },
+    {
+      "epoch": 0.34797964059416225,
+      "grad_norm": 0.38649681210517883,
+      "learning_rate": 7.298183328705436e-05,
+      "loss": 1.65,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3480835151137426,
+      "grad_norm": 0.3898165822029114,
+      "learning_rate": 7.296734120151394e-05,
+      "loss": 1.8236,
+      "step": 3351
+    },
+    {
+      "epoch": 0.34818738963332296,
+      "grad_norm": 0.37378284335136414,
+      "learning_rate": 7.295284667013083e-05,
+      "loss": 1.5903,
+      "step": 3352
+    },
+    {
+      "epoch": 0.3482912641529033,
+      "grad_norm": 0.39997485280036926,
+      "learning_rate": 7.293834969444861e-05,
+      "loss": 1.8107,
+      "step": 3353
+    },
+    {
+      "epoch": 0.3483951386724836,
+      "grad_norm": 0.37795597314834595,
+      "learning_rate": 7.292385027601105e-05,
+      "loss": 1.7677,
+      "step": 3354
+    },
+    {
+      "epoch": 0.348499013192064,
+      "grad_norm": 0.3679714500904083,
+      "learning_rate": 7.290934841636225e-05,
+      "loss": 1.6464,
+      "step": 3355
+    },
+    {
+      "epoch": 0.3486028877116443,
+      "grad_norm": 0.3939383029937744,
+      "learning_rate": 7.289484411704656e-05,
+      "loss": 1.7565,
+      "step": 3356
+    },
+    {
+      "epoch": 0.34870676223122465,
+      "grad_norm": 0.38755881786346436,
+      "learning_rate": 7.288033737960855e-05,
+      "loss": 1.5508,
+      "step": 3357
+    },
+    {
+      "epoch": 0.34881063675080504,
+      "grad_norm": 0.3855699598789215,
+      "learning_rate": 7.286582820559308e-05,
+      "loss": 1.6979,
+      "step": 3358
+    },
+    {
+      "epoch": 0.34891451127038536,
+      "grad_norm": 0.37389200925827026,
+      "learning_rate": 7.285131659654527e-05,
+      "loss": 1.703,
+      "step": 3359
+    },
+    {
+      "epoch": 0.34901838578996575,
+      "grad_norm": 0.3579116761684418,
+      "learning_rate": 7.283680255401049e-05,
+      "loss": 1.6348,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3491222603095461,
+      "grad_norm": 0.39986559748649597,
+      "learning_rate": 7.282228607953436e-05,
+      "loss": 1.7792,
+      "step": 3361
+    },
+    {
+      "epoch": 0.3492261348291264,
+      "grad_norm": 0.3745286166667938,
+      "learning_rate": 7.28077671746628e-05,
+      "loss": 1.7119,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3493300093487068,
+      "grad_norm": 0.45481473207473755,
+      "learning_rate": 7.279324584094194e-05,
+      "loss": 1.8279,
+      "step": 3363
+    },
+    {
+      "epoch": 0.3494338838682871,
+      "grad_norm": 0.42257899045944214,
+      "learning_rate": 7.277872207991818e-05,
+      "loss": 1.6366,
+      "step": 3364
+    },
+    {
+      "epoch": 0.34953775838786744,
+      "grad_norm": 0.3691715896129608,
+      "learning_rate": 7.276419589313821e-05,
+      "loss": 1.6741,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3496416329074478,
+      "grad_norm": 0.36931112408638,
+      "learning_rate": 7.274966728214895e-05,
+      "loss": 1.7447,
+      "step": 3366
+    },
+    {
+      "epoch": 0.34974550742702815,
+      "grad_norm": 0.4134025573730469,
+      "learning_rate": 7.273513624849757e-05,
+      "loss": 1.9171,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3498493819466085,
+      "grad_norm": 0.4000683128833771,
+      "learning_rate": 7.272060279373152e-05,
+      "loss": 1.7925,
+      "step": 3368
+    },
+    {
+      "epoch": 0.34995325646618886,
+      "grad_norm": 0.3902685344219208,
+      "learning_rate": 7.27060669193985e-05,
+      "loss": 1.6959,
+      "step": 3369
+    },
+    {
+      "epoch": 0.3500571309857692,
+      "grad_norm": 0.40209463238716125,
+      "learning_rate": 7.269152862704647e-05,
+      "loss": 1.7667,
+      "step": 3370
+    },
+    {
+      "epoch": 0.35016100550534957,
+      "grad_norm": 0.46758002042770386,
+      "learning_rate": 7.267698791822363e-05,
+      "loss": 1.8386,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3502648800249299,
+      "grad_norm": 0.4505844712257385,
+      "learning_rate": 7.266244479447847e-05,
+      "loss": 1.9704,
+      "step": 3372
+    },
+    {
+      "epoch": 0.3503687545445102,
+      "grad_norm": 0.4215162694454193,
+      "learning_rate": 7.264789925735974e-05,
+      "loss": 1.9084,
+      "step": 3373
+    },
+    {
+      "epoch": 0.3504726290640906,
+      "grad_norm": 0.4192196726799011,
+      "learning_rate": 7.263335130841638e-05,
+      "loss": 1.7802,
+      "step": 3374
+    },
+    {
+      "epoch": 0.35057650358367093,
+      "grad_norm": 0.4384503662586212,
+      "learning_rate": 7.261880094919765e-05,
+      "loss": 1.6945,
+      "step": 3375
+    },
+    {
+      "epoch": 0.35068037810325126,
+      "grad_norm": 0.4039691984653473,
+      "learning_rate": 7.260424818125305e-05,
+      "loss": 1.7436,
+      "step": 3376
+    },
+    {
+      "epoch": 0.35078425262283164,
+      "grad_norm": 0.3905966579914093,
+      "learning_rate": 7.258969300613233e-05,
+      "loss": 1.7358,
+      "step": 3377
+    },
+    {
+      "epoch": 0.35088812714241197,
+      "grad_norm": 0.4261592626571655,
+      "learning_rate": 7.257513542538553e-05,
+      "loss": 1.8172,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3509920016619923,
+      "grad_norm": 0.39150550961494446,
+      "learning_rate": 7.25605754405629e-05,
+      "loss": 1.6052,
+      "step": 3379
+    },
+    {
+      "epoch": 0.3510958761815727,
+      "grad_norm": 0.3744257092475891,
+      "learning_rate": 7.254601305321496e-05,
+      "loss": 1.7486,
+      "step": 3380
+    },
+    {
+      "epoch": 0.351199750701153,
+      "grad_norm": 0.36264219880104065,
+      "learning_rate": 7.253144826489249e-05,
+      "loss": 1.6068,
+      "step": 3381
+    },
+    {
+      "epoch": 0.35130362522073333,
+      "grad_norm": 0.35563141107559204,
+      "learning_rate": 7.251688107714654e-05,
+      "loss": 1.6542,
+      "step": 3382
+    },
+    {
+      "epoch": 0.3514074997403137,
+      "grad_norm": 0.39678165316581726,
+      "learning_rate": 7.250231149152838e-05,
+      "loss": 1.8309,
+      "step": 3383
+    },
+    {
+      "epoch": 0.35151137425989404,
+      "grad_norm": 0.3613341748714447,
+      "learning_rate": 7.248773950958958e-05,
+      "loss": 1.735,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3516152487794744,
+      "grad_norm": 0.3558713495731354,
+      "learning_rate": 7.247316513288192e-05,
+      "loss": 1.5545,
+      "step": 3385
+    },
+    {
+      "epoch": 0.35171912329905475,
+      "grad_norm": 0.3634355366230011,
+      "learning_rate": 7.245858836295749e-05,
+      "loss": 1.748,
+      "step": 3386
+    },
+    {
+      "epoch": 0.3518229978186351,
+      "grad_norm": 0.3754010796546936,
+      "learning_rate": 7.244400920136858e-05,
+      "loss": 1.703,
+      "step": 3387
+    },
+    {
+      "epoch": 0.35192687233821546,
+      "grad_norm": 0.380953848361969,
+      "learning_rate": 7.242942764966776e-05,
+      "loss": 1.6779,
+      "step": 3388
+    },
+    {
+      "epoch": 0.3520307468577958,
+      "grad_norm": 0.4804019033908844,
+      "learning_rate": 7.241484370940784e-05,
+      "loss": 1.7014,
+      "step": 3389
+    },
+    {
+      "epoch": 0.3521346213773761,
+      "grad_norm": 0.41277310252189636,
+      "learning_rate": 7.240025738214192e-05,
+      "loss": 1.6532,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3522384958969565,
+      "grad_norm": 0.4273497760295868,
+      "learning_rate": 7.238566866942332e-05,
+      "loss": 1.9818,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3523423704165368,
+      "grad_norm": 0.38524994254112244,
+      "learning_rate": 7.23710775728056e-05,
+      "loss": 1.7702,
+      "step": 3392
+    },
+    {
+      "epoch": 0.35244624493611715,
+      "grad_norm": 0.37887367606163025,
+      "learning_rate": 7.235648409384263e-05,
+      "loss": 1.6614,
+      "step": 3393
+    },
+    {
+      "epoch": 0.35255011945569753,
+      "grad_norm": 0.3723335564136505,
+      "learning_rate": 7.234188823408851e-05,
+      "loss": 1.6692,
+      "step": 3394
+    },
+    {
+      "epoch": 0.35265399397527786,
+      "grad_norm": 0.38009339570999146,
+      "learning_rate": 7.232728999509754e-05,
+      "loss": 1.7263,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3527578684948582,
+      "grad_norm": 0.4187556505203247,
+      "learning_rate": 7.231268937842438e-05,
+      "loss": 1.7052,
+      "step": 3396
+    },
+    {
+      "epoch": 0.35286174301443857,
+      "grad_norm": 0.3810417056083679,
+      "learning_rate": 7.229808638562381e-05,
+      "loss": 1.659,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3529656175340189,
+      "grad_norm": 0.41524961590766907,
+      "learning_rate": 7.228348101825099e-05,
+      "loss": 1.5006,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3530694920535993,
+      "grad_norm": 0.38853874802589417,
+      "learning_rate": 7.226887327786125e-05,
+      "loss": 1.8323,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3531733665731796,
+      "grad_norm": 0.3678146004676819,
+      "learning_rate": 7.22542631660102e-05,
+      "loss": 1.6357,
+      "step": 3400
+    },
+    {
+      "epoch": 0.35327724109275993,
+      "grad_norm": 0.3867931365966797,
+      "learning_rate": 7.223965068425376e-05,
+      "loss": 1.7819,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3533811156123403,
+      "grad_norm": 0.37931546568870544,
+      "learning_rate": 7.222503583414795e-05,
+      "loss": 1.6478,
+      "step": 3402
+    },
+    {
+      "epoch": 0.35348499013192064,
+      "grad_norm": 0.43002453446388245,
+      "learning_rate": 7.22104186172492e-05,
+      "loss": 1.8901,
+      "step": 3403
+    },
+    {
+      "epoch": 0.35358886465150097,
+      "grad_norm": 0.36890125274658203,
+      "learning_rate": 7.219579903511412e-05,
+      "loss": 1.7524,
+      "step": 3404
+    },
+    {
+      "epoch": 0.35369273917108135,
+      "grad_norm": 0.4097067415714264,
+      "learning_rate": 7.218117708929957e-05,
+      "loss": 1.6646,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3537966136906617,
+      "grad_norm": 0.47631388902664185,
+      "learning_rate": 7.216655278136269e-05,
+      "loss": 1.8199,
+      "step": 3406
+    },
+    {
+      "epoch": 0.353900488210242,
+      "grad_norm": 0.4068619906902313,
+      "learning_rate": 7.215192611286083e-05,
+      "loss": 1.5446,
+      "step": 3407
+    },
+    {
+      "epoch": 0.3540043627298224,
+      "grad_norm": 0.36159148812294006,
+      "learning_rate": 7.213729708535164e-05,
+      "loss": 1.4702,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3541082372494027,
+      "grad_norm": 0.39345964789390564,
+      "learning_rate": 7.212266570039299e-05,
+      "loss": 1.7324,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3542121117689831,
+      "grad_norm": 0.4041385352611542,
+      "learning_rate": 7.2108031959543e-05,
+      "loss": 1.7712,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3543159862885634,
+      "grad_norm": 0.40899065136909485,
+      "learning_rate": 7.209339586436005e-05,
+      "loss": 1.5812,
+      "step": 3411
+    },
+    {
+      "epoch": 0.35441986080814375,
+      "grad_norm": 0.37464478611946106,
+      "learning_rate": 7.20787574164028e-05,
+      "loss": 1.4846,
+      "step": 3412
+    },
+    {
+      "epoch": 0.35452373532772413,
+      "grad_norm": 0.41557177901268005,
+      "learning_rate": 7.206411661723007e-05,
+      "loss": 1.7985,
+      "step": 3413
+    },
+    {
+      "epoch": 0.35462760984730446,
+      "grad_norm": 0.4185909032821655,
+      "learning_rate": 7.204947346840106e-05,
+      "loss": 1.7178,
+      "step": 3414
+    },
+    {
+      "epoch": 0.3547314843668848,
+      "grad_norm": 0.40856650471687317,
+      "learning_rate": 7.203482797147512e-05,
+      "loss": 1.5386,
+      "step": 3415
+    },
+    {
+      "epoch": 0.35483535888646517,
+      "grad_norm": 0.40840035676956177,
+      "learning_rate": 7.202018012801185e-05,
+      "loss": 1.8462,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3549392334060455,
+      "grad_norm": 0.4034786820411682,
+      "learning_rate": 7.200552993957117e-05,
+      "loss": 1.6743,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3550431079256258,
+      "grad_norm": 0.39444810152053833,
+      "learning_rate": 7.199087740771321e-05,
+      "loss": 1.6801,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3551469824452062,
+      "grad_norm": 0.4487433433532715,
+      "learning_rate": 7.197622253399833e-05,
+      "loss": 1.9052,
+      "step": 3419
+    },
+    {
+      "epoch": 0.35525085696478653,
+      "grad_norm": 0.37087568640708923,
+      "learning_rate": 7.196156531998718e-05,
+      "loss": 1.6895,
+      "step": 3420
+    },
+    {
+      "epoch": 0.35535473148436686,
+      "grad_norm": 0.3976334035396576,
+      "learning_rate": 7.194690576724064e-05,
+      "loss": 1.6898,
+      "step": 3421
+    },
+    {
+      "epoch": 0.35545860600394724,
+      "grad_norm": 0.3944483697414398,
+      "learning_rate": 7.19322438773198e-05,
+      "loss": 1.7735,
+      "step": 3422
+    },
+    {
+      "epoch": 0.35556248052352757,
+      "grad_norm": 0.4049598276615143,
+      "learning_rate": 7.191757965178609e-05,
+      "loss": 1.7066,
+      "step": 3423
+    },
+    {
+      "epoch": 0.35566635504310795,
+      "grad_norm": 0.3984638452529907,
+      "learning_rate": 7.19029130922011e-05,
+      "loss": 1.622,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3557702295626883,
+      "grad_norm": 0.37969374656677246,
+      "learning_rate": 7.188824420012671e-05,
+      "loss": 1.5406,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3558741040822686,
+      "grad_norm": 0.39994993805885315,
+      "learning_rate": 7.187357297712506e-05,
+      "loss": 1.6681,
+      "step": 3426
+    },
+    {
+      "epoch": 0.355977978601849,
+      "grad_norm": 0.38335686922073364,
+      "learning_rate": 7.185889942475851e-05,
+      "loss": 1.7488,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3560818531214293,
+      "grad_norm": 0.37399354577064514,
+      "learning_rate": 7.184422354458966e-05,
+      "loss": 1.6041,
+      "step": 3428
+    },
+    {
+      "epoch": 0.35618572764100964,
+      "grad_norm": 0.3858386278152466,
+      "learning_rate": 7.182954533818142e-05,
+      "loss": 1.7968,
+      "step": 3429
+    },
+    {
+      "epoch": 0.35628960216059,
+      "grad_norm": 0.39342349767684937,
+      "learning_rate": 7.181486480709687e-05,
+      "loss": 1.7446,
+      "step": 3430
+    },
+    {
+      "epoch": 0.35639347668017035,
+      "grad_norm": 0.3939875364303589,
+      "learning_rate": 7.180018195289937e-05,
+      "loss": 1.7684,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3564973511997507,
+      "grad_norm": 0.3789867162704468,
+      "learning_rate": 7.178549677715254e-05,
+      "loss": 1.7591,
+      "step": 3432
+    },
+    {
+      "epoch": 0.35660122571933106,
+      "grad_norm": 0.41745778918266296,
+      "learning_rate": 7.177080928142024e-05,
+      "loss": 1.8228,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3567051002389114,
+      "grad_norm": 0.3856961727142334,
+      "learning_rate": 7.175611946726657e-05,
+      "loss": 1.7026,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3568089747584917,
+      "grad_norm": 0.36199331283569336,
+      "learning_rate": 7.174142733625589e-05,
+      "loss": 1.4263,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3569128492780721,
+      "grad_norm": 0.38251036405563354,
+      "learning_rate": 7.172673288995277e-05,
+      "loss": 1.6076,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3570167237976524,
+      "grad_norm": 0.3627626895904541,
+      "learning_rate": 7.171203612992209e-05,
+      "loss": 1.7049,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3571205983172328,
+      "grad_norm": 0.37557271122932434,
+      "learning_rate": 7.169733705772891e-05,
+      "loss": 1.5924,
+      "step": 3438
+    },
+    {
+      "epoch": 0.35722447283681313,
+      "grad_norm": 0.4099326729774475,
+      "learning_rate": 7.168263567493859e-05,
+      "loss": 1.8338,
+      "step": 3439
+    },
+    {
+      "epoch": 0.35732834735639346,
+      "grad_norm": 0.3865790367126465,
+      "learning_rate": 7.16679319831167e-05,
+      "loss": 1.8652,
+      "step": 3440
+    },
+    {
+      "epoch": 0.35743222187597384,
+      "grad_norm": 0.3986462950706482,
+      "learning_rate": 7.165322598382905e-05,
+      "loss": 1.8941,
+      "step": 3441
+    },
+    {
+      "epoch": 0.35753609639555417,
+      "grad_norm": 0.3964523673057556,
+      "learning_rate": 7.163851767864176e-05,
+      "loss": 1.8183,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3576399709151345,
+      "grad_norm": 0.4160378575325012,
+      "learning_rate": 7.162380706912112e-05,
+      "loss": 1.7119,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3577438454347149,
+      "grad_norm": 0.3585418462753296,
+      "learning_rate": 7.16090941568337e-05,
+      "loss": 1.5949,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3578477199542952,
+      "grad_norm": 0.3526414930820465,
+      "learning_rate": 7.159437894334629e-05,
+      "loss": 1.5768,
+      "step": 3445
+    },
+    {
+      "epoch": 0.35795159447387553,
+      "grad_norm": 0.4531826674938202,
+      "learning_rate": 7.157966143022599e-05,
+      "loss": 1.7067,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3580554689934559,
+      "grad_norm": 0.3793008625507355,
+      "learning_rate": 7.156494161904005e-05,
+      "loss": 1.6634,
+      "step": 3447
+    },
+    {
+      "epoch": 0.35815934351303624,
+      "grad_norm": 0.3665551245212555,
+      "learning_rate": 7.155021951135605e-05,
+      "loss": 1.6025,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3582632180326166,
+      "grad_norm": 0.4132823646068573,
+      "learning_rate": 7.153549510874178e-05,
+      "loss": 1.709,
+      "step": 3449
+    },
+    {
+      "epoch": 0.35836709255219695,
+      "grad_norm": 0.38809409737586975,
+      "learning_rate": 7.152076841276527e-05,
+      "loss": 1.6638,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3584709670717773,
+      "grad_norm": 0.388092964887619,
+      "learning_rate": 7.150603942499478e-05,
+      "loss": 1.7261,
+      "step": 3451
+    },
+    {
+      "epoch": 0.35857484159135766,
+      "grad_norm": 0.4031296372413635,
+      "learning_rate": 7.149130814699884e-05,
+      "loss": 1.7373,
+      "step": 3452
+    },
+    {
+      "epoch": 0.358678716110938,
+      "grad_norm": 0.36466771364212036,
+      "learning_rate": 7.147657458034622e-05,
+      "loss": 1.6811,
+      "step": 3453
+    },
+    {
+      "epoch": 0.3587825906305183,
+      "grad_norm": 0.3705892562866211,
+      "learning_rate": 7.146183872660595e-05,
+      "loss": 1.6032,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3588864651500987,
+      "grad_norm": 0.3856612741947174,
+      "learning_rate": 7.144710058734724e-05,
+      "loss": 1.5488,
+      "step": 3455
+    },
+    {
+      "epoch": 0.358990339669679,
+      "grad_norm": 0.3789518177509308,
+      "learning_rate": 7.143236016413963e-05,
+      "loss": 1.6905,
+      "step": 3456
+    },
+    {
+      "epoch": 0.35909421418925935,
+      "grad_norm": 0.3953239321708679,
+      "learning_rate": 7.141761745855284e-05,
+      "loss": 1.7514,
+      "step": 3457
+    },
+    {
+      "epoch": 0.35919808870883974,
+      "grad_norm": 0.3837435841560364,
+      "learning_rate": 7.140287247215685e-05,
+      "loss": 1.3697,
+      "step": 3458
+    },
+    {
+      "epoch": 0.35930196322842006,
+      "grad_norm": 0.3627301752567291,
+      "learning_rate": 7.13881252065219e-05,
+      "loss": 1.6055,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3594058377480004,
+      "grad_norm": 0.41467034816741943,
+      "learning_rate": 7.137337566321844e-05,
+      "loss": 1.7294,
+      "step": 3460
+    },
+    {
+      "epoch": 0.35950971226758077,
+      "grad_norm": 0.4232648015022278,
+      "learning_rate": 7.135862384381717e-05,
+      "loss": 1.8147,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3596135867871611,
+      "grad_norm": 0.43227145075798035,
+      "learning_rate": 7.134386974988909e-05,
+      "loss": 1.5631,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3597174613067415,
+      "grad_norm": 0.4435618817806244,
+      "learning_rate": 7.132911338300537e-05,
+      "loss": 1.8052,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3598213358263218,
+      "grad_norm": 0.41515088081359863,
+      "learning_rate": 7.131435474473744e-05,
+      "loss": 1.8405,
+      "step": 3464
+    },
+    {
+      "epoch": 0.35992521034590214,
+      "grad_norm": 0.3956370949745178,
+      "learning_rate": 7.1299593836657e-05,
+      "loss": 1.6957,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3600290848654825,
+      "grad_norm": 0.4101237654685974,
+      "learning_rate": 7.128483066033594e-05,
+      "loss": 1.6149,
+      "step": 3466
+    },
+    {
+      "epoch": 0.36013295938506285,
+      "grad_norm": 0.39739376306533813,
+      "learning_rate": 7.127006521734646e-05,
+      "loss": 1.7836,
+      "step": 3467
+    },
+    {
+      "epoch": 0.36023683390464317,
+      "grad_norm": 0.4086010158061981,
+      "learning_rate": 7.125529750926094e-05,
+      "loss": 1.7712,
+      "step": 3468
+    },
+    {
+      "epoch": 0.36034070842422355,
+      "grad_norm": 0.3838006556034088,
+      "learning_rate": 7.124052753765202e-05,
+      "loss": 1.7435,
+      "step": 3469
+    },
+    {
+      "epoch": 0.3604445829438039,
+      "grad_norm": 0.40963372588157654,
+      "learning_rate": 7.122575530409262e-05,
+      "loss": 1.4758,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3605484574633842,
+      "grad_norm": 0.4056002199649811,
+      "learning_rate": 7.121098081015586e-05,
+      "loss": 1.6,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3606523319829646,
+      "grad_norm": 0.46054428815841675,
+      "learning_rate": 7.119620405741506e-05,
+      "loss": 1.845,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3607562065025449,
+      "grad_norm": 0.41523477435112,
+      "learning_rate": 7.118142504744389e-05,
+      "loss": 1.7532,
+      "step": 3473
+    },
+    {
+      "epoch": 0.36086008102212525,
+      "grad_norm": 0.40955179929733276,
+      "learning_rate": 7.116664378181616e-05,
+      "loss": 1.7787,
+      "step": 3474
+    },
+    {
+      "epoch": 0.36096395554170563,
+      "grad_norm": 0.39328327775001526,
+      "learning_rate": 7.115186026210597e-05,
+      "loss": 1.7785,
+      "step": 3475
+    },
+    {
+      "epoch": 0.36106783006128595,
+      "grad_norm": 0.380825012922287,
+      "learning_rate": 7.113707448988767e-05,
+      "loss": 1.6208,
+      "step": 3476
+    },
+    {
+      "epoch": 0.36117170458086634,
+      "grad_norm": 0.3730035722255707,
+      "learning_rate": 7.11222864667358e-05,
+      "loss": 1.66,
+      "step": 3477
+    },
+    {
+      "epoch": 0.36127557910044666,
+      "grad_norm": 0.39808517694473267,
+      "learning_rate": 7.110749619422519e-05,
+      "loss": 1.6975,
+      "step": 3478
+    },
+    {
+      "epoch": 0.361379453620027,
+      "grad_norm": 0.4094315767288208,
+      "learning_rate": 7.109270367393088e-05,
+      "loss": 1.7015,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3614833281396074,
+      "grad_norm": 0.41740068793296814,
+      "learning_rate": 7.107790890742815e-05,
+      "loss": 1.7778,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3615872026591877,
+      "grad_norm": 0.41227155923843384,
+      "learning_rate": 7.106311189629253e-05,
+      "loss": 1.8917,
+      "step": 3481
+    },
+    {
+      "epoch": 0.36169107717876803,
+      "grad_norm": 0.3613469898700714,
+      "learning_rate": 7.104831264209982e-05,
+      "loss": 1.6972,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3617949516983484,
+      "grad_norm": 0.402305543422699,
+      "learning_rate": 7.103351114642596e-05,
+      "loss": 1.5879,
+      "step": 3483
+    },
+    {
+      "epoch": 0.36189882621792874,
+      "grad_norm": 0.39771807193756104,
+      "learning_rate": 7.101870741084726e-05,
+      "loss": 1.7124,
+      "step": 3484
+    },
+    {
+      "epoch": 0.36200270073750906,
+      "grad_norm": 0.36613133549690247,
+      "learning_rate": 7.100390143694016e-05,
+      "loss": 1.681,
+      "step": 3485
+    },
+    {
+      "epoch": 0.36210657525708945,
+      "grad_norm": 0.3933365046977997,
+      "learning_rate": 7.098909322628138e-05,
+      "loss": 1.703,
+      "step": 3486
+    },
+    {
+      "epoch": 0.3622104497766698,
+      "grad_norm": 0.39445760846138,
+      "learning_rate": 7.097428278044792e-05,
+      "loss": 1.7333,
+      "step": 3487
+    },
+    {
+      "epoch": 0.36231432429625016,
+      "grad_norm": 0.3644881844520569,
+      "learning_rate": 7.095947010101693e-05,
+      "loss": 1.713,
+      "step": 3488
+    },
+    {
+      "epoch": 0.3624181988158305,
+      "grad_norm": 0.3871423304080963,
+      "learning_rate": 7.094465518956588e-05,
+      "loss": 1.7749,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3625220733354108,
+      "grad_norm": 0.38430991768836975,
+      "learning_rate": 7.092983804767242e-05,
+      "loss": 1.6216,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3626259478549912,
+      "grad_norm": 0.3718239367008209,
+      "learning_rate": 7.091501867691446e-05,
+      "loss": 1.6914,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3627298223745715,
+      "grad_norm": 0.39211010932922363,
+      "learning_rate": 7.090019707887017e-05,
+      "loss": 1.4819,
+      "step": 3492
+    },
+    {
+      "epoch": 0.36283369689415185,
+      "grad_norm": 0.3777989149093628,
+      "learning_rate": 7.08853732551179e-05,
+      "loss": 1.6425,
+      "step": 3493
+    },
+    {
+      "epoch": 0.36293757141373223,
+      "grad_norm": 0.3853297829627991,
+      "learning_rate": 7.087054720723629e-05,
+      "loss": 1.6724,
+      "step": 3494
+    },
+    {
+      "epoch": 0.36304144593331256,
+      "grad_norm": 0.3982897400856018,
+      "learning_rate": 7.085571893680421e-05,
+      "loss": 1.7549,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3631453204528929,
+      "grad_norm": 0.38394638895988464,
+      "learning_rate": 7.084088844540074e-05,
+      "loss": 1.7959,
+      "step": 3496
+    },
+    {
+      "epoch": 0.36324919497247327,
+      "grad_norm": 0.4506365656852722,
+      "learning_rate": 7.08260557346052e-05,
+      "loss": 1.907,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3633530694920536,
+      "grad_norm": 0.43902164697647095,
+      "learning_rate": 7.081122080599721e-05,
+      "loss": 1.7573,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3634569440116339,
+      "grad_norm": 0.4061184525489807,
+      "learning_rate": 7.079638366115653e-05,
+      "loss": 1.8399,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3635608185312143,
+      "grad_norm": 0.3644791841506958,
+      "learning_rate": 7.078154430166319e-05,
+      "loss": 1.5973,
+      "step": 3500
+    },
+    {
+      "epoch": 0.36366469305079463,
+      "grad_norm": 0.44396528601646423,
+      "learning_rate": 7.07667027290975e-05,
+      "loss": 1.8829,
+      "step": 3501
+    },
+    {
+      "epoch": 0.363768567570375,
+      "grad_norm": 0.44041764736175537,
+      "learning_rate": 7.075185894503993e-05,
+      "loss": 1.9544,
+      "step": 3502
+    },
+    {
+      "epoch": 0.36387244208995534,
+      "grad_norm": 0.3821714520454407,
+      "learning_rate": 7.073701295107128e-05,
+      "loss": 1.7017,
+      "step": 3503
+    },
+    {
+      "epoch": 0.36397631660953567,
+      "grad_norm": 0.3823404014110565,
+      "learning_rate": 7.072216474877249e-05,
+      "loss": 1.5732,
+      "step": 3504
+    },
+    {
+      "epoch": 0.36408019112911605,
+      "grad_norm": 0.38946443796157837,
+      "learning_rate": 7.070731433972481e-05,
+      "loss": 1.5651,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3641840656486964,
+      "grad_norm": 0.3907792270183563,
+      "learning_rate": 7.069246172550967e-05,
+      "loss": 1.5276,
+      "step": 3506
+    },
+    {
+      "epoch": 0.3642879401682767,
+      "grad_norm": 0.374189168214798,
+      "learning_rate": 7.067760690770876e-05,
+      "loss": 1.4874,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3643918146878571,
+      "grad_norm": 0.37600815296173096,
+      "learning_rate": 7.066274988790399e-05,
+      "loss": 1.6168,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3644956892074374,
+      "grad_norm": 0.39330026507377625,
+      "learning_rate": 7.064789066767755e-05,
+      "loss": 1.8023,
+      "step": 3509
+    },
+    {
+      "epoch": 0.36459956372701774,
+      "grad_norm": 0.4047221541404724,
+      "learning_rate": 7.063302924861182e-05,
+      "loss": 1.718,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3647034382465981,
+      "grad_norm": 0.3758139908313751,
+      "learning_rate": 7.06181656322894e-05,
+      "loss": 1.6952,
+      "step": 3511
+    },
+    {
+      "epoch": 0.36480731276617845,
+      "grad_norm": 0.3800835907459259,
+      "learning_rate": 7.060329982029317e-05,
+      "loss": 1.5578,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3649111872857588,
+      "grad_norm": 0.38068559765815735,
+      "learning_rate": 7.058843181420624e-05,
+      "loss": 1.7363,
+      "step": 3513
+    },
+    {
+      "epoch": 0.36501506180533916,
+      "grad_norm": 0.40776312351226807,
+      "learning_rate": 7.05735616156119e-05,
+      "loss": 1.5816,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3651189363249195,
+      "grad_norm": 0.40851452946662903,
+      "learning_rate": 7.055868922609374e-05,
+      "loss": 1.8035,
+      "step": 3515
+    },
+    {
+      "epoch": 0.36522281084449987,
+      "grad_norm": 0.4785706698894501,
+      "learning_rate": 7.054381464723554e-05,
+      "loss": 2.132,
+      "step": 3516
+    },
+    {
+      "epoch": 0.3653266853640802,
+      "grad_norm": 0.3917141556739807,
+      "learning_rate": 7.052893788062131e-05,
+      "loss": 1.7564,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3654305598836605,
+      "grad_norm": 0.35814887285232544,
+      "learning_rate": 7.051405892783536e-05,
+      "loss": 1.6212,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3655344344032409,
+      "grad_norm": 0.37529706954956055,
+      "learning_rate": 7.049917779046213e-05,
+      "loss": 1.6613,
+      "step": 3519
+    },
+    {
+      "epoch": 0.36563830892282123,
+      "grad_norm": 0.4299023449420929,
+      "learning_rate": 7.048429447008636e-05,
+      "loss": 1.7108,
+      "step": 3520
+    },
+    {
+      "epoch": 0.36574218344240156,
+      "grad_norm": 0.39884042739868164,
+      "learning_rate": 7.046940896829302e-05,
+      "loss": 1.722,
+      "step": 3521
+    },
+    {
+      "epoch": 0.36584605796198194,
+      "grad_norm": 0.3787887394428253,
+      "learning_rate": 7.04545212866673e-05,
+      "loss": 1.7693,
+      "step": 3522
+    },
+    {
+      "epoch": 0.36594993248156227,
+      "grad_norm": 0.37919625639915466,
+      "learning_rate": 7.043963142679459e-05,
+      "loss": 1.8372,
+      "step": 3523
+    },
+    {
+      "epoch": 0.3660538070011426,
+      "grad_norm": 0.4805675446987152,
+      "learning_rate": 7.04247393902606e-05,
+      "loss": 1.9524,
+      "step": 3524
+    },
+    {
+      "epoch": 0.366157681520723,
+      "grad_norm": 0.3908388316631317,
+      "learning_rate": 7.040984517865116e-05,
+      "loss": 1.7887,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3662615560403033,
+      "grad_norm": 0.39634791016578674,
+      "learning_rate": 7.039494879355242e-05,
+      "loss": 1.7503,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3663654305598837,
+      "grad_norm": 0.3950914740562439,
+      "learning_rate": 7.038005023655072e-05,
+      "loss": 1.6939,
+      "step": 3527
+    },
+    {
+      "epoch": 0.366469305079464,
+      "grad_norm": 0.3956228792667389,
+      "learning_rate": 7.036514950923264e-05,
+      "loss": 1.7095,
+      "step": 3528
+    },
+    {
+      "epoch": 0.36657317959904434,
+      "grad_norm": 0.38164466619491577,
+      "learning_rate": 7.0350246613185e-05,
+      "loss": 1.6409,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3666770541186247,
+      "grad_norm": 0.39597025513648987,
+      "learning_rate": 7.033534154999481e-05,
+      "loss": 1.781,
+      "step": 3530
+    },
+    {
+      "epoch": 0.36678092863820505,
+      "grad_norm": 0.3930729329586029,
+      "learning_rate": 7.032043432124939e-05,
+      "loss": 1.7551,
+      "step": 3531
+    },
+    {
+      "epoch": 0.3668848031577854,
+      "grad_norm": 0.38284942507743835,
+      "learning_rate": 7.030552492853621e-05,
+      "loss": 1.4163,
+      "step": 3532
+    },
+    {
+      "epoch": 0.36698867767736576,
+      "grad_norm": 0.3863504230976105,
+      "learning_rate": 7.029061337344303e-05,
+      "loss": 1.8067,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3670925521969461,
+      "grad_norm": 0.42066165804862976,
+      "learning_rate": 7.027569965755777e-05,
+      "loss": 1.819,
+      "step": 3534
+    },
+    {
+      "epoch": 0.3671964267165264,
+      "grad_norm": 0.3733902871608734,
+      "learning_rate": 7.026078378246868e-05,
+      "loss": 1.6656,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3673003012361068,
+      "grad_norm": 0.39036303758621216,
+      "learning_rate": 7.024586574976414e-05,
+      "loss": 1.7219,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3674041757556871,
+      "grad_norm": 0.3913772702217102,
+      "learning_rate": 7.023094556103283e-05,
+      "loss": 1.6209,
+      "step": 3537
+    },
+    {
+      "epoch": 0.36750805027526745,
+      "grad_norm": 0.3987481892108917,
+      "learning_rate": 7.021602321786362e-05,
+      "loss": 1.7411,
+      "step": 3538
+    },
+    {
+      "epoch": 0.36761192479484783,
+      "grad_norm": 0.38136929273605347,
+      "learning_rate": 7.020109872184563e-05,
+      "loss": 1.6651,
+      "step": 3539
+    },
+    {
+      "epoch": 0.36771579931442816,
+      "grad_norm": 0.3978997766971588,
+      "learning_rate": 7.018617207456821e-05,
+      "loss": 1.7601,
+      "step": 3540
+    },
+    {
+      "epoch": 0.36781967383400854,
+      "grad_norm": 0.39841562509536743,
+      "learning_rate": 7.01712432776209e-05,
+      "loss": 1.6426,
+      "step": 3541
+    },
+    {
+      "epoch": 0.36792354835358887,
+      "grad_norm": 0.40489330887794495,
+      "learning_rate": 7.015631233259354e-05,
+      "loss": 1.4646,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3680274228731692,
+      "grad_norm": 0.3978332579135895,
+      "learning_rate": 7.014137924107614e-05,
+      "loss": 1.7999,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3681312973927496,
+      "grad_norm": 0.4058266878128052,
+      "learning_rate": 7.012644400465895e-05,
+      "loss": 1.8085,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3682351719123299,
+      "grad_norm": 0.4187474846839905,
+      "learning_rate": 7.011150662493248e-05,
+      "loss": 1.851,
+      "step": 3545
+    },
+    {
+      "epoch": 0.36833904643191023,
+      "grad_norm": 0.43871933221817017,
+      "learning_rate": 7.009656710348742e-05,
+      "loss": 2.0409,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3684429209514906,
+      "grad_norm": 0.4018607437610626,
+      "learning_rate": 7.008162544191474e-05,
+      "loss": 1.7269,
+      "step": 3547
+    },
+    {
+      "epoch": 0.36854679547107094,
+      "grad_norm": 0.4518108665943146,
+      "learning_rate": 7.006668164180557e-05,
+      "loss": 1.8257,
+      "step": 3548
+    },
+    {
+      "epoch": 0.36865066999065127,
+      "grad_norm": 0.4045441150665283,
+      "learning_rate": 7.005173570475135e-05,
+      "loss": 1.8499,
+      "step": 3549
+    },
+    {
+      "epoch": 0.36875454451023165,
+      "grad_norm": 0.41423842310905457,
+      "learning_rate": 7.003678763234367e-05,
+      "loss": 1.8105,
+      "step": 3550
+    },
+    {
+      "epoch": 0.368858419029812,
+      "grad_norm": 0.37954044342041016,
+      "learning_rate": 7.002183742617441e-05,
+      "loss": 1.7043,
+      "step": 3551
+    },
+    {
+      "epoch": 0.36896229354939236,
+      "grad_norm": 0.39509761333465576,
+      "learning_rate": 7.000688508783564e-05,
+      "loss": 1.7842,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3690661680689727,
+      "grad_norm": 0.4210776686668396,
+      "learning_rate": 6.999193061891967e-05,
+      "loss": 1.5356,
+      "step": 3553
+    },
+    {
+      "epoch": 0.369170042588553,
+      "grad_norm": 0.4176965355873108,
+      "learning_rate": 6.997697402101904e-05,
+      "loss": 1.8425,
+      "step": 3554
+    },
+    {
+      "epoch": 0.3692739171081334,
+      "grad_norm": 0.4520731270313263,
+      "learning_rate": 6.996201529572648e-05,
+      "loss": 1.7272,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3693777916277137,
+      "grad_norm": 0.3699086606502533,
+      "learning_rate": 6.994705444463501e-05,
+      "loss": 1.661,
+      "step": 3556
+    },
+    {
+      "epoch": 0.36948166614729405,
+      "grad_norm": 0.38104352355003357,
+      "learning_rate": 6.993209146933784e-05,
+      "loss": 1.5626,
+      "step": 3557
+    },
+    {
+      "epoch": 0.36958554066687443,
+      "grad_norm": 0.4503292739391327,
+      "learning_rate": 6.991712637142839e-05,
+      "loss": 1.8327,
+      "step": 3558
+    },
+    {
+      "epoch": 0.36968941518645476,
+      "grad_norm": 0.43513697385787964,
+      "learning_rate": 6.990215915250035e-05,
+      "loss": 1.6256,
+      "step": 3559
+    },
+    {
+      "epoch": 0.3697932897060351,
+      "grad_norm": 0.4127349555492401,
+      "learning_rate": 6.988718981414761e-05,
+      "loss": 1.8755,
+      "step": 3560
+    },
+    {
+      "epoch": 0.36989716422561547,
+      "grad_norm": 0.4628101885318756,
+      "learning_rate": 6.987221835796429e-05,
+      "loss": 1.8136,
+      "step": 3561
+    },
+    {
+      "epoch": 0.3700010387451958,
+      "grad_norm": 0.4025733768939972,
+      "learning_rate": 6.98572447855447e-05,
+      "loss": 1.66,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3701049132647761,
+      "grad_norm": 0.3781670331954956,
+      "learning_rate": 6.984226909848346e-05,
+      "loss": 1.6159,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3702087877843565,
+      "grad_norm": 0.40306591987609863,
+      "learning_rate": 6.982729129837531e-05,
+      "loss": 1.888,
+      "step": 3564
+    },
+    {
+      "epoch": 0.37031266230393683,
+      "grad_norm": 0.4283866584300995,
+      "learning_rate": 6.981231138681531e-05,
+      "loss": 1.7992,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3704165368235172,
+      "grad_norm": 0.4207936227321625,
+      "learning_rate": 6.979732936539868e-05,
+      "loss": 1.7709,
+      "step": 3566
+    },
+    {
+      "epoch": 0.37052041134309754,
+      "grad_norm": 0.4613027274608612,
+      "learning_rate": 6.97823452357209e-05,
+      "loss": 1.765,
+      "step": 3567
+    },
+    {
+      "epoch": 0.37062428586267787,
+      "grad_norm": 0.40821146965026855,
+      "learning_rate": 6.976735899937767e-05,
+      "loss": 1.7137,
+      "step": 3568
+    },
+    {
+      "epoch": 0.37072816038225825,
+      "grad_norm": 0.3744267225265503,
+      "learning_rate": 6.97523706579649e-05,
+      "loss": 1.7768,
+      "step": 3569
+    },
+    {
+      "epoch": 0.3708320349018386,
+      "grad_norm": 0.410510778427124,
+      "learning_rate": 6.973738021307871e-05,
+      "loss": 1.4456,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3709359094214189,
+      "grad_norm": 0.4135860800743103,
+      "learning_rate": 6.97223876663155e-05,
+      "loss": 1.6854,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3710397839409993,
+      "grad_norm": 0.43364089727401733,
+      "learning_rate": 6.970739301927183e-05,
+      "loss": 1.7468,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3711436584605796,
+      "grad_norm": 0.37639757990837097,
+      "learning_rate": 6.969239627354453e-05,
+      "loss": 1.6043,
+      "step": 3573
+    },
+    {
+      "epoch": 0.37124753298015994,
+      "grad_norm": 0.415420800447464,
+      "learning_rate": 6.967739743073065e-05,
+      "loss": 1.7144,
+      "step": 3574
+    },
+    {
+      "epoch": 0.3713514074997403,
+      "grad_norm": 0.39380958676338196,
+      "learning_rate": 6.966239649242743e-05,
+      "loss": 1.8287,
+      "step": 3575
+    },
+    {
+      "epoch": 0.37145528201932065,
+      "grad_norm": 0.41520291566848755,
+      "learning_rate": 6.964739346023235e-05,
+      "loss": 1.8377,
+      "step": 3576
+    },
+    {
+      "epoch": 0.371559156538901,
+      "grad_norm": 0.4204116761684418,
+      "learning_rate": 6.963238833574312e-05,
+      "loss": 1.7471,
+      "step": 3577
+    },
+    {
+      "epoch": 0.37166303105848136,
+      "grad_norm": 0.4574458599090576,
+      "learning_rate": 6.961738112055767e-05,
+      "loss": 1.4237,
+      "step": 3578
+    },
+    {
+      "epoch": 0.3717669055780617,
+      "grad_norm": 0.3843066990375519,
+      "learning_rate": 6.960237181627416e-05,
+      "loss": 1.7035,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3718707800976421,
+      "grad_norm": 0.3904613256454468,
+      "learning_rate": 6.958736042449095e-05,
+      "loss": 1.8277,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3719746546172224,
+      "grad_norm": 0.3790266811847687,
+      "learning_rate": 6.957234694680663e-05,
+      "loss": 1.7919,
+      "step": 3581
+    },
+    {
+      "epoch": 0.3720785291368027,
+      "grad_norm": 0.3974626362323761,
+      "learning_rate": 6.955733138482006e-05,
+      "loss": 1.795,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3721824036563831,
+      "grad_norm": 0.3986896574497223,
+      "learning_rate": 6.954231374013022e-05,
+      "loss": 1.7831,
+      "step": 3583
+    },
+    {
+      "epoch": 0.37228627817596344,
+      "grad_norm": 0.4144607484340668,
+      "learning_rate": 6.952729401433642e-05,
+      "loss": 1.4946,
+      "step": 3584
+    },
+    {
+      "epoch": 0.37239015269554376,
+      "grad_norm": 0.4263852536678314,
+      "learning_rate": 6.951227220903813e-05,
+      "loss": 1.7404,
+      "step": 3585
+    },
+    {
+      "epoch": 0.37249402721512415,
+      "grad_norm": 0.4002395272254944,
+      "learning_rate": 6.949724832583504e-05,
+      "loss": 1.6906,
+      "step": 3586
+    },
+    {
+      "epoch": 0.3725979017347045,
+      "grad_norm": 0.41054767370224,
+      "learning_rate": 6.948222236632709e-05,
+      "loss": 1.7632,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3727017762542848,
+      "grad_norm": 0.3650989234447479,
+      "learning_rate": 6.946719433211442e-05,
+      "loss": 1.6058,
+      "step": 3588
+    },
+    {
+      "epoch": 0.3728056507738652,
+      "grad_norm": 0.4016493856906891,
+      "learning_rate": 6.945216422479741e-05,
+      "loss": 1.7819,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3729095252934455,
+      "grad_norm": 0.3928471505641937,
+      "learning_rate": 6.943713204597664e-05,
+      "loss": 1.8151,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3730133998130259,
+      "grad_norm": 0.39063897728919983,
+      "learning_rate": 6.942209779725292e-05,
+      "loss": 1.775,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3731172743326062,
+      "grad_norm": 0.38129961490631104,
+      "learning_rate": 6.94070614802273e-05,
+      "loss": 1.6656,
+      "step": 3592
+    },
+    {
+      "epoch": 0.37322114885218655,
+      "grad_norm": 0.438495934009552,
+      "learning_rate": 6.9392023096501e-05,
+      "loss": 1.7774,
+      "step": 3593
+    },
+    {
+      "epoch": 0.37332502337176693,
+      "grad_norm": 0.43866604566574097,
+      "learning_rate": 6.937698264767549e-05,
+      "loss": 1.8816,
+      "step": 3594
+    },
+    {
+      "epoch": 0.37342889789134726,
+      "grad_norm": 0.41461604833602905,
+      "learning_rate": 6.93619401353525e-05,
+      "loss": 1.7163,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3735327724109276,
+      "grad_norm": 0.36867401003837585,
+      "learning_rate": 6.934689556113391e-05,
+      "loss": 1.7318,
+      "step": 3596
+    },
+    {
+      "epoch": 0.37363664693050797,
+      "grad_norm": 0.41185227036476135,
+      "learning_rate": 6.933184892662185e-05,
+      "loss": 1.8164,
+      "step": 3597
+    },
+    {
+      "epoch": 0.3737405214500883,
+      "grad_norm": 0.3863617777824402,
+      "learning_rate": 6.931680023341867e-05,
+      "loss": 1.7503,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3738443959696686,
+      "grad_norm": 0.3842338025569916,
+      "learning_rate": 6.930174948312696e-05,
+      "loss": 1.6838,
+      "step": 3599
+    },
+    {
+      "epoch": 0.373948270489249,
+      "grad_norm": 0.3778021037578583,
+      "learning_rate": 6.928669667734945e-05,
+      "loss": 1.743,
+      "step": 3600
+    },
+    {
+      "epoch": 0.37405214500882933,
+      "grad_norm": 0.40780338644981384,
+      "learning_rate": 6.927164181768921e-05,
+      "loss": 1.868,
+      "step": 3601
+    },
+    {
+      "epoch": 0.37415601952840966,
+      "grad_norm": 0.4100748300552368,
+      "learning_rate": 6.925658490574945e-05,
+      "loss": 1.8172,
+      "step": 3602
+    },
+    {
+      "epoch": 0.37425989404799004,
+      "grad_norm": 0.3898886740207672,
+      "learning_rate": 6.92415259431336e-05,
+      "loss": 1.671,
+      "step": 3603
+    },
+    {
+      "epoch": 0.37436376856757037,
+      "grad_norm": 0.4094226062297821,
+      "learning_rate": 6.922646493144532e-05,
+      "loss": 1.6493,
+      "step": 3604
+    },
+    {
+      "epoch": 0.37446764308715075,
+      "grad_norm": 0.36988019943237305,
+      "learning_rate": 6.92114018722885e-05,
+      "loss": 1.6539,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3745715176067311,
+      "grad_norm": 0.3859472870826721,
+      "learning_rate": 6.919633676726722e-05,
+      "loss": 1.5816,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3746753921263114,
+      "grad_norm": 0.4098130464553833,
+      "learning_rate": 6.918126961798583e-05,
+      "loss": 1.7291,
+      "step": 3607
+    },
+    {
+      "epoch": 0.3747792666458918,
+      "grad_norm": 0.40433475375175476,
+      "learning_rate": 6.916620042604883e-05,
+      "loss": 1.7515,
+      "step": 3608
+    },
+    {
+      "epoch": 0.3748831411654721,
+      "grad_norm": 0.38681918382644653,
+      "learning_rate": 6.915112919306101e-05,
+      "loss": 1.6408,
+      "step": 3609
+    },
+    {
+      "epoch": 0.37498701568505244,
+      "grad_norm": 0.43381497263908386,
+      "learning_rate": 6.913605592062731e-05,
+      "loss": 1.7996,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3750908902046328,
+      "grad_norm": 0.37756863236427307,
+      "learning_rate": 6.91209806103529e-05,
+      "loss": 1.6188,
+      "step": 3611
+    },
+    {
+      "epoch": 0.37519476472421315,
+      "grad_norm": 0.44089293479919434,
+      "learning_rate": 6.910590326384323e-05,
+      "loss": 1.8074,
+      "step": 3612
+    },
+    {
+      "epoch": 0.3752986392437935,
+      "grad_norm": 0.3893734812736511,
+      "learning_rate": 6.909082388270389e-05,
+      "loss": 1.6607,
+      "step": 3613
+    },
+    {
+      "epoch": 0.37540251376337386,
+      "grad_norm": 0.40464848279953003,
+      "learning_rate": 6.90757424685407e-05,
+      "loss": 1.7711,
+      "step": 3614
+    },
+    {
+      "epoch": 0.3755063882829542,
+      "grad_norm": 0.3943893313407898,
+      "learning_rate": 6.906065902295975e-05,
+      "loss": 1.8252,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3756102628025345,
+      "grad_norm": 0.38202980160713196,
+      "learning_rate": 6.904557354756728e-05,
+      "loss": 1.7832,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3757141373221149,
+      "grad_norm": 0.38370367884635925,
+      "learning_rate": 6.903048604396981e-05,
+      "loss": 1.6776,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3758180118416952,
+      "grad_norm": 0.39055919647216797,
+      "learning_rate": 6.9015396513774e-05,
+      "loss": 1.7664,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3759218863612756,
+      "grad_norm": 0.3747326731681824,
+      "learning_rate": 6.900030495858681e-05,
+      "loss": 1.5807,
+      "step": 3619
+    },
+    {
+      "epoch": 0.37602576088085593,
+      "grad_norm": 0.4175710380077362,
+      "learning_rate": 6.898521138001533e-05,
+      "loss": 1.8023,
+      "step": 3620
+    },
+    {
+      "epoch": 0.37612963540043626,
+      "grad_norm": 0.4446984529495239,
+      "learning_rate": 6.897011577966693e-05,
+      "loss": 1.7831,
+      "step": 3621
+    },
+    {
+      "epoch": 0.37623350992001664,
+      "grad_norm": 0.3818349242210388,
+      "learning_rate": 6.895501815914917e-05,
+      "loss": 1.7742,
+      "step": 3622
+    },
+    {
+      "epoch": 0.37633738443959697,
+      "grad_norm": 0.4104137122631073,
+      "learning_rate": 6.893991852006983e-05,
+      "loss": 1.6418,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3764412589591773,
+      "grad_norm": 0.3920380771160126,
+      "learning_rate": 6.892481686403694e-05,
+      "loss": 1.6307,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3765451334787577,
+      "grad_norm": 0.36929798126220703,
+      "learning_rate": 6.890971319265863e-05,
+      "loss": 1.6734,
+      "step": 3625
+    },
+    {
+      "epoch": 0.376649007998338,
+      "grad_norm": 0.39406248927116394,
+      "learning_rate": 6.889460750754337e-05,
+      "loss": 1.6584,
+      "step": 3626
+    },
+    {
+      "epoch": 0.37675288251791833,
+      "grad_norm": 0.4494403898715973,
+      "learning_rate": 6.887949981029981e-05,
+      "loss": 1.9331,
+      "step": 3627
+    },
+    {
+      "epoch": 0.3768567570374987,
+      "grad_norm": 0.4188593924045563,
+      "learning_rate": 6.886439010253678e-05,
+      "loss": 1.8305,
+      "step": 3628
+    },
+    {
+      "epoch": 0.37696063155707904,
+      "grad_norm": 0.3847233057022095,
+      "learning_rate": 6.884927838586336e-05,
+      "loss": 1.675,
+      "step": 3629
+    },
+    {
+      "epoch": 0.3770645060766594,
+      "grad_norm": 0.3903995156288147,
+      "learning_rate": 6.88341646618888e-05,
+      "loss": 1.6902,
+      "step": 3630
+    },
+    {
+      "epoch": 0.37716838059623975,
+      "grad_norm": 0.38415199518203735,
+      "learning_rate": 6.881904893222265e-05,
+      "loss": 1.593,
+      "step": 3631
+    },
+    {
+      "epoch": 0.3772722551158201,
+      "grad_norm": 0.3710866868495941,
+      "learning_rate": 6.880393119847456e-05,
+      "loss": 1.5337,
+      "step": 3632
+    },
+    {
+      "epoch": 0.37737612963540046,
+      "grad_norm": 0.3745911419391632,
+      "learning_rate": 6.878881146225447e-05,
+      "loss": 1.7957,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3774800041549808,
+      "grad_norm": 0.42726603150367737,
+      "learning_rate": 6.877368972517252e-05,
+      "loss": 1.783,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3775838786745611,
+      "grad_norm": 0.36636191606521606,
+      "learning_rate": 6.875856598883907e-05,
+      "loss": 1.756,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3776877531941415,
+      "grad_norm": 0.40645745396614075,
+      "learning_rate": 6.874344025486464e-05,
+      "loss": 1.7357,
+      "step": 3636
+    },
+    {
+      "epoch": 0.3777916277137218,
+      "grad_norm": 0.42991214990615845,
+      "learning_rate": 6.872831252486005e-05,
+      "loss": 1.9177,
+      "step": 3637
+    },
+    {
+      "epoch": 0.37789550223330215,
+      "grad_norm": 0.39431822299957275,
+      "learning_rate": 6.871318280043626e-05,
+      "loss": 1.6861,
+      "step": 3638
+    },
+    {
+      "epoch": 0.37799937675288253,
+      "grad_norm": 0.4022711515426636,
+      "learning_rate": 6.869805108320446e-05,
+      "loss": 1.7634,
+      "step": 3639
+    },
+    {
+      "epoch": 0.37810325127246286,
+      "grad_norm": 0.3846032917499542,
+      "learning_rate": 6.868291737477608e-05,
+      "loss": 1.616,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3782071257920432,
+      "grad_norm": 0.3910578787326813,
+      "learning_rate": 6.866778167676272e-05,
+      "loss": 1.6158,
+      "step": 3641
+    },
+    {
+      "epoch": 0.37831100031162357,
+      "grad_norm": 0.39738887548446655,
+      "learning_rate": 6.865264399077623e-05,
+      "loss": 1.657,
+      "step": 3642
+    },
+    {
+      "epoch": 0.3784148748312039,
+      "grad_norm": 0.3639686703681946,
+      "learning_rate": 6.863750431842865e-05,
+      "loss": 1.536,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3785187493507843,
+      "grad_norm": 0.4142943322658539,
+      "learning_rate": 6.862236266133225e-05,
+      "loss": 1.8389,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3786226238703646,
+      "grad_norm": 0.42220360040664673,
+      "learning_rate": 6.860721902109948e-05,
+      "loss": 1.6854,
+      "step": 3645
+    },
+    {
+      "epoch": 0.37872649838994493,
+      "grad_norm": 0.43140527606010437,
+      "learning_rate": 6.859207339934302e-05,
+      "loss": 1.8258,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3788303729095253,
+      "grad_norm": 0.40784284472465515,
+      "learning_rate": 6.857692579767576e-05,
+      "loss": 1.7949,
+      "step": 3647
+    },
+    {
+      "epoch": 0.37893424742910564,
+      "grad_norm": 0.39824676513671875,
+      "learning_rate": 6.856177621771083e-05,
+      "loss": 1.6723,
+      "step": 3648
+    },
+    {
+      "epoch": 0.37903812194868597,
+      "grad_norm": 0.3684784471988678,
+      "learning_rate": 6.854662466106151e-05,
+      "loss": 1.7018,
+      "step": 3649
+    },
+    {
+      "epoch": 0.37914199646826635,
+      "grad_norm": 0.4189784824848175,
+      "learning_rate": 6.853147112934132e-05,
+      "loss": 1.7914,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3792458709878467,
+      "grad_norm": 0.41028210520744324,
+      "learning_rate": 6.851631562416402e-05,
+      "loss": 1.7757,
+      "step": 3651
+    },
+    {
+      "epoch": 0.379349745507427,
+      "grad_norm": 0.37003424763679504,
+      "learning_rate": 6.850115814714355e-05,
+      "loss": 1.5974,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3794536200270074,
+      "grad_norm": 0.4214622676372528,
+      "learning_rate": 6.848599869989403e-05,
+      "loss": 1.8257,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3795574945465877,
+      "grad_norm": 0.4845244586467743,
+      "learning_rate": 6.847083728402986e-05,
+      "loss": 1.5854,
+      "step": 3654
+    },
+    {
+      "epoch": 0.37966136906616804,
+      "grad_norm": 0.4533901512622833,
+      "learning_rate": 6.845567390116558e-05,
+      "loss": 1.8799,
+      "step": 3655
+    },
+    {
+      "epoch": 0.3797652435857484,
+      "grad_norm": 0.43582406640052795,
+      "learning_rate": 6.8440508552916e-05,
+      "loss": 1.9001,
+      "step": 3656
+    },
+    {
+      "epoch": 0.37986911810532875,
+      "grad_norm": 0.3865016996860504,
+      "learning_rate": 6.84253412408961e-05,
+      "loss": 1.5654,
+      "step": 3657
+    },
+    {
+      "epoch": 0.37997299262490913,
+      "grad_norm": 0.3525942862033844,
+      "learning_rate": 6.841017196672109e-05,
+      "loss": 1.5015,
+      "step": 3658
+    },
+    {
+      "epoch": 0.38007686714448946,
+      "grad_norm": 0.3993504047393799,
+      "learning_rate": 6.839500073200636e-05,
+      "loss": 1.7024,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3801807416640698,
+      "grad_norm": 0.4344191253185272,
+      "learning_rate": 6.837982753836755e-05,
+      "loss": 1.7241,
+      "step": 3660
+    },
+    {
+      "epoch": 0.38028461618365017,
+      "grad_norm": 0.3942391574382782,
+      "learning_rate": 6.836465238742045e-05,
+      "loss": 1.6898,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3803884907032305,
+      "grad_norm": 0.4017980396747589,
+      "learning_rate": 6.834947528078114e-05,
+      "loss": 1.8323,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3804923652228108,
+      "grad_norm": 0.4126379191875458,
+      "learning_rate": 6.833429622006584e-05,
+      "loss": 1.7061,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3805962397423912,
+      "grad_norm": 0.4057486951351166,
+      "learning_rate": 6.8319115206891e-05,
+      "loss": 1.8332,
+      "step": 3664
+    },
+    {
+      "epoch": 0.38070011426197153,
+      "grad_norm": 0.4050893783569336,
+      "learning_rate": 6.83039322428733e-05,
+      "loss": 1.6706,
+      "step": 3665
+    },
+    {
+      "epoch": 0.38080398878155186,
+      "grad_norm": 0.37479984760284424,
+      "learning_rate": 6.828874732962958e-05,
+      "loss": 1.4825,
+      "step": 3666
+    },
+    {
+      "epoch": 0.38090786330113224,
+      "grad_norm": 0.39387476444244385,
+      "learning_rate": 6.827356046877693e-05,
+      "loss": 1.6615,
+      "step": 3667
+    },
+    {
+      "epoch": 0.38101173782071257,
+      "grad_norm": 0.35786718130111694,
+      "learning_rate": 6.825837166193263e-05,
+      "loss": 1.5442,
+      "step": 3668
+    },
+    {
+      "epoch": 0.38111561234029295,
+      "grad_norm": 0.3664378225803375,
+      "learning_rate": 6.824318091071416e-05,
+      "loss": 1.6057,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3812194868598733,
+      "grad_norm": 0.4093641936779022,
+      "learning_rate": 6.822798821673923e-05,
+      "loss": 1.7737,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3813233613794536,
+      "grad_norm": 0.3681464195251465,
+      "learning_rate": 6.821279358162575e-05,
+      "loss": 1.506,
+      "step": 3671
+    },
+    {
+      "epoch": 0.381427235899034,
+      "grad_norm": 0.4376184344291687,
+      "learning_rate": 6.819759700699181e-05,
+      "loss": 1.7086,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3815311104186143,
+      "grad_norm": 0.4148000180721283,
+      "learning_rate": 6.818239849445574e-05,
+      "loss": 1.8619,
+      "step": 3673
+    },
+    {
+      "epoch": 0.38163498493819464,
+      "grad_norm": 0.39664286375045776,
+      "learning_rate": 6.816719804563606e-05,
+      "loss": 1.7338,
+      "step": 3674
+    },
+    {
+      "epoch": 0.381738859457775,
+      "grad_norm": 0.37723055481910706,
+      "learning_rate": 6.815199566215149e-05,
+      "loss": 1.5691,
+      "step": 3675
+    },
+    {
+      "epoch": 0.38184273397735535,
+      "grad_norm": 0.42841121554374695,
+      "learning_rate": 6.813679134562098e-05,
+      "loss": 1.8749,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3819466084969357,
+      "grad_norm": 0.3727743625640869,
+      "learning_rate": 6.812158509766366e-05,
+      "loss": 1.6885,
+      "step": 3677
+    },
+    {
+      "epoch": 0.38205048301651606,
+      "grad_norm": 0.401883989572525,
+      "learning_rate": 6.810637691989887e-05,
+      "loss": 1.7689,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3821543575360964,
+      "grad_norm": 0.3472015857696533,
+      "learning_rate": 6.809116681394618e-05,
+      "loss": 1.6354,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3822582320556767,
+      "grad_norm": 0.3772028684616089,
+      "learning_rate": 6.807595478142536e-05,
+      "loss": 1.6749,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3823621065752571,
+      "grad_norm": 0.37408995628356934,
+      "learning_rate": 6.806074082395633e-05,
+      "loss": 1.7558,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3824659810948374,
+      "grad_norm": 0.4021287262439728,
+      "learning_rate": 6.80455249431593e-05,
+      "loss": 1.6639,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3825698556144178,
+      "grad_norm": 0.3701534867286682,
+      "learning_rate": 6.803030714065461e-05,
+      "loss": 1.6497,
+      "step": 3683
+    },
+    {
+      "epoch": 0.38267373013399814,
+      "grad_norm": 0.4467286765575409,
+      "learning_rate": 6.801508741806285e-05,
+      "loss": 1.86,
+      "step": 3684
+    },
+    {
+      "epoch": 0.38277760465357846,
+      "grad_norm": 0.40490084886550903,
+      "learning_rate": 6.799986577700481e-05,
+      "loss": 1.7865,
+      "step": 3685
+    },
+    {
+      "epoch": 0.38288147917315885,
+      "grad_norm": 0.39086583256721497,
+      "learning_rate": 6.798464221910147e-05,
+      "loss": 1.7091,
+      "step": 3686
+    },
+    {
+      "epoch": 0.38298535369273917,
+      "grad_norm": 0.4162215292453766,
+      "learning_rate": 6.796941674597402e-05,
+      "loss": 1.7886,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3830892282123195,
+      "grad_norm": 0.3935025930404663,
+      "learning_rate": 6.795418935924384e-05,
+      "loss": 1.8681,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3831931027318999,
+      "grad_norm": 0.40464386343955994,
+      "learning_rate": 6.793896006053255e-05,
+      "loss": 1.8301,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3832969772514802,
+      "grad_norm": 0.3767825961112976,
+      "learning_rate": 6.792372885146195e-05,
+      "loss": 1.6537,
+      "step": 3690
+    },
+    {
+      "epoch": 0.38340085177106054,
+      "grad_norm": 0.36534345149993896,
+      "learning_rate": 6.790849573365404e-05,
+      "loss": 1.5565,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3835047262906409,
+      "grad_norm": 0.38759198784828186,
+      "learning_rate": 6.7893260708731e-05,
+      "loss": 1.8307,
+      "step": 3692
+    },
+    {
+      "epoch": 0.38360860081022125,
+      "grad_norm": 0.38123390078544617,
+      "learning_rate": 6.78780237783153e-05,
+      "loss": 1.6495,
+      "step": 3693
+    },
+    {
+      "epoch": 0.38371247532980157,
+      "grad_norm": 0.4280903935432434,
+      "learning_rate": 6.786278494402952e-05,
+      "loss": 1.8522,
+      "step": 3694
+    },
+    {
+      "epoch": 0.38381634984938195,
+      "grad_norm": 0.42416825890541077,
+      "learning_rate": 6.784754420749645e-05,
+      "loss": 1.7361,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3839202243689623,
+      "grad_norm": 0.37206903100013733,
+      "learning_rate": 6.783230157033915e-05,
+      "loss": 1.7802,
+      "step": 3696
+    },
+    {
+      "epoch": 0.38402409888854266,
+      "grad_norm": 0.4096907079219818,
+      "learning_rate": 6.781705703418084e-05,
+      "loss": 1.7411,
+      "step": 3697
+    },
+    {
+      "epoch": 0.384127973408123,
+      "grad_norm": 0.38418519496917725,
+      "learning_rate": 6.780181060064493e-05,
+      "loss": 1.4685,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3842318479277033,
+      "grad_norm": 0.3541015088558197,
+      "learning_rate": 6.778656227135505e-05,
+      "loss": 1.4894,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3843357224472837,
+      "grad_norm": 0.3923327326774597,
+      "learning_rate": 6.777131204793502e-05,
+      "loss": 1.5972,
+      "step": 3700
+    },
+    {
+      "epoch": 0.38443959696686403,
+      "grad_norm": 0.38584020733833313,
+      "learning_rate": 6.775605993200889e-05,
+      "loss": 1.474,
+      "step": 3701
+    },
+    {
+      "epoch": 0.38454347148644435,
+      "grad_norm": 0.42586928606033325,
+      "learning_rate": 6.774080592520088e-05,
+      "loss": 1.8876,
+      "step": 3702
+    },
+    {
+      "epoch": 0.38464734600602474,
+      "grad_norm": 0.4268626570701599,
+      "learning_rate": 6.77255500291354e-05,
+      "loss": 1.8743,
+      "step": 3703
+    },
+    {
+      "epoch": 0.38475122052560506,
+      "grad_norm": 0.43447503447532654,
+      "learning_rate": 6.771029224543713e-05,
+      "loss": 1.9229,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3848550950451854,
+      "grad_norm": 0.36980974674224854,
+      "learning_rate": 6.769503257573089e-05,
+      "loss": 1.5463,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3849589695647658,
+      "grad_norm": 0.3850055932998657,
+      "learning_rate": 6.767977102164167e-05,
+      "loss": 1.724,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3850628440843461,
+      "grad_norm": 0.4217112064361572,
+      "learning_rate": 6.766450758479479e-05,
+      "loss": 1.965,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3851667186039265,
+      "grad_norm": 0.3799114525318146,
+      "learning_rate": 6.76492422668156e-05,
+      "loss": 1.6066,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3852705931235068,
+      "grad_norm": 0.39665165543556213,
+      "learning_rate": 6.76339750693298e-05,
+      "loss": 1.8164,
+      "step": 3709
+    },
+    {
+      "epoch": 0.38537446764308714,
+      "grad_norm": 0.38629379868507385,
+      "learning_rate": 6.76187059939632e-05,
+      "loss": 1.6843,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3854783421626675,
+      "grad_norm": 0.4092472493648529,
+      "learning_rate": 6.760343504234184e-05,
+      "loss": 1.5059,
+      "step": 3711
+    },
+    {
+      "epoch": 0.38558221668224785,
+      "grad_norm": 0.40238040685653687,
+      "learning_rate": 6.758816221609196e-05,
+      "loss": 1.8893,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3856860912018282,
+      "grad_norm": 0.39806345105171204,
+      "learning_rate": 6.757288751684e-05,
+      "loss": 1.7954,
+      "step": 3713
+    },
+    {
+      "epoch": 0.38578996572140856,
+      "grad_norm": 0.41928762197494507,
+      "learning_rate": 6.75576109462126e-05,
+      "loss": 1.8645,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3858938402409889,
+      "grad_norm": 0.4039596915245056,
+      "learning_rate": 6.75423325058366e-05,
+      "loss": 1.8103,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3859977147605692,
+      "grad_norm": 0.400309681892395,
+      "learning_rate": 6.7527052197339e-05,
+      "loss": 1.6896,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3861015892801496,
+      "grad_norm": 0.43688875436782837,
+      "learning_rate": 6.751177002234706e-05,
+      "loss": 1.9697,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3862054637997299,
+      "grad_norm": 0.5165666937828064,
+      "learning_rate": 6.749648598248821e-05,
+      "loss": 1.9304,
+      "step": 3718
+    },
+    {
+      "epoch": 0.38630933831931025,
+      "grad_norm": 0.4006554186344147,
+      "learning_rate": 6.74812000793901e-05,
+      "loss": 1.6306,
+      "step": 3719
+    },
+    {
+      "epoch": 0.38641321283889063,
+      "grad_norm": 0.3924461007118225,
+      "learning_rate": 6.74659123146805e-05,
+      "loss": 1.6424,
+      "step": 3720
+    },
+    {
+      "epoch": 0.38651708735847096,
+      "grad_norm": 0.35347139835357666,
+      "learning_rate": 6.745062268998752e-05,
+      "loss": 1.4823,
+      "step": 3721
+    },
+    {
+      "epoch": 0.38662096187805134,
+      "grad_norm": 0.41199618577957153,
+      "learning_rate": 6.743533120693932e-05,
+      "loss": 1.6857,
+      "step": 3722
+    },
+    {
+      "epoch": 0.38672483639763167,
+      "grad_norm": 0.40314194560050964,
+      "learning_rate": 6.742003786716435e-05,
+      "loss": 1.7333,
+      "step": 3723
+    },
+    {
+      "epoch": 0.386828710917212,
+      "grad_norm": 0.402383416891098,
+      "learning_rate": 6.740474267229123e-05,
+      "loss": 1.7304,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3869325854367924,
+      "grad_norm": 0.4002920687198639,
+      "learning_rate": 6.738944562394878e-05,
+      "loss": 1.6984,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3870364599563727,
+      "grad_norm": 0.3783153295516968,
+      "learning_rate": 6.737414672376601e-05,
+      "loss": 1.6048,
+      "step": 3726
+    },
+    {
+      "epoch": 0.38714033447595303,
+      "grad_norm": 0.4129243791103363,
+      "learning_rate": 6.735884597337214e-05,
+      "loss": 1.8856,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3872442089955334,
+      "grad_norm": 0.36337247490882874,
+      "learning_rate": 6.73435433743966e-05,
+      "loss": 1.6613,
+      "step": 3728
+    },
+    {
+      "epoch": 0.38734808351511374,
+      "grad_norm": 0.38264912366867065,
+      "learning_rate": 6.732823892846895e-05,
+      "loss": 1.6385,
+      "step": 3729
+    },
+    {
+      "epoch": 0.38745195803469407,
+      "grad_norm": 0.394731342792511,
+      "learning_rate": 6.731293263721902e-05,
+      "loss": 1.6883,
+      "step": 3730
+    },
+    {
+      "epoch": 0.38755583255427445,
+      "grad_norm": 0.3739268183708191,
+      "learning_rate": 6.72976245022768e-05,
+      "loss": 1.677,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3876597070738548,
+      "grad_norm": 0.37496718764305115,
+      "learning_rate": 6.728231452527251e-05,
+      "loss": 1.8645,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3877635815934351,
+      "grad_norm": 0.39028218388557434,
+      "learning_rate": 6.726700270783655e-05,
+      "loss": 1.7683,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3878674561130155,
+      "grad_norm": 0.4083595871925354,
+      "learning_rate": 6.725168905159945e-05,
+      "loss": 1.821,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3879713306325958,
+      "grad_norm": 0.3999233543872833,
+      "learning_rate": 6.723637355819206e-05,
+      "loss": 1.7861,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3880752051521762,
+      "grad_norm": 0.3761618435382843,
+      "learning_rate": 6.722105622924533e-05,
+      "loss": 1.7543,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3881790796717565,
+      "grad_norm": 0.4310716390609741,
+      "learning_rate": 6.720573706639044e-05,
+      "loss": 1.7804,
+      "step": 3737
+    },
+    {
+      "epoch": 0.38828295419133685,
+      "grad_norm": 0.3662225902080536,
+      "learning_rate": 6.719041607125877e-05,
+      "loss": 1.5168,
+      "step": 3738
+    },
+    {
+      "epoch": 0.38838682871091723,
+      "grad_norm": 0.3875160813331604,
+      "learning_rate": 6.717509324548187e-05,
+      "loss": 1.801,
+      "step": 3739
+    },
+    {
+      "epoch": 0.38849070323049756,
+      "grad_norm": 0.379233181476593,
+      "learning_rate": 6.715976859069151e-05,
+      "loss": 1.6002,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3885945777500779,
+      "grad_norm": 0.3968057632446289,
+      "learning_rate": 6.714444210851964e-05,
+      "loss": 1.5895,
+      "step": 3741
+    },
+    {
+      "epoch": 0.38869845226965827,
+      "grad_norm": 0.45486825704574585,
+      "learning_rate": 6.712911380059845e-05,
+      "loss": 1.8872,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3888023267892386,
+      "grad_norm": 0.38296419382095337,
+      "learning_rate": 6.711378366856023e-05,
+      "loss": 1.7437,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3889062013088189,
+      "grad_norm": 0.394843190908432,
+      "learning_rate": 6.709845171403754e-05,
+      "loss": 1.7002,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3890100758283993,
+      "grad_norm": 0.4002247154712677,
+      "learning_rate": 6.70831179386631e-05,
+      "loss": 1.7328,
+      "step": 3745
+    },
+    {
+      "epoch": 0.38911395034797963,
+      "grad_norm": 0.45309674739837646,
+      "learning_rate": 6.706778234406987e-05,
+      "loss": 1.7358,
+      "step": 3746
+    },
+    {
+      "epoch": 0.38921782486756,
+      "grad_norm": 0.40047159790992737,
+      "learning_rate": 6.705244493189094e-05,
+      "loss": 1.7627,
+      "step": 3747
+    },
+    {
+      "epoch": 0.38932169938714034,
+      "grad_norm": 0.3902556300163269,
+      "learning_rate": 6.703710570375966e-05,
+      "loss": 1.6005,
+      "step": 3748
+    },
+    {
+      "epoch": 0.38942557390672067,
+      "grad_norm": 0.4025695025920868,
+      "learning_rate": 6.702176466130953e-05,
+      "loss": 1.6763,
+      "step": 3749
+    },
+    {
+      "epoch": 0.38952944842630105,
+      "grad_norm": 0.40774810314178467,
+      "learning_rate": 6.70064218061742e-05,
+      "loss": 1.8209,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3896333229458814,
+      "grad_norm": 0.36347082257270813,
+      "learning_rate": 6.699107713998762e-05,
+      "loss": 1.4882,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3897371974654617,
+      "grad_norm": 0.39139434695243835,
+      "learning_rate": 6.697573066438387e-05,
+      "loss": 1.713,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3898410719850421,
+      "grad_norm": 0.4328778386116028,
+      "learning_rate": 6.69603823809972e-05,
+      "loss": 1.9039,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3899449465046224,
+      "grad_norm": 0.38652610778808594,
+      "learning_rate": 6.694503229146212e-05,
+      "loss": 1.7179,
+      "step": 3754
+    },
+    {
+      "epoch": 0.39004882102420274,
+      "grad_norm": 0.36628860235214233,
+      "learning_rate": 6.692968039741328e-05,
+      "loss": 1.5286,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3901526955437831,
+      "grad_norm": 0.3640875220298767,
+      "learning_rate": 6.691432670048552e-05,
+      "loss": 1.6235,
+      "step": 3756
+    },
+    {
+      "epoch": 0.39025657006336345,
+      "grad_norm": 0.3808233439922333,
+      "learning_rate": 6.689897120231389e-05,
+      "loss": 1.6554,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3903604445829438,
+      "grad_norm": 0.4116235375404358,
+      "learning_rate": 6.688361390453368e-05,
+      "loss": 1.7059,
+      "step": 3758
+    },
+    {
+      "epoch": 0.39046431910252416,
+      "grad_norm": 0.39567288756370544,
+      "learning_rate": 6.686825480878026e-05,
+      "loss": 1.5193,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3905681936221045,
+      "grad_norm": 0.39557141065597534,
+      "learning_rate": 6.685289391668929e-05,
+      "loss": 1.7383,
+      "step": 3760
+    },
+    {
+      "epoch": 0.39067206814168487,
+      "grad_norm": 0.40352147817611694,
+      "learning_rate": 6.683753122989656e-05,
+      "loss": 1.6945,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3907759426612652,
+      "grad_norm": 0.38249486684799194,
+      "learning_rate": 6.682216675003811e-05,
+      "loss": 1.5765,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3908798171808455,
+      "grad_norm": 0.3883412480354309,
+      "learning_rate": 6.680680047875014e-05,
+      "loss": 1.7523,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3909836917004259,
+      "grad_norm": 0.38825148344039917,
+      "learning_rate": 6.679143241766898e-05,
+      "loss": 1.7129,
+      "step": 3764
+    },
+    {
+      "epoch": 0.39108756622000623,
+      "grad_norm": 0.4497055113315582,
+      "learning_rate": 6.677606256843126e-05,
+      "loss": 1.7823,
+      "step": 3765
+    },
+    {
+      "epoch": 0.39119144073958656,
+      "grad_norm": 0.39453697204589844,
+      "learning_rate": 6.676069093267374e-05,
+      "loss": 1.7184,
+      "step": 3766
+    },
+    {
+      "epoch": 0.39129531525916694,
+      "grad_norm": 0.4167402982711792,
+      "learning_rate": 6.674531751203335e-05,
+      "loss": 1.921,
+      "step": 3767
+    },
+    {
+      "epoch": 0.39139918977874727,
+      "grad_norm": 0.39788275957107544,
+      "learning_rate": 6.672994230814729e-05,
+      "loss": 1.6948,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3915030642983276,
+      "grad_norm": 0.3702066242694855,
+      "learning_rate": 6.671456532265285e-05,
+      "loss": 1.6143,
+      "step": 3769
+    },
+    {
+      "epoch": 0.391606938817908,
+      "grad_norm": 0.4112524390220642,
+      "learning_rate": 6.669918655718759e-05,
+      "loss": 1.6526,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3917108133374883,
+      "grad_norm": 0.4594886898994446,
+      "learning_rate": 6.668380601338924e-05,
+      "loss": 1.8916,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3918146878570687,
+      "grad_norm": 0.3847801089286804,
+      "learning_rate": 6.666842369289566e-05,
+      "loss": 1.803,
+      "step": 3772
+    },
+    {
+      "epoch": 0.391918562376649,
+      "grad_norm": 0.413424551486969,
+      "learning_rate": 6.665303959734499e-05,
+      "loss": 1.7646,
+      "step": 3773
+    },
+    {
+      "epoch": 0.39202243689622934,
+      "grad_norm": 0.42914092540740967,
+      "learning_rate": 6.66376537283755e-05,
+      "loss": 1.8358,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3921263114158097,
+      "grad_norm": 0.42516449093818665,
+      "learning_rate": 6.662226608762566e-05,
+      "loss": 1.7087,
+      "step": 3775
+    },
+    {
+      "epoch": 0.39223018593539005,
+      "grad_norm": 0.40384045243263245,
+      "learning_rate": 6.660687667673416e-05,
+      "loss": 1.81,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3923340604549704,
+      "grad_norm": 0.3911544978618622,
+      "learning_rate": 6.659148549733983e-05,
+      "loss": 1.6589,
+      "step": 3777
+    },
+    {
+      "epoch": 0.39243793497455076,
+      "grad_norm": 0.3848823606967926,
+      "learning_rate": 6.657609255108169e-05,
+      "loss": 1.653,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3925418094941311,
+      "grad_norm": 0.38892778754234314,
+      "learning_rate": 6.656069783959901e-05,
+      "loss": 1.7404,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3926456840137114,
+      "grad_norm": 0.4406839609146118,
+      "learning_rate": 6.654530136453118e-05,
+      "loss": 2.1317,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3927495585332918,
+      "grad_norm": 0.39131468534469604,
+      "learning_rate": 6.652990312751781e-05,
+      "loss": 1.864,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3928534330528721,
+      "grad_norm": 0.38605618476867676,
+      "learning_rate": 6.65145031301987e-05,
+      "loss": 1.7456,
+      "step": 3782
+    },
+    {
+      "epoch": 0.39295730757245245,
+      "grad_norm": 0.3929480314254761,
+      "learning_rate": 6.649910137421383e-05,
+      "loss": 1.7215,
+      "step": 3783
+    },
+    {
+      "epoch": 0.39306118209203283,
+      "grad_norm": 0.4117045998573303,
+      "learning_rate": 6.648369786120336e-05,
+      "loss": 1.8767,
+      "step": 3784
+    },
+    {
+      "epoch": 0.39316505661161316,
+      "grad_norm": 0.4199231266975403,
+      "learning_rate": 6.646829259280765e-05,
+      "loss": 1.7678,
+      "step": 3785
+    },
+    {
+      "epoch": 0.39326893113119354,
+      "grad_norm": 0.3810809254646301,
+      "learning_rate": 6.645288557066722e-05,
+      "loss": 1.6417,
+      "step": 3786
+    },
+    {
+      "epoch": 0.39337280565077387,
+      "grad_norm": 0.40853750705718994,
+      "learning_rate": 6.643747679642282e-05,
+      "loss": 1.7405,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3934766801703542,
+      "grad_norm": 0.3977396488189697,
+      "learning_rate": 6.642206627171536e-05,
+      "loss": 1.7114,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3935805546899346,
+      "grad_norm": 0.40110906958580017,
+      "learning_rate": 6.640665399818594e-05,
+      "loss": 1.433,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3936844292095149,
+      "grad_norm": 0.37747594714164734,
+      "learning_rate": 6.639123997747585e-05,
+      "loss": 1.6693,
+      "step": 3790
+    },
+    {
+      "epoch": 0.39378830372909523,
+      "grad_norm": 0.39845001697540283,
+      "learning_rate": 6.637582421122657e-05,
+      "loss": 1.7434,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3938921782486756,
+      "grad_norm": 0.3821636438369751,
+      "learning_rate": 6.636040670107972e-05,
+      "loss": 1.6637,
+      "step": 3792
+    },
+    {
+      "epoch": 0.39399605276825594,
+      "grad_norm": 0.4459479749202728,
+      "learning_rate": 6.63449874486772e-05,
+      "loss": 1.9512,
+      "step": 3793
+    },
+    {
+      "epoch": 0.39409992728783627,
+      "grad_norm": 0.35890012979507446,
+      "learning_rate": 6.632956645566099e-05,
+      "loss": 1.5369,
+      "step": 3794
+    },
+    {
+      "epoch": 0.39420380180741665,
+      "grad_norm": 0.40800046920776367,
+      "learning_rate": 6.631414372367333e-05,
+      "loss": 1.7759,
+      "step": 3795
+    },
+    {
+      "epoch": 0.394307676326997,
+      "grad_norm": 0.42685750126838684,
+      "learning_rate": 6.629871925435663e-05,
+      "loss": 1.8498,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3944115508465773,
+      "grad_norm": 0.3911482095718384,
+      "learning_rate": 6.628329304935346e-05,
+      "loss": 1.7042,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3945154253661577,
+      "grad_norm": 0.3812938332557678,
+      "learning_rate": 6.62678651103066e-05,
+      "loss": 1.6208,
+      "step": 3798
+    },
+    {
+      "epoch": 0.394619299885738,
+      "grad_norm": 0.4290582239627838,
+      "learning_rate": 6.625243543885899e-05,
+      "loss": 1.8095,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3947231744053184,
+      "grad_norm": 0.38044747710227966,
+      "learning_rate": 6.623700403665378e-05,
+      "loss": 1.821,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3948270489248987,
+      "grad_norm": 0.42807307839393616,
+      "learning_rate": 6.622157090533431e-05,
+      "loss": 1.9329,
+      "step": 3801
+    },
+    {
+      "epoch": 0.39493092344447905,
+      "grad_norm": 0.3865020275115967,
+      "learning_rate": 6.620613604654405e-05,
+      "loss": 1.641,
+      "step": 3802
+    },
+    {
+      "epoch": 0.39503479796405944,
+      "grad_norm": 0.4715527892112732,
+      "learning_rate": 6.619069946192672e-05,
+      "loss": 1.8927,
+      "step": 3803
+    },
+    {
+      "epoch": 0.39513867248363976,
+      "grad_norm": 0.3802943229675293,
+      "learning_rate": 6.61752611531262e-05,
+      "loss": 1.7063,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3952425470032201,
+      "grad_norm": 0.38618597388267517,
+      "learning_rate": 6.615982112178653e-05,
+      "loss": 1.6576,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3953464215228005,
+      "grad_norm": 0.4029250741004944,
+      "learning_rate": 6.614437936955196e-05,
+      "loss": 1.7814,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3954502960423808,
+      "grad_norm": 0.3960702121257782,
+      "learning_rate": 6.612893589806693e-05,
+      "loss": 1.7102,
+      "step": 3807
+    },
+    {
+      "epoch": 0.3955541705619611,
+      "grad_norm": 0.38714170455932617,
+      "learning_rate": 6.611349070897604e-05,
+      "loss": 1.6924,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3956580450815415,
+      "grad_norm": 0.4107656180858612,
+      "learning_rate": 6.609804380392406e-05,
+      "loss": 1.5255,
+      "step": 3809
+    },
+    {
+      "epoch": 0.39576191960112184,
+      "grad_norm": 0.40633097290992737,
+      "learning_rate": 6.6082595184556e-05,
+      "loss": 1.734,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3958657941207022,
+      "grad_norm": 0.4659360945224762,
+      "learning_rate": 6.606714485251699e-05,
+      "loss": 1.7854,
+      "step": 3811
+    },
+    {
+      "epoch": 0.39596966864028255,
+      "grad_norm": 0.4087457060813904,
+      "learning_rate": 6.605169280945238e-05,
+      "loss": 1.7713,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3960735431598629,
+      "grad_norm": 0.38341912627220154,
+      "learning_rate": 6.603623905700771e-05,
+      "loss": 1.5804,
+      "step": 3813
+    },
+    {
+      "epoch": 0.39617741767944326,
+      "grad_norm": 0.41718804836273193,
+      "learning_rate": 6.602078359682865e-05,
+      "loss": 1.7562,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3962812921990236,
+      "grad_norm": 0.4072135388851166,
+      "learning_rate": 6.600532643056112e-05,
+      "loss": 1.6313,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3963851667186039,
+      "grad_norm": 0.4200040400028229,
+      "learning_rate": 6.598986755985116e-05,
+      "loss": 1.7867,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3964890412381843,
+      "grad_norm": 0.38370397686958313,
+      "learning_rate": 6.597440698634503e-05,
+      "loss": 1.5696,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3965929157577646,
+      "grad_norm": 0.4030844271183014,
+      "learning_rate": 6.595894471168917e-05,
+      "loss": 1.688,
+      "step": 3818
+    },
+    {
+      "epoch": 0.39669679027734495,
+      "grad_norm": 0.4080040454864502,
+      "learning_rate": 6.594348073753019e-05,
+      "loss": 1.7786,
+      "step": 3819
+    },
+    {
+      "epoch": 0.39680066479692533,
+      "grad_norm": 0.36491355299949646,
+      "learning_rate": 6.592801506551486e-05,
+      "loss": 1.5053,
+      "step": 3820
+    },
+    {
+      "epoch": 0.39690453931650566,
+      "grad_norm": 0.4053652882575989,
+      "learning_rate": 6.591254769729018e-05,
+      "loss": 1.7091,
+      "step": 3821
+    },
+    {
+      "epoch": 0.397008413836086,
+      "grad_norm": 0.40529873967170715,
+      "learning_rate": 6.589707863450327e-05,
+      "loss": 1.6866,
+      "step": 3822
+    },
+    {
+      "epoch": 0.39711228835566637,
+      "grad_norm": 0.3853190243244171,
+      "learning_rate": 6.588160787880151e-05,
+      "loss": 1.703,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3972161628752467,
+      "grad_norm": 0.39480873942375183,
+      "learning_rate": 6.586613543183237e-05,
+      "loss": 1.6685,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3973200373948271,
+      "grad_norm": 0.38072872161865234,
+      "learning_rate": 6.58506612952436e-05,
+      "loss": 1.6512,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3974239119144074,
+      "grad_norm": 0.42218974232673645,
+      "learning_rate": 6.583518547068302e-05,
+      "loss": 1.7624,
+      "step": 3826
+    },
+    {
+      "epoch": 0.39752778643398773,
+      "grad_norm": 0.34545183181762695,
+      "learning_rate": 6.581970795979871e-05,
+      "loss": 1.4234,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3976316609535681,
+      "grad_norm": 0.39277178049087524,
+      "learning_rate": 6.580422876423891e-05,
+      "loss": 1.7441,
+      "step": 3828
+    },
+    {
+      "epoch": 0.39773553547314844,
+      "grad_norm": 0.39970070123672485,
+      "learning_rate": 6.578874788565201e-05,
+      "loss": 1.6835,
+      "step": 3829
+    },
+    {
+      "epoch": 0.39783940999272877,
+      "grad_norm": 0.38639700412750244,
+      "learning_rate": 6.577326532568663e-05,
+      "loss": 1.5985,
+      "step": 3830
+    },
+    {
+      "epoch": 0.39794328451230915,
+      "grad_norm": 0.3735407292842865,
+      "learning_rate": 6.575778108599151e-05,
+      "loss": 1.734,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3980471590318895,
+      "grad_norm": 0.38914212584495544,
+      "learning_rate": 6.574229516821564e-05,
+      "loss": 1.6279,
+      "step": 3832
+    },
+    {
+      "epoch": 0.3981510335514698,
+      "grad_norm": 0.38888975977897644,
+      "learning_rate": 6.572680757400813e-05,
+      "loss": 1.7271,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3982549080710502,
+      "grad_norm": 0.4465731382369995,
+      "learning_rate": 6.571131830501827e-05,
+      "loss": 1.8058,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3983587825906305,
+      "grad_norm": 0.3945710062980652,
+      "learning_rate": 6.569582736289556e-05,
+      "loss": 1.6509,
+      "step": 3835
+    },
+    {
+      "epoch": 0.39846265711021084,
+      "grad_norm": 0.39435771107673645,
+      "learning_rate": 6.56803347492897e-05,
+      "loss": 1.5513,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3985665316297912,
+      "grad_norm": 0.3789151906967163,
+      "learning_rate": 6.566484046585047e-05,
+      "loss": 1.5777,
+      "step": 3837
+    },
+    {
+      "epoch": 0.39867040614937155,
+      "grad_norm": 0.39071643352508545,
+      "learning_rate": 6.564934451422793e-05,
+      "loss": 1.8612,
+      "step": 3838
+    },
+    {
+      "epoch": 0.39877428066895193,
+      "grad_norm": 0.3789994716644287,
+      "learning_rate": 6.563384689607227e-05,
+      "loss": 1.7228,
+      "step": 3839
+    },
+    {
+      "epoch": 0.39887815518853226,
+      "grad_norm": 0.3888130486011505,
+      "learning_rate": 6.561834761303386e-05,
+      "loss": 1.7733,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3989820297081126,
+      "grad_norm": 0.41186729073524475,
+      "learning_rate": 6.560284666676327e-05,
+      "loss": 1.8526,
+      "step": 3841
+    },
+    {
+      "epoch": 0.39908590422769297,
+      "grad_norm": 0.3710212707519531,
+      "learning_rate": 6.558734405891121e-05,
+      "loss": 1.4567,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3991897787472733,
+      "grad_norm": 0.4276915490627289,
+      "learning_rate": 6.557183979112859e-05,
+      "loss": 1.8229,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3992936532668536,
+      "grad_norm": 0.4046017527580261,
+      "learning_rate": 6.55563338650665e-05,
+      "loss": 1.7953,
+      "step": 3844
+    },
+    {
+      "epoch": 0.399397527786434,
+      "grad_norm": 0.36397168040275574,
+      "learning_rate": 6.554082628237619e-05,
+      "loss": 1.4815,
+      "step": 3845
+    },
+    {
+      "epoch": 0.39950140230601433,
+      "grad_norm": 0.49856171011924744,
+      "learning_rate": 6.552531704470911e-05,
+      "loss": 1.9411,
+      "step": 3846
+    },
+    {
+      "epoch": 0.39960527682559466,
+      "grad_norm": 0.36472898721694946,
+      "learning_rate": 6.550980615371687e-05,
+      "loss": 1.6889,
+      "step": 3847
+    },
+    {
+      "epoch": 0.39970915134517504,
+      "grad_norm": 0.37358102202415466,
+      "learning_rate": 6.549429361105126e-05,
+      "loss": 1.6639,
+      "step": 3848
+    },
+    {
+      "epoch": 0.39981302586475537,
+      "grad_norm": 0.3778083920478821,
+      "learning_rate": 6.547877941836424e-05,
+      "loss": 1.7107,
+      "step": 3849
+    },
+    {
+      "epoch": 0.39991690038433575,
+      "grad_norm": 0.424234002828598,
+      "learning_rate": 6.546326357730794e-05,
+      "loss": 1.7377,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4000207749039161,
+      "grad_norm": 0.4264697730541229,
+      "learning_rate": 6.544774608953468e-05,
+      "loss": 1.8146,
+      "step": 3851
+    },
+    {
+      "epoch": 0.4001246494234964,
+      "grad_norm": 0.4023953676223755,
+      "learning_rate": 6.543222695669697e-05,
+      "loss": 1.6474,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4002285239430768,
+      "grad_norm": 0.41954219341278076,
+      "learning_rate": 6.541670618044748e-05,
+      "loss": 1.8214,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4003323984626571,
+      "grad_norm": 0.41983938217163086,
+      "learning_rate": 6.540118376243903e-05,
+      "loss": 1.7134,
+      "step": 3854
+    },
+    {
+      "epoch": 0.40043627298223744,
+      "grad_norm": 0.4455628991127014,
+      "learning_rate": 6.538565970432463e-05,
+      "loss": 1.974,
+      "step": 3855
+    },
+    {
+      "epoch": 0.4005401475018178,
+      "grad_norm": 0.41283226013183594,
+      "learning_rate": 6.537013400775748e-05,
+      "loss": 1.5895,
+      "step": 3856
+    },
+    {
+      "epoch": 0.40064402202139815,
+      "grad_norm": 0.4059913158416748,
+      "learning_rate": 6.535460667439097e-05,
+      "loss": 1.6404,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4007478965409785,
+      "grad_norm": 0.4247495234012604,
+      "learning_rate": 6.53390777058786e-05,
+      "loss": 1.8849,
+      "step": 3858
+    },
+    {
+      "epoch": 0.40085177106055886,
+      "grad_norm": 0.4334213137626648,
+      "learning_rate": 6.532354710387411e-05,
+      "loss": 1.7472,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4009556455801392,
+      "grad_norm": 0.3892272114753723,
+      "learning_rate": 6.530801487003139e-05,
+      "loss": 1.601,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4010595200997195,
+      "grad_norm": 0.3871516287326813,
+      "learning_rate": 6.529248100600449e-05,
+      "loss": 1.7275,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4011633946192999,
+      "grad_norm": 0.3974870443344116,
+      "learning_rate": 6.527694551344765e-05,
+      "loss": 1.6634,
+      "step": 3862
+    },
+    {
+      "epoch": 0.4012672691388802,
+      "grad_norm": 0.3988457918167114,
+      "learning_rate": 6.526140839401527e-05,
+      "loss": 1.6647,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4013711436584606,
+      "grad_norm": 0.42209574580192566,
+      "learning_rate": 6.524586964936194e-05,
+      "loss": 1.8112,
+      "step": 3864
+    },
+    {
+      "epoch": 0.40147501817804093,
+      "grad_norm": 0.4116060733795166,
+      "learning_rate": 6.523032928114242e-05,
+      "loss": 1.7179,
+      "step": 3865
+    },
+    {
+      "epoch": 0.40157889269762126,
+      "grad_norm": 0.341496080160141,
+      "learning_rate": 6.521478729101164e-05,
+      "loss": 1.4234,
+      "step": 3866
+    },
+    {
+      "epoch": 0.40168276721720164,
+      "grad_norm": 0.3907933831214905,
+      "learning_rate": 6.519924368062468e-05,
+      "loss": 1.7079,
+      "step": 3867
+    },
+    {
+      "epoch": 0.40178664173678197,
+      "grad_norm": 0.4084981679916382,
+      "learning_rate": 6.518369845163683e-05,
+      "loss": 1.6538,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4018905162563623,
+      "grad_norm": 0.37673652172088623,
+      "learning_rate": 6.516815160570355e-05,
+      "loss": 1.7034,
+      "step": 3869
+    },
+    {
+      "epoch": 0.4019943907759427,
+      "grad_norm": 0.36309683322906494,
+      "learning_rate": 6.515260314448041e-05,
+      "loss": 1.5712,
+      "step": 3870
+    },
+    {
+      "epoch": 0.402098265295523,
+      "grad_norm": 0.40556076169013977,
+      "learning_rate": 6.513705306962325e-05,
+      "loss": 1.6909,
+      "step": 3871
+    },
+    {
+      "epoch": 0.40220213981510333,
+      "grad_norm": 0.39010584354400635,
+      "learning_rate": 6.5121501382788e-05,
+      "loss": 1.5812,
+      "step": 3872
+    },
+    {
+      "epoch": 0.4023060143346837,
+      "grad_norm": 0.4054809510707855,
+      "learning_rate": 6.51059480856308e-05,
+      "loss": 1.6611,
+      "step": 3873
+    },
+    {
+      "epoch": 0.40240988885426404,
+      "grad_norm": 0.3722795844078064,
+      "learning_rate": 6.509039317980798e-05,
+      "loss": 1.5964,
+      "step": 3874
+    },
+    {
+      "epoch": 0.40251376337384437,
+      "grad_norm": 0.41028568148612976,
+      "learning_rate": 6.5074836666976e-05,
+      "loss": 1.6734,
+      "step": 3875
+    },
+    {
+      "epoch": 0.40261763789342475,
+      "grad_norm": 0.39521047472953796,
+      "learning_rate": 6.505927854879148e-05,
+      "loss": 1.5993,
+      "step": 3876
+    },
+    {
+      "epoch": 0.4027215124130051,
+      "grad_norm": 0.3818994462490082,
+      "learning_rate": 6.504371882691126e-05,
+      "loss": 1.6791,
+      "step": 3877
+    },
+    {
+      "epoch": 0.40282538693258546,
+      "grad_norm": 0.4162769615650177,
+      "learning_rate": 6.502815750299235e-05,
+      "loss": 1.7729,
+      "step": 3878
+    },
+    {
+      "epoch": 0.4029292614521658,
+      "grad_norm": 0.38468343019485474,
+      "learning_rate": 6.501259457869187e-05,
+      "loss": 1.7063,
+      "step": 3879
+    },
+    {
+      "epoch": 0.4030331359717461,
+      "grad_norm": 0.3765242397785187,
+      "learning_rate": 6.49970300556672e-05,
+      "loss": 1.546,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4031370104913265,
+      "grad_norm": 0.3874824345111847,
+      "learning_rate": 6.498146393557578e-05,
+      "loss": 1.7616,
+      "step": 3881
+    },
+    {
+      "epoch": 0.4032408850109068,
+      "grad_norm": 0.39522698521614075,
+      "learning_rate": 6.496589622007532e-05,
+      "loss": 1.7472,
+      "step": 3882
+    },
+    {
+      "epoch": 0.40334475953048715,
+      "grad_norm": 0.3761104941368103,
+      "learning_rate": 6.495032691082367e-05,
+      "loss": 1.5378,
+      "step": 3883
+    },
+    {
+      "epoch": 0.40344863405006753,
+      "grad_norm": 0.4124884605407715,
+      "learning_rate": 6.493475600947879e-05,
+      "loss": 1.6228,
+      "step": 3884
+    },
+    {
+      "epoch": 0.40355250856964786,
+      "grad_norm": 0.3794013559818268,
+      "learning_rate": 6.491918351769891e-05,
+      "loss": 1.8077,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4036563830892282,
+      "grad_norm": 0.43968185782432556,
+      "learning_rate": 6.490360943714235e-05,
+      "loss": 1.8393,
+      "step": 3886
+    },
+    {
+      "epoch": 0.40376025760880857,
+      "grad_norm": 0.4507423937320709,
+      "learning_rate": 6.488803376946764e-05,
+      "loss": 1.9139,
+      "step": 3887
+    },
+    {
+      "epoch": 0.4038641321283889,
+      "grad_norm": 0.39006999135017395,
+      "learning_rate": 6.48724565163335e-05,
+      "loss": 1.7191,
+      "step": 3888
+    },
+    {
+      "epoch": 0.4039680066479693,
+      "grad_norm": 0.3846725523471832,
+      "learning_rate": 6.48568776793987e-05,
+      "loss": 1.6855,
+      "step": 3889
+    },
+    {
+      "epoch": 0.4040718811675496,
+      "grad_norm": 0.4519991874694824,
+      "learning_rate": 6.484129726032233e-05,
+      "loss": 1.7529,
+      "step": 3890
+    },
+    {
+      "epoch": 0.40417575568712993,
+      "grad_norm": 0.36061856150627136,
+      "learning_rate": 6.482571526076357e-05,
+      "loss": 1.5992,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4042796302067103,
+      "grad_norm": 0.4295485317707062,
+      "learning_rate": 6.481013168238178e-05,
+      "loss": 1.7882,
+      "step": 3892
+    },
+    {
+      "epoch": 0.40438350472629064,
+      "grad_norm": 0.4181903600692749,
+      "learning_rate": 6.479454652683649e-05,
+      "loss": 1.9548,
+      "step": 3893
+    },
+    {
+      "epoch": 0.40448737924587097,
+      "grad_norm": 0.37351468205451965,
+      "learning_rate": 6.47789597957874e-05,
+      "loss": 1.5784,
+      "step": 3894
+    },
+    {
+      "epoch": 0.40459125376545135,
+      "grad_norm": 0.505387544631958,
+      "learning_rate": 6.476337149089438e-05,
+      "loss": 1.9224,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4046951282850317,
+      "grad_norm": 0.3897416293621063,
+      "learning_rate": 6.474778161381743e-05,
+      "loss": 1.4597,
+      "step": 3896
+    },
+    {
+      "epoch": 0.404799002804612,
+      "grad_norm": 0.44174665212631226,
+      "learning_rate": 6.47321901662168e-05,
+      "loss": 1.9036,
+      "step": 3897
+    },
+    {
+      "epoch": 0.4049028773241924,
+      "grad_norm": 0.39586788415908813,
+      "learning_rate": 6.471659714975282e-05,
+      "loss": 1.802,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4050067518437727,
+      "grad_norm": 0.40716326236724854,
+      "learning_rate": 6.470100256608603e-05,
+      "loss": 1.7556,
+      "step": 3899
+    },
+    {
+      "epoch": 0.40511062636335304,
+      "grad_norm": 0.3839757740497589,
+      "learning_rate": 6.468540641687716e-05,
+      "loss": 1.6419,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4052145008829334,
+      "grad_norm": 0.3808140754699707,
+      "learning_rate": 6.466980870378704e-05,
+      "loss": 1.7802,
+      "step": 3901
+    },
+    {
+      "epoch": 0.40531837540251375,
+      "grad_norm": 0.4827220141887665,
+      "learning_rate": 6.465420942847673e-05,
+      "loss": 1.9368,
+      "step": 3902
+    },
+    {
+      "epoch": 0.40542224992209414,
+      "grad_norm": 0.42109206318855286,
+      "learning_rate": 6.463860859260742e-05,
+      "loss": 1.8801,
+      "step": 3903
+    },
+    {
+      "epoch": 0.40552612444167446,
+      "grad_norm": 0.41569140553474426,
+      "learning_rate": 6.462300619784048e-05,
+      "loss": 1.8356,
+      "step": 3904
+    },
+    {
+      "epoch": 0.4056299989612548,
+      "grad_norm": 0.3784182071685791,
+      "learning_rate": 6.460740224583746e-05,
+      "loss": 1.5941,
+      "step": 3905
+    },
+    {
+      "epoch": 0.40573387348083517,
+      "grad_norm": 0.38103383779525757,
+      "learning_rate": 6.459179673826001e-05,
+      "loss": 1.8269,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4058377480004155,
+      "grad_norm": 0.38035640120506287,
+      "learning_rate": 6.457618967677004e-05,
+      "loss": 1.7104,
+      "step": 3907
+    },
+    {
+      "epoch": 0.4059416225199958,
+      "grad_norm": 0.4427489638328552,
+      "learning_rate": 6.456058106302958e-05,
+      "loss": 2.0598,
+      "step": 3908
+    },
+    {
+      "epoch": 0.4060454970395762,
+      "grad_norm": 0.38776329159736633,
+      "learning_rate": 6.454497089870082e-05,
+      "loss": 1.8005,
+      "step": 3909
+    },
+    {
+      "epoch": 0.40614937155915654,
+      "grad_norm": 0.39295095205307007,
+      "learning_rate": 6.45293591854461e-05,
+      "loss": 1.5732,
+      "step": 3910
+    },
+    {
+      "epoch": 0.40625324607873686,
+      "grad_norm": 0.4256858229637146,
+      "learning_rate": 6.451374592492796e-05,
+      "loss": 1.8357,
+      "step": 3911
+    },
+    {
+      "epoch": 0.40635712059831725,
+      "grad_norm": 0.45658764243125916,
+      "learning_rate": 6.449813111880909e-05,
+      "loss": 1.851,
+      "step": 3912
+    },
+    {
+      "epoch": 0.40646099511789757,
+      "grad_norm": 0.40819069743156433,
+      "learning_rate": 6.448251476875235e-05,
+      "loss": 1.4937,
+      "step": 3913
+    },
+    {
+      "epoch": 0.4065648696374779,
+      "grad_norm": 0.35700273513793945,
+      "learning_rate": 6.446689687642076e-05,
+      "loss": 1.6035,
+      "step": 3914
+    },
+    {
+      "epoch": 0.4066687441570583,
+      "grad_norm": 0.3629886209964752,
+      "learning_rate": 6.445127744347748e-05,
+      "loss": 1.5591,
+      "step": 3915
+    },
+    {
+      "epoch": 0.4067726186766386,
+      "grad_norm": 0.38591131567955017,
+      "learning_rate": 6.443565647158589e-05,
+      "loss": 1.5891,
+      "step": 3916
+    },
+    {
+      "epoch": 0.406876493196219,
+      "grad_norm": 0.41065141558647156,
+      "learning_rate": 6.442003396240949e-05,
+      "loss": 1.8029,
+      "step": 3917
+    },
+    {
+      "epoch": 0.4069803677157993,
+      "grad_norm": 0.4433719515800476,
+      "learning_rate": 6.440440991761195e-05,
+      "loss": 1.8338,
+      "step": 3918
+    },
+    {
+      "epoch": 0.40708424223537965,
+      "grad_norm": 0.40136539936065674,
+      "learning_rate": 6.438878433885711e-05,
+      "loss": 1.5756,
+      "step": 3919
+    },
+    {
+      "epoch": 0.40718811675496003,
+      "grad_norm": 0.39997127652168274,
+      "learning_rate": 6.437315722780898e-05,
+      "loss": 1.6578,
+      "step": 3920
+    },
+    {
+      "epoch": 0.40729199127454035,
+      "grad_norm": 0.41028451919555664,
+      "learning_rate": 6.435752858613171e-05,
+      "loss": 1.7394,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4073958657941207,
+      "grad_norm": 0.3836668133735657,
+      "learning_rate": 6.434189841548965e-05,
+      "loss": 1.7368,
+      "step": 3922
+    },
+    {
+      "epoch": 0.40749974031370106,
+      "grad_norm": 0.38762718439102173,
+      "learning_rate": 6.432626671754729e-05,
+      "loss": 1.6122,
+      "step": 3923
+    },
+    {
+      "epoch": 0.4076036148332814,
+      "grad_norm": 0.37233954668045044,
+      "learning_rate": 6.431063349396926e-05,
+      "loss": 1.5676,
+      "step": 3924
+    },
+    {
+      "epoch": 0.4077074893528617,
+      "grad_norm": 0.3751562237739563,
+      "learning_rate": 6.429499874642038e-05,
+      "loss": 1.5929,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4078113638724421,
+      "grad_norm": 0.41311609745025635,
+      "learning_rate": 6.427936247656565e-05,
+      "loss": 1.4869,
+      "step": 3926
+    },
+    {
+      "epoch": 0.40791523839202243,
+      "grad_norm": 0.37297797203063965,
+      "learning_rate": 6.42637246860702e-05,
+      "loss": 1.6425,
+      "step": 3927
+    },
+    {
+      "epoch": 0.4080191129116028,
+      "grad_norm": 0.4266083240509033,
+      "learning_rate": 6.424808537659932e-05,
+      "loss": 1.7858,
+      "step": 3928
+    },
+    {
+      "epoch": 0.40812298743118314,
+      "grad_norm": 0.46376118063926697,
+      "learning_rate": 6.423244454981851e-05,
+      "loss": 1.7072,
+      "step": 3929
+    },
+    {
+      "epoch": 0.40822686195076346,
+      "grad_norm": 0.4833356440067291,
+      "learning_rate": 6.421680220739336e-05,
+      "loss": 1.792,
+      "step": 3930
+    },
+    {
+      "epoch": 0.40833073647034385,
+      "grad_norm": 0.42818766832351685,
+      "learning_rate": 6.420115835098967e-05,
+      "loss": 1.9283,
+      "step": 3931
+    },
+    {
+      "epoch": 0.4084346109899242,
+      "grad_norm": 0.41609877347946167,
+      "learning_rate": 6.418551298227338e-05,
+      "loss": 1.6607,
+      "step": 3932
+    },
+    {
+      "epoch": 0.4085384855095045,
+      "grad_norm": 0.3854401111602783,
+      "learning_rate": 6.416986610291064e-05,
+      "loss": 1.7383,
+      "step": 3933
+    },
+    {
+      "epoch": 0.4086423600290849,
+      "grad_norm": 0.4421563744544983,
+      "learning_rate": 6.415421771456766e-05,
+      "loss": 1.5812,
+      "step": 3934
+    },
+    {
+      "epoch": 0.4087462345486652,
+      "grad_norm": 0.40960294008255005,
+      "learning_rate": 6.41385678189109e-05,
+      "loss": 1.7514,
+      "step": 3935
+    },
+    {
+      "epoch": 0.40885010906824554,
+      "grad_norm": 0.37743479013442993,
+      "learning_rate": 6.412291641760697e-05,
+      "loss": 1.6141,
+      "step": 3936
+    },
+    {
+      "epoch": 0.4089539835878259,
+      "grad_norm": 0.3768305480480194,
+      "learning_rate": 6.410726351232259e-05,
+      "loss": 1.6868,
+      "step": 3937
+    },
+    {
+      "epoch": 0.40905785810740625,
+      "grad_norm": 0.40039536356925964,
+      "learning_rate": 6.409160910472469e-05,
+      "loss": 1.7051,
+      "step": 3938
+    },
+    {
+      "epoch": 0.4091617326269866,
+      "grad_norm": 0.4146733283996582,
+      "learning_rate": 6.407595319648034e-05,
+      "loss": 1.7432,
+      "step": 3939
+    },
+    {
+      "epoch": 0.40926560714656696,
+      "grad_norm": 0.3897091746330261,
+      "learning_rate": 6.406029578925677e-05,
+      "loss": 1.6319,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4093694816661473,
+      "grad_norm": 0.3977634608745575,
+      "learning_rate": 6.404463688472138e-05,
+      "loss": 1.6741,
+      "step": 3941
+    },
+    {
+      "epoch": 0.40947335618572767,
+      "grad_norm": 0.3753795027732849,
+      "learning_rate": 6.402897648454171e-05,
+      "loss": 1.6135,
+      "step": 3942
+    },
+    {
+      "epoch": 0.409577230705308,
+      "grad_norm": 0.40746620297431946,
+      "learning_rate": 6.401331459038546e-05,
+      "loss": 1.6742,
+      "step": 3943
+    },
+    {
+      "epoch": 0.4096811052248883,
+      "grad_norm": 0.4224250614643097,
+      "learning_rate": 6.399765120392054e-05,
+      "loss": 1.7462,
+      "step": 3944
+    },
+    {
+      "epoch": 0.4097849797444687,
+      "grad_norm": 0.4034820795059204,
+      "learning_rate": 6.398198632681494e-05,
+      "loss": 1.6549,
+      "step": 3945
+    },
+    {
+      "epoch": 0.40988885426404903,
+      "grad_norm": 0.40801072120666504,
+      "learning_rate": 6.396631996073686e-05,
+      "loss": 1.5801,
+      "step": 3946
+    },
+    {
+      "epoch": 0.40999272878362936,
+      "grad_norm": 0.4579329192638397,
+      "learning_rate": 6.395065210735464e-05,
+      "loss": 1.6841,
+      "step": 3947
+    },
+    {
+      "epoch": 0.41009660330320974,
+      "grad_norm": 0.4300961494445801,
+      "learning_rate": 6.39349827683368e-05,
+      "loss": 1.8659,
+      "step": 3948
+    },
+    {
+      "epoch": 0.41020047782279007,
+      "grad_norm": 0.4603518843650818,
+      "learning_rate": 6.391931194535199e-05,
+      "loss": 1.82,
+      "step": 3949
+    },
+    {
+      "epoch": 0.4103043523423704,
+      "grad_norm": 0.36023688316345215,
+      "learning_rate": 6.390363964006903e-05,
+      "loss": 1.5473,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4104082268619508,
+      "grad_norm": 0.36991775035858154,
+      "learning_rate": 6.38879658541569e-05,
+      "loss": 1.554,
+      "step": 3951
+    },
+    {
+      "epoch": 0.4105121013815311,
+      "grad_norm": 0.3790716528892517,
+      "learning_rate": 6.387229058928475e-05,
+      "loss": 1.6563,
+      "step": 3952
+    },
+    {
+      "epoch": 0.41061597590111143,
+      "grad_norm": 0.38919469714164734,
+      "learning_rate": 6.385661384712185e-05,
+      "loss": 1.6243,
+      "step": 3953
+    },
+    {
+      "epoch": 0.4107198504206918,
+      "grad_norm": 0.3934738337993622,
+      "learning_rate": 6.384093562933765e-05,
+      "loss": 1.6521,
+      "step": 3954
+    },
+    {
+      "epoch": 0.41082372494027214,
+      "grad_norm": 0.39593058824539185,
+      "learning_rate": 6.382525593760178e-05,
+      "loss": 1.7887,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4109275994598525,
+      "grad_norm": 0.47822320461273193,
+      "learning_rate": 6.380957477358399e-05,
+      "loss": 1.8647,
+      "step": 3956
+    },
+    {
+      "epoch": 0.41103147397943285,
+      "grad_norm": 0.44262126088142395,
+      "learning_rate": 6.379389213895421e-05,
+      "loss": 1.8746,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4111353484990132,
+      "grad_norm": 0.37435340881347656,
+      "learning_rate": 6.377820803538253e-05,
+      "loss": 1.6782,
+      "step": 3958
+    },
+    {
+      "epoch": 0.41123922301859356,
+      "grad_norm": 0.41944101452827454,
+      "learning_rate": 6.376252246453914e-05,
+      "loss": 1.7188,
+      "step": 3959
+    },
+    {
+      "epoch": 0.4113430975381739,
+      "grad_norm": 0.43215543031692505,
+      "learning_rate": 6.374683542809447e-05,
+      "loss": 1.8221,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4114469720577542,
+      "grad_norm": 0.40915408730506897,
+      "learning_rate": 6.373114692771906e-05,
+      "loss": 1.7294,
+      "step": 3961
+    },
+    {
+      "epoch": 0.4115508465773346,
+      "grad_norm": 0.41122907400131226,
+      "learning_rate": 6.371545696508358e-05,
+      "loss": 1.708,
+      "step": 3962
+    },
+    {
+      "epoch": 0.4116547210969149,
+      "grad_norm": 0.3763563930988312,
+      "learning_rate": 6.369976554185896e-05,
+      "loss": 1.8057,
+      "step": 3963
+    },
+    {
+      "epoch": 0.41175859561649525,
+      "grad_norm": 0.4643213748931885,
+      "learning_rate": 6.368407265971616e-05,
+      "loss": 1.6913,
+      "step": 3964
+    },
+    {
+      "epoch": 0.41186247013607563,
+      "grad_norm": 0.3934086263179779,
+      "learning_rate": 6.366837832032635e-05,
+      "loss": 1.8158,
+      "step": 3965
+    },
+    {
+      "epoch": 0.41196634465565596,
+      "grad_norm": 0.35763460397720337,
+      "learning_rate": 6.365268252536087e-05,
+      "loss": 1.5065,
+      "step": 3966
+    },
+    {
+      "epoch": 0.41207021917523634,
+      "grad_norm": 0.381234347820282,
+      "learning_rate": 6.36369852764912e-05,
+      "loss": 1.672,
+      "step": 3967
+    },
+    {
+      "epoch": 0.41217409369481667,
+      "grad_norm": 0.4128655195236206,
+      "learning_rate": 6.362128657538897e-05,
+      "loss": 1.8548,
+      "step": 3968
+    },
+    {
+      "epoch": 0.412277968214397,
+      "grad_norm": 0.42314839363098145,
+      "learning_rate": 6.360558642372597e-05,
+      "loss": 1.8469,
+      "step": 3969
+    },
+    {
+      "epoch": 0.4123818427339774,
+      "grad_norm": 0.3665439188480377,
+      "learning_rate": 6.358988482317414e-05,
+      "loss": 1.6893,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4124857172535577,
+      "grad_norm": 0.3973866105079651,
+      "learning_rate": 6.357418177540558e-05,
+      "loss": 1.5002,
+      "step": 3971
+    },
+    {
+      "epoch": 0.41258959177313803,
+      "grad_norm": 0.40862172842025757,
+      "learning_rate": 6.355847728209257e-05,
+      "loss": 1.817,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4126934662927184,
+      "grad_norm": 0.439984530210495,
+      "learning_rate": 6.354277134490746e-05,
+      "loss": 1.7698,
+      "step": 3973
+    },
+    {
+      "epoch": 0.41279734081229874,
+      "grad_norm": 0.3773249685764313,
+      "learning_rate": 6.352706396552285e-05,
+      "loss": 1.4967,
+      "step": 3974
+    },
+    {
+      "epoch": 0.41290121533187907,
+      "grad_norm": 0.4068536162376404,
+      "learning_rate": 6.351135514561144e-05,
+      "loss": 1.6282,
+      "step": 3975
+    },
+    {
+      "epoch": 0.41300508985145945,
+      "grad_norm": 0.3690778613090515,
+      "learning_rate": 6.34956448868461e-05,
+      "loss": 1.637,
+      "step": 3976
+    },
+    {
+      "epoch": 0.4131089643710398,
+      "grad_norm": 0.4202233552932739,
+      "learning_rate": 6.347993319089985e-05,
+      "loss": 1.9794,
+      "step": 3977
+    },
+    {
+      "epoch": 0.4132128388906201,
+      "grad_norm": 0.44085201621055603,
+      "learning_rate": 6.346422005944586e-05,
+      "loss": 1.8328,
+      "step": 3978
+    },
+    {
+      "epoch": 0.4133167134102005,
+      "grad_norm": 0.39549145102500916,
+      "learning_rate": 6.344850549415746e-05,
+      "loss": 1.7727,
+      "step": 3979
+    },
+    {
+      "epoch": 0.4134205879297808,
+      "grad_norm": 0.39016398787498474,
+      "learning_rate": 6.343278949670812e-05,
+      "loss": 1.6293,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4135244624493612,
+      "grad_norm": 0.40789923071861267,
+      "learning_rate": 6.341707206877149e-05,
+      "loss": 1.8089,
+      "step": 3981
+    },
+    {
+      "epoch": 0.4136283369689415,
+      "grad_norm": 0.41951802372932434,
+      "learning_rate": 6.340135321202134e-05,
+      "loss": 1.793,
+      "step": 3982
+    },
+    {
+      "epoch": 0.41373221148852185,
+      "grad_norm": 0.41457945108413696,
+      "learning_rate": 6.33856329281316e-05,
+      "loss": 1.8366,
+      "step": 3983
+    },
+    {
+      "epoch": 0.41383608600810223,
+      "grad_norm": 0.4352897107601166,
+      "learning_rate": 6.336991121877637e-05,
+      "loss": 1.7176,
+      "step": 3984
+    },
+    {
+      "epoch": 0.41393996052768256,
+      "grad_norm": 0.414348840713501,
+      "learning_rate": 6.335418808562988e-05,
+      "loss": 1.8262,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4140438350472629,
+      "grad_norm": 0.3896345794200897,
+      "learning_rate": 6.333846353036654e-05,
+      "loss": 1.8285,
+      "step": 3986
+    },
+    {
+      "epoch": 0.41414770956684327,
+      "grad_norm": 0.4062201678752899,
+      "learning_rate": 6.332273755466087e-05,
+      "loss": 1.709,
+      "step": 3987
+    },
+    {
+      "epoch": 0.4142515840864236,
+      "grad_norm": 0.4064597189426422,
+      "learning_rate": 6.330701016018757e-05,
+      "loss": 1.7283,
+      "step": 3988
+    },
+    {
+      "epoch": 0.4143554586060039,
+      "grad_norm": 0.42126670479774475,
+      "learning_rate": 6.329128134862148e-05,
+      "loss": 1.7227,
+      "step": 3989
+    },
+    {
+      "epoch": 0.4144593331255843,
+      "grad_norm": 0.38570138812065125,
+      "learning_rate": 6.32755511216376e-05,
+      "loss": 1.7244,
+      "step": 3990
+    },
+    {
+      "epoch": 0.41456320764516463,
+      "grad_norm": 0.4073387682437897,
+      "learning_rate": 6.32598194809111e-05,
+      "loss": 1.6718,
+      "step": 3991
+    },
+    {
+      "epoch": 0.414667082164745,
+      "grad_norm": 0.39126601815223694,
+      "learning_rate": 6.324408642811726e-05,
+      "loss": 1.584,
+      "step": 3992
+    },
+    {
+      "epoch": 0.41477095668432534,
+      "grad_norm": 0.38054293394088745,
+      "learning_rate": 6.32283519649315e-05,
+      "loss": 1.5411,
+      "step": 3993
+    },
+    {
+      "epoch": 0.41487483120390567,
+      "grad_norm": 0.42494648694992065,
+      "learning_rate": 6.321261609302945e-05,
+      "loss": 1.8188,
+      "step": 3994
+    },
+    {
+      "epoch": 0.41497870572348605,
+      "grad_norm": 0.3690442442893982,
+      "learning_rate": 6.319687881408683e-05,
+      "loss": 1.6147,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4150825802430664,
+      "grad_norm": 0.41196659207344055,
+      "learning_rate": 6.318114012977958e-05,
+      "loss": 1.917,
+      "step": 3996
+    },
+    {
+      "epoch": 0.4151864547626467,
+      "grad_norm": 0.38468247652053833,
+      "learning_rate": 6.316540004178371e-05,
+      "loss": 1.7175,
+      "step": 3997
+    },
+    {
+      "epoch": 0.4152903292822271,
+      "grad_norm": 0.36904019117355347,
+      "learning_rate": 6.314965855177544e-05,
+      "loss": 1.6544,
+      "step": 3998
+    },
+    {
+      "epoch": 0.4153942038018074,
+      "grad_norm": 0.3496026396751404,
+      "learning_rate": 6.31339156614311e-05,
+      "loss": 1.4105,
+      "step": 3999
+    },
+    {
+      "epoch": 0.41549807832138774,
+      "grad_norm": 0.3861815333366394,
+      "learning_rate": 6.311817137242721e-05,
+      "loss": 1.7974,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4156019528409681,
+      "grad_norm": 0.37794098258018494,
+      "learning_rate": 6.310242568644035e-05,
+      "loss": 1.5556,
+      "step": 4001
+    },
+    {
+      "epoch": 0.41570582736054845,
+      "grad_norm": 0.42547518014907837,
+      "learning_rate": 6.30866786051474e-05,
+      "loss": 1.6694,
+      "step": 4002
+    },
+    {
+      "epoch": 0.4158097018801288,
+      "grad_norm": 0.3894122540950775,
+      "learning_rate": 6.307093013022525e-05,
+      "loss": 1.5479,
+      "step": 4003
+    },
+    {
+      "epoch": 0.41591357639970916,
+      "grad_norm": 0.36968937516212463,
+      "learning_rate": 6.305518026335097e-05,
+      "loss": 1.6548,
+      "step": 4004
+    },
+    {
+      "epoch": 0.4160174509192895,
+      "grad_norm": 0.40135595202445984,
+      "learning_rate": 6.303942900620185e-05,
+      "loss": 1.5989,
+      "step": 4005
+    },
+    {
+      "epoch": 0.41612132543886987,
+      "grad_norm": 0.39152655005455017,
+      "learning_rate": 6.302367636045526e-05,
+      "loss": 1.6618,
+      "step": 4006
+    },
+    {
+      "epoch": 0.4162251999584502,
+      "grad_norm": 0.382112592458725,
+      "learning_rate": 6.300792232778872e-05,
+      "loss": 1.5055,
+      "step": 4007
+    },
+    {
+      "epoch": 0.4163290744780305,
+      "grad_norm": 0.4156250059604645,
+      "learning_rate": 6.29921669098799e-05,
+      "loss": 1.9369,
+      "step": 4008
+    },
+    {
+      "epoch": 0.4164329489976109,
+      "grad_norm": 0.3940093219280243,
+      "learning_rate": 6.297641010840665e-05,
+      "loss": 1.6275,
+      "step": 4009
+    },
+    {
+      "epoch": 0.41653682351719123,
+      "grad_norm": 0.4581923186779022,
+      "learning_rate": 6.296065192504697e-05,
+      "loss": 1.8641,
+      "step": 4010
+    },
+    {
+      "epoch": 0.41664069803677156,
+      "grad_norm": 0.3822265863418579,
+      "learning_rate": 6.294489236147894e-05,
+      "loss": 1.7272,
+      "step": 4011
+    },
+    {
+      "epoch": 0.41674457255635194,
+      "grad_norm": 0.3892137110233307,
+      "learning_rate": 6.292913141938084e-05,
+      "loss": 1.683,
+      "step": 4012
+    },
+    {
+      "epoch": 0.41684844707593227,
+      "grad_norm": 0.3944839835166931,
+      "learning_rate": 6.29133691004311e-05,
+      "loss": 1.6662,
+      "step": 4013
+    },
+    {
+      "epoch": 0.4169523215955126,
+      "grad_norm": 0.39415454864501953,
+      "learning_rate": 6.289760540630829e-05,
+      "loss": 1.6862,
+      "step": 4014
+    },
+    {
+      "epoch": 0.417056196115093,
+      "grad_norm": 0.39324238896369934,
+      "learning_rate": 6.288184033869109e-05,
+      "loss": 1.7041,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4171600706346733,
+      "grad_norm": 0.4160117506980896,
+      "learning_rate": 6.28660738992584e-05,
+      "loss": 1.7689,
+      "step": 4016
+    },
+    {
+      "epoch": 0.41726394515425363,
+      "grad_norm": 0.4204423725605011,
+      "learning_rate": 6.285030608968918e-05,
+      "loss": 1.7707,
+      "step": 4017
+    },
+    {
+      "epoch": 0.417367819673834,
+      "grad_norm": 0.3935635983943939,
+      "learning_rate": 6.28345369116626e-05,
+      "loss": 1.5884,
+      "step": 4018
+    },
+    {
+      "epoch": 0.41747169419341434,
+      "grad_norm": 0.41393959522247314,
+      "learning_rate": 6.281876636685795e-05,
+      "loss": 1.6574,
+      "step": 4019
+    },
+    {
+      "epoch": 0.4175755687129947,
+      "grad_norm": 0.4115196764469147,
+      "learning_rate": 6.280299445695469e-05,
+      "loss": 1.8304,
+      "step": 4020
+    },
+    {
+      "epoch": 0.41767944323257505,
+      "grad_norm": 0.3949635922908783,
+      "learning_rate": 6.278722118363237e-05,
+      "loss": 1.5268,
+      "step": 4021
+    },
+    {
+      "epoch": 0.4177833177521554,
+      "grad_norm": 0.41173237562179565,
+      "learning_rate": 6.277144654857074e-05,
+      "loss": 1.8324,
+      "step": 4022
+    },
+    {
+      "epoch": 0.41788719227173576,
+      "grad_norm": 0.3953053951263428,
+      "learning_rate": 6.275567055344967e-05,
+      "loss": 1.6818,
+      "step": 4023
+    },
+    {
+      "epoch": 0.4179910667913161,
+      "grad_norm": 0.46665164828300476,
+      "learning_rate": 6.27398931999492e-05,
+      "loss": 1.8911,
+      "step": 4024
+    },
+    {
+      "epoch": 0.4180949413108964,
+      "grad_norm": 0.4181447923183441,
+      "learning_rate": 6.272411448974947e-05,
+      "loss": 1.7229,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4181988158304768,
+      "grad_norm": 0.3725024461746216,
+      "learning_rate": 6.270833442453079e-05,
+      "loss": 1.664,
+      "step": 4026
+    },
+    {
+      "epoch": 0.4183026903500571,
+      "grad_norm": 0.430210679769516,
+      "learning_rate": 6.269255300597364e-05,
+      "loss": 1.6161,
+      "step": 4027
+    },
+    {
+      "epoch": 0.41840656486963745,
+      "grad_norm": 0.3929589092731476,
+      "learning_rate": 6.267677023575859e-05,
+      "loss": 1.7089,
+      "step": 4028
+    },
+    {
+      "epoch": 0.41851043938921784,
+      "grad_norm": 0.3702670931816101,
+      "learning_rate": 6.26609861155664e-05,
+      "loss": 1.6041,
+      "step": 4029
+    },
+    {
+      "epoch": 0.41861431390879816,
+      "grad_norm": 0.3990215063095093,
+      "learning_rate": 6.264520064707795e-05,
+      "loss": 1.7304,
+      "step": 4030
+    },
+    {
+      "epoch": 0.41871818842837855,
+      "grad_norm": 0.4187242388725281,
+      "learning_rate": 6.262941383197425e-05,
+      "loss": 1.813,
+      "step": 4031
+    },
+    {
+      "epoch": 0.4188220629479589,
+      "grad_norm": 0.46325740218162537,
+      "learning_rate": 6.261362567193651e-05,
+      "loss": 1.8182,
+      "step": 4032
+    },
+    {
+      "epoch": 0.4189259374675392,
+      "grad_norm": 0.3769034445285797,
+      "learning_rate": 6.259783616864601e-05,
+      "loss": 1.733,
+      "step": 4033
+    },
+    {
+      "epoch": 0.4190298119871196,
+      "grad_norm": 0.42521029710769653,
+      "learning_rate": 6.258204532378423e-05,
+      "loss": 1.8581,
+      "step": 4034
+    },
+    {
+      "epoch": 0.4191336865066999,
+      "grad_norm": 0.3773384988307953,
+      "learning_rate": 6.256625313903278e-05,
+      "loss": 1.6578,
+      "step": 4035
+    },
+    {
+      "epoch": 0.41923756102628024,
+      "grad_norm": 0.4068077504634857,
+      "learning_rate": 6.255045961607338e-05,
+      "loss": 1.7315,
+      "step": 4036
+    },
+    {
+      "epoch": 0.4193414355458606,
+      "grad_norm": 0.38190507888793945,
+      "learning_rate": 6.253466475658792e-05,
+      "loss": 1.5881,
+      "step": 4037
+    },
+    {
+      "epoch": 0.41944531006544095,
+      "grad_norm": 0.373384028673172,
+      "learning_rate": 6.251886856225844e-05,
+      "loss": 1.719,
+      "step": 4038
+    },
+    {
+      "epoch": 0.4195491845850213,
+      "grad_norm": 0.3872770369052887,
+      "learning_rate": 6.250307103476712e-05,
+      "loss": 1.6688,
+      "step": 4039
+    },
+    {
+      "epoch": 0.41965305910460166,
+      "grad_norm": 0.4040100574493408,
+      "learning_rate": 6.248727217579625e-05,
+      "loss": 1.7883,
+      "step": 4040
+    },
+    {
+      "epoch": 0.419756933624182,
+      "grad_norm": 0.5108729004859924,
+      "learning_rate": 6.247147198702832e-05,
+      "loss": 1.8919,
+      "step": 4041
+    },
+    {
+      "epoch": 0.4198608081437623,
+      "grad_norm": 0.4295211732387543,
+      "learning_rate": 6.245567047014589e-05,
+      "loss": 1.7379,
+      "step": 4042
+    },
+    {
+      "epoch": 0.4199646826633427,
+      "grad_norm": 0.4218069911003113,
+      "learning_rate": 6.243986762683171e-05,
+      "loss": 1.7343,
+      "step": 4043
+    },
+    {
+      "epoch": 0.420068557182923,
+      "grad_norm": 0.40920522809028625,
+      "learning_rate": 6.242406345876866e-05,
+      "loss": 1.8147,
+      "step": 4044
+    },
+    {
+      "epoch": 0.4201724317025034,
+      "grad_norm": 0.3913504481315613,
+      "learning_rate": 6.240825796763977e-05,
+      "loss": 1.6202,
+      "step": 4045
+    },
+    {
+      "epoch": 0.42027630622208373,
+      "grad_norm": 0.387791246175766,
+      "learning_rate": 6.239245115512818e-05,
+      "loss": 1.6385,
+      "step": 4046
+    },
+    {
+      "epoch": 0.42038018074166406,
+      "grad_norm": 0.3972698450088501,
+      "learning_rate": 6.237664302291722e-05,
+      "loss": 1.6763,
+      "step": 4047
+    },
+    {
+      "epoch": 0.42048405526124444,
+      "grad_norm": 0.35716134309768677,
+      "learning_rate": 6.236083357269033e-05,
+      "loss": 1.6527,
+      "step": 4048
+    },
+    {
+      "epoch": 0.42058792978082477,
+      "grad_norm": 0.40162917971611023,
+      "learning_rate": 6.234502280613108e-05,
+      "loss": 1.7658,
+      "step": 4049
+    },
+    {
+      "epoch": 0.4206918043004051,
+      "grad_norm": 0.3818826377391815,
+      "learning_rate": 6.232921072492318e-05,
+      "loss": 1.7382,
+      "step": 4050
+    },
+    {
+      "epoch": 0.4207956788199855,
+      "grad_norm": 0.4017144739627838,
+      "learning_rate": 6.23133973307505e-05,
+      "loss": 1.7945,
+      "step": 4051
+    },
+    {
+      "epoch": 0.4208995533395658,
+      "grad_norm": 0.3896738290786743,
+      "learning_rate": 6.229758262529708e-05,
+      "loss": 1.8404,
+      "step": 4052
+    },
+    {
+      "epoch": 0.42100342785914613,
+      "grad_norm": 0.38667070865631104,
+      "learning_rate": 6.228176661024703e-05,
+      "loss": 1.619,
+      "step": 4053
+    },
+    {
+      "epoch": 0.4211073023787265,
+      "grad_norm": 0.38369256258010864,
+      "learning_rate": 6.226594928728462e-05,
+      "loss": 1.6359,
+      "step": 4054
+    },
+    {
+      "epoch": 0.42121117689830684,
+      "grad_norm": 0.44246816635131836,
+      "learning_rate": 6.225013065809431e-05,
+      "loss": 1.8973,
+      "step": 4055
+    },
+    {
+      "epoch": 0.42131505141788717,
+      "grad_norm": 0.42337489128112793,
+      "learning_rate": 6.223431072436063e-05,
+      "loss": 1.6915,
+      "step": 4056
+    },
+    {
+      "epoch": 0.42141892593746755,
+      "grad_norm": 0.419403612613678,
+      "learning_rate": 6.221848948776828e-05,
+      "loss": 1.6772,
+      "step": 4057
+    },
+    {
+      "epoch": 0.4215228004570479,
+      "grad_norm": 0.3898250460624695,
+      "learning_rate": 6.220266695000211e-05,
+      "loss": 1.7213,
+      "step": 4058
+    },
+    {
+      "epoch": 0.42162667497662826,
+      "grad_norm": 0.4045391380786896,
+      "learning_rate": 6.21868431127471e-05,
+      "loss": 1.9761,
+      "step": 4059
+    },
+    {
+      "epoch": 0.4217305494962086,
+      "grad_norm": 0.3763904273509979,
+      "learning_rate": 6.217101797768837e-05,
+      "loss": 1.7037,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4218344240157889,
+      "grad_norm": 0.397461861371994,
+      "learning_rate": 6.215519154651116e-05,
+      "loss": 1.7218,
+      "step": 4061
+    },
+    {
+      "epoch": 0.4219382985353693,
+      "grad_norm": 0.4115433990955353,
+      "learning_rate": 6.213936382090085e-05,
+      "loss": 1.7957,
+      "step": 4062
+    },
+    {
+      "epoch": 0.4220421730549496,
+      "grad_norm": 0.393285870552063,
+      "learning_rate": 6.2123534802543e-05,
+      "loss": 1.7508,
+      "step": 4063
+    },
+    {
+      "epoch": 0.42214604757452995,
+      "grad_norm": 0.3760216236114502,
+      "learning_rate": 6.210770449312326e-05,
+      "loss": 1.6243,
+      "step": 4064
+    },
+    {
+      "epoch": 0.42224992209411033,
+      "grad_norm": 0.4444178342819214,
+      "learning_rate": 6.209187289432742e-05,
+      "loss": 1.8161,
+      "step": 4065
+    },
+    {
+      "epoch": 0.42235379661369066,
+      "grad_norm": 0.37198397517204285,
+      "learning_rate": 6.207604000784143e-05,
+      "loss": 1.6786,
+      "step": 4066
+    },
+    {
+      "epoch": 0.422457671133271,
+      "grad_norm": 0.39783281087875366,
+      "learning_rate": 6.206020583535141e-05,
+      "loss": 1.7703,
+      "step": 4067
+    },
+    {
+      "epoch": 0.42256154565285137,
+      "grad_norm": 0.45979025959968567,
+      "learning_rate": 6.204437037854351e-05,
+      "loss": 1.8734,
+      "step": 4068
+    },
+    {
+      "epoch": 0.4226654201724317,
+      "grad_norm": 0.42977213859558105,
+      "learning_rate": 6.202853363910413e-05,
+      "loss": 1.7048,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4227692946920121,
+      "grad_norm": 0.41636922955513,
+      "learning_rate": 6.201269561871975e-05,
+      "loss": 1.6683,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4228731692115924,
+      "grad_norm": 0.4053545594215393,
+      "learning_rate": 6.199685631907697e-05,
+      "loss": 1.6983,
+      "step": 4071
+    },
+    {
+      "epoch": 0.42297704373117273,
+      "grad_norm": 0.378132164478302,
+      "learning_rate": 6.19810157418626e-05,
+      "loss": 1.6345,
+      "step": 4072
+    },
+    {
+      "epoch": 0.4230809182507531,
+      "grad_norm": 0.4450819492340088,
+      "learning_rate": 6.196517388876348e-05,
+      "loss": 1.6669,
+      "step": 4073
+    },
+    {
+      "epoch": 0.42318479277033344,
+      "grad_norm": 0.3707204759120941,
+      "learning_rate": 6.194933076146668e-05,
+      "loss": 1.5462,
+      "step": 4074
+    },
+    {
+      "epoch": 0.42328866728991377,
+      "grad_norm": 0.41867223381996155,
+      "learning_rate": 6.193348636165937e-05,
+      "loss": 1.8317,
+      "step": 4075
+    },
+    {
+      "epoch": 0.42339254180949415,
+      "grad_norm": 0.4516044557094574,
+      "learning_rate": 6.191764069102883e-05,
+      "loss": 1.8396,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4234964163290745,
+      "grad_norm": 0.421350359916687,
+      "learning_rate": 6.190179375126254e-05,
+      "loss": 1.7317,
+      "step": 4077
+    },
+    {
+      "epoch": 0.4236002908486548,
+      "grad_norm": 0.3848811984062195,
+      "learning_rate": 6.188594554404804e-05,
+      "loss": 1.6922,
+      "step": 4078
+    },
+    {
+      "epoch": 0.4237041653682352,
+      "grad_norm": 0.41102609038352966,
+      "learning_rate": 6.187009607107304e-05,
+      "loss": 1.7322,
+      "step": 4079
+    },
+    {
+      "epoch": 0.4238080398878155,
+      "grad_norm": 0.4180265963077545,
+      "learning_rate": 6.185424533402543e-05,
+      "loss": 1.8053,
+      "step": 4080
+    },
+    {
+      "epoch": 0.42391191440739584,
+      "grad_norm": 0.4037413001060486,
+      "learning_rate": 6.183839333459315e-05,
+      "loss": 1.6599,
+      "step": 4081
+    },
+    {
+      "epoch": 0.4240157889269762,
+      "grad_norm": 0.40133136510849,
+      "learning_rate": 6.18225400744643e-05,
+      "loss": 1.7394,
+      "step": 4082
+    },
+    {
+      "epoch": 0.42411966344655655,
+      "grad_norm": 0.37302759289741516,
+      "learning_rate": 6.180668555532719e-05,
+      "loss": 1.5259,
+      "step": 4083
+    },
+    {
+      "epoch": 0.42422353796613693,
+      "grad_norm": 0.39756691455841064,
+      "learning_rate": 6.179082977887013e-05,
+      "loss": 1.6746,
+      "step": 4084
+    },
+    {
+      "epoch": 0.42432741248571726,
+      "grad_norm": 0.37990039587020874,
+      "learning_rate": 6.177497274678168e-05,
+      "loss": 1.6812,
+      "step": 4085
+    },
+    {
+      "epoch": 0.4244312870052976,
+      "grad_norm": 0.3902164697647095,
+      "learning_rate": 6.175911446075049e-05,
+      "loss": 1.6457,
+      "step": 4086
+    },
+    {
+      "epoch": 0.42453516152487797,
+      "grad_norm": 0.4064858555793762,
+      "learning_rate": 6.174325492246531e-05,
+      "loss": 1.7168,
+      "step": 4087
+    },
+    {
+      "epoch": 0.4246390360444583,
+      "grad_norm": 0.40128055214881897,
+      "learning_rate": 6.17273941336151e-05,
+      "loss": 1.7847,
+      "step": 4088
+    },
+    {
+      "epoch": 0.4247429105640386,
+      "grad_norm": 0.4626508951187134,
+      "learning_rate": 6.171153209588891e-05,
+      "loss": 1.951,
+      "step": 4089
+    },
+    {
+      "epoch": 0.424846785083619,
+      "grad_norm": 0.40969690680503845,
+      "learning_rate": 6.169566881097588e-05,
+      "loss": 1.4734,
+      "step": 4090
+    },
+    {
+      "epoch": 0.42495065960319933,
+      "grad_norm": 0.3807416558265686,
+      "learning_rate": 6.167980428056537e-05,
+      "loss": 1.5504,
+      "step": 4091
+    },
+    {
+      "epoch": 0.42505453412277966,
+      "grad_norm": 0.3962438702583313,
+      "learning_rate": 6.16639385063468e-05,
+      "loss": 1.8144,
+      "step": 4092
+    },
+    {
+      "epoch": 0.42515840864236004,
+      "grad_norm": 0.3545083701610565,
+      "learning_rate": 6.164807149000974e-05,
+      "loss": 1.5118,
+      "step": 4093
+    },
+    {
+      "epoch": 0.42526228316194037,
+      "grad_norm": 0.43927130103111267,
+      "learning_rate": 6.163220323324395e-05,
+      "loss": 1.8641,
+      "step": 4094
+    },
+    {
+      "epoch": 0.4253661576815207,
+      "grad_norm": 0.41815078258514404,
+      "learning_rate": 6.161633373773925e-05,
+      "loss": 1.7244,
+      "step": 4095
+    },
+    {
+      "epoch": 0.4254700322011011,
+      "grad_norm": 0.4570396840572357,
+      "learning_rate": 6.16004630051856e-05,
+      "loss": 1.8696,
+      "step": 4096
+    },
+    {
+      "epoch": 0.4255739067206814,
+      "grad_norm": 0.3970779478549957,
+      "learning_rate": 6.158459103727316e-05,
+      "loss": 1.7826,
+      "step": 4097
+    },
+    {
+      "epoch": 0.4256777812402618,
+      "grad_norm": 0.437796026468277,
+      "learning_rate": 6.15687178356921e-05,
+      "loss": 1.6452,
+      "step": 4098
+    },
+    {
+      "epoch": 0.4257816557598421,
+      "grad_norm": 0.4131401777267456,
+      "learning_rate": 6.155284340213285e-05,
+      "loss": 1.6827,
+      "step": 4099
+    },
+    {
+      "epoch": 0.42588553027942244,
+      "grad_norm": 0.4367378056049347,
+      "learning_rate": 6.15369677382859e-05,
+      "loss": 1.9042,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4259894047990028,
+      "grad_norm": 0.39361101388931274,
+      "learning_rate": 6.152109084584186e-05,
+      "loss": 1.643,
+      "step": 4101
+    },
+    {
+      "epoch": 0.42609327931858315,
+      "grad_norm": 0.37626150250434875,
+      "learning_rate": 6.150521272649152e-05,
+      "loss": 1.5454,
+      "step": 4102
+    },
+    {
+      "epoch": 0.4261971538381635,
+      "grad_norm": 0.3899744749069214,
+      "learning_rate": 6.148933338192578e-05,
+      "loss": 1.6159,
+      "step": 4103
+    },
+    {
+      "epoch": 0.42630102835774386,
+      "grad_norm": 0.4217323362827301,
+      "learning_rate": 6.147345281383565e-05,
+      "loss": 1.8407,
+      "step": 4104
+    },
+    {
+      "epoch": 0.4264049028773242,
+      "grad_norm": 0.42654165625572205,
+      "learning_rate": 6.145757102391227e-05,
+      "loss": 1.7704,
+      "step": 4105
+    },
+    {
+      "epoch": 0.4265087773969045,
+      "grad_norm": 0.3943133056163788,
+      "learning_rate": 6.144168801384698e-05,
+      "loss": 1.7799,
+      "step": 4106
+    },
+    {
+      "epoch": 0.4266126519164849,
+      "grad_norm": 0.37911903858184814,
+      "learning_rate": 6.142580378533114e-05,
+      "loss": 1.6106,
+      "step": 4107
+    },
+    {
+      "epoch": 0.4267165264360652,
+      "grad_norm": 0.42597952485084534,
+      "learning_rate": 6.140991834005632e-05,
+      "loss": 1.6198,
+      "step": 4108
+    },
+    {
+      "epoch": 0.4268204009556456,
+      "grad_norm": 0.43222105503082275,
+      "learning_rate": 6.139403167971422e-05,
+      "loss": 1.7884,
+      "step": 4109
+    },
+    {
+      "epoch": 0.42692427547522593,
+      "grad_norm": 0.4335835576057434,
+      "learning_rate": 6.13781438059966e-05,
+      "loss": 1.7371,
+      "step": 4110
+    },
+    {
+      "epoch": 0.42702814999480626,
+      "grad_norm": 0.41362127661705017,
+      "learning_rate": 6.136225472059545e-05,
+      "loss": 1.6047,
+      "step": 4111
+    },
+    {
+      "epoch": 0.42713202451438664,
+      "grad_norm": 0.38855838775634766,
+      "learning_rate": 6.134636442520277e-05,
+      "loss": 1.6329,
+      "step": 4112
+    },
+    {
+      "epoch": 0.42723589903396697,
+      "grad_norm": 0.39782068133354187,
+      "learning_rate": 6.13304729215108e-05,
+      "loss": 1.711,
+      "step": 4113
+    },
+    {
+      "epoch": 0.4273397735535473,
+      "grad_norm": 0.4341132342815399,
+      "learning_rate": 6.131458021121184e-05,
+      "loss": 1.6623,
+      "step": 4114
+    },
+    {
+      "epoch": 0.4274436480731277,
+      "grad_norm": 0.4523240625858307,
+      "learning_rate": 6.129868629599832e-05,
+      "loss": 1.8319,
+      "step": 4115
+    },
+    {
+      "epoch": 0.427547522592708,
+      "grad_norm": 0.3760691285133362,
+      "learning_rate": 6.128279117756288e-05,
+      "loss": 1.8514,
+      "step": 4116
+    },
+    {
+      "epoch": 0.42765139711228833,
+      "grad_norm": 0.45057788491249084,
+      "learning_rate": 6.126689485759818e-05,
+      "loss": 1.7907,
+      "step": 4117
+    },
+    {
+      "epoch": 0.4277552716318687,
+      "grad_norm": 0.4334897994995117,
+      "learning_rate": 6.125099733779706e-05,
+      "loss": 1.5299,
+      "step": 4118
+    },
+    {
+      "epoch": 0.42785914615144904,
+      "grad_norm": 0.4173983633518219,
+      "learning_rate": 6.12350986198525e-05,
+      "loss": 1.6515,
+      "step": 4119
+    },
+    {
+      "epoch": 0.42796302067102937,
+      "grad_norm": 0.4509902000427246,
+      "learning_rate": 6.121919870545755e-05,
+      "loss": 1.9308,
+      "step": 4120
+    },
+    {
+      "epoch": 0.42806689519060975,
+      "grad_norm": 0.433244526386261,
+      "learning_rate": 6.120329759630546e-05,
+      "loss": 1.9438,
+      "step": 4121
+    },
+    {
+      "epoch": 0.4281707697101901,
+      "grad_norm": 0.4360440671443939,
+      "learning_rate": 6.118739529408956e-05,
+      "loss": 1.8285,
+      "step": 4122
+    },
+    {
+      "epoch": 0.42827464422977046,
+      "grad_norm": 0.44721856713294983,
+      "learning_rate": 6.117149180050332e-05,
+      "loss": 1.7948,
+      "step": 4123
+    },
+    {
+      "epoch": 0.4283785187493508,
+      "grad_norm": 0.4097500741481781,
+      "learning_rate": 6.115558711724036e-05,
+      "loss": 1.6341,
+      "step": 4124
+    },
+    {
+      "epoch": 0.4284823932689311,
+      "grad_norm": 0.3732147514820099,
+      "learning_rate": 6.113968124599439e-05,
+      "loss": 1.6512,
+      "step": 4125
+    },
+    {
+      "epoch": 0.4285862677885115,
+      "grad_norm": 0.407043993473053,
+      "learning_rate": 6.112377418845925e-05,
+      "loss": 1.7734,
+      "step": 4126
+    },
+    {
+      "epoch": 0.4286901423080918,
+      "grad_norm": 0.4379003047943115,
+      "learning_rate": 6.110786594632892e-05,
+      "loss": 1.8773,
+      "step": 4127
+    },
+    {
+      "epoch": 0.42879401682767215,
+      "grad_norm": 0.49205613136291504,
+      "learning_rate": 6.109195652129753e-05,
+      "loss": 1.6428,
+      "step": 4128
+    },
+    {
+      "epoch": 0.42889789134725254,
+      "grad_norm": 0.40176907181739807,
+      "learning_rate": 6.107604591505928e-05,
+      "loss": 1.5291,
+      "step": 4129
+    },
+    {
+      "epoch": 0.42900176586683286,
+      "grad_norm": 0.4463737905025482,
+      "learning_rate": 6.106013412930853e-05,
+      "loss": 1.8709,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4291056403864132,
+      "grad_norm": 0.49634161591529846,
+      "learning_rate": 6.10442211657398e-05,
+      "loss": 2.1084,
+      "step": 4131
+    },
+    {
+      "epoch": 0.42920951490599357,
+      "grad_norm": 0.3978555500507355,
+      "learning_rate": 6.1028307026047624e-05,
+      "loss": 1.7182,
+      "step": 4132
+    },
+    {
+      "epoch": 0.4293133894255739,
+      "grad_norm": 0.39746221899986267,
+      "learning_rate": 6.1012391711926795e-05,
+      "loss": 1.7323,
+      "step": 4133
+    },
+    {
+      "epoch": 0.4294172639451542,
+      "grad_norm": 0.39065563678741455,
+      "learning_rate": 6.099647522507216e-05,
+      "loss": 1.7496,
+      "step": 4134
+    },
+    {
+      "epoch": 0.4295211384647346,
+      "grad_norm": 0.41617366671562195,
+      "learning_rate": 6.098055756717866e-05,
+      "loss": 1.781,
+      "step": 4135
+    },
+    {
+      "epoch": 0.42962501298431494,
+      "grad_norm": 0.41195693612098694,
+      "learning_rate": 6.0964638739941447e-05,
+      "loss": 1.7705,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4297288875038953,
+      "grad_norm": 0.397640585899353,
+      "learning_rate": 6.094871874505572e-05,
+      "loss": 1.7518,
+      "step": 4137
+    },
+    {
+      "epoch": 0.42983276202347565,
+      "grad_norm": 0.37663549184799194,
+      "learning_rate": 6.093279758421687e-05,
+      "loss": 1.6859,
+      "step": 4138
+    },
+    {
+      "epoch": 0.42993663654305597,
+      "grad_norm": 0.3834432065486908,
+      "learning_rate": 6.0916875259120366e-05,
+      "loss": 1.5105,
+      "step": 4139
+    },
+    {
+      "epoch": 0.43004051106263635,
+      "grad_norm": 0.4278203248977661,
+      "learning_rate": 6.090095177146178e-05,
+      "loss": 1.7621,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4301443855822167,
+      "grad_norm": 0.4155001938343048,
+      "learning_rate": 6.088502712293687e-05,
+      "loss": 1.8687,
+      "step": 4141
+    },
+    {
+      "epoch": 0.430248260101797,
+      "grad_norm": 0.4091452658176422,
+      "learning_rate": 6.086910131524147e-05,
+      "loss": 1.6906,
+      "step": 4142
+    },
+    {
+      "epoch": 0.4303521346213774,
+      "grad_norm": 0.4341062605381012,
+      "learning_rate": 6.085317435007156e-05,
+      "loss": 1.7431,
+      "step": 4143
+    },
+    {
+      "epoch": 0.4304560091409577,
+      "grad_norm": 0.4048025906085968,
+      "learning_rate": 6.0837246229123246e-05,
+      "loss": 1.7628,
+      "step": 4144
+    },
+    {
+      "epoch": 0.43055988366053805,
+      "grad_norm": 0.3721714913845062,
+      "learning_rate": 6.0821316954092745e-05,
+      "loss": 1.4422,
+      "step": 4145
+    },
+    {
+      "epoch": 0.43066375818011843,
+      "grad_norm": 0.39783287048339844,
+      "learning_rate": 6.08053865266764e-05,
+      "loss": 1.6521,
+      "step": 4146
+    },
+    {
+      "epoch": 0.43076763269969875,
+      "grad_norm": 0.44611310958862305,
+      "learning_rate": 6.078945494857068e-05,
+      "loss": 1.9144,
+      "step": 4147
+    },
+    {
+      "epoch": 0.43087150721927914,
+      "grad_norm": 0.3994828462600708,
+      "learning_rate": 6.077352222147214e-05,
+      "loss": 1.6819,
+      "step": 4148
+    },
+    {
+      "epoch": 0.43097538173885946,
+      "grad_norm": 0.4165269434452057,
+      "learning_rate": 6.075758834707754e-05,
+      "loss": 1.7586,
+      "step": 4149
+    },
+    {
+      "epoch": 0.4310792562584398,
+      "grad_norm": 0.4540640413761139,
+      "learning_rate": 6.0741653327083703e-05,
+      "loss": 1.8504,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4311831307780202,
+      "grad_norm": 0.4224712550640106,
+      "learning_rate": 6.072571716318756e-05,
+      "loss": 1.7859,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4312870052976005,
+      "grad_norm": 0.4140393137931824,
+      "learning_rate": 6.0709779857086214e-05,
+      "loss": 1.6534,
+      "step": 4152
+    },
+    {
+      "epoch": 0.43139087981718083,
+      "grad_norm": 0.37748152017593384,
+      "learning_rate": 6.069384141047686e-05,
+      "loss": 1.5697,
+      "step": 4153
+    },
+    {
+      "epoch": 0.4314947543367612,
+      "grad_norm": 0.4263037145137787,
+      "learning_rate": 6.067790182505678e-05,
+      "loss": 1.6565,
+      "step": 4154
+    },
+    {
+      "epoch": 0.43159862885634154,
+      "grad_norm": 0.38402366638183594,
+      "learning_rate": 6.066196110252347e-05,
+      "loss": 1.6159,
+      "step": 4155
+    },
+    {
+      "epoch": 0.43170250337592186,
+      "grad_norm": 0.3785112798213959,
+      "learning_rate": 6.064601924457447e-05,
+      "loss": 1.4971,
+      "step": 4156
+    },
+    {
+      "epoch": 0.43180637789550225,
+      "grad_norm": 0.3928913474082947,
+      "learning_rate": 6.0630076252907445e-05,
+      "loss": 1.6329,
+      "step": 4157
+    },
+    {
+      "epoch": 0.4319102524150826,
+      "grad_norm": 0.42711085081100464,
+      "learning_rate": 6.061413212922022e-05,
+      "loss": 1.7333,
+      "step": 4158
+    },
+    {
+      "epoch": 0.4320141269346629,
+      "grad_norm": 0.40235477685928345,
+      "learning_rate": 6.059818687521074e-05,
+      "loss": 1.6936,
+      "step": 4159
+    },
+    {
+      "epoch": 0.4321180014542433,
+      "grad_norm": 0.4238249957561493,
+      "learning_rate": 6.0582240492577015e-05,
+      "loss": 1.7692,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4322218759738236,
+      "grad_norm": 0.4094812870025635,
+      "learning_rate": 6.056629298301722e-05,
+      "loss": 1.6549,
+      "step": 4161
+    },
+    {
+      "epoch": 0.432325750493404,
+      "grad_norm": 0.4390838146209717,
+      "learning_rate": 6.0550344348229626e-05,
+      "loss": 1.8607,
+      "step": 4162
+    },
+    {
+      "epoch": 0.4324296250129843,
+      "grad_norm": 0.3718215823173523,
+      "learning_rate": 6.053439458991268e-05,
+      "loss": 1.6305,
+      "step": 4163
+    },
+    {
+      "epoch": 0.43253349953256465,
+      "grad_norm": 0.42819488048553467,
+      "learning_rate": 6.051844370976487e-05,
+      "loss": 1.5788,
+      "step": 4164
+    },
+    {
+      "epoch": 0.43263737405214503,
+      "grad_norm": 0.37748944759368896,
+      "learning_rate": 6.050249170948484e-05,
+      "loss": 1.6356,
+      "step": 4165
+    },
+    {
+      "epoch": 0.43274124857172536,
+      "grad_norm": 0.4419516623020172,
+      "learning_rate": 6.048653859077138e-05,
+      "loss": 1.7449,
+      "step": 4166
+    },
+    {
+      "epoch": 0.4328451230913057,
+      "grad_norm": 0.40298861265182495,
+      "learning_rate": 6.0470584355323375e-05,
+      "loss": 1.648,
+      "step": 4167
+    },
+    {
+      "epoch": 0.43294899761088607,
+      "grad_norm": 0.41812488436698914,
+      "learning_rate": 6.045462900483978e-05,
+      "loss": 1.8709,
+      "step": 4168
+    },
+    {
+      "epoch": 0.4330528721304664,
+      "grad_norm": 0.4024185538291931,
+      "learning_rate": 6.0438672541019755e-05,
+      "loss": 1.7013,
+      "step": 4169
+    },
+    {
+      "epoch": 0.4331567466500467,
+      "grad_norm": 0.3921603262424469,
+      "learning_rate": 6.042271496556254e-05,
+      "loss": 1.6795,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4332606211696271,
+      "grad_norm": 0.41133809089660645,
+      "learning_rate": 6.040675628016746e-05,
+      "loss": 1.7468,
+      "step": 4171
+    },
+    {
+      "epoch": 0.43336449568920743,
+      "grad_norm": 0.4003753960132599,
+      "learning_rate": 6.039079648653403e-05,
+      "loss": 1.7661,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4334683702087878,
+      "grad_norm": 0.3852844536304474,
+      "learning_rate": 6.037483558636183e-05,
+      "loss": 1.6341,
+      "step": 4173
+    },
+    {
+      "epoch": 0.43357224472836814,
+      "grad_norm": 0.40245676040649414,
+      "learning_rate": 6.035887358135056e-05,
+      "loss": 1.7788,
+      "step": 4174
+    },
+    {
+      "epoch": 0.43367611924794847,
+      "grad_norm": 0.4260011613368988,
+      "learning_rate": 6.0342910473200054e-05,
+      "loss": 1.8262,
+      "step": 4175
+    },
+    {
+      "epoch": 0.43377999376752885,
+      "grad_norm": 0.4148736298084259,
+      "learning_rate": 6.032694626361025e-05,
+      "loss": 1.7661,
+      "step": 4176
+    },
+    {
+      "epoch": 0.4338838682871092,
+      "grad_norm": 0.38179653882980347,
+      "learning_rate": 6.0310980954281234e-05,
+      "loss": 1.5742,
+      "step": 4177
+    },
+    {
+      "epoch": 0.4339877428066895,
+      "grad_norm": 0.4099988639354706,
+      "learning_rate": 6.029501454691318e-05,
+      "loss": 1.66,
+      "step": 4178
+    },
+    {
+      "epoch": 0.4340916173262699,
+      "grad_norm": 0.3874553442001343,
+      "learning_rate": 6.027904704320636e-05,
+      "loss": 1.6484,
+      "step": 4179
+    },
+    {
+      "epoch": 0.4341954918458502,
+      "grad_norm": 0.390216588973999,
+      "learning_rate": 6.026307844486123e-05,
+      "loss": 1.7898,
+      "step": 4180
+    },
+    {
+      "epoch": 0.43429936636543054,
+      "grad_norm": 0.38214191794395447,
+      "learning_rate": 6.024710875357831e-05,
+      "loss": 1.6147,
+      "step": 4181
+    },
+    {
+      "epoch": 0.4344032408850109,
+      "grad_norm": 0.37591472268104553,
+      "learning_rate": 6.02311379710582e-05,
+      "loss": 1.685,
+      "step": 4182
+    },
+    {
+      "epoch": 0.43450711540459125,
+      "grad_norm": 0.42642828822135925,
+      "learning_rate": 6.021516609900173e-05,
+      "loss": 1.8715,
+      "step": 4183
+    },
+    {
+      "epoch": 0.4346109899241716,
+      "grad_norm": 0.4037899672985077,
+      "learning_rate": 6.019919313910976e-05,
+      "loss": 1.6407,
+      "step": 4184
+    },
+    {
+      "epoch": 0.43471486444375196,
+      "grad_norm": 0.40876585245132446,
+      "learning_rate": 6.0183219093083244e-05,
+      "loss": 1.6688,
+      "step": 4185
+    },
+    {
+      "epoch": 0.4348187389633323,
+      "grad_norm": 0.4056430757045746,
+      "learning_rate": 6.016724396262335e-05,
+      "loss": 1.8047,
+      "step": 4186
+    },
+    {
+      "epoch": 0.43492261348291267,
+      "grad_norm": 0.4015723168849945,
+      "learning_rate": 6.015126774943128e-05,
+      "loss": 1.7502,
+      "step": 4187
+    },
+    {
+      "epoch": 0.435026488002493,
+      "grad_norm": 0.4126836359500885,
+      "learning_rate": 6.013529045520838e-05,
+      "loss": 1.8191,
+      "step": 4188
+    },
+    {
+      "epoch": 0.4351303625220733,
+      "grad_norm": 0.39362239837646484,
+      "learning_rate": 6.011931208165611e-05,
+      "loss": 1.6365,
+      "step": 4189
+    },
+    {
+      "epoch": 0.4352342370416537,
+      "grad_norm": 0.37851396203041077,
+      "learning_rate": 6.0103332630476017e-05,
+      "loss": 1.7451,
+      "step": 4190
+    },
+    {
+      "epoch": 0.43533811156123403,
+      "grad_norm": 0.3897380530834198,
+      "learning_rate": 6.008735210336982e-05,
+      "loss": 1.7397,
+      "step": 4191
+    },
+    {
+      "epoch": 0.43544198608081436,
+      "grad_norm": 0.38429588079452515,
+      "learning_rate": 6.00713705020393e-05,
+      "loss": 1.5615,
+      "step": 4192
+    },
+    {
+      "epoch": 0.43554586060039474,
+      "grad_norm": 0.4072874188423157,
+      "learning_rate": 6.005538782818638e-05,
+      "loss": 1.8237,
+      "step": 4193
+    },
+    {
+      "epoch": 0.43564973511997507,
+      "grad_norm": 0.4054878354072571,
+      "learning_rate": 6.003940408351311e-05,
+      "loss": 1.7932,
+      "step": 4194
+    },
+    {
+      "epoch": 0.4357536096395554,
+      "grad_norm": 0.38141578435897827,
+      "learning_rate": 6.0023419269721613e-05,
+      "loss": 1.6768,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4358574841591358,
+      "grad_norm": 0.4173438251018524,
+      "learning_rate": 6.000743338851413e-05,
+      "loss": 1.513,
+      "step": 4196
+    },
+    {
+      "epoch": 0.4359613586787161,
+      "grad_norm": 0.4424532651901245,
+      "learning_rate": 5.999144644159307e-05,
+      "loss": 1.7438,
+      "step": 4197
+    },
+    {
+      "epoch": 0.43606523319829643,
+      "grad_norm": 0.43783673644065857,
+      "learning_rate": 5.997545843066089e-05,
+      "loss": 1.7892,
+      "step": 4198
+    },
+    {
+      "epoch": 0.4361691077178768,
+      "grad_norm": 0.3705306649208069,
+      "learning_rate": 5.995946935742019e-05,
+      "loss": 1.6238,
+      "step": 4199
+    },
+    {
+      "epoch": 0.43627298223745714,
+      "grad_norm": 0.38308608531951904,
+      "learning_rate": 5.994347922357372e-05,
+      "loss": 1.5195,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4363768567570375,
+      "grad_norm": 0.43744367361068726,
+      "learning_rate": 5.992748803082425e-05,
+      "loss": 1.9181,
+      "step": 4201
+    },
+    {
+      "epoch": 0.43648073127661785,
+      "grad_norm": 0.42308345437049866,
+      "learning_rate": 5.991149578087476e-05,
+      "loss": 1.767,
+      "step": 4202
+    },
+    {
+      "epoch": 0.4365846057961982,
+      "grad_norm": 0.4409851133823395,
+      "learning_rate": 5.9895502475428265e-05,
+      "loss": 1.8537,
+      "step": 4203
+    },
+    {
+      "epoch": 0.43668848031577856,
+      "grad_norm": 0.377109169960022,
+      "learning_rate": 5.9879508116187947e-05,
+      "loss": 1.6446,
+      "step": 4204
+    },
+    {
+      "epoch": 0.4367923548353589,
+      "grad_norm": 0.4228600263595581,
+      "learning_rate": 5.9863512704857085e-05,
+      "loss": 1.7255,
+      "step": 4205
+    },
+    {
+      "epoch": 0.4368962293549392,
+      "grad_norm": 0.410178542137146,
+      "learning_rate": 5.9847516243139055e-05,
+      "loss": 1.6686,
+      "step": 4206
+    },
+    {
+      "epoch": 0.4370001038745196,
+      "grad_norm": 0.37821877002716064,
+      "learning_rate": 5.9831518732737344e-05,
+      "loss": 1.563,
+      "step": 4207
+    },
+    {
+      "epoch": 0.4371039783940999,
+      "grad_norm": 0.3783068358898163,
+      "learning_rate": 5.981552017535561e-05,
+      "loss": 1.6068,
+      "step": 4208
+    },
+    {
+      "epoch": 0.43720785291368025,
+      "grad_norm": 0.4072893261909485,
+      "learning_rate": 5.979952057269752e-05,
+      "loss": 1.7952,
+      "step": 4209
+    },
+    {
+      "epoch": 0.43731172743326063,
+      "grad_norm": 0.3989700376987457,
+      "learning_rate": 5.9783519926466924e-05,
+      "loss": 1.66,
+      "step": 4210
+    },
+    {
+      "epoch": 0.43741560195284096,
+      "grad_norm": 0.4692555367946625,
+      "learning_rate": 5.976751823836778e-05,
+      "loss": 1.6255,
+      "step": 4211
+    },
+    {
+      "epoch": 0.43751947647242134,
+      "grad_norm": 0.3792388439178467,
+      "learning_rate": 5.975151551010412e-05,
+      "loss": 1.5188,
+      "step": 4212
+    },
+    {
+      "epoch": 0.43762335099200167,
+      "grad_norm": 0.38823702931404114,
+      "learning_rate": 5.973551174338014e-05,
+      "loss": 1.6951,
+      "step": 4213
+    },
+    {
+      "epoch": 0.437727225511582,
+      "grad_norm": 0.37780869007110596,
+      "learning_rate": 5.9719506939900094e-05,
+      "loss": 1.5638,
+      "step": 4214
+    },
+    {
+      "epoch": 0.4378311000311624,
+      "grad_norm": 0.45048758387565613,
+      "learning_rate": 5.9703501101368374e-05,
+      "loss": 1.8339,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4379349745507427,
+      "grad_norm": 0.3920362889766693,
+      "learning_rate": 5.968749422948947e-05,
+      "loss": 1.723,
+      "step": 4216
+    },
+    {
+      "epoch": 0.43803884907032303,
+      "grad_norm": 0.3886549770832062,
+      "learning_rate": 5.967148632596801e-05,
+      "loss": 1.6502,
+      "step": 4217
+    },
+    {
+      "epoch": 0.4381427235899034,
+      "grad_norm": 0.4173612594604492,
+      "learning_rate": 5.965547739250867e-05,
+      "loss": 1.8315,
+      "step": 4218
+    },
+    {
+      "epoch": 0.43824659810948374,
+      "grad_norm": 0.45074644684791565,
+      "learning_rate": 5.963946743081633e-05,
+      "loss": 1.7273,
+      "step": 4219
+    },
+    {
+      "epoch": 0.43835047262906407,
+      "grad_norm": 0.42388424277305603,
+      "learning_rate": 5.962345644259589e-05,
+      "loss": 1.7344,
+      "step": 4220
+    },
+    {
+      "epoch": 0.43845434714864445,
+      "grad_norm": 0.44487425684928894,
+      "learning_rate": 5.9607444429552405e-05,
+      "loss": 1.771,
+      "step": 4221
+    },
+    {
+      "epoch": 0.4385582216682248,
+      "grad_norm": 0.4247962236404419,
+      "learning_rate": 5.959143139339104e-05,
+      "loss": 1.7933,
+      "step": 4222
+    },
+    {
+      "epoch": 0.4386620961878051,
+      "grad_norm": 0.4087913930416107,
+      "learning_rate": 5.957541733581704e-05,
+      "loss": 1.667,
+      "step": 4223
+    },
+    {
+      "epoch": 0.4387659707073855,
+      "grad_norm": 0.3898204267024994,
+      "learning_rate": 5.955940225853577e-05,
+      "loss": 1.6835,
+      "step": 4224
+    },
+    {
+      "epoch": 0.4388698452269658,
+      "grad_norm": 0.3906581997871399,
+      "learning_rate": 5.9543386163252744e-05,
+      "loss": 1.7532,
+      "step": 4225
+    },
+    {
+      "epoch": 0.4389737197465462,
+      "grad_norm": 0.37566760182380676,
+      "learning_rate": 5.952736905167352e-05,
+      "loss": 1.5813,
+      "step": 4226
+    },
+    {
+      "epoch": 0.4390775942661265,
+      "grad_norm": 0.42560824751853943,
+      "learning_rate": 5.951135092550382e-05,
+      "loss": 1.8791,
+      "step": 4227
+    },
+    {
+      "epoch": 0.43918146878570685,
+      "grad_norm": 0.4161139130592346,
+      "learning_rate": 5.949533178644943e-05,
+      "loss": 1.6862,
+      "step": 4228
+    },
+    {
+      "epoch": 0.43928534330528723,
+      "grad_norm": 0.398384690284729,
+      "learning_rate": 5.9479311636216274e-05,
+      "loss": 1.7341,
+      "step": 4229
+    },
+    {
+      "epoch": 0.43938921782486756,
+      "grad_norm": 0.36606839299201965,
+      "learning_rate": 5.946329047651037e-05,
+      "loss": 1.6987,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4394930923444479,
+      "grad_norm": 0.41310790181159973,
+      "learning_rate": 5.944726830903785e-05,
+      "loss": 1.7809,
+      "step": 4231
+    },
+    {
+      "epoch": 0.43959696686402827,
+      "grad_norm": 0.41169822216033936,
+      "learning_rate": 5.9431245135504934e-05,
+      "loss": 1.7338,
+      "step": 4232
+    },
+    {
+      "epoch": 0.4397008413836086,
+      "grad_norm": 0.3810160756111145,
+      "learning_rate": 5.941522095761799e-05,
+      "loss": 1.5948,
+      "step": 4233
+    },
+    {
+      "epoch": 0.4398047159031889,
+      "grad_norm": 0.3669174909591675,
+      "learning_rate": 5.939919577708346e-05,
+      "loss": 1.2893,
+      "step": 4234
+    },
+    {
+      "epoch": 0.4399085904227693,
+      "grad_norm": 0.4046458601951599,
+      "learning_rate": 5.938316959560788e-05,
+      "loss": 1.6906,
+      "step": 4235
+    },
+    {
+      "epoch": 0.44001246494234963,
+      "grad_norm": 0.40834760665893555,
+      "learning_rate": 5.9367142414897945e-05,
+      "loss": 1.7074,
+      "step": 4236
+    },
+    {
+      "epoch": 0.44011633946192996,
+      "grad_norm": 0.425662636756897,
+      "learning_rate": 5.935111423666041e-05,
+      "loss": 1.7486,
+      "step": 4237
+    },
+    {
+      "epoch": 0.44022021398151034,
+      "grad_norm": 0.37841665744781494,
+      "learning_rate": 5.933508506260214e-05,
+      "loss": 1.7472,
+      "step": 4238
+    },
+    {
+      "epoch": 0.44032408850109067,
+      "grad_norm": 0.4464881420135498,
+      "learning_rate": 5.9319054894430126e-05,
+      "loss": 1.8697,
+      "step": 4239
+    },
+    {
+      "epoch": 0.44042796302067105,
+      "grad_norm": 0.4150855541229248,
+      "learning_rate": 5.930302373385145e-05,
+      "loss": 1.5873,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4405318375402514,
+      "grad_norm": 0.38565129041671753,
+      "learning_rate": 5.928699158257333e-05,
+      "loss": 1.71,
+      "step": 4241
+    },
+    {
+      "epoch": 0.4406357120598317,
+      "grad_norm": 0.44895511865615845,
+      "learning_rate": 5.9270958442303035e-05,
+      "loss": 1.8128,
+      "step": 4242
+    },
+    {
+      "epoch": 0.4407395865794121,
+      "grad_norm": 0.3845813572406769,
+      "learning_rate": 5.9254924314748e-05,
+      "loss": 1.7892,
+      "step": 4243
+    },
+    {
+      "epoch": 0.4408434610989924,
+      "grad_norm": 0.4151462912559509,
+      "learning_rate": 5.923888920161571e-05,
+      "loss": 1.7573,
+      "step": 4244
+    },
+    {
+      "epoch": 0.44094733561857274,
+      "grad_norm": 0.38922443985939026,
+      "learning_rate": 5.922285310461378e-05,
+      "loss": 1.7848,
+      "step": 4245
+    },
+    {
+      "epoch": 0.4410512101381531,
+      "grad_norm": 0.38882389664649963,
+      "learning_rate": 5.920681602544993e-05,
+      "loss": 1.7333,
+      "step": 4246
+    },
+    {
+      "epoch": 0.44115508465773345,
+      "grad_norm": 0.38732343912124634,
+      "learning_rate": 5.9190777965832e-05,
+      "loss": 1.6265,
+      "step": 4247
+    },
+    {
+      "epoch": 0.4412589591773138,
+      "grad_norm": 0.3835653066635132,
+      "learning_rate": 5.917473892746791e-05,
+      "loss": 1.5993,
+      "step": 4248
+    },
+    {
+      "epoch": 0.44136283369689416,
+      "grad_norm": 0.41864073276519775,
+      "learning_rate": 5.915869891206567e-05,
+      "loss": 1.7911,
+      "step": 4249
+    },
+    {
+      "epoch": 0.4414667082164745,
+      "grad_norm": 0.43092262744903564,
+      "learning_rate": 5.9142657921333466e-05,
+      "loss": 1.487,
+      "step": 4250
+    },
+    {
+      "epoch": 0.4415705827360549,
+      "grad_norm": 0.3756438195705414,
+      "learning_rate": 5.912661595697948e-05,
+      "loss": 1.5246,
+      "step": 4251
+    },
+    {
+      "epoch": 0.4416744572556352,
+      "grad_norm": 0.3791937828063965,
+      "learning_rate": 5.91105730207121e-05,
+      "loss": 1.7199,
+      "step": 4252
+    },
+    {
+      "epoch": 0.4417783317752155,
+      "grad_norm": 0.381509393453598,
+      "learning_rate": 5.909452911423976e-05,
+      "loss": 1.6264,
+      "step": 4253
+    },
+    {
+      "epoch": 0.4418822062947959,
+      "grad_norm": 0.3744488060474396,
+      "learning_rate": 5.9078484239271e-05,
+      "loss": 1.5363,
+      "step": 4254
+    },
+    {
+      "epoch": 0.44198608081437624,
+      "grad_norm": 0.3968122601509094,
+      "learning_rate": 5.9062438397514496e-05,
+      "loss": 1.7037,
+      "step": 4255
+    },
+    {
+      "epoch": 0.44208995533395656,
+      "grad_norm": 0.4114841818809509,
+      "learning_rate": 5.904639159067898e-05,
+      "loss": 1.7508,
+      "step": 4256
+    },
+    {
+      "epoch": 0.44219382985353695,
+      "grad_norm": 0.37198910117149353,
+      "learning_rate": 5.9030343820473335e-05,
+      "loss": 1.7378,
+      "step": 4257
+    },
+    {
+      "epoch": 0.4422977043731173,
+      "grad_norm": 0.47801533341407776,
+      "learning_rate": 5.9014295088606505e-05,
+      "loss": 1.838,
+      "step": 4258
+    },
+    {
+      "epoch": 0.4424015788926976,
+      "grad_norm": 0.43482860922813416,
+      "learning_rate": 5.899824539678758e-05,
+      "loss": 1.8126,
+      "step": 4259
+    },
+    {
+      "epoch": 0.442505453412278,
+      "grad_norm": 0.3933789134025574,
+      "learning_rate": 5.898219474672568e-05,
+      "loss": 1.5948,
+      "step": 4260
+    },
+    {
+      "epoch": 0.4426093279318583,
+      "grad_norm": 0.369711697101593,
+      "learning_rate": 5.896614314013012e-05,
+      "loss": 1.5256,
+      "step": 4261
+    },
+    {
+      "epoch": 0.44271320245143864,
+      "grad_norm": 0.43509915471076965,
+      "learning_rate": 5.8950090578710246e-05,
+      "loss": 1.762,
+      "step": 4262
+    },
+    {
+      "epoch": 0.442817076971019,
+      "grad_norm": 0.4086291193962097,
+      "learning_rate": 5.893403706417554e-05,
+      "loss": 1.8018,
+      "step": 4263
+    },
+    {
+      "epoch": 0.44292095149059935,
+      "grad_norm": 0.38804736733436584,
+      "learning_rate": 5.89179825982356e-05,
+      "loss": 1.717,
+      "step": 4264
+    },
+    {
+      "epoch": 0.44302482601017973,
+      "grad_norm": 0.4058080017566681,
+      "learning_rate": 5.8901927182600035e-05,
+      "loss": 1.8017,
+      "step": 4265
+    },
+    {
+      "epoch": 0.44312870052976006,
+      "grad_norm": 0.381388783454895,
+      "learning_rate": 5.88858708189787e-05,
+      "loss": 1.8363,
+      "step": 4266
+    },
+    {
+      "epoch": 0.4432325750493404,
+      "grad_norm": 0.3509836196899414,
+      "learning_rate": 5.886981350908142e-05,
+      "loss": 1.5213,
+      "step": 4267
+    },
+    {
+      "epoch": 0.44333644956892077,
+      "grad_norm": 0.39383694529533386,
+      "learning_rate": 5.885375525461817e-05,
+      "loss": 1.6166,
+      "step": 4268
+    },
+    {
+      "epoch": 0.4434403240885011,
+      "grad_norm": 0.4015539586544037,
+      "learning_rate": 5.883769605729907e-05,
+      "loss": 1.7148,
+      "step": 4269
+    },
+    {
+      "epoch": 0.4435441986080814,
+      "grad_norm": 0.3899979293346405,
+      "learning_rate": 5.882163591883427e-05,
+      "loss": 1.5109,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4436480731276618,
+      "grad_norm": 0.41530346870422363,
+      "learning_rate": 5.8805574840934067e-05,
+      "loss": 1.6773,
+      "step": 4271
+    },
+    {
+      "epoch": 0.44375194764724213,
+      "grad_norm": 0.4475281536579132,
+      "learning_rate": 5.878951282530881e-05,
+      "loss": 1.6551,
+      "step": 4272
+    },
+    {
+      "epoch": 0.44385582216682246,
+      "grad_norm": 0.442122220993042,
+      "learning_rate": 5.877344987366902e-05,
+      "loss": 1.7879,
+      "step": 4273
+    },
+    {
+      "epoch": 0.44395969668640284,
+      "grad_norm": 0.39350566267967224,
+      "learning_rate": 5.875738598772522e-05,
+      "loss": 1.705,
+      "step": 4274
+    },
+    {
+      "epoch": 0.44406357120598317,
+      "grad_norm": 0.43882155418395996,
+      "learning_rate": 5.874132116918816e-05,
+      "loss": 1.7977,
+      "step": 4275
+    },
+    {
+      "epoch": 0.4441674457255635,
+      "grad_norm": 0.42583081126213074,
+      "learning_rate": 5.8725255419768565e-05,
+      "loss": 1.6302,
+      "step": 4276
+    },
+    {
+      "epoch": 0.4442713202451439,
+      "grad_norm": 0.40676143765449524,
+      "learning_rate": 5.870918874117731e-05,
+      "loss": 1.8449,
+      "step": 4277
+    },
+    {
+      "epoch": 0.4443751947647242,
+      "grad_norm": 0.42294740676879883,
+      "learning_rate": 5.869312113512542e-05,
+      "loss": 1.7231,
+      "step": 4278
+    },
+    {
+      "epoch": 0.4444790692843046,
+      "grad_norm": 0.41314196586608887,
+      "learning_rate": 5.867705260332391e-05,
+      "loss": 1.6797,
+      "step": 4279
+    },
+    {
+      "epoch": 0.4445829438038849,
+      "grad_norm": 0.395781934261322,
+      "learning_rate": 5.866098314748401e-05,
+      "loss": 1.7341,
+      "step": 4280
+    },
+    {
+      "epoch": 0.44468681832346524,
+      "grad_norm": 0.412775456905365,
+      "learning_rate": 5.864491276931694e-05,
+      "loss": 1.685,
+      "step": 4281
+    },
+    {
+      "epoch": 0.4447906928430456,
+      "grad_norm": 0.38872653245925903,
+      "learning_rate": 5.862884147053409e-05,
+      "loss": 1.7268,
+      "step": 4282
+    },
+    {
+      "epoch": 0.44489456736262595,
+      "grad_norm": 0.36840468645095825,
+      "learning_rate": 5.861276925284694e-05,
+      "loss": 1.5863,
+      "step": 4283
+    },
+    {
+      "epoch": 0.4449984418822063,
+      "grad_norm": 0.37796318531036377,
+      "learning_rate": 5.8596696117967054e-05,
+      "loss": 1.6242,
+      "step": 4284
+    },
+    {
+      "epoch": 0.44510231640178666,
+      "grad_norm": 0.3686732351779938,
+      "learning_rate": 5.8580622067606083e-05,
+      "loss": 1.5881,
+      "step": 4285
+    },
+    {
+      "epoch": 0.445206190921367,
+      "grad_norm": 0.4148024320602417,
+      "learning_rate": 5.8564547103475795e-05,
+      "loss": 1.815,
+      "step": 4286
+    },
+    {
+      "epoch": 0.4453100654409473,
+      "grad_norm": 0.39075860381126404,
+      "learning_rate": 5.854847122728805e-05,
+      "loss": 1.6837,
+      "step": 4287
+    },
+    {
+      "epoch": 0.4454139399605277,
+      "grad_norm": 0.40239056944847107,
+      "learning_rate": 5.853239444075479e-05,
+      "loss": 1.7351,
+      "step": 4288
+    },
+    {
+      "epoch": 0.445517814480108,
+      "grad_norm": 0.4210759401321411,
+      "learning_rate": 5.8516316745588085e-05,
+      "loss": 1.6621,
+      "step": 4289
+    },
+    {
+      "epoch": 0.4456216889996884,
+      "grad_norm": 0.39281001687049866,
+      "learning_rate": 5.850023814350006e-05,
+      "loss": 1.7465,
+      "step": 4290
+    },
+    {
+      "epoch": 0.44572556351926873,
+      "grad_norm": 0.4055309295654297,
+      "learning_rate": 5.848415863620299e-05,
+      "loss": 1.6812,
+      "step": 4291
+    },
+    {
+      "epoch": 0.44582943803884906,
+      "grad_norm": 0.3883981704711914,
+      "learning_rate": 5.846807822540922e-05,
+      "loss": 1.6243,
+      "step": 4292
+    },
+    {
+      "epoch": 0.44593331255842944,
+      "grad_norm": 0.394539475440979,
+      "learning_rate": 5.8451996912831156e-05,
+      "loss": 1.7954,
+      "step": 4293
+    },
+    {
+      "epoch": 0.44603718707800977,
+      "grad_norm": 0.40347257256507874,
+      "learning_rate": 5.8435914700181347e-05,
+      "loss": 1.6049,
+      "step": 4294
+    },
+    {
+      "epoch": 0.4461410615975901,
+      "grad_norm": 0.47316792607307434,
+      "learning_rate": 5.8419831589172426e-05,
+      "loss": 1.9193,
+      "step": 4295
+    },
+    {
+      "epoch": 0.4462449361171705,
+      "grad_norm": 0.427937775850296,
+      "learning_rate": 5.840374758151711e-05,
+      "loss": 1.6742,
+      "step": 4296
+    },
+    {
+      "epoch": 0.4463488106367508,
+      "grad_norm": 0.44503235816955566,
+      "learning_rate": 5.838766267892825e-05,
+      "loss": 1.9509,
+      "step": 4297
+    },
+    {
+      "epoch": 0.44645268515633113,
+      "grad_norm": 0.4180712103843689,
+      "learning_rate": 5.837157688311873e-05,
+      "loss": 1.8231,
+      "step": 4298
+    },
+    {
+      "epoch": 0.4465565596759115,
+      "grad_norm": 0.4312383532524109,
+      "learning_rate": 5.8355490195801566e-05,
+      "loss": 1.7526,
+      "step": 4299
+    },
+    {
+      "epoch": 0.44666043419549184,
+      "grad_norm": 0.38129255175590515,
+      "learning_rate": 5.8339402618689885e-05,
+      "loss": 1.749,
+      "step": 4300
+    },
+    {
+      "epoch": 0.44676430871507217,
+      "grad_norm": 0.47581759095191956,
+      "learning_rate": 5.832331415349687e-05,
+      "loss": 2.0137,
+      "step": 4301
+    },
+    {
+      "epoch": 0.44686818323465255,
+      "grad_norm": 0.3956213593482971,
+      "learning_rate": 5.83072248019358e-05,
+      "loss": 1.7437,
+      "step": 4302
+    },
+    {
+      "epoch": 0.4469720577542329,
+      "grad_norm": 0.42220327258110046,
+      "learning_rate": 5.829113456572012e-05,
+      "loss": 1.9,
+      "step": 4303
+    },
+    {
+      "epoch": 0.44707593227381326,
+      "grad_norm": 0.39744284749031067,
+      "learning_rate": 5.827504344656326e-05,
+      "loss": 1.5549,
+      "step": 4304
+    },
+    {
+      "epoch": 0.4471798067933936,
+      "grad_norm": 0.3954363763332367,
+      "learning_rate": 5.825895144617883e-05,
+      "loss": 1.7203,
+      "step": 4305
+    },
+    {
+      "epoch": 0.4472836813129739,
+      "grad_norm": 0.39236709475517273,
+      "learning_rate": 5.82428585662805e-05,
+      "loss": 1.7229,
+      "step": 4306
+    },
+    {
+      "epoch": 0.4473875558325543,
+      "grad_norm": 0.3697808086872101,
+      "learning_rate": 5.822676480858201e-05,
+      "loss": 1.6764,
+      "step": 4307
+    },
+    {
+      "epoch": 0.4474914303521346,
+      "grad_norm": 0.45007967948913574,
+      "learning_rate": 5.821067017479727e-05,
+      "loss": 1.8381,
+      "step": 4308
+    },
+    {
+      "epoch": 0.44759530487171495,
+      "grad_norm": 0.49184536933898926,
+      "learning_rate": 5.819457466664018e-05,
+      "loss": 2.1298,
+      "step": 4309
+    },
+    {
+      "epoch": 0.44769917939129533,
+      "grad_norm": 0.3896716833114624,
+      "learning_rate": 5.81784782858248e-05,
+      "loss": 1.7621,
+      "step": 4310
+    },
+    {
+      "epoch": 0.44780305391087566,
+      "grad_norm": 0.39417698979377747,
+      "learning_rate": 5.816238103406529e-05,
+      "loss": 1.7275,
+      "step": 4311
+    },
+    {
+      "epoch": 0.447906928430456,
+      "grad_norm": 0.45076748728752136,
+      "learning_rate": 5.814628291307587e-05,
+      "loss": 1.7021,
+      "step": 4312
+    },
+    {
+      "epoch": 0.44801080295003637,
+      "grad_norm": 0.43015792965888977,
+      "learning_rate": 5.813018392457086e-05,
+      "loss": 1.8654,
+      "step": 4313
+    },
+    {
+      "epoch": 0.4481146774696167,
+      "grad_norm": 0.3831145167350769,
+      "learning_rate": 5.8114084070264686e-05,
+      "loss": 1.7,
+      "step": 4314
+    },
+    {
+      "epoch": 0.448218551989197,
+      "grad_norm": 0.3929171860218048,
+      "learning_rate": 5.8097983351871844e-05,
+      "loss": 1.611,
+      "step": 4315
+    },
+    {
+      "epoch": 0.4483224265087774,
+      "grad_norm": 0.41042855381965637,
+      "learning_rate": 5.808188177110694e-05,
+      "loss": 2.0493,
+      "step": 4316
+    },
+    {
+      "epoch": 0.44842630102835773,
+      "grad_norm": 0.4465937316417694,
+      "learning_rate": 5.806577932968467e-05,
+      "loss": 1.8463,
+      "step": 4317
+    },
+    {
+      "epoch": 0.4485301755479381,
+      "grad_norm": 0.4185175597667694,
+      "learning_rate": 5.804967602931981e-05,
+      "loss": 1.649,
+      "step": 4318
+    },
+    {
+      "epoch": 0.44863405006751844,
+      "grad_norm": 0.413787841796875,
+      "learning_rate": 5.803357187172726e-05,
+      "loss": 1.8159,
+      "step": 4319
+    },
+    {
+      "epoch": 0.44873792458709877,
+      "grad_norm": 0.4084557890892029,
+      "learning_rate": 5.801746685862197e-05,
+      "loss": 1.6689,
+      "step": 4320
+    },
+    {
+      "epoch": 0.44884179910667915,
+      "grad_norm": 0.4004693329334259,
+      "learning_rate": 5.8001360991718976e-05,
+      "loss": 1.5459,
+      "step": 4321
+    },
+    {
+      "epoch": 0.4489456736262595,
+      "grad_norm": 0.40109360218048096,
+      "learning_rate": 5.798525427273347e-05,
+      "loss": 1.6249,
+      "step": 4322
+    },
+    {
+      "epoch": 0.4490495481458398,
+      "grad_norm": 0.4452827274799347,
+      "learning_rate": 5.796914670338067e-05,
+      "loss": 1.7092,
+      "step": 4323
+    },
+    {
+      "epoch": 0.4491534226654202,
+      "grad_norm": 0.4333063066005707,
+      "learning_rate": 5.795303828537589e-05,
+      "loss": 1.8151,
+      "step": 4324
+    },
+    {
+      "epoch": 0.4492572971850005,
+      "grad_norm": 0.4180525243282318,
+      "learning_rate": 5.7936929020434585e-05,
+      "loss": 1.7309,
+      "step": 4325
+    },
+    {
+      "epoch": 0.44936117170458084,
+      "grad_norm": 0.40769898891448975,
+      "learning_rate": 5.792081891027224e-05,
+      "loss": 1.7786,
+      "step": 4326
+    },
+    {
+      "epoch": 0.4494650462241612,
+      "grad_norm": 0.46367326378822327,
+      "learning_rate": 5.790470795660447e-05,
+      "loss": 1.9491,
+      "step": 4327
+    },
+    {
+      "epoch": 0.44956892074374155,
+      "grad_norm": 0.40147140622138977,
+      "learning_rate": 5.788859616114697e-05,
+      "loss": 1.7364,
+      "step": 4328
+    },
+    {
+      "epoch": 0.44967279526332193,
+      "grad_norm": 0.4153495132923126,
+      "learning_rate": 5.787248352561549e-05,
+      "loss": 1.7383,
+      "step": 4329
+    },
+    {
+      "epoch": 0.44977666978290226,
+      "grad_norm": 0.40633097290992737,
+      "learning_rate": 5.785637005172595e-05,
+      "loss": 1.4243,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4498805443024826,
+      "grad_norm": 0.3852026164531708,
+      "learning_rate": 5.784025574119427e-05,
+      "loss": 1.6868,
+      "step": 4331
+    },
+    {
+      "epoch": 0.44998441882206297,
+      "grad_norm": 0.3848346769809723,
+      "learning_rate": 5.78241405957365e-05,
+      "loss": 1.5136,
+      "step": 4332
+    },
+    {
+      "epoch": 0.4500882933416433,
+      "grad_norm": 0.3910854160785675,
+      "learning_rate": 5.78080246170688e-05,
+      "loss": 1.6864,
+      "step": 4333
+    },
+    {
+      "epoch": 0.4501921678612236,
+      "grad_norm": 0.3685462176799774,
+      "learning_rate": 5.7791907806907395e-05,
+      "loss": 1.6707,
+      "step": 4334
+    },
+    {
+      "epoch": 0.450296042380804,
+      "grad_norm": 0.4090143144130707,
+      "learning_rate": 5.777579016696856e-05,
+      "loss": 1.5942,
+      "step": 4335
+    },
+    {
+      "epoch": 0.45039991690038433,
+      "grad_norm": 0.377214252948761,
+      "learning_rate": 5.7759671698968745e-05,
+      "loss": 1.5599,
+      "step": 4336
+    },
+    {
+      "epoch": 0.45050379141996466,
+      "grad_norm": 0.41875123977661133,
+      "learning_rate": 5.774355240462441e-05,
+      "loss": 1.7982,
+      "step": 4337
+    },
+    {
+      "epoch": 0.45060766593954504,
+      "grad_norm": 0.3935871422290802,
+      "learning_rate": 5.772743228565215e-05,
+      "loss": 1.7028,
+      "step": 4338
+    },
+    {
+      "epoch": 0.45071154045912537,
+      "grad_norm": 0.406841903924942,
+      "learning_rate": 5.771131134376863e-05,
+      "loss": 1.7354,
+      "step": 4339
+    },
+    {
+      "epoch": 0.4508154149787057,
+      "grad_norm": 0.37955769896507263,
+      "learning_rate": 5.76951895806906e-05,
+      "loss": 1.647,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4509192894982861,
+      "grad_norm": 0.42771968245506287,
+      "learning_rate": 5.76790669981349e-05,
+      "loss": 1.825,
+      "step": 4341
+    },
+    {
+      "epoch": 0.4510231640178664,
+      "grad_norm": 0.3990226686000824,
+      "learning_rate": 5.766294359781848e-05,
+      "loss": 1.611,
+      "step": 4342
+    },
+    {
+      "epoch": 0.4511270385374468,
+      "grad_norm": 0.4627864360809326,
+      "learning_rate": 5.764681938145832e-05,
+      "loss": 1.7403,
+      "step": 4343
+    },
+    {
+      "epoch": 0.4512309130570271,
+      "grad_norm": 0.3694743812084198,
+      "learning_rate": 5.763069435077155e-05,
+      "loss": 1.5556,
+      "step": 4344
+    },
+    {
+      "epoch": 0.45133478757660744,
+      "grad_norm": 0.37064674496650696,
+      "learning_rate": 5.761456850747536e-05,
+      "loss": 1.6155,
+      "step": 4345
+    },
+    {
+      "epoch": 0.4514386620961878,
+      "grad_norm": 0.37530389428138733,
+      "learning_rate": 5.759844185328701e-05,
+      "loss": 1.462,
+      "step": 4346
+    },
+    {
+      "epoch": 0.45154253661576815,
+      "grad_norm": 0.4151467978954315,
+      "learning_rate": 5.7582314389923876e-05,
+      "loss": 1.6177,
+      "step": 4347
+    },
+    {
+      "epoch": 0.4516464111353485,
+      "grad_norm": 0.4127391278743744,
+      "learning_rate": 5.756618611910343e-05,
+      "loss": 1.782,
+      "step": 4348
+    },
+    {
+      "epoch": 0.45175028565492886,
+      "grad_norm": 0.41236239671707153,
+      "learning_rate": 5.7550057042543137e-05,
+      "loss": 1.7076,
+      "step": 4349
+    },
+    {
+      "epoch": 0.4518541601745092,
+      "grad_norm": 0.3952932357788086,
+      "learning_rate": 5.753392716196069e-05,
+      "loss": 1.6074,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4519580346940895,
+      "grad_norm": 0.5258936285972595,
+      "learning_rate": 5.751779647907376e-05,
+      "loss": 1.8871,
+      "step": 4351
+    },
+    {
+      "epoch": 0.4520619092136699,
+      "grad_norm": 0.3987799882888794,
+      "learning_rate": 5.7501664995600134e-05,
+      "loss": 1.7228,
+      "step": 4352
+    },
+    {
+      "epoch": 0.4521657837332502,
+      "grad_norm": 0.3863909840583801,
+      "learning_rate": 5.748553271325772e-05,
+      "loss": 1.5237,
+      "step": 4353
+    },
+    {
+      "epoch": 0.45226965825283055,
+      "grad_norm": 0.40286675095558167,
+      "learning_rate": 5.746939963376445e-05,
+      "loss": 1.6773,
+      "step": 4354
+    },
+    {
+      "epoch": 0.45237353277241094,
+      "grad_norm": 0.4835149645805359,
+      "learning_rate": 5.745326575883839e-05,
+      "loss": 1.7949,
+      "step": 4355
+    },
+    {
+      "epoch": 0.45247740729199126,
+      "grad_norm": 0.3882110118865967,
+      "learning_rate": 5.743713109019766e-05,
+      "loss": 1.6454,
+      "step": 4356
+    },
+    {
+      "epoch": 0.45258128181157165,
+      "grad_norm": 0.4016772508621216,
+      "learning_rate": 5.742099562956048e-05,
+      "loss": 1.6265,
+      "step": 4357
+    },
+    {
+      "epoch": 0.45268515633115197,
+      "grad_norm": 0.4201604425907135,
+      "learning_rate": 5.740485937864515e-05,
+      "loss": 1.6872,
+      "step": 4358
+    },
+    {
+      "epoch": 0.4527890308507323,
+      "grad_norm": 0.45017364621162415,
+      "learning_rate": 5.7388722339170066e-05,
+      "loss": 1.8138,
+      "step": 4359
+    },
+    {
+      "epoch": 0.4528929053703127,
+      "grad_norm": 0.39483213424682617,
+      "learning_rate": 5.7372584512853665e-05,
+      "loss": 1.5788,
+      "step": 4360
+    },
+    {
+      "epoch": 0.452996779889893,
+      "grad_norm": 0.4778917133808136,
+      "learning_rate": 5.7356445901414545e-05,
+      "loss": 1.8203,
+      "step": 4361
+    },
+    {
+      "epoch": 0.45310065440947334,
+      "grad_norm": 0.3956323564052582,
+      "learning_rate": 5.734030650657132e-05,
+      "loss": 1.6614,
+      "step": 4362
+    },
+    {
+      "epoch": 0.4532045289290537,
+      "grad_norm": 0.42811745405197144,
+      "learning_rate": 5.73241663300427e-05,
+      "loss": 1.851,
+      "step": 4363
+    },
+    {
+      "epoch": 0.45330840344863405,
+      "grad_norm": 0.39979517459869385,
+      "learning_rate": 5.730802537354749e-05,
+      "loss": 1.6255,
+      "step": 4364
+    },
+    {
+      "epoch": 0.45341227796821437,
+      "grad_norm": 0.4086802899837494,
+      "learning_rate": 5.729188363880459e-05,
+      "loss": 1.4243,
+      "step": 4365
+    },
+    {
+      "epoch": 0.45351615248779475,
+      "grad_norm": 0.43692517280578613,
+      "learning_rate": 5.727574112753296e-05,
+      "loss": 1.8354,
+      "step": 4366
+    },
+    {
+      "epoch": 0.4536200270073751,
+      "grad_norm": 0.37770721316337585,
+      "learning_rate": 5.725959784145165e-05,
+      "loss": 1.5944,
+      "step": 4367
+    },
+    {
+      "epoch": 0.45372390152695546,
+      "grad_norm": 0.4468797445297241,
+      "learning_rate": 5.72434537822798e-05,
+      "loss": 1.7066,
+      "step": 4368
+    },
+    {
+      "epoch": 0.4538277760465358,
+      "grad_norm": 0.4045311510562897,
+      "learning_rate": 5.722730895173662e-05,
+      "loss": 1.7554,
+      "step": 4369
+    },
+    {
+      "epoch": 0.4539316505661161,
+      "grad_norm": 0.40388810634613037,
+      "learning_rate": 5.721116335154142e-05,
+      "loss": 1.4915,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4540355250856965,
+      "grad_norm": 0.4059275686740875,
+      "learning_rate": 5.719501698341356e-05,
+      "loss": 1.6267,
+      "step": 4371
+    },
+    {
+      "epoch": 0.45413939960527683,
+      "grad_norm": 0.3910108804702759,
+      "learning_rate": 5.7178869849072526e-05,
+      "loss": 1.6264,
+      "step": 4372
+    },
+    {
+      "epoch": 0.45424327412485715,
+      "grad_norm": 0.445516437292099,
+      "learning_rate": 5.716272195023785e-05,
+      "loss": 1.7784,
+      "step": 4373
+    },
+    {
+      "epoch": 0.45434714864443754,
+      "grad_norm": 0.4089578688144684,
+      "learning_rate": 5.7146573288629145e-05,
+      "loss": 1.7885,
+      "step": 4374
+    },
+    {
+      "epoch": 0.45445102316401786,
+      "grad_norm": 0.4145938754081726,
+      "learning_rate": 5.713042386596614e-05,
+      "loss": 1.6938,
+      "step": 4375
+    },
+    {
+      "epoch": 0.4545548976835982,
+      "grad_norm": 0.38517993688583374,
+      "learning_rate": 5.7114273683968625e-05,
+      "loss": 1.5404,
+      "step": 4376
+    },
+    {
+      "epoch": 0.4546587722031786,
+      "grad_norm": 0.40502604842185974,
+      "learning_rate": 5.709812274435643e-05,
+      "loss": 1.6948,
+      "step": 4377
+    },
+    {
+      "epoch": 0.4547626467227589,
+      "grad_norm": 0.3674617409706116,
+      "learning_rate": 5.708197104884955e-05,
+      "loss": 1.6687,
+      "step": 4378
+    },
+    {
+      "epoch": 0.45486652124233923,
+      "grad_norm": 0.38975459337234497,
+      "learning_rate": 5.706581859916799e-05,
+      "loss": 1.5973,
+      "step": 4379
+    },
+    {
+      "epoch": 0.4549703957619196,
+      "grad_norm": 0.3994758725166321,
+      "learning_rate": 5.704966539703185e-05,
+      "loss": 1.6266,
+      "step": 4380
+    },
+    {
+      "epoch": 0.45507427028149994,
+      "grad_norm": 0.39510855078697205,
+      "learning_rate": 5.7033511444161355e-05,
+      "loss": 1.7119,
+      "step": 4381
+    },
+    {
+      "epoch": 0.4551781448010803,
+      "grad_norm": 0.4508286416530609,
+      "learning_rate": 5.701735674227675e-05,
+      "loss": 1.9857,
+      "step": 4382
+    },
+    {
+      "epoch": 0.45528201932066065,
+      "grad_norm": 0.4357827603816986,
+      "learning_rate": 5.7001201293098386e-05,
+      "loss": 1.9188,
+      "step": 4383
+    },
+    {
+      "epoch": 0.455385893840241,
+      "grad_norm": 0.4613460898399353,
+      "learning_rate": 5.69850450983467e-05,
+      "loss": 1.7152,
+      "step": 4384
+    },
+    {
+      "epoch": 0.45548976835982136,
+      "grad_norm": 0.43691280484199524,
+      "learning_rate": 5.6968888159742184e-05,
+      "loss": 1.7644,
+      "step": 4385
+    },
+    {
+      "epoch": 0.4555936428794017,
+      "grad_norm": 0.40065327286720276,
+      "learning_rate": 5.695273047900544e-05,
+      "loss": 1.6811,
+      "step": 4386
+    },
+    {
+      "epoch": 0.455697517398982,
+      "grad_norm": 0.4466818869113922,
+      "learning_rate": 5.693657205785715e-05,
+      "loss": 1.8516,
+      "step": 4387
+    },
+    {
+      "epoch": 0.4558013919185624,
+      "grad_norm": 0.4043349623680115,
+      "learning_rate": 5.6920412898018026e-05,
+      "loss": 1.8134,
+      "step": 4388
+    },
+    {
+      "epoch": 0.4559052664381427,
+      "grad_norm": 0.42350390553474426,
+      "learning_rate": 5.690425300120893e-05,
+      "loss": 1.8296,
+      "step": 4389
+    },
+    {
+      "epoch": 0.45600914095772305,
+      "grad_norm": 0.38826295733451843,
+      "learning_rate": 5.6888092369150734e-05,
+      "loss": 1.7926,
+      "step": 4390
+    },
+    {
+      "epoch": 0.45611301547730343,
+      "grad_norm": 0.39671480655670166,
+      "learning_rate": 5.6871931003564414e-05,
+      "loss": 1.8088,
+      "step": 4391
+    },
+    {
+      "epoch": 0.45621688999688376,
+      "grad_norm": 0.38095131516456604,
+      "learning_rate": 5.685576890617107e-05,
+      "loss": 1.5183,
+      "step": 4392
+    },
+    {
+      "epoch": 0.45632076451646414,
+      "grad_norm": 0.37252774834632874,
+      "learning_rate": 5.683960607869179e-05,
+      "loss": 1.6363,
+      "step": 4393
+    },
+    {
+      "epoch": 0.45642463903604447,
+      "grad_norm": 0.43346065282821655,
+      "learning_rate": 5.6823442522847835e-05,
+      "loss": 1.5937,
+      "step": 4394
+    },
+    {
+      "epoch": 0.4565285135556248,
+      "grad_norm": 0.3876819610595703,
+      "learning_rate": 5.680727824036046e-05,
+      "loss": 1.723,
+      "step": 4395
+    },
+    {
+      "epoch": 0.4566323880752052,
+      "grad_norm": 0.3917519152164459,
+      "learning_rate": 5.6791113232951063e-05,
+      "loss": 1.5769,
+      "step": 4396
+    },
+    {
+      "epoch": 0.4567362625947855,
+      "grad_norm": 0.4223952889442444,
+      "learning_rate": 5.677494750234108e-05,
+      "loss": 1.7898,
+      "step": 4397
+    },
+    {
+      "epoch": 0.45684013711436583,
+      "grad_norm": 0.39051714539527893,
+      "learning_rate": 5.675878105025203e-05,
+      "loss": 1.7079,
+      "step": 4398
+    },
+    {
+      "epoch": 0.4569440116339462,
+      "grad_norm": 0.3843806982040405,
+      "learning_rate": 5.674261387840551e-05,
+      "loss": 1.5216,
+      "step": 4399
+    },
+    {
+      "epoch": 0.45704788615352654,
+      "grad_norm": 0.4141395092010498,
+      "learning_rate": 5.6726445988523224e-05,
+      "loss": 1.6415,
+      "step": 4400
+    },
+    {
+      "epoch": 0.45715176067310687,
+      "grad_norm": 0.4137116074562073,
+      "learning_rate": 5.67102773823269e-05,
+      "loss": 1.7626,
+      "step": 4401
+    },
+    {
+      "epoch": 0.45725563519268725,
+      "grad_norm": 0.3695942461490631,
+      "learning_rate": 5.669410806153838e-05,
+      "loss": 1.6487,
+      "step": 4402
+    },
+    {
+      "epoch": 0.4573595097122676,
+      "grad_norm": 0.3959349989891052,
+      "learning_rate": 5.667793802787957e-05,
+      "loss": 1.7431,
+      "step": 4403
+    },
+    {
+      "epoch": 0.4574633842318479,
+      "grad_norm": 0.39805370569229126,
+      "learning_rate": 5.6661767283072444e-05,
+      "loss": 1.6595,
+      "step": 4404
+    },
+    {
+      "epoch": 0.4575672587514283,
+      "grad_norm": 0.38343602418899536,
+      "learning_rate": 5.664559582883906e-05,
+      "loss": 1.8235,
+      "step": 4405
+    },
+    {
+      "epoch": 0.4576711332710086,
+      "grad_norm": 0.4133247435092926,
+      "learning_rate": 5.662942366690157e-05,
+      "loss": 1.7427,
+      "step": 4406
+    },
+    {
+      "epoch": 0.457775007790589,
+      "grad_norm": 0.44640305638313293,
+      "learning_rate": 5.6613250798982156e-05,
+      "loss": 1.7745,
+      "step": 4407
+    },
+    {
+      "epoch": 0.4578788823101693,
+      "grad_norm": 0.4178631603717804,
+      "learning_rate": 5.659707722680313e-05,
+      "loss": 1.7777,
+      "step": 4408
+    },
+    {
+      "epoch": 0.45798275682974965,
+      "grad_norm": 0.44724735617637634,
+      "learning_rate": 5.6580902952086836e-05,
+      "loss": 1.5845,
+      "step": 4409
+    },
+    {
+      "epoch": 0.45808663134933003,
+      "grad_norm": 0.39849719405174255,
+      "learning_rate": 5.656472797655571e-05,
+      "loss": 1.5669,
+      "step": 4410
+    },
+    {
+      "epoch": 0.45819050586891036,
+      "grad_norm": 0.3746553957462311,
+      "learning_rate": 5.6548552301932265e-05,
+      "loss": 1.5851,
+      "step": 4411
+    },
+    {
+      "epoch": 0.4582943803884907,
+      "grad_norm": 0.4282001852989197,
+      "learning_rate": 5.6532375929939075e-05,
+      "loss": 1.8875,
+      "step": 4412
+    },
+    {
+      "epoch": 0.45839825490807107,
+      "grad_norm": 0.3972803056240082,
+      "learning_rate": 5.65161988622988e-05,
+      "loss": 1.6262,
+      "step": 4413
+    },
+    {
+      "epoch": 0.4585021294276514,
+      "grad_norm": 0.38488489389419556,
+      "learning_rate": 5.650002110073418e-05,
+      "loss": 1.5766,
+      "step": 4414
+    },
+    {
+      "epoch": 0.4586060039472317,
+      "grad_norm": 0.45945605635643005,
+      "learning_rate": 5.648384264696802e-05,
+      "loss": 1.5854,
+      "step": 4415
+    },
+    {
+      "epoch": 0.4587098784668121,
+      "grad_norm": 0.3760433793067932,
+      "learning_rate": 5.6467663502723175e-05,
+      "loss": 1.6934,
+      "step": 4416
+    },
+    {
+      "epoch": 0.45881375298639243,
+      "grad_norm": 0.38754063844680786,
+      "learning_rate": 5.645148366972264e-05,
+      "loss": 1.7554,
+      "step": 4417
+    },
+    {
+      "epoch": 0.45891762750597276,
+      "grad_norm": 0.388735830783844,
+      "learning_rate": 5.643530314968941e-05,
+      "loss": 1.4804,
+      "step": 4418
+    },
+    {
+      "epoch": 0.45902150202555314,
+      "grad_norm": 0.39035165309906006,
+      "learning_rate": 5.6419121944346585e-05,
+      "loss": 1.6639,
+      "step": 4419
+    },
+    {
+      "epoch": 0.45912537654513347,
+      "grad_norm": 0.3769480586051941,
+      "learning_rate": 5.640294005541735e-05,
+      "loss": 1.7626,
+      "step": 4420
+    },
+    {
+      "epoch": 0.45922925106471385,
+      "grad_norm": 0.41671907901763916,
+      "learning_rate": 5.638675748462493e-05,
+      "loss": 1.7363,
+      "step": 4421
+    },
+    {
+      "epoch": 0.4593331255842942,
+      "grad_norm": 0.4125288426876068,
+      "learning_rate": 5.637057423369268e-05,
+      "loss": 1.725,
+      "step": 4422
+    },
+    {
+      "epoch": 0.4594370001038745,
+      "grad_norm": 0.44852516055107117,
+      "learning_rate": 5.635439030434395e-05,
+      "loss": 1.8253,
+      "step": 4423
+    },
+    {
+      "epoch": 0.4595408746234549,
+      "grad_norm": 0.4015621542930603,
+      "learning_rate": 5.6338205698302224e-05,
+      "loss": 1.6797,
+      "step": 4424
+    },
+    {
+      "epoch": 0.4596447491430352,
+      "grad_norm": 0.40786075592041016,
+      "learning_rate": 5.6322020417291034e-05,
+      "loss": 1.7569,
+      "step": 4425
+    },
+    {
+      "epoch": 0.45974862366261554,
+      "grad_norm": 0.4462954103946686,
+      "learning_rate": 5.630583446303399e-05,
+      "loss": 1.8999,
+      "step": 4426
+    },
+    {
+      "epoch": 0.4598524981821959,
+      "grad_norm": 0.40880078077316284,
+      "learning_rate": 5.6289647837254744e-05,
+      "loss": 1.5374,
+      "step": 4427
+    },
+    {
+      "epoch": 0.45995637270177625,
+      "grad_norm": 0.4236305356025696,
+      "learning_rate": 5.627346054167707e-05,
+      "loss": 1.7143,
+      "step": 4428
+    },
+    {
+      "epoch": 0.4600602472213566,
+      "grad_norm": 0.3998420834541321,
+      "learning_rate": 5.625727257802479e-05,
+      "loss": 1.8177,
+      "step": 4429
+    },
+    {
+      "epoch": 0.46016412174093696,
+      "grad_norm": 0.4473719000816345,
+      "learning_rate": 5.624108394802178e-05,
+      "loss": 1.8531,
+      "step": 4430
+    },
+    {
+      "epoch": 0.4602679962605173,
+      "grad_norm": 0.39816030859947205,
+      "learning_rate": 5.6224894653392035e-05,
+      "loss": 1.6969,
+      "step": 4431
+    },
+    {
+      "epoch": 0.46037187078009767,
+      "grad_norm": 0.36577850580215454,
+      "learning_rate": 5.6208704695859535e-05,
+      "loss": 1.5994,
+      "step": 4432
+    },
+    {
+      "epoch": 0.460475745299678,
+      "grad_norm": 0.3840119540691376,
+      "learning_rate": 5.619251407714843e-05,
+      "loss": 1.6662,
+      "step": 4433
+    },
+    {
+      "epoch": 0.4605796198192583,
+      "grad_norm": 0.4154397249221802,
+      "learning_rate": 5.617632279898288e-05,
+      "loss": 1.7509,
+      "step": 4434
+    },
+    {
+      "epoch": 0.4606834943388387,
+      "grad_norm": 0.3793880045413971,
+      "learning_rate": 5.6160130863087115e-05,
+      "loss": 1.6896,
+      "step": 4435
+    },
+    {
+      "epoch": 0.46078736885841903,
+      "grad_norm": 0.3721482455730438,
+      "learning_rate": 5.614393827118548e-05,
+      "loss": 1.6231,
+      "step": 4436
+    },
+    {
+      "epoch": 0.46089124337799936,
+      "grad_norm": 0.37379753589630127,
+      "learning_rate": 5.6127745025002344e-05,
+      "loss": 1.6131,
+      "step": 4437
+    },
+    {
+      "epoch": 0.46099511789757974,
+      "grad_norm": 0.3776273727416992,
+      "learning_rate": 5.611155112626215e-05,
+      "loss": 1.7206,
+      "step": 4438
+    },
+    {
+      "epoch": 0.46109899241716007,
+      "grad_norm": 0.37556153535842896,
+      "learning_rate": 5.609535657668945e-05,
+      "loss": 1.6558,
+      "step": 4439
+    },
+    {
+      "epoch": 0.4612028669367404,
+      "grad_norm": 0.41758623719215393,
+      "learning_rate": 5.6079161378008805e-05,
+      "loss": 1.7243,
+      "step": 4440
+    },
+    {
+      "epoch": 0.4613067414563208,
+      "grad_norm": 0.39208123087882996,
+      "learning_rate": 5.606296553194489e-05,
+      "loss": 1.7576,
+      "step": 4441
+    },
+    {
+      "epoch": 0.4614106159759011,
+      "grad_norm": 0.3936588764190674,
+      "learning_rate": 5.6046769040222446e-05,
+      "loss": 1.5866,
+      "step": 4442
+    },
+    {
+      "epoch": 0.46151449049548143,
+      "grad_norm": 0.3708806335926056,
+      "learning_rate": 5.603057190456628e-05,
+      "loss": 1.6697,
+      "step": 4443
+    },
+    {
+      "epoch": 0.4616183650150618,
+      "grad_norm": 0.4200730621814728,
+      "learning_rate": 5.601437412670121e-05,
+      "loss": 1.4743,
+      "step": 4444
+    },
+    {
+      "epoch": 0.46172223953464214,
+      "grad_norm": 0.40892651677131653,
+      "learning_rate": 5.5998175708352255e-05,
+      "loss": 1.7079,
+      "step": 4445
+    },
+    {
+      "epoch": 0.4618261140542225,
+      "grad_norm": 0.41610532999038696,
+      "learning_rate": 5.598197665124434e-05,
+      "loss": 1.6898,
+      "step": 4446
+    },
+    {
+      "epoch": 0.46192998857380285,
+      "grad_norm": 0.39250999689102173,
+      "learning_rate": 5.596577695710258e-05,
+      "loss": 1.686,
+      "step": 4447
+    },
+    {
+      "epoch": 0.4620338630933832,
+      "grad_norm": 0.3861941695213318,
+      "learning_rate": 5.594957662765211e-05,
+      "loss": 1.6314,
+      "step": 4448
+    },
+    {
+      "epoch": 0.46213773761296356,
+      "grad_norm": 0.4138748049736023,
+      "learning_rate": 5.5933375664618134e-05,
+      "loss": 1.6829,
+      "step": 4449
+    },
+    {
+      "epoch": 0.4622416121325439,
+      "grad_norm": 0.45050138235092163,
+      "learning_rate": 5.591717406972594e-05,
+      "loss": 1.9202,
+      "step": 4450
+    },
+    {
+      "epoch": 0.4623454866521242,
+      "grad_norm": 0.40642380714416504,
+      "learning_rate": 5.590097184470087e-05,
+      "loss": 1.7383,
+      "step": 4451
+    },
+    {
+      "epoch": 0.4624493611717046,
+      "grad_norm": 0.4135674238204956,
+      "learning_rate": 5.588476899126832e-05,
+      "loss": 1.7811,
+      "step": 4452
+    },
+    {
+      "epoch": 0.4625532356912849,
+      "grad_norm": 0.4031025767326355,
+      "learning_rate": 5.586856551115378e-05,
+      "loss": 1.6378,
+      "step": 4453
+    },
+    {
+      "epoch": 0.46265711021086525,
+      "grad_norm": 0.41000255942344666,
+      "learning_rate": 5.585236140608279e-05,
+      "loss": 1.7076,
+      "step": 4454
+    },
+    {
+      "epoch": 0.46276098473044563,
+      "grad_norm": 0.41616103053092957,
+      "learning_rate": 5.583615667778094e-05,
+      "loss": 1.7697,
+      "step": 4455
+    },
+    {
+      "epoch": 0.46286485925002596,
+      "grad_norm": 0.4099687337875366,
+      "learning_rate": 5.581995132797394e-05,
+      "loss": 1.5478,
+      "step": 4456
+    },
+    {
+      "epoch": 0.4629687337696063,
+      "grad_norm": 0.41153401136398315,
+      "learning_rate": 5.5803745358387526e-05,
+      "loss": 1.7247,
+      "step": 4457
+    },
+    {
+      "epoch": 0.46307260828918667,
+      "grad_norm": 0.41319242119789124,
+      "learning_rate": 5.578753877074749e-05,
+      "loss": 1.768,
+      "step": 4458
+    },
+    {
+      "epoch": 0.463176482808767,
+      "grad_norm": 0.3985736072063446,
+      "learning_rate": 5.577133156677975e-05,
+      "loss": 1.7182,
+      "step": 4459
+    },
+    {
+      "epoch": 0.4632803573283474,
+      "grad_norm": 0.4140629470348358,
+      "learning_rate": 5.575512374821018e-05,
+      "loss": 1.8411,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4633842318479277,
+      "grad_norm": 0.40386584401130676,
+      "learning_rate": 5.573891531676484e-05,
+      "loss": 1.6013,
+      "step": 4461
+    },
+    {
+      "epoch": 0.46348810636750803,
+      "grad_norm": 0.40092045068740845,
+      "learning_rate": 5.572270627416979e-05,
+      "loss": 1.8274,
+      "step": 4462
+    },
+    {
+      "epoch": 0.4635919808870884,
+      "grad_norm": 0.4397892355918884,
+      "learning_rate": 5.5706496622151136e-05,
+      "loss": 1.8865,
+      "step": 4463
+    },
+    {
+      "epoch": 0.46369585540666874,
+      "grad_norm": 0.41136595606803894,
+      "learning_rate": 5.569028636243513e-05,
+      "loss": 1.6921,
+      "step": 4464
+    },
+    {
+      "epoch": 0.46379972992624907,
+      "grad_norm": 0.3688594698905945,
+      "learning_rate": 5.5674075496748e-05,
+      "loss": 1.6099,
+      "step": 4465
+    },
+    {
+      "epoch": 0.46390360444582945,
+      "grad_norm": 0.41220369935035706,
+      "learning_rate": 5.56578640268161e-05,
+      "loss": 1.7032,
+      "step": 4466
+    },
+    {
+      "epoch": 0.4640074789654098,
+      "grad_norm": 0.38784199953079224,
+      "learning_rate": 5.5641651954365803e-05,
+      "loss": 1.7474,
+      "step": 4467
+    },
+    {
+      "epoch": 0.4641113534849901,
+      "grad_norm": 0.3877417743206024,
+      "learning_rate": 5.562543928112358e-05,
+      "loss": 1.7868,
+      "step": 4468
+    },
+    {
+      "epoch": 0.4642152280045705,
+      "grad_norm": 0.41094866394996643,
+      "learning_rate": 5.560922600881595e-05,
+      "loss": 1.684,
+      "step": 4469
+    },
+    {
+      "epoch": 0.4643191025241508,
+      "grad_norm": 0.3839239180088043,
+      "learning_rate": 5.559301213916952e-05,
+      "loss": 1.601,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4644229770437312,
+      "grad_norm": 0.3839179277420044,
+      "learning_rate": 5.557679767391091e-05,
+      "loss": 1.773,
+      "step": 4471
+    },
+    {
+      "epoch": 0.4645268515633115,
+      "grad_norm": 0.3948630392551422,
+      "learning_rate": 5.5560582614766845e-05,
+      "loss": 1.7059,
+      "step": 4472
+    },
+    {
+      "epoch": 0.46463072608289185,
+      "grad_norm": 0.4155755639076233,
+      "learning_rate": 5.5544366963464134e-05,
+      "loss": 1.7463,
+      "step": 4473
+    },
+    {
+      "epoch": 0.46473460060247224,
+      "grad_norm": 0.3817596733570099,
+      "learning_rate": 5.552815072172955e-05,
+      "loss": 1.6543,
+      "step": 4474
+    },
+    {
+      "epoch": 0.46483847512205256,
+      "grad_norm": 0.40604111552238464,
+      "learning_rate": 5.5511933891290056e-05,
+      "loss": 1.6462,
+      "step": 4475
+    },
+    {
+      "epoch": 0.4649423496416329,
+      "grad_norm": 0.39084017276763916,
+      "learning_rate": 5.5495716473872604e-05,
+      "loss": 1.656,
+      "step": 4476
+    },
+    {
+      "epoch": 0.4650462241612133,
+      "grad_norm": 0.4190458059310913,
+      "learning_rate": 5.5479498471204196e-05,
+      "loss": 1.7544,
+      "step": 4477
+    },
+    {
+      "epoch": 0.4651500986807936,
+      "grad_norm": 0.37424778938293457,
+      "learning_rate": 5.546327988501196e-05,
+      "loss": 1.5076,
+      "step": 4478
+    },
+    {
+      "epoch": 0.4652539732003739,
+      "grad_norm": 0.4260268807411194,
+      "learning_rate": 5.544706071702302e-05,
+      "loss": 1.638,
+      "step": 4479
+    },
+    {
+      "epoch": 0.4653578477199543,
+      "grad_norm": 0.3845141530036926,
+      "learning_rate": 5.5430840968964615e-05,
+      "loss": 1.5985,
+      "step": 4480
+    },
+    {
+      "epoch": 0.46546172223953464,
+      "grad_norm": 0.385220468044281,
+      "learning_rate": 5.541462064256401e-05,
+      "loss": 1.7247,
+      "step": 4481
+    },
+    {
+      "epoch": 0.46556559675911496,
+      "grad_norm": 0.41517341136932373,
+      "learning_rate": 5.5398399739548524e-05,
+      "loss": 1.8184,
+      "step": 4482
+    },
+    {
+      "epoch": 0.46566947127869535,
+      "grad_norm": 0.4744667112827301,
+      "learning_rate": 5.538217826164558e-05,
+      "loss": 1.8839,
+      "step": 4483
+    },
+    {
+      "epoch": 0.4657733457982757,
+      "grad_norm": 0.40122199058532715,
+      "learning_rate": 5.5365956210582634e-05,
+      "loss": 1.5911,
+      "step": 4484
+    },
+    {
+      "epoch": 0.46587722031785606,
+      "grad_norm": 0.4111073911190033,
+      "learning_rate": 5.5349733588087216e-05,
+      "loss": 1.7894,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4659810948374364,
+      "grad_norm": 0.38711994886398315,
+      "learning_rate": 5.533351039588689e-05,
+      "loss": 1.7112,
+      "step": 4486
+    },
+    {
+      "epoch": 0.4660849693570167,
+      "grad_norm": 0.4004124104976654,
+      "learning_rate": 5.531728663570933e-05,
+      "loss": 1.7559,
+      "step": 4487
+    },
+    {
+      "epoch": 0.4661888438765971,
+      "grad_norm": 0.3998907804489136,
+      "learning_rate": 5.53010623092822e-05,
+      "loss": 1.6631,
+      "step": 4488
+    },
+    {
+      "epoch": 0.4662927183961774,
+      "grad_norm": 0.3848699927330017,
+      "learning_rate": 5.5284837418333294e-05,
+      "loss": 1.736,
+      "step": 4489
+    },
+    {
+      "epoch": 0.46639659291575775,
+      "grad_norm": 0.41114541888237,
+      "learning_rate": 5.5268611964590425e-05,
+      "loss": 1.8165,
+      "step": 4490
+    },
+    {
+      "epoch": 0.46650046743533813,
+      "grad_norm": 0.475700706243515,
+      "learning_rate": 5.525238594978146e-05,
+      "loss": 1.8559,
+      "step": 4491
+    },
+    {
+      "epoch": 0.46660434195491846,
+      "grad_norm": 0.37533336877822876,
+      "learning_rate": 5.523615937563438e-05,
+      "loss": 1.5622,
+      "step": 4492
+    },
+    {
+      "epoch": 0.4667082164744988,
+      "grad_norm": 0.38787853717803955,
+      "learning_rate": 5.521993224387717e-05,
+      "loss": 1.6647,
+      "step": 4493
+    },
+    {
+      "epoch": 0.46681209099407917,
+      "grad_norm": 0.4438975155353546,
+      "learning_rate": 5.520370455623789e-05,
+      "loss": 1.7663,
+      "step": 4494
+    },
+    {
+      "epoch": 0.4669159655136595,
+      "grad_norm": 0.4348212480545044,
+      "learning_rate": 5.518747631444467e-05,
+      "loss": 1.8218,
+      "step": 4495
+    },
+    {
+      "epoch": 0.4670198400332398,
+      "grad_norm": 0.4120256304740906,
+      "learning_rate": 5.517124752022569e-05,
+      "loss": 1.6427,
+      "step": 4496
+    },
+    {
+      "epoch": 0.4671237145528202,
+      "grad_norm": 0.3732791244983673,
+      "learning_rate": 5.515501817530917e-05,
+      "loss": 1.5317,
+      "step": 4497
+    },
+    {
+      "epoch": 0.46722758907240053,
+      "grad_norm": 0.40980592370033264,
+      "learning_rate": 5.513878828142344e-05,
+      "loss": 1.736,
+      "step": 4498
+    },
+    {
+      "epoch": 0.4673314635919809,
+      "grad_norm": 0.42405858635902405,
+      "learning_rate": 5.5122557840296815e-05,
+      "loss": 1.7212,
+      "step": 4499
+    },
+    {
+      "epoch": 0.46743533811156124,
+      "grad_norm": 0.380643367767334,
+      "learning_rate": 5.5106326853657764e-05,
+      "loss": 1.6732,
+      "step": 4500
+    },
+    {
+      "epoch": 0.46753921263114157,
+      "grad_norm": 0.4484168589115143,
+      "learning_rate": 5.5090095323234735e-05,
+      "loss": 1.8326,
+      "step": 4501
+    },
+    {
+      "epoch": 0.46764308715072195,
+      "grad_norm": 0.44353044033050537,
+      "learning_rate": 5.507386325075625e-05,
+      "loss": 1.8159,
+      "step": 4502
+    },
+    {
+      "epoch": 0.4677469616703023,
+      "grad_norm": 0.40523630380630493,
+      "learning_rate": 5.505763063795091e-05,
+      "loss": 1.8045,
+      "step": 4503
+    },
+    {
+      "epoch": 0.4678508361898826,
+      "grad_norm": 0.41718611121177673,
+      "learning_rate": 5.504139748654735e-05,
+      "loss": 1.7684,
+      "step": 4504
+    },
+    {
+      "epoch": 0.467954710709463,
+      "grad_norm": 0.3799745440483093,
+      "learning_rate": 5.5025163798274283e-05,
+      "loss": 1.7299,
+      "step": 4505
+    },
+    {
+      "epoch": 0.4680585852290433,
+      "grad_norm": 0.5379766821861267,
+      "learning_rate": 5.500892957486048e-05,
+      "loss": 1.8735,
+      "step": 4506
+    },
+    {
+      "epoch": 0.46816245974862364,
+      "grad_norm": 0.42083024978637695,
+      "learning_rate": 5.499269481803474e-05,
+      "loss": 1.6187,
+      "step": 4507
+    },
+    {
+      "epoch": 0.468266334268204,
+      "grad_norm": 0.44988521933555603,
+      "learning_rate": 5.497645952952596e-05,
+      "loss": 1.8866,
+      "step": 4508
+    },
+    {
+      "epoch": 0.46837020878778435,
+      "grad_norm": 0.37519097328186035,
+      "learning_rate": 5.496022371106304e-05,
+      "loss": 1.6358,
+      "step": 4509
+    },
+    {
+      "epoch": 0.46847408330736473,
+      "grad_norm": 0.40404897928237915,
+      "learning_rate": 5.4943987364374985e-05,
+      "loss": 1.6527,
+      "step": 4510
+    },
+    {
+      "epoch": 0.46857795782694506,
+      "grad_norm": 0.3824155032634735,
+      "learning_rate": 5.492775049119085e-05,
+      "loss": 1.571,
+      "step": 4511
+    },
+    {
+      "epoch": 0.4686818323465254,
+      "grad_norm": 0.38465747237205505,
+      "learning_rate": 5.491151309323973e-05,
+      "loss": 1.8012,
+      "step": 4512
+    },
+    {
+      "epoch": 0.46878570686610577,
+      "grad_norm": 0.41244107484817505,
+      "learning_rate": 5.489527517225076e-05,
+      "loss": 1.7915,
+      "step": 4513
+    },
+    {
+      "epoch": 0.4688895813856861,
+      "grad_norm": 0.39716869592666626,
+      "learning_rate": 5.487903672995318e-05,
+      "loss": 1.6541,
+      "step": 4514
+    },
+    {
+      "epoch": 0.4689934559052664,
+      "grad_norm": 0.3766006529331207,
+      "learning_rate": 5.486279776807626e-05,
+      "loss": 1.7457,
+      "step": 4515
+    },
+    {
+      "epoch": 0.4690973304248468,
+      "grad_norm": 0.3792710602283478,
+      "learning_rate": 5.484655828834929e-05,
+      "loss": 1.5523,
+      "step": 4516
+    },
+    {
+      "epoch": 0.46920120494442713,
+      "grad_norm": 0.4360962212085724,
+      "learning_rate": 5.483031829250167e-05,
+      "loss": 1.7419,
+      "step": 4517
+    },
+    {
+      "epoch": 0.46930507946400746,
+      "grad_norm": 0.4073307514190674,
+      "learning_rate": 5.481407778226284e-05,
+      "loss": 1.7789,
+      "step": 4518
+    },
+    {
+      "epoch": 0.46940895398358784,
+      "grad_norm": 0.3982420563697815,
+      "learning_rate": 5.4797836759362256e-05,
+      "loss": 1.6478,
+      "step": 4519
+    },
+    {
+      "epoch": 0.46951282850316817,
+      "grad_norm": 0.3903031051158905,
+      "learning_rate": 5.478159522552949e-05,
+      "loss": 1.5748,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4696167030227485,
+      "grad_norm": 0.4410886764526367,
+      "learning_rate": 5.476535318249414e-05,
+      "loss": 1.7942,
+      "step": 4521
+    },
+    {
+      "epoch": 0.4697205775423289,
+      "grad_norm": 0.3934234380722046,
+      "learning_rate": 5.4749110631985846e-05,
+      "loss": 1.6272,
+      "step": 4522
+    },
+    {
+      "epoch": 0.4698244520619092,
+      "grad_norm": 0.4124247431755066,
+      "learning_rate": 5.4732867575734314e-05,
+      "loss": 1.7511,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4699283265814896,
+      "grad_norm": 0.4178972840309143,
+      "learning_rate": 5.4716624015469284e-05,
+      "loss": 1.7224,
+      "step": 4524
+    },
+    {
+      "epoch": 0.4700322011010699,
+      "grad_norm": 0.44093725085258484,
+      "learning_rate": 5.470037995292061e-05,
+      "loss": 1.6125,
+      "step": 4525
+    },
+    {
+      "epoch": 0.47013607562065024,
+      "grad_norm": 0.3923119008541107,
+      "learning_rate": 5.468413538981814e-05,
+      "loss": 1.6888,
+      "step": 4526
+    },
+    {
+      "epoch": 0.4702399501402306,
+      "grad_norm": 0.4040331244468689,
+      "learning_rate": 5.466789032789176e-05,
+      "loss": 1.7642,
+      "step": 4527
+    },
+    {
+      "epoch": 0.47034382465981095,
+      "grad_norm": 0.4173428416252136,
+      "learning_rate": 5.465164476887149e-05,
+      "loss": 1.6372,
+      "step": 4528
+    },
+    {
+      "epoch": 0.4704476991793913,
+      "grad_norm": 0.36699429154396057,
+      "learning_rate": 5.463539871448735e-05,
+      "loss": 1.7026,
+      "step": 4529
+    },
+    {
+      "epoch": 0.47055157369897166,
+      "grad_norm": 0.37579965591430664,
+      "learning_rate": 5.461915216646938e-05,
+      "loss": 1.513,
+      "step": 4530
+    },
+    {
+      "epoch": 0.470655448218552,
+      "grad_norm": 0.39434462785720825,
+      "learning_rate": 5.4602905126547744e-05,
+      "loss": 1.6832,
+      "step": 4531
+    },
+    {
+      "epoch": 0.4707593227381323,
+      "grad_norm": 0.40880143642425537,
+      "learning_rate": 5.4586657596452615e-05,
+      "loss": 1.5767,
+      "step": 4532
+    },
+    {
+      "epoch": 0.4708631972577127,
+      "grad_norm": 0.4027133584022522,
+      "learning_rate": 5.4570409577914215e-05,
+      "loss": 1.6025,
+      "step": 4533
+    },
+    {
+      "epoch": 0.470967071777293,
+      "grad_norm": 0.37982282042503357,
+      "learning_rate": 5.455416107266287e-05,
+      "loss": 1.6899,
+      "step": 4534
+    },
+    {
+      "epoch": 0.47107094629687335,
+      "grad_norm": 0.4183795154094696,
+      "learning_rate": 5.453791208242889e-05,
+      "loss": 1.7355,
+      "step": 4535
+    },
+    {
+      "epoch": 0.47117482081645373,
+      "grad_norm": 0.4518531560897827,
+      "learning_rate": 5.452166260894267e-05,
+      "loss": 1.9148,
+      "step": 4536
+    },
+    {
+      "epoch": 0.47127869533603406,
+      "grad_norm": 0.41706570982933044,
+      "learning_rate": 5.450541265393464e-05,
+      "loss": 1.7175,
+      "step": 4537
+    },
+    {
+      "epoch": 0.47138256985561444,
+      "grad_norm": 0.38298308849334717,
+      "learning_rate": 5.4489162219135314e-05,
+      "loss": 1.8146,
+      "step": 4538
+    },
+    {
+      "epoch": 0.47148644437519477,
+      "grad_norm": 0.40449774265289307,
+      "learning_rate": 5.447291130627522e-05,
+      "loss": 1.8353,
+      "step": 4539
+    },
+    {
+      "epoch": 0.4715903188947751,
+      "grad_norm": 0.40344521403312683,
+      "learning_rate": 5.4456659917084975e-05,
+      "loss": 1.672,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4716941934143555,
+      "grad_norm": 0.3761052191257477,
+      "learning_rate": 5.44404080532952e-05,
+      "loss": 1.7119,
+      "step": 4541
+    },
+    {
+      "epoch": 0.4717980679339358,
+      "grad_norm": 0.45389243960380554,
+      "learning_rate": 5.442415571663661e-05,
+      "loss": 1.9569,
+      "step": 4542
+    },
+    {
+      "epoch": 0.47190194245351613,
+      "grad_norm": 0.3938455581665039,
+      "learning_rate": 5.440790290883997e-05,
+      "loss": 1.5407,
+      "step": 4543
+    },
+    {
+      "epoch": 0.4720058169730965,
+      "grad_norm": 0.39464959502220154,
+      "learning_rate": 5.439164963163601e-05,
+      "loss": 1.7799,
+      "step": 4544
+    },
+    {
+      "epoch": 0.47210969149267684,
+      "grad_norm": 0.3845635652542114,
+      "learning_rate": 5.437539588675564e-05,
+      "loss": 1.6902,
+      "step": 4545
+    },
+    {
+      "epoch": 0.47221356601225717,
+      "grad_norm": 0.41250425577163696,
+      "learning_rate": 5.435914167592975e-05,
+      "loss": 1.7341,
+      "step": 4546
+    },
+    {
+      "epoch": 0.47231744053183755,
+      "grad_norm": 0.42640382051467896,
+      "learning_rate": 5.434288700088925e-05,
+      "loss": 1.7448,
+      "step": 4547
+    },
+    {
+      "epoch": 0.4724213150514179,
+      "grad_norm": 0.3739955425262451,
+      "learning_rate": 5.432663186336517e-05,
+      "loss": 1.6007,
+      "step": 4548
+    },
+    {
+      "epoch": 0.47252518957099826,
+      "grad_norm": 0.3853756785392761,
+      "learning_rate": 5.431037626508856e-05,
+      "loss": 1.6348,
+      "step": 4549
+    },
+    {
+      "epoch": 0.4726290640905786,
+      "grad_norm": 0.4050140082836151,
+      "learning_rate": 5.429412020779049e-05,
+      "loss": 1.6318,
+      "step": 4550
+    },
+    {
+      "epoch": 0.4727329386101589,
+      "grad_norm": 0.39661628007888794,
+      "learning_rate": 5.427786369320211e-05,
+      "loss": 1.7079,
+      "step": 4551
+    },
+    {
+      "epoch": 0.4728368131297393,
+      "grad_norm": 0.3745506703853607,
+      "learning_rate": 5.4261606723054604e-05,
+      "loss": 1.5058,
+      "step": 4552
+    },
+    {
+      "epoch": 0.4729406876493196,
+      "grad_norm": 0.45182400941848755,
+      "learning_rate": 5.424534929907924e-05,
+      "loss": 1.7662,
+      "step": 4553
+    },
+    {
+      "epoch": 0.47304456216889995,
+      "grad_norm": 0.39401260018348694,
+      "learning_rate": 5.422909142300729e-05,
+      "loss": 1.6387,
+      "step": 4554
+    },
+    {
+      "epoch": 0.47314843668848033,
+      "grad_norm": 0.43190351128578186,
+      "learning_rate": 5.4212833096570085e-05,
+      "loss": 1.9062,
+      "step": 4555
+    },
+    {
+      "epoch": 0.47325231120806066,
+      "grad_norm": 0.3953791558742523,
+      "learning_rate": 5.419657432149903e-05,
+      "loss": 1.779,
+      "step": 4556
+    },
+    {
+      "epoch": 0.473356185727641,
+      "grad_norm": 0.4370119273662567,
+      "learning_rate": 5.418031509952554e-05,
+      "loss": 1.7205,
+      "step": 4557
+    },
+    {
+      "epoch": 0.47346006024722137,
+      "grad_norm": 0.46050959825515747,
+      "learning_rate": 5.4164055432381076e-05,
+      "loss": 1.8187,
+      "step": 4558
+    },
+    {
+      "epoch": 0.4735639347668017,
+      "grad_norm": 0.3929772675037384,
+      "learning_rate": 5.414779532179721e-05,
+      "loss": 1.7266,
+      "step": 4559
+    },
+    {
+      "epoch": 0.473667809286382,
+      "grad_norm": 0.3913351595401764,
+      "learning_rate": 5.41315347695055e-05,
+      "loss": 1.6847,
+      "step": 4560
+    },
+    {
+      "epoch": 0.4737716838059624,
+      "grad_norm": 0.37702706456184387,
+      "learning_rate": 5.4115273777237554e-05,
+      "loss": 1.5548,
+      "step": 4561
+    },
+    {
+      "epoch": 0.47387555832554273,
+      "grad_norm": 0.4072859585285187,
+      "learning_rate": 5.4099012346725054e-05,
+      "loss": 1.6328,
+      "step": 4562
+    },
+    {
+      "epoch": 0.4739794328451231,
+      "grad_norm": 0.4485922157764435,
+      "learning_rate": 5.4082750479699715e-05,
+      "loss": 1.7856,
+      "step": 4563
+    },
+    {
+      "epoch": 0.47408330736470344,
+      "grad_norm": 0.4083961248397827,
+      "learning_rate": 5.406648817789329e-05,
+      "loss": 1.6751,
+      "step": 4564
+    },
+    {
+      "epoch": 0.47418718188428377,
+      "grad_norm": 0.4277319312095642,
+      "learning_rate": 5.40502254430376e-05,
+      "loss": 1.793,
+      "step": 4565
+    },
+    {
+      "epoch": 0.47429105640386415,
+      "grad_norm": 0.4535433053970337,
+      "learning_rate": 5.403396227686449e-05,
+      "loss": 1.5738,
+      "step": 4566
+    },
+    {
+      "epoch": 0.4743949309234445,
+      "grad_norm": 0.41116324067115784,
+      "learning_rate": 5.4017698681105865e-05,
+      "loss": 1.7443,
+      "step": 4567
+    },
+    {
+      "epoch": 0.4744988054430248,
+      "grad_norm": 0.38053932785987854,
+      "learning_rate": 5.400143465749368e-05,
+      "loss": 1.6522,
+      "step": 4568
+    },
+    {
+      "epoch": 0.4746026799626052,
+      "grad_norm": 0.40427184104919434,
+      "learning_rate": 5.398517020775989e-05,
+      "loss": 1.7336,
+      "step": 4569
+    },
+    {
+      "epoch": 0.4747065544821855,
+      "grad_norm": 0.4339055120944977,
+      "learning_rate": 5.39689053336366e-05,
+      "loss": 1.7395,
+      "step": 4570
+    },
+    {
+      "epoch": 0.47481042900176584,
+      "grad_norm": 0.415379136800766,
+      "learning_rate": 5.3952640036855825e-05,
+      "loss": 1.5726,
+      "step": 4571
+    },
+    {
+      "epoch": 0.4749143035213462,
+      "grad_norm": 0.4650099575519562,
+      "learning_rate": 5.3936374319149705e-05,
+      "loss": 2.0233,
+      "step": 4572
+    },
+    {
+      "epoch": 0.47501817804092655,
+      "grad_norm": 0.429559588432312,
+      "learning_rate": 5.392010818225044e-05,
+      "loss": 1.539,
+      "step": 4573
+    },
+    {
+      "epoch": 0.4751220525605069,
+      "grad_norm": 0.4143827259540558,
+      "learning_rate": 5.390384162789024e-05,
+      "loss": 1.767,
+      "step": 4574
+    },
+    {
+      "epoch": 0.47522592708008726,
+      "grad_norm": 0.4033125042915344,
+      "learning_rate": 5.3887574657801346e-05,
+      "loss": 1.7017,
+      "step": 4575
+    },
+    {
+      "epoch": 0.4753298015996676,
+      "grad_norm": 0.41067859530448914,
+      "learning_rate": 5.3871307273716074e-05,
+      "loss": 1.6445,
+      "step": 4576
+    },
+    {
+      "epoch": 0.47543367611924797,
+      "grad_norm": 0.41794803738594055,
+      "learning_rate": 5.3855039477366775e-05,
+      "loss": 1.7336,
+      "step": 4577
+    },
+    {
+      "epoch": 0.4755375506388283,
+      "grad_norm": 0.38265594840049744,
+      "learning_rate": 5.3838771270485854e-05,
+      "loss": 1.6598,
+      "step": 4578
+    },
+    {
+      "epoch": 0.4756414251584086,
+      "grad_norm": 0.3966656029224396,
+      "learning_rate": 5.3822502654805725e-05,
+      "loss": 1.5692,
+      "step": 4579
+    },
+    {
+      "epoch": 0.475745299677989,
+      "grad_norm": 0.4140309989452362,
+      "learning_rate": 5.3806233632058876e-05,
+      "loss": 1.687,
+      "step": 4580
+    },
+    {
+      "epoch": 0.47584917419756934,
+      "grad_norm": 0.43315640091896057,
+      "learning_rate": 5.378996420397784e-05,
+      "loss": 1.6507,
+      "step": 4581
+    },
+    {
+      "epoch": 0.47595304871714966,
+      "grad_norm": 0.4187996983528137,
+      "learning_rate": 5.377369437229518e-05,
+      "loss": 1.782,
+      "step": 4582
+    },
+    {
+      "epoch": 0.47605692323673005,
+      "grad_norm": 0.41366416215896606,
+      "learning_rate": 5.375742413874351e-05,
+      "loss": 1.7035,
+      "step": 4583
+    },
+    {
+      "epoch": 0.47616079775631037,
+      "grad_norm": 0.4386520981788635,
+      "learning_rate": 5.3741153505055475e-05,
+      "loss": 1.6678,
+      "step": 4584
+    },
+    {
+      "epoch": 0.4762646722758907,
+      "grad_norm": 0.4309309124946594,
+      "learning_rate": 5.372488247296378e-05,
+      "loss": 1.7662,
+      "step": 4585
+    },
+    {
+      "epoch": 0.4763685467954711,
+      "grad_norm": 0.4261937737464905,
+      "learning_rate": 5.370861104420114e-05,
+      "loss": 1.7769,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4764724213150514,
+      "grad_norm": 0.3945866823196411,
+      "learning_rate": 5.369233922050036e-05,
+      "loss": 1.4977,
+      "step": 4587
+    },
+    {
+      "epoch": 0.4765762958346318,
+      "grad_norm": 0.38651934266090393,
+      "learning_rate": 5.3676067003594255e-05,
+      "loss": 1.7031,
+      "step": 4588
+    },
+    {
+      "epoch": 0.4766801703542121,
+      "grad_norm": 0.43587467074394226,
+      "learning_rate": 5.365979439521569e-05,
+      "loss": 1.6961,
+      "step": 4589
+    },
+    {
+      "epoch": 0.47678404487379245,
+      "grad_norm": 0.43202391266822815,
+      "learning_rate": 5.364352139709758e-05,
+      "loss": 1.8117,
+      "step": 4590
+    },
+    {
+      "epoch": 0.47688791939337283,
+      "grad_norm": 0.4378650486469269,
+      "learning_rate": 5.362724801097284e-05,
+      "loss": 1.6322,
+      "step": 4591
+    },
+    {
+      "epoch": 0.47699179391295315,
+      "grad_norm": 0.4283282160758972,
+      "learning_rate": 5.3610974238574495e-05,
+      "loss": 1.7273,
+      "step": 4592
+    },
+    {
+      "epoch": 0.4770956684325335,
+      "grad_norm": 0.42612189054489136,
+      "learning_rate": 5.359470008163556e-05,
+      "loss": 1.824,
+      "step": 4593
+    },
+    {
+      "epoch": 0.47719954295211386,
+      "grad_norm": 0.38064736127853394,
+      "learning_rate": 5.35784255418891e-05,
+      "loss": 1.5924,
+      "step": 4594
+    },
+    {
+      "epoch": 0.4773034174716942,
+      "grad_norm": 0.40858712792396545,
+      "learning_rate": 5.356215062106823e-05,
+      "loss": 1.7402,
+      "step": 4595
+    },
+    {
+      "epoch": 0.4774072919912745,
+      "grad_norm": 0.4329719543457031,
+      "learning_rate": 5.354587532090611e-05,
+      "loss": 1.861,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4775111665108549,
+      "grad_norm": 0.4516226649284363,
+      "learning_rate": 5.352959964313592e-05,
+      "loss": 1.8382,
+      "step": 4597
+    },
+    {
+      "epoch": 0.47761504103043523,
+      "grad_norm": 0.4081841707229614,
+      "learning_rate": 5.3513323589490924e-05,
+      "loss": 1.6581,
+      "step": 4598
+    },
+    {
+      "epoch": 0.47771891555001555,
+      "grad_norm": 0.3968779444694519,
+      "learning_rate": 5.349704716170435e-05,
+      "loss": 1.5309,
+      "step": 4599
+    },
+    {
+      "epoch": 0.47782279006959594,
+      "grad_norm": 0.4048573076725006,
+      "learning_rate": 5.348077036150954e-05,
+      "loss": 1.7319,
+      "step": 4600
+    },
+    {
+      "epoch": 0.47792666458917626,
+      "grad_norm": 0.4110606014728546,
+      "learning_rate": 5.346449319063983e-05,
+      "loss": 1.7603,
+      "step": 4601
+    },
+    {
+      "epoch": 0.47803053910875665,
+      "grad_norm": 0.4106254577636719,
+      "learning_rate": 5.34482156508286e-05,
+      "loss": 1.5944,
+      "step": 4602
+    },
+    {
+      "epoch": 0.478134413628337,
+      "grad_norm": 0.3852187693119049,
+      "learning_rate": 5.343193774380933e-05,
+      "loss": 1.5459,
+      "step": 4603
+    },
+    {
+      "epoch": 0.4782382881479173,
+      "grad_norm": 0.40513893961906433,
+      "learning_rate": 5.341565947131546e-05,
+      "loss": 1.6581,
+      "step": 4604
+    },
+    {
+      "epoch": 0.4783421626674977,
+      "grad_norm": 0.38794079422950745,
+      "learning_rate": 5.3399380835080493e-05,
+      "loss": 1.6475,
+      "step": 4605
+    },
+    {
+      "epoch": 0.478446037187078,
+      "grad_norm": 0.3935708999633789,
+      "learning_rate": 5.338310183683799e-05,
+      "loss": 1.7491,
+      "step": 4606
+    },
+    {
+      "epoch": 0.47854991170665834,
+      "grad_norm": 0.5326302647590637,
+      "learning_rate": 5.3366822478321524e-05,
+      "loss": 1.7748,
+      "step": 4607
+    },
+    {
+      "epoch": 0.4786537862262387,
+      "grad_norm": 0.4071970283985138,
+      "learning_rate": 5.335054276126471e-05,
+      "loss": 1.6662,
+      "step": 4608
+    },
+    {
+      "epoch": 0.47875766074581905,
+      "grad_norm": 0.43349170684814453,
+      "learning_rate": 5.333426268740125e-05,
+      "loss": 1.8815,
+      "step": 4609
+    },
+    {
+      "epoch": 0.4788615352653994,
+      "grad_norm": 0.372646689414978,
+      "learning_rate": 5.331798225846482e-05,
+      "loss": 1.5864,
+      "step": 4610
+    },
+    {
+      "epoch": 0.47896540978497976,
+      "grad_norm": 0.4014649987220764,
+      "learning_rate": 5.330170147618914e-05,
+      "loss": 1.6708,
+      "step": 4611
+    },
+    {
+      "epoch": 0.4790692843045601,
+      "grad_norm": 0.4292800724506378,
+      "learning_rate": 5.3285420342308035e-05,
+      "loss": 1.8315,
+      "step": 4612
+    },
+    {
+      "epoch": 0.47917315882414047,
+      "grad_norm": 0.41147077083587646,
+      "learning_rate": 5.326913885855527e-05,
+      "loss": 1.6449,
+      "step": 4613
+    },
+    {
+      "epoch": 0.4792770333437208,
+      "grad_norm": 0.35840991139411926,
+      "learning_rate": 5.325285702666473e-05,
+      "loss": 1.5407,
+      "step": 4614
+    },
+    {
+      "epoch": 0.4793809078633011,
+      "grad_norm": 0.4264986515045166,
+      "learning_rate": 5.323657484837028e-05,
+      "loss": 1.6642,
+      "step": 4615
+    },
+    {
+      "epoch": 0.4794847823828815,
+      "grad_norm": 0.4104459881782532,
+      "learning_rate": 5.3220292325405854e-05,
+      "loss": 1.6084,
+      "step": 4616
+    },
+    {
+      "epoch": 0.47958865690246183,
+      "grad_norm": 0.4270613193511963,
+      "learning_rate": 5.320400945950542e-05,
+      "loss": 1.9207,
+      "step": 4617
+    },
+    {
+      "epoch": 0.47969253142204216,
+      "grad_norm": 0.4374467730522156,
+      "learning_rate": 5.318772625240298e-05,
+      "loss": 1.6418,
+      "step": 4618
+    },
+    {
+      "epoch": 0.47979640594162254,
+      "grad_norm": 0.3923068940639496,
+      "learning_rate": 5.3171442705832554e-05,
+      "loss": 1.6963,
+      "step": 4619
+    },
+    {
+      "epoch": 0.47990028046120287,
+      "grad_norm": 0.3642514646053314,
+      "learning_rate": 5.315515882152822e-05,
+      "loss": 1.5201,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4800041549807832,
+      "grad_norm": 0.41580015420913696,
+      "learning_rate": 5.313887460122408e-05,
+      "loss": 1.5179,
+      "step": 4621
+    },
+    {
+      "epoch": 0.4801080295003636,
+      "grad_norm": 0.42775875329971313,
+      "learning_rate": 5.312259004665426e-05,
+      "loss": 1.7967,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4802119040199439,
+      "grad_norm": 0.3977086544036865,
+      "learning_rate": 5.310630515955297e-05,
+      "loss": 1.4864,
+      "step": 4623
+    },
+    {
+      "epoch": 0.48031577853952423,
+      "grad_norm": 0.41823333501815796,
+      "learning_rate": 5.309001994165442e-05,
+      "loss": 1.7526,
+      "step": 4624
+    },
+    {
+      "epoch": 0.4804196530591046,
+      "grad_norm": 0.41474875807762146,
+      "learning_rate": 5.307373439469283e-05,
+      "loss": 1.5733,
+      "step": 4625
+    },
+    {
+      "epoch": 0.48052352757868494,
+      "grad_norm": 0.370061457157135,
+      "learning_rate": 5.305744852040253e-05,
+      "loss": 1.541,
+      "step": 4626
+    },
+    {
+      "epoch": 0.4806274020982653,
+      "grad_norm": 0.36893734335899353,
+      "learning_rate": 5.304116232051779e-05,
+      "loss": 1.6238,
+      "step": 4627
+    },
+    {
+      "epoch": 0.48073127661784565,
+      "grad_norm": 0.4357246458530426,
+      "learning_rate": 5.3024875796773e-05,
+      "loss": 1.9604,
+      "step": 4628
+    },
+    {
+      "epoch": 0.480835151137426,
+      "grad_norm": 0.3936040997505188,
+      "learning_rate": 5.3008588950902524e-05,
+      "loss": 1.4643,
+      "step": 4629
+    },
+    {
+      "epoch": 0.48093902565700636,
+      "grad_norm": 0.38085660338401794,
+      "learning_rate": 5.2992301784640785e-05,
+      "loss": 1.6453,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4810429001765867,
+      "grad_norm": 0.3864935338497162,
+      "learning_rate": 5.2976014299722263e-05,
+      "loss": 1.6807,
+      "step": 4631
+    },
+    {
+      "epoch": 0.481146774696167,
+      "grad_norm": 0.405137836933136,
+      "learning_rate": 5.2959726497881425e-05,
+      "loss": 1.7978,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4812506492157474,
+      "grad_norm": 0.3816376328468323,
+      "learning_rate": 5.29434383808528e-05,
+      "loss": 1.7011,
+      "step": 4633
+    },
+    {
+      "epoch": 0.4813545237353277,
+      "grad_norm": 0.3807872533798218,
+      "learning_rate": 5.292714995037096e-05,
+      "loss": 1.6109,
+      "step": 4634
+    },
+    {
+      "epoch": 0.48145839825490805,
+      "grad_norm": 0.3880075514316559,
+      "learning_rate": 5.291086120817047e-05,
+      "loss": 1.6175,
+      "step": 4635
+    },
+    {
+      "epoch": 0.48156227277448843,
+      "grad_norm": 0.40148478746414185,
+      "learning_rate": 5.289457215598595e-05,
+      "loss": 1.5225,
+      "step": 4636
+    },
+    {
+      "epoch": 0.48166614729406876,
+      "grad_norm": 0.4223523437976837,
+      "learning_rate": 5.28782827955521e-05,
+      "loss": 1.7701,
+      "step": 4637
+    },
+    {
+      "epoch": 0.4817700218136491,
+      "grad_norm": 0.41836312413215637,
+      "learning_rate": 5.286199312860358e-05,
+      "loss": 1.6871,
+      "step": 4638
+    },
+    {
+      "epoch": 0.48187389633322947,
+      "grad_norm": 0.41826310753822327,
+      "learning_rate": 5.2845703156875106e-05,
+      "loss": 1.7912,
+      "step": 4639
+    },
+    {
+      "epoch": 0.4819777708528098,
+      "grad_norm": 0.3970772325992584,
+      "learning_rate": 5.282941288210146e-05,
+      "loss": 1.6757,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4820816453723902,
+      "grad_norm": 0.4118634760379791,
+      "learning_rate": 5.2813122306017396e-05,
+      "loss": 1.8312,
+      "step": 4641
+    },
+    {
+      "epoch": 0.4821855198919705,
+      "grad_norm": 0.38596856594085693,
+      "learning_rate": 5.279683143035775e-05,
+      "loss": 1.6572,
+      "step": 4642
+    },
+    {
+      "epoch": 0.48228939441155083,
+      "grad_norm": 0.3717338442802429,
+      "learning_rate": 5.278054025685739e-05,
+      "loss": 1.6622,
+      "step": 4643
+    },
+    {
+      "epoch": 0.4823932689311312,
+      "grad_norm": 0.38928812742233276,
+      "learning_rate": 5.276424878725116e-05,
+      "loss": 1.578,
+      "step": 4644
+    },
+    {
+      "epoch": 0.48249714345071154,
+      "grad_norm": 0.4005277156829834,
+      "learning_rate": 5.2747957023274005e-05,
+      "loss": 1.6593,
+      "step": 4645
+    },
+    {
+      "epoch": 0.48260101797029187,
+      "grad_norm": 0.37568166851997375,
+      "learning_rate": 5.2731664966660864e-05,
+      "loss": 1.6828,
+      "step": 4646
+    },
+    {
+      "epoch": 0.48270489248987225,
+      "grad_norm": 0.4337156116962433,
+      "learning_rate": 5.271537261914672e-05,
+      "loss": 1.8237,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4828087670094526,
+      "grad_norm": 0.4006423056125641,
+      "learning_rate": 5.269907998246656e-05,
+      "loss": 1.6222,
+      "step": 4648
+    },
+    {
+      "epoch": 0.4829126415290329,
+      "grad_norm": 0.3779585361480713,
+      "learning_rate": 5.268278705835544e-05,
+      "loss": 1.6672,
+      "step": 4649
+    },
+    {
+      "epoch": 0.4830165160486133,
+      "grad_norm": 0.4423217475414276,
+      "learning_rate": 5.266649384854841e-05,
+      "loss": 1.7731,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4831203905681936,
+      "grad_norm": 0.4306303560733795,
+      "learning_rate": 5.265020035478061e-05,
+      "loss": 1.7711,
+      "step": 4651
+    },
+    {
+      "epoch": 0.483224265087774,
+      "grad_norm": 0.38442516326904297,
+      "learning_rate": 5.263390657878712e-05,
+      "loss": 1.4701,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4833281396073543,
+      "grad_norm": 0.38074007630348206,
+      "learning_rate": 5.2617612522303136e-05,
+      "loss": 1.6486,
+      "step": 4653
+    },
+    {
+      "epoch": 0.48343201412693465,
+      "grad_norm": 0.3747223913669586,
+      "learning_rate": 5.2601318187063855e-05,
+      "loss": 1.6641,
+      "step": 4654
+    },
+    {
+      "epoch": 0.48353588864651503,
+      "grad_norm": 0.48334217071533203,
+      "learning_rate": 5.258502357480445e-05,
+      "loss": 1.5608,
+      "step": 4655
+    },
+    {
+      "epoch": 0.48363976316609536,
+      "grad_norm": 0.4297388792037964,
+      "learning_rate": 5.256872868726021e-05,
+      "loss": 1.6617,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4837436376856757,
+      "grad_norm": 0.433788001537323,
+      "learning_rate": 5.255243352616641e-05,
+      "loss": 1.8598,
+      "step": 4657
+    },
+    {
+      "epoch": 0.48384751220525607,
+      "grad_norm": 0.4389197528362274,
+      "learning_rate": 5.253613809325834e-05,
+      "loss": 1.8134,
+      "step": 4658
+    },
+    {
+      "epoch": 0.4839513867248364,
+      "grad_norm": 0.39111611247062683,
+      "learning_rate": 5.251984239027133e-05,
+      "loss": 1.5018,
+      "step": 4659
+    },
+    {
+      "epoch": 0.4840552612444167,
+      "grad_norm": 0.39344149827957153,
+      "learning_rate": 5.2503546418940784e-05,
+      "loss": 1.7079,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4841591357639971,
+      "grad_norm": 0.42502087354660034,
+      "learning_rate": 5.248725018100207e-05,
+      "loss": 1.7178,
+      "step": 4661
+    },
+    {
+      "epoch": 0.48426301028357743,
+      "grad_norm": 0.4289288818836212,
+      "learning_rate": 5.2470953678190615e-05,
+      "loss": 1.455,
+      "step": 4662
+    },
+    {
+      "epoch": 0.48436688480315776,
+      "grad_norm": 0.41218411922454834,
+      "learning_rate": 5.2454656912241864e-05,
+      "loss": 1.8562,
+      "step": 4663
+    },
+    {
+      "epoch": 0.48447075932273814,
+      "grad_norm": 0.38502606749534607,
+      "learning_rate": 5.2438359884891285e-05,
+      "loss": 1.5738,
+      "step": 4664
+    },
+    {
+      "epoch": 0.48457463384231847,
+      "grad_norm": 0.4311594069004059,
+      "learning_rate": 5.242206259787441e-05,
+      "loss": 1.69,
+      "step": 4665
+    },
+    {
+      "epoch": 0.48467850836189885,
+      "grad_norm": 0.4154076874256134,
+      "learning_rate": 5.2405765052926755e-05,
+      "loss": 1.7648,
+      "step": 4666
+    },
+    {
+      "epoch": 0.4847823828814792,
+      "grad_norm": 0.4217148721218109,
+      "learning_rate": 5.23894672517839e-05,
+      "loss": 1.6886,
+      "step": 4667
+    },
+    {
+      "epoch": 0.4848862574010595,
+      "grad_norm": 0.39737677574157715,
+      "learning_rate": 5.237316919618143e-05,
+      "loss": 1.6023,
+      "step": 4668
+    },
+    {
+      "epoch": 0.4849901319206399,
+      "grad_norm": 0.3839234709739685,
+      "learning_rate": 5.2356870887854926e-05,
+      "loss": 1.7436,
+      "step": 4669
+    },
+    {
+      "epoch": 0.4850940064402202,
+      "grad_norm": 0.39807283878326416,
+      "learning_rate": 5.234057232854006e-05,
+      "loss": 1.7422,
+      "step": 4670
+    },
+    {
+      "epoch": 0.48519788095980054,
+      "grad_norm": 0.40253087878227234,
+      "learning_rate": 5.232427351997251e-05,
+      "loss": 1.6793,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4853017554793809,
+      "grad_norm": 0.41623252630233765,
+      "learning_rate": 5.2307974463887944e-05,
+      "loss": 1.6998,
+      "step": 4672
+    },
+    {
+      "epoch": 0.48540562999896125,
+      "grad_norm": 0.39853668212890625,
+      "learning_rate": 5.229167516202211e-05,
+      "loss": 1.7607,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4855095045185416,
+      "grad_norm": 0.433289110660553,
+      "learning_rate": 5.2275375616110755e-05,
+      "loss": 1.707,
+      "step": 4674
+    },
+    {
+      "epoch": 0.48561337903812196,
+      "grad_norm": 0.37123411893844604,
+      "learning_rate": 5.2259075827889644e-05,
+      "loss": 1.765,
+      "step": 4675
+    },
+    {
+      "epoch": 0.4857172535577023,
+      "grad_norm": 0.3944990634918213,
+      "learning_rate": 5.224277579909458e-05,
+      "loss": 1.6677,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4858211280772826,
+      "grad_norm": 0.38303902745246887,
+      "learning_rate": 5.222647553146138e-05,
+      "loss": 1.5869,
+      "step": 4677
+    },
+    {
+      "epoch": 0.485925002596863,
+      "grad_norm": 0.40630918741226196,
+      "learning_rate": 5.221017502672589e-05,
+      "loss": 1.6596,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4860288771164433,
+      "grad_norm": 0.38045310974121094,
+      "learning_rate": 5.219387428662402e-05,
+      "loss": 1.75,
+      "step": 4679
+    },
+    {
+      "epoch": 0.4861327516360237,
+      "grad_norm": 0.36957138776779175,
+      "learning_rate": 5.2177573312891646e-05,
+      "loss": 1.5169,
+      "step": 4680
+    },
+    {
+      "epoch": 0.48623662615560403,
+      "grad_norm": 0.40403348207473755,
+      "learning_rate": 5.21612721072647e-05,
+      "loss": 1.8167,
+      "step": 4681
+    },
+    {
+      "epoch": 0.48634050067518436,
+      "grad_norm": 0.4342941641807556,
+      "learning_rate": 5.214497067147914e-05,
+      "loss": 1.6114,
+      "step": 4682
+    },
+    {
+      "epoch": 0.48644437519476474,
+      "grad_norm": 0.40665194392204285,
+      "learning_rate": 5.212866900727093e-05,
+      "loss": 1.6924,
+      "step": 4683
+    },
+    {
+      "epoch": 0.48654824971434507,
+      "grad_norm": 0.397088885307312,
+      "learning_rate": 5.211236711637607e-05,
+      "loss": 1.6505,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4866521242339254,
+      "grad_norm": 0.4288082718849182,
+      "learning_rate": 5.209606500053059e-05,
+      "loss": 1.9258,
+      "step": 4685
+    },
+    {
+      "epoch": 0.4867559987535058,
+      "grad_norm": 0.43021494150161743,
+      "learning_rate": 5.207976266147053e-05,
+      "loss": 1.7348,
+      "step": 4686
+    },
+    {
+      "epoch": 0.4868598732730861,
+      "grad_norm": 0.4119536280632019,
+      "learning_rate": 5.206346010093198e-05,
+      "loss": 1.6395,
+      "step": 4687
+    },
+    {
+      "epoch": 0.48696374779266643,
+      "grad_norm": 0.4928002655506134,
+      "learning_rate": 5.204715732065103e-05,
+      "loss": 1.846,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4870676223122468,
+      "grad_norm": 0.3708318769931793,
+      "learning_rate": 5.203085432236379e-05,
+      "loss": 1.4251,
+      "step": 4689
+    },
+    {
+      "epoch": 0.48717149683182714,
+      "grad_norm": 0.3991853892803192,
+      "learning_rate": 5.201455110780641e-05,
+      "loss": 1.6869,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4872753713514075,
+      "grad_norm": 0.4395716190338135,
+      "learning_rate": 5.199824767871503e-05,
+      "loss": 1.8586,
+      "step": 4691
+    },
+    {
+      "epoch": 0.48737924587098785,
+      "grad_norm": 0.4379446804523468,
+      "learning_rate": 5.198194403682588e-05,
+      "loss": 1.6928,
+      "step": 4692
+    },
+    {
+      "epoch": 0.4874831203905682,
+      "grad_norm": 0.4161565601825714,
+      "learning_rate": 5.1965640183875144e-05,
+      "loss": 1.8182,
+      "step": 4693
+    },
+    {
+      "epoch": 0.48758699491014856,
+      "grad_norm": 0.41332149505615234,
+      "learning_rate": 5.194933612159906e-05,
+      "loss": 1.8115,
+      "step": 4694
+    },
+    {
+      "epoch": 0.4876908694297289,
+      "grad_norm": 0.38883647322654724,
+      "learning_rate": 5.193303185173388e-05,
+      "loss": 1.7725,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4877947439493092,
+      "grad_norm": 0.4268670976161957,
+      "learning_rate": 5.191672737601591e-05,
+      "loss": 1.823,
+      "step": 4696
+    },
+    {
+      "epoch": 0.4878986184688896,
+      "grad_norm": 0.4016038775444031,
+      "learning_rate": 5.1900422696181396e-05,
+      "loss": 1.5653,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4880024929884699,
+      "grad_norm": 0.43578583002090454,
+      "learning_rate": 5.1884117813966705e-05,
+      "loss": 1.7653,
+      "step": 4698
+    },
+    {
+      "epoch": 0.48810636750805025,
+      "grad_norm": 0.45143744349479675,
+      "learning_rate": 5.186781273110816e-05,
+      "loss": 1.7976,
+      "step": 4699
+    },
+    {
+      "epoch": 0.48821024202763064,
+      "grad_norm": 0.415283739566803,
+      "learning_rate": 5.1851507449342104e-05,
+      "loss": 1.6952,
+      "step": 4700
+    },
+    {
+      "epoch": 0.48831411654721096,
+      "grad_norm": 0.38803666830062866,
+      "learning_rate": 5.183520197040498e-05,
+      "loss": 1.6622,
+      "step": 4701
+    },
+    {
+      "epoch": 0.4884179910667913,
+      "grad_norm": 0.400741845369339,
+      "learning_rate": 5.1818896296033145e-05,
+      "loss": 1.7436,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4885218655863717,
+      "grad_norm": 0.39219218492507935,
+      "learning_rate": 5.180259042796306e-05,
+      "loss": 1.6381,
+      "step": 4703
+    },
+    {
+      "epoch": 0.488625740105952,
+      "grad_norm": 0.44362738728523254,
+      "learning_rate": 5.178628436793115e-05,
+      "loss": 1.8482,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4887296146255324,
+      "grad_norm": 0.4290272295475006,
+      "learning_rate": 5.176997811767387e-05,
+      "loss": 1.7622,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4888334891451127,
+      "grad_norm": 0.38011881709098816,
+      "learning_rate": 5.175367167892774e-05,
+      "loss": 1.7822,
+      "step": 4706
+    },
+    {
+      "epoch": 0.48893736366469304,
+      "grad_norm": 0.4040461778640747,
+      "learning_rate": 5.1737365053429254e-05,
+      "loss": 1.4567,
+      "step": 4707
+    },
+    {
+      "epoch": 0.4890412381842734,
+      "grad_norm": 0.40654054284095764,
+      "learning_rate": 5.1721058242914935e-05,
+      "loss": 1.7415,
+      "step": 4708
+    },
+    {
+      "epoch": 0.48914511270385375,
+      "grad_norm": 0.3765581548213959,
+      "learning_rate": 5.170475124912134e-05,
+      "loss": 1.6306,
+      "step": 4709
+    },
+    {
+      "epoch": 0.4892489872234341,
+      "grad_norm": 0.4085083305835724,
+      "learning_rate": 5.168844407378506e-05,
+      "loss": 1.6247,
+      "step": 4710
+    },
+    {
+      "epoch": 0.48935286174301446,
+      "grad_norm": 0.4177039861679077,
+      "learning_rate": 5.167213671864263e-05,
+      "loss": 1.6903,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4894567362625948,
+      "grad_norm": 0.449349582195282,
+      "learning_rate": 5.1655829185430685e-05,
+      "loss": 1.6055,
+      "step": 4712
+    },
+    {
+      "epoch": 0.4895606107821751,
+      "grad_norm": 0.4039447605609894,
+      "learning_rate": 5.1639521475885854e-05,
+      "loss": 1.6719,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4896644853017555,
+      "grad_norm": 0.39503759145736694,
+      "learning_rate": 5.162321359174476e-05,
+      "loss": 1.7133,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4897683598213358,
+      "grad_norm": 0.4664006531238556,
+      "learning_rate": 5.16069055347441e-05,
+      "loss": 1.805,
+      "step": 4715
+    },
+    {
+      "epoch": 0.48987223434091615,
+      "grad_norm": 0.39584633708000183,
+      "learning_rate": 5.1590597306620525e-05,
+      "loss": 1.5628,
+      "step": 4716
+    },
+    {
+      "epoch": 0.48997610886049653,
+      "grad_norm": 0.3739977478981018,
+      "learning_rate": 5.1574288909110744e-05,
+      "loss": 1.5699,
+      "step": 4717
+    },
+    {
+      "epoch": 0.49007998338007686,
+      "grad_norm": 0.3966865539550781,
+      "learning_rate": 5.155798034395147e-05,
+      "loss": 1.7159,
+      "step": 4718
+    },
+    {
+      "epoch": 0.49018385789965724,
+      "grad_norm": 0.48280176520347595,
+      "learning_rate": 5.154167161287945e-05,
+      "loss": 1.8649,
+      "step": 4719
+    },
+    {
+      "epoch": 0.49028773241923757,
+      "grad_norm": 0.43992406129837036,
+      "learning_rate": 5.152536271763143e-05,
+      "loss": 1.7603,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4903916069388179,
+      "grad_norm": 0.40175652503967285,
+      "learning_rate": 5.1509053659944175e-05,
+      "loss": 1.664,
+      "step": 4721
+    },
+    {
+      "epoch": 0.4904954814583983,
+      "grad_norm": 0.3840475082397461,
+      "learning_rate": 5.149274444155446e-05,
+      "loss": 1.5613,
+      "step": 4722
+    },
+    {
+      "epoch": 0.4905993559779786,
+      "grad_norm": 0.42482373118400574,
+      "learning_rate": 5.147643506419914e-05,
+      "loss": 1.7978,
+      "step": 4723
+    },
+    {
+      "epoch": 0.49070323049755893,
+      "grad_norm": 0.40053901076316833,
+      "learning_rate": 5.146012552961499e-05,
+      "loss": 1.736,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4908071050171393,
+      "grad_norm": 0.3966452479362488,
+      "learning_rate": 5.1443815839538845e-05,
+      "loss": 1.7367,
+      "step": 4725
+    },
+    {
+      "epoch": 0.49091097953671964,
+      "grad_norm": 0.39208799600601196,
+      "learning_rate": 5.1427505995707594e-05,
+      "loss": 1.4424,
+      "step": 4726
+    },
+    {
+      "epoch": 0.49101485405629997,
+      "grad_norm": 0.42873436212539673,
+      "learning_rate": 5.1411195999858074e-05,
+      "loss": 1.8207,
+      "step": 4727
+    },
+    {
+      "epoch": 0.49111872857588035,
+      "grad_norm": 0.44238024950027466,
+      "learning_rate": 5.1394885853727195e-05,
+      "loss": 1.756,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4912226030954607,
+      "grad_norm": 0.38255882263183594,
+      "learning_rate": 5.137857555905184e-05,
+      "loss": 1.5362,
+      "step": 4729
+    },
+    {
+      "epoch": 0.49132647761504106,
+      "grad_norm": 0.4030906558036804,
+      "learning_rate": 5.1362265117568965e-05,
+      "loss": 1.6924,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4914303521346214,
+      "grad_norm": 0.4658985733985901,
+      "learning_rate": 5.134595453101547e-05,
+      "loss": 1.7965,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4915342266542017,
+      "grad_norm": 0.3776882290840149,
+      "learning_rate": 5.1329643801128324e-05,
+      "loss": 1.5735,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4916381011737821,
+      "grad_norm": 0.3989698588848114,
+      "learning_rate": 5.131333292964447e-05,
+      "loss": 1.7203,
+      "step": 4733
+    },
+    {
+      "epoch": 0.4917419756933624,
+      "grad_norm": 0.4176570177078247,
+      "learning_rate": 5.12970219183009e-05,
+      "loss": 1.7932,
+      "step": 4734
+    },
+    {
+      "epoch": 0.49184585021294275,
+      "grad_norm": 0.42140617966651917,
+      "learning_rate": 5.128071076883463e-05,
+      "loss": 1.8202,
+      "step": 4735
+    },
+    {
+      "epoch": 0.49194972473252313,
+      "grad_norm": 0.41681942343711853,
+      "learning_rate": 5.126439948298264e-05,
+      "loss": 1.6721,
+      "step": 4736
+    },
+    {
+      "epoch": 0.49205359925210346,
+      "grad_norm": 0.465843141078949,
+      "learning_rate": 5.124808806248199e-05,
+      "loss": 1.6901,
+      "step": 4737
+    },
+    {
+      "epoch": 0.4921574737716838,
+      "grad_norm": 0.41670602560043335,
+      "learning_rate": 5.123177650906968e-05,
+      "loss": 1.745,
+      "step": 4738
+    },
+    {
+      "epoch": 0.49226134829126417,
+      "grad_norm": 0.415339857339859,
+      "learning_rate": 5.1215464824482775e-05,
+      "loss": 1.7013,
+      "step": 4739
+    },
+    {
+      "epoch": 0.4923652228108445,
+      "grad_norm": 0.3894515931606293,
+      "learning_rate": 5.119915301045836e-05,
+      "loss": 1.5359,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4924690973304248,
+      "grad_norm": 0.37234368920326233,
+      "learning_rate": 5.118284106873349e-05,
+      "loss": 1.6351,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4925729718500052,
+      "grad_norm": 0.47200119495391846,
+      "learning_rate": 5.116652900104527e-05,
+      "loss": 1.9368,
+      "step": 4742
+    },
+    {
+      "epoch": 0.49267684636958553,
+      "grad_norm": 0.430601567029953,
+      "learning_rate": 5.1150216809130826e-05,
+      "loss": 1.6977,
+      "step": 4743
+    },
+    {
+      "epoch": 0.4927807208891659,
+      "grad_norm": 0.4601489305496216,
+      "learning_rate": 5.113390449472726e-05,
+      "loss": 1.8549,
+      "step": 4744
+    },
+    {
+      "epoch": 0.49288459540874624,
+      "grad_norm": 0.42277416586875916,
+      "learning_rate": 5.11175920595717e-05,
+      "loss": 1.7991,
+      "step": 4745
+    },
+    {
+      "epoch": 0.49298846992832657,
+      "grad_norm": 0.3993331789970398,
+      "learning_rate": 5.1101279505401325e-05,
+      "loss": 1.7089,
+      "step": 4746
+    },
+    {
+      "epoch": 0.49309234444790695,
+      "grad_norm": 0.4037160277366638,
+      "learning_rate": 5.108496683395324e-05,
+      "loss": 1.6423,
+      "step": 4747
+    },
+    {
+      "epoch": 0.4931962189674873,
+      "grad_norm": 0.3775770664215088,
+      "learning_rate": 5.106865404696468e-05,
+      "loss": 1.6201,
+      "step": 4748
+    },
+    {
+      "epoch": 0.4933000934870676,
+      "grad_norm": 0.4403548836708069,
+      "learning_rate": 5.1052341146172785e-05,
+      "loss": 1.7299,
+      "step": 4749
+    },
+    {
+      "epoch": 0.493403968006648,
+      "grad_norm": 0.4192941188812256,
+      "learning_rate": 5.103602813331476e-05,
+      "loss": 1.7541,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4935078425262283,
+      "grad_norm": 0.3859345018863678,
+      "learning_rate": 5.101971501012785e-05,
+      "loss": 1.3569,
+      "step": 4751
+    },
+    {
+      "epoch": 0.49361171704580864,
+      "grad_norm": 0.4286668598651886,
+      "learning_rate": 5.1003401778349224e-05,
+      "loss": 1.8115,
+      "step": 4752
+    },
+    {
+      "epoch": 0.493715591565389,
+      "grad_norm": 0.4297914206981659,
+      "learning_rate": 5.0987088439716134e-05,
+      "loss": 1.6486,
+      "step": 4753
+    },
+    {
+      "epoch": 0.49381946608496935,
+      "grad_norm": 0.4609547555446625,
+      "learning_rate": 5.097077499596583e-05,
+      "loss": 1.8906,
+      "step": 4754
+    },
+    {
+      "epoch": 0.4939233406045497,
+      "grad_norm": 0.3923378884792328,
+      "learning_rate": 5.095446144883556e-05,
+      "loss": 1.5989,
+      "step": 4755
+    },
+    {
+      "epoch": 0.49402721512413006,
+      "grad_norm": 0.42669469118118286,
+      "learning_rate": 5.0938147800062586e-05,
+      "loss": 1.5584,
+      "step": 4756
+    },
+    {
+      "epoch": 0.4941310896437104,
+      "grad_norm": 0.4073004126548767,
+      "learning_rate": 5.092183405138419e-05,
+      "loss": 1.6068,
+      "step": 4757
+    },
+    {
+      "epoch": 0.49423496416329077,
+      "grad_norm": 0.4162704050540924,
+      "learning_rate": 5.090552020453766e-05,
+      "loss": 1.8445,
+      "step": 4758
+    },
+    {
+      "epoch": 0.4943388386828711,
+      "grad_norm": 0.45715436339378357,
+      "learning_rate": 5.08892062612603e-05,
+      "loss": 1.8761,
+      "step": 4759
+    },
+    {
+      "epoch": 0.4944427132024514,
+      "grad_norm": 0.38431254029273987,
+      "learning_rate": 5.08728922232894e-05,
+      "loss": 1.5639,
+      "step": 4760
+    },
+    {
+      "epoch": 0.4945465877220318,
+      "grad_norm": 0.4815351963043213,
+      "learning_rate": 5.0856578092362285e-05,
+      "loss": 1.7764,
+      "step": 4761
+    },
+    {
+      "epoch": 0.49465046224161213,
+      "grad_norm": 0.41645729541778564,
+      "learning_rate": 5.084026387021629e-05,
+      "loss": 1.6924,
+      "step": 4762
+    },
+    {
+      "epoch": 0.49475433676119246,
+      "grad_norm": 0.4079589247703552,
+      "learning_rate": 5.0823949558588746e-05,
+      "loss": 1.7549,
+      "step": 4763
+    },
+    {
+      "epoch": 0.49485821128077284,
+      "grad_norm": 0.40971922874450684,
+      "learning_rate": 5.0807635159216993e-05,
+      "loss": 1.5389,
+      "step": 4764
+    },
+    {
+      "epoch": 0.49496208580035317,
+      "grad_norm": 0.41493481397628784,
+      "learning_rate": 5.079132067383842e-05,
+      "loss": 1.7683,
+      "step": 4765
+    },
+    {
+      "epoch": 0.4950659603199335,
+      "grad_norm": 0.3941478729248047,
+      "learning_rate": 5.077500610419036e-05,
+      "loss": 1.7123,
+      "step": 4766
+    },
+    {
+      "epoch": 0.4951698348395139,
+      "grad_norm": 0.4111056327819824,
+      "learning_rate": 5.0758691452010175e-05,
+      "loss": 1.7361,
+      "step": 4767
+    },
+    {
+      "epoch": 0.4952737093590942,
+      "grad_norm": 0.4429363012313843,
+      "learning_rate": 5.0742376719035276e-05,
+      "loss": 1.8205,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4953775838786746,
+      "grad_norm": 0.3898932635784149,
+      "learning_rate": 5.0726061907003045e-05,
+      "loss": 1.6695,
+      "step": 4769
+    },
+    {
+      "epoch": 0.4954814583982549,
+      "grad_norm": 0.43110302090644836,
+      "learning_rate": 5.070974701765089e-05,
+      "loss": 1.6963,
+      "step": 4770
+    },
+    {
+      "epoch": 0.49558533291783524,
+      "grad_norm": 0.4207279682159424,
+      "learning_rate": 5.069343205271622e-05,
+      "loss": 1.5213,
+      "step": 4771
+    },
+    {
+      "epoch": 0.4956892074374156,
+      "grad_norm": 0.423676073551178,
+      "learning_rate": 5.067711701393644e-05,
+      "loss": 1.5632,
+      "step": 4772
+    },
+    {
+      "epoch": 0.49579308195699595,
+      "grad_norm": 0.39001235365867615,
+      "learning_rate": 5.066080190304898e-05,
+      "loss": 1.717,
+      "step": 4773
+    },
+    {
+      "epoch": 0.4958969564765763,
+      "grad_norm": 0.3835489749908447,
+      "learning_rate": 5.0644486721791273e-05,
+      "loss": 1.7107,
+      "step": 4774
+    },
+    {
+      "epoch": 0.49600083099615666,
+      "grad_norm": 0.4401542842388153,
+      "learning_rate": 5.0628171471900756e-05,
+      "loss": 1.8646,
+      "step": 4775
+    },
+    {
+      "epoch": 0.496104705515737,
+      "grad_norm": 0.38474228978157043,
+      "learning_rate": 5.0611856155114876e-05,
+      "loss": 1.6334,
+      "step": 4776
+    },
+    {
+      "epoch": 0.4962085800353173,
+      "grad_norm": 0.4321017563343048,
+      "learning_rate": 5.059554077317109e-05,
+      "loss": 1.7811,
+      "step": 4777
+    },
+    {
+      "epoch": 0.4963124545548977,
+      "grad_norm": 0.38683265447616577,
+      "learning_rate": 5.0579225327806847e-05,
+      "loss": 1.6125,
+      "step": 4778
+    },
+    {
+      "epoch": 0.496416329074478,
+      "grad_norm": 0.3820945918560028,
+      "learning_rate": 5.0562909820759654e-05,
+      "loss": 1.554,
+      "step": 4779
+    },
+    {
+      "epoch": 0.49652020359405835,
+      "grad_norm": 0.38485878705978394,
+      "learning_rate": 5.054659425376693e-05,
+      "loss": 1.6951,
+      "step": 4780
+    },
+    {
+      "epoch": 0.49662407811363873,
+      "grad_norm": 0.40460067987442017,
+      "learning_rate": 5.0530278628566175e-05,
+      "loss": 1.7489,
+      "step": 4781
+    },
+    {
+      "epoch": 0.49672795263321906,
+      "grad_norm": 0.4411293566226959,
+      "learning_rate": 5.05139629468949e-05,
+      "loss": 1.7431,
+      "step": 4782
+    },
+    {
+      "epoch": 0.49683182715279944,
+      "grad_norm": 0.41115111112594604,
+      "learning_rate": 5.0497647210490565e-05,
+      "loss": 1.5739,
+      "step": 4783
+    },
+    {
+      "epoch": 0.49693570167237977,
+      "grad_norm": 0.4054515063762665,
+      "learning_rate": 5.048133142109069e-05,
+      "loss": 1.6096,
+      "step": 4784
+    },
+    {
+      "epoch": 0.4970395761919601,
+      "grad_norm": 0.45801469683647156,
+      "learning_rate": 5.046501558043278e-05,
+      "loss": 1.8216,
+      "step": 4785
+    },
+    {
+      "epoch": 0.4971434507115405,
+      "grad_norm": 0.3939175009727478,
+      "learning_rate": 5.044869969025434e-05,
+      "loss": 1.6871,
+      "step": 4786
+    },
+    {
+      "epoch": 0.4972473252311208,
+      "grad_norm": 0.38540032505989075,
+      "learning_rate": 5.043238375229289e-05,
+      "loss": 1.6358,
+      "step": 4787
+    },
+    {
+      "epoch": 0.49735119975070113,
+      "grad_norm": 0.4039050340652466,
+      "learning_rate": 5.0416067768285924e-05,
+      "loss": 1.6841,
+      "step": 4788
+    },
+    {
+      "epoch": 0.4974550742702815,
+      "grad_norm": 0.3593752384185791,
+      "learning_rate": 5.039975173997099e-05,
+      "loss": 1.4662,
+      "step": 4789
+    },
+    {
+      "epoch": 0.49755894878986184,
+      "grad_norm": 0.4602617919445038,
+      "learning_rate": 5.038343566908562e-05,
+      "loss": 1.8741,
+      "step": 4790
+    },
+    {
+      "epoch": 0.49766282330944217,
+      "grad_norm": 0.39395564794540405,
+      "learning_rate": 5.036711955736734e-05,
+      "loss": 1.6235,
+      "step": 4791
+    },
+    {
+      "epoch": 0.49776669782902255,
+      "grad_norm": 0.3926306664943695,
+      "learning_rate": 5.0350803406553684e-05,
+      "loss": 1.4756,
+      "step": 4792
+    },
+    {
+      "epoch": 0.4978705723486029,
+      "grad_norm": 0.4372478723526001,
+      "learning_rate": 5.033448721838222e-05,
+      "loss": 1.6785,
+      "step": 4793
+    },
+    {
+      "epoch": 0.4979744468681832,
+      "grad_norm": 0.41775140166282654,
+      "learning_rate": 5.0318170994590464e-05,
+      "loss": 1.6019,
+      "step": 4794
+    },
+    {
+      "epoch": 0.4980783213877636,
+      "grad_norm": 0.4097503423690796,
+      "learning_rate": 5.030185473691597e-05,
+      "loss": 1.8253,
+      "step": 4795
+    },
+    {
+      "epoch": 0.4981821959073439,
+      "grad_norm": 0.38628101348876953,
+      "learning_rate": 5.0285538447096315e-05,
+      "loss": 1.6124,
+      "step": 4796
+    },
+    {
+      "epoch": 0.4982860704269243,
+      "grad_norm": 0.4230360984802246,
+      "learning_rate": 5.026922212686902e-05,
+      "loss": 1.67,
+      "step": 4797
+    },
+    {
+      "epoch": 0.4983899449465046,
+      "grad_norm": 0.40452295541763306,
+      "learning_rate": 5.025290577797169e-05,
+      "loss": 1.6923,
+      "step": 4798
+    },
+    {
+      "epoch": 0.49849381946608495,
+      "grad_norm": 0.38569211959838867,
+      "learning_rate": 5.0236589402141864e-05,
+      "loss": 1.585,
+      "step": 4799
+    },
+    {
+      "epoch": 0.49859769398566534,
+      "grad_norm": 0.4305466413497925,
+      "learning_rate": 5.022027300111711e-05,
+      "loss": 1.4631,
+      "step": 4800
+    },
+    {
+      "epoch": 0.49870156850524566,
+      "grad_norm": 0.42307230830192566,
+      "learning_rate": 5.0203956576635015e-05,
+      "loss": 1.6969,
+      "step": 4801
+    },
+    {
+      "epoch": 0.498805443024826,
+      "grad_norm": 0.44654223322868347,
+      "learning_rate": 5.018764013043312e-05,
+      "loss": 1.8514,
+      "step": 4802
+    },
+    {
+      "epoch": 0.49890931754440637,
+      "grad_norm": 0.4396657943725586,
+      "learning_rate": 5.0171323664249005e-05,
+      "loss": 1.7264,
+      "step": 4803
+    },
+    {
+      "epoch": 0.4990131920639867,
+      "grad_norm": 0.39395204186439514,
+      "learning_rate": 5.0155007179820266e-05,
+      "loss": 1.6963,
+      "step": 4804
+    },
+    {
+      "epoch": 0.499117066583567,
+      "grad_norm": 0.4187118411064148,
+      "learning_rate": 5.013869067888448e-05,
+      "loss": 1.7839,
+      "step": 4805
+    },
+    {
+      "epoch": 0.4992209411031474,
+      "grad_norm": 0.40489673614501953,
+      "learning_rate": 5.01223741631792e-05,
+      "loss": 1.6467,
+      "step": 4806
+    },
+    {
+      "epoch": 0.49932481562272774,
+      "grad_norm": 0.41493701934814453,
+      "learning_rate": 5.010605763444205e-05,
+      "loss": 1.8405,
+      "step": 4807
+    },
+    {
+      "epoch": 0.4994286901423081,
+      "grad_norm": 0.4441494643688202,
+      "learning_rate": 5.008974109441057e-05,
+      "loss": 1.6736,
+      "step": 4808
+    },
+    {
+      "epoch": 0.49953256466188845,
+      "grad_norm": 0.38821157813072205,
+      "learning_rate": 5.007342454482238e-05,
+      "loss": 1.6041,
+      "step": 4809
+    },
+    {
+      "epoch": 0.49963643918146877,
+      "grad_norm": 0.4361104965209961,
+      "learning_rate": 5.005710798741503e-05,
+      "loss": 1.7678,
+      "step": 4810
+    },
+    {
+      "epoch": 0.49974031370104915,
+      "grad_norm": 0.39289024472236633,
+      "learning_rate": 5.0040791423926125e-05,
+      "loss": 1.5475,
+      "step": 4811
+    },
+    {
+      "epoch": 0.4998441882206295,
+      "grad_norm": 0.35669323801994324,
+      "learning_rate": 5.002447485609327e-05,
+      "loss": 1.4978,
+      "step": 4812
+    },
+    {
+      "epoch": 0.4999480627402098,
+      "grad_norm": 0.38865354657173157,
+      "learning_rate": 5.0008158285654027e-05,
+      "loss": 1.7025,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5000519372597901,
+      "grad_norm": 0.42419496178627014,
+      "learning_rate": 4.9991841714345985e-05,
+      "loss": 1.7601,
+      "step": 4814
+    },
+    {
+      "epoch": 0.5001558117793705,
+      "grad_norm": 0.39531058073043823,
+      "learning_rate": 4.997552514390674e-05,
+      "loss": 1.5614,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5002596862989509,
+      "grad_norm": 0.415740430355072,
+      "learning_rate": 4.9959208576073866e-05,
+      "loss": 1.7504,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5003635608185312,
+      "grad_norm": 0.4087817966938019,
+      "learning_rate": 4.994289201258498e-05,
+      "loss": 1.6628,
+      "step": 4817
+    },
+    {
+      "epoch": 0.5004674353381116,
+      "grad_norm": 0.39313584566116333,
+      "learning_rate": 4.992657545517764e-05,
+      "loss": 1.7745,
+      "step": 4818
+    },
+    {
+      "epoch": 0.5005713098576919,
+      "grad_norm": 0.3950080871582031,
+      "learning_rate": 4.9910258905589427e-05,
+      "loss": 1.6339,
+      "step": 4819
+    },
+    {
+      "epoch": 0.5006751843772722,
+      "grad_norm": 0.37852367758750916,
+      "learning_rate": 4.989394236555797e-05,
+      "loss": 1.6569,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5007790588968526,
+      "grad_norm": 0.43450725078582764,
+      "learning_rate": 4.9877625836820806e-05,
+      "loss": 1.602,
+      "step": 4821
+    },
+    {
+      "epoch": 0.500882933416433,
+      "grad_norm": 0.42665791511535645,
+      "learning_rate": 4.986130932111553e-05,
+      "loss": 1.6783,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5009868079360132,
+      "grad_norm": 0.396518737077713,
+      "learning_rate": 4.984499282017975e-05,
+      "loss": 1.6118,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5010906824555936,
+      "grad_norm": 0.4622015655040741,
+      "learning_rate": 4.9828676335751006e-05,
+      "loss": 1.8231,
+      "step": 4824
+    },
+    {
+      "epoch": 0.501194556975174,
+      "grad_norm": 0.3886905908584595,
+      "learning_rate": 4.9812359869566896e-05,
+      "loss": 1.667,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5012984314947543,
+      "grad_norm": 0.4054078161716461,
+      "learning_rate": 4.979604342336501e-05,
+      "loss": 1.6415,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5014023060143347,
+      "grad_norm": 0.38207703828811646,
+      "learning_rate": 4.977972699888289e-05,
+      "loss": 1.7066,
+      "step": 4827
+    },
+    {
+      "epoch": 0.501506180533915,
+      "grad_norm": 0.43799397349357605,
+      "learning_rate": 4.976341059785815e-05,
+      "loss": 1.6469,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5016100550534954,
+      "grad_norm": 0.5063971281051636,
+      "learning_rate": 4.974709422202831e-05,
+      "loss": 1.9436,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5017139295730757,
+      "grad_norm": 0.39013898372650146,
+      "learning_rate": 4.973077787313099e-05,
+      "loss": 1.5765,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5018178040926561,
+      "grad_norm": 0.4318859875202179,
+      "learning_rate": 4.9714461552903704e-05,
+      "loss": 1.8306,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5019216786122365,
+      "grad_norm": 0.4619205892086029,
+      "learning_rate": 4.9698145263084036e-05,
+      "loss": 1.7797,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5020255531318167,
+      "grad_norm": 0.3963909447193146,
+      "learning_rate": 4.968182900540956e-05,
+      "loss": 1.646,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5021294276513971,
+      "grad_norm": 0.3879808187484741,
+      "learning_rate": 4.966551278161779e-05,
+      "loss": 1.6855,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5022333021709775,
+      "grad_norm": 0.393045037984848,
+      "learning_rate": 4.964919659344632e-05,
+      "loss": 1.678,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5023371766905578,
+      "grad_norm": 0.38027775287628174,
+      "learning_rate": 4.963288044263268e-05,
+      "loss": 1.6203,
+      "step": 4836
+    },
+    {
+      "epoch": 0.5024410512101382,
+      "grad_norm": 0.4136490821838379,
+      "learning_rate": 4.96165643309144e-05,
+      "loss": 1.772,
+      "step": 4837
+    },
+    {
+      "epoch": 0.5025449257297185,
+      "grad_norm": 0.3826773762702942,
+      "learning_rate": 4.9600248260029016e-05,
+      "loss": 1.8326,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5026488002492988,
+      "grad_norm": 0.4194203317165375,
+      "learning_rate": 4.9583932231714095e-05,
+      "loss": 1.5652,
+      "step": 4839
+    },
+    {
+      "epoch": 0.5027526747688792,
+      "grad_norm": 0.4387541115283966,
+      "learning_rate": 4.956761624770713e-05,
+      "loss": 1.8814,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5028565492884596,
+      "grad_norm": 0.3808096945285797,
+      "learning_rate": 4.9551300309745666e-05,
+      "loss": 1.748,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5029604238080398,
+      "grad_norm": 0.4226476550102234,
+      "learning_rate": 4.953498441956723e-05,
+      "loss": 1.6055,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5030642983276202,
+      "grad_norm": 0.4065408706665039,
+      "learning_rate": 4.951866857890931e-05,
+      "loss": 1.7516,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5031681728472006,
+      "grad_norm": 0.3814634084701538,
+      "learning_rate": 4.950235278950943e-05,
+      "loss": 1.521,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5032720473667809,
+      "grad_norm": 0.4002365469932556,
+      "learning_rate": 4.9486037053105125e-05,
+      "loss": 1.7387,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5033759218863613,
+      "grad_norm": 0.39226168394088745,
+      "learning_rate": 4.946972137143383e-05,
+      "loss": 1.6565,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5034797964059416,
+      "grad_norm": 0.425890177488327,
+      "learning_rate": 4.9453405746233076e-05,
+      "loss": 1.7404,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5035836709255219,
+      "grad_norm": 0.41715189814567566,
+      "learning_rate": 4.943709017924038e-05,
+      "loss": 1.7237,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5036875454451023,
+      "grad_norm": 0.4025704860687256,
+      "learning_rate": 4.9420774672193165e-05,
+      "loss": 1.4753,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5037914199646827,
+      "grad_norm": 0.4412018060684204,
+      "learning_rate": 4.940445922682892e-05,
+      "loss": 1.9151,
+      "step": 4850
+    },
+    {
+      "epoch": 0.503895294484263,
+      "grad_norm": 0.3976673483848572,
+      "learning_rate": 4.938814384488515e-05,
+      "loss": 1.707,
+      "step": 4851
+    },
+    {
+      "epoch": 0.5039991690038433,
+      "grad_norm": 0.40271249413490295,
+      "learning_rate": 4.937182852809926e-05,
+      "loss": 1.658,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5041030435234237,
+      "grad_norm": 0.4666357934474945,
+      "learning_rate": 4.935551327820873e-05,
+      "loss": 1.8472,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5042069180430041,
+      "grad_norm": 0.42836159467697144,
+      "learning_rate": 4.933919809695103e-05,
+      "loss": 1.7557,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5043107925625844,
+      "grad_norm": 0.44381582736968994,
+      "learning_rate": 4.9322882986063566e-05,
+      "loss": 1.7145,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5044146670821648,
+      "grad_norm": 0.4695770740509033,
+      "learning_rate": 4.9306567947283785e-05,
+      "loss": 1.5709,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5045185416017451,
+      "grad_norm": 0.40107208490371704,
+      "learning_rate": 4.92902529823491e-05,
+      "loss": 1.6956,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5046224161213254,
+      "grad_norm": 0.39944395422935486,
+      "learning_rate": 4.927393809299696e-05,
+      "loss": 1.6442,
+      "step": 4858
+    },
+    {
+      "epoch": 0.5047262906409058,
+      "grad_norm": 0.42077764868736267,
+      "learning_rate": 4.925762328096473e-05,
+      "loss": 1.7561,
+      "step": 4859
+    },
+    {
+      "epoch": 0.5048301651604862,
+      "grad_norm": 0.4306979477405548,
+      "learning_rate": 4.924130854798983e-05,
+      "loss": 1.6888,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5049340396800664,
+      "grad_norm": 0.41281387209892273,
+      "learning_rate": 4.922499389580967e-05,
+      "loss": 1.8061,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5050379141996468,
+      "grad_norm": 0.4295770227909088,
+      "learning_rate": 4.920867932616159e-05,
+      "loss": 1.7401,
+      "step": 4862
+    },
+    {
+      "epoch": 0.5051417887192272,
+      "grad_norm": 0.37192273139953613,
+      "learning_rate": 4.9192364840783e-05,
+      "loss": 1.5993,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5052456632388075,
+      "grad_norm": 0.4081692695617676,
+      "learning_rate": 4.917605044141127e-05,
+      "loss": 1.7359,
+      "step": 4864
+    },
+    {
+      "epoch": 0.5053495377583879,
+      "grad_norm": 0.40556600689888,
+      "learning_rate": 4.915973612978372e-05,
+      "loss": 1.6507,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5054534122779683,
+      "grad_norm": 0.41496628522872925,
+      "learning_rate": 4.914342190763771e-05,
+      "loss": 1.6881,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5055572867975485,
+      "grad_norm": 0.3876533806324005,
+      "learning_rate": 4.912710777671062e-05,
+      "loss": 1.5866,
+      "step": 4867
+    },
+    {
+      "epoch": 0.5056611613171289,
+      "grad_norm": 0.3975103795528412,
+      "learning_rate": 4.9110793738739715e-05,
+      "loss": 1.7478,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5057650358367093,
+      "grad_norm": 0.37800195813179016,
+      "learning_rate": 4.909447979546235e-05,
+      "loss": 1.6304,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5058689103562896,
+      "grad_norm": 0.3874056041240692,
+      "learning_rate": 4.907816594861582e-05,
+      "loss": 1.6151,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5059727848758699,
+      "grad_norm": 0.42271357774734497,
+      "learning_rate": 4.906185219993743e-05,
+      "loss": 1.7042,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5060766593954503,
+      "grad_norm": 0.44538614153862,
+      "learning_rate": 4.904553855116445e-05,
+      "loss": 1.7608,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5061805339150306,
+      "grad_norm": 0.41812676191329956,
+      "learning_rate": 4.90292250040342e-05,
+      "loss": 1.7,
+      "step": 4873
+    },
+    {
+      "epoch": 0.506284408434611,
+      "grad_norm": 0.3845331370830536,
+      "learning_rate": 4.901291156028388e-05,
+      "loss": 1.5532,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5063882829541914,
+      "grad_norm": 0.4065440595149994,
+      "learning_rate": 4.899659822165079e-05,
+      "loss": 1.7019,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5064921574737716,
+      "grad_norm": 0.42397770285606384,
+      "learning_rate": 4.898028498987217e-05,
+      "loss": 1.7414,
+      "step": 4876
+    },
+    {
+      "epoch": 0.506596031993352,
+      "grad_norm": 0.38539576530456543,
+      "learning_rate": 4.8963971866685245e-05,
+      "loss": 1.7377,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5066999065129324,
+      "grad_norm": 0.373605877161026,
+      "learning_rate": 4.894765885382722e-05,
+      "loss": 1.6698,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5068037810325128,
+      "grad_norm": 0.36256158351898193,
+      "learning_rate": 4.8931345953035326e-05,
+      "loss": 1.5029,
+      "step": 4879
+    },
+    {
+      "epoch": 0.506907655552093,
+      "grad_norm": 0.39593711495399475,
+      "learning_rate": 4.891503316604677e-05,
+      "loss": 1.7055,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5070115300716734,
+      "grad_norm": 0.3991319239139557,
+      "learning_rate": 4.889872049459869e-05,
+      "loss": 1.6002,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5071154045912538,
+      "grad_norm": 0.3885091543197632,
+      "learning_rate": 4.88824079404283e-05,
+      "loss": 1.6086,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5072192791108341,
+      "grad_norm": 0.41479891538619995,
+      "learning_rate": 4.8866095505272754e-05,
+      "loss": 1.4958,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5073231536304145,
+      "grad_norm": 0.4320788085460663,
+      "learning_rate": 4.8849783190869186e-05,
+      "loss": 1.6549,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5074270281499949,
+      "grad_norm": 0.49210458993911743,
+      "learning_rate": 4.883347099895473e-05,
+      "loss": 1.9594,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5075309026695751,
+      "grad_norm": 0.38703057169914246,
+      "learning_rate": 4.8817158931266534e-05,
+      "loss": 1.5258,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5076347771891555,
+      "grad_norm": 0.387260764837265,
+      "learning_rate": 4.880084698954166e-05,
+      "loss": 1.5779,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5077386517087359,
+      "grad_norm": 0.42436152696609497,
+      "learning_rate": 4.878453517551722e-05,
+      "loss": 1.874,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5078425262283162,
+      "grad_norm": 0.3863707184791565,
+      "learning_rate": 4.876822349093034e-05,
+      "loss": 1.4991,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5079464007478965,
+      "grad_norm": 0.40059852600097656,
+      "learning_rate": 4.875191193751802e-05,
+      "loss": 1.5891,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5080502752674769,
+      "grad_norm": 0.45580723881721497,
+      "learning_rate": 4.873560051701736e-05,
+      "loss": 1.8611,
+      "step": 4891
+    },
+    {
+      "epoch": 0.5081541497870572,
+      "grad_norm": 0.3890886902809143,
+      "learning_rate": 4.871928923116539e-05,
+      "loss": 1.5606,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5082580243066376,
+      "grad_norm": 0.42675158381462097,
+      "learning_rate": 4.8702978081699104e-05,
+      "loss": 1.7001,
+      "step": 4893
+    },
+    {
+      "epoch": 0.508361898826218,
+      "grad_norm": 0.44091951847076416,
+      "learning_rate": 4.868666707035554e-05,
+      "loss": 1.8272,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5084657733457982,
+      "grad_norm": 0.3957894444465637,
+      "learning_rate": 4.86703561988717e-05,
+      "loss": 1.8023,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5085696478653786,
+      "grad_norm": 0.37731900811195374,
+      "learning_rate": 4.865404546898454e-05,
+      "loss": 1.6024,
+      "step": 4896
+    },
+    {
+      "epoch": 0.508673522384959,
+      "grad_norm": 0.41856059432029724,
+      "learning_rate": 4.863773488243105e-05,
+      "loss": 1.6815,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5087773969045393,
+      "grad_norm": 0.45765843987464905,
+      "learning_rate": 4.862142444094816e-05,
+      "loss": 1.8893,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5088812714241197,
+      "grad_norm": 0.4164111018180847,
+      "learning_rate": 4.8605114146272824e-05,
+      "loss": 1.6191,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5089851459437,
+      "grad_norm": 0.4314156770706177,
+      "learning_rate": 4.858880400014194e-05,
+      "loss": 1.9152,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5090890204632803,
+      "grad_norm": 0.42690524458885193,
+      "learning_rate": 4.857249400429244e-05,
+      "loss": 1.787,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5091928949828607,
+      "grad_norm": 0.3987329602241516,
+      "learning_rate": 4.8556184160461167e-05,
+      "loss": 1.6278,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5092967695024411,
+      "grad_norm": 0.427369624376297,
+      "learning_rate": 4.8539874470385024e-05,
+      "loss": 1.8381,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5094006440220213,
+      "grad_norm": 0.39449864625930786,
+      "learning_rate": 4.852356493580088e-05,
+      "loss": 1.7702,
+      "step": 4904
+    },
+    {
+      "epoch": 0.5095045185416017,
+      "grad_norm": 0.4443662464618683,
+      "learning_rate": 4.850725555844555e-05,
+      "loss": 1.8187,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5096083930611821,
+      "grad_norm": 0.42031386494636536,
+      "learning_rate": 4.8490946340055837e-05,
+      "loss": 1.7612,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5097122675807625,
+      "grad_norm": 0.44448190927505493,
+      "learning_rate": 4.8474637282368577e-05,
+      "loss": 1.6873,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5098161421003428,
+      "grad_norm": 0.4955507516860962,
+      "learning_rate": 4.8458328387120565e-05,
+      "loss": 2.1091,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5099200166199231,
+      "grad_norm": 0.3980197012424469,
+      "learning_rate": 4.844201965604853e-05,
+      "loss": 1.6583,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5100238911395035,
+      "grad_norm": 0.41697457432746887,
+      "learning_rate": 4.842571109088927e-05,
+      "loss": 1.5176,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5101277656590838,
+      "grad_norm": 0.412379652261734,
+      "learning_rate": 4.840940269337949e-05,
+      "loss": 1.5699,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5102316401786642,
+      "grad_norm": 0.420980840921402,
+      "learning_rate": 4.839309446525592e-05,
+      "loss": 1.6266,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5103355146982446,
+      "grad_norm": 0.4162856340408325,
+      "learning_rate": 4.837678640825524e-05,
+      "loss": 1.6759,
+      "step": 4913
+    },
+    {
+      "epoch": 0.5104393892178248,
+      "grad_norm": 0.4442167580127716,
+      "learning_rate": 4.836047852411417e-05,
+      "loss": 1.7818,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5105432637374052,
+      "grad_norm": 0.41404595971107483,
+      "learning_rate": 4.834417081456933e-05,
+      "loss": 1.7876,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5106471382569856,
+      "grad_norm": 0.4493081271648407,
+      "learning_rate": 4.8327863281357376e-05,
+      "loss": 1.9106,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5107510127765659,
+      "grad_norm": 0.379351407289505,
+      "learning_rate": 4.8311555926214966e-05,
+      "loss": 1.6729,
+      "step": 4917
+    },
+    {
+      "epoch": 0.5108548872961463,
+      "grad_norm": 0.40297406911849976,
+      "learning_rate": 4.829524875087867e-05,
+      "loss": 1.7575,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5109587618157266,
+      "grad_norm": 0.4127398133277893,
+      "learning_rate": 4.8278941757085063e-05,
+      "loss": 1.5132,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5110626363353069,
+      "grad_norm": 0.38783156871795654,
+      "learning_rate": 4.8262634946570764e-05,
+      "loss": 1.6706,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5111665108548873,
+      "grad_norm": 0.39855116605758667,
+      "learning_rate": 4.824632832107227e-05,
+      "loss": 1.7722,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5112703853744677,
+      "grad_norm": 0.38658004999160767,
+      "learning_rate": 4.8230021882326135e-05,
+      "loss": 1.7226,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5113742598940479,
+      "grad_norm": 0.4114481210708618,
+      "learning_rate": 4.821371563206888e-05,
+      "loss": 1.7858,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5114781344136283,
+      "grad_norm": 0.3694150745868683,
+      "learning_rate": 4.819740957203696e-05,
+      "loss": 1.4821,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5115820089332087,
+      "grad_norm": 0.40538889169692993,
+      "learning_rate": 4.818110370396686e-05,
+      "loss": 1.6467,
+      "step": 4925
+    },
+    {
+      "epoch": 0.511685883452789,
+      "grad_norm": 0.4382588863372803,
+      "learning_rate": 4.816479802959503e-05,
+      "loss": 1.6146,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5117897579723694,
+      "grad_norm": 0.41980645060539246,
+      "learning_rate": 4.81484925506579e-05,
+      "loss": 1.6783,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5118936324919497,
+      "grad_norm": 0.413632333278656,
+      "learning_rate": 4.813218726889186e-05,
+      "loss": 1.749,
+      "step": 4928
+    },
+    {
+      "epoch": 0.51199750701153,
+      "grad_norm": 0.40364977717399597,
+      "learning_rate": 4.8115882186033326e-05,
+      "loss": 1.5492,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5121013815311104,
+      "grad_norm": 0.3991188406944275,
+      "learning_rate": 4.8099577303818616e-05,
+      "loss": 1.7461,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5122052560506908,
+      "grad_norm": 0.4893375337123871,
+      "learning_rate": 4.8083272623984104e-05,
+      "loss": 1.9981,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5123091305702712,
+      "grad_norm": 0.40781837701797485,
+      "learning_rate": 4.8066968148266116e-05,
+      "loss": 1.7219,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5124130050898514,
+      "grad_norm": 0.39397352933883667,
+      "learning_rate": 4.805066387840096e-05,
+      "loss": 1.6126,
+      "step": 4933
+    },
+    {
+      "epoch": 0.5125168796094318,
+      "grad_norm": 0.41246309876441956,
+      "learning_rate": 4.803435981612486e-05,
+      "loss": 1.7436,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5126207541290122,
+      "grad_norm": 0.3805597424507141,
+      "learning_rate": 4.801805596317412e-05,
+      "loss": 1.6401,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5127246286485925,
+      "grad_norm": 0.404367059469223,
+      "learning_rate": 4.800175232128498e-05,
+      "loss": 1.6504,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5128285031681729,
+      "grad_norm": 0.3783699870109558,
+      "learning_rate": 4.79854488921936e-05,
+      "loss": 1.6495,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5129323776877532,
+      "grad_norm": 0.48037418723106384,
+      "learning_rate": 4.7969145677636215e-05,
+      "loss": 1.6088,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5130362522073335,
+      "grad_norm": 0.4683350622653961,
+      "learning_rate": 4.795284267934898e-05,
+      "loss": 1.8945,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5131401267269139,
+      "grad_norm": 0.4192398488521576,
+      "learning_rate": 4.793653989906802e-05,
+      "loss": 1.8459,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5132440012464943,
+      "grad_norm": 0.3951408565044403,
+      "learning_rate": 4.792023733852947e-05,
+      "loss": 1.6314,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5133478757660745,
+      "grad_norm": 0.38078486919403076,
+      "learning_rate": 4.790393499946943e-05,
+      "loss": 1.6349,
+      "step": 4942
+    },
+    {
+      "epoch": 0.5134517502856549,
+      "grad_norm": 0.4152866005897522,
+      "learning_rate": 4.788763288362394e-05,
+      "loss": 1.7195,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5135556248052353,
+      "grad_norm": 0.37676650285720825,
+      "learning_rate": 4.787133099272908e-05,
+      "loss": 1.5917,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5136594993248156,
+      "grad_norm": 0.3816780149936676,
+      "learning_rate": 4.7855029328520876e-05,
+      "loss": 1.6664,
+      "step": 4945
+    },
+    {
+      "epoch": 0.513763373844396,
+      "grad_norm": 0.3960273265838623,
+      "learning_rate": 4.7838727892735315e-05,
+      "loss": 1.6255,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5138672483639763,
+      "grad_norm": 0.41314899921417236,
+      "learning_rate": 4.782242668710836e-05,
+      "loss": 1.7316,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5139711228835566,
+      "grad_norm": 0.3930761516094208,
+      "learning_rate": 4.7806125713376e-05,
+      "loss": 1.6513,
+      "step": 4948
+    },
+    {
+      "epoch": 0.514074997403137,
+      "grad_norm": 0.4532455503940582,
+      "learning_rate": 4.7789824973274114e-05,
+      "loss": 1.6415,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5141788719227174,
+      "grad_norm": 0.4238525927066803,
+      "learning_rate": 4.7773524468538624e-05,
+      "loss": 1.8161,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5142827464422977,
+      "grad_norm": 0.40888962149620056,
+      "learning_rate": 4.7757224200905444e-05,
+      "loss": 1.7179,
+      "step": 4951
+    },
+    {
+      "epoch": 0.514386620961878,
+      "grad_norm": 0.38617759943008423,
+      "learning_rate": 4.774092417211037e-05,
+      "loss": 1.6069,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5144904954814584,
+      "grad_norm": 0.41479814052581787,
+      "learning_rate": 4.7724624383889256e-05,
+      "loss": 1.7158,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5145943700010387,
+      "grad_norm": 0.38144171237945557,
+      "learning_rate": 4.770832483797789e-05,
+      "loss": 1.6917,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5146982445206191,
+      "grad_norm": 0.41436558961868286,
+      "learning_rate": 4.769202553611206e-05,
+      "loss": 1.817,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5148021190401995,
+      "grad_norm": 0.43558332324028015,
+      "learning_rate": 4.76757264800275e-05,
+      "loss": 1.7038,
+      "step": 4956
+    },
+    {
+      "epoch": 0.5149059935597798,
+      "grad_norm": 0.3971516788005829,
+      "learning_rate": 4.765942767145994e-05,
+      "loss": 1.6663,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5150098680793601,
+      "grad_norm": 0.39739498496055603,
+      "learning_rate": 4.764312911214509e-05,
+      "loss": 1.5594,
+      "step": 4958
+    },
+    {
+      "epoch": 0.5151137425989405,
+      "grad_norm": 0.408383846282959,
+      "learning_rate": 4.762683080381859e-05,
+      "loss": 1.7743,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5152176171185209,
+      "grad_norm": 0.39290642738342285,
+      "learning_rate": 4.761053274821611e-05,
+      "loss": 1.6531,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5153214916381011,
+      "grad_norm": 0.46999290585517883,
+      "learning_rate": 4.7594234947073256e-05,
+      "loss": 1.8683,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5154253661576815,
+      "grad_norm": 0.40802252292633057,
+      "learning_rate": 4.75779374021256e-05,
+      "loss": 1.5823,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5155292406772619,
+      "grad_norm": 0.42017194628715515,
+      "learning_rate": 4.756164011510871e-05,
+      "loss": 1.802,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5156331151968422,
+      "grad_norm": 0.4172336161136627,
+      "learning_rate": 4.754534308775816e-05,
+      "loss": 1.7046,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5157369897164226,
+      "grad_norm": 0.38955143094062805,
+      "learning_rate": 4.75290463218094e-05,
+      "loss": 1.5606,
+      "step": 4965
+    },
+    {
+      "epoch": 0.515840864236003,
+      "grad_norm": 0.47302696108818054,
+      "learning_rate": 4.7512749818997936e-05,
+      "loss": 1.6814,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5159447387555832,
+      "grad_norm": 0.40039142966270447,
+      "learning_rate": 4.749645358105923e-05,
+      "loss": 1.6354,
+      "step": 4967
+    },
+    {
+      "epoch": 0.5160486132751636,
+      "grad_norm": 0.3926571011543274,
+      "learning_rate": 4.748015760972867e-05,
+      "loss": 1.6332,
+      "step": 4968
+    },
+    {
+      "epoch": 0.516152487794744,
+      "grad_norm": 0.4233894348144531,
+      "learning_rate": 4.7463861906741666e-05,
+      "loss": 1.3089,
+      "step": 4969
+    },
+    {
+      "epoch": 0.5162563623143243,
+      "grad_norm": 0.3955030143260956,
+      "learning_rate": 4.7447566473833615e-05,
+      "loss": 1.5917,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5163602368339046,
+      "grad_norm": 0.3940108120441437,
+      "learning_rate": 4.743127131273979e-05,
+      "loss": 1.7515,
+      "step": 4971
+    },
+    {
+      "epoch": 0.516464111353485,
+      "grad_norm": 0.44134819507598877,
+      "learning_rate": 4.7414976425195544e-05,
+      "loss": 1.5085,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5165679858730653,
+      "grad_norm": 0.44649553298950195,
+      "learning_rate": 4.739868181293616e-05,
+      "loss": 1.8201,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5166718603926457,
+      "grad_norm": 0.4249851703643799,
+      "learning_rate": 4.738238747769687e-05,
+      "loss": 1.7515,
+      "step": 4974
+    },
+    {
+      "epoch": 0.5167757349122261,
+      "grad_norm": 0.41049546003341675,
+      "learning_rate": 4.7366093421212884e-05,
+      "loss": 1.6677,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5168796094318063,
+      "grad_norm": 0.4584263563156128,
+      "learning_rate": 4.7349799645219415e-05,
+      "loss": 1.7249,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5169834839513867,
+      "grad_norm": 0.4268789291381836,
+      "learning_rate": 4.73335061514516e-05,
+      "loss": 1.6639,
+      "step": 4977
+    },
+    {
+      "epoch": 0.5170873584709671,
+      "grad_norm": 0.4349389672279358,
+      "learning_rate": 4.731721294164457e-05,
+      "loss": 1.6422,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5171912329905474,
+      "grad_norm": 0.43998467922210693,
+      "learning_rate": 4.730092001753346e-05,
+      "loss": 1.8851,
+      "step": 4979
+    },
+    {
+      "epoch": 0.5172951075101277,
+      "grad_norm": 0.4167867600917816,
+      "learning_rate": 4.72846273808533e-05,
+      "loss": 1.6397,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5173989820297081,
+      "grad_norm": 0.40250730514526367,
+      "learning_rate": 4.726833503333915e-05,
+      "loss": 1.4696,
+      "step": 4981
+    },
+    {
+      "epoch": 0.5175028565492885,
+      "grad_norm": 0.44303545355796814,
+      "learning_rate": 4.7252042976726e-05,
+      "loss": 1.814,
+      "step": 4982
+    },
+    {
+      "epoch": 0.5176067310688688,
+      "grad_norm": 0.3853762745857239,
+      "learning_rate": 4.723575121274885e-05,
+      "loss": 1.6521,
+      "step": 4983
+    },
+    {
+      "epoch": 0.5177106055884492,
+      "grad_norm": 0.39856797456741333,
+      "learning_rate": 4.721945974314263e-05,
+      "loss": 1.6085,
+      "step": 4984
+    },
+    {
+      "epoch": 0.5178144801080296,
+      "grad_norm": 0.380560040473938,
+      "learning_rate": 4.720316856964224e-05,
+      "loss": 1.6189,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5179183546276098,
+      "grad_norm": 0.4493720531463623,
+      "learning_rate": 4.7186877693982616e-05,
+      "loss": 1.7474,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5180222291471902,
+      "grad_norm": 0.38229069113731384,
+      "learning_rate": 4.717058711789855e-05,
+      "loss": 1.5625,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5181261036667706,
+      "grad_norm": 0.3856019675731659,
+      "learning_rate": 4.715429684312489e-05,
+      "loss": 1.6004,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5182299781863509,
+      "grad_norm": 0.3876418471336365,
+      "learning_rate": 4.713800687139644e-05,
+      "loss": 1.6596,
+      "step": 4989
+    },
+    {
+      "epoch": 0.5183338527059312,
+      "grad_norm": 0.41648218035697937,
+      "learning_rate": 4.712171720444791e-05,
+      "loss": 1.8412,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5184377272255116,
+      "grad_norm": 0.43493372201919556,
+      "learning_rate": 4.7105427844014046e-05,
+      "loss": 1.8457,
+      "step": 4991
+    },
+    {
+      "epoch": 0.5185416017450919,
+      "grad_norm": 0.37969422340393066,
+      "learning_rate": 4.7089138791829554e-05,
+      "loss": 1.6451,
+      "step": 4992
+    },
+    {
+      "epoch": 0.5186454762646723,
+      "grad_norm": 0.4082360863685608,
+      "learning_rate": 4.707285004962906e-05,
+      "loss": 1.8016,
+      "step": 4993
+    },
+    {
+      "epoch": 0.5187493507842527,
+      "grad_norm": 0.42352405190467834,
+      "learning_rate": 4.70565616191472e-05,
+      "loss": 1.5526,
+      "step": 4994
+    },
+    {
+      "epoch": 0.5188532253038329,
+      "grad_norm": 0.3847419321537018,
+      "learning_rate": 4.704027350211859e-05,
+      "loss": 1.6518,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5189570998234133,
+      "grad_norm": 0.40276169776916504,
+      "learning_rate": 4.702398570027775e-05,
+      "loss": 1.5955,
+      "step": 4996
+    },
+    {
+      "epoch": 0.5190609743429937,
+      "grad_norm": 0.44198933243751526,
+      "learning_rate": 4.700769821535921e-05,
+      "loss": 1.7202,
+      "step": 4997
+    },
+    {
+      "epoch": 0.519164848862574,
+      "grad_norm": 0.39979052543640137,
+      "learning_rate": 4.6991411049097494e-05,
+      "loss": 1.6186,
+      "step": 4998
+    },
+    {
+      "epoch": 0.5192687233821544,
+      "grad_norm": 0.39757975935935974,
+      "learning_rate": 4.697512420322701e-05,
+      "loss": 1.7077,
+      "step": 4999
+    },
+    {
+      "epoch": 0.5193725979017347,
+      "grad_norm": 0.3819902837276459,
+      "learning_rate": 4.69588376794822e-05,
+      "loss": 1.5771,
+      "step": 5000
+    },
+    {
+      "epoch": 0.519476472421315,
+      "grad_norm": 0.3874730169773102,
+      "learning_rate": 4.694255147959749e-05,
+      "loss": 1.6497,
+      "step": 5001
+    },
+    {
+      "epoch": 0.5195803469408954,
+      "grad_norm": 0.38986441493034363,
+      "learning_rate": 4.6926265605307175e-05,
+      "loss": 1.6801,
+      "step": 5002
+    },
+    {
+      "epoch": 0.5196842214604758,
+      "grad_norm": 0.4043687880039215,
+      "learning_rate": 4.690998005834559e-05,
+      "loss": 1.5736,
+      "step": 5003
+    },
+    {
+      "epoch": 0.519788095980056,
+      "grad_norm": 0.4174960255622864,
+      "learning_rate": 4.689369484044705e-05,
+      "loss": 1.7306,
+      "step": 5004
+    },
+    {
+      "epoch": 0.5198919704996364,
+      "grad_norm": 0.43060052394866943,
+      "learning_rate": 4.687740995334575e-05,
+      "loss": 1.7073,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5199958450192168,
+      "grad_norm": 0.405344158411026,
+      "learning_rate": 4.686112539877594e-05,
+      "loss": 1.7753,
+      "step": 5006
+    },
+    {
+      "epoch": 0.5200997195387971,
+      "grad_norm": 0.39156869053840637,
+      "learning_rate": 4.684484117847181e-05,
+      "loss": 1.5766,
+      "step": 5007
+    },
+    {
+      "epoch": 0.5202035940583775,
+      "grad_norm": 0.40544578433036804,
+      "learning_rate": 4.6828557294167465e-05,
+      "loss": 1.6717,
+      "step": 5008
+    },
+    {
+      "epoch": 0.5203074685779578,
+      "grad_norm": 0.41265055537223816,
+      "learning_rate": 4.681227374759704e-05,
+      "loss": 1.7227,
+      "step": 5009
+    },
+    {
+      "epoch": 0.5204113430975382,
+      "grad_norm": 0.3951095640659332,
+      "learning_rate": 4.679599054049458e-05,
+      "loss": 1.6448,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5205152176171185,
+      "grad_norm": 0.4113246500492096,
+      "learning_rate": 4.677970767459416e-05,
+      "loss": 1.7294,
+      "step": 5011
+    },
+    {
+      "epoch": 0.5206190921366989,
+      "grad_norm": 0.40176793932914734,
+      "learning_rate": 4.676342515162973e-05,
+      "loss": 1.6219,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5207229666562793,
+      "grad_norm": 0.6003035306930542,
+      "learning_rate": 4.674714297333527e-05,
+      "loss": 1.99,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5208268411758595,
+      "grad_norm": 0.36669179797172546,
+      "learning_rate": 4.673086114144474e-05,
+      "loss": 1.5943,
+      "step": 5014
+    },
+    {
+      "epoch": 0.5209307156954399,
+      "grad_norm": 0.3950868546962738,
+      "learning_rate": 4.671457965769198e-05,
+      "loss": 1.6115,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5210345902150203,
+      "grad_norm": 0.4209156036376953,
+      "learning_rate": 4.6698298523810854e-05,
+      "loss": 1.6077,
+      "step": 5016
+    },
+    {
+      "epoch": 0.5211384647346006,
+      "grad_norm": 0.38038328289985657,
+      "learning_rate": 4.668201774153521e-05,
+      "loss": 1.5769,
+      "step": 5017
+    },
+    {
+      "epoch": 0.521242339254181,
+      "grad_norm": 0.4059074819087982,
+      "learning_rate": 4.666573731259877e-05,
+      "loss": 1.6872,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5213462137737613,
+      "grad_norm": 0.3864630162715912,
+      "learning_rate": 4.664945723873529e-05,
+      "loss": 1.6696,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5214500882933416,
+      "grad_norm": 0.3989625871181488,
+      "learning_rate": 4.66331775216785e-05,
+      "loss": 1.716,
+      "step": 5020
+    },
+    {
+      "epoch": 0.521553962812922,
+      "grad_norm": 0.376228392124176,
+      "learning_rate": 4.6616898163162024e-05,
+      "loss": 1.6107,
+      "step": 5021
+    },
+    {
+      "epoch": 0.5216578373325024,
+      "grad_norm": 0.40415439009666443,
+      "learning_rate": 4.660061916491952e-05,
+      "loss": 1.6419,
+      "step": 5022
+    },
+    {
+      "epoch": 0.5217617118520826,
+      "grad_norm": 0.4512626528739929,
+      "learning_rate": 4.658434052868455e-05,
+      "loss": 1.8734,
+      "step": 5023
+    },
+    {
+      "epoch": 0.521865586371663,
+      "grad_norm": 0.5184199810028076,
+      "learning_rate": 4.656806225619067e-05,
+      "loss": 1.8781,
+      "step": 5024
+    },
+    {
+      "epoch": 0.5219694608912434,
+      "grad_norm": 0.4308519661426544,
+      "learning_rate": 4.655178434917139e-05,
+      "loss": 1.7415,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5220733354108237,
+      "grad_norm": 0.41828837990760803,
+      "learning_rate": 4.65355068093602e-05,
+      "loss": 1.6995,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5221772099304041,
+      "grad_norm": 0.3959868252277374,
+      "learning_rate": 4.651922963849048e-05,
+      "loss": 1.7536,
+      "step": 5027
+    },
+    {
+      "epoch": 0.5222810844499844,
+      "grad_norm": 0.427030473947525,
+      "learning_rate": 4.6502952838295654e-05,
+      "loss": 1.6309,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5223849589695647,
+      "grad_norm": 0.3931408226490021,
+      "learning_rate": 4.64866764105091e-05,
+      "loss": 1.6537,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5224888334891451,
+      "grad_norm": 0.343965619802475,
+      "learning_rate": 4.647040035686409e-05,
+      "loss": 1.3474,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5225927080087255,
+      "grad_norm": 0.40805351734161377,
+      "learning_rate": 4.64541246790939e-05,
+      "loss": 1.6824,
+      "step": 5031
+    },
+    {
+      "epoch": 0.5226965825283058,
+      "grad_norm": 0.39621418714523315,
+      "learning_rate": 4.643784937893179e-05,
+      "loss": 1.6969,
+      "step": 5032
+    },
+    {
+      "epoch": 0.5228004570478861,
+      "grad_norm": 0.38394787907600403,
+      "learning_rate": 4.642157445811092e-05,
+      "loss": 1.614,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5229043315674665,
+      "grad_norm": 0.3906930685043335,
+      "learning_rate": 4.6405299918364456e-05,
+      "loss": 1.7664,
+      "step": 5034
+    },
+    {
+      "epoch": 0.5230082060870469,
+      "grad_norm": 0.37266916036605835,
+      "learning_rate": 4.638902576142552e-05,
+      "loss": 1.4807,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5231120806066272,
+      "grad_norm": 0.4071597456932068,
+      "learning_rate": 4.637275198902717e-05,
+      "loss": 1.4604,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5232159551262076,
+      "grad_norm": 0.4267309606075287,
+      "learning_rate": 4.635647860290244e-05,
+      "loss": 1.7642,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5233198296457879,
+      "grad_norm": 0.4255138635635376,
+      "learning_rate": 4.634020560478431e-05,
+      "loss": 1.7285,
+      "step": 5038
+    },
+    {
+      "epoch": 0.5234237041653682,
+      "grad_norm": 0.3941444456577301,
+      "learning_rate": 4.6323932996405764e-05,
+      "loss": 1.665,
+      "step": 5039
+    },
+    {
+      "epoch": 0.5235275786849486,
+      "grad_norm": 0.4022713005542755,
+      "learning_rate": 4.630766077949965e-05,
+      "loss": 1.731,
+      "step": 5040
+    },
+    {
+      "epoch": 0.523631453204529,
+      "grad_norm": 0.3884924352169037,
+      "learning_rate": 4.629138895579886e-05,
+      "loss": 1.541,
+      "step": 5041
+    },
+    {
+      "epoch": 0.5237353277241092,
+      "grad_norm": 0.41130009293556213,
+      "learning_rate": 4.6275117527036246e-05,
+      "loss": 1.6621,
+      "step": 5042
+    },
+    {
+      "epoch": 0.5238392022436896,
+      "grad_norm": 0.3775813579559326,
+      "learning_rate": 4.625884649494453e-05,
+      "loss": 1.5298,
+      "step": 5043
+    },
+    {
+      "epoch": 0.52394307676327,
+      "grad_norm": 0.4054543375968933,
+      "learning_rate": 4.6242575861256496e-05,
+      "loss": 1.8429,
+      "step": 5044
+    },
+    {
+      "epoch": 0.5240469512828503,
+      "grad_norm": 0.3855196535587311,
+      "learning_rate": 4.622630562770483e-05,
+      "loss": 1.589,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5241508258024307,
+      "grad_norm": 0.3769000470638275,
+      "learning_rate": 4.6210035796022166e-05,
+      "loss": 1.6403,
+      "step": 5046
+    },
+    {
+      "epoch": 0.524254700322011,
+      "grad_norm": 0.4284866750240326,
+      "learning_rate": 4.619376636794113e-05,
+      "loss": 1.6255,
+      "step": 5047
+    },
+    {
+      "epoch": 0.5243585748415913,
+      "grad_norm": 0.40864747762680054,
+      "learning_rate": 4.61774973451943e-05,
+      "loss": 1.627,
+      "step": 5048
+    },
+    {
+      "epoch": 0.5244624493611717,
+      "grad_norm": 0.4019078016281128,
+      "learning_rate": 4.6161228729514164e-05,
+      "loss": 1.5215,
+      "step": 5049
+    },
+    {
+      "epoch": 0.5245663238807521,
+      "grad_norm": 0.40582525730133057,
+      "learning_rate": 4.614496052263324e-05,
+      "loss": 1.615,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5246701984003324,
+      "grad_norm": 0.4704711437225342,
+      "learning_rate": 4.612869272628394e-05,
+      "loss": 1.8979,
+      "step": 5051
+    },
+    {
+      "epoch": 0.5247740729199127,
+      "grad_norm": 0.4247235059738159,
+      "learning_rate": 4.611242534219867e-05,
+      "loss": 1.7807,
+      "step": 5052
+    },
+    {
+      "epoch": 0.5248779474394931,
+      "grad_norm": 0.4381190240383148,
+      "learning_rate": 4.609615837210977e-05,
+      "loss": 1.6944,
+      "step": 5053
+    },
+    {
+      "epoch": 0.5249818219590734,
+      "grad_norm": 0.37743279337882996,
+      "learning_rate": 4.6079891817749576e-05,
+      "loss": 1.5408,
+      "step": 5054
+    },
+    {
+      "epoch": 0.5250856964786538,
+      "grad_norm": 0.41715410351753235,
+      "learning_rate": 4.6063625680850306e-05,
+      "loss": 1.8648,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5251895709982342,
+      "grad_norm": 0.3836834132671356,
+      "learning_rate": 4.6047359963144187e-05,
+      "loss": 1.5844,
+      "step": 5056
+    },
+    {
+      "epoch": 0.5252934455178144,
+      "grad_norm": 0.4763341248035431,
+      "learning_rate": 4.6031094666363426e-05,
+      "loss": 1.9145,
+      "step": 5057
+    },
+    {
+      "epoch": 0.5253973200373948,
+      "grad_norm": 0.5105146765708923,
+      "learning_rate": 4.601482979224012e-05,
+      "loss": 1.9522,
+      "step": 5058
+    },
+    {
+      "epoch": 0.5255011945569752,
+      "grad_norm": 0.4624611735343933,
+      "learning_rate": 4.599856534250633e-05,
+      "loss": 1.7487,
+      "step": 5059
+    },
+    {
+      "epoch": 0.5256050690765556,
+      "grad_norm": 0.46096524596214294,
+      "learning_rate": 4.5982301318894126e-05,
+      "loss": 1.7626,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5257089435961358,
+      "grad_norm": 0.42811891436576843,
+      "learning_rate": 4.596603772313552e-05,
+      "loss": 1.9035,
+      "step": 5061
+    },
+    {
+      "epoch": 0.5258128181157162,
+      "grad_norm": 0.38237836956977844,
+      "learning_rate": 4.59497745569624e-05,
+      "loss": 1.654,
+      "step": 5062
+    },
+    {
+      "epoch": 0.5259166926352966,
+      "grad_norm": 0.38058823347091675,
+      "learning_rate": 4.593351182210671e-05,
+      "loss": 1.6587,
+      "step": 5063
+    },
+    {
+      "epoch": 0.5260205671548769,
+      "grad_norm": 0.4481144845485687,
+      "learning_rate": 4.59172495203003e-05,
+      "loss": 1.7732,
+      "step": 5064
+    },
+    {
+      "epoch": 0.5261244416744573,
+      "grad_norm": 0.3913651406764984,
+      "learning_rate": 4.590098765327496e-05,
+      "loss": 1.7154,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5262283161940376,
+      "grad_norm": 0.3957405388355255,
+      "learning_rate": 4.588472622276245e-05,
+      "loss": 1.7209,
+      "step": 5066
+    },
+    {
+      "epoch": 0.5263321907136179,
+      "grad_norm": 0.4275793135166168,
+      "learning_rate": 4.5868465230494525e-05,
+      "loss": 1.7113,
+      "step": 5067
+    },
+    {
+      "epoch": 0.5264360652331983,
+      "grad_norm": 0.39351120591163635,
+      "learning_rate": 4.58522046782028e-05,
+      "loss": 1.7318,
+      "step": 5068
+    },
+    {
+      "epoch": 0.5265399397527787,
+      "grad_norm": 0.42391449213027954,
+      "learning_rate": 4.5835944567618916e-05,
+      "loss": 1.721,
+      "step": 5069
+    },
+    {
+      "epoch": 0.526643814272359,
+      "grad_norm": 0.42262449860572815,
+      "learning_rate": 4.5819684900474486e-05,
+      "loss": 1.7771,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5267476887919393,
+      "grad_norm": 0.4281655251979828,
+      "learning_rate": 4.580342567850098e-05,
+      "loss": 1.6757,
+      "step": 5071
+    },
+    {
+      "epoch": 0.5268515633115197,
+      "grad_norm": 0.4202190935611725,
+      "learning_rate": 4.578716690342992e-05,
+      "loss": 1.597,
+      "step": 5072
+    },
+    {
+      "epoch": 0.5269554378311,
+      "grad_norm": 0.39920350909233093,
+      "learning_rate": 4.577090857699273e-05,
+      "loss": 1.4776,
+      "step": 5073
+    },
+    {
+      "epoch": 0.5270593123506804,
+      "grad_norm": 0.4190593361854553,
+      "learning_rate": 4.575465070092077e-05,
+      "loss": 1.7005,
+      "step": 5074
+    },
+    {
+      "epoch": 0.5271631868702608,
+      "grad_norm": 0.4298344850540161,
+      "learning_rate": 4.573839327694539e-05,
+      "loss": 1.7374,
+      "step": 5075
+    },
+    {
+      "epoch": 0.527267061389841,
+      "grad_norm": 0.3944896161556244,
+      "learning_rate": 4.5722136306797914e-05,
+      "loss": 1.7146,
+      "step": 5076
+    },
+    {
+      "epoch": 0.5273709359094214,
+      "grad_norm": 0.44215551018714905,
+      "learning_rate": 4.570587979220953e-05,
+      "loss": 1.7662,
+      "step": 5077
+    },
+    {
+      "epoch": 0.5274748104290018,
+      "grad_norm": 0.3969592750072479,
+      "learning_rate": 4.568962373491146e-05,
+      "loss": 1.5801,
+      "step": 5078
+    },
+    {
+      "epoch": 0.5275786849485821,
+      "grad_norm": 0.4148666262626648,
+      "learning_rate": 4.567336813663484e-05,
+      "loss": 1.6804,
+      "step": 5079
+    },
+    {
+      "epoch": 0.5276825594681624,
+      "grad_norm": 0.3804900050163269,
+      "learning_rate": 4.565711299911076e-05,
+      "loss": 1.6895,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5277864339877428,
+      "grad_norm": 0.4792127013206482,
+      "learning_rate": 4.564085832407026e-05,
+      "loss": 1.9221,
+      "step": 5081
+    },
+    {
+      "epoch": 0.5278903085073231,
+      "grad_norm": 0.36079883575439453,
+      "learning_rate": 4.5624604113244376e-05,
+      "loss": 1.6225,
+      "step": 5082
+    },
+    {
+      "epoch": 0.5279941830269035,
+      "grad_norm": 0.3568095266819,
+      "learning_rate": 4.5608350368363995e-05,
+      "loss": 1.4058,
+      "step": 5083
+    },
+    {
+      "epoch": 0.5280980575464839,
+      "grad_norm": 0.43193015456199646,
+      "learning_rate": 4.559209709116005e-05,
+      "loss": 1.4616,
+      "step": 5084
+    },
+    {
+      "epoch": 0.5282019320660641,
+      "grad_norm": 0.43650293350219727,
+      "learning_rate": 4.557584428336339e-05,
+      "loss": 1.7984,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5283058065856445,
+      "grad_norm": 0.3841239809989929,
+      "learning_rate": 4.555959194670481e-05,
+      "loss": 1.6124,
+      "step": 5086
+    },
+    {
+      "epoch": 0.5284096811052249,
+      "grad_norm": 0.3851770758628845,
+      "learning_rate": 4.554334008291503e-05,
+      "loss": 1.629,
+      "step": 5087
+    },
+    {
+      "epoch": 0.5285135556248053,
+      "grad_norm": 0.4100612998008728,
+      "learning_rate": 4.5527088693724775e-05,
+      "loss": 1.6488,
+      "step": 5088
+    },
+    {
+      "epoch": 0.5286174301443856,
+      "grad_norm": 0.42137911915779114,
+      "learning_rate": 4.5510837780864705e-05,
+      "loss": 1.8,
+      "step": 5089
+    },
+    {
+      "epoch": 0.5287213046639659,
+      "grad_norm": 0.5062525868415833,
+      "learning_rate": 4.549458734606537e-05,
+      "loss": 1.6027,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5288251791835463,
+      "grad_norm": 0.3984263241291046,
+      "learning_rate": 4.547833739105735e-05,
+      "loss": 1.4691,
+      "step": 5091
+    },
+    {
+      "epoch": 0.5289290537031266,
+      "grad_norm": 0.4082936644554138,
+      "learning_rate": 4.5462087917571124e-05,
+      "loss": 1.7615,
+      "step": 5092
+    },
+    {
+      "epoch": 0.529032928222707,
+      "grad_norm": 0.43450236320495605,
+      "learning_rate": 4.544583892733714e-05,
+      "loss": 1.6093,
+      "step": 5093
+    },
+    {
+      "epoch": 0.5291368027422874,
+      "grad_norm": 0.4008505344390869,
+      "learning_rate": 4.5429590422085776e-05,
+      "loss": 1.7545,
+      "step": 5094
+    },
+    {
+      "epoch": 0.5292406772618676,
+      "grad_norm": 0.41801589727401733,
+      "learning_rate": 4.5413342403547403e-05,
+      "loss": 1.6709,
+      "step": 5095
+    },
+    {
+      "epoch": 0.529344551781448,
+      "grad_norm": 0.39262089133262634,
+      "learning_rate": 4.539709487345227e-05,
+      "loss": 1.4997,
+      "step": 5096
+    },
+    {
+      "epoch": 0.5294484263010284,
+      "grad_norm": 0.41034695506095886,
+      "learning_rate": 4.5380847833530625e-05,
+      "loss": 1.803,
+      "step": 5097
+    },
+    {
+      "epoch": 0.5295523008206087,
+      "grad_norm": 0.42729073762893677,
+      "learning_rate": 4.5364601285512674e-05,
+      "loss": 1.589,
+      "step": 5098
+    },
+    {
+      "epoch": 0.529656175340189,
+      "grad_norm": 0.3822977542877197,
+      "learning_rate": 4.534835523112852e-05,
+      "loss": 1.6351,
+      "step": 5099
+    },
+    {
+      "epoch": 0.5297600498597694,
+      "grad_norm": 0.4156366288661957,
+      "learning_rate": 4.533210967210824e-05,
+      "loss": 1.7417,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5298639243793497,
+      "grad_norm": 0.4239443242549896,
+      "learning_rate": 4.5315864610181894e-05,
+      "loss": 1.9393,
+      "step": 5101
+    },
+    {
+      "epoch": 0.5299677988989301,
+      "grad_norm": 0.40841466188430786,
+      "learning_rate": 4.52996200470794e-05,
+      "loss": 1.5193,
+      "step": 5102
+    },
+    {
+      "epoch": 0.5300716734185105,
+      "grad_norm": 0.4005943238735199,
+      "learning_rate": 4.528337598453071e-05,
+      "loss": 1.6073,
+      "step": 5103
+    },
+    {
+      "epoch": 0.5301755479380907,
+      "grad_norm": 0.3967158794403076,
+      "learning_rate": 4.5267132424265705e-05,
+      "loss": 1.6722,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5302794224576711,
+      "grad_norm": 0.3896438181400299,
+      "learning_rate": 4.5250889368014166e-05,
+      "loss": 1.684,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5303832969772515,
+      "grad_norm": 0.4113629460334778,
+      "learning_rate": 4.523464681750587e-05,
+      "loss": 1.5668,
+      "step": 5106
+    },
+    {
+      "epoch": 0.5304871714968318,
+      "grad_norm": 0.39420047402381897,
+      "learning_rate": 4.521840477447051e-05,
+      "loss": 1.8714,
+      "step": 5107
+    },
+    {
+      "epoch": 0.5305910460164122,
+      "grad_norm": 0.3833891451358795,
+      "learning_rate": 4.5202163240637755e-05,
+      "loss": 1.6501,
+      "step": 5108
+    },
+    {
+      "epoch": 0.5306949205359925,
+      "grad_norm": 0.40579354763031006,
+      "learning_rate": 4.5185922217737175e-05,
+      "loss": 1.5287,
+      "step": 5109
+    },
+    {
+      "epoch": 0.5307987950555728,
+      "grad_norm": 0.39992040395736694,
+      "learning_rate": 4.5169681707498354e-05,
+      "loss": 1.7683,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5309026695751532,
+      "grad_norm": 0.3796159327030182,
+      "learning_rate": 4.515344171165072e-05,
+      "loss": 1.4853,
+      "step": 5111
+    },
+    {
+      "epoch": 0.5310065440947336,
+      "grad_norm": 0.3873627185821533,
+      "learning_rate": 4.5137202231923744e-05,
+      "loss": 1.7049,
+      "step": 5112
+    },
+    {
+      "epoch": 0.531110418614314,
+      "grad_norm": 0.42043253779411316,
+      "learning_rate": 4.512096327004682e-05,
+      "loss": 1.822,
+      "step": 5113
+    },
+    {
+      "epoch": 0.5312142931338942,
+      "grad_norm": 0.4138578176498413,
+      "learning_rate": 4.5104724827749254e-05,
+      "loss": 1.8201,
+      "step": 5114
+    },
+    {
+      "epoch": 0.5313181676534746,
+      "grad_norm": 0.4510040879249573,
+      "learning_rate": 4.508848690676028e-05,
+      "loss": 1.8475,
+      "step": 5115
+    },
+    {
+      "epoch": 0.531422042173055,
+      "grad_norm": 0.45895758271217346,
+      "learning_rate": 4.507224950880915e-05,
+      "loss": 1.8352,
+      "step": 5116
+    },
+    {
+      "epoch": 0.5315259166926353,
+      "grad_norm": 0.3979472815990448,
+      "learning_rate": 4.505601263562502e-05,
+      "loss": 1.7358,
+      "step": 5117
+    },
+    {
+      "epoch": 0.5316297912122157,
+      "grad_norm": 0.3930369019508362,
+      "learning_rate": 4.503977628893697e-05,
+      "loss": 1.7055,
+      "step": 5118
+    },
+    {
+      "epoch": 0.531733665731796,
+      "grad_norm": 0.3967008590698242,
+      "learning_rate": 4.502354047047406e-05,
+      "loss": 1.6146,
+      "step": 5119
+    },
+    {
+      "epoch": 0.5318375402513763,
+      "grad_norm": 0.4116523563861847,
+      "learning_rate": 4.500730518196527e-05,
+      "loss": 1.6813,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5319414147709567,
+      "grad_norm": 0.4032840430736542,
+      "learning_rate": 4.499107042513953e-05,
+      "loss": 1.566,
+      "step": 5121
+    },
+    {
+      "epoch": 0.5320452892905371,
+      "grad_norm": 0.3959857225418091,
+      "learning_rate": 4.4974836201725715e-05,
+      "loss": 1.5526,
+      "step": 5122
+    },
+    {
+      "epoch": 0.5321491638101173,
+      "grad_norm": 0.399959534406662,
+      "learning_rate": 4.495860251345266e-05,
+      "loss": 1.6931,
+      "step": 5123
+    },
+    {
+      "epoch": 0.5322530383296977,
+      "grad_norm": 0.4058084487915039,
+      "learning_rate": 4.49423693620491e-05,
+      "loss": 1.5507,
+      "step": 5124
+    },
+    {
+      "epoch": 0.5323569128492781,
+      "grad_norm": 0.4492541551589966,
+      "learning_rate": 4.4926136749243754e-05,
+      "loss": 1.7912,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5324607873688584,
+      "grad_norm": 0.4223318099975586,
+      "learning_rate": 4.490990467676528e-05,
+      "loss": 1.7354,
+      "step": 5126
+    },
+    {
+      "epoch": 0.5325646618884388,
+      "grad_norm": 0.38780099153518677,
+      "learning_rate": 4.489367314634225e-05,
+      "loss": 1.5337,
+      "step": 5127
+    },
+    {
+      "epoch": 0.5326685364080191,
+      "grad_norm": 0.40995070338249207,
+      "learning_rate": 4.487744215970318e-05,
+      "loss": 1.7209,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5327724109275994,
+      "grad_norm": 0.45095205307006836,
+      "learning_rate": 4.486121171857659e-05,
+      "loss": 1.7835,
+      "step": 5129
+    },
+    {
+      "epoch": 0.5328762854471798,
+      "grad_norm": 0.42928487062454224,
+      "learning_rate": 4.4844981824690844e-05,
+      "loss": 1.7334,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5329801599667602,
+      "grad_norm": 0.38315919041633606,
+      "learning_rate": 4.4828752479774324e-05,
+      "loss": 1.5904,
+      "step": 5131
+    },
+    {
+      "epoch": 0.5330840344863405,
+      "grad_norm": 0.3942658603191376,
+      "learning_rate": 4.481252368555535e-05,
+      "loss": 1.6793,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5331879090059208,
+      "grad_norm": 0.419535756111145,
+      "learning_rate": 4.4796295443762116e-05,
+      "loss": 1.7484,
+      "step": 5133
+    },
+    {
+      "epoch": 0.5332917835255012,
+      "grad_norm": 0.378946989774704,
+      "learning_rate": 4.478006775612284e-05,
+      "loss": 1.6216,
+      "step": 5134
+    },
+    {
+      "epoch": 0.5333956580450815,
+      "grad_norm": 0.43386486172676086,
+      "learning_rate": 4.476384062436563e-05,
+      "loss": 1.8178,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5334995325646619,
+      "grad_norm": 0.4382666051387787,
+      "learning_rate": 4.474761405021855e-05,
+      "loss": 1.8414,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5336034070842423,
+      "grad_norm": 0.4107944667339325,
+      "learning_rate": 4.4731388035409586e-05,
+      "loss": 1.7848,
+      "step": 5137
+    },
+    {
+      "epoch": 0.5337072816038226,
+      "grad_norm": 0.4328972399234772,
+      "learning_rate": 4.471516258166671e-05,
+      "loss": 1.8443,
+      "step": 5138
+    },
+    {
+      "epoch": 0.5338111561234029,
+      "grad_norm": 0.3964642584323883,
+      "learning_rate": 4.4698937690717816e-05,
+      "loss": 1.6394,
+      "step": 5139
+    },
+    {
+      "epoch": 0.5339150306429833,
+      "grad_norm": 0.37310969829559326,
+      "learning_rate": 4.468271336429068e-05,
+      "loss": 1.5261,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5340189051625637,
+      "grad_norm": 0.3795417547225952,
+      "learning_rate": 4.466648960411311e-05,
+      "loss": 1.512,
+      "step": 5141
+    },
+    {
+      "epoch": 0.5341227796821439,
+      "grad_norm": 0.40631794929504395,
+      "learning_rate": 4.46502664119128e-05,
+      "loss": 1.8027,
+      "step": 5142
+    },
+    {
+      "epoch": 0.5342266542017243,
+      "grad_norm": 0.47585329413414,
+      "learning_rate": 4.463404378941737e-05,
+      "loss": 1.8224,
+      "step": 5143
+    },
+    {
+      "epoch": 0.5343305287213047,
+      "grad_norm": 0.4149931073188782,
+      "learning_rate": 4.461782173835442e-05,
+      "loss": 1.6209,
+      "step": 5144
+    },
+    {
+      "epoch": 0.534434403240885,
+      "grad_norm": 0.41741567850112915,
+      "learning_rate": 4.4601600260451494e-05,
+      "loss": 1.8086,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5345382777604654,
+      "grad_norm": 0.43015965819358826,
+      "learning_rate": 4.458537935743601e-05,
+      "loss": 1.7298,
+      "step": 5146
+    },
+    {
+      "epoch": 0.5346421522800457,
+      "grad_norm": 0.3866709768772125,
+      "learning_rate": 4.45691590310354e-05,
+      "loss": 1.6128,
+      "step": 5147
+    },
+    {
+      "epoch": 0.534746026799626,
+      "grad_norm": 0.4770145118236542,
+      "learning_rate": 4.455293928297698e-05,
+      "loss": 1.8353,
+      "step": 5148
+    },
+    {
+      "epoch": 0.5348499013192064,
+      "grad_norm": 0.39021778106689453,
+      "learning_rate": 4.453672011498805e-05,
+      "loss": 1.5159,
+      "step": 5149
+    },
+    {
+      "epoch": 0.5349537758387868,
+      "grad_norm": 0.4521084129810333,
+      "learning_rate": 4.45205015287958e-05,
+      "loss": 1.7056,
+      "step": 5150
+    },
+    {
+      "epoch": 0.535057650358367,
+      "grad_norm": 0.3977662920951843,
+      "learning_rate": 4.450428352612742e-05,
+      "loss": 1.5728,
+      "step": 5151
+    },
+    {
+      "epoch": 0.5351615248779474,
+      "grad_norm": 0.4670353829860687,
+      "learning_rate": 4.448806610870995e-05,
+      "loss": 1.8235,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5352653993975278,
+      "grad_norm": 0.4064006507396698,
+      "learning_rate": 4.4471849278270446e-05,
+      "loss": 1.6757,
+      "step": 5153
+    },
+    {
+      "epoch": 0.5353692739171081,
+      "grad_norm": 0.41022297739982605,
+      "learning_rate": 4.445563303653589e-05,
+      "loss": 1.7759,
+      "step": 5154
+    },
+    {
+      "epoch": 0.5354731484366885,
+      "grad_norm": 0.37725940346717834,
+      "learning_rate": 4.443941738523316e-05,
+      "loss": 1.6766,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5355770229562689,
+      "grad_norm": 0.4157141447067261,
+      "learning_rate": 4.4423202326089095e-05,
+      "loss": 1.5477,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5356808974758491,
+      "grad_norm": 0.43074604868888855,
+      "learning_rate": 4.4406987860830505e-05,
+      "loss": 1.6997,
+      "step": 5157
+    },
+    {
+      "epoch": 0.5357847719954295,
+      "grad_norm": 0.4153660833835602,
+      "learning_rate": 4.439077399118405e-05,
+      "loss": 1.7147,
+      "step": 5158
+    },
+    {
+      "epoch": 0.5358886465150099,
+      "grad_norm": 0.40564560890197754,
+      "learning_rate": 4.4374560718876426e-05,
+      "loss": 1.6825,
+      "step": 5159
+    },
+    {
+      "epoch": 0.5359925210345902,
+      "grad_norm": 0.41882357001304626,
+      "learning_rate": 4.4358348045634215e-05,
+      "loss": 1.6888,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5360963955541705,
+      "grad_norm": 0.4029204249382019,
+      "learning_rate": 4.4342135973183915e-05,
+      "loss": 1.7675,
+      "step": 5161
+    },
+    {
+      "epoch": 0.5362002700737509,
+      "grad_norm": 0.42406782507896423,
+      "learning_rate": 4.432592450325201e-05,
+      "loss": 1.7902,
+      "step": 5162
+    },
+    {
+      "epoch": 0.5363041445933313,
+      "grad_norm": 0.4365594983100891,
+      "learning_rate": 4.430971363756487e-05,
+      "loss": 1.8162,
+      "step": 5163
+    },
+    {
+      "epoch": 0.5364080191129116,
+      "grad_norm": 0.3890252709388733,
+      "learning_rate": 4.429350337784887e-05,
+      "loss": 1.7314,
+      "step": 5164
+    },
+    {
+      "epoch": 0.536511893632492,
+      "grad_norm": 0.3934305012226105,
+      "learning_rate": 4.427729372583023e-05,
+      "loss": 1.7803,
+      "step": 5165
+    },
+    {
+      "epoch": 0.5366157681520723,
+      "grad_norm": 0.40188831090927124,
+      "learning_rate": 4.426108468323516e-05,
+      "loss": 1.6121,
+      "step": 5166
+    },
+    {
+      "epoch": 0.5367196426716526,
+      "grad_norm": 0.37544122338294983,
+      "learning_rate": 4.424487625178983e-05,
+      "loss": 1.6125,
+      "step": 5167
+    },
+    {
+      "epoch": 0.536823517191233,
+      "grad_norm": 0.43579837679862976,
+      "learning_rate": 4.4228668433220264e-05,
+      "loss": 1.806,
+      "step": 5168
+    },
+    {
+      "epoch": 0.5369273917108134,
+      "grad_norm": 0.37916016578674316,
+      "learning_rate": 4.421246122925251e-05,
+      "loss": 1.4683,
+      "step": 5169
+    },
+    {
+      "epoch": 0.5370312662303937,
+      "grad_norm": 0.3701176345348358,
+      "learning_rate": 4.419625464161249e-05,
+      "loss": 1.7148,
+      "step": 5170
+    },
+    {
+      "epoch": 0.537135140749974,
+      "grad_norm": 0.39556387066841125,
+      "learning_rate": 4.4180048672026066e-05,
+      "loss": 1.5598,
+      "step": 5171
+    },
+    {
+      "epoch": 0.5372390152695544,
+      "grad_norm": 0.413783460855484,
+      "learning_rate": 4.416384332221906e-05,
+      "loss": 1.6101,
+      "step": 5172
+    },
+    {
+      "epoch": 0.5373428897891347,
+      "grad_norm": 0.41170206665992737,
+      "learning_rate": 4.414763859391724e-05,
+      "loss": 1.7957,
+      "step": 5173
+    },
+    {
+      "epoch": 0.5374467643087151,
+      "grad_norm": 0.3944395184516907,
+      "learning_rate": 4.4131434488846236e-05,
+      "loss": 1.5504,
+      "step": 5174
+    },
+    {
+      "epoch": 0.5375506388282955,
+      "grad_norm": 0.47782790660858154,
+      "learning_rate": 4.4115231008731696e-05,
+      "loss": 1.7433,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5376545133478757,
+      "grad_norm": 0.42878782749176025,
+      "learning_rate": 4.409902815529914e-05,
+      "loss": 1.7387,
+      "step": 5176
+    },
+    {
+      "epoch": 0.5377583878674561,
+      "grad_norm": 0.42631304264068604,
+      "learning_rate": 4.4082825930274066e-05,
+      "loss": 1.8365,
+      "step": 5177
+    },
+    {
+      "epoch": 0.5378622623870365,
+      "grad_norm": 0.4046842157840729,
+      "learning_rate": 4.4066624335381864e-05,
+      "loss": 1.6611,
+      "step": 5178
+    },
+    {
+      "epoch": 0.5379661369066168,
+      "grad_norm": 0.41138967871665955,
+      "learning_rate": 4.40504233723479e-05,
+      "loss": 1.7386,
+      "step": 5179
+    },
+    {
+      "epoch": 0.5380700114261971,
+      "grad_norm": 0.4140818417072296,
+      "learning_rate": 4.4034223042897423e-05,
+      "loss": 1.7121,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5381738859457775,
+      "grad_norm": 0.4164121150970459,
+      "learning_rate": 4.4018023348755666e-05,
+      "loss": 1.746,
+      "step": 5181
+    },
+    {
+      "epoch": 0.5382777604653578,
+      "grad_norm": 0.4518551528453827,
+      "learning_rate": 4.400182429164778e-05,
+      "loss": 1.5526,
+      "step": 5182
+    },
+    {
+      "epoch": 0.5383816349849382,
+      "grad_norm": 0.39879488945007324,
+      "learning_rate": 4.39856258732988e-05,
+      "loss": 1.7516,
+      "step": 5183
+    },
+    {
+      "epoch": 0.5384855095045186,
+      "grad_norm": 0.43033257126808167,
+      "learning_rate": 4.396942809543374e-05,
+      "loss": 1.7992,
+      "step": 5184
+    },
+    {
+      "epoch": 0.5385893840240988,
+      "grad_norm": 0.4049583673477173,
+      "learning_rate": 4.395323095977757e-05,
+      "loss": 1.6651,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5386932585436792,
+      "grad_norm": 0.43411314487457275,
+      "learning_rate": 4.393703446805512e-05,
+      "loss": 1.7299,
+      "step": 5186
+    },
+    {
+      "epoch": 0.5387971330632596,
+      "grad_norm": 0.47748497128486633,
+      "learning_rate": 4.39208386219912e-05,
+      "loss": 1.8707,
+      "step": 5187
+    },
+    {
+      "epoch": 0.5389010075828399,
+      "grad_norm": 0.3915466070175171,
+      "learning_rate": 4.3904643423310574e-05,
+      "loss": 1.6647,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5390048821024203,
+      "grad_norm": 0.46273308992385864,
+      "learning_rate": 4.388844887373785e-05,
+      "loss": 1.7202,
+      "step": 5189
+    },
+    {
+      "epoch": 0.5391087566220006,
+      "grad_norm": 0.37621983885765076,
+      "learning_rate": 4.387225497499767e-05,
+      "loss": 1.7193,
+      "step": 5190
+    },
+    {
+      "epoch": 0.539212631141581,
+      "grad_norm": 0.4030202627182007,
+      "learning_rate": 4.385606172881452e-05,
+      "loss": 1.6711,
+      "step": 5191
+    },
+    {
+      "epoch": 0.5393165056611613,
+      "grad_norm": 0.4407753646373749,
+      "learning_rate": 4.383986913691289e-05,
+      "loss": 1.7142,
+      "step": 5192
+    },
+    {
+      "epoch": 0.5394203801807417,
+      "grad_norm": 0.40765175223350525,
+      "learning_rate": 4.382367720101713e-05,
+      "loss": 1.7295,
+      "step": 5193
+    },
+    {
+      "epoch": 0.5395242547003221,
+      "grad_norm": 0.40376102924346924,
+      "learning_rate": 4.3807485922851574e-05,
+      "loss": 1.574,
+      "step": 5194
+    },
+    {
+      "epoch": 0.5396281292199023,
+      "grad_norm": 0.41411998867988586,
+      "learning_rate": 4.3791295304140476e-05,
+      "loss": 1.6128,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5397320037394827,
+      "grad_norm": 0.4627225399017334,
+      "learning_rate": 4.3775105346607983e-05,
+      "loss": 1.8546,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5398358782590631,
+      "grad_norm": 0.4189532399177551,
+      "learning_rate": 4.375891605197822e-05,
+      "loss": 1.7621,
+      "step": 5197
+    },
+    {
+      "epoch": 0.5399397527786434,
+      "grad_norm": 0.40579670667648315,
+      "learning_rate": 4.3742727421975224e-05,
+      "loss": 1.6286,
+      "step": 5198
+    },
+    {
+      "epoch": 0.5400436272982237,
+      "grad_norm": 0.4409949481487274,
+      "learning_rate": 4.3726539458322943e-05,
+      "loss": 1.8302,
+      "step": 5199
+    },
+    {
+      "epoch": 0.5401475018178041,
+      "grad_norm": 0.4035548269748688,
+      "learning_rate": 4.371035216274526e-05,
+      "loss": 1.7163,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5402513763373844,
+      "grad_norm": 0.41622182726860046,
+      "learning_rate": 4.369416553696604e-05,
+      "loss": 1.5436,
+      "step": 5201
+    },
+    {
+      "epoch": 0.5403552508569648,
+      "grad_norm": 0.41848766803741455,
+      "learning_rate": 4.367797958270897e-05,
+      "loss": 1.6028,
+      "step": 5202
+    },
+    {
+      "epoch": 0.5404591253765452,
+      "grad_norm": 0.4080602824687958,
+      "learning_rate": 4.366179430169779e-05,
+      "loss": 1.6886,
+      "step": 5203
+    },
+    {
+      "epoch": 0.5405629998961254,
+      "grad_norm": 0.40305790305137634,
+      "learning_rate": 4.3645609695656056e-05,
+      "loss": 1.604,
+      "step": 5204
+    },
+    {
+      "epoch": 0.5406668744157058,
+      "grad_norm": 0.42248713970184326,
+      "learning_rate": 4.3629425766307334e-05,
+      "loss": 1.6635,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5407707489352862,
+      "grad_norm": 0.41928306221961975,
+      "learning_rate": 4.361324251537506e-05,
+      "loss": 1.5922,
+      "step": 5206
+    },
+    {
+      "epoch": 0.5408746234548665,
+      "grad_norm": 0.45712658762931824,
+      "learning_rate": 4.359705994458267e-05,
+      "loss": 1.7722,
+      "step": 5207
+    },
+    {
+      "epoch": 0.5409784979744469,
+      "grad_norm": 0.3815540373325348,
+      "learning_rate": 4.358087805565343e-05,
+      "loss": 1.6145,
+      "step": 5208
+    },
+    {
+      "epoch": 0.5410823724940272,
+      "grad_norm": 0.41365593671798706,
+      "learning_rate": 4.35646968503106e-05,
+      "loss": 1.5844,
+      "step": 5209
+    },
+    {
+      "epoch": 0.5411862470136075,
+      "grad_norm": 0.4178517162799835,
+      "learning_rate": 4.354851633027738e-05,
+      "loss": 1.7365,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5412901215331879,
+      "grad_norm": 0.4436066448688507,
+      "learning_rate": 4.3532336497276836e-05,
+      "loss": 1.7656,
+      "step": 5211
+    },
+    {
+      "epoch": 0.5413939960527683,
+      "grad_norm": 0.4040657877922058,
+      "learning_rate": 4.351615735303199e-05,
+      "loss": 1.7024,
+      "step": 5212
+    },
+    {
+      "epoch": 0.5414978705723485,
+      "grad_norm": 0.4756530821323395,
+      "learning_rate": 4.3499978899265844e-05,
+      "loss": 1.9407,
+      "step": 5213
+    },
+    {
+      "epoch": 0.5416017450919289,
+      "grad_norm": 0.43450602889060974,
+      "learning_rate": 4.348380113770121e-05,
+      "loss": 1.7781,
+      "step": 5214
+    },
+    {
+      "epoch": 0.5417056196115093,
+      "grad_norm": 0.4165419638156891,
+      "learning_rate": 4.346762407006093e-05,
+      "loss": 1.6077,
+      "step": 5215
+    },
+    {
+      "epoch": 0.5418094941310897,
+      "grad_norm": 0.4009828269481659,
+      "learning_rate": 4.345144769806775e-05,
+      "loss": 1.5677,
+      "step": 5216
+    },
+    {
+      "epoch": 0.54191336865067,
+      "grad_norm": 0.37052732706069946,
+      "learning_rate": 4.3435272023444304e-05,
+      "loss": 1.5191,
+      "step": 5217
+    },
+    {
+      "epoch": 0.5420172431702504,
+      "grad_norm": 0.4109463095664978,
+      "learning_rate": 4.3419097047913175e-05,
+      "loss": 1.5929,
+      "step": 5218
+    },
+    {
+      "epoch": 0.5421211176898307,
+      "grad_norm": 0.4167434275150299,
+      "learning_rate": 4.340292277319687e-05,
+      "loss": 1.612,
+      "step": 5219
+    },
+    {
+      "epoch": 0.542224992209411,
+      "grad_norm": 0.39904865622520447,
+      "learning_rate": 4.3386749201017856e-05,
+      "loss": 1.6542,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5423288667289914,
+      "grad_norm": 0.4138846695423126,
+      "learning_rate": 4.3370576333098446e-05,
+      "loss": 1.6344,
+      "step": 5221
+    },
+    {
+      "epoch": 0.5424327412485718,
+      "grad_norm": 0.38887113332748413,
+      "learning_rate": 4.335440417116094e-05,
+      "loss": 1.6036,
+      "step": 5222
+    },
+    {
+      "epoch": 0.542536615768152,
+      "grad_norm": 0.4017149806022644,
+      "learning_rate": 4.3338232716927574e-05,
+      "loss": 1.6746,
+      "step": 5223
+    },
+    {
+      "epoch": 0.5426404902877324,
+      "grad_norm": 0.39271867275238037,
+      "learning_rate": 4.3322061972120445e-05,
+      "loss": 1.6285,
+      "step": 5224
+    },
+    {
+      "epoch": 0.5427443648073128,
+      "grad_norm": 0.4323449432849884,
+      "learning_rate": 4.3305891938461626e-05,
+      "loss": 1.5169,
+      "step": 5225
+    },
+    {
+      "epoch": 0.5428482393268931,
+      "grad_norm": 0.37628036737442017,
+      "learning_rate": 4.3289722617673115e-05,
+      "loss": 1.6436,
+      "step": 5226
+    },
+    {
+      "epoch": 0.5429521138464735,
+      "grad_norm": 0.43014097213745117,
+      "learning_rate": 4.327355401147679e-05,
+      "loss": 1.7777,
+      "step": 5227
+    },
+    {
+      "epoch": 0.5430559883660538,
+      "grad_norm": 0.3823567032814026,
+      "learning_rate": 4.3257386121594485e-05,
+      "loss": 1.5837,
+      "step": 5228
+    },
+    {
+      "epoch": 0.5431598628856341,
+      "grad_norm": 0.4464382827281952,
+      "learning_rate": 4.324121894974799e-05,
+      "loss": 1.8451,
+      "step": 5229
+    },
+    {
+      "epoch": 0.5432637374052145,
+      "grad_norm": 0.37794092297554016,
+      "learning_rate": 4.322505249765893e-05,
+      "loss": 1.7015,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5433676119247949,
+      "grad_norm": 0.40177834033966064,
+      "learning_rate": 4.320888676704895e-05,
+      "loss": 1.7247,
+      "step": 5231
+    },
+    {
+      "epoch": 0.5434714864443752,
+      "grad_norm": 0.3933226764202118,
+      "learning_rate": 4.3192721759639545e-05,
+      "loss": 1.6449,
+      "step": 5232
+    },
+    {
+      "epoch": 0.5435753609639555,
+      "grad_norm": 0.4006556272506714,
+      "learning_rate": 4.317655747715218e-05,
+      "loss": 1.5498,
+      "step": 5233
+    },
+    {
+      "epoch": 0.5436792354835359,
+      "grad_norm": 0.4249705970287323,
+      "learning_rate": 4.3160393921308204e-05,
+      "loss": 1.7594,
+      "step": 5234
+    },
+    {
+      "epoch": 0.5437831100031162,
+      "grad_norm": 0.44793421030044556,
+      "learning_rate": 4.3144231093828954e-05,
+      "loss": 1.8968,
+      "step": 5235
+    },
+    {
+      "epoch": 0.5438869845226966,
+      "grad_norm": 0.3769989311695099,
+      "learning_rate": 4.312806899643559e-05,
+      "loss": 1.5737,
+      "step": 5236
+    },
+    {
+      "epoch": 0.543990859042277,
+      "grad_norm": 0.41074395179748535,
+      "learning_rate": 4.311190763084928e-05,
+      "loss": 1.6493,
+      "step": 5237
+    },
+    {
+      "epoch": 0.5440947335618572,
+      "grad_norm": 0.4340376555919647,
+      "learning_rate": 4.309574699879109e-05,
+      "loss": 1.7359,
+      "step": 5238
+    },
+    {
+      "epoch": 0.5441986080814376,
+      "grad_norm": 0.4758669435977936,
+      "learning_rate": 4.3079587101981985e-05,
+      "loss": 1.8211,
+      "step": 5239
+    },
+    {
+      "epoch": 0.544302482601018,
+      "grad_norm": 0.3941514790058136,
+      "learning_rate": 4.306342794214286e-05,
+      "loss": 1.4555,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5444063571205984,
+      "grad_norm": 0.40955716371536255,
+      "learning_rate": 4.3047269520994555e-05,
+      "loss": 1.6498,
+      "step": 5241
+    },
+    {
+      "epoch": 0.5445102316401786,
+      "grad_norm": 0.45448553562164307,
+      "learning_rate": 4.3031111840257834e-05,
+      "loss": 1.7542,
+      "step": 5242
+    },
+    {
+      "epoch": 0.544614106159759,
+      "grad_norm": 0.4177002012729645,
+      "learning_rate": 4.301495490165331e-05,
+      "loss": 1.6508,
+      "step": 5243
+    },
+    {
+      "epoch": 0.5447179806793394,
+      "grad_norm": 0.431246280670166,
+      "learning_rate": 4.299879870690162e-05,
+      "loss": 1.8128,
+      "step": 5244
+    },
+    {
+      "epoch": 0.5448218551989197,
+      "grad_norm": 0.4549204707145691,
+      "learning_rate": 4.298264325772327e-05,
+      "loss": 1.6395,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5449257297185001,
+      "grad_norm": 0.4020931124687195,
+      "learning_rate": 4.2966488555838656e-05,
+      "loss": 1.6735,
+      "step": 5246
+    },
+    {
+      "epoch": 0.5450296042380804,
+      "grad_norm": 0.4968905448913574,
+      "learning_rate": 4.295033460296814e-05,
+      "loss": 1.7222,
+      "step": 5247
+    },
+    {
+      "epoch": 0.5451334787576607,
+      "grad_norm": 0.3912610709667206,
+      "learning_rate": 4.293418140083203e-05,
+      "loss": 1.6682,
+      "step": 5248
+    },
+    {
+      "epoch": 0.5452373532772411,
+      "grad_norm": 0.39999935030937195,
+      "learning_rate": 4.291802895115046e-05,
+      "loss": 1.746,
+      "step": 5249
+    },
+    {
+      "epoch": 0.5453412277968215,
+      "grad_norm": 0.4106389582157135,
+      "learning_rate": 4.290187725564356e-05,
+      "loss": 1.5329,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5454451023164018,
+      "grad_norm": 0.4551447927951813,
+      "learning_rate": 4.28857263160314e-05,
+      "loss": 1.7031,
+      "step": 5251
+    },
+    {
+      "epoch": 0.5455489768359821,
+      "grad_norm": 0.39867910742759705,
+      "learning_rate": 4.286957613403387e-05,
+      "loss": 1.5662,
+      "step": 5252
+    },
+    {
+      "epoch": 0.5456528513555625,
+      "grad_norm": 0.38527026772499084,
+      "learning_rate": 4.285342671137086e-05,
+      "loss": 1.5221,
+      "step": 5253
+    },
+    {
+      "epoch": 0.5457567258751428,
+      "grad_norm": 0.40201330184936523,
+      "learning_rate": 4.283727804976217e-05,
+      "loss": 1.5841,
+      "step": 5254
+    },
+    {
+      "epoch": 0.5458606003947232,
+      "grad_norm": 0.40640905499458313,
+      "learning_rate": 4.2821130150927486e-05,
+      "loss": 1.6659,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5459644749143036,
+      "grad_norm": 0.39006564021110535,
+      "learning_rate": 4.280498301658644e-05,
+      "loss": 1.463,
+      "step": 5256
+    },
+    {
+      "epoch": 0.5460683494338838,
+      "grad_norm": 0.44832733273506165,
+      "learning_rate": 4.27888366484586e-05,
+      "loss": 1.941,
+      "step": 5257
+    },
+    {
+      "epoch": 0.5461722239534642,
+      "grad_norm": 0.41479620337486267,
+      "learning_rate": 4.277269104826339e-05,
+      "loss": 1.7727,
+      "step": 5258
+    },
+    {
+      "epoch": 0.5462760984730446,
+      "grad_norm": 0.40158823132514954,
+      "learning_rate": 4.275654621772021e-05,
+      "loss": 1.6807,
+      "step": 5259
+    },
+    {
+      "epoch": 0.5463799729926249,
+      "grad_norm": 0.39162763953208923,
+      "learning_rate": 4.2740402158548366e-05,
+      "loss": 1.6696,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5464838475122052,
+      "grad_norm": 0.41060155630111694,
+      "learning_rate": 4.272425887246706e-05,
+      "loss": 1.7343,
+      "step": 5261
+    },
+    {
+      "epoch": 0.5465877220317856,
+      "grad_norm": 0.3723122179508209,
+      "learning_rate": 4.270811636119542e-05,
+      "loss": 1.6486,
+      "step": 5262
+    },
+    {
+      "epoch": 0.5466915965513659,
+      "grad_norm": 0.4098038375377655,
+      "learning_rate": 4.269197462645253e-05,
+      "loss": 1.691,
+      "step": 5263
+    },
+    {
+      "epoch": 0.5467954710709463,
+      "grad_norm": 0.38569438457489014,
+      "learning_rate": 4.267583366995732e-05,
+      "loss": 1.6651,
+      "step": 5264
+    },
+    {
+      "epoch": 0.5468993455905267,
+      "grad_norm": 0.42734023928642273,
+      "learning_rate": 4.2659693493428684e-05,
+      "loss": 1.8253,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5470032201101069,
+      "grad_norm": 0.4274364709854126,
+      "learning_rate": 4.264355409858547e-05,
+      "loss": 1.6397,
+      "step": 5266
+    },
+    {
+      "epoch": 0.5471070946296873,
+      "grad_norm": 0.38870930671691895,
+      "learning_rate": 4.2627415487146346e-05,
+      "loss": 1.5403,
+      "step": 5267
+    },
+    {
+      "epoch": 0.5472109691492677,
+      "grad_norm": 0.3967702388763428,
+      "learning_rate": 4.261127766082995e-05,
+      "loss": 1.8323,
+      "step": 5268
+    },
+    {
+      "epoch": 0.5473148436688481,
+      "grad_norm": 0.42525243759155273,
+      "learning_rate": 4.2595140621354846e-05,
+      "loss": 1.7435,
+      "step": 5269
+    },
+    {
+      "epoch": 0.5474187181884284,
+      "grad_norm": 0.41377174854278564,
+      "learning_rate": 4.257900437043954e-05,
+      "loss": 1.5484,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5475225927080087,
+      "grad_norm": 0.36868613958358765,
+      "learning_rate": 4.256286890980234e-05,
+      "loss": 1.6659,
+      "step": 5271
+    },
+    {
+      "epoch": 0.5476264672275891,
+      "grad_norm": 0.43371888995170593,
+      "learning_rate": 4.254673424116162e-05,
+      "loss": 1.6907,
+      "step": 5272
+    },
+    {
+      "epoch": 0.5477303417471694,
+      "grad_norm": 0.4247437119483948,
+      "learning_rate": 4.2530600366235554e-05,
+      "loss": 1.7444,
+      "step": 5273
+    },
+    {
+      "epoch": 0.5478342162667498,
+      "grad_norm": 0.43270236253738403,
+      "learning_rate": 4.251446728674229e-05,
+      "loss": 1.6692,
+      "step": 5274
+    },
+    {
+      "epoch": 0.5479380907863302,
+      "grad_norm": 0.40581294894218445,
+      "learning_rate": 4.2498335004399864e-05,
+      "loss": 1.7245,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5480419653059104,
+      "grad_norm": 0.40520188212394714,
+      "learning_rate": 4.248220352092626e-05,
+      "loss": 1.6344,
+      "step": 5276
+    },
+    {
+      "epoch": 0.5481458398254908,
+      "grad_norm": 0.41677728295326233,
+      "learning_rate": 4.246607283803932e-05,
+      "loss": 1.7355,
+      "step": 5277
+    },
+    {
+      "epoch": 0.5482497143450712,
+      "grad_norm": 0.4243401885032654,
+      "learning_rate": 4.2449942957456855e-05,
+      "loss": 1.7528,
+      "step": 5278
+    },
+    {
+      "epoch": 0.5483535888646515,
+      "grad_norm": 0.44001367688179016,
+      "learning_rate": 4.24338138808966e-05,
+      "loss": 1.7903,
+      "step": 5279
+    },
+    {
+      "epoch": 0.5484574633842318,
+      "grad_norm": 0.40032702684402466,
+      "learning_rate": 4.241768561007613e-05,
+      "loss": 1.6569,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5485613379038122,
+      "grad_norm": 0.4183237850666046,
+      "learning_rate": 4.2401558146712994e-05,
+      "loss": 1.6377,
+      "step": 5281
+    },
+    {
+      "epoch": 0.5486652124233925,
+      "grad_norm": 0.4048320949077606,
+      "learning_rate": 4.238543149252466e-05,
+      "loss": 1.5852,
+      "step": 5282
+    },
+    {
+      "epoch": 0.5487690869429729,
+      "grad_norm": 0.4144987165927887,
+      "learning_rate": 4.2369305649228455e-05,
+      "loss": 1.7107,
+      "step": 5283
+    },
+    {
+      "epoch": 0.5488729614625533,
+      "grad_norm": 0.40201154351234436,
+      "learning_rate": 4.235318061854168e-05,
+      "loss": 1.7281,
+      "step": 5284
+    },
+    {
+      "epoch": 0.5489768359821335,
+      "grad_norm": 0.41914767026901245,
+      "learning_rate": 4.233705640218154e-05,
+      "loss": 1.7443,
+      "step": 5285
+    },
+    {
+      "epoch": 0.5490807105017139,
+      "grad_norm": 0.39670661091804504,
+      "learning_rate": 4.2320933001865106e-05,
+      "loss": 1.5151,
+      "step": 5286
+    },
+    {
+      "epoch": 0.5491845850212943,
+      "grad_norm": 0.49306583404541016,
+      "learning_rate": 4.230481041930941e-05,
+      "loss": 2.0025,
+      "step": 5287
+    },
+    {
+      "epoch": 0.5492884595408746,
+      "grad_norm": 0.4066086709499359,
+      "learning_rate": 4.2288688656231384e-05,
+      "loss": 1.6324,
+      "step": 5288
+    },
+    {
+      "epoch": 0.549392334060455,
+      "grad_norm": 0.43332695960998535,
+      "learning_rate": 4.227256771434787e-05,
+      "loss": 1.7869,
+      "step": 5289
+    },
+    {
+      "epoch": 0.5494962085800353,
+      "grad_norm": 0.37857580184936523,
+      "learning_rate": 4.2256447595375596e-05,
+      "loss": 1.6478,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5496000830996156,
+      "grad_norm": 0.45964285731315613,
+      "learning_rate": 4.224032830103128e-05,
+      "loss": 1.7415,
+      "step": 5291
+    },
+    {
+      "epoch": 0.549703957619196,
+      "grad_norm": 0.3959317207336426,
+      "learning_rate": 4.2224209833031445e-05,
+      "loss": 1.6643,
+      "step": 5292
+    },
+    {
+      "epoch": 0.5498078321387764,
+      "grad_norm": 0.48961135745048523,
+      "learning_rate": 4.220809219309262e-05,
+      "loss": 1.76,
+      "step": 5293
+    },
+    {
+      "epoch": 0.5499117066583568,
+      "grad_norm": 0.38995257019996643,
+      "learning_rate": 4.21919753829312e-05,
+      "loss": 1.6642,
+      "step": 5294
+    },
+    {
+      "epoch": 0.550015581177937,
+      "grad_norm": 0.4197613000869751,
+      "learning_rate": 4.217585940426351e-05,
+      "loss": 1.7614,
+      "step": 5295
+    },
+    {
+      "epoch": 0.5501194556975174,
+      "grad_norm": 0.40901002287864685,
+      "learning_rate": 4.2159744258805736e-05,
+      "loss": 1.8191,
+      "step": 5296
+    },
+    {
+      "epoch": 0.5502233302170978,
+      "grad_norm": 0.4012939929962158,
+      "learning_rate": 4.214362994827405e-05,
+      "loss": 1.4622,
+      "step": 5297
+    },
+    {
+      "epoch": 0.5503272047366781,
+      "grad_norm": 0.4541564881801605,
+      "learning_rate": 4.212751647438451e-05,
+      "loss": 1.9298,
+      "step": 5298
+    },
+    {
+      "epoch": 0.5504310792562584,
+      "grad_norm": 0.3447628915309906,
+      "learning_rate": 4.2111403838853034e-05,
+      "loss": 1.3871,
+      "step": 5299
+    },
+    {
+      "epoch": 0.5505349537758388,
+      "grad_norm": 0.4184315502643585,
+      "learning_rate": 4.209529204339553e-05,
+      "loss": 1.5378,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5506388282954191,
+      "grad_norm": 0.49825429916381836,
+      "learning_rate": 4.207918108972776e-05,
+      "loss": 1.8026,
+      "step": 5301
+    },
+    {
+      "epoch": 0.5507427028149995,
+      "grad_norm": 0.385919451713562,
+      "learning_rate": 4.2063070979565426e-05,
+      "loss": 1.6863,
+      "step": 5302
+    },
+    {
+      "epoch": 0.5508465773345799,
+      "grad_norm": 0.40161100029945374,
+      "learning_rate": 4.204696171462411e-05,
+      "loss": 1.5964,
+      "step": 5303
+    },
+    {
+      "epoch": 0.5509504518541601,
+      "grad_norm": 0.4033989906311035,
+      "learning_rate": 4.203085329661935e-05,
+      "loss": 1.7334,
+      "step": 5304
+    },
+    {
+      "epoch": 0.5510543263737405,
+      "grad_norm": 0.416106641292572,
+      "learning_rate": 4.201474572726654e-05,
+      "loss": 1.7121,
+      "step": 5305
+    },
+    {
+      "epoch": 0.5511582008933209,
+      "grad_norm": 0.43745073676109314,
+      "learning_rate": 4.199863900828102e-05,
+      "loss": 1.7883,
+      "step": 5306
+    },
+    {
+      "epoch": 0.5512620754129012,
+      "grad_norm": 0.4253213703632355,
+      "learning_rate": 4.1982533141378056e-05,
+      "loss": 1.7145,
+      "step": 5307
+    },
+    {
+      "epoch": 0.5513659499324816,
+      "grad_norm": 0.4426640570163727,
+      "learning_rate": 4.1966428128272754e-05,
+      "loss": 1.7868,
+      "step": 5308
+    },
+    {
+      "epoch": 0.5514698244520619,
+      "grad_norm": 0.3622012436389923,
+      "learning_rate": 4.195032397068019e-05,
+      "loss": 1.5348,
+      "step": 5309
+    },
+    {
+      "epoch": 0.5515736989716422,
+      "grad_norm": 0.45180484652519226,
+      "learning_rate": 4.193422067031535e-05,
+      "loss": 1.8076,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5516775734912226,
+      "grad_norm": 0.41505420207977295,
+      "learning_rate": 4.191811822889307e-05,
+      "loss": 1.7471,
+      "step": 5311
+    },
+    {
+      "epoch": 0.551781448010803,
+      "grad_norm": 0.42771202325820923,
+      "learning_rate": 4.190201664812816e-05,
+      "loss": 1.6095,
+      "step": 5312
+    },
+    {
+      "epoch": 0.5518853225303832,
+      "grad_norm": 0.41721728444099426,
+      "learning_rate": 4.188591592973534e-05,
+      "loss": 1.7719,
+      "step": 5313
+    },
+    {
+      "epoch": 0.5519891970499636,
+      "grad_norm": 0.4197467267513275,
+      "learning_rate": 4.1869816075429145e-05,
+      "loss": 1.7216,
+      "step": 5314
+    },
+    {
+      "epoch": 0.552093071569544,
+      "grad_norm": 0.43040528893470764,
+      "learning_rate": 4.1853717086924136e-05,
+      "loss": 1.7245,
+      "step": 5315
+    },
+    {
+      "epoch": 0.5521969460891243,
+      "grad_norm": 0.46574804186820984,
+      "learning_rate": 4.183761896593472e-05,
+      "loss": 1.8018,
+      "step": 5316
+    },
+    {
+      "epoch": 0.5523008206087047,
+      "grad_norm": 0.41133713722229004,
+      "learning_rate": 4.182152171417521e-05,
+      "loss": 1.628,
+      "step": 5317
+    },
+    {
+      "epoch": 0.552404695128285,
+      "grad_norm": 0.3880963623523712,
+      "learning_rate": 4.180542533335984e-05,
+      "loss": 1.6717,
+      "step": 5318
+    },
+    {
+      "epoch": 0.5525085696478654,
+      "grad_norm": 0.3985794484615326,
+      "learning_rate": 4.1789329825202736e-05,
+      "loss": 1.7005,
+      "step": 5319
+    },
+    {
+      "epoch": 0.5526124441674457,
+      "grad_norm": 0.43377330899238586,
+      "learning_rate": 4.177323519141799e-05,
+      "loss": 1.7466,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5527163186870261,
+      "grad_norm": 0.44802311062812805,
+      "learning_rate": 4.1757141433719506e-05,
+      "loss": 1.6648,
+      "step": 5321
+    },
+    {
+      "epoch": 0.5528201932066065,
+      "grad_norm": 0.4492226839065552,
+      "learning_rate": 4.174104855382117e-05,
+      "loss": 1.5708,
+      "step": 5322
+    },
+    {
+      "epoch": 0.5529240677261867,
+      "grad_norm": 0.4240928590297699,
+      "learning_rate": 4.1724956553436756e-05,
+      "loss": 1.6883,
+      "step": 5323
+    },
+    {
+      "epoch": 0.5530279422457671,
+      "grad_norm": 0.4252764582633972,
+      "learning_rate": 4.17088654342799e-05,
+      "loss": 1.7223,
+      "step": 5324
+    },
+    {
+      "epoch": 0.5531318167653475,
+      "grad_norm": 0.44476190209388733,
+      "learning_rate": 4.1692775198064194e-05,
+      "loss": 1.8179,
+      "step": 5325
+    },
+    {
+      "epoch": 0.5532356912849278,
+      "grad_norm": 0.38664236664772034,
+      "learning_rate": 4.167668584650316e-05,
+      "loss": 1.5664,
+      "step": 5326
+    },
+    {
+      "epoch": 0.5533395658045082,
+      "grad_norm": 0.3723793923854828,
+      "learning_rate": 4.166059738131013e-05,
+      "loss": 1.6096,
+      "step": 5327
+    },
+    {
+      "epoch": 0.5534434403240885,
+      "grad_norm": 0.4321557283401489,
+      "learning_rate": 4.164450980419844e-05,
+      "loss": 1.8148,
+      "step": 5328
+    },
+    {
+      "epoch": 0.5535473148436688,
+      "grad_norm": 0.40069714188575745,
+      "learning_rate": 4.162842311688129e-05,
+      "loss": 1.6699,
+      "step": 5329
+    },
+    {
+      "epoch": 0.5536511893632492,
+      "grad_norm": 0.4085390269756317,
+      "learning_rate": 4.161233732107177e-05,
+      "loss": 1.7729,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5537550638828296,
+      "grad_norm": 0.4812339246273041,
+      "learning_rate": 4.1596252418482884e-05,
+      "loss": 1.963,
+      "step": 5331
+    },
+    {
+      "epoch": 0.5538589384024099,
+      "grad_norm": 0.458366721868515,
+      "learning_rate": 4.158016841082759e-05,
+      "loss": 1.8112,
+      "step": 5332
+    },
+    {
+      "epoch": 0.5539628129219902,
+      "grad_norm": 0.4018287658691406,
+      "learning_rate": 4.1564085299818665e-05,
+      "loss": 1.7265,
+      "step": 5333
+    },
+    {
+      "epoch": 0.5540666874415706,
+      "grad_norm": 0.41217106580734253,
+      "learning_rate": 4.154800308716885e-05,
+      "loss": 1.7181,
+      "step": 5334
+    },
+    {
+      "epoch": 0.5541705619611509,
+      "grad_norm": 0.5003406405448914,
+      "learning_rate": 4.15319217745908e-05,
+      "loss": 1.8835,
+      "step": 5335
+    },
+    {
+      "epoch": 0.5542744364807313,
+      "grad_norm": 0.4172942340373993,
+      "learning_rate": 4.151584136379702e-05,
+      "loss": 1.6447,
+      "step": 5336
+    },
+    {
+      "epoch": 0.5543783110003117,
+      "grad_norm": 0.4157778322696686,
+      "learning_rate": 4.149976185649994e-05,
+      "loss": 1.7282,
+      "step": 5337
+    },
+    {
+      "epoch": 0.5544821855198919,
+      "grad_norm": 0.40136539936065674,
+      "learning_rate": 4.148368325441194e-05,
+      "loss": 1.7406,
+      "step": 5338
+    },
+    {
+      "epoch": 0.5545860600394723,
+      "grad_norm": 0.4425347149372101,
+      "learning_rate": 4.1467605559245226e-05,
+      "loss": 1.7426,
+      "step": 5339
+    },
+    {
+      "epoch": 0.5546899345590527,
+      "grad_norm": 0.44530102610588074,
+      "learning_rate": 4.145152877271196e-05,
+      "loss": 1.8979,
+      "step": 5340
+    },
+    {
+      "epoch": 0.554793809078633,
+      "grad_norm": 0.41977494955062866,
+      "learning_rate": 4.143545289652422e-05,
+      "loss": 1.7588,
+      "step": 5341
+    },
+    {
+      "epoch": 0.5548976835982133,
+      "grad_norm": 0.43756383657455444,
+      "learning_rate": 4.141937793239392e-05,
+      "loss": 1.6979,
+      "step": 5342
+    },
+    {
+      "epoch": 0.5550015581177937,
+      "grad_norm": 0.42378005385398865,
+      "learning_rate": 4.140330388203295e-05,
+      "loss": 1.6303,
+      "step": 5343
+    },
+    {
+      "epoch": 0.555105432637374,
+      "grad_norm": 0.39726606011390686,
+      "learning_rate": 4.138723074715306e-05,
+      "loss": 1.6592,
+      "step": 5344
+    },
+    {
+      "epoch": 0.5552093071569544,
+      "grad_norm": 0.4202333092689514,
+      "learning_rate": 4.1371158529465916e-05,
+      "loss": 1.7364,
+      "step": 5345
+    },
+    {
+      "epoch": 0.5553131816765348,
+      "grad_norm": 0.40389859676361084,
+      "learning_rate": 4.1355087230683065e-05,
+      "loss": 1.5759,
+      "step": 5346
+    },
+    {
+      "epoch": 0.5554170561961151,
+      "grad_norm": 0.37404200434684753,
+      "learning_rate": 4.1339016852516e-05,
+      "loss": 1.544,
+      "step": 5347
+    },
+    {
+      "epoch": 0.5555209307156954,
+      "grad_norm": 0.45083087682724,
+      "learning_rate": 4.132294739667609e-05,
+      "loss": 1.8926,
+      "step": 5348
+    },
+    {
+      "epoch": 0.5556248052352758,
+      "grad_norm": 0.4377882480621338,
+      "learning_rate": 4.130687886487459e-05,
+      "loss": 1.741,
+      "step": 5349
+    },
+    {
+      "epoch": 0.5557286797548562,
+      "grad_norm": 0.39583009481430054,
+      "learning_rate": 4.1290811258822685e-05,
+      "loss": 1.6806,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5558325542744365,
+      "grad_norm": 0.48825472593307495,
+      "learning_rate": 4.127474458023146e-05,
+      "loss": 1.9815,
+      "step": 5351
+    },
+    {
+      "epoch": 0.5559364287940168,
+      "grad_norm": 0.39083537459373474,
+      "learning_rate": 4.125867883081186e-05,
+      "loss": 1.7249,
+      "step": 5352
+    },
+    {
+      "epoch": 0.5560403033135972,
+      "grad_norm": 0.385078489780426,
+      "learning_rate": 4.124261401227477e-05,
+      "loss": 1.546,
+      "step": 5353
+    },
+    {
+      "epoch": 0.5561441778331775,
+      "grad_norm": 0.41559967398643494,
+      "learning_rate": 4.122655012633101e-05,
+      "loss": 1.6632,
+      "step": 5354
+    },
+    {
+      "epoch": 0.5562480523527579,
+      "grad_norm": 0.4069918990135193,
+      "learning_rate": 4.121048717469119e-05,
+      "loss": 1.6128,
+      "step": 5355
+    },
+    {
+      "epoch": 0.5563519268723383,
+      "grad_norm": 0.43804723024368286,
+      "learning_rate": 4.119442515906595e-05,
+      "loss": 1.618,
+      "step": 5356
+    },
+    {
+      "epoch": 0.5564558013919185,
+      "grad_norm": 0.46744030714035034,
+      "learning_rate": 4.117836408116574e-05,
+      "loss": 1.787,
+      "step": 5357
+    },
+    {
+      "epoch": 0.5565596759114989,
+      "grad_norm": 0.4054070711135864,
+      "learning_rate": 4.116230394270094e-05,
+      "loss": 1.6505,
+      "step": 5358
+    },
+    {
+      "epoch": 0.5566635504310793,
+      "grad_norm": 0.4278983771800995,
+      "learning_rate": 4.114624474538183e-05,
+      "loss": 1.7579,
+      "step": 5359
+    },
+    {
+      "epoch": 0.5567674249506596,
+      "grad_norm": 0.4365118443965912,
+      "learning_rate": 4.113018649091861e-05,
+      "loss": 1.7672,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5568712994702399,
+      "grad_norm": 0.4227050244808197,
+      "learning_rate": 4.111412918102132e-05,
+      "loss": 1.8521,
+      "step": 5361
+    },
+    {
+      "epoch": 0.5569751739898203,
+      "grad_norm": 0.41989371180534363,
+      "learning_rate": 4.1098072817399956e-05,
+      "loss": 1.6816,
+      "step": 5362
+    },
+    {
+      "epoch": 0.5570790485094006,
+      "grad_norm": 0.43567296862602234,
+      "learning_rate": 4.108201740176443e-05,
+      "loss": 1.7746,
+      "step": 5363
+    },
+    {
+      "epoch": 0.557182923028981,
+      "grad_norm": 0.48267829418182373,
+      "learning_rate": 4.106596293582446e-05,
+      "loss": 1.7837,
+      "step": 5364
+    },
+    {
+      "epoch": 0.5572867975485614,
+      "grad_norm": 0.44921717047691345,
+      "learning_rate": 4.104990942128976e-05,
+      "loss": 1.8066,
+      "step": 5365
+    },
+    {
+      "epoch": 0.5573906720681416,
+      "grad_norm": 0.4056643843650818,
+      "learning_rate": 4.10338568598699e-05,
+      "loss": 1.472,
+      "step": 5366
+    },
+    {
+      "epoch": 0.557494546587722,
+      "grad_norm": 0.4002752900123596,
+      "learning_rate": 4.1017805253274336e-05,
+      "loss": 1.6328,
+      "step": 5367
+    },
+    {
+      "epoch": 0.5575984211073024,
+      "grad_norm": 0.42509204149246216,
+      "learning_rate": 4.100175460321244e-05,
+      "loss": 1.5223,
+      "step": 5368
+    },
+    {
+      "epoch": 0.5577022956268827,
+      "grad_norm": 0.391414612531662,
+      "learning_rate": 4.0985704911393506e-05,
+      "loss": 1.685,
+      "step": 5369
+    },
+    {
+      "epoch": 0.557806170146463,
+      "grad_norm": 0.4597155749797821,
+      "learning_rate": 4.096965617952667e-05,
+      "loss": 1.7821,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5579100446660434,
+      "grad_norm": 0.4139951467514038,
+      "learning_rate": 4.0953608409321024e-05,
+      "loss": 1.7831,
+      "step": 5371
+    },
+    {
+      "epoch": 0.5580139191856238,
+      "grad_norm": 0.4334876239299774,
+      "learning_rate": 4.093756160248551e-05,
+      "loss": 1.6776,
+      "step": 5372
+    },
+    {
+      "epoch": 0.5581177937052041,
+      "grad_norm": 0.47372686862945557,
+      "learning_rate": 4.092151576072901e-05,
+      "loss": 1.8743,
+      "step": 5373
+    },
+    {
+      "epoch": 0.5582216682247845,
+      "grad_norm": 0.39616334438323975,
+      "learning_rate": 4.0905470885760244e-05,
+      "loss": 1.6113,
+      "step": 5374
+    },
+    {
+      "epoch": 0.5583255427443649,
+      "grad_norm": 0.44439664483070374,
+      "learning_rate": 4.088942697928789e-05,
+      "loss": 1.842,
+      "step": 5375
+    },
+    {
+      "epoch": 0.5584294172639451,
+      "grad_norm": 0.3774793744087219,
+      "learning_rate": 4.087338404302052e-05,
+      "loss": 1.7603,
+      "step": 5376
+    },
+    {
+      "epoch": 0.5585332917835255,
+      "grad_norm": 0.4046783745288849,
+      "learning_rate": 4.0857342078666546e-05,
+      "loss": 1.597,
+      "step": 5377
+    },
+    {
+      "epoch": 0.5586371663031059,
+      "grad_norm": 0.41748616099357605,
+      "learning_rate": 4.0841301087934326e-05,
+      "loss": 1.7326,
+      "step": 5378
+    },
+    {
+      "epoch": 0.5587410408226862,
+      "grad_norm": 0.3903839886188507,
+      "learning_rate": 4.082526107253212e-05,
+      "loss": 1.5858,
+      "step": 5379
+    },
+    {
+      "epoch": 0.5588449153422665,
+      "grad_norm": 0.4938502311706543,
+      "learning_rate": 4.080922203416801e-05,
+      "loss": 1.5894,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5589487898618469,
+      "grad_norm": 0.4077463150024414,
+      "learning_rate": 4.079318397455007e-05,
+      "loss": 1.6721,
+      "step": 5381
+    },
+    {
+      "epoch": 0.5590526643814272,
+      "grad_norm": 0.42645683884620667,
+      "learning_rate": 4.077714689538624e-05,
+      "loss": 1.8013,
+      "step": 5382
+    },
+    {
+      "epoch": 0.5591565389010076,
+      "grad_norm": 0.3947215676307678,
+      "learning_rate": 4.076111079838431e-05,
+      "loss": 1.6274,
+      "step": 5383
+    },
+    {
+      "epoch": 0.559260413420588,
+      "grad_norm": 0.43122047185897827,
+      "learning_rate": 4.074507568525201e-05,
+      "loss": 1.7273,
+      "step": 5384
+    },
+    {
+      "epoch": 0.5593642879401682,
+      "grad_norm": 0.41860494017601013,
+      "learning_rate": 4.072904155769697e-05,
+      "loss": 1.6412,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5594681624597486,
+      "grad_norm": 0.4793721139431,
+      "learning_rate": 4.0713008417426676e-05,
+      "loss": 1.8271,
+      "step": 5386
+    },
+    {
+      "epoch": 0.559572036979329,
+      "grad_norm": 0.4130531847476959,
+      "learning_rate": 4.069697626614855e-05,
+      "loss": 1.7992,
+      "step": 5387
+    },
+    {
+      "epoch": 0.5596759114989093,
+      "grad_norm": 0.40088602900505066,
+      "learning_rate": 4.068094510556989e-05,
+      "loss": 1.4903,
+      "step": 5388
+    },
+    {
+      "epoch": 0.5597797860184897,
+      "grad_norm": 0.40667760372161865,
+      "learning_rate": 4.066491493739788e-05,
+      "loss": 1.5853,
+      "step": 5389
+    },
+    {
+      "epoch": 0.55988366053807,
+      "grad_norm": 0.44721919298171997,
+      "learning_rate": 4.064888576333961e-05,
+      "loss": 1.7506,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5599875350576503,
+      "grad_norm": 0.4468611478805542,
+      "learning_rate": 4.063285758510207e-05,
+      "loss": 1.6431,
+      "step": 5391
+    },
+    {
+      "epoch": 0.5600914095772307,
+      "grad_norm": 0.39779260754585266,
+      "learning_rate": 4.061683040439214e-05,
+      "loss": 1.7715,
+      "step": 5392
+    },
+    {
+      "epoch": 0.5601952840968111,
+      "grad_norm": 0.41888242959976196,
+      "learning_rate": 4.060080422291656e-05,
+      "loss": 1.6373,
+      "step": 5393
+    },
+    {
+      "epoch": 0.5602991586163913,
+      "grad_norm": 0.4231390953063965,
+      "learning_rate": 4.058477904238203e-05,
+      "loss": 1.7707,
+      "step": 5394
+    },
+    {
+      "epoch": 0.5604030331359717,
+      "grad_norm": 0.4443194568157196,
+      "learning_rate": 4.056875486449507e-05,
+      "loss": 1.8821,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5605069076555521,
+      "grad_norm": 0.4171590507030487,
+      "learning_rate": 4.055273169096216e-05,
+      "loss": 1.7404,
+      "step": 5396
+    },
+    {
+      "epoch": 0.5606107821751325,
+      "grad_norm": 0.41153210401535034,
+      "learning_rate": 4.053670952348964e-05,
+      "loss": 1.679,
+      "step": 5397
+    },
+    {
+      "epoch": 0.5607146566947128,
+      "grad_norm": 0.40622779726982117,
+      "learning_rate": 4.052068836378374e-05,
+      "loss": 1.69,
+      "step": 5398
+    },
+    {
+      "epoch": 0.5608185312142931,
+      "grad_norm": 0.40083175897598267,
+      "learning_rate": 4.050466821355058e-05,
+      "loss": 1.6595,
+      "step": 5399
+    },
+    {
+      "epoch": 0.5609224057338735,
+      "grad_norm": 0.4463566243648529,
+      "learning_rate": 4.0488649074496186e-05,
+      "loss": 1.7722,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5610262802534538,
+      "grad_norm": 0.39217326045036316,
+      "learning_rate": 4.04726309483265e-05,
+      "loss": 1.5485,
+      "step": 5401
+    },
+    {
+      "epoch": 0.5611301547730342,
+      "grad_norm": 0.39511242508888245,
+      "learning_rate": 4.045661383674727e-05,
+      "loss": 1.6444,
+      "step": 5402
+    },
+    {
+      "epoch": 0.5612340292926146,
+      "grad_norm": 0.411687970161438,
+      "learning_rate": 4.044059774146423e-05,
+      "loss": 1.8602,
+      "step": 5403
+    },
+    {
+      "epoch": 0.5613379038121948,
+      "grad_norm": 0.38304659724235535,
+      "learning_rate": 4.0424582664182986e-05,
+      "loss": 1.6232,
+      "step": 5404
+    },
+    {
+      "epoch": 0.5614417783317752,
+      "grad_norm": 0.4158017635345459,
+      "learning_rate": 4.040856860660897e-05,
+      "loss": 1.6325,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5615456528513556,
+      "grad_norm": 0.43518683314323425,
+      "learning_rate": 4.03925555704476e-05,
+      "loss": 1.6961,
+      "step": 5406
+    },
+    {
+      "epoch": 0.5616495273709359,
+      "grad_norm": 0.3899015486240387,
+      "learning_rate": 4.0376543557404125e-05,
+      "loss": 1.6689,
+      "step": 5407
+    },
+    {
+      "epoch": 0.5617534018905163,
+      "grad_norm": 0.3894193470478058,
+      "learning_rate": 4.036053256918368e-05,
+      "loss": 1.5847,
+      "step": 5408
+    },
+    {
+      "epoch": 0.5618572764100966,
+      "grad_norm": 0.41329947113990784,
+      "learning_rate": 4.034452260749132e-05,
+      "loss": 1.7453,
+      "step": 5409
+    },
+    {
+      "epoch": 0.5619611509296769,
+      "grad_norm": 0.43754690885543823,
+      "learning_rate": 4.032851367403201e-05,
+      "loss": 1.7059,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5620650254492573,
+      "grad_norm": 0.423709899187088,
+      "learning_rate": 4.031250577051053e-05,
+      "loss": 1.6802,
+      "step": 5411
+    },
+    {
+      "epoch": 0.5621688999688377,
+      "grad_norm": 0.40563976764678955,
+      "learning_rate": 4.0296498898631644e-05,
+      "loss": 1.7492,
+      "step": 5412
+    },
+    {
+      "epoch": 0.562272774488418,
+      "grad_norm": 0.40074846148490906,
+      "learning_rate": 4.028049306009992e-05,
+      "loss": 1.7393,
+      "step": 5413
+    },
+    {
+      "epoch": 0.5623766490079983,
+      "grad_norm": 0.4513157308101654,
+      "learning_rate": 4.026448825661987e-05,
+      "loss": 1.7905,
+      "step": 5414
+    },
+    {
+      "epoch": 0.5624805235275787,
+      "grad_norm": 0.4043268859386444,
+      "learning_rate": 4.024848448989588e-05,
+      "loss": 1.8213,
+      "step": 5415
+    },
+    {
+      "epoch": 0.562584398047159,
+      "grad_norm": 0.36234188079833984,
+      "learning_rate": 4.023248176163224e-05,
+      "loss": 1.4981,
+      "step": 5416
+    },
+    {
+      "epoch": 0.5626882725667394,
+      "grad_norm": 0.42101240158081055,
+      "learning_rate": 4.021648007353309e-05,
+      "loss": 1.6849,
+      "step": 5417
+    },
+    {
+      "epoch": 0.5627921470863197,
+      "grad_norm": 0.3881211280822754,
+      "learning_rate": 4.0200479427302486e-05,
+      "loss": 1.5648,
+      "step": 5418
+    },
+    {
+      "epoch": 0.5628960216059,
+      "grad_norm": 0.4609657824039459,
+      "learning_rate": 4.0184479824644413e-05,
+      "loss": 1.8176,
+      "step": 5419
+    },
+    {
+      "epoch": 0.5629998961254804,
+      "grad_norm": 0.4178425669670105,
+      "learning_rate": 4.016848126726266e-05,
+      "loss": 1.6165,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5631037706450608,
+      "grad_norm": 0.390349805355072,
+      "learning_rate": 4.0152483756860956e-05,
+      "loss": 1.5583,
+      "step": 5421
+    },
+    {
+      "epoch": 0.5632076451646412,
+      "grad_norm": 0.40884721279144287,
+      "learning_rate": 4.0136487295142913e-05,
+      "loss": 1.7998,
+      "step": 5422
+    },
+    {
+      "epoch": 0.5633115196842214,
+      "grad_norm": 0.4485437273979187,
+      "learning_rate": 4.0120491883812065e-05,
+      "loss": 2.054,
+      "step": 5423
+    },
+    {
+      "epoch": 0.5634153942038018,
+      "grad_norm": 0.43038633465766907,
+      "learning_rate": 4.010449752457175e-05,
+      "loss": 1.7295,
+      "step": 5424
+    },
+    {
+      "epoch": 0.5635192687233822,
+      "grad_norm": 0.36314454674720764,
+      "learning_rate": 4.0088504219125253e-05,
+      "loss": 1.21,
+      "step": 5425
+    },
+    {
+      "epoch": 0.5636231432429625,
+      "grad_norm": 0.3857851028442383,
+      "learning_rate": 4.007251196917576e-05,
+      "loss": 1.6524,
+      "step": 5426
+    },
+    {
+      "epoch": 0.5637270177625429,
+      "grad_norm": 0.3993614912033081,
+      "learning_rate": 4.00565207764263e-05,
+      "loss": 1.6198,
+      "step": 5427
+    },
+    {
+      "epoch": 0.5638308922821232,
+      "grad_norm": 0.41268518567085266,
+      "learning_rate": 4.00405306425798e-05,
+      "loss": 1.6297,
+      "step": 5428
+    },
+    {
+      "epoch": 0.5639347668017035,
+      "grad_norm": 0.410853773355484,
+      "learning_rate": 4.0024541569339126e-05,
+      "loss": 1.6958,
+      "step": 5429
+    },
+    {
+      "epoch": 0.5640386413212839,
+      "grad_norm": 0.39320501685142517,
+      "learning_rate": 4.0008553558406945e-05,
+      "loss": 1.5767,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5641425158408643,
+      "grad_norm": 0.444997102022171,
+      "learning_rate": 3.9992566611485866e-05,
+      "loss": 1.6399,
+      "step": 5431
+    },
+    {
+      "epoch": 0.5642463903604445,
+      "grad_norm": 0.4061078131198883,
+      "learning_rate": 3.9976580730278405e-05,
+      "loss": 1.485,
+      "step": 5432
+    },
+    {
+      "epoch": 0.5643502648800249,
+      "grad_norm": 0.38394030928611755,
+      "learning_rate": 3.99605959164869e-05,
+      "loss": 1.6078,
+      "step": 5433
+    },
+    {
+      "epoch": 0.5644541393996053,
+      "grad_norm": 0.46738219261169434,
+      "learning_rate": 3.9944612171813616e-05,
+      "loss": 1.7178,
+      "step": 5434
+    },
+    {
+      "epoch": 0.5645580139191856,
+      "grad_norm": 0.47596266865730286,
+      "learning_rate": 3.992862949796072e-05,
+      "loss": 1.7114,
+      "step": 5435
+    },
+    {
+      "epoch": 0.564661888438766,
+      "grad_norm": 0.4068540930747986,
+      "learning_rate": 3.991264789663019e-05,
+      "loss": 1.5976,
+      "step": 5436
+    },
+    {
+      "epoch": 0.5647657629583464,
+      "grad_norm": 0.4116705060005188,
+      "learning_rate": 3.989666736952399e-05,
+      "loss": 1.7029,
+      "step": 5437
+    },
+    {
+      "epoch": 0.5648696374779266,
+      "grad_norm": 0.37593719363212585,
+      "learning_rate": 3.9880687918343916e-05,
+      "loss": 1.5416,
+      "step": 5438
+    },
+    {
+      "epoch": 0.564973511997507,
+      "grad_norm": 0.43204763531684875,
+      "learning_rate": 3.986470954479163e-05,
+      "loss": 1.7622,
+      "step": 5439
+    },
+    {
+      "epoch": 0.5650773865170874,
+      "grad_norm": 0.4367099106311798,
+      "learning_rate": 3.984873225056873e-05,
+      "loss": 1.6935,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5651812610366677,
+      "grad_norm": 0.3884550631046295,
+      "learning_rate": 3.983275603737666e-05,
+      "loss": 1.6477,
+      "step": 5441
+    },
+    {
+      "epoch": 0.565285135556248,
+      "grad_norm": 0.40880388021469116,
+      "learning_rate": 3.981678090691676e-05,
+      "loss": 1.7526,
+      "step": 5442
+    },
+    {
+      "epoch": 0.5653890100758284,
+      "grad_norm": 0.40109533071517944,
+      "learning_rate": 3.980080686089026e-05,
+      "loss": 1.7972,
+      "step": 5443
+    },
+    {
+      "epoch": 0.5654928845954087,
+      "grad_norm": 0.4425828456878662,
+      "learning_rate": 3.9784833900998286e-05,
+      "loss": 1.738,
+      "step": 5444
+    },
+    {
+      "epoch": 0.5655967591149891,
+      "grad_norm": 0.4051485061645508,
+      "learning_rate": 3.97688620289418e-05,
+      "loss": 1.6437,
+      "step": 5445
+    },
+    {
+      "epoch": 0.5657006336345695,
+      "grad_norm": 0.3986642360687256,
+      "learning_rate": 3.975289124642171e-05,
+      "loss": 1.6672,
+      "step": 5446
+    },
+    {
+      "epoch": 0.5658045081541497,
+      "grad_norm": 0.3762451112270355,
+      "learning_rate": 3.9736921555138787e-05,
+      "loss": 1.6666,
+      "step": 5447
+    },
+    {
+      "epoch": 0.5659083826737301,
+      "grad_norm": 0.3796166479587555,
+      "learning_rate": 3.972095295679365e-05,
+      "loss": 1.6013,
+      "step": 5448
+    },
+    {
+      "epoch": 0.5660122571933105,
+      "grad_norm": 0.4417518377304077,
+      "learning_rate": 3.9704985453086833e-05,
+      "loss": 1.6795,
+      "step": 5449
+    },
+    {
+      "epoch": 0.5661161317128909,
+      "grad_norm": 0.3697144389152527,
+      "learning_rate": 3.968901904571877e-05,
+      "loss": 1.4858,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5662200062324712,
+      "grad_norm": 0.3945258557796478,
+      "learning_rate": 3.967305373638976e-05,
+      "loss": 1.5509,
+      "step": 5451
+    },
+    {
+      "epoch": 0.5663238807520515,
+      "grad_norm": 0.39280179142951965,
+      "learning_rate": 3.965708952679996e-05,
+      "loss": 1.6271,
+      "step": 5452
+    },
+    {
+      "epoch": 0.5664277552716319,
+      "grad_norm": 0.4188973605632782,
+      "learning_rate": 3.9641126418649456e-05,
+      "loss": 1.656,
+      "step": 5453
+    },
+    {
+      "epoch": 0.5665316297912122,
+      "grad_norm": 0.39895448088645935,
+      "learning_rate": 3.962516441363818e-05,
+      "loss": 1.6666,
+      "step": 5454
+    },
+    {
+      "epoch": 0.5666355043107926,
+      "grad_norm": 0.38310202956199646,
+      "learning_rate": 3.9609203513465974e-05,
+      "loss": 1.5975,
+      "step": 5455
+    },
+    {
+      "epoch": 0.566739378830373,
+      "grad_norm": 0.43473097681999207,
+      "learning_rate": 3.9593243719832536e-05,
+      "loss": 1.7687,
+      "step": 5456
+    },
+    {
+      "epoch": 0.5668432533499532,
+      "grad_norm": 0.4295366406440735,
+      "learning_rate": 3.957728503443748e-05,
+      "loss": 1.6263,
+      "step": 5457
+    },
+    {
+      "epoch": 0.5669471278695336,
+      "grad_norm": 0.4174087643623352,
+      "learning_rate": 3.956132745898025e-05,
+      "loss": 1.6705,
+      "step": 5458
+    },
+    {
+      "epoch": 0.567051002389114,
+      "grad_norm": 0.41141003370285034,
+      "learning_rate": 3.954537099516021e-05,
+      "loss": 1.6423,
+      "step": 5459
+    },
+    {
+      "epoch": 0.5671548769086943,
+      "grad_norm": 0.44283923506736755,
+      "learning_rate": 3.952941564467665e-05,
+      "loss": 1.7335,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5672587514282746,
+      "grad_norm": 0.4794785678386688,
+      "learning_rate": 3.951346140922863e-05,
+      "loss": 1.7162,
+      "step": 5461
+    },
+    {
+      "epoch": 0.567362625947855,
+      "grad_norm": 0.39859795570373535,
+      "learning_rate": 3.949750829051516e-05,
+      "loss": 1.7051,
+      "step": 5462
+    },
+    {
+      "epoch": 0.5674665004674353,
+      "grad_norm": 0.39325210452079773,
+      "learning_rate": 3.948155629023515e-05,
+      "loss": 1.5073,
+      "step": 5463
+    },
+    {
+      "epoch": 0.5675703749870157,
+      "grad_norm": 0.37789732217788696,
+      "learning_rate": 3.946560541008734e-05,
+      "loss": 1.7606,
+      "step": 5464
+    },
+    {
+      "epoch": 0.5676742495065961,
+      "grad_norm": 0.4179360270500183,
+      "learning_rate": 3.9449655651770365e-05,
+      "loss": 1.576,
+      "step": 5465
+    },
+    {
+      "epoch": 0.5677781240261763,
+      "grad_norm": 0.38552039861679077,
+      "learning_rate": 3.943370701698281e-05,
+      "loss": 1.6677,
+      "step": 5466
+    },
+    {
+      "epoch": 0.5678819985457567,
+      "grad_norm": 0.4086502194404602,
+      "learning_rate": 3.9417759507423004e-05,
+      "loss": 1.7007,
+      "step": 5467
+    },
+    {
+      "epoch": 0.5679858730653371,
+      "grad_norm": 0.43782469630241394,
+      "learning_rate": 3.940181312478928e-05,
+      "loss": 1.788,
+      "step": 5468
+    },
+    {
+      "epoch": 0.5680897475849174,
+      "grad_norm": 0.40869951248168945,
+      "learning_rate": 3.938586787077978e-05,
+      "loss": 1.7881,
+      "step": 5469
+    },
+    {
+      "epoch": 0.5681936221044978,
+      "grad_norm": 0.4343576729297638,
+      "learning_rate": 3.936992374709256e-05,
+      "loss": 1.8009,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5682974966240781,
+      "grad_norm": 0.4356035590171814,
+      "learning_rate": 3.935398075542554e-05,
+      "loss": 1.7802,
+      "step": 5471
+    },
+    {
+      "epoch": 0.5684013711436584,
+      "grad_norm": 0.4507029950618744,
+      "learning_rate": 3.933803889747655e-05,
+      "loss": 1.5211,
+      "step": 5472
+    },
+    {
+      "epoch": 0.5685052456632388,
+      "grad_norm": 0.3965437114238739,
+      "learning_rate": 3.9322098174943226e-05,
+      "loss": 1.6319,
+      "step": 5473
+    },
+    {
+      "epoch": 0.5686091201828192,
+      "grad_norm": 0.4066412150859833,
+      "learning_rate": 3.930615858952316e-05,
+      "loss": 1.4707,
+      "step": 5474
+    },
+    {
+      "epoch": 0.5687129947023996,
+      "grad_norm": 0.4407905638217926,
+      "learning_rate": 3.929022014291379e-05,
+      "loss": 1.7746,
+      "step": 5475
+    },
+    {
+      "epoch": 0.5688168692219798,
+      "grad_norm": 0.43420854210853577,
+      "learning_rate": 3.927428283681245e-05,
+      "loss": 1.7012,
+      "step": 5476
+    },
+    {
+      "epoch": 0.5689207437415602,
+      "grad_norm": 0.39704859256744385,
+      "learning_rate": 3.925834667291631e-05,
+      "loss": 1.623,
+      "step": 5477
+    },
+    {
+      "epoch": 0.5690246182611406,
+      "grad_norm": 0.39629414677619934,
+      "learning_rate": 3.924241165292245e-05,
+      "loss": 1.6304,
+      "step": 5478
+    },
+    {
+      "epoch": 0.5691284927807209,
+      "grad_norm": 0.4737340211868286,
+      "learning_rate": 3.922647777852786e-05,
+      "loss": 1.7869,
+      "step": 5479
+    },
+    {
+      "epoch": 0.5692323673003012,
+      "grad_norm": 0.41576650738716125,
+      "learning_rate": 3.9210545051429337e-05,
+      "loss": 1.7863,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5693362418198816,
+      "grad_norm": 0.4058329463005066,
+      "learning_rate": 3.919461347332361e-05,
+      "loss": 1.7295,
+      "step": 5481
+    },
+    {
+      "epoch": 0.5694401163394619,
+      "grad_norm": 0.38741639256477356,
+      "learning_rate": 3.917868304590726e-05,
+      "loss": 1.5988,
+      "step": 5482
+    },
+    {
+      "epoch": 0.5695439908590423,
+      "grad_norm": 0.40862467885017395,
+      "learning_rate": 3.916275377087676e-05,
+      "loss": 1.6312,
+      "step": 5483
+    },
+    {
+      "epoch": 0.5696478653786227,
+      "grad_norm": 0.38968440890312195,
+      "learning_rate": 3.9146825649928446e-05,
+      "loss": 1.7144,
+      "step": 5484
+    },
+    {
+      "epoch": 0.5697517398982029,
+      "grad_norm": 0.41511067748069763,
+      "learning_rate": 3.913089868475855e-05,
+      "loss": 1.8591,
+      "step": 5485
+    },
+    {
+      "epoch": 0.5698556144177833,
+      "grad_norm": 0.4132608473300934,
+      "learning_rate": 3.911497287706315e-05,
+      "loss": 1.7331,
+      "step": 5486
+    },
+    {
+      "epoch": 0.5699594889373637,
+      "grad_norm": 0.41503503918647766,
+      "learning_rate": 3.909904822853823e-05,
+      "loss": 1.867,
+      "step": 5487
+    },
+    {
+      "epoch": 0.570063363456944,
+      "grad_norm": 0.39798980951309204,
+      "learning_rate": 3.908312474087966e-05,
+      "loss": 1.7591,
+      "step": 5488
+    },
+    {
+      "epoch": 0.5701672379765244,
+      "grad_norm": 0.4197697639465332,
+      "learning_rate": 3.906720241578314e-05,
+      "loss": 1.7974,
+      "step": 5489
+    },
+    {
+      "epoch": 0.5702711124961047,
+      "grad_norm": 0.47081831097602844,
+      "learning_rate": 3.905128125494427e-05,
+      "loss": 1.7685,
+      "step": 5490
+    },
+    {
+      "epoch": 0.570374987015685,
+      "grad_norm": 0.42261427640914917,
+      "learning_rate": 3.903536126005858e-05,
+      "loss": 1.7431,
+      "step": 5491
+    },
+    {
+      "epoch": 0.5704788615352654,
+      "grad_norm": 0.39490261673927307,
+      "learning_rate": 3.901944243282135e-05,
+      "loss": 1.8147,
+      "step": 5492
+    },
+    {
+      "epoch": 0.5705827360548458,
+      "grad_norm": 0.41960594058036804,
+      "learning_rate": 3.900352477492786e-05,
+      "loss": 1.7233,
+      "step": 5493
+    },
+    {
+      "epoch": 0.570686610574426,
+      "grad_norm": 0.3711908161640167,
+      "learning_rate": 3.8987608288073216e-05,
+      "loss": 1.5517,
+      "step": 5494
+    },
+    {
+      "epoch": 0.5707904850940064,
+      "grad_norm": 0.433907151222229,
+      "learning_rate": 3.897169297395238e-05,
+      "loss": 1.6967,
+      "step": 5495
+    },
+    {
+      "epoch": 0.5708943596135868,
+      "grad_norm": 0.40768545866012573,
+      "learning_rate": 3.895577883426022e-05,
+      "loss": 1.4462,
+      "step": 5496
+    },
+    {
+      "epoch": 0.5709982341331671,
+      "grad_norm": 0.4261188209056854,
+      "learning_rate": 3.893986587069147e-05,
+      "loss": 1.8624,
+      "step": 5497
+    },
+    {
+      "epoch": 0.5711021086527475,
+      "grad_norm": 0.3955959379673004,
+      "learning_rate": 3.892395408494073e-05,
+      "loss": 1.6179,
+      "step": 5498
+    },
+    {
+      "epoch": 0.5712059831723278,
+      "grad_norm": 0.4365078806877136,
+      "learning_rate": 3.890804347870248e-05,
+      "loss": 1.7387,
+      "step": 5499
+    },
+    {
+      "epoch": 0.5713098576919082,
+      "grad_norm": 0.4353174865245819,
+      "learning_rate": 3.889213405367107e-05,
+      "loss": 1.4776,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5714137322114885,
+      "grad_norm": 0.39252567291259766,
+      "learning_rate": 3.887622581154077e-05,
+      "loss": 1.6338,
+      "step": 5501
+    },
+    {
+      "epoch": 0.5715176067310689,
+      "grad_norm": 0.49222537875175476,
+      "learning_rate": 3.886031875400562e-05,
+      "loss": 1.9512,
+      "step": 5502
+    },
+    {
+      "epoch": 0.5716214812506493,
+      "grad_norm": 0.41813385486602783,
+      "learning_rate": 3.884441288275965e-05,
+      "loss": 1.7856,
+      "step": 5503
+    },
+    {
+      "epoch": 0.5717253557702295,
+      "grad_norm": 0.4292755424976349,
+      "learning_rate": 3.8828508199496694e-05,
+      "loss": 1.6817,
+      "step": 5504
+    },
+    {
+      "epoch": 0.5718292302898099,
+      "grad_norm": 0.4089149832725525,
+      "learning_rate": 3.8812604705910454e-05,
+      "loss": 1.803,
+      "step": 5505
+    },
+    {
+      "epoch": 0.5719331048093903,
+      "grad_norm": 0.44375675916671753,
+      "learning_rate": 3.8796702403694554e-05,
+      "loss": 1.5654,
+      "step": 5506
+    },
+    {
+      "epoch": 0.5720369793289706,
+      "grad_norm": 0.3902745246887207,
+      "learning_rate": 3.8780801294542476e-05,
+      "loss": 1.6287,
+      "step": 5507
+    },
+    {
+      "epoch": 0.572140853848551,
+      "grad_norm": 0.38409364223480225,
+      "learning_rate": 3.876490138014752e-05,
+      "loss": 1.5963,
+      "step": 5508
+    },
+    {
+      "epoch": 0.5722447283681313,
+      "grad_norm": 0.41062119603157043,
+      "learning_rate": 3.874900266220295e-05,
+      "loss": 1.6819,
+      "step": 5509
+    },
+    {
+      "epoch": 0.5723486028877116,
+      "grad_norm": 0.38653069734573364,
+      "learning_rate": 3.873310514240183e-05,
+      "loss": 1.5166,
+      "step": 5510
+    },
+    {
+      "epoch": 0.572452477407292,
+      "grad_norm": 0.38248497247695923,
+      "learning_rate": 3.871720882243712e-05,
+      "loss": 1.505,
+      "step": 5511
+    },
+    {
+      "epoch": 0.5725563519268724,
+      "grad_norm": 0.41657835245132446,
+      "learning_rate": 3.870131370400166e-05,
+      "loss": 1.6057,
+      "step": 5512
+    },
+    {
+      "epoch": 0.5726602264464526,
+      "grad_norm": 0.4237949252128601,
+      "learning_rate": 3.868541978878818e-05,
+      "loss": 1.6186,
+      "step": 5513
+    },
+    {
+      "epoch": 0.572764100966033,
+      "grad_norm": 0.4268006682395935,
+      "learning_rate": 3.866952707848922e-05,
+      "loss": 1.5518,
+      "step": 5514
+    },
+    {
+      "epoch": 0.5728679754856134,
+      "grad_norm": 0.46735864877700806,
+      "learning_rate": 3.865363557479723e-05,
+      "loss": 1.7631,
+      "step": 5515
+    },
+    {
+      "epoch": 0.5729718500051937,
+      "grad_norm": 0.45263731479644775,
+      "learning_rate": 3.863774527940457e-05,
+      "loss": 1.7801,
+      "step": 5516
+    },
+    {
+      "epoch": 0.5730757245247741,
+      "grad_norm": 0.38436025381088257,
+      "learning_rate": 3.86218561940034e-05,
+      "loss": 1.4836,
+      "step": 5517
+    },
+    {
+      "epoch": 0.5731795990443544,
+      "grad_norm": 0.3812609910964966,
+      "learning_rate": 3.860596832028579e-05,
+      "loss": 1.5733,
+      "step": 5518
+    },
+    {
+      "epoch": 0.5732834735639347,
+      "grad_norm": 0.4150224030017853,
+      "learning_rate": 3.8590081659943694e-05,
+      "loss": 1.7095,
+      "step": 5519
+    },
+    {
+      "epoch": 0.5733873480835151,
+      "grad_norm": 0.4313552975654602,
+      "learning_rate": 3.8574196214668876e-05,
+      "loss": 1.7818,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5734912226030955,
+      "grad_norm": 0.3813556730747223,
+      "learning_rate": 3.855831198615304e-05,
+      "loss": 1.5572,
+      "step": 5521
+    },
+    {
+      "epoch": 0.5735950971226758,
+      "grad_norm": 0.3927074670791626,
+      "learning_rate": 3.854242897608775e-05,
+      "loss": 1.4641,
+      "step": 5522
+    },
+    {
+      "epoch": 0.5736989716422561,
+      "grad_norm": 0.38922983407974243,
+      "learning_rate": 3.852654718616438e-05,
+      "loss": 1.3964,
+      "step": 5523
+    },
+    {
+      "epoch": 0.5738028461618365,
+      "grad_norm": 0.39198037981987,
+      "learning_rate": 3.851066661807424e-05,
+      "loss": 1.66,
+      "step": 5524
+    },
+    {
+      "epoch": 0.5739067206814168,
+      "grad_norm": 0.41385430097579956,
+      "learning_rate": 3.849478727350849e-05,
+      "loss": 1.7209,
+      "step": 5525
+    },
+    {
+      "epoch": 0.5740105952009972,
+      "grad_norm": 0.4008021652698517,
+      "learning_rate": 3.847890915415815e-05,
+      "loss": 1.6168,
+      "step": 5526
+    },
+    {
+      "epoch": 0.5741144697205776,
+      "grad_norm": 0.42012637853622437,
+      "learning_rate": 3.8463032261714114e-05,
+      "loss": 1.7073,
+      "step": 5527
+    },
+    {
+      "epoch": 0.5742183442401579,
+      "grad_norm": 0.428591251373291,
+      "learning_rate": 3.8447156597867153e-05,
+      "loss": 1.7536,
+      "step": 5528
+    },
+    {
+      "epoch": 0.5743222187597382,
+      "grad_norm": 0.45808881521224976,
+      "learning_rate": 3.843128216430791e-05,
+      "loss": 1.6032,
+      "step": 5529
+    },
+    {
+      "epoch": 0.5744260932793186,
+      "grad_norm": 0.41718247532844543,
+      "learning_rate": 3.841540896272685e-05,
+      "loss": 1.6722,
+      "step": 5530
+    },
+    {
+      "epoch": 0.574529967798899,
+      "grad_norm": 0.4867742955684662,
+      "learning_rate": 3.83995369948144e-05,
+      "loss": 1.7742,
+      "step": 5531
+    },
+    {
+      "epoch": 0.5746338423184792,
+      "grad_norm": 0.4175654947757721,
+      "learning_rate": 3.838366626226078e-05,
+      "loss": 1.5992,
+      "step": 5532
+    },
+    {
+      "epoch": 0.5747377168380596,
+      "grad_norm": 0.3990960121154785,
+      "learning_rate": 3.836779676675606e-05,
+      "loss": 1.7132,
+      "step": 5533
+    },
+    {
+      "epoch": 0.57484159135764,
+      "grad_norm": 0.43652915954589844,
+      "learning_rate": 3.8351928509990255e-05,
+      "loss": 1.8061,
+      "step": 5534
+    },
+    {
+      "epoch": 0.5749454658772203,
+      "grad_norm": 0.4044228196144104,
+      "learning_rate": 3.833606149365323e-05,
+      "loss": 1.7402,
+      "step": 5535
+    },
+    {
+      "epoch": 0.5750493403968007,
+      "grad_norm": 0.41677069664001465,
+      "learning_rate": 3.832019571943465e-05,
+      "loss": 1.6522,
+      "step": 5536
+    },
+    {
+      "epoch": 0.575153214916381,
+      "grad_norm": 0.40722280740737915,
+      "learning_rate": 3.8304331189024127e-05,
+      "loss": 1.8343,
+      "step": 5537
+    },
+    {
+      "epoch": 0.5752570894359613,
+      "grad_norm": 0.40876513719558716,
+      "learning_rate": 3.828846790411111e-05,
+      "loss": 1.6463,
+      "step": 5538
+    },
+    {
+      "epoch": 0.5753609639555417,
+      "grad_norm": 0.4222007095813751,
+      "learning_rate": 3.82726058663849e-05,
+      "loss": 1.672,
+      "step": 5539
+    },
+    {
+      "epoch": 0.5754648384751221,
+      "grad_norm": 0.4677537679672241,
+      "learning_rate": 3.8256745077534684e-05,
+      "loss": 1.8548,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5755687129947024,
+      "grad_norm": 0.4388614594936371,
+      "learning_rate": 3.824088553924954e-05,
+      "loss": 1.8835,
+      "step": 5541
+    },
+    {
+      "epoch": 0.5756725875142827,
+      "grad_norm": 0.43357276916503906,
+      "learning_rate": 3.822502725321833e-05,
+      "loss": 1.831,
+      "step": 5542
+    },
+    {
+      "epoch": 0.5757764620338631,
+      "grad_norm": 0.40327322483062744,
+      "learning_rate": 3.8209170221129875e-05,
+      "loss": 1.7199,
+      "step": 5543
+    },
+    {
+      "epoch": 0.5758803365534434,
+      "grad_norm": 0.4302258789539337,
+      "learning_rate": 3.819331444467284e-05,
+      "loss": 1.8026,
+      "step": 5544
+    },
+    {
+      "epoch": 0.5759842110730238,
+      "grad_norm": 0.4435931444168091,
+      "learning_rate": 3.817745992553571e-05,
+      "loss": 1.6854,
+      "step": 5545
+    },
+    {
+      "epoch": 0.5760880855926042,
+      "grad_norm": 0.40510040521621704,
+      "learning_rate": 3.816160666540687e-05,
+      "loss": 1.6003,
+      "step": 5546
+    },
+    {
+      "epoch": 0.5761919601121844,
+      "grad_norm": 0.43098416924476624,
+      "learning_rate": 3.814575466597459e-05,
+      "loss": 1.8288,
+      "step": 5547
+    },
+    {
+      "epoch": 0.5762958346317648,
+      "grad_norm": 0.4217774569988251,
+      "learning_rate": 3.8129903928926956e-05,
+      "loss": 1.8433,
+      "step": 5548
+    },
+    {
+      "epoch": 0.5763997091513452,
+      "grad_norm": 0.4399735629558563,
+      "learning_rate": 3.811405445595197e-05,
+      "loss": 1.6735,
+      "step": 5549
+    },
+    {
+      "epoch": 0.5765035836709255,
+      "grad_norm": 0.39600399136543274,
+      "learning_rate": 3.809820624873748e-05,
+      "loss": 1.5637,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5766074581905059,
+      "grad_norm": 0.4116421937942505,
+      "learning_rate": 3.808235930897118e-05,
+      "loss": 1.6252,
+      "step": 5551
+    },
+    {
+      "epoch": 0.5767113327100862,
+      "grad_norm": 0.4399339258670807,
+      "learning_rate": 3.806651363834065e-05,
+      "loss": 1.7843,
+      "step": 5552
+    },
+    {
+      "epoch": 0.5768152072296666,
+      "grad_norm": 0.3872789442539215,
+      "learning_rate": 3.805066923853333e-05,
+      "loss": 1.547,
+      "step": 5553
+    },
+    {
+      "epoch": 0.5769190817492469,
+      "grad_norm": 0.38812029361724854,
+      "learning_rate": 3.803482611123654e-05,
+      "loss": 1.5512,
+      "step": 5554
+    },
+    {
+      "epoch": 0.5770229562688273,
+      "grad_norm": 0.4194970428943634,
+      "learning_rate": 3.8018984258137423e-05,
+      "loss": 1.5616,
+      "step": 5555
+    },
+    {
+      "epoch": 0.5771268307884077,
+      "grad_norm": 0.44213446974754333,
+      "learning_rate": 3.800314368092302e-05,
+      "loss": 1.8645,
+      "step": 5556
+    },
+    {
+      "epoch": 0.5772307053079879,
+      "grad_norm": 0.39991211891174316,
+      "learning_rate": 3.798730438128027e-05,
+      "loss": 1.7879,
+      "step": 5557
+    },
+    {
+      "epoch": 0.5773345798275683,
+      "grad_norm": 0.4132145047187805,
+      "learning_rate": 3.797146636089587e-05,
+      "loss": 1.8618,
+      "step": 5558
+    },
+    {
+      "epoch": 0.5774384543471487,
+      "grad_norm": 0.4226401150226593,
+      "learning_rate": 3.795562962145649e-05,
+      "loss": 1.5837,
+      "step": 5559
+    },
+    {
+      "epoch": 0.577542328866729,
+      "grad_norm": 0.3981073498725891,
+      "learning_rate": 3.793979416464862e-05,
+      "loss": 1.6511,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5776462033863093,
+      "grad_norm": 0.399758517742157,
+      "learning_rate": 3.792395999215857e-05,
+      "loss": 1.5685,
+      "step": 5561
+    },
+    {
+      "epoch": 0.5777500779058897,
+      "grad_norm": 0.4009932279586792,
+      "learning_rate": 3.790812710567259e-05,
+      "loss": 1.6031,
+      "step": 5562
+    },
+    {
+      "epoch": 0.57785395242547,
+      "grad_norm": 0.4045742154121399,
+      "learning_rate": 3.789229550687676e-05,
+      "loss": 1.6291,
+      "step": 5563
+    },
+    {
+      "epoch": 0.5779578269450504,
+      "grad_norm": 0.4436100423336029,
+      "learning_rate": 3.7876465197457014e-05,
+      "loss": 1.7213,
+      "step": 5564
+    },
+    {
+      "epoch": 0.5780617014646308,
+      "grad_norm": 0.41849127411842346,
+      "learning_rate": 3.7860636179099153e-05,
+      "loss": 1.7851,
+      "step": 5565
+    },
+    {
+      "epoch": 0.578165575984211,
+      "grad_norm": 0.44967445731163025,
+      "learning_rate": 3.784480845348886e-05,
+      "loss": 1.7522,
+      "step": 5566
+    },
+    {
+      "epoch": 0.5782694505037914,
+      "grad_norm": 0.4041469693183899,
+      "learning_rate": 3.7828982022311645e-05,
+      "loss": 1.6278,
+      "step": 5567
+    },
+    {
+      "epoch": 0.5783733250233718,
+      "grad_norm": 0.3950141668319702,
+      "learning_rate": 3.7813156887252896e-05,
+      "loss": 1.566,
+      "step": 5568
+    },
+    {
+      "epoch": 0.5784771995429521,
+      "grad_norm": 0.4024663269519806,
+      "learning_rate": 3.77973330499979e-05,
+      "loss": 1.6743,
+      "step": 5569
+    },
+    {
+      "epoch": 0.5785810740625325,
+      "grad_norm": 0.4183747172355652,
+      "learning_rate": 3.778151051223173e-05,
+      "loss": 1.6308,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5786849485821128,
+      "grad_norm": 0.3978317081928253,
+      "learning_rate": 3.7765689275639374e-05,
+      "loss": 1.6298,
+      "step": 5571
+    },
+    {
+      "epoch": 0.5787888231016931,
+      "grad_norm": 0.4176594018936157,
+      "learning_rate": 3.774986934190571e-05,
+      "loss": 1.7289,
+      "step": 5572
+    },
+    {
+      "epoch": 0.5788926976212735,
+      "grad_norm": 0.4060553014278412,
+      "learning_rate": 3.7734050712715394e-05,
+      "loss": 1.7287,
+      "step": 5573
+    },
+    {
+      "epoch": 0.5789965721408539,
+      "grad_norm": 0.40676042437553406,
+      "learning_rate": 3.7718233389752986e-05,
+      "loss": 1.4959,
+      "step": 5574
+    },
+    {
+      "epoch": 0.5791004466604341,
+      "grad_norm": 0.4172111451625824,
+      "learning_rate": 3.770241737470295e-05,
+      "loss": 1.669,
+      "step": 5575
+    },
+    {
+      "epoch": 0.5792043211800145,
+      "grad_norm": 0.4500811994075775,
+      "learning_rate": 3.7686602669249505e-05,
+      "loss": 1.8687,
+      "step": 5576
+    },
+    {
+      "epoch": 0.5793081956995949,
+      "grad_norm": 0.4102492928504944,
+      "learning_rate": 3.767078927507683e-05,
+      "loss": 1.6422,
+      "step": 5577
+    },
+    {
+      "epoch": 0.5794120702191753,
+      "grad_norm": 0.42576107382774353,
+      "learning_rate": 3.765497719386894e-05,
+      "loss": 1.7986,
+      "step": 5578
+    },
+    {
+      "epoch": 0.5795159447387556,
+      "grad_norm": 0.43037348985671997,
+      "learning_rate": 3.763916642730969e-05,
+      "loss": 1.52,
+      "step": 5579
+    },
+    {
+      "epoch": 0.5796198192583359,
+      "grad_norm": 0.4052847623825073,
+      "learning_rate": 3.762335697708279e-05,
+      "loss": 1.8281,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5797236937779163,
+      "grad_norm": 0.42516037821769714,
+      "learning_rate": 3.760754884487182e-05,
+      "loss": 1.7839,
+      "step": 5581
+    },
+    {
+      "epoch": 0.5798275682974966,
+      "grad_norm": 0.4137275815010071,
+      "learning_rate": 3.759174203236025e-05,
+      "loss": 1.643,
+      "step": 5582
+    },
+    {
+      "epoch": 0.579931442817077,
+      "grad_norm": 0.45627397298812866,
+      "learning_rate": 3.757593654123135e-05,
+      "loss": 1.8582,
+      "step": 5583
+    },
+    {
+      "epoch": 0.5800353173366574,
+      "grad_norm": 0.420109361410141,
+      "learning_rate": 3.75601323731683e-05,
+      "loss": 1.7523,
+      "step": 5584
+    },
+    {
+      "epoch": 0.5801391918562376,
+      "grad_norm": 0.4100499749183655,
+      "learning_rate": 3.754432952985413e-05,
+      "loss": 1.6175,
+      "step": 5585
+    },
+    {
+      "epoch": 0.580243066375818,
+      "grad_norm": 0.4295746088027954,
+      "learning_rate": 3.7528528012971694e-05,
+      "loss": 1.7074,
+      "step": 5586
+    },
+    {
+      "epoch": 0.5803469408953984,
+      "grad_norm": 0.40844690799713135,
+      "learning_rate": 3.7512727824203754e-05,
+      "loss": 1.6747,
+      "step": 5587
+    },
+    {
+      "epoch": 0.5804508154149787,
+      "grad_norm": 0.45334941148757935,
+      "learning_rate": 3.7496928965232905e-05,
+      "loss": 1.6918,
+      "step": 5588
+    },
+    {
+      "epoch": 0.580554689934559,
+      "grad_norm": 0.35644862055778503,
+      "learning_rate": 3.7481131437741566e-05,
+      "loss": 1.5046,
+      "step": 5589
+    },
+    {
+      "epoch": 0.5806585644541394,
+      "grad_norm": 0.41652175784111023,
+      "learning_rate": 3.746533524341208e-05,
+      "loss": 1.4811,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5807624389737197,
+      "grad_norm": 0.40654313564300537,
+      "learning_rate": 3.7449540383926646e-05,
+      "loss": 1.6301,
+      "step": 5591
+    },
+    {
+      "epoch": 0.5808663134933001,
+      "grad_norm": 0.40000343322753906,
+      "learning_rate": 3.7433746860967235e-05,
+      "loss": 1.6732,
+      "step": 5592
+    },
+    {
+      "epoch": 0.5809701880128805,
+      "grad_norm": 0.4352266788482666,
+      "learning_rate": 3.741795467621578e-05,
+      "loss": 1.4708,
+      "step": 5593
+    },
+    {
+      "epoch": 0.5810740625324607,
+      "grad_norm": 0.40756914019584656,
+      "learning_rate": 3.7402163831354e-05,
+      "loss": 1.647,
+      "step": 5594
+    },
+    {
+      "epoch": 0.5811779370520411,
+      "grad_norm": 0.4151524603366852,
+      "learning_rate": 3.738637432806351e-05,
+      "loss": 1.5459,
+      "step": 5595
+    },
+    {
+      "epoch": 0.5812818115716215,
+      "grad_norm": 0.39485087990760803,
+      "learning_rate": 3.737058616802575e-05,
+      "loss": 1.6415,
+      "step": 5596
+    },
+    {
+      "epoch": 0.5813856860912018,
+      "grad_norm": 0.4184524118900299,
+      "learning_rate": 3.735479935292208e-05,
+      "loss": 1.4745,
+      "step": 5597
+    },
+    {
+      "epoch": 0.5814895606107822,
+      "grad_norm": 0.42042723298072815,
+      "learning_rate": 3.733901388443361e-05,
+      "loss": 1.6336,
+      "step": 5598
+    },
+    {
+      "epoch": 0.5815934351303625,
+      "grad_norm": 0.39631494879722595,
+      "learning_rate": 3.732322976424141e-05,
+      "loss": 1.4979,
+      "step": 5599
+    },
+    {
+      "epoch": 0.5816973096499428,
+      "grad_norm": 0.45113274455070496,
+      "learning_rate": 3.7307446994026375e-05,
+      "loss": 1.826,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5818011841695232,
+      "grad_norm": 0.43801307678222656,
+      "learning_rate": 3.7291665575469216e-05,
+      "loss": 2.0605,
+      "step": 5601
+    },
+    {
+      "epoch": 0.5819050586891036,
+      "grad_norm": 0.4118470549583435,
+      "learning_rate": 3.727588551025053e-05,
+      "loss": 1.6777,
+      "step": 5602
+    },
+    {
+      "epoch": 0.582008933208684,
+      "grad_norm": 0.40802454948425293,
+      "learning_rate": 3.72601068000508e-05,
+      "loss": 1.6875,
+      "step": 5603
+    },
+    {
+      "epoch": 0.5821128077282642,
+      "grad_norm": 0.3772849142551422,
+      "learning_rate": 3.724432944655033e-05,
+      "loss": 1.7246,
+      "step": 5604
+    },
+    {
+      "epoch": 0.5822166822478446,
+      "grad_norm": 0.42528003454208374,
+      "learning_rate": 3.722855345142926e-05,
+      "loss": 1.7337,
+      "step": 5605
+    },
+    {
+      "epoch": 0.582320556767425,
+      "grad_norm": 0.4423576593399048,
+      "learning_rate": 3.721277881636764e-05,
+      "loss": 1.8097,
+      "step": 5606
+    },
+    {
+      "epoch": 0.5824244312870053,
+      "grad_norm": 0.4138087034225464,
+      "learning_rate": 3.719700554304533e-05,
+      "loss": 1.5108,
+      "step": 5607
+    },
+    {
+      "epoch": 0.5825283058065857,
+      "grad_norm": 0.4109199047088623,
+      "learning_rate": 3.718123363314206e-05,
+      "loss": 1.541,
+      "step": 5608
+    },
+    {
+      "epoch": 0.582632180326166,
+      "grad_norm": 0.38805916905403137,
+      "learning_rate": 3.7165463088337406e-05,
+      "loss": 1.7062,
+      "step": 5609
+    },
+    {
+      "epoch": 0.5827360548457463,
+      "grad_norm": 0.39964035153388977,
+      "learning_rate": 3.714969391031084e-05,
+      "loss": 1.6754,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5828399293653267,
+      "grad_norm": 0.41137054562568665,
+      "learning_rate": 3.713392610074162e-05,
+      "loss": 1.6984,
+      "step": 5611
+    },
+    {
+      "epoch": 0.5829438038849071,
+      "grad_norm": 0.3872339129447937,
+      "learning_rate": 3.7118159661308906e-05,
+      "loss": 1.4976,
+      "step": 5612
+    },
+    {
+      "epoch": 0.5830476784044873,
+      "grad_norm": 0.4146081507205963,
+      "learning_rate": 3.7102394593691724e-05,
+      "loss": 1.5536,
+      "step": 5613
+    },
+    {
+      "epoch": 0.5831515529240677,
+      "grad_norm": 0.4401085674762726,
+      "learning_rate": 3.7086630899568906e-05,
+      "loss": 1.8147,
+      "step": 5614
+    },
+    {
+      "epoch": 0.5832554274436481,
+      "grad_norm": 0.46493473649024963,
+      "learning_rate": 3.707086858061917e-05,
+      "loss": 1.934,
+      "step": 5615
+    },
+    {
+      "epoch": 0.5833593019632284,
+      "grad_norm": 0.4143672585487366,
+      "learning_rate": 3.7055107638521086e-05,
+      "loss": 1.6462,
+      "step": 5616
+    },
+    {
+      "epoch": 0.5834631764828088,
+      "grad_norm": 0.41151854395866394,
+      "learning_rate": 3.703934807495305e-05,
+      "loss": 1.6508,
+      "step": 5617
+    },
+    {
+      "epoch": 0.5835670510023891,
+      "grad_norm": 0.47356948256492615,
+      "learning_rate": 3.7023589891593344e-05,
+      "loss": 1.9211,
+      "step": 5618
+    },
+    {
+      "epoch": 0.5836709255219694,
+      "grad_norm": 0.4014083445072174,
+      "learning_rate": 3.7007833090120116e-05,
+      "loss": 1.7667,
+      "step": 5619
+    },
+    {
+      "epoch": 0.5837748000415498,
+      "grad_norm": 0.3790227174758911,
+      "learning_rate": 3.69920776722113e-05,
+      "loss": 1.4863,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5838786745611302,
+      "grad_norm": 0.43195411562919617,
+      "learning_rate": 3.697632363954475e-05,
+      "loss": 1.7431,
+      "step": 5621
+    },
+    {
+      "epoch": 0.5839825490807105,
+      "grad_norm": 0.4108622372150421,
+      "learning_rate": 3.696057099379816e-05,
+      "loss": 1.6647,
+      "step": 5622
+    },
+    {
+      "epoch": 0.5840864236002908,
+      "grad_norm": 0.3950975835323334,
+      "learning_rate": 3.694481973664904e-05,
+      "loss": 1.4736,
+      "step": 5623
+    },
+    {
+      "epoch": 0.5841902981198712,
+      "grad_norm": 0.41740620136260986,
+      "learning_rate": 3.6929069869774766e-05,
+      "loss": 1.6726,
+      "step": 5624
+    },
+    {
+      "epoch": 0.5842941726394515,
+      "grad_norm": 0.45372968912124634,
+      "learning_rate": 3.6913321394852616e-05,
+      "loss": 1.7569,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5843980471590319,
+      "grad_norm": 0.3958111107349396,
+      "learning_rate": 3.689757431355964e-05,
+      "loss": 1.7056,
+      "step": 5626
+    },
+    {
+      "epoch": 0.5845019216786123,
+      "grad_norm": 0.4113248288631439,
+      "learning_rate": 3.68818286275728e-05,
+      "loss": 1.7895,
+      "step": 5627
+    },
+    {
+      "epoch": 0.5846057961981925,
+      "grad_norm": 0.4527917802333832,
+      "learning_rate": 3.686608433856891e-05,
+      "loss": 1.7048,
+      "step": 5628
+    },
+    {
+      "epoch": 0.5847096707177729,
+      "grad_norm": 0.43237176537513733,
+      "learning_rate": 3.6850341448224574e-05,
+      "loss": 1.6414,
+      "step": 5629
+    },
+    {
+      "epoch": 0.5848135452373533,
+      "grad_norm": 0.4343053102493286,
+      "learning_rate": 3.6834599958216286e-05,
+      "loss": 1.7855,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5849174197569337,
+      "grad_norm": 0.44802016019821167,
+      "learning_rate": 3.6818859870220415e-05,
+      "loss": 1.8118,
+      "step": 5631
+    },
+    {
+      "epoch": 0.585021294276514,
+      "grad_norm": 0.4242957532405853,
+      "learning_rate": 3.680312118591317e-05,
+      "loss": 1.6769,
+      "step": 5632
+    },
+    {
+      "epoch": 0.5851251687960943,
+      "grad_norm": 0.4112863540649414,
+      "learning_rate": 3.6787383906970564e-05,
+      "loss": 1.6765,
+      "step": 5633
+    },
+    {
+      "epoch": 0.5852290433156747,
+      "grad_norm": 0.435183048248291,
+      "learning_rate": 3.677164803506851e-05,
+      "loss": 1.6958,
+      "step": 5634
+    },
+    {
+      "epoch": 0.585332917835255,
+      "grad_norm": 0.4290895462036133,
+      "learning_rate": 3.675591357188276e-05,
+      "loss": 1.6542,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5854367923548354,
+      "grad_norm": 0.44175586104393005,
+      "learning_rate": 3.6740180519088905e-05,
+      "loss": 1.802,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5855406668744157,
+      "grad_norm": 0.3649253249168396,
+      "learning_rate": 3.672444887836239e-05,
+      "loss": 1.4738,
+      "step": 5637
+    },
+    {
+      "epoch": 0.585644541393996,
+      "grad_norm": 0.4127234220504761,
+      "learning_rate": 3.6708718651378534e-05,
+      "loss": 1.522,
+      "step": 5638
+    },
+    {
+      "epoch": 0.5857484159135764,
+      "grad_norm": 0.41260048747062683,
+      "learning_rate": 3.6692989839812444e-05,
+      "loss": 1.6182,
+      "step": 5639
+    },
+    {
+      "epoch": 0.5858522904331568,
+      "grad_norm": 0.410636305809021,
+      "learning_rate": 3.667726244533913e-05,
+      "loss": 1.6908,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5859561649527371,
+      "grad_norm": 0.393493115901947,
+      "learning_rate": 3.6661536469633474e-05,
+      "loss": 1.5382,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5860600394723174,
+      "grad_norm": 0.4044821560382843,
+      "learning_rate": 3.6645811914370126e-05,
+      "loss": 1.683,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5861639139918978,
+      "grad_norm": 0.39565759897232056,
+      "learning_rate": 3.663008878122364e-05,
+      "loss": 1.6154,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5862677885114781,
+      "grad_norm": 0.4338008165359497,
+      "learning_rate": 3.661436707186842e-05,
+      "loss": 1.8497,
+      "step": 5644
+    },
+    {
+      "epoch": 0.5863716630310585,
+      "grad_norm": 0.4520065188407898,
+      "learning_rate": 3.6598646787978676e-05,
+      "loss": 1.6919,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5864755375506389,
+      "grad_norm": 0.3839634656906128,
+      "learning_rate": 3.6582927931228516e-05,
+      "loss": 1.6516,
+      "step": 5646
+    },
+    {
+      "epoch": 0.5865794120702191,
+      "grad_norm": 0.4146488606929779,
+      "learning_rate": 3.656721050329189e-05,
+      "loss": 1.7285,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5866832865897995,
+      "grad_norm": 0.43894267082214355,
+      "learning_rate": 3.6551494505842554e-05,
+      "loss": 1.6571,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5867871611093799,
+      "grad_norm": 0.42978256940841675,
+      "learning_rate": 3.653577994055416e-05,
+      "loss": 1.6641,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5868910356289602,
+      "grad_norm": 0.40996453166007996,
+      "learning_rate": 3.652006680910017e-05,
+      "loss": 1.5068,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5869949101485405,
+      "grad_norm": 0.4839547574520111,
+      "learning_rate": 3.650435511315392e-05,
+      "loss": 1.7859,
+      "step": 5651
+    },
+    {
+      "epoch": 0.5870987846681209,
+      "grad_norm": 0.40157848596572876,
+      "learning_rate": 3.6488644854388577e-05,
+      "loss": 1.5551,
+      "step": 5652
+    },
+    {
+      "epoch": 0.5872026591877012,
+      "grad_norm": 0.46978557109832764,
+      "learning_rate": 3.647293603447718e-05,
+      "loss": 1.8654,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5873065337072816,
+      "grad_norm": 0.39479735493659973,
+      "learning_rate": 3.6457228655092554e-05,
+      "loss": 1.6824,
+      "step": 5654
+    },
+    {
+      "epoch": 0.587410408226862,
+      "grad_norm": 0.4402686357498169,
+      "learning_rate": 3.6441522717907443e-05,
+      "loss": 1.7669,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5875142827464424,
+      "grad_norm": 0.41756048798561096,
+      "learning_rate": 3.642581822459441e-05,
+      "loss": 1.7688,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5876181572660226,
+      "grad_norm": 0.42584848403930664,
+      "learning_rate": 3.641011517682588e-05,
+      "loss": 1.6345,
+      "step": 5657
+    },
+    {
+      "epoch": 0.587722031785603,
+      "grad_norm": 0.40674230456352234,
+      "learning_rate": 3.639441357627404e-05,
+      "loss": 1.6559,
+      "step": 5658
+    },
+    {
+      "epoch": 0.5878259063051834,
+      "grad_norm": 0.40528520941734314,
+      "learning_rate": 3.637871342461103e-05,
+      "loss": 1.6813,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5879297808247637,
+      "grad_norm": 0.4171191453933716,
+      "learning_rate": 3.6363014723508813e-05,
+      "loss": 1.7832,
+      "step": 5660
+    },
+    {
+      "epoch": 0.588033655344344,
+      "grad_norm": 0.3836202919483185,
+      "learning_rate": 3.634731747463913e-05,
+      "loss": 1.6288,
+      "step": 5661
+    },
+    {
+      "epoch": 0.5881375298639244,
+      "grad_norm": 0.45273473858833313,
+      "learning_rate": 3.633162167967366e-05,
+      "loss": 1.6388,
+      "step": 5662
+    },
+    {
+      "epoch": 0.5882414043835047,
+      "grad_norm": 0.42023298144340515,
+      "learning_rate": 3.6315927340283855e-05,
+      "loss": 1.6475,
+      "step": 5663
+    },
+    {
+      "epoch": 0.5883452789030851,
+      "grad_norm": 0.4072648286819458,
+      "learning_rate": 3.630023445814106e-05,
+      "loss": 1.7276,
+      "step": 5664
+    },
+    {
+      "epoch": 0.5884491534226655,
+      "grad_norm": 0.37127307057380676,
+      "learning_rate": 3.6284543034916405e-05,
+      "loss": 1.449,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5885530279422457,
+      "grad_norm": 0.40578770637512207,
+      "learning_rate": 3.6268853072280966e-05,
+      "loss": 1.692,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5886569024618261,
+      "grad_norm": 0.4521169364452362,
+      "learning_rate": 3.625316457190555e-05,
+      "loss": 1.7896,
+      "step": 5667
+    },
+    {
+      "epoch": 0.5887607769814065,
+      "grad_norm": 0.4111865758895874,
+      "learning_rate": 3.623747753546086e-05,
+      "loss": 1.8176,
+      "step": 5668
+    },
+    {
+      "epoch": 0.5888646515009868,
+      "grad_norm": 0.4034864008426666,
+      "learning_rate": 3.62217919646175e-05,
+      "loss": 1.7778,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5889685260205672,
+      "grad_norm": 0.405782550573349,
+      "learning_rate": 3.6206107861045804e-05,
+      "loss": 1.8,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5890724005401475,
+      "grad_norm": 0.3821369707584381,
+      "learning_rate": 3.619042522641601e-05,
+      "loss": 1.4904,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5891762750597278,
+      "grad_norm": 0.42174914479255676,
+      "learning_rate": 3.617474406239824e-05,
+      "loss": 1.6863,
+      "step": 5672
+    },
+    {
+      "epoch": 0.5892801495793082,
+      "grad_norm": 0.4899470806121826,
+      "learning_rate": 3.615906437066236e-05,
+      "loss": 1.5951,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5893840240988886,
+      "grad_norm": 0.3886786699295044,
+      "learning_rate": 3.614338615287816e-05,
+      "loss": 1.5416,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5894878986184688,
+      "grad_norm": 0.3970104455947876,
+      "learning_rate": 3.612770941071527e-05,
+      "loss": 1.6345,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5895917731380492,
+      "grad_norm": 0.4441760182380676,
+      "learning_rate": 3.61120341458431e-05,
+      "loss": 1.7968,
+      "step": 5676
+    },
+    {
+      "epoch": 0.5896956476576296,
+      "grad_norm": 0.4187043607234955,
+      "learning_rate": 3.6096360359930984e-05,
+      "loss": 1.458,
+      "step": 5677
+    },
+    {
+      "epoch": 0.5897995221772099,
+      "grad_norm": 0.39509010314941406,
+      "learning_rate": 3.608068805464802e-05,
+      "loss": 1.8428,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5899033966967903,
+      "grad_norm": 0.41668543219566345,
+      "learning_rate": 3.6065017231663214e-05,
+      "loss": 1.4289,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5900072712163706,
+      "grad_norm": 0.4090832471847534,
+      "learning_rate": 3.6049347892645356e-05,
+      "loss": 1.7567,
+      "step": 5680
+    },
+    {
+      "epoch": 0.590111145735951,
+      "grad_norm": 0.4536181390285492,
+      "learning_rate": 3.603368003926314e-05,
+      "loss": 1.6062,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5902150202555313,
+      "grad_norm": 0.4117638170719147,
+      "learning_rate": 3.601801367318507e-05,
+      "loss": 1.6996,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5903188947751117,
+      "grad_norm": 0.4338860809803009,
+      "learning_rate": 3.600234879607947e-05,
+      "loss": 1.6467,
+      "step": 5683
+    },
+    {
+      "epoch": 0.5904227692946921,
+      "grad_norm": 0.4608552157878876,
+      "learning_rate": 3.5986685409614526e-05,
+      "loss": 1.8217,
+      "step": 5684
+    },
+    {
+      "epoch": 0.5905266438142723,
+      "grad_norm": 0.4324304163455963,
+      "learning_rate": 3.597102351545831e-05,
+      "loss": 1.7504,
+      "step": 5685
+    },
+    {
+      "epoch": 0.5906305183338527,
+      "grad_norm": 0.3990081250667572,
+      "learning_rate": 3.5955363115278626e-05,
+      "loss": 1.6519,
+      "step": 5686
+    },
+    {
+      "epoch": 0.5907343928534331,
+      "grad_norm": 0.4149520695209503,
+      "learning_rate": 3.593970421074323e-05,
+      "loss": 1.6564,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5908382673730134,
+      "grad_norm": 0.4176534116268158,
+      "learning_rate": 3.5924046803519675e-05,
+      "loss": 1.5858,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5909421418925938,
+      "grad_norm": 0.40572646260261536,
+      "learning_rate": 3.590839089527531e-05,
+      "loss": 1.6495,
+      "step": 5689
+    },
+    {
+      "epoch": 0.5910460164121741,
+      "grad_norm": 0.38206931948661804,
+      "learning_rate": 3.589273648767742e-05,
+      "loss": 1.6129,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5911498909317544,
+      "grad_norm": 0.4322497248649597,
+      "learning_rate": 3.5877083582393044e-05,
+      "loss": 1.7354,
+      "step": 5691
+    },
+    {
+      "epoch": 0.5912537654513348,
+      "grad_norm": 0.41732117533683777,
+      "learning_rate": 3.586143218108911e-05,
+      "loss": 1.891,
+      "step": 5692
+    },
+    {
+      "epoch": 0.5913576399709152,
+      "grad_norm": 0.3857957422733307,
+      "learning_rate": 3.584578228543235e-05,
+      "loss": 1.6225,
+      "step": 5693
+    },
+    {
+      "epoch": 0.5914615144904954,
+      "grad_norm": 0.408961683511734,
+      "learning_rate": 3.5830133897089394e-05,
+      "loss": 1.671,
+      "step": 5694
+    },
+    {
+      "epoch": 0.5915653890100758,
+      "grad_norm": 0.4022825062274933,
+      "learning_rate": 3.5814487017726616e-05,
+      "loss": 1.648,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5916692635296562,
+      "grad_norm": 0.468014657497406,
+      "learning_rate": 3.579884164901033e-05,
+      "loss": 1.8003,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5917731380492365,
+      "grad_norm": 0.3964448869228363,
+      "learning_rate": 3.5783197792606654e-05,
+      "loss": 1.7439,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5918770125688169,
+      "grad_norm": 0.39556026458740234,
+      "learning_rate": 3.576755545018151e-05,
+      "loss": 1.6194,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5919808870883972,
+      "grad_norm": 0.3882671594619751,
+      "learning_rate": 3.5751914623400675e-05,
+      "loss": 1.599,
+      "step": 5699
+    },
+    {
+      "epoch": 0.5920847616079775,
+      "grad_norm": 0.4119251072406769,
+      "learning_rate": 3.573627531392982e-05,
+      "loss": 1.8195,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5921886361275579,
+      "grad_norm": 0.4111797511577606,
+      "learning_rate": 3.572063752343436e-05,
+      "loss": 1.6844,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5922925106471383,
+      "grad_norm": 0.3847576379776001,
+      "learning_rate": 3.5705001253579615e-05,
+      "loss": 1.652,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5923963851667186,
+      "grad_norm": 0.44039520621299744,
+      "learning_rate": 3.568936650603076e-05,
+      "loss": 1.6053,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5925002596862989,
+      "grad_norm": 0.4001440703868866,
+      "learning_rate": 3.5673733282452725e-05,
+      "loss": 1.3555,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5926041342058793,
+      "grad_norm": 0.4470379054546356,
+      "learning_rate": 3.565810158451036e-05,
+      "loss": 1.7809,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5927080087254596,
+      "grad_norm": 0.46440574526786804,
+      "learning_rate": 3.5642471413868305e-05,
+      "loss": 1.903,
+      "step": 5706
+    },
+    {
+      "epoch": 0.59281188324504,
+      "grad_norm": 0.4026218056678772,
+      "learning_rate": 3.562684277219104e-05,
+      "loss": 1.5994,
+      "step": 5707
+    },
+    {
+      "epoch": 0.5929157577646204,
+      "grad_norm": 0.47202712297439575,
+      "learning_rate": 3.5611215661142897e-05,
+      "loss": 1.8916,
+      "step": 5708
+    },
+    {
+      "epoch": 0.5930196322842007,
+      "grad_norm": 0.4305804967880249,
+      "learning_rate": 3.559559008238805e-05,
+      "loss": 1.4593,
+      "step": 5709
+    },
+    {
+      "epoch": 0.593123506803781,
+      "grad_norm": 0.4332740306854248,
+      "learning_rate": 3.5579966037590524e-05,
+      "loss": 1.7235,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5932273813233614,
+      "grad_norm": 0.40714386105537415,
+      "learning_rate": 3.556434352841411e-05,
+      "loss": 1.631,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5933312558429418,
+      "grad_norm": 0.4484121799468994,
+      "learning_rate": 3.554872255652252e-05,
+      "loss": 1.8386,
+      "step": 5712
+    },
+    {
+      "epoch": 0.593435130362522,
+      "grad_norm": 0.446119099855423,
+      "learning_rate": 3.553310312357927e-05,
+      "loss": 1.838,
+      "step": 5713
+    },
+    {
+      "epoch": 0.5935390048821024,
+      "grad_norm": 0.45277395844459534,
+      "learning_rate": 3.551748523124766e-05,
+      "loss": 1.7349,
+      "step": 5714
+    },
+    {
+      "epoch": 0.5936428794016828,
+      "grad_norm": 0.3999378979206085,
+      "learning_rate": 3.550186888119091e-05,
+      "loss": 1.6032,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5937467539212631,
+      "grad_norm": 0.4385007619857788,
+      "learning_rate": 3.5486254075072056e-05,
+      "loss": 1.8558,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5938506284408435,
+      "grad_norm": 0.42009690403938293,
+      "learning_rate": 3.547064081455391e-05,
+      "loss": 1.7783,
+      "step": 5717
+    },
+    {
+      "epoch": 0.5939545029604238,
+      "grad_norm": 0.3846227526664734,
+      "learning_rate": 3.5455029101299196e-05,
+      "loss": 1.4527,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5940583774800041,
+      "grad_norm": 0.4681823253631592,
+      "learning_rate": 3.5439418936970425e-05,
+      "loss": 1.8712,
+      "step": 5719
+    },
+    {
+      "epoch": 0.5941622519995845,
+      "grad_norm": 0.4000855088233948,
+      "learning_rate": 3.542381032322997e-05,
+      "loss": 1.548,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5942661265191649,
+      "grad_norm": 0.3869953453540802,
+      "learning_rate": 3.540820326173999e-05,
+      "loss": 1.5979,
+      "step": 5721
+    },
+    {
+      "epoch": 0.5943700010387452,
+      "grad_norm": 0.43959927558898926,
+      "learning_rate": 3.539259775416257e-05,
+      "loss": 1.7001,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5944738755583255,
+      "grad_norm": 0.40868815779685974,
+      "learning_rate": 3.537699380215953e-05,
+      "loss": 1.7146,
+      "step": 5723
+    },
+    {
+      "epoch": 0.5945777500779059,
+      "grad_norm": 0.39394938945770264,
+      "learning_rate": 3.5361391407392583e-05,
+      "loss": 1.711,
+      "step": 5724
+    },
+    {
+      "epoch": 0.5946816245974862,
+      "grad_norm": 0.4582231044769287,
+      "learning_rate": 3.5345790571523286e-05,
+      "loss": 1.9054,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5947854991170666,
+      "grad_norm": 0.40016230940818787,
+      "learning_rate": 3.5330191296212976e-05,
+      "loss": 1.7314,
+      "step": 5726
+    },
+    {
+      "epoch": 0.594889373636647,
+      "grad_norm": 0.45281219482421875,
+      "learning_rate": 3.5314593583122854e-05,
+      "loss": 1.8133,
+      "step": 5727
+    },
+    {
+      "epoch": 0.5949932481562272,
+      "grad_norm": 0.41264450550079346,
+      "learning_rate": 3.529899743391398e-05,
+      "loss": 1.7283,
+      "step": 5728
+    },
+    {
+      "epoch": 0.5950971226758076,
+      "grad_norm": 0.41202911734580994,
+      "learning_rate": 3.5283402850247194e-05,
+      "loss": 1.3802,
+      "step": 5729
+    },
+    {
+      "epoch": 0.595200997195388,
+      "grad_norm": 0.4194154143333435,
+      "learning_rate": 3.526780983378321e-05,
+      "loss": 1.6485,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5953048717149683,
+      "grad_norm": 0.4135974943637848,
+      "learning_rate": 3.5252218386182585e-05,
+      "loss": 1.793,
+      "step": 5731
+    },
+    {
+      "epoch": 0.5954087462345486,
+      "grad_norm": 0.41171103715896606,
+      "learning_rate": 3.523662850910564e-05,
+      "loss": 1.7508,
+      "step": 5732
+    },
+    {
+      "epoch": 0.595512620754129,
+      "grad_norm": 0.4255639314651489,
+      "learning_rate": 3.5221040204212606e-05,
+      "loss": 1.6849,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5956164952737094,
+      "grad_norm": 0.4412103295326233,
+      "learning_rate": 3.5205453473163506e-05,
+      "loss": 1.6842,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5957203697932897,
+      "grad_norm": 0.397868275642395,
+      "learning_rate": 3.518986831761824e-05,
+      "loss": 1.6491,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5958242443128701,
+      "grad_norm": 0.4192911684513092,
+      "learning_rate": 3.517428473923644e-05,
+      "loss": 1.8276,
+      "step": 5736
+    },
+    {
+      "epoch": 0.5959281188324504,
+      "grad_norm": 0.39529258012771606,
+      "learning_rate": 3.515870273967767e-05,
+      "loss": 1.6447,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5960319933520307,
+      "grad_norm": 0.39592933654785156,
+      "learning_rate": 3.514312232060132e-05,
+      "loss": 1.6278,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5961358678716111,
+      "grad_norm": 0.4018373489379883,
+      "learning_rate": 3.512754348366652e-05,
+      "loss": 1.7605,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5962397423911915,
+      "grad_norm": 0.47401878237724304,
+      "learning_rate": 3.5111966230532355e-05,
+      "loss": 2.0161,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5963436169107718,
+      "grad_norm": 0.4021179676055908,
+      "learning_rate": 3.509639056285766e-05,
+      "loss": 1.6192,
+      "step": 5741
+    },
+    {
+      "epoch": 0.5964474914303521,
+      "grad_norm": 0.427116334438324,
+      "learning_rate": 3.50808164823011e-05,
+      "loss": 1.6114,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5965513659499325,
+      "grad_norm": 0.41006430983543396,
+      "learning_rate": 3.5065243990521205e-05,
+      "loss": 1.7771,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5966552404695128,
+      "grad_norm": 0.39242881536483765,
+      "learning_rate": 3.504967308917635e-05,
+      "loss": 1.6355,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5967591149890932,
+      "grad_norm": 0.39691224694252014,
+      "learning_rate": 3.5034103779924674e-05,
+      "loss": 1.5866,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5968629895086736,
+      "grad_norm": 0.4471852481365204,
+      "learning_rate": 3.5018536064424224e-05,
+      "loss": 1.663,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5969668640282538,
+      "grad_norm": 0.39760124683380127,
+      "learning_rate": 3.500296994433282e-05,
+      "loss": 1.7701,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5970707385478342,
+      "grad_norm": 0.3909447193145752,
+      "learning_rate": 3.498740542130813e-05,
+      "loss": 1.5167,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5971746130674146,
+      "grad_norm": 0.4225844442844391,
+      "learning_rate": 3.497184249700766e-05,
+      "loss": 1.6903,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5972784875869949,
+      "grad_norm": 0.4088863432407379,
+      "learning_rate": 3.495628117308875e-05,
+      "loss": 1.3562,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5973823621065752,
+      "grad_norm": 0.39334914088249207,
+      "learning_rate": 3.4940721451208535e-05,
+      "loss": 1.6846,
+      "step": 5751
+    },
+    {
+      "epoch": 0.5974862366261556,
+      "grad_norm": 0.4623134434223175,
+      "learning_rate": 3.4925163333024015e-05,
+      "loss": 1.6667,
+      "step": 5752
+    },
+    {
+      "epoch": 0.5975901111457359,
+      "grad_norm": 0.4101615846157074,
+      "learning_rate": 3.490960682019203e-05,
+      "loss": 1.594,
+      "step": 5753
+    },
+    {
+      "epoch": 0.5976939856653163,
+      "grad_norm": 0.4440319240093231,
+      "learning_rate": 3.489405191436921e-05,
+      "loss": 1.8576,
+      "step": 5754
+    },
+    {
+      "epoch": 0.5977978601848967,
+      "grad_norm": 0.45121869444847107,
+      "learning_rate": 3.487849861721201e-05,
+      "loss": 1.7881,
+      "step": 5755
+    },
+    {
+      "epoch": 0.5979017347044769,
+      "grad_norm": 0.4189302623271942,
+      "learning_rate": 3.4862946930376775e-05,
+      "loss": 1.5814,
+      "step": 5756
+    },
+    {
+      "epoch": 0.5980056092240573,
+      "grad_norm": 0.45438921451568604,
+      "learning_rate": 3.4847396855519604e-05,
+      "loss": 1.688,
+      "step": 5757
+    },
+    {
+      "epoch": 0.5981094837436377,
+      "grad_norm": 0.40121012926101685,
+      "learning_rate": 3.483184839429647e-05,
+      "loss": 1.7056,
+      "step": 5758
+    },
+    {
+      "epoch": 0.5982133582632181,
+      "grad_norm": 0.4294279217720032,
+      "learning_rate": 3.481630154836317e-05,
+      "loss": 1.6278,
+      "step": 5759
+    },
+    {
+      "epoch": 0.5983172327827984,
+      "grad_norm": 0.3801029324531555,
+      "learning_rate": 3.4800756319375325e-05,
+      "loss": 1.5913,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5984211073023787,
+      "grad_norm": 0.43568864464759827,
+      "learning_rate": 3.478521270898838e-05,
+      "loss": 1.7894,
+      "step": 5761
+    },
+    {
+      "epoch": 0.5985249818219591,
+      "grad_norm": 0.4160782992839813,
+      "learning_rate": 3.476967071885758e-05,
+      "loss": 1.6332,
+      "step": 5762
+    },
+    {
+      "epoch": 0.5986288563415394,
+      "grad_norm": 0.4249245524406433,
+      "learning_rate": 3.475413035063807e-05,
+      "loss": 1.6174,
+      "step": 5763
+    },
+    {
+      "epoch": 0.5987327308611198,
+      "grad_norm": 0.38329437375068665,
+      "learning_rate": 3.473859160598474e-05,
+      "loss": 1.7204,
+      "step": 5764
+    },
+    {
+      "epoch": 0.5988366053807002,
+      "grad_norm": 0.4142932593822479,
+      "learning_rate": 3.4723054486552354e-05,
+      "loss": 1.7932,
+      "step": 5765
+    },
+    {
+      "epoch": 0.5989404799002804,
+      "grad_norm": 0.44899502396583557,
+      "learning_rate": 3.4707518993995524e-05,
+      "loss": 1.7966,
+      "step": 5766
+    },
+    {
+      "epoch": 0.5990443544198608,
+      "grad_norm": 0.40485885739326477,
+      "learning_rate": 3.469198512996863e-05,
+      "loss": 1.6783,
+      "step": 5767
+    },
+    {
+      "epoch": 0.5991482289394412,
+      "grad_norm": 0.39931273460388184,
+      "learning_rate": 3.467645289612589e-05,
+      "loss": 1.5305,
+      "step": 5768
+    },
+    {
+      "epoch": 0.5992521034590215,
+      "grad_norm": 0.381760835647583,
+      "learning_rate": 3.4660922294121416e-05,
+      "loss": 1.4935,
+      "step": 5769
+    },
+    {
+      "epoch": 0.5993559779786019,
+      "grad_norm": 0.4411904215812683,
+      "learning_rate": 3.4645393325609046e-05,
+      "loss": 1.5182,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5994598524981822,
+      "grad_norm": 0.39932867884635925,
+      "learning_rate": 3.462986599224252e-05,
+      "loss": 1.7623,
+      "step": 5771
+    },
+    {
+      "epoch": 0.5995637270177625,
+      "grad_norm": 0.40389779210090637,
+      "learning_rate": 3.46143402956754e-05,
+      "loss": 1.5783,
+      "step": 5772
+    },
+    {
+      "epoch": 0.5996676015373429,
+      "grad_norm": 0.41848430037498474,
+      "learning_rate": 3.4598816237561e-05,
+      "loss": 1.7221,
+      "step": 5773
+    },
+    {
+      "epoch": 0.5997714760569233,
+      "grad_norm": 0.4139845669269562,
+      "learning_rate": 3.4583293819552534e-05,
+      "loss": 1.5198,
+      "step": 5774
+    },
+    {
+      "epoch": 0.5998753505765035,
+      "grad_norm": 0.4001912474632263,
+      "learning_rate": 3.456777304330303e-05,
+      "loss": 1.6871,
+      "step": 5775
+    },
+    {
+      "epoch": 0.5999792250960839,
+      "grad_norm": 0.44707363843917847,
+      "learning_rate": 3.455225391046533e-05,
+      "loss": 1.6126,
+      "step": 5776
+    },
+    {
+      "epoch": 0.6000830996156643,
+      "grad_norm": 0.3965167701244354,
+      "learning_rate": 3.4536736422692076e-05,
+      "loss": 1.4743,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6001869741352446,
+      "grad_norm": 0.42330119013786316,
+      "learning_rate": 3.4521220581635785e-05,
+      "loss": 1.4636,
+      "step": 5778
+    },
+    {
+      "epoch": 0.600290848654825,
+      "grad_norm": 0.3950074315071106,
+      "learning_rate": 3.4505706388948755e-05,
+      "loss": 1.7064,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6003947231744053,
+      "grad_norm": 0.42809754610061646,
+      "learning_rate": 3.4490193846283125e-05,
+      "loss": 1.7697,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6004985976939856,
+      "grad_norm": 0.4212953448295593,
+      "learning_rate": 3.44746829552909e-05,
+      "loss": 1.7538,
+      "step": 5781
+    },
+    {
+      "epoch": 0.600602472213566,
+      "grad_norm": 0.4184706211090088,
+      "learning_rate": 3.445917371762382e-05,
+      "loss": 1.8034,
+      "step": 5782
+    },
+    {
+      "epoch": 0.6007063467331464,
+      "grad_norm": 0.42354482412338257,
+      "learning_rate": 3.444366613493351e-05,
+      "loss": 1.6961,
+      "step": 5783
+    },
+    {
+      "epoch": 0.6008102212527268,
+      "grad_norm": 0.388049840927124,
+      "learning_rate": 3.4428160208871414e-05,
+      "loss": 1.6964,
+      "step": 5784
+    },
+    {
+      "epoch": 0.600914095772307,
+      "grad_norm": 0.4392579197883606,
+      "learning_rate": 3.441265594108881e-05,
+      "loss": 1.5853,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6010179702918874,
+      "grad_norm": 0.3931465148925781,
+      "learning_rate": 3.4397153333236735e-05,
+      "loss": 1.5617,
+      "step": 5786
+    },
+    {
+      "epoch": 0.6011218448114678,
+      "grad_norm": 0.3917011320590973,
+      "learning_rate": 3.438165238696614e-05,
+      "loss": 1.642,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6012257193310481,
+      "grad_norm": 0.41590529680252075,
+      "learning_rate": 3.436615310392773e-05,
+      "loss": 1.5589,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6013295938506285,
+      "grad_norm": 0.46075764298439026,
+      "learning_rate": 3.4350655485772074e-05,
+      "loss": 1.8877,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6014334683702088,
+      "grad_norm": 0.4078013300895691,
+      "learning_rate": 3.433515953414953e-05,
+      "loss": 1.5837,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6015373428897891,
+      "grad_norm": 0.4370696246623993,
+      "learning_rate": 3.431966525071033e-05,
+      "loss": 1.7953,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6016412174093695,
+      "grad_norm": 0.3889653980731964,
+      "learning_rate": 3.4304172637104436e-05,
+      "loss": 1.6294,
+      "step": 5792
+    },
+    {
+      "epoch": 0.6017450919289499,
+      "grad_norm": 0.3979250192642212,
+      "learning_rate": 3.4288681694981735e-05,
+      "loss": 1.686,
+      "step": 5793
+    },
+    {
+      "epoch": 0.6018489664485301,
+      "grad_norm": 0.428313672542572,
+      "learning_rate": 3.42731924259919e-05,
+      "loss": 1.6336,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6019528409681105,
+      "grad_norm": 0.4451068043708801,
+      "learning_rate": 3.4257704831784375e-05,
+      "loss": 1.6593,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6020567154876909,
+      "grad_norm": 0.47442716360092163,
+      "learning_rate": 3.42422189140085e-05,
+      "loss": 1.8737,
+      "step": 5796
+    },
+    {
+      "epoch": 0.6021605900072712,
+      "grad_norm": 0.4260861277580261,
+      "learning_rate": 3.42267346743134e-05,
+      "loss": 1.7137,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6022644645268516,
+      "grad_norm": 0.43714046478271484,
+      "learning_rate": 3.4211252114348e-05,
+      "loss": 1.7122,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6023683390464319,
+      "grad_norm": 0.4364699721336365,
+      "learning_rate": 3.41957712357611e-05,
+      "loss": 1.6723,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6024722135660122,
+      "grad_norm": 0.39638301730155945,
+      "learning_rate": 3.41802920402013e-05,
+      "loss": 1.6197,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6025760880855926,
+      "grad_norm": 0.4492684602737427,
+      "learning_rate": 3.416481452931699e-05,
+      "loss": 1.5572,
+      "step": 5801
+    },
+    {
+      "epoch": 0.602679962605173,
+      "grad_norm": 0.3939664661884308,
+      "learning_rate": 3.4149338704756416e-05,
+      "loss": 1.6363,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6027838371247533,
+      "grad_norm": 0.5074815154075623,
+      "learning_rate": 3.413386456816763e-05,
+      "loss": 1.8986,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6028877116443336,
+      "grad_norm": 0.4267491102218628,
+      "learning_rate": 3.411839212119851e-05,
+      "loss": 1.64,
+      "step": 5804
+    },
+    {
+      "epoch": 0.602991586163914,
+      "grad_norm": 0.39998719096183777,
+      "learning_rate": 3.410292136549673e-05,
+      "loss": 1.577,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6030954606834943,
+      "grad_norm": 0.3894689381122589,
+      "learning_rate": 3.408745230270985e-05,
+      "loss": 1.6308,
+      "step": 5806
+    },
+    {
+      "epoch": 0.6031993352030747,
+      "grad_norm": 0.4221261441707611,
+      "learning_rate": 3.407198493448516e-05,
+      "loss": 1.6313,
+      "step": 5807
+    },
+    {
+      "epoch": 0.603303209722655,
+      "grad_norm": 0.4228569269180298,
+      "learning_rate": 3.405651926246982e-05,
+      "loss": 1.7215,
+      "step": 5808
+    },
+    {
+      "epoch": 0.6034070842422353,
+      "grad_norm": 0.3922746479511261,
+      "learning_rate": 3.4041055288310844e-05,
+      "loss": 1.665,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6035109587618157,
+      "grad_norm": 0.39183309674263,
+      "learning_rate": 3.402559301365498e-05,
+      "loss": 1.6458,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6036148332813961,
+      "grad_norm": 0.4340969920158386,
+      "learning_rate": 3.401013244014884e-05,
+      "loss": 1.6737,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6037187078009765,
+      "grad_norm": 0.4317123591899872,
+      "learning_rate": 3.3994673569438886e-05,
+      "loss": 1.702,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6038225823205567,
+      "grad_norm": 0.4379931092262268,
+      "learning_rate": 3.397921640317136e-05,
+      "loss": 1.6435,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6039264568401371,
+      "grad_norm": 0.4368740916252136,
+      "learning_rate": 3.39637609429923e-05,
+      "loss": 1.7202,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6040303313597175,
+      "grad_norm": 0.4536083936691284,
+      "learning_rate": 3.3948307190547624e-05,
+      "loss": 1.664,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6041342058792978,
+      "grad_norm": 0.40802082419395447,
+      "learning_rate": 3.3932855147483024e-05,
+      "loss": 1.6268,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6042380803988782,
+      "grad_norm": 0.4093196392059326,
+      "learning_rate": 3.391740481544402e-05,
+      "loss": 1.7967,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6043419549184585,
+      "grad_norm": 0.41186216473579407,
+      "learning_rate": 3.3901956196075945e-05,
+      "loss": 1.5787,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6044458294380388,
+      "grad_norm": 0.431708961725235,
+      "learning_rate": 3.3886509291023986e-05,
+      "loss": 1.6593,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6045497039576192,
+      "grad_norm": 0.42467281222343445,
+      "learning_rate": 3.387106410193308e-05,
+      "loss": 1.6942,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6046535784771996,
+      "grad_norm": 0.3974582254886627,
+      "learning_rate": 3.385562063044803e-05,
+      "loss": 1.5867,
+      "step": 5821
+    },
+    {
+      "epoch": 0.6047574529967799,
+      "grad_norm": 0.4653453528881073,
+      "learning_rate": 3.384017887821348e-05,
+      "loss": 1.8252,
+      "step": 5822
+    },
+    {
+      "epoch": 0.6048613275163602,
+      "grad_norm": 0.4493721127510071,
+      "learning_rate": 3.382473884687382e-05,
+      "loss": 1.8101,
+      "step": 5823
+    },
+    {
+      "epoch": 0.6049652020359406,
+      "grad_norm": 0.3836005926132202,
+      "learning_rate": 3.380930053807328e-05,
+      "loss": 1.6124,
+      "step": 5824
+    },
+    {
+      "epoch": 0.6050690765555209,
+      "grad_norm": 0.38905587792396545,
+      "learning_rate": 3.379386395345596e-05,
+      "loss": 1.575,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6051729510751013,
+      "grad_norm": 0.4363052248954773,
+      "learning_rate": 3.377842909466571e-05,
+      "loss": 1.6505,
+      "step": 5826
+    },
+    {
+      "epoch": 0.6052768255946817,
+      "grad_norm": 0.40476155281066895,
+      "learning_rate": 3.376299596334622e-05,
+      "loss": 1.7677,
+      "step": 5827
+    },
+    {
+      "epoch": 0.6053807001142619,
+      "grad_norm": 0.42141294479370117,
+      "learning_rate": 3.374756456114102e-05,
+      "loss": 1.5649,
+      "step": 5828
+    },
+    {
+      "epoch": 0.6054845746338423,
+      "grad_norm": 0.45798051357269287,
+      "learning_rate": 3.373213488969341e-05,
+      "loss": 1.8967,
+      "step": 5829
+    },
+    {
+      "epoch": 0.6055884491534227,
+      "grad_norm": 0.4245013892650604,
+      "learning_rate": 3.371670695064655e-05,
+      "loss": 1.6626,
+      "step": 5830
+    },
+    {
+      "epoch": 0.605692323673003,
+      "grad_norm": 0.4293268918991089,
+      "learning_rate": 3.370128074564338e-05,
+      "loss": 1.9178,
+      "step": 5831
+    },
+    {
+      "epoch": 0.6057961981925833,
+      "grad_norm": 0.42448824644088745,
+      "learning_rate": 3.368585627632668e-05,
+      "loss": 1.7219,
+      "step": 5832
+    },
+    {
+      "epoch": 0.6059000727121637,
+      "grad_norm": 0.3986135423183441,
+      "learning_rate": 3.3670433544339014e-05,
+      "loss": 1.6541,
+      "step": 5833
+    },
+    {
+      "epoch": 0.606003947231744,
+      "grad_norm": 0.44702157378196716,
+      "learning_rate": 3.365501255132283e-05,
+      "loss": 1.5446,
+      "step": 5834
+    },
+    {
+      "epoch": 0.6061078217513244,
+      "grad_norm": 0.39528048038482666,
+      "learning_rate": 3.363959329892029e-05,
+      "loss": 1.5639,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6062116962709048,
+      "grad_norm": 0.40425539016723633,
+      "learning_rate": 3.362417578877344e-05,
+      "loss": 1.5162,
+      "step": 5836
+    },
+    {
+      "epoch": 0.6063155707904851,
+      "grad_norm": 0.3955405056476593,
+      "learning_rate": 3.360876002252415e-05,
+      "loss": 1.6176,
+      "step": 5837
+    },
+    {
+      "epoch": 0.6064194453100654,
+      "grad_norm": 0.4061652421951294,
+      "learning_rate": 3.359334600181407e-05,
+      "loss": 1.7371,
+      "step": 5838
+    },
+    {
+      "epoch": 0.6065233198296458,
+      "grad_norm": 0.41885098814964294,
+      "learning_rate": 3.357793372828464e-05,
+      "loss": 1.4975,
+      "step": 5839
+    },
+    {
+      "epoch": 0.6066271943492262,
+      "grad_norm": 0.3937050998210907,
+      "learning_rate": 3.356252320357718e-05,
+      "loss": 1.623,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6067310688688065,
+      "grad_norm": 0.4089513123035431,
+      "learning_rate": 3.354711442933279e-05,
+      "loss": 1.5882,
+      "step": 5841
+    },
+    {
+      "epoch": 0.6068349433883868,
+      "grad_norm": 0.43716487288475037,
+      "learning_rate": 3.353170740719236e-05,
+      "loss": 1.814,
+      "step": 5842
+    },
+    {
+      "epoch": 0.6069388179079672,
+      "grad_norm": 0.4473409056663513,
+      "learning_rate": 3.351630213879665e-05,
+      "loss": 1.5899,
+      "step": 5843
+    },
+    {
+      "epoch": 0.6070426924275475,
+      "grad_norm": 0.40252211689949036,
+      "learning_rate": 3.3500898625786184e-05,
+      "loss": 1.6746,
+      "step": 5844
+    },
+    {
+      "epoch": 0.6071465669471279,
+      "grad_norm": 0.406857967376709,
+      "learning_rate": 3.34854968698013e-05,
+      "loss": 1.5783,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6072504414667083,
+      "grad_norm": 0.41453415155410767,
+      "learning_rate": 3.347009687248219e-05,
+      "loss": 1.7318,
+      "step": 5846
+    },
+    {
+      "epoch": 0.6073543159862885,
+      "grad_norm": 0.3779454231262207,
+      "learning_rate": 3.345469863546884e-05,
+      "loss": 1.5555,
+      "step": 5847
+    },
+    {
+      "epoch": 0.6074581905058689,
+      "grad_norm": 0.40654996037483215,
+      "learning_rate": 3.3439302160401e-05,
+      "loss": 1.5914,
+      "step": 5848
+    },
+    {
+      "epoch": 0.6075620650254493,
+      "grad_norm": 0.4251600205898285,
+      "learning_rate": 3.342390744891831e-05,
+      "loss": 1.7131,
+      "step": 5849
+    },
+    {
+      "epoch": 0.6076659395450296,
+      "grad_norm": 0.45632246136665344,
+      "learning_rate": 3.340851450266019e-05,
+      "loss": 1.7986,
+      "step": 5850
+    },
+    {
+      "epoch": 0.60776981406461,
+      "grad_norm": 0.43041008710861206,
+      "learning_rate": 3.3393123323265855e-05,
+      "loss": 1.7041,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6078736885841903,
+      "grad_norm": 0.3932364284992218,
+      "learning_rate": 3.337773391237433e-05,
+      "loss": 1.5044,
+      "step": 5852
+    },
+    {
+      "epoch": 0.6079775631037706,
+      "grad_norm": 0.3887140154838562,
+      "learning_rate": 3.336234627162451e-05,
+      "loss": 1.6597,
+      "step": 5853
+    },
+    {
+      "epoch": 0.608081437623351,
+      "grad_norm": 0.41823363304138184,
+      "learning_rate": 3.3346960402655016e-05,
+      "loss": 1.6794,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6081853121429314,
+      "grad_norm": 0.4427807033061981,
+      "learning_rate": 3.333157630710434e-05,
+      "loss": 1.8548,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6082891866625116,
+      "grad_norm": 0.4366241693496704,
+      "learning_rate": 3.331619398661078e-05,
+      "loss": 1.597,
+      "step": 5856
+    },
+    {
+      "epoch": 0.608393061182092,
+      "grad_norm": 0.4059387147426605,
+      "learning_rate": 3.3300813442812406e-05,
+      "loss": 1.5453,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6084969357016724,
+      "grad_norm": 0.40862488746643066,
+      "learning_rate": 3.328543467734715e-05,
+      "loss": 1.6159,
+      "step": 5858
+    },
+    {
+      "epoch": 0.6086008102212527,
+      "grad_norm": 0.4672967195510864,
+      "learning_rate": 3.327005769185273e-05,
+      "loss": 1.9154,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6087046847408331,
+      "grad_norm": 0.4559909403324127,
+      "learning_rate": 3.325468248796666e-05,
+      "loss": 1.6842,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6088085592604134,
+      "grad_norm": 0.3882789611816406,
+      "learning_rate": 3.323930906732628e-05,
+      "loss": 1.5068,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6089124337799938,
+      "grad_norm": 0.4709301292896271,
+      "learning_rate": 3.322393743156874e-05,
+      "loss": 1.8432,
+      "step": 5862
+    },
+    {
+      "epoch": 0.6090163082995741,
+      "grad_norm": 0.43151330947875977,
+      "learning_rate": 3.3208567582331034e-05,
+      "loss": 1.7516,
+      "step": 5863
+    },
+    {
+      "epoch": 0.6091201828191545,
+      "grad_norm": 0.3997372090816498,
+      "learning_rate": 3.3193199521249876e-05,
+      "loss": 1.7277,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6092240573387349,
+      "grad_norm": 0.44914302229881287,
+      "learning_rate": 3.3177833249961886e-05,
+      "loss": 1.5975,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6093279318583151,
+      "grad_norm": 0.3897424638271332,
+      "learning_rate": 3.316246877010344e-05,
+      "loss": 1.6415,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6094318063778955,
+      "grad_norm": 0.4456162452697754,
+      "learning_rate": 3.314710608331072e-05,
+      "loss": 1.7596,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6095356808974759,
+      "grad_norm": 0.4402760863304138,
+      "learning_rate": 3.3131745191219744e-05,
+      "loss": 1.7909,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6096395554170562,
+      "grad_norm": 0.41520991921424866,
+      "learning_rate": 3.311638609546635e-05,
+      "loss": 1.6518,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6097434299366365,
+      "grad_norm": 0.40946778655052185,
+      "learning_rate": 3.3101028797686106e-05,
+      "loss": 1.6841,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6098473044562169,
+      "grad_norm": 0.4258642792701721,
+      "learning_rate": 3.3085673299514496e-05,
+      "loss": 1.7182,
+      "step": 5871
+    },
+    {
+      "epoch": 0.6099511789757972,
+      "grad_norm": 0.38915473222732544,
+      "learning_rate": 3.307031960258674e-05,
+      "loss": 1.5292,
+      "step": 5872
+    },
+    {
+      "epoch": 0.6100550534953776,
+      "grad_norm": 0.39618629217147827,
+      "learning_rate": 3.30549677085379e-05,
+      "loss": 1.6406,
+      "step": 5873
+    },
+    {
+      "epoch": 0.610158928014958,
+      "grad_norm": 0.3926134407520294,
+      "learning_rate": 3.30396176190028e-05,
+      "loss": 1.5848,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6102628025345382,
+      "grad_norm": 0.41277050971984863,
+      "learning_rate": 3.3024269335616155e-05,
+      "loss": 1.6565,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6103666770541186,
+      "grad_norm": 0.40988320112228394,
+      "learning_rate": 3.300892286001238e-05,
+      "loss": 1.72,
+      "step": 5876
+    },
+    {
+      "epoch": 0.610470551573699,
+      "grad_norm": 0.4124213457107544,
+      "learning_rate": 3.29935781938258e-05,
+      "loss": 1.5959,
+      "step": 5877
+    },
+    {
+      "epoch": 0.6105744260932793,
+      "grad_norm": 0.3976089060306549,
+      "learning_rate": 3.29782353386905e-05,
+      "loss": 1.7009,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6106783006128597,
+      "grad_norm": 0.40462133288383484,
+      "learning_rate": 3.296289429624035e-05,
+      "loss": 1.5514,
+      "step": 5879
+    },
+    {
+      "epoch": 0.61078217513244,
+      "grad_norm": 0.42904216051101685,
+      "learning_rate": 3.294755506810905e-05,
+      "loss": 1.6511,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6108860496520203,
+      "grad_norm": 0.44326838850975037,
+      "learning_rate": 3.293221765593015e-05,
+      "loss": 1.4797,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6109899241716007,
+      "grad_norm": 0.3995533883571625,
+      "learning_rate": 3.291688206133691e-05,
+      "loss": 1.77,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6110937986911811,
+      "grad_norm": 0.3896981179714203,
+      "learning_rate": 3.290154828596248e-05,
+      "loss": 1.6022,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6111976732107613,
+      "grad_norm": 0.38104161620140076,
+      "learning_rate": 3.28862163314398e-05,
+      "loss": 1.5761,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6113015477303417,
+      "grad_norm": 0.4309033155441284,
+      "learning_rate": 3.287088619940158e-05,
+      "loss": 1.7047,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6114054222499221,
+      "grad_norm": 0.38424256443977356,
+      "learning_rate": 3.285555789148036e-05,
+      "loss": 1.5804,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6115092967695024,
+      "grad_norm": 0.39635035395622253,
+      "learning_rate": 3.284023140930851e-05,
+      "loss": 1.7646,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6116131712890828,
+      "grad_norm": 0.40174224972724915,
+      "learning_rate": 3.2824906754518146e-05,
+      "loss": 1.6731,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6117170458086632,
+      "grad_norm": 0.42028430104255676,
+      "learning_rate": 3.280958392874124e-05,
+      "loss": 1.7342,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6118209203282435,
+      "grad_norm": 0.40835410356521606,
+      "learning_rate": 3.2794262933609555e-05,
+      "loss": 1.6333,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6119247948478238,
+      "grad_norm": 0.43254995346069336,
+      "learning_rate": 3.277894377075468e-05,
+      "loss": 1.8507,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6120286693674042,
+      "grad_norm": 0.41028037667274475,
+      "learning_rate": 3.276362644180794e-05,
+      "loss": 1.712,
+      "step": 5892
+    },
+    {
+      "epoch": 0.6121325438869846,
+      "grad_norm": 0.428543359041214,
+      "learning_rate": 3.2748310948400545e-05,
+      "loss": 1.6575,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6122364184065648,
+      "grad_norm": 0.46985751390457153,
+      "learning_rate": 3.273299729216348e-05,
+      "loss": 1.9697,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6123402929261452,
+      "grad_norm": 0.4072946012020111,
+      "learning_rate": 3.2717685474727495e-05,
+      "loss": 1.7102,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6124441674457256,
+      "grad_norm": 0.41395875811576843,
+      "learning_rate": 3.270237549772319e-05,
+      "loss": 1.6602,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6125480419653059,
+      "grad_norm": 0.40566155314445496,
+      "learning_rate": 3.2687067362780997e-05,
+      "loss": 1.6355,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6126519164848863,
+      "grad_norm": 0.40978339314460754,
+      "learning_rate": 3.267176107153107e-05,
+      "loss": 1.5367,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6127557910044666,
+      "grad_norm": 0.39154794812202454,
+      "learning_rate": 3.2656456625603425e-05,
+      "loss": 1.7275,
+      "step": 5899
+    },
+    {
+      "epoch": 0.6128596655240469,
+      "grad_norm": 0.4286908209323883,
+      "learning_rate": 3.2641154026627866e-05,
+      "loss": 1.7106,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6129635400436273,
+      "grad_norm": 0.39364275336265564,
+      "learning_rate": 3.2625853276234e-05,
+      "loss": 1.6203,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6130674145632077,
+      "grad_norm": 0.43890032172203064,
+      "learning_rate": 3.2610554376051224e-05,
+      "loss": 1.6936,
+      "step": 5902
+    },
+    {
+      "epoch": 0.613171289082788,
+      "grad_norm": 0.4209052622318268,
+      "learning_rate": 3.2595257327708785e-05,
+      "loss": 1.9501,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6132751636023683,
+      "grad_norm": 0.4081374704837799,
+      "learning_rate": 3.257996213283566e-05,
+      "loss": 1.5555,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6133790381219487,
+      "grad_norm": 0.3956921696662903,
+      "learning_rate": 3.256466879306068e-05,
+      "loss": 1.5927,
+      "step": 5905
+    },
+    {
+      "epoch": 0.613482912641529,
+      "grad_norm": 0.4056268334388733,
+      "learning_rate": 3.25493773100125e-05,
+      "loss": 1.6676,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6135867871611094,
+      "grad_norm": 0.4204394221305847,
+      "learning_rate": 3.2534087685319506e-05,
+      "loss": 1.5935,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6136906616806898,
+      "grad_norm": 0.4562225043773651,
+      "learning_rate": 3.2518799920609924e-05,
+      "loss": 1.7057,
+      "step": 5908
+    },
+    {
+      "epoch": 0.61379453620027,
+      "grad_norm": 0.4130711257457733,
+      "learning_rate": 3.2503514017511806e-05,
+      "loss": 1.6022,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6138984107198504,
+      "grad_norm": 0.4055189788341522,
+      "learning_rate": 3.2488229977652954e-05,
+      "loss": 1.7709,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6140022852394308,
+      "grad_norm": 0.406704306602478,
+      "learning_rate": 3.2472947802661005e-05,
+      "loss": 1.702,
+      "step": 5911
+    },
+    {
+      "epoch": 0.6141061597590111,
+      "grad_norm": 0.41046005487442017,
+      "learning_rate": 3.245766749416342e-05,
+      "loss": 1.6426,
+      "step": 5912
+    },
+    {
+      "epoch": 0.6142100342785914,
+      "grad_norm": 0.40672361850738525,
+      "learning_rate": 3.24423890537874e-05,
+      "loss": 1.6559,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6143139087981718,
+      "grad_norm": 0.3978021442890167,
+      "learning_rate": 3.242711248316e-05,
+      "loss": 1.7074,
+      "step": 5914
+    },
+    {
+      "epoch": 0.6144177833177522,
+      "grad_norm": 0.43113160133361816,
+      "learning_rate": 3.241183778390803e-05,
+      "loss": 1.8111,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6145216578373325,
+      "grad_norm": 0.3847864270210266,
+      "learning_rate": 3.239656495765817e-05,
+      "loss": 1.5673,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6146255323569129,
+      "grad_norm": 0.406105101108551,
+      "learning_rate": 3.23812940060368e-05,
+      "loss": 1.6545,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6147294068764932,
+      "grad_norm": 0.4283885657787323,
+      "learning_rate": 3.2366024930670194e-05,
+      "loss": 1.6233,
+      "step": 5918
+    },
+    {
+      "epoch": 0.6148332813960735,
+      "grad_norm": 0.39741960167884827,
+      "learning_rate": 3.23507577331844e-05,
+      "loss": 1.7184,
+      "step": 5919
+    },
+    {
+      "epoch": 0.6149371559156539,
+      "grad_norm": 0.4278057813644409,
+      "learning_rate": 3.2335492415205224e-05,
+      "loss": 1.6877,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6150410304352343,
+      "grad_norm": 0.3887667655944824,
+      "learning_rate": 3.2320228978358324e-05,
+      "loss": 1.6024,
+      "step": 5921
+    },
+    {
+      "epoch": 0.6151449049548146,
+      "grad_norm": 0.4222593903541565,
+      "learning_rate": 3.230496742426914e-05,
+      "loss": 1.5523,
+      "step": 5922
+    },
+    {
+      "epoch": 0.6152487794743949,
+      "grad_norm": 0.4194217026233673,
+      "learning_rate": 3.228970775456288e-05,
+      "loss": 1.6793,
+      "step": 5923
+    },
+    {
+      "epoch": 0.6153526539939753,
+      "grad_norm": 0.41720110177993774,
+      "learning_rate": 3.2274449970864595e-05,
+      "loss": 1.6962,
+      "step": 5924
+    },
+    {
+      "epoch": 0.6154565285135556,
+      "grad_norm": 0.43170827627182007,
+      "learning_rate": 3.225919407479914e-05,
+      "loss": 1.7432,
+      "step": 5925
+    },
+    {
+      "epoch": 0.615560403033136,
+      "grad_norm": 0.41001245379447937,
+      "learning_rate": 3.224394006799112e-05,
+      "loss": 1.6198,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6156642775527164,
+      "grad_norm": 0.39082086086273193,
+      "learning_rate": 3.222868795206498e-05,
+      "loss": 1.5267,
+      "step": 5927
+    },
+    {
+      "epoch": 0.6157681520722966,
+      "grad_norm": 0.4277300536632538,
+      "learning_rate": 3.221343772864497e-05,
+      "loss": 1.8421,
+      "step": 5928
+    },
+    {
+      "epoch": 0.615872026591877,
+      "grad_norm": 0.4021863639354706,
+      "learning_rate": 3.2198189399355085e-05,
+      "loss": 1.701,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6159759011114574,
+      "grad_norm": 0.39132073521614075,
+      "learning_rate": 3.218294296581916e-05,
+      "loss": 1.5238,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6160797756310377,
+      "grad_norm": 0.44258078932762146,
+      "learning_rate": 3.2167698429660866e-05,
+      "loss": 1.7779,
+      "step": 5931
+    },
+    {
+      "epoch": 0.616183650150618,
+      "grad_norm": 0.4172841012477875,
+      "learning_rate": 3.215245579250355e-05,
+      "loss": 1.6755,
+      "step": 5932
+    },
+    {
+      "epoch": 0.6162875246701984,
+      "grad_norm": 0.46141964197158813,
+      "learning_rate": 3.2137215055970496e-05,
+      "loss": 1.8098,
+      "step": 5933
+    },
+    {
+      "epoch": 0.6163913991897787,
+      "grad_norm": 0.39755067229270935,
+      "learning_rate": 3.2121976221684714e-05,
+      "loss": 1.5704,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6164952737093591,
+      "grad_norm": 0.44859442114830017,
+      "learning_rate": 3.2106739291269005e-05,
+      "loss": 1.7943,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6165991482289395,
+      "grad_norm": 0.4175950288772583,
+      "learning_rate": 3.209150426634597e-05,
+      "loss": 1.7347,
+      "step": 5936
+    },
+    {
+      "epoch": 0.6167030227485197,
+      "grad_norm": 0.42605000734329224,
+      "learning_rate": 3.2076271148538066e-05,
+      "loss": 1.6469,
+      "step": 5937
+    },
+    {
+      "epoch": 0.6168068972681001,
+      "grad_norm": 0.42592230439186096,
+      "learning_rate": 3.2061039939467454e-05,
+      "loss": 1.7029,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6169107717876805,
+      "grad_norm": 0.4934481978416443,
+      "learning_rate": 3.204581064075616e-05,
+      "loss": 2.0553,
+      "step": 5939
+    },
+    {
+      "epoch": 0.6170146463072609,
+      "grad_norm": 0.4315035939216614,
+      "learning_rate": 3.203058325402599e-05,
+      "loss": 1.7076,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6171185208268412,
+      "grad_norm": 0.4047625958919525,
+      "learning_rate": 3.201535778089853e-05,
+      "loss": 1.5993,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6172223953464215,
+      "grad_norm": 0.4070419371128082,
+      "learning_rate": 3.20001342229952e-05,
+      "loss": 1.3709,
+      "step": 5942
+    },
+    {
+      "epoch": 0.6173262698660019,
+      "grad_norm": 0.4009752869606018,
+      "learning_rate": 3.1984912581937146e-05,
+      "loss": 1.6183,
+      "step": 5943
+    },
+    {
+      "epoch": 0.6174301443855822,
+      "grad_norm": 0.40829452872276306,
+      "learning_rate": 3.196969285934541e-05,
+      "loss": 1.5723,
+      "step": 5944
+    },
+    {
+      "epoch": 0.6175340189051626,
+      "grad_norm": 0.43150803446769714,
+      "learning_rate": 3.1954475056840714e-05,
+      "loss": 1.593,
+      "step": 5945
+    },
+    {
+      "epoch": 0.617637893424743,
+      "grad_norm": 0.4301930367946625,
+      "learning_rate": 3.193925917604367e-05,
+      "loss": 1.7796,
+      "step": 5946
+    },
+    {
+      "epoch": 0.6177417679443232,
+      "grad_norm": 0.4352579414844513,
+      "learning_rate": 3.1924045218574654e-05,
+      "loss": 1.8134,
+      "step": 5947
+    },
+    {
+      "epoch": 0.6178456424639036,
+      "grad_norm": 0.3833911716938019,
+      "learning_rate": 3.190883318605382e-05,
+      "loss": 1.5344,
+      "step": 5948
+    },
+    {
+      "epoch": 0.617949516983484,
+      "grad_norm": 0.4389069676399231,
+      "learning_rate": 3.189362308010113e-05,
+      "loss": 1.6706,
+      "step": 5949
+    },
+    {
+      "epoch": 0.6180533915030643,
+      "grad_norm": 0.3941787779331207,
+      "learning_rate": 3.187841490233636e-05,
+      "loss": 1.5388,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6181572660226446,
+      "grad_norm": 0.44791027903556824,
+      "learning_rate": 3.1863208654379036e-05,
+      "loss": 1.7468,
+      "step": 5951
+    },
+    {
+      "epoch": 0.618261140542225,
+      "grad_norm": 0.41120895743370056,
+      "learning_rate": 3.184800433784851e-05,
+      "loss": 1.5909,
+      "step": 5952
+    },
+    {
+      "epoch": 0.6183650150618053,
+      "grad_norm": 0.41934871673583984,
+      "learning_rate": 3.1832801954363955e-05,
+      "loss": 1.8014,
+      "step": 5953
+    },
+    {
+      "epoch": 0.6184688895813857,
+      "grad_norm": 0.38895368576049805,
+      "learning_rate": 3.1817601505544264e-05,
+      "loss": 1.6452,
+      "step": 5954
+    },
+    {
+      "epoch": 0.6185727641009661,
+      "grad_norm": 0.38875988125801086,
+      "learning_rate": 3.18024029930082e-05,
+      "loss": 1.6638,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6186766386205463,
+      "grad_norm": 0.4126511216163635,
+      "learning_rate": 3.178720641837426e-05,
+      "loss": 1.6237,
+      "step": 5956
+    },
+    {
+      "epoch": 0.6187805131401267,
+      "grad_norm": 0.4127541184425354,
+      "learning_rate": 3.177201178326077e-05,
+      "loss": 1.6954,
+      "step": 5957
+    },
+    {
+      "epoch": 0.6188843876597071,
+      "grad_norm": 0.4332854449748993,
+      "learning_rate": 3.175681908928584e-05,
+      "loss": 1.7631,
+      "step": 5958
+    },
+    {
+      "epoch": 0.6189882621792874,
+      "grad_norm": 0.401528537273407,
+      "learning_rate": 3.174162833806739e-05,
+      "loss": 1.6,
+      "step": 5959
+    },
+    {
+      "epoch": 0.6190921366988678,
+      "grad_norm": 0.4269302785396576,
+      "learning_rate": 3.172643953122308e-05,
+      "loss": 1.7953,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6191960112184481,
+      "grad_norm": 0.4833371341228485,
+      "learning_rate": 3.171125267037042e-05,
+      "loss": 1.779,
+      "step": 5961
+    },
+    {
+      "epoch": 0.6192998857380284,
+      "grad_norm": 0.4078036844730377,
+      "learning_rate": 3.169606775712672e-05,
+      "loss": 1.7738,
+      "step": 5962
+    },
+    {
+      "epoch": 0.6194037602576088,
+      "grad_norm": 0.3845091164112091,
+      "learning_rate": 3.1680884793109014e-05,
+      "loss": 1.488,
+      "step": 5963
+    },
+    {
+      "epoch": 0.6195076347771892,
+      "grad_norm": 0.4441048800945282,
+      "learning_rate": 3.166570377993417e-05,
+      "loss": 1.6613,
+      "step": 5964
+    },
+    {
+      "epoch": 0.6196115092967694,
+      "grad_norm": 0.4266035556793213,
+      "learning_rate": 3.165052471921888e-05,
+      "loss": 1.6224,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6197153838163498,
+      "grad_norm": 0.4394325315952301,
+      "learning_rate": 3.163534761257956e-05,
+      "loss": 1.6947,
+      "step": 5966
+    },
+    {
+      "epoch": 0.6198192583359302,
+      "grad_norm": 0.4144458472728729,
+      "learning_rate": 3.162017246163247e-05,
+      "loss": 1.7144,
+      "step": 5967
+    },
+    {
+      "epoch": 0.6199231328555106,
+      "grad_norm": 0.39400461316108704,
+      "learning_rate": 3.160499926799365e-05,
+      "loss": 1.6563,
+      "step": 5968
+    },
+    {
+      "epoch": 0.6200270073750909,
+      "grad_norm": 0.40855416655540466,
+      "learning_rate": 3.1589828033278925e-05,
+      "loss": 1.7237,
+      "step": 5969
+    },
+    {
+      "epoch": 0.6201308818946712,
+      "grad_norm": 0.39257362484931946,
+      "learning_rate": 3.157465875910391e-05,
+      "loss": 1.6751,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6202347564142516,
+      "grad_norm": 0.3722376525402069,
+      "learning_rate": 3.1559491447084e-05,
+      "loss": 1.547,
+      "step": 5971
+    },
+    {
+      "epoch": 0.6203386309338319,
+      "grad_norm": 0.38416385650634766,
+      "learning_rate": 3.1544326098834433e-05,
+      "loss": 1.6595,
+      "step": 5972
+    },
+    {
+      "epoch": 0.6204425054534123,
+      "grad_norm": 0.42782196402549744,
+      "learning_rate": 3.152916271597015e-05,
+      "loss": 1.755,
+      "step": 5973
+    },
+    {
+      "epoch": 0.6205463799729927,
+      "grad_norm": 0.3950657248497009,
+      "learning_rate": 3.151400130010597e-05,
+      "loss": 1.6521,
+      "step": 5974
+    },
+    {
+      "epoch": 0.6206502544925729,
+      "grad_norm": 0.39494791626930237,
+      "learning_rate": 3.149884185285647e-05,
+      "loss": 1.6084,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6207541290121533,
+      "grad_norm": 0.4145883321762085,
+      "learning_rate": 3.148368437583599e-05,
+      "loss": 1.5866,
+      "step": 5976
+    },
+    {
+      "epoch": 0.6208580035317337,
+      "grad_norm": 0.5039044618606567,
+      "learning_rate": 3.146852887065868e-05,
+      "loss": 1.8135,
+      "step": 5977
+    },
+    {
+      "epoch": 0.620961878051314,
+      "grad_norm": 0.42017415165901184,
+      "learning_rate": 3.145337533893851e-05,
+      "loss": 1.6958,
+      "step": 5978
+    },
+    {
+      "epoch": 0.6210657525708944,
+      "grad_norm": 0.41060763597488403,
+      "learning_rate": 3.143822378228918e-05,
+      "loss": 1.5397,
+      "step": 5979
+    },
+    {
+      "epoch": 0.6211696270904747,
+      "grad_norm": 0.42383599281311035,
+      "learning_rate": 3.142307420232423e-05,
+      "loss": 1.7797,
+      "step": 5980
+    },
+    {
+      "epoch": 0.621273501610055,
+      "grad_norm": 0.446959912776947,
+      "learning_rate": 3.1407926600657e-05,
+      "loss": 1.8221,
+      "step": 5981
+    },
+    {
+      "epoch": 0.6213773761296354,
+      "grad_norm": 0.403238445520401,
+      "learning_rate": 3.1392780978900535e-05,
+      "loss": 1.6687,
+      "step": 5982
+    },
+    {
+      "epoch": 0.6214812506492158,
+      "grad_norm": 0.4446134567260742,
+      "learning_rate": 3.1377637338667764e-05,
+      "loss": 1.719,
+      "step": 5983
+    },
+    {
+      "epoch": 0.621585125168796,
+      "grad_norm": 0.4344415068626404,
+      "learning_rate": 3.136249568157136e-05,
+      "loss": 1.6722,
+      "step": 5984
+    },
+    {
+      "epoch": 0.6216889996883764,
+      "grad_norm": 0.38726866245269775,
+      "learning_rate": 3.1347356009223786e-05,
+      "loss": 1.6432,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6217928742079568,
+      "grad_norm": 0.43266814947128296,
+      "learning_rate": 3.133221832323728e-05,
+      "loss": 1.747,
+      "step": 5986
+    },
+    {
+      "epoch": 0.6218967487275371,
+      "grad_norm": 0.39049196243286133,
+      "learning_rate": 3.131708262522395e-05,
+      "loss": 1.6543,
+      "step": 5987
+    },
+    {
+      "epoch": 0.6220006232471175,
+      "grad_norm": 0.40162894129753113,
+      "learning_rate": 3.1301948916795554e-05,
+      "loss": 1.5967,
+      "step": 5988
+    },
+    {
+      "epoch": 0.6221044977666979,
+      "grad_norm": 0.4087611734867096,
+      "learning_rate": 3.128681719956375e-05,
+      "loss": 1.647,
+      "step": 5989
+    },
+    {
+      "epoch": 0.6222083722862781,
+      "grad_norm": 0.4542302191257477,
+      "learning_rate": 3.1271687475139966e-05,
+      "loss": 1.7831,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6223122468058585,
+      "grad_norm": 0.40611404180526733,
+      "learning_rate": 3.125655974513537e-05,
+      "loss": 1.6214,
+      "step": 5991
+    },
+    {
+      "epoch": 0.6224161213254389,
+      "grad_norm": 0.4047275483608246,
+      "learning_rate": 3.124143401116094e-05,
+      "loss": 1.7274,
+      "step": 5992
+    },
+    {
+      "epoch": 0.6225199958450193,
+      "grad_norm": 0.4009588062763214,
+      "learning_rate": 3.1226310274827473e-05,
+      "loss": 1.7371,
+      "step": 5993
+    },
+    {
+      "epoch": 0.6226238703645995,
+      "grad_norm": 0.398447185754776,
+      "learning_rate": 3.121118853774554e-05,
+      "loss": 1.5706,
+      "step": 5994
+    },
+    {
+      "epoch": 0.6227277448841799,
+      "grad_norm": 0.46983155608177185,
+      "learning_rate": 3.119606880152545e-05,
+      "loss": 1.7618,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6228316194037603,
+      "grad_norm": 0.42344680428504944,
+      "learning_rate": 3.1180951067777365e-05,
+      "loss": 1.6264,
+      "step": 5996
+    },
+    {
+      "epoch": 0.6229354939233406,
+      "grad_norm": 0.4075701832771301,
+      "learning_rate": 3.11658353381112e-05,
+      "loss": 1.7759,
+      "step": 5997
+    },
+    {
+      "epoch": 0.623039368442921,
+      "grad_norm": 0.40149906277656555,
+      "learning_rate": 3.1150721614136655e-05,
+      "loss": 1.7063,
+      "step": 5998
+    },
+    {
+      "epoch": 0.6231432429625013,
+      "grad_norm": 0.3998561203479767,
+      "learning_rate": 3.113560989746322e-05,
+      "loss": 1.4695,
+      "step": 5999
+    },
+    {
+      "epoch": 0.6232471174820816,
+      "grad_norm": 0.4251110553741455,
+      "learning_rate": 3.11205001897002e-05,
+      "loss": 1.7207,
+      "step": 6000
+    },
+    {
+      "epoch": 0.623350992001662,
+      "grad_norm": 0.41832244396209717,
+      "learning_rate": 3.110539249245663e-05,
+      "loss": 1.4959,
+      "step": 6001
+    },
+    {
+      "epoch": 0.6234548665212424,
+      "grad_norm": 0.3928918242454529,
+      "learning_rate": 3.1090286807341374e-05,
+      "loss": 1.7322,
+      "step": 6002
+    },
+    {
+      "epoch": 0.6235587410408227,
+      "grad_norm": 0.4026903212070465,
+      "learning_rate": 3.107518313596309e-05,
+      "loss": 1.8031,
+      "step": 6003
+    },
+    {
+      "epoch": 0.623662615560403,
+      "grad_norm": 0.4026026129722595,
+      "learning_rate": 3.106008147993017e-05,
+      "loss": 1.5662,
+      "step": 6004
+    },
+    {
+      "epoch": 0.6237664900799834,
+      "grad_norm": 0.4900519549846649,
+      "learning_rate": 3.1044981840850826e-05,
+      "loss": 1.7387,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6238703645995637,
+      "grad_norm": 0.4159107804298401,
+      "learning_rate": 3.102988422033309e-05,
+      "loss": 1.6261,
+      "step": 6006
+    },
+    {
+      "epoch": 0.6239742391191441,
+      "grad_norm": 0.46456992626190186,
+      "learning_rate": 3.101478861998469e-05,
+      "loss": 1.8022,
+      "step": 6007
+    },
+    {
+      "epoch": 0.6240781136387245,
+      "grad_norm": 0.4258232116699219,
+      "learning_rate": 3.09996950414132e-05,
+      "loss": 1.7152,
+      "step": 6008
+    },
+    {
+      "epoch": 0.6241819881583047,
+      "grad_norm": 0.4385705888271332,
+      "learning_rate": 3.098460348622601e-05,
+      "loss": 1.7273,
+      "step": 6009
+    },
+    {
+      "epoch": 0.6242858626778851,
+      "grad_norm": 0.3911799192428589,
+      "learning_rate": 3.09695139560302e-05,
+      "loss": 1.525,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6243897371974655,
+      "grad_norm": 0.45774349570274353,
+      "learning_rate": 3.095442645243272e-05,
+      "loss": 1.8,
+      "step": 6011
+    },
+    {
+      "epoch": 0.6244936117170458,
+      "grad_norm": 0.3881503939628601,
+      "learning_rate": 3.0939340977040267e-05,
+      "loss": 1.7008,
+      "step": 6012
+    },
+    {
+      "epoch": 0.6245974862366261,
+      "grad_norm": 0.41384172439575195,
+      "learning_rate": 3.092425753145931e-05,
+      "loss": 1.5997,
+      "step": 6013
+    },
+    {
+      "epoch": 0.6247013607562065,
+      "grad_norm": 0.4150152802467346,
+      "learning_rate": 3.0909176117296126e-05,
+      "loss": 1.6308,
+      "step": 6014
+    },
+    {
+      "epoch": 0.6248052352757868,
+      "grad_norm": 0.4042709767818451,
+      "learning_rate": 3.08940967361568e-05,
+      "loss": 1.6706,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6249091097953672,
+      "grad_norm": 0.4141731560230255,
+      "learning_rate": 3.08790193896471e-05,
+      "loss": 1.6422,
+      "step": 6016
+    },
+    {
+      "epoch": 0.6250129843149476,
+      "grad_norm": 0.40485629439353943,
+      "learning_rate": 3.0863944079372694e-05,
+      "loss": 1.6133,
+      "step": 6017
+    },
+    {
+      "epoch": 0.6251168588345279,
+      "grad_norm": 0.4418627917766571,
+      "learning_rate": 3.0848870806938995e-05,
+      "loss": 1.741,
+      "step": 6018
+    },
+    {
+      "epoch": 0.6252207333541082,
+      "grad_norm": 0.40159109234809875,
+      "learning_rate": 3.083379957395117e-05,
+      "loss": 1.618,
+      "step": 6019
+    },
+    {
+      "epoch": 0.6253246078736886,
+      "grad_norm": 0.4042385220527649,
+      "learning_rate": 3.081873038201418e-05,
+      "loss": 1.7609,
+      "step": 6020
+    },
+    {
+      "epoch": 0.625428482393269,
+      "grad_norm": 0.4162798523902893,
+      "learning_rate": 3.0803663232732775e-05,
+      "loss": 1.7222,
+      "step": 6021
+    },
+    {
+      "epoch": 0.6255323569128493,
+      "grad_norm": 0.3897026777267456,
+      "learning_rate": 3.078859812771152e-05,
+      "loss": 1.5955,
+      "step": 6022
+    },
+    {
+      "epoch": 0.6256362314324296,
+      "grad_norm": 0.4212419092655182,
+      "learning_rate": 3.0773535068554695e-05,
+      "loss": 1.6507,
+      "step": 6023
+    },
+    {
+      "epoch": 0.62574010595201,
+      "grad_norm": 0.4059465229511261,
+      "learning_rate": 3.075847405686641e-05,
+      "loss": 1.6731,
+      "step": 6024
+    },
+    {
+      "epoch": 0.6258439804715903,
+      "grad_norm": 0.4359121024608612,
+      "learning_rate": 3.074341509425056e-05,
+      "loss": 1.5718,
+      "step": 6025
+    },
+    {
+      "epoch": 0.6259478549911707,
+      "grad_norm": 0.4287969470024109,
+      "learning_rate": 3.07283581823108e-05,
+      "loss": 1.6525,
+      "step": 6026
+    },
+    {
+      "epoch": 0.626051729510751,
+      "grad_norm": 0.4397551715373993,
+      "learning_rate": 3.071330332265054e-05,
+      "loss": 1.6681,
+      "step": 6027
+    },
+    {
+      "epoch": 0.6261556040303313,
+      "grad_norm": 0.4228843152523041,
+      "learning_rate": 3.069825051687307e-05,
+      "loss": 1.68,
+      "step": 6028
+    },
+    {
+      "epoch": 0.6262594785499117,
+      "grad_norm": 0.4212433993816376,
+      "learning_rate": 3.068319976658134e-05,
+      "loss": 1.7587,
+      "step": 6029
+    },
+    {
+      "epoch": 0.6263633530694921,
+      "grad_norm": 0.41605517268180847,
+      "learning_rate": 3.066815107337815e-05,
+      "loss": 1.6564,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6264672275890724,
+      "grad_norm": 0.4286707639694214,
+      "learning_rate": 3.0653104438866104e-05,
+      "loss": 1.7046,
+      "step": 6031
+    },
+    {
+      "epoch": 0.6265711021086527,
+      "grad_norm": 0.4247523248195648,
+      "learning_rate": 3.063805986464751e-05,
+      "loss": 1.7082,
+      "step": 6032
+    },
+    {
+      "epoch": 0.6266749766282331,
+      "grad_norm": 0.4271533489227295,
+      "learning_rate": 3.06230173523245e-05,
+      "loss": 1.7456,
+      "step": 6033
+    },
+    {
+      "epoch": 0.6267788511478134,
+      "grad_norm": 0.43358635902404785,
+      "learning_rate": 3.060797690349902e-05,
+      "loss": 1.5457,
+      "step": 6034
+    },
+    {
+      "epoch": 0.6268827256673938,
+      "grad_norm": 0.3621548116207123,
+      "learning_rate": 3.0592938519772715e-05,
+      "loss": 1.3377,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6269866001869742,
+      "grad_norm": 0.4215681850910187,
+      "learning_rate": 3.057790220274707e-05,
+      "loss": 1.5863,
+      "step": 6036
+    },
+    {
+      "epoch": 0.6270904747065544,
+      "grad_norm": 0.425849586725235,
+      "learning_rate": 3.056286795402337e-05,
+      "loss": 1.6182,
+      "step": 6037
+    },
+    {
+      "epoch": 0.6271943492261348,
+      "grad_norm": 0.4131641089916229,
+      "learning_rate": 3.0547835775202597e-05,
+      "loss": 1.6674,
+      "step": 6038
+    },
+    {
+      "epoch": 0.6272982237457152,
+      "grad_norm": 0.41090524196624756,
+      "learning_rate": 3.053280566788559e-05,
+      "loss": 1.5321,
+      "step": 6039
+    },
+    {
+      "epoch": 0.6274020982652955,
+      "grad_norm": 0.443823903799057,
+      "learning_rate": 3.051777763367293e-05,
+      "loss": 1.5838,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6275059727848759,
+      "grad_norm": 0.41161811351776123,
+      "learning_rate": 3.0502751674164976e-05,
+      "loss": 1.6502,
+      "step": 6041
+    },
+    {
+      "epoch": 0.6276098473044562,
+      "grad_norm": 0.4168041944503784,
+      "learning_rate": 3.0487727790961884e-05,
+      "loss": 1.6534,
+      "step": 6042
+    },
+    {
+      "epoch": 0.6277137218240366,
+      "grad_norm": 0.43634623289108276,
+      "learning_rate": 3.0472705985663575e-05,
+      "loss": 1.6755,
+      "step": 6043
+    },
+    {
+      "epoch": 0.6278175963436169,
+      "grad_norm": 0.45974037051200867,
+      "learning_rate": 3.045768625986979e-05,
+      "loss": 1.8234,
+      "step": 6044
+    },
+    {
+      "epoch": 0.6279214708631973,
+      "grad_norm": 0.4236984848976135,
+      "learning_rate": 3.0442668615179953e-05,
+      "loss": 1.794,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6280253453827777,
+      "grad_norm": 0.3920854330062866,
+      "learning_rate": 3.0427653053193366e-05,
+      "loss": 1.6738,
+      "step": 6046
+    },
+    {
+      "epoch": 0.6281292199023579,
+      "grad_norm": 0.44960126280784607,
+      "learning_rate": 3.0412639575509073e-05,
+      "loss": 1.6268,
+      "step": 6047
+    },
+    {
+      "epoch": 0.6282330944219383,
+      "grad_norm": 0.4008250832557678,
+      "learning_rate": 3.0397628183725856e-05,
+      "loss": 1.4388,
+      "step": 6048
+    },
+    {
+      "epoch": 0.6283369689415187,
+      "grad_norm": 0.39700058102607727,
+      "learning_rate": 3.038261887944233e-05,
+      "loss": 1.613,
+      "step": 6049
+    },
+    {
+      "epoch": 0.628440843461099,
+      "grad_norm": 0.46325239539146423,
+      "learning_rate": 3.0367611664256895e-05,
+      "loss": 1.8034,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6285447179806793,
+      "grad_norm": 0.434199720621109,
+      "learning_rate": 3.0352606539767657e-05,
+      "loss": 1.669,
+      "step": 6051
+    },
+    {
+      "epoch": 0.6286485925002597,
+      "grad_norm": 0.3849222958087921,
+      "learning_rate": 3.033760350757258e-05,
+      "loss": 1.5992,
+      "step": 6052
+    },
+    {
+      "epoch": 0.62875246701984,
+      "grad_norm": 0.4169051945209503,
+      "learning_rate": 3.0322602569269355e-05,
+      "loss": 1.5102,
+      "step": 6053
+    },
+    {
+      "epoch": 0.6288563415394204,
+      "grad_norm": 0.42412588000297546,
+      "learning_rate": 3.030760372645547e-05,
+      "loss": 1.5752,
+      "step": 6054
+    },
+    {
+      "epoch": 0.6289602160590008,
+      "grad_norm": 0.4000975787639618,
+      "learning_rate": 3.0292606980728165e-05,
+      "loss": 1.715,
+      "step": 6055
+    },
+    {
+      "epoch": 0.629064090578581,
+      "grad_norm": 0.3950633704662323,
+      "learning_rate": 3.0277612333684526e-05,
+      "loss": 1.6543,
+      "step": 6056
+    },
+    {
+      "epoch": 0.6291679650981614,
+      "grad_norm": 0.4239039421081543,
+      "learning_rate": 3.02626197869213e-05,
+      "loss": 1.7277,
+      "step": 6057
+    },
+    {
+      "epoch": 0.6292718396177418,
+      "grad_norm": 0.44476318359375,
+      "learning_rate": 3.024762934203511e-05,
+      "loss": 1.8181,
+      "step": 6058
+    },
+    {
+      "epoch": 0.6293757141373221,
+      "grad_norm": 0.3775947093963623,
+      "learning_rate": 3.023264100062234e-05,
+      "loss": 1.5467,
+      "step": 6059
+    },
+    {
+      "epoch": 0.6294795886569025,
+      "grad_norm": 0.4439935088157654,
+      "learning_rate": 3.0217654764279114e-05,
+      "loss": 1.9913,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6295834631764828,
+      "grad_norm": 0.4108879566192627,
+      "learning_rate": 3.0202670634601325e-05,
+      "loss": 1.6593,
+      "step": 6061
+    },
+    {
+      "epoch": 0.6296873376960631,
+      "grad_norm": 0.4360349476337433,
+      "learning_rate": 3.0187688613184717e-05,
+      "loss": 1.4867,
+      "step": 6062
+    },
+    {
+      "epoch": 0.6297912122156435,
+      "grad_norm": 0.4250946044921875,
+      "learning_rate": 3.0172708701624704e-05,
+      "loss": 1.6254,
+      "step": 6063
+    },
+    {
+      "epoch": 0.6298950867352239,
+      "grad_norm": 0.44005441665649414,
+      "learning_rate": 3.0157730901516556e-05,
+      "loss": 1.7055,
+      "step": 6064
+    },
+    {
+      "epoch": 0.6299989612548041,
+      "grad_norm": 0.40063902735710144,
+      "learning_rate": 3.014275521445531e-05,
+      "loss": 1.4714,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6301028357743845,
+      "grad_norm": 0.3904891610145569,
+      "learning_rate": 3.0127781642035725e-05,
+      "loss": 1.5278,
+      "step": 6066
+    },
+    {
+      "epoch": 0.6302067102939649,
+      "grad_norm": 0.36918964982032776,
+      "learning_rate": 3.0112810185852393e-05,
+      "loss": 1.4645,
+      "step": 6067
+    },
+    {
+      "epoch": 0.6303105848135452,
+      "grad_norm": 0.43034636974334717,
+      "learning_rate": 3.009784084749966e-05,
+      "loss": 1.6989,
+      "step": 6068
+    },
+    {
+      "epoch": 0.6304144593331256,
+      "grad_norm": 0.43391627073287964,
+      "learning_rate": 3.0082873628571616e-05,
+      "loss": 1.7468,
+      "step": 6069
+    },
+    {
+      "epoch": 0.630518333852706,
+      "grad_norm": 0.47783637046813965,
+      "learning_rate": 3.006790853066217e-05,
+      "loss": 1.7116,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6306222083722863,
+      "grad_norm": 0.46756497025489807,
+      "learning_rate": 3.0052945555364987e-05,
+      "loss": 1.8164,
+      "step": 6071
+    },
+    {
+      "epoch": 0.6307260828918666,
+      "grad_norm": 0.4332909882068634,
+      "learning_rate": 3.0037984704273536e-05,
+      "loss": 1.7011,
+      "step": 6072
+    },
+    {
+      "epoch": 0.630829957411447,
+      "grad_norm": 0.46313124895095825,
+      "learning_rate": 3.0023025978980977e-05,
+      "loss": 1.8097,
+      "step": 6073
+    },
+    {
+      "epoch": 0.6309338319310274,
+      "grad_norm": 0.4317631125450134,
+      "learning_rate": 3.000806938108034e-05,
+      "loss": 1.7535,
+      "step": 6074
+    },
+    {
+      "epoch": 0.6310377064506076,
+      "grad_norm": 0.39492112398147583,
+      "learning_rate": 2.9993114912164378e-05,
+      "loss": 1.6294,
+      "step": 6075
+    },
+    {
+      "epoch": 0.631141580970188,
+      "grad_norm": 0.3940942883491516,
+      "learning_rate": 2.9978162573825597e-05,
+      "loss": 1.7187,
+      "step": 6076
+    },
+    {
+      "epoch": 0.6312454554897684,
+      "grad_norm": 0.4159606695175171,
+      "learning_rate": 2.996321236765633e-05,
+      "loss": 1.7397,
+      "step": 6077
+    },
+    {
+      "epoch": 0.6313493300093487,
+      "grad_norm": 0.38286933302879333,
+      "learning_rate": 2.994826429524867e-05,
+      "loss": 1.5252,
+      "step": 6078
+    },
+    {
+      "epoch": 0.6314532045289291,
+      "grad_norm": 0.424221396446228,
+      "learning_rate": 2.9933318358194427e-05,
+      "loss": 1.4488,
+      "step": 6079
+    },
+    {
+      "epoch": 0.6315570790485094,
+      "grad_norm": 0.4659724831581116,
+      "learning_rate": 2.9918374558085272e-05,
+      "loss": 1.705,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6316609535680897,
+      "grad_norm": 0.42719754576683044,
+      "learning_rate": 2.9903432896512583e-05,
+      "loss": 1.5949,
+      "step": 6081
+    },
+    {
+      "epoch": 0.6317648280876701,
+      "grad_norm": 0.4064619243144989,
+      "learning_rate": 2.988849337506753e-05,
+      "loss": 1.5759,
+      "step": 6082
+    },
+    {
+      "epoch": 0.6318687026072505,
+      "grad_norm": 0.4120051860809326,
+      "learning_rate": 2.9873555995341045e-05,
+      "loss": 1.6129,
+      "step": 6083
+    },
+    {
+      "epoch": 0.6319725771268307,
+      "grad_norm": 0.41194257140159607,
+      "learning_rate": 2.9858620758923872e-05,
+      "loss": 1.6637,
+      "step": 6084
+    },
+    {
+      "epoch": 0.6320764516464111,
+      "grad_norm": 0.4339029788970947,
+      "learning_rate": 2.9843687667406466e-05,
+      "loss": 1.7608,
+      "step": 6085
+    },
+    {
+      "epoch": 0.6321803261659915,
+      "grad_norm": 0.4774200916290283,
+      "learning_rate": 2.982875672237909e-05,
+      "loss": 1.5836,
+      "step": 6086
+    },
+    {
+      "epoch": 0.6322842006855718,
+      "grad_norm": 0.40862688422203064,
+      "learning_rate": 2.9813827925431813e-05,
+      "loss": 1.6473,
+      "step": 6087
+    },
+    {
+      "epoch": 0.6323880752051522,
+      "grad_norm": 0.42833518981933594,
+      "learning_rate": 2.9798901278154384e-05,
+      "loss": 1.6633,
+      "step": 6088
+    },
+    {
+      "epoch": 0.6324919497247325,
+      "grad_norm": 0.42553073167800903,
+      "learning_rate": 2.9783976782136385e-05,
+      "loss": 1.6624,
+      "step": 6089
+    },
+    {
+      "epoch": 0.6325958242443128,
+      "grad_norm": 0.3767523169517517,
+      "learning_rate": 2.976905443896719e-05,
+      "loss": 1.516,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6326996987638932,
+      "grad_norm": 0.44192999601364136,
+      "learning_rate": 2.975413425023587e-05,
+      "loss": 1.8359,
+      "step": 6091
+    },
+    {
+      "epoch": 0.6328035732834736,
+      "grad_norm": 0.4179973304271698,
+      "learning_rate": 2.973921621753133e-05,
+      "loss": 1.5251,
+      "step": 6092
+    },
+    {
+      "epoch": 0.6329074478030539,
+      "grad_norm": 0.40900513529777527,
+      "learning_rate": 2.972430034244224e-05,
+      "loss": 1.7986,
+      "step": 6093
+    },
+    {
+      "epoch": 0.6330113223226342,
+      "grad_norm": 0.42563167214393616,
+      "learning_rate": 2.970938662655699e-05,
+      "loss": 1.6558,
+      "step": 6094
+    },
+    {
+      "epoch": 0.6331151968422146,
+      "grad_norm": 0.4261634349822998,
+      "learning_rate": 2.9694475071463797e-05,
+      "loss": 1.6296,
+      "step": 6095
+    },
+    {
+      "epoch": 0.633219071361795,
+      "grad_norm": 0.41434627771377563,
+      "learning_rate": 2.9679565678750608e-05,
+      "loss": 1.6559,
+      "step": 6096
+    },
+    {
+      "epoch": 0.6333229458813753,
+      "grad_norm": 0.4356957972049713,
+      "learning_rate": 2.9664658450005195e-05,
+      "loss": 1.6724,
+      "step": 6097
+    },
+    {
+      "epoch": 0.6334268204009557,
+      "grad_norm": 0.4227178692817688,
+      "learning_rate": 2.9649753386815017e-05,
+      "loss": 1.7473,
+      "step": 6098
+    },
+    {
+      "epoch": 0.633530694920536,
+      "grad_norm": 0.45652738213539124,
+      "learning_rate": 2.963485049076736e-05,
+      "loss": 1.7247,
+      "step": 6099
+    },
+    {
+      "epoch": 0.6336345694401163,
+      "grad_norm": 0.4394940435886383,
+      "learning_rate": 2.961994976344929e-05,
+      "loss": 1.7161,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6337384439596967,
+      "grad_norm": 0.4043090343475342,
+      "learning_rate": 2.9605051206447582e-05,
+      "loss": 1.6737,
+      "step": 6101
+    },
+    {
+      "epoch": 0.6338423184792771,
+      "grad_norm": 0.4564562141895294,
+      "learning_rate": 2.959015482134885e-05,
+      "loss": 1.8074,
+      "step": 6102
+    },
+    {
+      "epoch": 0.6339461929988573,
+      "grad_norm": 0.4548121392726898,
+      "learning_rate": 2.9575260609739423e-05,
+      "loss": 1.7882,
+      "step": 6103
+    },
+    {
+      "epoch": 0.6340500675184377,
+      "grad_norm": 0.41030994057655334,
+      "learning_rate": 2.956036857320541e-05,
+      "loss": 1.6736,
+      "step": 6104
+    },
+    {
+      "epoch": 0.6341539420380181,
+      "grad_norm": 0.4318486154079437,
+      "learning_rate": 2.9545478713332707e-05,
+      "loss": 1.7257,
+      "step": 6105
+    },
+    {
+      "epoch": 0.6342578165575984,
+      "grad_norm": 0.4724993109703064,
+      "learning_rate": 2.9530591031706988e-05,
+      "loss": 1.6384,
+      "step": 6106
+    },
+    {
+      "epoch": 0.6343616910771788,
+      "grad_norm": 0.40737417340278625,
+      "learning_rate": 2.9515705529913644e-05,
+      "loss": 1.7593,
+      "step": 6107
+    },
+    {
+      "epoch": 0.6344655655967592,
+      "grad_norm": 0.4164734482765198,
+      "learning_rate": 2.9500822209537883e-05,
+      "loss": 1.7383,
+      "step": 6108
+    },
+    {
+      "epoch": 0.6345694401163394,
+      "grad_norm": 0.40095841884613037,
+      "learning_rate": 2.9485941072164656e-05,
+      "loss": 1.6363,
+      "step": 6109
+    },
+    {
+      "epoch": 0.6346733146359198,
+      "grad_norm": 0.4260212481021881,
+      "learning_rate": 2.9471062119378688e-05,
+      "loss": 1.7249,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6347771891555002,
+      "grad_norm": 0.4054844081401825,
+      "learning_rate": 2.9456185352764465e-05,
+      "loss": 1.4506,
+      "step": 6111
+    },
+    {
+      "epoch": 0.6348810636750805,
+      "grad_norm": 0.39739441871643066,
+      "learning_rate": 2.944131077390627e-05,
+      "loss": 1.6051,
+      "step": 6112
+    },
+    {
+      "epoch": 0.6349849381946608,
+      "grad_norm": 0.4330110549926758,
+      "learning_rate": 2.94264383843881e-05,
+      "loss": 1.7506,
+      "step": 6113
+    },
+    {
+      "epoch": 0.6350888127142412,
+      "grad_norm": 0.4423455595970154,
+      "learning_rate": 2.9411568185793757e-05,
+      "loss": 1.7382,
+      "step": 6114
+    },
+    {
+      "epoch": 0.6351926872338215,
+      "grad_norm": 0.41518527269363403,
+      "learning_rate": 2.939670017970683e-05,
+      "loss": 1.4555,
+      "step": 6115
+    },
+    {
+      "epoch": 0.6352965617534019,
+      "grad_norm": 0.40972813963890076,
+      "learning_rate": 2.9381834367710616e-05,
+      "loss": 1.7941,
+      "step": 6116
+    },
+    {
+      "epoch": 0.6354004362729823,
+      "grad_norm": 0.43250906467437744,
+      "learning_rate": 2.9366970751388196e-05,
+      "loss": 1.7317,
+      "step": 6117
+    },
+    {
+      "epoch": 0.6355043107925625,
+      "grad_norm": 0.4202265441417694,
+      "learning_rate": 2.9352109332322468e-05,
+      "loss": 1.652,
+      "step": 6118
+    },
+    {
+      "epoch": 0.6356081853121429,
+      "grad_norm": 0.40753960609436035,
+      "learning_rate": 2.9337250112096016e-05,
+      "loss": 1.6144,
+      "step": 6119
+    },
+    {
+      "epoch": 0.6357120598317233,
+      "grad_norm": 0.39751505851745605,
+      "learning_rate": 2.9322393092291256e-05,
+      "loss": 1.7013,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6358159343513037,
+      "grad_norm": 0.4368493854999542,
+      "learning_rate": 2.9307538274490348e-05,
+      "loss": 1.7454,
+      "step": 6121
+    },
+    {
+      "epoch": 0.635919808870884,
+      "grad_norm": 0.40719810128211975,
+      "learning_rate": 2.9292685660275204e-05,
+      "loss": 1.7023,
+      "step": 6122
+    },
+    {
+      "epoch": 0.6360236833904643,
+      "grad_norm": 0.4638088047504425,
+      "learning_rate": 2.9277835251227514e-05,
+      "loss": 1.7887,
+      "step": 6123
+    },
+    {
+      "epoch": 0.6361275579100447,
+      "grad_norm": 0.454287588596344,
+      "learning_rate": 2.9262987048928725e-05,
+      "loss": 1.8606,
+      "step": 6124
+    },
+    {
+      "epoch": 0.636231432429625,
+      "grad_norm": 0.3964051902294159,
+      "learning_rate": 2.9248141054960076e-05,
+      "loss": 1.5617,
+      "step": 6125
+    },
+    {
+      "epoch": 0.6363353069492054,
+      "grad_norm": 0.39127108454704285,
+      "learning_rate": 2.9233297270902514e-05,
+      "loss": 1.5637,
+      "step": 6126
+    },
+    {
+      "epoch": 0.6364391814687858,
+      "grad_norm": 0.4058896005153656,
+      "learning_rate": 2.921845569833681e-05,
+      "loss": 1.6716,
+      "step": 6127
+    },
+    {
+      "epoch": 0.636543055988366,
+      "grad_norm": 0.41195571422576904,
+      "learning_rate": 2.9203616338843486e-05,
+      "loss": 1.5734,
+      "step": 6128
+    },
+    {
+      "epoch": 0.6366469305079464,
+      "grad_norm": 0.45315390825271606,
+      "learning_rate": 2.9188779194002803e-05,
+      "loss": 1.6421,
+      "step": 6129
+    },
+    {
+      "epoch": 0.6367508050275268,
+      "grad_norm": 0.4393003284931183,
+      "learning_rate": 2.917394426539477e-05,
+      "loss": 1.777,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6368546795471071,
+      "grad_norm": 0.4733230769634247,
+      "learning_rate": 2.915911155459927e-05,
+      "loss": 1.7351,
+      "step": 6131
+    },
+    {
+      "epoch": 0.6369585540666874,
+      "grad_norm": 0.44397008419036865,
+      "learning_rate": 2.914428106319579e-05,
+      "loss": 1.7189,
+      "step": 6132
+    },
+    {
+      "epoch": 0.6370624285862678,
+      "grad_norm": 0.4575912654399872,
+      "learning_rate": 2.912945279276371e-05,
+      "loss": 2.0055,
+      "step": 6133
+    },
+    {
+      "epoch": 0.6371663031058481,
+      "grad_norm": 0.43883270025253296,
+      "learning_rate": 2.9114626744882128e-05,
+      "loss": 1.8263,
+      "step": 6134
+    },
+    {
+      "epoch": 0.6372701776254285,
+      "grad_norm": 0.4244493842124939,
+      "learning_rate": 2.909980292112985e-05,
+      "loss": 1.7515,
+      "step": 6135
+    },
+    {
+      "epoch": 0.6373740521450089,
+      "grad_norm": 0.45086610317230225,
+      "learning_rate": 2.9084981323085547e-05,
+      "loss": 1.7123,
+      "step": 6136
+    },
+    {
+      "epoch": 0.6374779266645891,
+      "grad_norm": 0.43869614601135254,
+      "learning_rate": 2.90701619523276e-05,
+      "loss": 1.6792,
+      "step": 6137
+    },
+    {
+      "epoch": 0.6375818011841695,
+      "grad_norm": 0.43083521723747253,
+      "learning_rate": 2.9055344810434137e-05,
+      "loss": 1.6872,
+      "step": 6138
+    },
+    {
+      "epoch": 0.6376856757037499,
+      "grad_norm": 0.5012997984886169,
+      "learning_rate": 2.904052989898308e-05,
+      "loss": 1.908,
+      "step": 6139
+    },
+    {
+      "epoch": 0.6377895502233302,
+      "grad_norm": 0.4534330368041992,
+      "learning_rate": 2.90257172195521e-05,
+      "loss": 1.6729,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6378934247429106,
+      "grad_norm": 0.43945181369781494,
+      "learning_rate": 2.9010906773718627e-05,
+      "loss": 1.6487,
+      "step": 6141
+    },
+    {
+      "epoch": 0.6379972992624909,
+      "grad_norm": 0.40094152092933655,
+      "learning_rate": 2.899609856305985e-05,
+      "loss": 1.508,
+      "step": 6142
+    },
+    {
+      "epoch": 0.6381011737820712,
+      "grad_norm": 0.4558815658092499,
+      "learning_rate": 2.8981292589152768e-05,
+      "loss": 1.8111,
+      "step": 6143
+    },
+    {
+      "epoch": 0.6382050483016516,
+      "grad_norm": 0.43613293766975403,
+      "learning_rate": 2.8966488853574036e-05,
+      "loss": 1.6826,
+      "step": 6144
+    },
+    {
+      "epoch": 0.638308922821232,
+      "grad_norm": 0.43343204259872437,
+      "learning_rate": 2.8951687357900203e-05,
+      "loss": 1.7693,
+      "step": 6145
+    },
+    {
+      "epoch": 0.6384127973408122,
+      "grad_norm": 0.45726385712623596,
+      "learning_rate": 2.8936888103707473e-05,
+      "loss": 1.883,
+      "step": 6146
+    },
+    {
+      "epoch": 0.6385166718603926,
+      "grad_norm": 0.3765776455402374,
+      "learning_rate": 2.8922091092571862e-05,
+      "loss": 1.5675,
+      "step": 6147
+    },
+    {
+      "epoch": 0.638620546379973,
+      "grad_norm": 0.4046104848384857,
+      "learning_rate": 2.8907296326069137e-05,
+      "loss": 1.7282,
+      "step": 6148
+    },
+    {
+      "epoch": 0.6387244208995534,
+      "grad_norm": 0.40840303897857666,
+      "learning_rate": 2.8892503805774806e-05,
+      "loss": 1.6623,
+      "step": 6149
+    },
+    {
+      "epoch": 0.6388282954191337,
+      "grad_norm": 0.4642704427242279,
+      "learning_rate": 2.8877713533264217e-05,
+      "loss": 1.6275,
+      "step": 6150
+    },
+    {
+      "epoch": 0.638932169938714,
+      "grad_norm": 0.41992318630218506,
+      "learning_rate": 2.8862925510112327e-05,
+      "loss": 1.6549,
+      "step": 6151
+    },
+    {
+      "epoch": 0.6390360444582944,
+      "grad_norm": 0.4473896324634552,
+      "learning_rate": 2.884813973789402e-05,
+      "loss": 1.7372,
+      "step": 6152
+    },
+    {
+      "epoch": 0.6391399189778747,
+      "grad_norm": 0.39671966433525085,
+      "learning_rate": 2.8833356218183843e-05,
+      "loss": 1.5238,
+      "step": 6153
+    },
+    {
+      "epoch": 0.6392437934974551,
+      "grad_norm": 0.4355907440185547,
+      "learning_rate": 2.881857495255612e-05,
+      "loss": 1.8191,
+      "step": 6154
+    },
+    {
+      "epoch": 0.6393476680170355,
+      "grad_norm": 0.4201485812664032,
+      "learning_rate": 2.8803795942584945e-05,
+      "loss": 1.4469,
+      "step": 6155
+    },
+    {
+      "epoch": 0.6394515425366157,
+      "grad_norm": 0.40149399638175964,
+      "learning_rate": 2.8789019189844163e-05,
+      "loss": 1.5771,
+      "step": 6156
+    },
+    {
+      "epoch": 0.6395554170561961,
+      "grad_norm": 0.40967410802841187,
+      "learning_rate": 2.877424469590739e-05,
+      "loss": 1.6283,
+      "step": 6157
+    },
+    {
+      "epoch": 0.6396592915757765,
+      "grad_norm": 0.40948042273521423,
+      "learning_rate": 2.8759472462347965e-05,
+      "loss": 1.6039,
+      "step": 6158
+    },
+    {
+      "epoch": 0.6397631660953568,
+      "grad_norm": 0.4414249658584595,
+      "learning_rate": 2.8744702490739077e-05,
+      "loss": 1.6585,
+      "step": 6159
+    },
+    {
+      "epoch": 0.6398670406149372,
+      "grad_norm": 0.39458298683166504,
+      "learning_rate": 2.8729934782653545e-05,
+      "loss": 1.6049,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6399709151345175,
+      "grad_norm": 0.4163827896118164,
+      "learning_rate": 2.8715169339664062e-05,
+      "loss": 1.4499,
+      "step": 6161
+    },
+    {
+      "epoch": 0.6400747896540978,
+      "grad_norm": 0.4390014708042145,
+      "learning_rate": 2.870040616334303e-05,
+      "loss": 1.7916,
+      "step": 6162
+    },
+    {
+      "epoch": 0.6401786641736782,
+      "grad_norm": 0.38799330592155457,
+      "learning_rate": 2.8685645255262562e-05,
+      "loss": 1.6107,
+      "step": 6163
+    },
+    {
+      "epoch": 0.6402825386932586,
+      "grad_norm": 0.4088546633720398,
+      "learning_rate": 2.867088661699464e-05,
+      "loss": 1.7303,
+      "step": 6164
+    },
+    {
+      "epoch": 0.6403864132128388,
+      "grad_norm": 0.39084142446517944,
+      "learning_rate": 2.8656130250110913e-05,
+      "loss": 1.5763,
+      "step": 6165
+    },
+    {
+      "epoch": 0.6404902877324192,
+      "grad_norm": 0.4360373318195343,
+      "learning_rate": 2.864137615618283e-05,
+      "loss": 1.6337,
+      "step": 6166
+    },
+    {
+      "epoch": 0.6405941622519996,
+      "grad_norm": 0.39488184452056885,
+      "learning_rate": 2.862662433678158e-05,
+      "loss": 1.5719,
+      "step": 6167
+    },
+    {
+      "epoch": 0.6406980367715799,
+      "grad_norm": 0.402828186750412,
+      "learning_rate": 2.861187479347813e-05,
+      "loss": 1.4238,
+      "step": 6168
+    },
+    {
+      "epoch": 0.6408019112911603,
+      "grad_norm": 0.48173683881759644,
+      "learning_rate": 2.8597127527843172e-05,
+      "loss": 1.7344,
+      "step": 6169
+    },
+    {
+      "epoch": 0.6409057858107406,
+      "grad_norm": 0.4502374529838562,
+      "learning_rate": 2.858238254144716e-05,
+      "loss": 1.8152,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6410096603303209,
+      "grad_norm": 0.41789987683296204,
+      "learning_rate": 2.8567639835860394e-05,
+      "loss": 1.612,
+      "step": 6171
+    },
+    {
+      "epoch": 0.6411135348499013,
+      "grad_norm": 0.4217549264431,
+      "learning_rate": 2.8552899412652757e-05,
+      "loss": 1.7688,
+      "step": 6172
+    },
+    {
+      "epoch": 0.6412174093694817,
+      "grad_norm": 0.41339951753616333,
+      "learning_rate": 2.853816127339406e-05,
+      "loss": 1.779,
+      "step": 6173
+    },
+    {
+      "epoch": 0.6413212838890621,
+      "grad_norm": 0.4295401871204376,
+      "learning_rate": 2.8523425419653784e-05,
+      "loss": 1.723,
+      "step": 6174
+    },
+    {
+      "epoch": 0.6414251584086423,
+      "grad_norm": 0.44703397154808044,
+      "learning_rate": 2.8508691853001174e-05,
+      "loss": 1.74,
+      "step": 6175
+    },
+    {
+      "epoch": 0.6415290329282227,
+      "grad_norm": 0.4364173710346222,
+      "learning_rate": 2.8493960575005242e-05,
+      "loss": 1.8088,
+      "step": 6176
+    },
+    {
+      "epoch": 0.6416329074478031,
+      "grad_norm": 0.41096925735473633,
+      "learning_rate": 2.8479231587234734e-05,
+      "loss": 1.6215,
+      "step": 6177
+    },
+    {
+      "epoch": 0.6417367819673834,
+      "grad_norm": 0.42458489537239075,
+      "learning_rate": 2.8464504891258237e-05,
+      "loss": 1.6694,
+      "step": 6178
+    },
+    {
+      "epoch": 0.6418406564869638,
+      "grad_norm": 0.4292949438095093,
+      "learning_rate": 2.8449780488643944e-05,
+      "loss": 1.8489,
+      "step": 6179
+    },
+    {
+      "epoch": 0.6419445310065441,
+      "grad_norm": 0.4265950918197632,
+      "learning_rate": 2.8435058380959956e-05,
+      "loss": 1.691,
+      "step": 6180
+    },
+    {
+      "epoch": 0.6420484055261244,
+      "grad_norm": 0.44009000062942505,
+      "learning_rate": 2.842033856977403e-05,
+      "loss": 1.7029,
+      "step": 6181
+    },
+    {
+      "epoch": 0.6421522800457048,
+      "grad_norm": 0.46439558267593384,
+      "learning_rate": 2.840562105665372e-05,
+      "loss": 1.7289,
+      "step": 6182
+    },
+    {
+      "epoch": 0.6422561545652852,
+      "grad_norm": 0.4735531210899353,
+      "learning_rate": 2.8390905843166326e-05,
+      "loss": 1.8544,
+      "step": 6183
+    },
+    {
+      "epoch": 0.6423600290848654,
+      "grad_norm": 0.4040323793888092,
+      "learning_rate": 2.8376192930878898e-05,
+      "loss": 1.6042,
+      "step": 6184
+    },
+    {
+      "epoch": 0.6424639036044458,
+      "grad_norm": 0.399845689535141,
+      "learning_rate": 2.8361482321358256e-05,
+      "loss": 1.7571,
+      "step": 6185
+    },
+    {
+      "epoch": 0.6425677781240262,
+      "grad_norm": 0.4226529002189636,
+      "learning_rate": 2.834677401617094e-05,
+      "loss": 1.7529,
+      "step": 6186
+    },
+    {
+      "epoch": 0.6426716526436065,
+      "grad_norm": 0.45261451601982117,
+      "learning_rate": 2.833206801688333e-05,
+      "loss": 1.7456,
+      "step": 6187
+    },
+    {
+      "epoch": 0.6427755271631869,
+      "grad_norm": 0.4298214912414551,
+      "learning_rate": 2.8317364325061412e-05,
+      "loss": 1.5975,
+      "step": 6188
+    },
+    {
+      "epoch": 0.6428794016827672,
+      "grad_norm": 0.3926003575325012,
+      "learning_rate": 2.8302662942271098e-05,
+      "loss": 1.6589,
+      "step": 6189
+    },
+    {
+      "epoch": 0.6429832762023475,
+      "grad_norm": 0.43051791191101074,
+      "learning_rate": 2.828796387007794e-05,
+      "loss": 1.7522,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6430871507219279,
+      "grad_norm": 0.3769807815551758,
+      "learning_rate": 2.8273267110047234e-05,
+      "loss": 1.4194,
+      "step": 6191
+    },
+    {
+      "epoch": 0.6431910252415083,
+      "grad_norm": 0.4164545238018036,
+      "learning_rate": 2.8258572663744132e-05,
+      "loss": 1.6241,
+      "step": 6192
+    },
+    {
+      "epoch": 0.6432948997610886,
+      "grad_norm": 0.41669386625289917,
+      "learning_rate": 2.824388053273344e-05,
+      "loss": 1.6012,
+      "step": 6193
+    },
+    {
+      "epoch": 0.6433987742806689,
+      "grad_norm": 0.3818405568599701,
+      "learning_rate": 2.8229190718579777e-05,
+      "loss": 1.5466,
+      "step": 6194
+    },
+    {
+      "epoch": 0.6435026488002493,
+      "grad_norm": 0.3918877840042114,
+      "learning_rate": 2.8214503222847477e-05,
+      "loss": 1.6249,
+      "step": 6195
+    },
+    {
+      "epoch": 0.6436065233198296,
+      "grad_norm": 0.45735591650009155,
+      "learning_rate": 2.8199818047100656e-05,
+      "loss": 1.8056,
+      "step": 6196
+    },
+    {
+      "epoch": 0.64371039783941,
+      "grad_norm": 0.45919516682624817,
+      "learning_rate": 2.8185135192903163e-05,
+      "loss": 1.9505,
+      "step": 6197
+    },
+    {
+      "epoch": 0.6438142723589904,
+      "grad_norm": 0.38972949981689453,
+      "learning_rate": 2.8170454661818583e-05,
+      "loss": 1.4729,
+      "step": 6198
+    },
+    {
+      "epoch": 0.6439181468785707,
+      "grad_norm": 0.479105681180954,
+      "learning_rate": 2.815577645541033e-05,
+      "loss": 1.8712,
+      "step": 6199
+    },
+    {
+      "epoch": 0.644022021398151,
+      "grad_norm": 0.38174498081207275,
+      "learning_rate": 2.8141100575241498e-05,
+      "loss": 1.5689,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6441258959177314,
+      "grad_norm": 0.40346476435661316,
+      "learning_rate": 2.8126427022874948e-05,
+      "loss": 1.7155,
+      "step": 6201
+    },
+    {
+      "epoch": 0.6442297704373118,
+      "grad_norm": 0.5447202324867249,
+      "learning_rate": 2.811175579987329e-05,
+      "loss": 1.9476,
+      "step": 6202
+    },
+    {
+      "epoch": 0.644333644956892,
+      "grad_norm": 0.4525381326675415,
+      "learning_rate": 2.809708690779891e-05,
+      "loss": 1.6353,
+      "step": 6203
+    },
+    {
+      "epoch": 0.6444375194764724,
+      "grad_norm": 0.4123556911945343,
+      "learning_rate": 2.8082420348213922e-05,
+      "loss": 1.5309,
+      "step": 6204
+    },
+    {
+      "epoch": 0.6445413939960528,
+      "grad_norm": 0.40919628739356995,
+      "learning_rate": 2.806775612268019e-05,
+      "loss": 1.5944,
+      "step": 6205
+    },
+    {
+      "epoch": 0.6446452685156331,
+      "grad_norm": 0.4005577266216278,
+      "learning_rate": 2.8053094232759393e-05,
+      "loss": 1.5079,
+      "step": 6206
+    },
+    {
+      "epoch": 0.6447491430352135,
+      "grad_norm": 0.41373926401138306,
+      "learning_rate": 2.8038434680012814e-05,
+      "loss": 1.7176,
+      "step": 6207
+    },
+    {
+      "epoch": 0.6448530175547939,
+      "grad_norm": 0.4339597821235657,
+      "learning_rate": 2.8023777466001666e-05,
+      "loss": 1.7649,
+      "step": 6208
+    },
+    {
+      "epoch": 0.6449568920743741,
+      "grad_norm": 0.40516406297683716,
+      "learning_rate": 2.8009122592286795e-05,
+      "loss": 1.6787,
+      "step": 6209
+    },
+    {
+      "epoch": 0.6450607665939545,
+      "grad_norm": 0.435835599899292,
+      "learning_rate": 2.7994470060428834e-05,
+      "loss": 1.626,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6451646411135349,
+      "grad_norm": 0.44071394205093384,
+      "learning_rate": 2.7979819871988166e-05,
+      "loss": 1.6221,
+      "step": 6211
+    },
+    {
+      "epoch": 0.6452685156331152,
+      "grad_norm": 0.40906620025634766,
+      "learning_rate": 2.796517202852491e-05,
+      "loss": 1.6874,
+      "step": 6212
+    },
+    {
+      "epoch": 0.6453723901526955,
+      "grad_norm": 0.4259955585002899,
+      "learning_rate": 2.795052653159896e-05,
+      "loss": 1.6113,
+      "step": 6213
+    },
+    {
+      "epoch": 0.6454762646722759,
+      "grad_norm": 0.41644105315208435,
+      "learning_rate": 2.7935883382769918e-05,
+      "loss": 1.5794,
+      "step": 6214
+    },
+    {
+      "epoch": 0.6455801391918562,
+      "grad_norm": 0.42759883403778076,
+      "learning_rate": 2.7921242583597228e-05,
+      "loss": 1.716,
+      "step": 6215
+    },
+    {
+      "epoch": 0.6456840137114366,
+      "grad_norm": 0.4132600426673889,
+      "learning_rate": 2.7906604135639945e-05,
+      "loss": 1.6549,
+      "step": 6216
+    },
+    {
+      "epoch": 0.645787888231017,
+      "grad_norm": 0.42022904753685,
+      "learning_rate": 2.7891968040457007e-05,
+      "loss": 1.7255,
+      "step": 6217
+    },
+    {
+      "epoch": 0.6458917627505972,
+      "grad_norm": 0.5371679663658142,
+      "learning_rate": 2.7877334299607037e-05,
+      "loss": 1.993,
+      "step": 6218
+    },
+    {
+      "epoch": 0.6459956372701776,
+      "grad_norm": 0.4094901978969574,
+      "learning_rate": 2.7862702914648364e-05,
+      "loss": 1.6315,
+      "step": 6219
+    },
+    {
+      "epoch": 0.646099511789758,
+      "grad_norm": 0.3900490999221802,
+      "learning_rate": 2.784807388713918e-05,
+      "loss": 1.521,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6462033863093383,
+      "grad_norm": 0.4084405303001404,
+      "learning_rate": 2.7833447218637332e-05,
+      "loss": 1.6055,
+      "step": 6221
+    },
+    {
+      "epoch": 0.6463072608289187,
+      "grad_norm": 0.38741302490234375,
+      "learning_rate": 2.7818822910700442e-05,
+      "loss": 1.6364,
+      "step": 6222
+    },
+    {
+      "epoch": 0.646411135348499,
+      "grad_norm": 0.443694531917572,
+      "learning_rate": 2.7804200964885897e-05,
+      "loss": 1.6882,
+      "step": 6223
+    },
+    {
+      "epoch": 0.6465150098680794,
+      "grad_norm": 0.47792142629623413,
+      "learning_rate": 2.7789581382750794e-05,
+      "loss": 1.6535,
+      "step": 6224
+    },
+    {
+      "epoch": 0.6466188843876597,
+      "grad_norm": 0.40224647521972656,
+      "learning_rate": 2.777496416585207e-05,
+      "loss": 1.5526,
+      "step": 6225
+    },
+    {
+      "epoch": 0.6467227589072401,
+      "grad_norm": 0.4233601987361908,
+      "learning_rate": 2.776034931574626e-05,
+      "loss": 1.6387,
+      "step": 6226
+    },
+    {
+      "epoch": 0.6468266334268205,
+      "grad_norm": 0.4132138788700104,
+      "learning_rate": 2.7745736833989788e-05,
+      "loss": 1.7418,
+      "step": 6227
+    },
+    {
+      "epoch": 0.6469305079464007,
+      "grad_norm": 0.4106374979019165,
+      "learning_rate": 2.7731126722138756e-05,
+      "loss": 1.7758,
+      "step": 6228
+    },
+    {
+      "epoch": 0.6470343824659811,
+      "grad_norm": 0.40899068117141724,
+      "learning_rate": 2.7716518981749022e-05,
+      "loss": 1.7876,
+      "step": 6229
+    },
+    {
+      "epoch": 0.6471382569855615,
+      "grad_norm": 0.4388611614704132,
+      "learning_rate": 2.77019136143762e-05,
+      "loss": 1.5814,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6472421315051418,
+      "grad_norm": 0.4218350946903229,
+      "learning_rate": 2.7687310621575646e-05,
+      "loss": 1.6095,
+      "step": 6231
+    },
+    {
+      "epoch": 0.6473460060247221,
+      "grad_norm": 0.43606144189834595,
+      "learning_rate": 2.767271000490247e-05,
+      "loss": 1.5995,
+      "step": 6232
+    },
+    {
+      "epoch": 0.6474498805443025,
+      "grad_norm": 0.4526714086532593,
+      "learning_rate": 2.7658111765911486e-05,
+      "loss": 1.8424,
+      "step": 6233
+    },
+    {
+      "epoch": 0.6475537550638828,
+      "grad_norm": 0.43624788522720337,
+      "learning_rate": 2.764351590615738e-05,
+      "loss": 1.544,
+      "step": 6234
+    },
+    {
+      "epoch": 0.6476576295834632,
+      "grad_norm": 0.409190833568573,
+      "learning_rate": 2.762892242719439e-05,
+      "loss": 1.6898,
+      "step": 6235
+    },
+    {
+      "epoch": 0.6477615041030436,
+      "grad_norm": 0.40407395362854004,
+      "learning_rate": 2.7614331330576694e-05,
+      "loss": 1.4979,
+      "step": 6236
+    },
+    {
+      "epoch": 0.6478653786226238,
+      "grad_norm": 0.4525948464870453,
+      "learning_rate": 2.7599742617858088e-05,
+      "loss": 1.8237,
+      "step": 6237
+    },
+    {
+      "epoch": 0.6479692531422042,
+      "grad_norm": 0.44162291288375854,
+      "learning_rate": 2.7585156290592164e-05,
+      "loss": 1.482,
+      "step": 6238
+    },
+    {
+      "epoch": 0.6480731276617846,
+      "grad_norm": 0.44789817929267883,
+      "learning_rate": 2.7570572350332258e-05,
+      "loss": 1.7348,
+      "step": 6239
+    },
+    {
+      "epoch": 0.6481770021813649,
+      "grad_norm": 0.42044568061828613,
+      "learning_rate": 2.7555990798631436e-05,
+      "loss": 1.6357,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6482808767009453,
+      "grad_norm": 0.4480580985546112,
+      "learning_rate": 2.7541411637042526e-05,
+      "loss": 1.8034,
+      "step": 6241
+    },
+    {
+      "epoch": 0.6483847512205256,
+      "grad_norm": 0.4233933389186859,
+      "learning_rate": 2.752683486711807e-05,
+      "loss": 1.6212,
+      "step": 6242
+    },
+    {
+      "epoch": 0.6484886257401059,
+      "grad_norm": 0.38887619972229004,
+      "learning_rate": 2.7512260490410448e-05,
+      "loss": 1.5956,
+      "step": 6243
+    },
+    {
+      "epoch": 0.6485925002596863,
+      "grad_norm": 0.3876674771308899,
+      "learning_rate": 2.7497688508471625e-05,
+      "loss": 1.5114,
+      "step": 6244
+    },
+    {
+      "epoch": 0.6486963747792667,
+      "grad_norm": 0.4232475459575653,
+      "learning_rate": 2.748311892285348e-05,
+      "loss": 1.4739,
+      "step": 6245
+    },
+    {
+      "epoch": 0.6488002492988469,
+      "grad_norm": 0.4148775041103363,
+      "learning_rate": 2.746855173510754e-05,
+      "loss": 1.5095,
+      "step": 6246
+    },
+    {
+      "epoch": 0.6489041238184273,
+      "grad_norm": 0.4203701913356781,
+      "learning_rate": 2.745398694678506e-05,
+      "loss": 1.8714,
+      "step": 6247
+    },
+    {
+      "epoch": 0.6490079983380077,
+      "grad_norm": 0.424444317817688,
+      "learning_rate": 2.7439424559437122e-05,
+      "loss": 1.7226,
+      "step": 6248
+    },
+    {
+      "epoch": 0.649111872857588,
+      "grad_norm": 0.3754339814186096,
+      "learning_rate": 2.7424864574614484e-05,
+      "loss": 1.5442,
+      "step": 6249
+    },
+    {
+      "epoch": 0.6492157473771684,
+      "grad_norm": 0.506229043006897,
+      "learning_rate": 2.7410306993867675e-05,
+      "loss": 1.713,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6493196218967487,
+      "grad_norm": 0.43399345874786377,
+      "learning_rate": 2.7395751818746972e-05,
+      "loss": 1.6123,
+      "step": 6251
+    },
+    {
+      "epoch": 0.6494234964163291,
+      "grad_norm": 0.40855225920677185,
+      "learning_rate": 2.738119905080235e-05,
+      "loss": 1.642,
+      "step": 6252
+    },
+    {
+      "epoch": 0.6495273709359094,
+      "grad_norm": 0.3835819959640503,
+      "learning_rate": 2.7366648691583646e-05,
+      "loss": 1.4608,
+      "step": 6253
+    },
+    {
+      "epoch": 0.6496312454554898,
+      "grad_norm": 0.41114479303359985,
+      "learning_rate": 2.735210074264027e-05,
+      "loss": 1.5948,
+      "step": 6254
+    },
+    {
+      "epoch": 0.6497351199750702,
+      "grad_norm": 0.45714348554611206,
+      "learning_rate": 2.7337555205521514e-05,
+      "loss": 1.5191,
+      "step": 6255
+    },
+    {
+      "epoch": 0.6498389944946504,
+      "grad_norm": 0.4060697853565216,
+      "learning_rate": 2.7323012081776368e-05,
+      "loss": 1.5952,
+      "step": 6256
+    },
+    {
+      "epoch": 0.6499428690142308,
+      "grad_norm": 0.4719262421131134,
+      "learning_rate": 2.7308471372953538e-05,
+      "loss": 1.644,
+      "step": 6257
+    },
+    {
+      "epoch": 0.6500467435338112,
+      "grad_norm": 0.40513306856155396,
+      "learning_rate": 2.7293933080601507e-05,
+      "loss": 1.7363,
+      "step": 6258
+    },
+    {
+      "epoch": 0.6501506180533915,
+      "grad_norm": 0.4476475417613983,
+      "learning_rate": 2.727939720626849e-05,
+      "loss": 1.7536,
+      "step": 6259
+    },
+    {
+      "epoch": 0.6502544925729719,
+      "grad_norm": 0.44684842228889465,
+      "learning_rate": 2.7264863751502446e-05,
+      "loss": 1.8819,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6503583670925522,
+      "grad_norm": 0.40098902583122253,
+      "learning_rate": 2.725033271785105e-05,
+      "loss": 1.6615,
+      "step": 6261
+    },
+    {
+      "epoch": 0.6504622416121325,
+      "grad_norm": 0.4347749948501587,
+      "learning_rate": 2.7235804106861807e-05,
+      "loss": 1.5312,
+      "step": 6262
+    },
+    {
+      "epoch": 0.6505661161317129,
+      "grad_norm": 0.4464839994907379,
+      "learning_rate": 2.7221277920081818e-05,
+      "loss": 1.8898,
+      "step": 6263
+    },
+    {
+      "epoch": 0.6506699906512933,
+      "grad_norm": 0.4316708743572235,
+      "learning_rate": 2.720675415905807e-05,
+      "loss": 1.779,
+      "step": 6264
+    },
+    {
+      "epoch": 0.6507738651708735,
+      "grad_norm": 0.42350926995277405,
+      "learning_rate": 2.7192232825337206e-05,
+      "loss": 1.6765,
+      "step": 6265
+    },
+    {
+      "epoch": 0.6508777396904539,
+      "grad_norm": 0.39864397048950195,
+      "learning_rate": 2.7177713920465643e-05,
+      "loss": 1.631,
+      "step": 6266
+    },
+    {
+      "epoch": 0.6509816142100343,
+      "grad_norm": 0.40547481179237366,
+      "learning_rate": 2.7163197445989518e-05,
+      "loss": 1.5956,
+      "step": 6267
+    },
+    {
+      "epoch": 0.6510854887296146,
+      "grad_norm": 0.4294564127922058,
+      "learning_rate": 2.7148683403454746e-05,
+      "loss": 1.5979,
+      "step": 6268
+    },
+    {
+      "epoch": 0.651189363249195,
+      "grad_norm": 0.41150328516960144,
+      "learning_rate": 2.7134171794406932e-05,
+      "loss": 1.546,
+      "step": 6269
+    },
+    {
+      "epoch": 0.6512932377687753,
+      "grad_norm": 0.4225398004055023,
+      "learning_rate": 2.711966262039145e-05,
+      "loss": 1.7254,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6513971122883556,
+      "grad_norm": 0.42999348044395447,
+      "learning_rate": 2.710515588295346e-05,
+      "loss": 1.7585,
+      "step": 6271
+    },
+    {
+      "epoch": 0.651500986807936,
+      "grad_norm": 0.39667728543281555,
+      "learning_rate": 2.709065158363774e-05,
+      "loss": 1.5776,
+      "step": 6272
+    },
+    {
+      "epoch": 0.6516048613275164,
+      "grad_norm": 0.41493138670921326,
+      "learning_rate": 2.7076149723988954e-05,
+      "loss": 1.7011,
+      "step": 6273
+    },
+    {
+      "epoch": 0.6517087358470967,
+      "grad_norm": 0.45391151309013367,
+      "learning_rate": 2.7061650305551424e-05,
+      "loss": 1.7816,
+      "step": 6274
+    },
+    {
+      "epoch": 0.651812610366677,
+      "grad_norm": 0.4255356788635254,
+      "learning_rate": 2.7047153329869167e-05,
+      "loss": 1.8013,
+      "step": 6275
+    },
+    {
+      "epoch": 0.6519164848862574,
+      "grad_norm": 0.3899310827255249,
+      "learning_rate": 2.7032658798486072e-05,
+      "loss": 1.4257,
+      "step": 6276
+    },
+    {
+      "epoch": 0.6520203594058378,
+      "grad_norm": 0.4353606402873993,
+      "learning_rate": 2.7018166712945635e-05,
+      "loss": 1.732,
+      "step": 6277
+    },
+    {
+      "epoch": 0.6521242339254181,
+      "grad_norm": 0.4526216685771942,
+      "learning_rate": 2.700367707479122e-05,
+      "loss": 1.7406,
+      "step": 6278
+    },
+    {
+      "epoch": 0.6522281084449985,
+      "grad_norm": 0.44765564799308777,
+      "learning_rate": 2.6989189885565803e-05,
+      "loss": 1.7469,
+      "step": 6279
+    },
+    {
+      "epoch": 0.6523319829645788,
+      "grad_norm": 0.4167086184024811,
+      "learning_rate": 2.697470514681214e-05,
+      "loss": 1.6946,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6524358574841591,
+      "grad_norm": 0.40468063950538635,
+      "learning_rate": 2.6960222860072826e-05,
+      "loss": 1.5709,
+      "step": 6281
+    },
+    {
+      "epoch": 0.6525397320037395,
+      "grad_norm": 0.411888062953949,
+      "learning_rate": 2.6945743026890012e-05,
+      "loss": 1.4393,
+      "step": 6282
+    },
+    {
+      "epoch": 0.6526436065233199,
+      "grad_norm": 0.41102951765060425,
+      "learning_rate": 2.693126564880576e-05,
+      "loss": 1.68,
+      "step": 6283
+    },
+    {
+      "epoch": 0.6527474810429001,
+      "grad_norm": 0.43322500586509705,
+      "learning_rate": 2.6916790727361773e-05,
+      "loss": 1.6189,
+      "step": 6284
+    },
+    {
+      "epoch": 0.6528513555624805,
+      "grad_norm": 0.47928208112716675,
+      "learning_rate": 2.6902318264099517e-05,
+      "loss": 1.8985,
+      "step": 6285
+    },
+    {
+      "epoch": 0.6529552300820609,
+      "grad_norm": 0.4368571639060974,
+      "learning_rate": 2.68878482605602e-05,
+      "loss": 1.7447,
+      "step": 6286
+    },
+    {
+      "epoch": 0.6530591046016412,
+      "grad_norm": 0.39063751697540283,
+      "learning_rate": 2.6873380718284757e-05,
+      "loss": 1.5522,
+      "step": 6287
+    },
+    {
+      "epoch": 0.6531629791212216,
+      "grad_norm": 0.4288500249385834,
+      "learning_rate": 2.6858915638813874e-05,
+      "loss": 1.8171,
+      "step": 6288
+    },
+    {
+      "epoch": 0.653266853640802,
+      "grad_norm": 0.42578479647636414,
+      "learning_rate": 2.6844453023687956e-05,
+      "loss": 1.5586,
+      "step": 6289
+    },
+    {
+      "epoch": 0.6533707281603822,
+      "grad_norm": 0.43467986583709717,
+      "learning_rate": 2.6829992874447196e-05,
+      "loss": 1.8163,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6534746026799626,
+      "grad_norm": 0.39412596821784973,
+      "learning_rate": 2.6815535192631437e-05,
+      "loss": 1.4798,
+      "step": 6291
+    },
+    {
+      "epoch": 0.653578477199543,
+      "grad_norm": 0.40805160999298096,
+      "learning_rate": 2.6801079979780343e-05,
+      "loss": 1.7423,
+      "step": 6292
+    },
+    {
+      "epoch": 0.6536823517191233,
+      "grad_norm": 0.4715184271335602,
+      "learning_rate": 2.6786627237433275e-05,
+      "loss": 1.7352,
+      "step": 6293
+    },
+    {
+      "epoch": 0.6537862262387036,
+      "grad_norm": 0.469553142786026,
+      "learning_rate": 2.6772176967129336e-05,
+      "loss": 1.892,
+      "step": 6294
+    },
+    {
+      "epoch": 0.653890100758284,
+      "grad_norm": 0.45803409814834595,
+      "learning_rate": 2.6757729170407368e-05,
+      "loss": 1.7635,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6539939752778643,
+      "grad_norm": 0.4040502607822418,
+      "learning_rate": 2.6743283848805945e-05,
+      "loss": 1.7023,
+      "step": 6296
+    },
+    {
+      "epoch": 0.6540978497974447,
+      "grad_norm": 0.4045460522174835,
+      "learning_rate": 2.672884100386338e-05,
+      "loss": 1.6528,
+      "step": 6297
+    },
+    {
+      "epoch": 0.6542017243170251,
+      "grad_norm": 0.44102489948272705,
+      "learning_rate": 2.6714400637117708e-05,
+      "loss": 1.7864,
+      "step": 6298
+    },
+    {
+      "epoch": 0.6543055988366053,
+      "grad_norm": 0.43990832567214966,
+      "learning_rate": 2.669996275010678e-05,
+      "loss": 1.7531,
+      "step": 6299
+    },
+    {
+      "epoch": 0.6544094733561857,
+      "grad_norm": 0.46511977910995483,
+      "learning_rate": 2.6685527344368022e-05,
+      "loss": 1.7993,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6545133478757661,
+      "grad_norm": 0.40895703434944153,
+      "learning_rate": 2.667109442143877e-05,
+      "loss": 1.7012,
+      "step": 6301
+    },
+    {
+      "epoch": 0.6546172223953465,
+      "grad_norm": 0.41711392998695374,
+      "learning_rate": 2.665666398285599e-05,
+      "loss": 1.6973,
+      "step": 6302
+    },
+    {
+      "epoch": 0.6547210969149267,
+      "grad_norm": 0.44164448976516724,
+      "learning_rate": 2.6642236030156405e-05,
+      "loss": 1.6828,
+      "step": 6303
+    },
+    {
+      "epoch": 0.6548249714345071,
+      "grad_norm": 0.45560404658317566,
+      "learning_rate": 2.6627810564876488e-05,
+      "loss": 1.7191,
+      "step": 6304
+    },
+    {
+      "epoch": 0.6549288459540875,
+      "grad_norm": 0.40289151668548584,
+      "learning_rate": 2.6613387588552406e-05,
+      "loss": 1.6014,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6550327204736678,
+      "grad_norm": 0.44911909103393555,
+      "learning_rate": 2.6598967102720174e-05,
+      "loss": 1.794,
+      "step": 6306
+    },
+    {
+      "epoch": 0.6551365949932482,
+      "grad_norm": 0.4513666331768036,
+      "learning_rate": 2.658454910891539e-05,
+      "loss": 1.697,
+      "step": 6307
+    },
+    {
+      "epoch": 0.6552404695128285,
+      "grad_norm": 0.4077112674713135,
+      "learning_rate": 2.657013360867346e-05,
+      "loss": 1.5579,
+      "step": 6308
+    },
+    {
+      "epoch": 0.6553443440324088,
+      "grad_norm": 0.4188466966152191,
+      "learning_rate": 2.6555720603529578e-05,
+      "loss": 1.61,
+      "step": 6309
+    },
+    {
+      "epoch": 0.6554482185519892,
+      "grad_norm": 0.38735076785087585,
+      "learning_rate": 2.6541310095018533e-05,
+      "loss": 1.4753,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6555520930715696,
+      "grad_norm": 0.4412975609302521,
+      "learning_rate": 2.6526902084675008e-05,
+      "loss": 1.7289,
+      "step": 6311
+    },
+    {
+      "epoch": 0.6556559675911499,
+      "grad_norm": 0.3914285898208618,
+      "learning_rate": 2.6512496574033307e-05,
+      "loss": 1.5964,
+      "step": 6312
+    },
+    {
+      "epoch": 0.6557598421107302,
+      "grad_norm": 0.4345444142818451,
+      "learning_rate": 2.6498093564627515e-05,
+      "loss": 1.5983,
+      "step": 6313
+    },
+    {
+      "epoch": 0.6558637166303106,
+      "grad_norm": 0.4521616995334625,
+      "learning_rate": 2.6483693057991433e-05,
+      "loss": 1.719,
+      "step": 6314
+    },
+    {
+      "epoch": 0.6559675911498909,
+      "grad_norm": 0.465675413608551,
+      "learning_rate": 2.646929505565861e-05,
+      "loss": 1.7625,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6560714656694713,
+      "grad_norm": 0.4356716573238373,
+      "learning_rate": 2.6454899559162316e-05,
+      "loss": 1.6035,
+      "step": 6316
+    },
+    {
+      "epoch": 0.6561753401890517,
+      "grad_norm": 0.3972066640853882,
+      "learning_rate": 2.644050657003554e-05,
+      "loss": 1.8204,
+      "step": 6317
+    },
+    {
+      "epoch": 0.6562792147086319,
+      "grad_norm": 0.40507668256759644,
+      "learning_rate": 2.642611608981108e-05,
+      "loss": 1.5586,
+      "step": 6318
+    },
+    {
+      "epoch": 0.6563830892282123,
+      "grad_norm": 0.4741031229496002,
+      "learning_rate": 2.641172812002134e-05,
+      "loss": 1.9185,
+      "step": 6319
+    },
+    {
+      "epoch": 0.6564869637477927,
+      "grad_norm": 0.4094397723674774,
+      "learning_rate": 2.6397342662198578e-05,
+      "loss": 1.7196,
+      "step": 6320
+    },
+    {
+      "epoch": 0.656590838267373,
+      "grad_norm": 0.43364858627319336,
+      "learning_rate": 2.638295971787471e-05,
+      "loss": 1.7506,
+      "step": 6321
+    },
+    {
+      "epoch": 0.6566947127869533,
+      "grad_norm": 0.43040645122528076,
+      "learning_rate": 2.6368579288581414e-05,
+      "loss": 1.5984,
+      "step": 6322
+    },
+    {
+      "epoch": 0.6567985873065337,
+      "grad_norm": 0.41142383217811584,
+      "learning_rate": 2.635420137585009e-05,
+      "loss": 1.5823,
+      "step": 6323
+    },
+    {
+      "epoch": 0.656902461826114,
+      "grad_norm": 0.4649263918399811,
+      "learning_rate": 2.6339825981211873e-05,
+      "loss": 1.8616,
+      "step": 6324
+    },
+    {
+      "epoch": 0.6570063363456944,
+      "grad_norm": 0.42940422892570496,
+      "learning_rate": 2.632545310619764e-05,
+      "loss": 1.4848,
+      "step": 6325
+    },
+    {
+      "epoch": 0.6571102108652748,
+      "grad_norm": 0.4017314016819,
+      "learning_rate": 2.6311082752337957e-05,
+      "loss": 1.4897,
+      "step": 6326
+    },
+    {
+      "epoch": 0.657214085384855,
+      "grad_norm": 0.4207054674625397,
+      "learning_rate": 2.6296714921163224e-05,
+      "loss": 1.5465,
+      "step": 6327
+    },
+    {
+      "epoch": 0.6573179599044354,
+      "grad_norm": 0.3997343182563782,
+      "learning_rate": 2.6282349614203415e-05,
+      "loss": 1.4289,
+      "step": 6328
+    },
+    {
+      "epoch": 0.6574218344240158,
+      "grad_norm": 0.43719449639320374,
+      "learning_rate": 2.626798683298839e-05,
+      "loss": 1.64,
+      "step": 6329
+    },
+    {
+      "epoch": 0.6575257089435962,
+      "grad_norm": 0.47477173805236816,
+      "learning_rate": 2.6253626579047654e-05,
+      "loss": 1.4736,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6576295834631765,
+      "grad_norm": 0.39647629857063293,
+      "learning_rate": 2.6239268853910452e-05,
+      "loss": 1.5444,
+      "step": 6331
+    },
+    {
+      "epoch": 0.6577334579827568,
+      "grad_norm": 0.3856428563594818,
+      "learning_rate": 2.6224913659105775e-05,
+      "loss": 1.5782,
+      "step": 6332
+    },
+    {
+      "epoch": 0.6578373325023372,
+      "grad_norm": 0.41423386335372925,
+      "learning_rate": 2.621056099616233e-05,
+      "loss": 1.5784,
+      "step": 6333
+    },
+    {
+      "epoch": 0.6579412070219175,
+      "grad_norm": 0.38329392671585083,
+      "learning_rate": 2.6196210866608606e-05,
+      "loss": 1.6442,
+      "step": 6334
+    },
+    {
+      "epoch": 0.6580450815414979,
+      "grad_norm": 0.39722350239753723,
+      "learning_rate": 2.6181863271972735e-05,
+      "loss": 1.6947,
+      "step": 6335
+    },
+    {
+      "epoch": 0.6581489560610783,
+      "grad_norm": 0.4223669469356537,
+      "learning_rate": 2.616751821378261e-05,
+      "loss": 1.6055,
+      "step": 6336
+    },
+    {
+      "epoch": 0.6582528305806585,
+      "grad_norm": 0.39463213086128235,
+      "learning_rate": 2.6153175693565933e-05,
+      "loss": 1.5574,
+      "step": 6337
+    },
+    {
+      "epoch": 0.6583567051002389,
+      "grad_norm": 0.40961316227912903,
+      "learning_rate": 2.6138835712850003e-05,
+      "loss": 1.6478,
+      "step": 6338
+    },
+    {
+      "epoch": 0.6584605796198193,
+      "grad_norm": 0.39676934480667114,
+      "learning_rate": 2.6124498273161957e-05,
+      "loss": 1.6545,
+      "step": 6339
+    },
+    {
+      "epoch": 0.6585644541393996,
+      "grad_norm": 0.397554874420166,
+      "learning_rate": 2.6110163376028613e-05,
+      "loss": 1.5861,
+      "step": 6340
+    },
+    {
+      "epoch": 0.65866832865898,
+      "grad_norm": 0.40185415744781494,
+      "learning_rate": 2.6095831022976514e-05,
+      "loss": 1.6366,
+      "step": 6341
+    },
+    {
+      "epoch": 0.6587722031785603,
+      "grad_norm": 0.3959345519542694,
+      "learning_rate": 2.6081501215531955e-05,
+      "loss": 1.6307,
+      "step": 6342
+    },
+    {
+      "epoch": 0.6588760776981406,
+      "grad_norm": 0.41983965039253235,
+      "learning_rate": 2.606717395522095e-05,
+      "loss": 1.6515,
+      "step": 6343
+    },
+    {
+      "epoch": 0.658979952217721,
+      "grad_norm": 0.41705140471458435,
+      "learning_rate": 2.6052849243569233e-05,
+      "loss": 1.6791,
+      "step": 6344
+    },
+    {
+      "epoch": 0.6590838267373014,
+      "grad_norm": 0.5571948885917664,
+      "learning_rate": 2.603852708210226e-05,
+      "loss": 1.7317,
+      "step": 6345
+    },
+    {
+      "epoch": 0.6591877012568816,
+      "grad_norm": 0.39944010972976685,
+      "learning_rate": 2.6024207472345273e-05,
+      "loss": 1.6033,
+      "step": 6346
+    },
+    {
+      "epoch": 0.659291575776462,
+      "grad_norm": 0.39192959666252136,
+      "learning_rate": 2.6009890415823147e-05,
+      "loss": 1.6101,
+      "step": 6347
+    },
+    {
+      "epoch": 0.6593954502960424,
+      "grad_norm": 0.41190558671951294,
+      "learning_rate": 2.5995575914060576e-05,
+      "loss": 1.5903,
+      "step": 6348
+    },
+    {
+      "epoch": 0.6594993248156227,
+      "grad_norm": 0.39744171500205994,
+      "learning_rate": 2.5981263968581925e-05,
+      "loss": 1.663,
+      "step": 6349
+    },
+    {
+      "epoch": 0.6596031993352031,
+      "grad_norm": 0.4570941925048828,
+      "learning_rate": 2.5966954580911314e-05,
+      "loss": 1.5257,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6597070738547834,
+      "grad_norm": 0.43954119086265564,
+      "learning_rate": 2.5952647752572566e-05,
+      "loss": 1.8657,
+      "step": 6351
+    },
+    {
+      "epoch": 0.6598109483743637,
+      "grad_norm": 0.4737524390220642,
+      "learning_rate": 2.5938343485089267e-05,
+      "loss": 1.9287,
+      "step": 6352
+    },
+    {
+      "epoch": 0.6599148228939441,
+      "grad_norm": 0.4555112421512604,
+      "learning_rate": 2.5924041779984692e-05,
+      "loss": 1.7726,
+      "step": 6353
+    },
+    {
+      "epoch": 0.6600186974135245,
+      "grad_norm": 0.4156988561153412,
+      "learning_rate": 2.5909742638781852e-05,
+      "loss": 1.6406,
+      "step": 6354
+    },
+    {
+      "epoch": 0.6601225719331049,
+      "grad_norm": 0.45365050435066223,
+      "learning_rate": 2.5895446063003526e-05,
+      "loss": 1.6795,
+      "step": 6355
+    },
+    {
+      "epoch": 0.6602264464526851,
+      "grad_norm": 0.5357591509819031,
+      "learning_rate": 2.5881152054172193e-05,
+      "loss": 1.6828,
+      "step": 6356
+    },
+    {
+      "epoch": 0.6603303209722655,
+      "grad_norm": 0.422297865152359,
+      "learning_rate": 2.586686061380999e-05,
+      "loss": 1.6548,
+      "step": 6357
+    },
+    {
+      "epoch": 0.6604341954918459,
+      "grad_norm": 0.40216442942619324,
+      "learning_rate": 2.5852571743438908e-05,
+      "loss": 1.5247,
+      "step": 6358
+    },
+    {
+      "epoch": 0.6605380700114262,
+      "grad_norm": 0.4673736095428467,
+      "learning_rate": 2.583828544458058e-05,
+      "loss": 1.6892,
+      "step": 6359
+    },
+    {
+      "epoch": 0.6606419445310066,
+      "grad_norm": 0.42048928141593933,
+      "learning_rate": 2.582400171875638e-05,
+      "loss": 1.6188,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6607458190505869,
+      "grad_norm": 0.4226718246936798,
+      "learning_rate": 2.5809720567487395e-05,
+      "loss": 1.4864,
+      "step": 6361
+    },
+    {
+      "epoch": 0.6608496935701672,
+      "grad_norm": 0.46937909722328186,
+      "learning_rate": 2.5795441992294523e-05,
+      "loss": 1.7432,
+      "step": 6362
+    },
+    {
+      "epoch": 0.6609535680897476,
+      "grad_norm": 0.43607574701309204,
+      "learning_rate": 2.578116599469825e-05,
+      "loss": 1.8233,
+      "step": 6363
+    },
+    {
+      "epoch": 0.661057442609328,
+      "grad_norm": 0.4324806034564972,
+      "learning_rate": 2.576689257621887e-05,
+      "loss": 1.7842,
+      "step": 6364
+    },
+    {
+      "epoch": 0.6611613171289082,
+      "grad_norm": 0.3860158920288086,
+      "learning_rate": 2.5752621738376447e-05,
+      "loss": 1.4591,
+      "step": 6365
+    },
+    {
+      "epoch": 0.6612651916484886,
+      "grad_norm": 0.3831734359264374,
+      "learning_rate": 2.5738353482690624e-05,
+      "loss": 1.527,
+      "step": 6366
+    },
+    {
+      "epoch": 0.661369066168069,
+      "grad_norm": 0.41983523964881897,
+      "learning_rate": 2.5724087810680937e-05,
+      "loss": 1.7582,
+      "step": 6367
+    },
+    {
+      "epoch": 0.6614729406876493,
+      "grad_norm": 0.4340680241584778,
+      "learning_rate": 2.570982472386653e-05,
+      "loss": 1.7625,
+      "step": 6368
+    },
+    {
+      "epoch": 0.6615768152072297,
+      "grad_norm": 0.4355176091194153,
+      "learning_rate": 2.569556422376632e-05,
+      "loss": 1.6903,
+      "step": 6369
+    },
+    {
+      "epoch": 0.66168068972681,
+      "grad_norm": 0.39546066522598267,
+      "learning_rate": 2.5681306311898945e-05,
+      "loss": 1.6276,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6617845642463903,
+      "grad_norm": 0.40482446551322937,
+      "learning_rate": 2.566705098978275e-05,
+      "loss": 1.7003,
+      "step": 6371
+    },
+    {
+      "epoch": 0.6618884387659707,
+      "grad_norm": 0.4112469255924225,
+      "learning_rate": 2.5652798258935825e-05,
+      "loss": 1.6817,
+      "step": 6372
+    },
+    {
+      "epoch": 0.6619923132855511,
+      "grad_norm": 0.4560087025165558,
+      "learning_rate": 2.5638548120875945e-05,
+      "loss": 1.7853,
+      "step": 6373
+    },
+    {
+      "epoch": 0.6620961878051314,
+      "grad_norm": 0.4203476905822754,
+      "learning_rate": 2.5624300577120703e-05,
+      "loss": 1.7259,
+      "step": 6374
+    },
+    {
+      "epoch": 0.6622000623247117,
+      "grad_norm": 0.4099465608596802,
+      "learning_rate": 2.5610055629187275e-05,
+      "loss": 1.6924,
+      "step": 6375
+    },
+    {
+      "epoch": 0.6623039368442921,
+      "grad_norm": 0.4466152489185333,
+      "learning_rate": 2.559581327859269e-05,
+      "loss": 1.8033,
+      "step": 6376
+    },
+    {
+      "epoch": 0.6624078113638724,
+      "grad_norm": 0.487020343542099,
+      "learning_rate": 2.558157352685363e-05,
+      "loss": 1.9298,
+      "step": 6377
+    },
+    {
+      "epoch": 0.6625116858834528,
+      "grad_norm": 0.5030232071876526,
+      "learning_rate": 2.5567336375486528e-05,
+      "loss": 1.6374,
+      "step": 6378
+    },
+    {
+      "epoch": 0.6626155604030332,
+      "grad_norm": 0.4117771089076996,
+      "learning_rate": 2.555310182600751e-05,
+      "loss": 1.6823,
+      "step": 6379
+    },
+    {
+      "epoch": 0.6627194349226135,
+      "grad_norm": 0.5117984414100647,
+      "learning_rate": 2.5538869879932438e-05,
+      "loss": 1.7883,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6628233094421938,
+      "grad_norm": 0.4503713548183441,
+      "learning_rate": 2.5524640538776955e-05,
+      "loss": 1.8058,
+      "step": 6381
+    },
+    {
+      "epoch": 0.6629271839617742,
+      "grad_norm": 0.4063549041748047,
+      "learning_rate": 2.551041380405631e-05,
+      "loss": 1.6395,
+      "step": 6382
+    },
+    {
+      "epoch": 0.6630310584813546,
+      "grad_norm": 0.4199891984462738,
+      "learning_rate": 2.5496189677285574e-05,
+      "loss": 1.6788,
+      "step": 6383
+    },
+    {
+      "epoch": 0.6631349330009348,
+      "grad_norm": 0.45948025584220886,
+      "learning_rate": 2.548196815997953e-05,
+      "loss": 1.8964,
+      "step": 6384
+    },
+    {
+      "epoch": 0.6632388075205152,
+      "grad_norm": 0.4091807007789612,
+      "learning_rate": 2.546774925365259e-05,
+      "loss": 1.7551,
+      "step": 6385
+    },
+    {
+      "epoch": 0.6633426820400956,
+      "grad_norm": 0.3907213807106018,
+      "learning_rate": 2.5453532959819017e-05,
+      "loss": 1.6075,
+      "step": 6386
+    },
+    {
+      "epoch": 0.6634465565596759,
+      "grad_norm": 0.4916064143180847,
+      "learning_rate": 2.5439319279992708e-05,
+      "loss": 1.8376,
+      "step": 6387
+    },
+    {
+      "epoch": 0.6635504310792563,
+      "grad_norm": 0.3837120532989502,
+      "learning_rate": 2.5425108215687314e-05,
+      "loss": 1.6351,
+      "step": 6388
+    },
+    {
+      "epoch": 0.6636543055988366,
+      "grad_norm": 0.4401141405105591,
+      "learning_rate": 2.5410899768416186e-05,
+      "loss": 1.7869,
+      "step": 6389
+    },
+    {
+      "epoch": 0.6637581801184169,
+      "grad_norm": 0.4023844599723816,
+      "learning_rate": 2.5396693939692474e-05,
+      "loss": 1.5495,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6638620546379973,
+      "grad_norm": 0.41856905817985535,
+      "learning_rate": 2.5382490731028918e-05,
+      "loss": 1.641,
+      "step": 6391
+    },
+    {
+      "epoch": 0.6639659291575777,
+      "grad_norm": 0.40514466166496277,
+      "learning_rate": 2.5368290143938062e-05,
+      "loss": 1.7506,
+      "step": 6392
+    },
+    {
+      "epoch": 0.664069803677158,
+      "grad_norm": 0.4336822032928467,
+      "learning_rate": 2.5354092179932204e-05,
+      "loss": 1.7111,
+      "step": 6393
+    },
+    {
+      "epoch": 0.6641736781967383,
+      "grad_norm": 0.4441390037536621,
+      "learning_rate": 2.533989684052325e-05,
+      "loss": 1.6557,
+      "step": 6394
+    },
+    {
+      "epoch": 0.6642775527163187,
+      "grad_norm": 0.4061375558376312,
+      "learning_rate": 2.5325704127222937e-05,
+      "loss": 1.5773,
+      "step": 6395
+    },
+    {
+      "epoch": 0.664381427235899,
+      "grad_norm": 0.4289427101612091,
+      "learning_rate": 2.5311514041542678e-05,
+      "loss": 1.8766,
+      "step": 6396
+    },
+    {
+      "epoch": 0.6644853017554794,
+      "grad_norm": 0.40495064854621887,
+      "learning_rate": 2.529732658499359e-05,
+      "loss": 1.582,
+      "step": 6397
+    },
+    {
+      "epoch": 0.6645891762750598,
+      "grad_norm": 0.4054896831512451,
+      "learning_rate": 2.5283141759086537e-05,
+      "loss": 1.6379,
+      "step": 6398
+    },
+    {
+      "epoch": 0.66469305079464,
+      "grad_norm": 0.3749221861362457,
+      "learning_rate": 2.5268959565332086e-05,
+      "loss": 1.4214,
+      "step": 6399
+    },
+    {
+      "epoch": 0.6647969253142204,
+      "grad_norm": 0.3932819068431854,
+      "learning_rate": 2.525478000524053e-05,
+      "loss": 1.6518,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6649007998338008,
+      "grad_norm": 0.4017762541770935,
+      "learning_rate": 2.5240603080321863e-05,
+      "loss": 1.578,
+      "step": 6401
+    },
+    {
+      "epoch": 0.6650046743533811,
+      "grad_norm": 0.47562339901924133,
+      "learning_rate": 2.5226428792085875e-05,
+      "loss": 2.0138,
+      "step": 6402
+    },
+    {
+      "epoch": 0.6651085488729614,
+      "grad_norm": 0.407149076461792,
+      "learning_rate": 2.5212257142041938e-05,
+      "loss": 1.5924,
+      "step": 6403
+    },
+    {
+      "epoch": 0.6652124233925418,
+      "grad_norm": 0.44227856397628784,
+      "learning_rate": 2.5198088131699294e-05,
+      "loss": 1.7202,
+      "step": 6404
+    },
+    {
+      "epoch": 0.6653162979121221,
+      "grad_norm": 0.44379758834838867,
+      "learning_rate": 2.5183921762566797e-05,
+      "loss": 1.8716,
+      "step": 6405
+    },
+    {
+      "epoch": 0.6654201724317025,
+      "grad_norm": 0.42266109585762024,
+      "learning_rate": 2.5169758036153056e-05,
+      "loss": 1.6601,
+      "step": 6406
+    },
+    {
+      "epoch": 0.6655240469512829,
+      "grad_norm": 0.5245758295059204,
+      "learning_rate": 2.5155596953966397e-05,
+      "loss": 1.7759,
+      "step": 6407
+    },
+    {
+      "epoch": 0.6656279214708632,
+      "grad_norm": 0.3796195089817047,
+      "learning_rate": 2.514143851751486e-05,
+      "loss": 1.4598,
+      "step": 6408
+    },
+    {
+      "epoch": 0.6657317959904435,
+      "grad_norm": 0.38661640882492065,
+      "learning_rate": 2.512728272830625e-05,
+      "loss": 1.5706,
+      "step": 6409
+    },
+    {
+      "epoch": 0.6658356705100239,
+      "grad_norm": 0.41082271933555603,
+      "learning_rate": 2.511312958784797e-05,
+      "loss": 1.7148,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6659395450296043,
+      "grad_norm": 0.42262521386146545,
+      "learning_rate": 2.5098979097647297e-05,
+      "loss": 1.561,
+      "step": 6411
+    },
+    {
+      "epoch": 0.6660434195491846,
+      "grad_norm": 0.3816385865211487,
+      "learning_rate": 2.508483125921113e-05,
+      "loss": 1.602,
+      "step": 6412
+    },
+    {
+      "epoch": 0.6661472940687649,
+      "grad_norm": 0.47888967394828796,
+      "learning_rate": 2.507068607404604e-05,
+      "loss": 1.9113,
+      "step": 6413
+    },
+    {
+      "epoch": 0.6662511685883453,
+      "grad_norm": 0.47457215189933777,
+      "learning_rate": 2.5056543543658452e-05,
+      "loss": 1.5093,
+      "step": 6414
+    },
+    {
+      "epoch": 0.6663550431079256,
+      "grad_norm": 0.3893379867076874,
+      "learning_rate": 2.5042403669554416e-05,
+      "loss": 1.5901,
+      "step": 6415
+    },
+    {
+      "epoch": 0.666458917627506,
+      "grad_norm": 0.4125143587589264,
+      "learning_rate": 2.5028266453239703e-05,
+      "loss": 1.6606,
+      "step": 6416
+    },
+    {
+      "epoch": 0.6665627921470864,
+      "grad_norm": 0.3965260088443756,
+      "learning_rate": 2.501413189621983e-05,
+      "loss": 1.6371,
+      "step": 6417
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.4429340660572052,
+      "learning_rate": 2.500000000000001e-05,
+      "loss": 1.7512,
+      "step": 6418
+    },
+    {
+      "epoch": 0.666770541186247,
+      "grad_norm": 0.471049427986145,
+      "learning_rate": 2.4985870766085185e-05,
+      "loss": 1.6877,
+      "step": 6419
+    },
+    {
+      "epoch": 0.6668744157058274,
+      "grad_norm": 0.4106440544128418,
+      "learning_rate": 2.4971744195979985e-05,
+      "loss": 1.6417,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6669782902254077,
+      "grad_norm": 0.4270106554031372,
+      "learning_rate": 2.4957620291188844e-05,
+      "loss": 1.7746,
+      "step": 6421
+    },
+    {
+      "epoch": 0.667082164744988,
+      "grad_norm": 0.39720797538757324,
+      "learning_rate": 2.4943499053215756e-05,
+      "loss": 1.5985,
+      "step": 6422
+    },
+    {
+      "epoch": 0.6671860392645684,
+      "grad_norm": 0.43498122692108154,
+      "learning_rate": 2.4929380483564602e-05,
+      "loss": 1.7645,
+      "step": 6423
+    },
+    {
+      "epoch": 0.6672899137841487,
+      "grad_norm": 0.3922276198863983,
+      "learning_rate": 2.4915264583738858e-05,
+      "loss": 1.5833,
+      "step": 6424
+    },
+    {
+      "epoch": 0.6673937883037291,
+      "grad_norm": 0.4027794599533081,
+      "learning_rate": 2.490115135524177e-05,
+      "loss": 1.6876,
+      "step": 6425
+    },
+    {
+      "epoch": 0.6674976628233095,
+      "grad_norm": 0.4081961214542389,
+      "learning_rate": 2.488704079957629e-05,
+      "loss": 1.7356,
+      "step": 6426
+    },
+    {
+      "epoch": 0.6676015373428897,
+      "grad_norm": 0.427378386259079,
+      "learning_rate": 2.487293291824507e-05,
+      "loss": 1.6506,
+      "step": 6427
+    },
+    {
+      "epoch": 0.6677054118624701,
+      "grad_norm": 0.4446810185909271,
+      "learning_rate": 2.4858827712750494e-05,
+      "loss": 1.8314,
+      "step": 6428
+    },
+    {
+      "epoch": 0.6678092863820505,
+      "grad_norm": 0.41210460662841797,
+      "learning_rate": 2.484472518459464e-05,
+      "loss": 1.6845,
+      "step": 6429
+    },
+    {
+      "epoch": 0.6679131609016308,
+      "grad_norm": 0.39390888810157776,
+      "learning_rate": 2.483062533527938e-05,
+      "loss": 1.4303,
+      "step": 6430
+    },
+    {
+      "epoch": 0.6680170354212112,
+      "grad_norm": 0.4102948307991028,
+      "learning_rate": 2.4816528166306146e-05,
+      "loss": 1.7178,
+      "step": 6431
+    },
+    {
+      "epoch": 0.6681209099407915,
+      "grad_norm": 0.45998239517211914,
+      "learning_rate": 2.4802433679176247e-05,
+      "loss": 1.69,
+      "step": 6432
+    },
+    {
+      "epoch": 0.6682247844603719,
+      "grad_norm": 0.41955703496932983,
+      "learning_rate": 2.4788341875390614e-05,
+      "loss": 1.6254,
+      "step": 6433
+    },
+    {
+      "epoch": 0.6683286589799522,
+      "grad_norm": 0.41680797934532166,
+      "learning_rate": 2.4774252756449912e-05,
+      "loss": 1.6673,
+      "step": 6434
+    },
+    {
+      "epoch": 0.6684325334995326,
+      "grad_norm": 0.43315428495407104,
+      "learning_rate": 2.4760166323854524e-05,
+      "loss": 1.8226,
+      "step": 6435
+    },
+    {
+      "epoch": 0.668536408019113,
+      "grad_norm": 0.4136770963668823,
+      "learning_rate": 2.4746082579104518e-05,
+      "loss": 1.6372,
+      "step": 6436
+    },
+    {
+      "epoch": 0.6686402825386932,
+      "grad_norm": 0.40513721108436584,
+      "learning_rate": 2.4732001523699772e-05,
+      "loss": 1.5601,
+      "step": 6437
+    },
+    {
+      "epoch": 0.6687441570582736,
+      "grad_norm": 0.4485640227794647,
+      "learning_rate": 2.471792315913973e-05,
+      "loss": 1.5854,
+      "step": 6438
+    },
+    {
+      "epoch": 0.668848031577854,
+      "grad_norm": 0.4075925946235657,
+      "learning_rate": 2.470384748692368e-05,
+      "loss": 1.5502,
+      "step": 6439
+    },
+    {
+      "epoch": 0.6689519060974343,
+      "grad_norm": 0.45135799050331116,
+      "learning_rate": 2.468977450855058e-05,
+      "loss": 1.836,
+      "step": 6440
+    },
+    {
+      "epoch": 0.6690557806170147,
+      "grad_norm": 0.41414105892181396,
+      "learning_rate": 2.4675704225519018e-05,
+      "loss": 1.6872,
+      "step": 6441
+    },
+    {
+      "epoch": 0.669159655136595,
+      "grad_norm": 0.39774543046951294,
+      "learning_rate": 2.4661636639327444e-05,
+      "loss": 1.5706,
+      "step": 6442
+    },
+    {
+      "epoch": 0.6692635296561753,
+      "grad_norm": 0.43322694301605225,
+      "learning_rate": 2.4647571751473924e-05,
+      "loss": 1.7595,
+      "step": 6443
+    },
+    {
+      "epoch": 0.6693674041757557,
+      "grad_norm": 0.4438508450984955,
+      "learning_rate": 2.4633509563456252e-05,
+      "loss": 1.5921,
+      "step": 6444
+    },
+    {
+      "epoch": 0.6694712786953361,
+      "grad_norm": 0.460419237613678,
+      "learning_rate": 2.461945007677195e-05,
+      "loss": 1.9157,
+      "step": 6445
+    },
+    {
+      "epoch": 0.6695751532149163,
+      "grad_norm": 0.4211447536945343,
+      "learning_rate": 2.4605393292918234e-05,
+      "loss": 1.7016,
+      "step": 6446
+    },
+    {
+      "epoch": 0.6696790277344967,
+      "grad_norm": 0.40026143193244934,
+      "learning_rate": 2.459133921339205e-05,
+      "loss": 1.5405,
+      "step": 6447
+    },
+    {
+      "epoch": 0.6697829022540771,
+      "grad_norm": 0.4232792258262634,
+      "learning_rate": 2.4577287839690022e-05,
+      "loss": 1.7017,
+      "step": 6448
+    },
+    {
+      "epoch": 0.6698867767736574,
+      "grad_norm": 0.42340609431266785,
+      "learning_rate": 2.456323917330857e-05,
+      "loss": 1.8029,
+      "step": 6449
+    },
+    {
+      "epoch": 0.6699906512932378,
+      "grad_norm": 0.3742263615131378,
+      "learning_rate": 2.4549193215743706e-05,
+      "loss": 1.6008,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6700945258128181,
+      "grad_norm": 0.41276246309280396,
+      "learning_rate": 2.453514996849125e-05,
+      "loss": 1.667,
+      "step": 6451
+    },
+    {
+      "epoch": 0.6701984003323984,
+      "grad_norm": 0.4266214668750763,
+      "learning_rate": 2.452110943304669e-05,
+      "loss": 1.5254,
+      "step": 6452
+    },
+    {
+      "epoch": 0.6703022748519788,
+      "grad_norm": 0.39875510334968567,
+      "learning_rate": 2.450707161090523e-05,
+      "loss": 1.5862,
+      "step": 6453
+    },
+    {
+      "epoch": 0.6704061493715592,
+      "grad_norm": 0.47201064229011536,
+      "learning_rate": 2.4493036503561795e-05,
+      "loss": 1.7249,
+      "step": 6454
+    },
+    {
+      "epoch": 0.6705100238911395,
+      "grad_norm": 0.43423768877983093,
+      "learning_rate": 2.4479004112511012e-05,
+      "loss": 1.5667,
+      "step": 6455
+    },
+    {
+      "epoch": 0.6706138984107198,
+      "grad_norm": 0.45527634024620056,
+      "learning_rate": 2.4464974439247213e-05,
+      "loss": 1.7565,
+      "step": 6456
+    },
+    {
+      "epoch": 0.6707177729303002,
+      "grad_norm": 0.4044038951396942,
+      "learning_rate": 2.4450947485264442e-05,
+      "loss": 1.6019,
+      "step": 6457
+    },
+    {
+      "epoch": 0.6708216474498806,
+      "grad_norm": 0.4240105450153351,
+      "learning_rate": 2.4436923252056486e-05,
+      "loss": 1.6025,
+      "step": 6458
+    },
+    {
+      "epoch": 0.6709255219694609,
+      "grad_norm": 0.45421627163887024,
+      "learning_rate": 2.442290174111681e-05,
+      "loss": 1.8032,
+      "step": 6459
+    },
+    {
+      "epoch": 0.6710293964890413,
+      "grad_norm": 0.4415995478630066,
+      "learning_rate": 2.440888295393859e-05,
+      "loss": 1.844,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6711332710086216,
+      "grad_norm": 0.410188227891922,
+      "learning_rate": 2.4394866892014716e-05,
+      "loss": 1.5926,
+      "step": 6461
+    },
+    {
+      "epoch": 0.6712371455282019,
+      "grad_norm": 0.46723711490631104,
+      "learning_rate": 2.438085355683779e-05,
+      "loss": 1.5847,
+      "step": 6462
+    },
+    {
+      "epoch": 0.6713410200477823,
+      "grad_norm": 0.3974010944366455,
+      "learning_rate": 2.436684294990012e-05,
+      "loss": 1.5292,
+      "step": 6463
+    },
+    {
+      "epoch": 0.6714448945673627,
+      "grad_norm": 0.39809471368789673,
+      "learning_rate": 2.4352835072693713e-05,
+      "loss": 1.6891,
+      "step": 6464
+    },
+    {
+      "epoch": 0.6715487690869429,
+      "grad_norm": 0.45135897397994995,
+      "learning_rate": 2.4338829926710356e-05,
+      "loss": 1.8254,
+      "step": 6465
+    },
+    {
+      "epoch": 0.6716526436065233,
+      "grad_norm": 0.3903135061264038,
+      "learning_rate": 2.4324827513441406e-05,
+      "loss": 1.5315,
+      "step": 6466
+    },
+    {
+      "epoch": 0.6717565181261037,
+      "grad_norm": 0.47816312313079834,
+      "learning_rate": 2.431082783437807e-05,
+      "loss": 1.6046,
+      "step": 6467
+    },
+    {
+      "epoch": 0.671860392645684,
+      "grad_norm": 0.47781482338905334,
+      "learning_rate": 2.4296830891011206e-05,
+      "loss": 1.7201,
+      "step": 6468
+    },
+    {
+      "epoch": 0.6719642671652644,
+      "grad_norm": 0.4250834584236145,
+      "learning_rate": 2.4282836684831324e-05,
+      "loss": 1.5979,
+      "step": 6469
+    },
+    {
+      "epoch": 0.6720681416848447,
+      "grad_norm": 0.4348602592945099,
+      "learning_rate": 2.426884521732875e-05,
+      "loss": 1.6024,
+      "step": 6470
+    },
+    {
+      "epoch": 0.672172016204425,
+      "grad_norm": 0.3877941370010376,
+      "learning_rate": 2.425485648999345e-05,
+      "loss": 1.4421,
+      "step": 6471
+    },
+    {
+      "epoch": 0.6722758907240054,
+      "grad_norm": 0.4232386350631714,
+      "learning_rate": 2.4240870504315117e-05,
+      "loss": 1.7063,
+      "step": 6472
+    },
+    {
+      "epoch": 0.6723797652435858,
+      "grad_norm": 0.4069366455078125,
+      "learning_rate": 2.4226887261783143e-05,
+      "loss": 1.6469,
+      "step": 6473
+    },
+    {
+      "epoch": 0.672483639763166,
+      "grad_norm": 0.43410545587539673,
+      "learning_rate": 2.4212906763886638e-05,
+      "loss": 1.6914,
+      "step": 6474
+    },
+    {
+      "epoch": 0.6725875142827464,
+      "grad_norm": 0.40854349732398987,
+      "learning_rate": 2.4198929012114413e-05,
+      "loss": 1.5814,
+      "step": 6475
+    },
+    {
+      "epoch": 0.6726913888023268,
+      "grad_norm": 0.4242352247238159,
+      "learning_rate": 2.4184954007954978e-05,
+      "loss": 1.6704,
+      "step": 6476
+    },
+    {
+      "epoch": 0.6727952633219071,
+      "grad_norm": 0.44728976488113403,
+      "learning_rate": 2.417098175289661e-05,
+      "loss": 1.7766,
+      "step": 6477
+    },
+    {
+      "epoch": 0.6728991378414875,
+      "grad_norm": 0.4239303171634674,
+      "learning_rate": 2.4157012248427173e-05,
+      "loss": 1.5764,
+      "step": 6478
+    },
+    {
+      "epoch": 0.6730030123610679,
+      "grad_norm": 0.406534880399704,
+      "learning_rate": 2.414304549603437e-05,
+      "loss": 1.7778,
+      "step": 6479
+    },
+    {
+      "epoch": 0.6731068868806481,
+      "grad_norm": 0.43634507060050964,
+      "learning_rate": 2.4129081497205535e-05,
+      "loss": 1.7988,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6732107614002285,
+      "grad_norm": 0.4149050712585449,
+      "learning_rate": 2.4115120253427715e-05,
+      "loss": 1.636,
+      "step": 6481
+    },
+    {
+      "epoch": 0.6733146359198089,
+      "grad_norm": 0.46954065561294556,
+      "learning_rate": 2.4101161766187673e-05,
+      "loss": 1.6532,
+      "step": 6482
+    },
+    {
+      "epoch": 0.6734185104393893,
+      "grad_norm": 0.4280845820903778,
+      "learning_rate": 2.4087206036971872e-05,
+      "loss": 1.5258,
+      "step": 6483
+    },
+    {
+      "epoch": 0.6735223849589695,
+      "grad_norm": 0.40671125054359436,
+      "learning_rate": 2.4073253067266532e-05,
+      "loss": 1.6615,
+      "step": 6484
+    },
+    {
+      "epoch": 0.6736262594785499,
+      "grad_norm": 0.45035362243652344,
+      "learning_rate": 2.4059302858557475e-05,
+      "loss": 1.7412,
+      "step": 6485
+    },
+    {
+      "epoch": 0.6737301339981303,
+      "grad_norm": 0.4099656045436859,
+      "learning_rate": 2.4045355412330328e-05,
+      "loss": 1.6766,
+      "step": 6486
+    },
+    {
+      "epoch": 0.6738340085177106,
+      "grad_norm": 0.39857202768325806,
+      "learning_rate": 2.403141073007038e-05,
+      "loss": 1.6006,
+      "step": 6487
+    },
+    {
+      "epoch": 0.673937883037291,
+      "grad_norm": 0.4072171151638031,
+      "learning_rate": 2.401746881326263e-05,
+      "loss": 1.6916,
+      "step": 6488
+    },
+    {
+      "epoch": 0.6740417575568713,
+      "grad_norm": 0.4422532916069031,
+      "learning_rate": 2.400352966339177e-05,
+      "loss": 1.7378,
+      "step": 6489
+    },
+    {
+      "epoch": 0.6741456320764516,
+      "grad_norm": 0.41221001744270325,
+      "learning_rate": 2.398959328194223e-05,
+      "loss": 1.6959,
+      "step": 6490
+    },
+    {
+      "epoch": 0.674249506596032,
+      "grad_norm": 0.4334010183811188,
+      "learning_rate": 2.3975659670398117e-05,
+      "loss": 1.7088,
+      "step": 6491
+    },
+    {
+      "epoch": 0.6743533811156124,
+      "grad_norm": 0.4333952069282532,
+      "learning_rate": 2.3961728830243224e-05,
+      "loss": 1.4458,
+      "step": 6492
+    },
+    {
+      "epoch": 0.6744572556351927,
+      "grad_norm": 0.40676429867744446,
+      "learning_rate": 2.3947800762961147e-05,
+      "loss": 1.68,
+      "step": 6493
+    },
+    {
+      "epoch": 0.674561130154773,
+      "grad_norm": 0.39929771423339844,
+      "learning_rate": 2.3933875470035038e-05,
+      "loss": 1.4526,
+      "step": 6494
+    },
+    {
+      "epoch": 0.6746650046743534,
+      "grad_norm": 0.4088972508907318,
+      "learning_rate": 2.391995295294788e-05,
+      "loss": 1.6881,
+      "step": 6495
+    },
+    {
+      "epoch": 0.6747688791939337,
+      "grad_norm": 0.47918689250946045,
+      "learning_rate": 2.3906033213182328e-05,
+      "loss": 1.8923,
+      "step": 6496
+    },
+    {
+      "epoch": 0.6748727537135141,
+      "grad_norm": 0.41081228852272034,
+      "learning_rate": 2.3892116252220653e-05,
+      "loss": 1.6732,
+      "step": 6497
+    },
+    {
+      "epoch": 0.6749766282330945,
+      "grad_norm": 0.3840208649635315,
+      "learning_rate": 2.3878202071544977e-05,
+      "loss": 1.5638,
+      "step": 6498
+    },
+    {
+      "epoch": 0.6750805027526747,
+      "grad_norm": 0.43148073554039,
+      "learning_rate": 2.3864290672637012e-05,
+      "loss": 1.7186,
+      "step": 6499
+    },
+    {
+      "epoch": 0.6751843772722551,
+      "grad_norm": 0.42979419231414795,
+      "learning_rate": 2.3850382056978228e-05,
+      "loss": 1.621,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6752882517918355,
+      "grad_norm": 0.3895018696784973,
+      "learning_rate": 2.3836476226049775e-05,
+      "loss": 1.5422,
+      "step": 6501
+    },
+    {
+      "epoch": 0.6753921263114158,
+      "grad_norm": 0.511653482913971,
+      "learning_rate": 2.3822573181332526e-05,
+      "loss": 1.7677,
+      "step": 6502
+    },
+    {
+      "epoch": 0.6754960008309961,
+      "grad_norm": 0.43084853887557983,
+      "learning_rate": 2.380867292430704e-05,
+      "loss": 1.7901,
+      "step": 6503
+    },
+    {
+      "epoch": 0.6755998753505765,
+      "grad_norm": 0.404482364654541,
+      "learning_rate": 2.3794775456453573e-05,
+      "loss": 1.708,
+      "step": 6504
+    },
+    {
+      "epoch": 0.6757037498701568,
+      "grad_norm": 0.4159182906150818,
+      "learning_rate": 2.378088077925214e-05,
+      "loss": 1.5471,
+      "step": 6505
+    },
+    {
+      "epoch": 0.6758076243897372,
+      "grad_norm": 0.4247626066207886,
+      "learning_rate": 2.3766988894182357e-05,
+      "loss": 1.8363,
+      "step": 6506
+    },
+    {
+      "epoch": 0.6759114989093176,
+      "grad_norm": 0.47585469484329224,
+      "learning_rate": 2.3753099802723643e-05,
+      "loss": 2.0124,
+      "step": 6507
+    },
+    {
+      "epoch": 0.6760153734288978,
+      "grad_norm": 0.44073712825775146,
+      "learning_rate": 2.3739213506355073e-05,
+      "loss": 1.698,
+      "step": 6508
+    },
+    {
+      "epoch": 0.6761192479484782,
+      "grad_norm": 0.4209062159061432,
+      "learning_rate": 2.372533000655542e-05,
+      "loss": 1.8448,
+      "step": 6509
+    },
+    {
+      "epoch": 0.6762231224680586,
+      "grad_norm": 0.40654614567756653,
+      "learning_rate": 2.3711449304803175e-05,
+      "loss": 1.6932,
+      "step": 6510
+    },
+    {
+      "epoch": 0.676326996987639,
+      "grad_norm": 0.42411383986473083,
+      "learning_rate": 2.3697571402576497e-05,
+      "loss": 1.6501,
+      "step": 6511
+    },
+    {
+      "epoch": 0.6764308715072193,
+      "grad_norm": 0.4248904883861542,
+      "learning_rate": 2.3683696301353337e-05,
+      "loss": 1.4518,
+      "step": 6512
+    },
+    {
+      "epoch": 0.6765347460267996,
+      "grad_norm": 0.39044690132141113,
+      "learning_rate": 2.3669824002611207e-05,
+      "loss": 1.6496,
+      "step": 6513
+    },
+    {
+      "epoch": 0.67663862054638,
+      "grad_norm": 0.4223690629005432,
+      "learning_rate": 2.3655954507827456e-05,
+      "loss": 1.6076,
+      "step": 6514
+    },
+    {
+      "epoch": 0.6767424950659603,
+      "grad_norm": 0.44539690017700195,
+      "learning_rate": 2.364208781847906e-05,
+      "loss": 1.7109,
+      "step": 6515
+    },
+    {
+      "epoch": 0.6768463695855407,
+      "grad_norm": 0.4467482566833496,
+      "learning_rate": 2.3628223936042708e-05,
+      "loss": 1.6478,
+      "step": 6516
+    },
+    {
+      "epoch": 0.6769502441051211,
+      "grad_norm": 0.3779411315917969,
+      "learning_rate": 2.3614362861994803e-05,
+      "loss": 1.5394,
+      "step": 6517
+    },
+    {
+      "epoch": 0.6770541186247013,
+      "grad_norm": 0.44637593626976013,
+      "learning_rate": 2.3600504597811436e-05,
+      "loss": 1.8076,
+      "step": 6518
+    },
+    {
+      "epoch": 0.6771579931442817,
+      "grad_norm": 0.43689000606536865,
+      "learning_rate": 2.3586649144968404e-05,
+      "loss": 1.718,
+      "step": 6519
+    },
+    {
+      "epoch": 0.6772618676638621,
+      "grad_norm": 0.4006705582141876,
+      "learning_rate": 2.357279650494119e-05,
+      "loss": 1.4616,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6773657421834424,
+      "grad_norm": 0.42330577969551086,
+      "learning_rate": 2.3558946679205046e-05,
+      "loss": 1.6517,
+      "step": 6521
+    },
+    {
+      "epoch": 0.6774696167030227,
+      "grad_norm": 0.41748082637786865,
+      "learning_rate": 2.3545099669234792e-05,
+      "loss": 1.5593,
+      "step": 6522
+    },
+    {
+      "epoch": 0.6775734912226031,
+      "grad_norm": 0.46182626485824585,
+      "learning_rate": 2.3531255476505086e-05,
+      "loss": 1.7158,
+      "step": 6523
+    },
+    {
+      "epoch": 0.6776773657421834,
+      "grad_norm": 0.444038450717926,
+      "learning_rate": 2.3517414102490232e-05,
+      "loss": 1.8144,
+      "step": 6524
+    },
+    {
+      "epoch": 0.6777812402617638,
+      "grad_norm": 0.46683838963508606,
+      "learning_rate": 2.350357554866416e-05,
+      "loss": 1.8248,
+      "step": 6525
+    },
+    {
+      "epoch": 0.6778851147813442,
+      "grad_norm": 0.4560922384262085,
+      "learning_rate": 2.3489739816500635e-05,
+      "loss": 1.7912,
+      "step": 6526
+    },
+    {
+      "epoch": 0.6779889893009244,
+      "grad_norm": 0.4331681430339813,
+      "learning_rate": 2.3475906907473027e-05,
+      "loss": 1.7446,
+      "step": 6527
+    },
+    {
+      "epoch": 0.6780928638205048,
+      "grad_norm": 0.42912784218788147,
+      "learning_rate": 2.3462076823054442e-05,
+      "loss": 1.6939,
+      "step": 6528
+    },
+    {
+      "epoch": 0.6781967383400852,
+      "grad_norm": 0.451789915561676,
+      "learning_rate": 2.3448249564717666e-05,
+      "loss": 1.7973,
+      "step": 6529
+    },
+    {
+      "epoch": 0.6783006128596655,
+      "grad_norm": 0.4173526465892792,
+      "learning_rate": 2.3434425133935207e-05,
+      "loss": 1.6438,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6784044873792459,
+      "grad_norm": 0.4266951084136963,
+      "learning_rate": 2.3420603532179246e-05,
+      "loss": 1.7301,
+      "step": 6531
+    },
+    {
+      "epoch": 0.6785083618988262,
+      "grad_norm": 0.48929300904273987,
+      "learning_rate": 2.3406784760921667e-05,
+      "loss": 1.7364,
+      "step": 6532
+    },
+    {
+      "epoch": 0.6786122364184065,
+      "grad_norm": 0.3748891055583954,
+      "learning_rate": 2.339296882163411e-05,
+      "loss": 1.608,
+      "step": 6533
+    },
+    {
+      "epoch": 0.6787161109379869,
+      "grad_norm": 0.42468810081481934,
+      "learning_rate": 2.3379155715787792e-05,
+      "loss": 1.8397,
+      "step": 6534
+    },
+    {
+      "epoch": 0.6788199854575673,
+      "grad_norm": 0.46891170740127563,
+      "learning_rate": 2.3365345444853763e-05,
+      "loss": 1.9061,
+      "step": 6535
+    },
+    {
+      "epoch": 0.6789238599771477,
+      "grad_norm": 0.42651164531707764,
+      "learning_rate": 2.3351538010302688e-05,
+      "loss": 1.6904,
+      "step": 6536
+    },
+    {
+      "epoch": 0.6790277344967279,
+      "grad_norm": 0.41323697566986084,
+      "learning_rate": 2.333773341360494e-05,
+      "loss": 1.6176,
+      "step": 6537
+    },
+    {
+      "epoch": 0.6791316090163083,
+      "grad_norm": 0.4378218948841095,
+      "learning_rate": 2.3323931656230614e-05,
+      "loss": 1.7018,
+      "step": 6538
+    },
+    {
+      "epoch": 0.6792354835358887,
+      "grad_norm": 0.42341065406799316,
+      "learning_rate": 2.331013273964947e-05,
+      "loss": 1.6211,
+      "step": 6539
+    },
+    {
+      "epoch": 0.679339358055469,
+      "grad_norm": 0.37841764092445374,
+      "learning_rate": 2.3296336665331032e-05,
+      "loss": 1.6159,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6794432325750493,
+      "grad_norm": 0.38354504108428955,
+      "learning_rate": 2.3282543434744407e-05,
+      "loss": 1.6425,
+      "step": 6541
+    },
+    {
+      "epoch": 0.6795471070946297,
+      "grad_norm": 0.4210819900035858,
+      "learning_rate": 2.3268753049358522e-05,
+      "loss": 1.6235,
+      "step": 6542
+    },
+    {
+      "epoch": 0.67965098161421,
+      "grad_norm": 0.4259316623210907,
+      "learning_rate": 2.3254965510641928e-05,
+      "loss": 1.8103,
+      "step": 6543
+    },
+    {
+      "epoch": 0.6797548561337904,
+      "grad_norm": 0.3991251289844513,
+      "learning_rate": 2.3241180820062885e-05,
+      "loss": 1.6595,
+      "step": 6544
+    },
+    {
+      "epoch": 0.6798587306533708,
+      "grad_norm": 0.3932139575481415,
+      "learning_rate": 2.3227398979089353e-05,
+      "loss": 1.586,
+      "step": 6545
+    },
+    {
+      "epoch": 0.679962605172951,
+      "grad_norm": 0.40890079736709595,
+      "learning_rate": 2.3213619989189006e-05,
+      "loss": 1.4932,
+      "step": 6546
+    },
+    {
+      "epoch": 0.6800664796925314,
+      "grad_norm": 0.42647114396095276,
+      "learning_rate": 2.319984385182918e-05,
+      "loss": 1.6931,
+      "step": 6547
+    },
+    {
+      "epoch": 0.6801703542121118,
+      "grad_norm": 0.39702990651130676,
+      "learning_rate": 2.318607056847692e-05,
+      "loss": 1.5665,
+      "step": 6548
+    },
+    {
+      "epoch": 0.6802742287316921,
+      "grad_norm": 0.3964666724205017,
+      "learning_rate": 2.3172300140599024e-05,
+      "loss": 1.7028,
+      "step": 6549
+    },
+    {
+      "epoch": 0.6803781032512725,
+      "grad_norm": 0.41276541352272034,
+      "learning_rate": 2.315853256966186e-05,
+      "loss": 1.5754,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6804819777708528,
+      "grad_norm": 0.40658214688301086,
+      "learning_rate": 2.314476785713163e-05,
+      "loss": 1.8211,
+      "step": 6551
+    },
+    {
+      "epoch": 0.6805858522904331,
+      "grad_norm": 0.43847519159317017,
+      "learning_rate": 2.3131006004474152e-05,
+      "loss": 1.5858,
+      "step": 6552
+    },
+    {
+      "epoch": 0.6806897268100135,
+      "grad_norm": 0.5137412548065186,
+      "learning_rate": 2.3117247013154918e-05,
+      "loss": 1.9818,
+      "step": 6553
+    },
+    {
+      "epoch": 0.6807936013295939,
+      "grad_norm": 0.41750749945640564,
+      "learning_rate": 2.31034908846392e-05,
+      "loss": 1.519,
+      "step": 6554
+    },
+    {
+      "epoch": 0.6808974758491741,
+      "grad_norm": 0.3900861144065857,
+      "learning_rate": 2.3089737620391906e-05,
+      "loss": 1.6836,
+      "step": 6555
+    },
+    {
+      "epoch": 0.6810013503687545,
+      "grad_norm": 0.4610309302806854,
+      "learning_rate": 2.307598722187765e-05,
+      "loss": 1.6183,
+      "step": 6556
+    },
+    {
+      "epoch": 0.6811052248883349,
+      "grad_norm": 0.4481363892555237,
+      "learning_rate": 2.306223969056073e-05,
+      "loss": 1.7593,
+      "step": 6557
+    },
+    {
+      "epoch": 0.6812090994079152,
+      "grad_norm": 0.37722769379615784,
+      "learning_rate": 2.3048495027905166e-05,
+      "loss": 1.5302,
+      "step": 6558
+    },
+    {
+      "epoch": 0.6813129739274956,
+      "grad_norm": 0.42121073603630066,
+      "learning_rate": 2.303475323537465e-05,
+      "loss": 1.5051,
+      "step": 6559
+    },
+    {
+      "epoch": 0.681416848447076,
+      "grad_norm": 0.49709874391555786,
+      "learning_rate": 2.302101431443256e-05,
+      "loss": 1.9087,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6815207229666563,
+      "grad_norm": 0.4022544026374817,
+      "learning_rate": 2.300727826654202e-05,
+      "loss": 1.6909,
+      "step": 6561
+    },
+    {
+      "epoch": 0.6816245974862366,
+      "grad_norm": 0.41728565096855164,
+      "learning_rate": 2.299354509316579e-05,
+      "loss": 1.6303,
+      "step": 6562
+    },
+    {
+      "epoch": 0.681728472005817,
+      "grad_norm": 0.439998060464859,
+      "learning_rate": 2.297981479576635e-05,
+      "loss": 1.8708,
+      "step": 6563
+    },
+    {
+      "epoch": 0.6818323465253974,
+      "grad_norm": 0.39185917377471924,
+      "learning_rate": 2.2966087375805874e-05,
+      "loss": 1.6224,
+      "step": 6564
+    },
+    {
+      "epoch": 0.6819362210449776,
+      "grad_norm": 0.38439640402793884,
+      "learning_rate": 2.2952362834746216e-05,
+      "loss": 1.5744,
+      "step": 6565
+    },
+    {
+      "epoch": 0.682040095564558,
+      "grad_norm": 0.4013480544090271,
+      "learning_rate": 2.2938641174048936e-05,
+      "loss": 1.5341,
+      "step": 6566
+    },
+    {
+      "epoch": 0.6821439700841384,
+      "grad_norm": 0.44190388917922974,
+      "learning_rate": 2.2924922395175268e-05,
+      "loss": 1.8871,
+      "step": 6567
+    },
+    {
+      "epoch": 0.6822478446037187,
+      "grad_norm": 0.4578564167022705,
+      "learning_rate": 2.2911206499586207e-05,
+      "loss": 1.8006,
+      "step": 6568
+    },
+    {
+      "epoch": 0.6823517191232991,
+      "grad_norm": 0.44078001379966736,
+      "learning_rate": 2.2897493488742317e-05,
+      "loss": 1.6603,
+      "step": 6569
+    },
+    {
+      "epoch": 0.6824555936428794,
+      "grad_norm": 0.4083844721317291,
+      "learning_rate": 2.2883783364103976e-05,
+      "loss": 1.5109,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6825594681624597,
+      "grad_norm": 0.4550290107727051,
+      "learning_rate": 2.2870076127131196e-05,
+      "loss": 1.7607,
+      "step": 6571
+    },
+    {
+      "epoch": 0.6826633426820401,
+      "grad_norm": 0.4314436912536621,
+      "learning_rate": 2.2856371779283692e-05,
+      "loss": 1.7517,
+      "step": 6572
+    },
+    {
+      "epoch": 0.6827672172016205,
+      "grad_norm": 0.4039313793182373,
+      "learning_rate": 2.2842670322020858e-05,
+      "loss": 1.7537,
+      "step": 6573
+    },
+    {
+      "epoch": 0.6828710917212008,
+      "grad_norm": 0.40721437335014343,
+      "learning_rate": 2.2828971756801802e-05,
+      "loss": 1.5869,
+      "step": 6574
+    },
+    {
+      "epoch": 0.6829749662407811,
+      "grad_norm": 0.38915568590164185,
+      "learning_rate": 2.2815276085085313e-05,
+      "loss": 1.5199,
+      "step": 6575
+    },
+    {
+      "epoch": 0.6830788407603615,
+      "grad_norm": 0.42462047934532166,
+      "learning_rate": 2.2801583308329848e-05,
+      "loss": 1.6698,
+      "step": 6576
+    },
+    {
+      "epoch": 0.6831827152799418,
+      "grad_norm": 0.42079266905784607,
+      "learning_rate": 2.2787893427993646e-05,
+      "loss": 1.582,
+      "step": 6577
+    },
+    {
+      "epoch": 0.6832865897995222,
+      "grad_norm": 0.45003542304039,
+      "learning_rate": 2.2774206445534495e-05,
+      "loss": 1.4429,
+      "step": 6578
+    },
+    {
+      "epoch": 0.6833904643191026,
+      "grad_norm": 0.38841044902801514,
+      "learning_rate": 2.2760522362410003e-05,
+      "loss": 1.5655,
+      "step": 6579
+    },
+    {
+      "epoch": 0.6834943388386828,
+      "grad_norm": 0.39281266927719116,
+      "learning_rate": 2.274684118007743e-05,
+      "loss": 1.6493,
+      "step": 6580
+    },
+    {
+      "epoch": 0.6835982133582632,
+      "grad_norm": 0.37398335337638855,
+      "learning_rate": 2.2733162899993642e-05,
+      "loss": 1.4814,
+      "step": 6581
+    },
+    {
+      "epoch": 0.6837020878778436,
+      "grad_norm": 0.43852871656417847,
+      "learning_rate": 2.271948752361534e-05,
+      "loss": 1.5412,
+      "step": 6582
+    },
+    {
+      "epoch": 0.6838059623974239,
+      "grad_norm": 0.4092741012573242,
+      "learning_rate": 2.2705815052398816e-05,
+      "loss": 1.6233,
+      "step": 6583
+    },
+    {
+      "epoch": 0.6839098369170042,
+      "grad_norm": 0.3938494026660919,
+      "learning_rate": 2.2692145487800093e-05,
+      "loss": 1.6233,
+      "step": 6584
+    },
+    {
+      "epoch": 0.6840137114365846,
+      "grad_norm": 0.5273641347885132,
+      "learning_rate": 2.2678478831274858e-05,
+      "loss": 1.8792,
+      "step": 6585
+    },
+    {
+      "epoch": 0.6841175859561649,
+      "grad_norm": 0.41497117280960083,
+      "learning_rate": 2.266481508427852e-05,
+      "loss": 1.6535,
+      "step": 6586
+    },
+    {
+      "epoch": 0.6842214604757453,
+      "grad_norm": 0.39837929606437683,
+      "learning_rate": 2.2651154248266145e-05,
+      "loss": 1.5587,
+      "step": 6587
+    },
+    {
+      "epoch": 0.6843253349953257,
+      "grad_norm": 0.504227340221405,
+      "learning_rate": 2.2637496324692496e-05,
+      "loss": 1.6661,
+      "step": 6588
+    },
+    {
+      "epoch": 0.684429209514906,
+      "grad_norm": 0.425836443901062,
+      "learning_rate": 2.2623841315012072e-05,
+      "loss": 1.7157,
+      "step": 6589
+    },
+    {
+      "epoch": 0.6845330840344863,
+      "grad_norm": 0.429581880569458,
+      "learning_rate": 2.2610189220679007e-05,
+      "loss": 1.7036,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6846369585540667,
+      "grad_norm": 0.4043353497982025,
+      "learning_rate": 2.2596540043147142e-05,
+      "loss": 1.6208,
+      "step": 6591
+    },
+    {
+      "epoch": 0.6847408330736471,
+      "grad_norm": 0.423662006855011,
+      "learning_rate": 2.2582893783870006e-05,
+      "loss": 1.8271,
+      "step": 6592
+    },
+    {
+      "epoch": 0.6848447075932274,
+      "grad_norm": 0.3973560631275177,
+      "learning_rate": 2.256925044430082e-05,
+      "loss": 1.638,
+      "step": 6593
+    },
+    {
+      "epoch": 0.6849485821128077,
+      "grad_norm": 0.41801849007606506,
+      "learning_rate": 2.2555610025892498e-05,
+      "loss": 1.5826,
+      "step": 6594
+    },
+    {
+      "epoch": 0.6850524566323881,
+      "grad_norm": 0.45296528935432434,
+      "learning_rate": 2.254197253009761e-05,
+      "loss": 1.5875,
+      "step": 6595
+    },
+    {
+      "epoch": 0.6851563311519684,
+      "grad_norm": 0.4170701205730438,
+      "learning_rate": 2.2528337958368507e-05,
+      "loss": 1.6933,
+      "step": 6596
+    },
+    {
+      "epoch": 0.6852602056715488,
+      "grad_norm": 0.3868260979652405,
+      "learning_rate": 2.2514706312157085e-05,
+      "loss": 1.5773,
+      "step": 6597
+    },
+    {
+      "epoch": 0.6853640801911292,
+      "grad_norm": 0.3958593010902405,
+      "learning_rate": 2.2501077592915065e-05,
+      "loss": 1.6426,
+      "step": 6598
+    },
+    {
+      "epoch": 0.6854679547107094,
+      "grad_norm": 0.3819239139556885,
+      "learning_rate": 2.2487451802093784e-05,
+      "loss": 1.5363,
+      "step": 6599
+    },
+    {
+      "epoch": 0.6855718292302898,
+      "grad_norm": 0.4158235192298889,
+      "learning_rate": 2.2473828941144276e-05,
+      "loss": 1.5895,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6856757037498702,
+      "grad_norm": 0.4286392331123352,
+      "learning_rate": 2.246020901151728e-05,
+      "loss": 1.6797,
+      "step": 6601
+    },
+    {
+      "epoch": 0.6857795782694505,
+      "grad_norm": 0.39033573865890503,
+      "learning_rate": 2.24465920146632e-05,
+      "loss": 1.6592,
+      "step": 6602
+    },
+    {
+      "epoch": 0.6858834527890308,
+      "grad_norm": 0.47178182005882263,
+      "learning_rate": 2.2432977952032148e-05,
+      "loss": 1.7151,
+      "step": 6603
+    },
+    {
+      "epoch": 0.6859873273086112,
+      "grad_norm": 0.4451124370098114,
+      "learning_rate": 2.2419366825073894e-05,
+      "loss": 1.8579,
+      "step": 6604
+    },
+    {
+      "epoch": 0.6860912018281915,
+      "grad_norm": 0.4096846878528595,
+      "learning_rate": 2.2405758635237973e-05,
+      "loss": 1.57,
+      "step": 6605
+    },
+    {
+      "epoch": 0.6861950763477719,
+      "grad_norm": 0.437682569026947,
+      "learning_rate": 2.2392153383973473e-05,
+      "loss": 1.7868,
+      "step": 6606
+    },
+    {
+      "epoch": 0.6862989508673523,
+      "grad_norm": 0.4488213062286377,
+      "learning_rate": 2.2378551072729314e-05,
+      "loss": 1.7668,
+      "step": 6607
+    },
+    {
+      "epoch": 0.6864028253869325,
+      "grad_norm": 0.4229215383529663,
+      "learning_rate": 2.2364951702954025e-05,
+      "loss": 1.7325,
+      "step": 6608
+    },
+    {
+      "epoch": 0.6865066999065129,
+      "grad_norm": 0.45291194319725037,
+      "learning_rate": 2.235135527609578e-05,
+      "loss": 1.8742,
+      "step": 6609
+    },
+    {
+      "epoch": 0.6866105744260933,
+      "grad_norm": 0.41478431224823,
+      "learning_rate": 2.233776179360255e-05,
+      "loss": 1.5874,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6867144489456736,
+      "grad_norm": 0.4467930495738983,
+      "learning_rate": 2.2324171256921918e-05,
+      "loss": 1.8026,
+      "step": 6611
+    },
+    {
+      "epoch": 0.686818323465254,
+      "grad_norm": 0.4063619375228882,
+      "learning_rate": 2.2310583667501163e-05,
+      "loss": 1.68,
+      "step": 6612
+    },
+    {
+      "epoch": 0.6869221979848343,
+      "grad_norm": 0.4245857000350952,
+      "learning_rate": 2.2296999026787262e-05,
+      "loss": 1.5025,
+      "step": 6613
+    },
+    {
+      "epoch": 0.6870260725044147,
+      "grad_norm": 0.44443392753601074,
+      "learning_rate": 2.2283417336226853e-05,
+      "loss": 1.6217,
+      "step": 6614
+    },
+    {
+      "epoch": 0.687129947023995,
+      "grad_norm": 0.4405082166194916,
+      "learning_rate": 2.2269838597266342e-05,
+      "loss": 1.5827,
+      "step": 6615
+    },
+    {
+      "epoch": 0.6872338215435754,
+      "grad_norm": 0.4452870190143585,
+      "learning_rate": 2.2256262811351676e-05,
+      "loss": 1.7894,
+      "step": 6616
+    },
+    {
+      "epoch": 0.6873376960631558,
+      "grad_norm": 0.37524208426475525,
+      "learning_rate": 2.2242689979928632e-05,
+      "loss": 1.5107,
+      "step": 6617
+    },
+    {
+      "epoch": 0.687441570582736,
+      "grad_norm": 0.4177568554878235,
+      "learning_rate": 2.222912010444259e-05,
+      "loss": 1.6793,
+      "step": 6618
+    },
+    {
+      "epoch": 0.6875454451023164,
+      "grad_norm": 0.39872753620147705,
+      "learning_rate": 2.2215553186338635e-05,
+      "loss": 1.6067,
+      "step": 6619
+    },
+    {
+      "epoch": 0.6876493196218968,
+      "grad_norm": 0.40704911947250366,
+      "learning_rate": 2.2201989227061535e-05,
+      "loss": 1.6807,
+      "step": 6620
+    },
+    {
+      "epoch": 0.6877531941414771,
+      "grad_norm": 0.43390125036239624,
+      "learning_rate": 2.2188428228055756e-05,
+      "loss": 1.5434,
+      "step": 6621
+    },
+    {
+      "epoch": 0.6878570686610574,
+      "grad_norm": 0.4030878245830536,
+      "learning_rate": 2.2174870190765428e-05,
+      "loss": 1.6909,
+      "step": 6622
+    },
+    {
+      "epoch": 0.6879609431806378,
+      "grad_norm": 0.45585277676582336,
+      "learning_rate": 2.2161315116634362e-05,
+      "loss": 1.7252,
+      "step": 6623
+    },
+    {
+      "epoch": 0.6880648177002181,
+      "grad_norm": 0.4133506715297699,
+      "learning_rate": 2.2147763007106126e-05,
+      "loss": 1.7813,
+      "step": 6624
+    },
+    {
+      "epoch": 0.6881686922197985,
+      "grad_norm": 0.45067742466926575,
+      "learning_rate": 2.213421386362383e-05,
+      "loss": 1.7338,
+      "step": 6625
+    },
+    {
+      "epoch": 0.6882725667393789,
+      "grad_norm": 0.4103121757507324,
+      "learning_rate": 2.212066768763042e-05,
+      "loss": 1.7215,
+      "step": 6626
+    },
+    {
+      "epoch": 0.6883764412589591,
+      "grad_norm": 0.43877947330474854,
+      "learning_rate": 2.2107124480568435e-05,
+      "loss": 1.5899,
+      "step": 6627
+    },
+    {
+      "epoch": 0.6884803157785395,
+      "grad_norm": 0.3948143720626831,
+      "learning_rate": 2.2093584243880116e-05,
+      "loss": 1.6213,
+      "step": 6628
+    },
+    {
+      "epoch": 0.6885841902981199,
+      "grad_norm": 0.39642539620399475,
+      "learning_rate": 2.20800469790074e-05,
+      "loss": 1.5516,
+      "step": 6629
+    },
+    {
+      "epoch": 0.6886880648177002,
+      "grad_norm": 0.4189082980155945,
+      "learning_rate": 2.2066512687391898e-05,
+      "loss": 1.6291,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6887919393372806,
+      "grad_norm": 0.48019567131996155,
+      "learning_rate": 2.20529813704749e-05,
+      "loss": 1.819,
+      "step": 6631
+    },
+    {
+      "epoch": 0.6888958138568609,
+      "grad_norm": 0.43770167231559753,
+      "learning_rate": 2.2039453029697378e-05,
+      "loss": 1.7297,
+      "step": 6632
+    },
+    {
+      "epoch": 0.6889996883764412,
+      "grad_norm": 0.3850918114185333,
+      "learning_rate": 2.202592766650004e-05,
+      "loss": 1.6097,
+      "step": 6633
+    },
+    {
+      "epoch": 0.6891035628960216,
+      "grad_norm": 0.4326091408729553,
+      "learning_rate": 2.2012405282323166e-05,
+      "loss": 1.4958,
+      "step": 6634
+    },
+    {
+      "epoch": 0.689207437415602,
+      "grad_norm": 0.3961119055747986,
+      "learning_rate": 2.1998885878606835e-05,
+      "loss": 1.5057,
+      "step": 6635
+    },
+    {
+      "epoch": 0.6893113119351822,
+      "grad_norm": 0.4198208153247833,
+      "learning_rate": 2.1985369456790757e-05,
+      "loss": 1.614,
+      "step": 6636
+    },
+    {
+      "epoch": 0.6894151864547626,
+      "grad_norm": 0.4884597659111023,
+      "learning_rate": 2.1971856018314274e-05,
+      "loss": 1.8599,
+      "step": 6637
+    },
+    {
+      "epoch": 0.689519060974343,
+      "grad_norm": 0.42158132791519165,
+      "learning_rate": 2.1958345564616515e-05,
+      "loss": 1.6569,
+      "step": 6638
+    },
+    {
+      "epoch": 0.6896229354939234,
+      "grad_norm": 0.41984039545059204,
+      "learning_rate": 2.1944838097136205e-05,
+      "loss": 1.6422,
+      "step": 6639
+    },
+    {
+      "epoch": 0.6897268100135037,
+      "grad_norm": 0.42424213886260986,
+      "learning_rate": 2.1931333617311833e-05,
+      "loss": 1.8098,
+      "step": 6640
+    },
+    {
+      "epoch": 0.689830684533084,
+      "grad_norm": 0.3981630206108093,
+      "learning_rate": 2.1917832126581467e-05,
+      "loss": 1.6439,
+      "step": 6641
+    },
+    {
+      "epoch": 0.6899345590526644,
+      "grad_norm": 0.42164191603660583,
+      "learning_rate": 2.1904333626382918e-05,
+      "loss": 1.6079,
+      "step": 6642
+    },
+    {
+      "epoch": 0.6900384335722447,
+      "grad_norm": 0.43774572014808655,
+      "learning_rate": 2.189083811815371e-05,
+      "loss": 1.6243,
+      "step": 6643
+    },
+    {
+      "epoch": 0.6901423080918251,
+      "grad_norm": 0.4788624942302704,
+      "learning_rate": 2.187734560333095e-05,
+      "loss": 1.8424,
+      "step": 6644
+    },
+    {
+      "epoch": 0.6902461826114055,
+      "grad_norm": 0.3973691761493683,
+      "learning_rate": 2.186385608335154e-05,
+      "loss": 1.5472,
+      "step": 6645
+    },
+    {
+      "epoch": 0.6903500571309857,
+      "grad_norm": 0.4038008451461792,
+      "learning_rate": 2.1850369559651984e-05,
+      "loss": 1.5287,
+      "step": 6646
+    },
+    {
+      "epoch": 0.6904539316505661,
+      "grad_norm": 0.38313379883766174,
+      "learning_rate": 2.1836886033668495e-05,
+      "loss": 1.5671,
+      "step": 6647
+    },
+    {
+      "epoch": 0.6905578061701465,
+      "grad_norm": 0.43707215785980225,
+      "learning_rate": 2.1823405506836962e-05,
+      "loss": 1.6821,
+      "step": 6648
+    },
+    {
+      "epoch": 0.6906616806897268,
+      "grad_norm": 0.40973857045173645,
+      "learning_rate": 2.180992798059296e-05,
+      "loss": 1.6621,
+      "step": 6649
+    },
+    {
+      "epoch": 0.6907655552093072,
+      "grad_norm": 0.3938702344894409,
+      "learning_rate": 2.1796453456371735e-05,
+      "loss": 1.5388,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6908694297288875,
+      "grad_norm": 0.3766863942146301,
+      "learning_rate": 2.17829819356082e-05,
+      "loss": 1.5155,
+      "step": 6651
+    },
+    {
+      "epoch": 0.6909733042484678,
+      "grad_norm": 0.4006540775299072,
+      "learning_rate": 2.1769513419737024e-05,
+      "loss": 1.768,
+      "step": 6652
+    },
+    {
+      "epoch": 0.6910771787680482,
+      "grad_norm": 0.4285779297351837,
+      "learning_rate": 2.1756047910192428e-05,
+      "loss": 1.6215,
+      "step": 6653
+    },
+    {
+      "epoch": 0.6911810532876286,
+      "grad_norm": 0.46323853731155396,
+      "learning_rate": 2.1742585408408438e-05,
+      "loss": 1.7715,
+      "step": 6654
+    },
+    {
+      "epoch": 0.6912849278072088,
+      "grad_norm": 0.46970510482788086,
+      "learning_rate": 2.1729125915818684e-05,
+      "loss": 1.7967,
+      "step": 6655
+    },
+    {
+      "epoch": 0.6913888023267892,
+      "grad_norm": 0.39832404255867004,
+      "learning_rate": 2.17156694338565e-05,
+      "loss": 1.6482,
+      "step": 6656
+    },
+    {
+      "epoch": 0.6914926768463696,
+      "grad_norm": 0.42489683628082275,
+      "learning_rate": 2.1702215963954893e-05,
+      "loss": 1.6526,
+      "step": 6657
+    },
+    {
+      "epoch": 0.6915965513659499,
+      "grad_norm": 0.39924055337905884,
+      "learning_rate": 2.1688765507546556e-05,
+      "loss": 1.3809,
+      "step": 6658
+    },
+    {
+      "epoch": 0.6917004258855303,
+      "grad_norm": 0.4202609062194824,
+      "learning_rate": 2.1675318066063855e-05,
+      "loss": 1.6426,
+      "step": 6659
+    },
+    {
+      "epoch": 0.6918043004051107,
+      "grad_norm": 0.42206838726997375,
+      "learning_rate": 2.166187364093882e-05,
+      "loss": 1.6819,
+      "step": 6660
+    },
+    {
+      "epoch": 0.6919081749246909,
+      "grad_norm": 0.4156840443611145,
+      "learning_rate": 2.1648432233603232e-05,
+      "loss": 1.8221,
+      "step": 6661
+    },
+    {
+      "epoch": 0.6920120494442713,
+      "grad_norm": 0.41110360622406006,
+      "learning_rate": 2.1634993845488426e-05,
+      "loss": 1.707,
+      "step": 6662
+    },
+    {
+      "epoch": 0.6921159239638517,
+      "grad_norm": 0.4039273262023926,
+      "learning_rate": 2.162155847802554e-05,
+      "loss": 1.6048,
+      "step": 6663
+    },
+    {
+      "epoch": 0.6922197984834321,
+      "grad_norm": 0.4477052688598633,
+      "learning_rate": 2.1608126132645317e-05,
+      "loss": 1.8689,
+      "step": 6664
+    },
+    {
+      "epoch": 0.6923236730030123,
+      "grad_norm": 0.40778693556785583,
+      "learning_rate": 2.1594696810778193e-05,
+      "loss": 1.7781,
+      "step": 6665
+    },
+    {
+      "epoch": 0.6924275475225927,
+      "grad_norm": 0.42754799127578735,
+      "learning_rate": 2.1581270513854283e-05,
+      "loss": 1.71,
+      "step": 6666
+    },
+    {
+      "epoch": 0.6925314220421731,
+      "grad_norm": 0.4134778380393982,
+      "learning_rate": 2.156784724330338e-05,
+      "loss": 1.7161,
+      "step": 6667
+    },
+    {
+      "epoch": 0.6926352965617534,
+      "grad_norm": 0.40109783411026,
+      "learning_rate": 2.1554427000555e-05,
+      "loss": 1.5071,
+      "step": 6668
+    },
+    {
+      "epoch": 0.6927391710813338,
+      "grad_norm": 0.49069786071777344,
+      "learning_rate": 2.154100978703824e-05,
+      "loss": 1.9258,
+      "step": 6669
+    },
+    {
+      "epoch": 0.6928430456009141,
+      "grad_norm": 0.4370437562465668,
+      "learning_rate": 2.1527595604181933e-05,
+      "loss": 1.4963,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6929469201204944,
+      "grad_norm": 0.45224136114120483,
+      "learning_rate": 2.151418445341463e-05,
+      "loss": 1.745,
+      "step": 6671
+    },
+    {
+      "epoch": 0.6930507946400748,
+      "grad_norm": 0.4150208532810211,
+      "learning_rate": 2.1500776336164456e-05,
+      "loss": 1.6489,
+      "step": 6672
+    },
+    {
+      "epoch": 0.6931546691596552,
+      "grad_norm": 0.4757159352302551,
+      "learning_rate": 2.1487371253859316e-05,
+      "loss": 1.8266,
+      "step": 6673
+    },
+    {
+      "epoch": 0.6932585436792355,
+      "grad_norm": 0.3998592793941498,
+      "learning_rate": 2.1473969207926737e-05,
+      "loss": 1.6169,
+      "step": 6674
+    },
+    {
+      "epoch": 0.6933624181988158,
+      "grad_norm": 0.42257049679756165,
+      "learning_rate": 2.1460570199793917e-05,
+      "loss": 1.7495,
+      "step": 6675
+    },
+    {
+      "epoch": 0.6934662927183962,
+      "grad_norm": 0.3936207592487335,
+      "learning_rate": 2.1447174230887757e-05,
+      "loss": 1.7681,
+      "step": 6676
+    },
+    {
+      "epoch": 0.6935701672379765,
+      "grad_norm": 0.4289858639240265,
+      "learning_rate": 2.143378130263482e-05,
+      "loss": 1.665,
+      "step": 6677
+    },
+    {
+      "epoch": 0.6936740417575569,
+      "grad_norm": 0.4393899738788605,
+      "learning_rate": 2.142039141646135e-05,
+      "loss": 1.6799,
+      "step": 6678
+    },
+    {
+      "epoch": 0.6937779162771373,
+      "grad_norm": 0.4026007056236267,
+      "learning_rate": 2.140700457379325e-05,
+      "loss": 1.6812,
+      "step": 6679
+    },
+    {
+      "epoch": 0.6938817907967175,
+      "grad_norm": 0.4645029306411743,
+      "learning_rate": 2.1393620776056163e-05,
+      "loss": 1.7876,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6939856653162979,
+      "grad_norm": 0.3936692774295807,
+      "learning_rate": 2.1380240024675285e-05,
+      "loss": 1.3614,
+      "step": 6681
+    },
+    {
+      "epoch": 0.6940895398358783,
+      "grad_norm": 0.4686059057712555,
+      "learning_rate": 2.136686232107562e-05,
+      "loss": 1.7599,
+      "step": 6682
+    },
+    {
+      "epoch": 0.6941934143554586,
+      "grad_norm": 0.4059823155403137,
+      "learning_rate": 2.1353487666681777e-05,
+      "loss": 1.663,
+      "step": 6683
+    },
+    {
+      "epoch": 0.6942972888750389,
+      "grad_norm": 0.4833250939846039,
+      "learning_rate": 2.134011606291804e-05,
+      "loss": 1.8039,
+      "step": 6684
+    },
+    {
+      "epoch": 0.6944011633946193,
+      "grad_norm": 0.41929182410240173,
+      "learning_rate": 2.132674751120839e-05,
+      "loss": 1.7361,
+      "step": 6685
+    },
+    {
+      "epoch": 0.6945050379141996,
+      "grad_norm": 0.4122251272201538,
+      "learning_rate": 2.131338201297647e-05,
+      "loss": 1.5977,
+      "step": 6686
+    },
+    {
+      "epoch": 0.69460891243378,
+      "grad_norm": 0.476242333650589,
+      "learning_rate": 2.1300019569645596e-05,
+      "loss": 1.9488,
+      "step": 6687
+    },
+    {
+      "epoch": 0.6947127869533604,
+      "grad_norm": 0.40492451190948486,
+      "learning_rate": 2.1286660182638756e-05,
+      "loss": 1.7338,
+      "step": 6688
+    },
+    {
+      "epoch": 0.6948166614729406,
+      "grad_norm": 0.4443548917770386,
+      "learning_rate": 2.1273303853378674e-05,
+      "loss": 1.5402,
+      "step": 6689
+    },
+    {
+      "epoch": 0.694920535992521,
+      "grad_norm": 0.40552589297294617,
+      "learning_rate": 2.125995058328763e-05,
+      "loss": 1.6692,
+      "step": 6690
+    },
+    {
+      "epoch": 0.6950244105121014,
+      "grad_norm": 0.5920501351356506,
+      "learning_rate": 2.124660037378765e-05,
+      "loss": 1.7801,
+      "step": 6691
+    },
+    {
+      "epoch": 0.6951282850316818,
+      "grad_norm": 0.4190227687358856,
+      "learning_rate": 2.1233253226300465e-05,
+      "loss": 1.5115,
+      "step": 6692
+    },
+    {
+      "epoch": 0.695232159551262,
+      "grad_norm": 0.47602778673171997,
+      "learning_rate": 2.1219909142247424e-05,
+      "loss": 1.8221,
+      "step": 6693
+    },
+    {
+      "epoch": 0.6953360340708424,
+      "grad_norm": 0.43550148606300354,
+      "learning_rate": 2.1206568123049564e-05,
+      "loss": 1.6134,
+      "step": 6694
+    },
+    {
+      "epoch": 0.6954399085904228,
+      "grad_norm": 0.4044070541858673,
+      "learning_rate": 2.119323017012759e-05,
+      "loss": 1.5809,
+      "step": 6695
+    },
+    {
+      "epoch": 0.6955437831100031,
+      "grad_norm": 0.4208220839500427,
+      "learning_rate": 2.1179895284901924e-05,
+      "loss": 1.5416,
+      "step": 6696
+    },
+    {
+      "epoch": 0.6956476576295835,
+      "grad_norm": 0.40507733821868896,
+      "learning_rate": 2.1166563468792594e-05,
+      "loss": 1.5112,
+      "step": 6697
+    },
+    {
+      "epoch": 0.6957515321491639,
+      "grad_norm": 0.4560292959213257,
+      "learning_rate": 2.1153234723219328e-05,
+      "loss": 1.6208,
+      "step": 6698
+    },
+    {
+      "epoch": 0.6958554066687441,
+      "grad_norm": 0.46594947576522827,
+      "learning_rate": 2.1139909049601587e-05,
+      "loss": 1.7745,
+      "step": 6699
+    },
+    {
+      "epoch": 0.6959592811883245,
+      "grad_norm": 0.4122544527053833,
+      "learning_rate": 2.1126586449358375e-05,
+      "loss": 1.5898,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6960631557079049,
+      "grad_norm": 0.45607662200927734,
+      "learning_rate": 2.111326692390851e-05,
+      "loss": 1.7258,
+      "step": 6701
+    },
+    {
+      "epoch": 0.6961670302274852,
+      "grad_norm": 0.4141891300678253,
+      "learning_rate": 2.1099950474670387e-05,
+      "loss": 1.6533,
+      "step": 6702
+    },
+    {
+      "epoch": 0.6962709047470655,
+      "grad_norm": 0.3780854046344757,
+      "learning_rate": 2.1086637103062107e-05,
+      "loss": 1.5623,
+      "step": 6703
+    },
+    {
+      "epoch": 0.6963747792666459,
+      "grad_norm": 0.4268392026424408,
+      "learning_rate": 2.1073326810501448e-05,
+      "loss": 1.6596,
+      "step": 6704
+    },
+    {
+      "epoch": 0.6964786537862262,
+      "grad_norm": 0.40509405732154846,
+      "learning_rate": 2.1060019598405843e-05,
+      "loss": 1.5612,
+      "step": 6705
+    },
+    {
+      "epoch": 0.6965825283058066,
+      "grad_norm": 0.38771677017211914,
+      "learning_rate": 2.1046715468192414e-05,
+      "loss": 1.5457,
+      "step": 6706
+    },
+    {
+      "epoch": 0.696686402825387,
+      "grad_norm": 0.4286300539970398,
+      "learning_rate": 2.1033414421277924e-05,
+      "loss": 1.7415,
+      "step": 6707
+    },
+    {
+      "epoch": 0.6967902773449672,
+      "grad_norm": 0.46014636754989624,
+      "learning_rate": 2.1020116459078887e-05,
+      "loss": 1.5627,
+      "step": 6708
+    },
+    {
+      "epoch": 0.6968941518645476,
+      "grad_norm": 0.42904940247535706,
+      "learning_rate": 2.1006821583011356e-05,
+      "loss": 1.6067,
+      "step": 6709
+    },
+    {
+      "epoch": 0.696998026384128,
+      "grad_norm": 0.39022296667099,
+      "learning_rate": 2.099352979449119e-05,
+      "loss": 1.5778,
+      "step": 6710
+    },
+    {
+      "epoch": 0.6971019009037083,
+      "grad_norm": 0.4107384979724884,
+      "learning_rate": 2.0980241094933845e-05,
+      "loss": 1.685,
+      "step": 6711
+    },
+    {
+      "epoch": 0.6972057754232887,
+      "grad_norm": 0.42475056648254395,
+      "learning_rate": 2.0966955485754457e-05,
+      "loss": 1.2405,
+      "step": 6712
+    },
+    {
+      "epoch": 0.697309649942869,
+      "grad_norm": 0.4216676950454712,
+      "learning_rate": 2.0953672968367844e-05,
+      "loss": 1.4806,
+      "step": 6713
+    },
+    {
+      "epoch": 0.6974135244624493,
+      "grad_norm": 0.37157952785491943,
+      "learning_rate": 2.094039354418849e-05,
+      "loss": 1.4771,
+      "step": 6714
+    },
+    {
+      "epoch": 0.6975173989820297,
+      "grad_norm": 0.44007498025894165,
+      "learning_rate": 2.0927117214630547e-05,
+      "loss": 1.8148,
+      "step": 6715
+    },
+    {
+      "epoch": 0.6976212735016101,
+      "grad_norm": 0.4631294012069702,
+      "learning_rate": 2.0913843981107835e-05,
+      "loss": 1.7156,
+      "step": 6716
+    },
+    {
+      "epoch": 0.6977251480211905,
+      "grad_norm": 0.42013266682624817,
+      "learning_rate": 2.0900573845033865e-05,
+      "loss": 1.6558,
+      "step": 6717
+    },
+    {
+      "epoch": 0.6978290225407707,
+      "grad_norm": 0.39242061972618103,
+      "learning_rate": 2.0887306807821822e-05,
+      "loss": 1.5849,
+      "step": 6718
+    },
+    {
+      "epoch": 0.6979328970603511,
+      "grad_norm": 0.41832900047302246,
+      "learning_rate": 2.0874042870884475e-05,
+      "loss": 1.5836,
+      "step": 6719
+    },
+    {
+      "epoch": 0.6980367715799315,
+      "grad_norm": 0.44263386726379395,
+      "learning_rate": 2.086078203563439e-05,
+      "loss": 1.7748,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6981406460995118,
+      "grad_norm": 0.4205859303474426,
+      "learning_rate": 2.0847524303483728e-05,
+      "loss": 1.7206,
+      "step": 6721
+    },
+    {
+      "epoch": 0.6982445206190921,
+      "grad_norm": 0.422676146030426,
+      "learning_rate": 2.0834269675844326e-05,
+      "loss": 1.7428,
+      "step": 6722
+    },
+    {
+      "epoch": 0.6983483951386725,
+      "grad_norm": 0.42799460887908936,
+      "learning_rate": 2.0821018154127686e-05,
+      "loss": 1.7212,
+      "step": 6723
+    },
+    {
+      "epoch": 0.6984522696582528,
+      "grad_norm": 0.37061259150505066,
+      "learning_rate": 2.0807769739745043e-05,
+      "loss": 1.326,
+      "step": 6724
+    },
+    {
+      "epoch": 0.6985561441778332,
+      "grad_norm": 0.48382559418678284,
+      "learning_rate": 2.0794524434107192e-05,
+      "loss": 1.6675,
+      "step": 6725
+    },
+    {
+      "epoch": 0.6986600186974136,
+      "grad_norm": 0.4368096888065338,
+      "learning_rate": 2.0781282238624666e-05,
+      "loss": 1.6125,
+      "step": 6726
+    },
+    {
+      "epoch": 0.6987638932169938,
+      "grad_norm": 0.424841970205307,
+      "learning_rate": 2.07680431547077e-05,
+      "loss": 1.567,
+      "step": 6727
+    },
+    {
+      "epoch": 0.6988677677365742,
+      "grad_norm": 0.40765881538391113,
+      "learning_rate": 2.0754807183766084e-05,
+      "loss": 1.6608,
+      "step": 6728
+    },
+    {
+      "epoch": 0.6989716422561546,
+      "grad_norm": 0.45474496483802795,
+      "learning_rate": 2.07415743272094e-05,
+      "loss": 1.672,
+      "step": 6729
+    },
+    {
+      "epoch": 0.6990755167757349,
+      "grad_norm": 0.3880652189254761,
+      "learning_rate": 2.0728344586446825e-05,
+      "loss": 1.6127,
+      "step": 6730
+    },
+    {
+      "epoch": 0.6991793912953153,
+      "grad_norm": 0.40850159525871277,
+      "learning_rate": 2.0715117962887224e-05,
+      "loss": 1.604,
+      "step": 6731
+    },
+    {
+      "epoch": 0.6992832658148956,
+      "grad_norm": 0.44652217626571655,
+      "learning_rate": 2.070189445793913e-05,
+      "loss": 1.6413,
+      "step": 6732
+    },
+    {
+      "epoch": 0.6993871403344759,
+      "grad_norm": 0.42982134222984314,
+      "learning_rate": 2.068867407301074e-05,
+      "loss": 1.6768,
+      "step": 6733
+    },
+    {
+      "epoch": 0.6994910148540563,
+      "grad_norm": 0.40679651498794556,
+      "learning_rate": 2.0675456809509925e-05,
+      "loss": 1.6151,
+      "step": 6734
+    },
+    {
+      "epoch": 0.6995948893736367,
+      "grad_norm": 0.427057683467865,
+      "learning_rate": 2.06622426688442e-05,
+      "loss": 1.6619,
+      "step": 6735
+    },
+    {
+      "epoch": 0.699698763893217,
+      "grad_norm": 0.4573904275894165,
+      "learning_rate": 2.064903165242083e-05,
+      "loss": 1.7536,
+      "step": 6736
+    },
+    {
+      "epoch": 0.6998026384127973,
+      "grad_norm": 0.4206405282020569,
+      "learning_rate": 2.06358237616466e-05,
+      "loss": 1.8067,
+      "step": 6737
+    },
+    {
+      "epoch": 0.6999065129323777,
+      "grad_norm": 0.4055739641189575,
+      "learning_rate": 2.062261899792811e-05,
+      "loss": 1.6172,
+      "step": 6738
+    },
+    {
+      "epoch": 0.700010387451958,
+      "grad_norm": 0.45139777660369873,
+      "learning_rate": 2.0609417362671545e-05,
+      "loss": 1.7188,
+      "step": 6739
+    },
+    {
+      "epoch": 0.7001142619715384,
+      "grad_norm": 0.4384258985519409,
+      "learning_rate": 2.0596218857282778e-05,
+      "loss": 1.5628,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7002181364911187,
+      "grad_norm": 0.39930063486099243,
+      "learning_rate": 2.058302348316734e-05,
+      "loss": 1.6419,
+      "step": 6741
+    },
+    {
+      "epoch": 0.7003220110106991,
+      "grad_norm": 0.40519270300865173,
+      "learning_rate": 2.0569831241730425e-05,
+      "loss": 1.5731,
+      "step": 6742
+    },
+    {
+      "epoch": 0.7004258855302794,
+      "grad_norm": 0.46715521812438965,
+      "learning_rate": 2.0556642134376953e-05,
+      "loss": 1.6441,
+      "step": 6743
+    },
+    {
+      "epoch": 0.7005297600498598,
+      "grad_norm": 0.4366108179092407,
+      "learning_rate": 2.0543456162511386e-05,
+      "loss": 1.7277,
+      "step": 6744
+    },
+    {
+      "epoch": 0.7006336345694402,
+      "grad_norm": 0.3887299597263336,
+      "learning_rate": 2.053027332753799e-05,
+      "loss": 1.4717,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7007375090890204,
+      "grad_norm": 0.4008408486843109,
+      "learning_rate": 2.0517093630860633e-05,
+      "loss": 1.619,
+      "step": 6746
+    },
+    {
+      "epoch": 0.7008413836086008,
+      "grad_norm": 0.43216922879219055,
+      "learning_rate": 2.050391707388279e-05,
+      "loss": 1.5183,
+      "step": 6747
+    },
+    {
+      "epoch": 0.7009452581281812,
+      "grad_norm": 0.4278073310852051,
+      "learning_rate": 2.049074365800772e-05,
+      "loss": 1.7367,
+      "step": 6748
+    },
+    {
+      "epoch": 0.7010491326477615,
+      "grad_norm": 0.4540071487426758,
+      "learning_rate": 2.0477573384638266e-05,
+      "loss": 1.8808,
+      "step": 6749
+    },
+    {
+      "epoch": 0.7011530071673419,
+      "grad_norm": 0.4321799874305725,
+      "learning_rate": 2.0464406255176966e-05,
+      "loss": 1.6206,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7012568816869222,
+      "grad_norm": 0.39823484420776367,
+      "learning_rate": 2.045124227102601e-05,
+      "loss": 1.6493,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7013607562065025,
+      "grad_norm": 0.3932490050792694,
+      "learning_rate": 2.043808143358727e-05,
+      "loss": 1.6487,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7014646307260829,
+      "grad_norm": 0.45285892486572266,
+      "learning_rate": 2.0424923744262258e-05,
+      "loss": 1.7022,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7015685052456633,
+      "grad_norm": 0.41068267822265625,
+      "learning_rate": 2.041176920445216e-05,
+      "loss": 1.7152,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7016723797652435,
+      "grad_norm": 0.4176632761955261,
+      "learning_rate": 2.0398617815557875e-05,
+      "loss": 1.5213,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7017762542848239,
+      "grad_norm": 0.3959343433380127,
+      "learning_rate": 2.038546957897986e-05,
+      "loss": 1.4975,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7018801288044043,
+      "grad_norm": 0.4066161513328552,
+      "learning_rate": 2.037232449611835e-05,
+      "loss": 1.5669,
+      "step": 6757
+    },
+    {
+      "epoch": 0.7019840033239846,
+      "grad_norm": 0.4088003635406494,
+      "learning_rate": 2.0359182568373176e-05,
+      "loss": 1.7064,
+      "step": 6758
+    },
+    {
+      "epoch": 0.702087877843565,
+      "grad_norm": 0.4266449213027954,
+      "learning_rate": 2.0346043797143854e-05,
+      "loss": 1.5934,
+      "step": 6759
+    },
+    {
+      "epoch": 0.7021917523631453,
+      "grad_norm": 0.418796181678772,
+      "learning_rate": 2.0332908183829556e-05,
+      "loss": 1.6623,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7022956268827256,
+      "grad_norm": 0.44862931966781616,
+      "learning_rate": 2.031977572982912e-05,
+      "loss": 1.8849,
+      "step": 6761
+    },
+    {
+      "epoch": 0.702399501402306,
+      "grad_norm": 0.44697731733322144,
+      "learning_rate": 2.030664643654106e-05,
+      "loss": 1.7985,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7025033759218864,
+      "grad_norm": 0.38466355204582214,
+      "learning_rate": 2.0293520305363522e-05,
+      "loss": 1.5767,
+      "step": 6763
+    },
+    {
+      "epoch": 0.7026072504414667,
+      "grad_norm": 0.40668144822120667,
+      "learning_rate": 2.028039733769438e-05,
+      "loss": 1.7073,
+      "step": 6764
+    },
+    {
+      "epoch": 0.702711124961047,
+      "grad_norm": 0.42051973938941956,
+      "learning_rate": 2.0267277534931066e-05,
+      "loss": 1.3946,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7028149994806274,
+      "grad_norm": 0.42236143350601196,
+      "learning_rate": 2.0254160898470792e-05,
+      "loss": 1.7273,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7029188740002077,
+      "grad_norm": 0.4152265191078186,
+      "learning_rate": 2.024104742971035e-05,
+      "loss": 1.4529,
+      "step": 6767
+    },
+    {
+      "epoch": 0.7030227485197881,
+      "grad_norm": 0.37839797139167786,
+      "learning_rate": 2.0227937130046237e-05,
+      "loss": 1.4317,
+      "step": 6768
+    },
+    {
+      "epoch": 0.7031266230393685,
+      "grad_norm": 0.4131396412849426,
+      "learning_rate": 2.0214830000874586e-05,
+      "loss": 1.659,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7032304975589488,
+      "grad_norm": 0.4527067244052887,
+      "learning_rate": 2.0201726043591186e-05,
+      "loss": 1.7416,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7033343720785291,
+      "grad_norm": 0.40972375869750977,
+      "learning_rate": 2.018862525959156e-05,
+      "loss": 1.6191,
+      "step": 6771
+    },
+    {
+      "epoch": 0.7034382465981095,
+      "grad_norm": 0.38629379868507385,
+      "learning_rate": 2.017552765027077e-05,
+      "loss": 1.6096,
+      "step": 6772
+    },
+    {
+      "epoch": 0.7035421211176899,
+      "grad_norm": 0.39593833684921265,
+      "learning_rate": 2.0162433217023667e-05,
+      "loss": 1.7234,
+      "step": 6773
+    },
+    {
+      "epoch": 0.7036459956372701,
+      "grad_norm": 0.3909621834754944,
+      "learning_rate": 2.0149341961244704e-05,
+      "loss": 1.5358,
+      "step": 6774
+    },
+    {
+      "epoch": 0.7037498701568505,
+      "grad_norm": 0.4053054749965668,
+      "learning_rate": 2.0136253884327937e-05,
+      "loss": 1.592,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7038537446764309,
+      "grad_norm": 0.4064863622188568,
+      "learning_rate": 2.0123168987667206e-05,
+      "loss": 1.6287,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7039576191960112,
+      "grad_norm": 0.422603040933609,
+      "learning_rate": 2.011008727265593e-05,
+      "loss": 1.6935,
+      "step": 6777
+    },
+    {
+      "epoch": 0.7040614937155916,
+      "grad_norm": 0.418403297662735,
+      "learning_rate": 2.009700874068721e-05,
+      "loss": 1.8375,
+      "step": 6778
+    },
+    {
+      "epoch": 0.704165368235172,
+      "grad_norm": 0.4110656678676605,
+      "learning_rate": 2.0083933393153808e-05,
+      "loss": 1.6755,
+      "step": 6779
+    },
+    {
+      "epoch": 0.7042692427547522,
+      "grad_norm": 0.47258806228637695,
+      "learning_rate": 2.007086123144814e-05,
+      "loss": 1.6924,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7043731172743326,
+      "grad_norm": 0.4690280854701996,
+      "learning_rate": 2.00577922569623e-05,
+      "loss": 1.8002,
+      "step": 6781
+    },
+    {
+      "epoch": 0.704476991793913,
+      "grad_norm": 0.40814927220344543,
+      "learning_rate": 2.0044726471088005e-05,
+      "loss": 1.6296,
+      "step": 6782
+    },
+    {
+      "epoch": 0.7045808663134933,
+      "grad_norm": 0.4353756606578827,
+      "learning_rate": 2.0031663875216717e-05,
+      "loss": 1.7662,
+      "step": 6783
+    },
+    {
+      "epoch": 0.7046847408330736,
+      "grad_norm": 0.41065043210983276,
+      "learning_rate": 2.0018604470739434e-05,
+      "loss": 1.5681,
+      "step": 6784
+    },
+    {
+      "epoch": 0.704788615352654,
+      "grad_norm": 0.4216291308403015,
+      "learning_rate": 2.000554825904692e-05,
+      "loss": 1.6114,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7048924898722343,
+      "grad_norm": 0.43526455760002136,
+      "learning_rate": 1.9992495241529556e-05,
+      "loss": 1.6702,
+      "step": 6786
+    },
+    {
+      "epoch": 0.7049963643918147,
+      "grad_norm": 0.43194445967674255,
+      "learning_rate": 1.997944541957738e-05,
+      "loss": 1.7412,
+      "step": 6787
+    },
+    {
+      "epoch": 0.7051002389113951,
+      "grad_norm": 0.4365564286708832,
+      "learning_rate": 1.99663987945801e-05,
+      "loss": 1.7816,
+      "step": 6788
+    },
+    {
+      "epoch": 0.7052041134309753,
+      "grad_norm": 0.4253787696361542,
+      "learning_rate": 1.9953355367927075e-05,
+      "loss": 1.6678,
+      "step": 6789
+    },
+    {
+      "epoch": 0.7053079879505557,
+      "grad_norm": 0.43567463755607605,
+      "learning_rate": 1.9940315141007327e-05,
+      "loss": 1.6948,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7054118624701361,
+      "grad_norm": 0.41654548048973083,
+      "learning_rate": 1.992727811520953e-05,
+      "loss": 1.6762,
+      "step": 6791
+    },
+    {
+      "epoch": 0.7055157369897164,
+      "grad_norm": 0.3824082612991333,
+      "learning_rate": 1.991424429192207e-05,
+      "loss": 1.495,
+      "step": 6792
+    },
+    {
+      "epoch": 0.7056196115092968,
+      "grad_norm": 0.4068382978439331,
+      "learning_rate": 1.9901213672532875e-05,
+      "loss": 1.6277,
+      "step": 6793
+    },
+    {
+      "epoch": 0.7057234860288771,
+      "grad_norm": 0.3962841033935547,
+      "learning_rate": 1.988818625842967e-05,
+      "loss": 1.6252,
+      "step": 6794
+    },
+    {
+      "epoch": 0.7058273605484575,
+      "grad_norm": 0.4224420487880707,
+      "learning_rate": 1.9875162050999747e-05,
+      "loss": 1.5978,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7059312350680378,
+      "grad_norm": 0.38469555974006653,
+      "learning_rate": 1.986214105163008e-05,
+      "loss": 1.521,
+      "step": 6796
+    },
+    {
+      "epoch": 0.7060351095876182,
+      "grad_norm": 0.43312588334083557,
+      "learning_rate": 1.9849123261707305e-05,
+      "loss": 1.7809,
+      "step": 6797
+    },
+    {
+      "epoch": 0.7061389841071986,
+      "grad_norm": 0.42561838030815125,
+      "learning_rate": 1.98361086826177e-05,
+      "loss": 1.6883,
+      "step": 6798
+    },
+    {
+      "epoch": 0.7062428586267788,
+      "grad_norm": 0.4514966309070587,
+      "learning_rate": 1.9823097315747267e-05,
+      "loss": 1.6939,
+      "step": 6799
+    },
+    {
+      "epoch": 0.7063467331463592,
+      "grad_norm": 0.4347332715988159,
+      "learning_rate": 1.9810089162481548e-05,
+      "loss": 1.6935,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7064506076659396,
+      "grad_norm": 0.45488476753234863,
+      "learning_rate": 1.9797084224205857e-05,
+      "loss": 1.8802,
+      "step": 6801
+    },
+    {
+      "epoch": 0.7065544821855199,
+      "grad_norm": 0.44390448927879333,
+      "learning_rate": 1.978408250230513e-05,
+      "loss": 1.7329,
+      "step": 6802
+    },
+    {
+      "epoch": 0.7066583567051002,
+      "grad_norm": 0.39649200439453125,
+      "learning_rate": 1.977108399816388e-05,
+      "loss": 1.5892,
+      "step": 6803
+    },
+    {
+      "epoch": 0.7067622312246806,
+      "grad_norm": 0.3859086036682129,
+      "learning_rate": 1.975808871316641e-05,
+      "loss": 1.5868,
+      "step": 6804
+    },
+    {
+      "epoch": 0.7068661057442609,
+      "grad_norm": 0.48838183283805847,
+      "learning_rate": 1.97450966486966e-05,
+      "loss": 1.7898,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7069699802638413,
+      "grad_norm": 0.4066251218318939,
+      "learning_rate": 1.9732107806137996e-05,
+      "loss": 1.662,
+      "step": 6806
+    },
+    {
+      "epoch": 0.7070738547834217,
+      "grad_norm": 0.4213266968727112,
+      "learning_rate": 1.971912218687381e-05,
+      "loss": 1.5625,
+      "step": 6807
+    },
+    {
+      "epoch": 0.7071777293030019,
+      "grad_norm": 0.38811829686164856,
+      "learning_rate": 1.9706139792286915e-05,
+      "loss": 1.4567,
+      "step": 6808
+    },
+    {
+      "epoch": 0.7072816038225823,
+      "grad_norm": 0.42099571228027344,
+      "learning_rate": 1.9693160623759826e-05,
+      "loss": 1.6887,
+      "step": 6809
+    },
+    {
+      "epoch": 0.7073854783421627,
+      "grad_norm": 0.45879611372947693,
+      "learning_rate": 1.968018468267472e-05,
+      "loss": 1.7088,
+      "step": 6810
+    },
+    {
+      "epoch": 0.707489352861743,
+      "grad_norm": 0.3958274722099304,
+      "learning_rate": 1.9667211970413468e-05,
+      "loss": 1.5636,
+      "step": 6811
+    },
+    {
+      "epoch": 0.7075932273813234,
+      "grad_norm": 0.4116361141204834,
+      "learning_rate": 1.96542424883575e-05,
+      "loss": 1.5771,
+      "step": 6812
+    },
+    {
+      "epoch": 0.7076971019009037,
+      "grad_norm": 0.42354995012283325,
+      "learning_rate": 1.9641276237888023e-05,
+      "loss": 1.8045,
+      "step": 6813
+    },
+    {
+      "epoch": 0.707800976420484,
+      "grad_norm": 0.4337577223777771,
+      "learning_rate": 1.9628313220385818e-05,
+      "loss": 1.7703,
+      "step": 6814
+    },
+    {
+      "epoch": 0.7079048509400644,
+      "grad_norm": 0.4336363673210144,
+      "learning_rate": 1.961535343723135e-05,
+      "loss": 1.4605,
+      "step": 6815
+    },
+    {
+      "epoch": 0.7080087254596448,
+      "grad_norm": 0.41159120202064514,
+      "learning_rate": 1.960239688980473e-05,
+      "loss": 1.6786,
+      "step": 6816
+    },
+    {
+      "epoch": 0.708112599979225,
+      "grad_norm": 0.4380592405796051,
+      "learning_rate": 1.958944357948573e-05,
+      "loss": 1.5045,
+      "step": 6817
+    },
+    {
+      "epoch": 0.7082164744988054,
+      "grad_norm": 0.3844902515411377,
+      "learning_rate": 1.9576493507653776e-05,
+      "loss": 1.653,
+      "step": 6818
+    },
+    {
+      "epoch": 0.7083203490183858,
+      "grad_norm": 0.41153958439826965,
+      "learning_rate": 1.9563546675687932e-05,
+      "loss": 1.7776,
+      "step": 6819
+    },
+    {
+      "epoch": 0.7084242235379662,
+      "grad_norm": 0.39045795798301697,
+      "learning_rate": 1.9550603084966973e-05,
+      "loss": 1.5983,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7085280980575465,
+      "grad_norm": 0.4701341688632965,
+      "learning_rate": 1.9537662736869272e-05,
+      "loss": 1.7198,
+      "step": 6821
+    },
+    {
+      "epoch": 0.7086319725771268,
+      "grad_norm": 0.4244726896286011,
+      "learning_rate": 1.9524725632772878e-05,
+      "loss": 1.6711,
+      "step": 6822
+    },
+    {
+      "epoch": 0.7087358470967072,
+      "grad_norm": 0.44290024042129517,
+      "learning_rate": 1.9511791774055482e-05,
+      "loss": 1.8374,
+      "step": 6823
+    },
+    {
+      "epoch": 0.7088397216162875,
+      "grad_norm": 0.4390253722667694,
+      "learning_rate": 1.9498861162094446e-05,
+      "loss": 1.6595,
+      "step": 6824
+    },
+    {
+      "epoch": 0.7089435961358679,
+      "grad_norm": 0.3866863548755646,
+      "learning_rate": 1.948593379826678e-05,
+      "loss": 1.5316,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7090474706554483,
+      "grad_norm": 0.4000120162963867,
+      "learning_rate": 1.947300968394913e-05,
+      "loss": 1.5832,
+      "step": 6826
+    },
+    {
+      "epoch": 0.7091513451750285,
+      "grad_norm": 0.4002024531364441,
+      "learning_rate": 1.946008882051786e-05,
+      "loss": 1.6259,
+      "step": 6827
+    },
+    {
+      "epoch": 0.7092552196946089,
+      "grad_norm": 0.43123364448547363,
+      "learning_rate": 1.9447171209348875e-05,
+      "loss": 1.6728,
+      "step": 6828
+    },
+    {
+      "epoch": 0.7093590942141893,
+      "grad_norm": 0.39023008942604065,
+      "learning_rate": 1.943425685181785e-05,
+      "loss": 1.5225,
+      "step": 6829
+    },
+    {
+      "epoch": 0.7094629687337696,
+      "grad_norm": 0.40491533279418945,
+      "learning_rate": 1.9421345749300073e-05,
+      "loss": 1.6512,
+      "step": 6830
+    },
+    {
+      "epoch": 0.70956684325335,
+      "grad_norm": 0.4594792127609253,
+      "learning_rate": 1.940843790317041e-05,
+      "loss": 1.6372,
+      "step": 6831
+    },
+    {
+      "epoch": 0.7096707177729303,
+      "grad_norm": 0.4281672537326813,
+      "learning_rate": 1.939553331480351e-05,
+      "loss": 1.5949,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7097745922925106,
+      "grad_norm": 0.4785386025905609,
+      "learning_rate": 1.9382631985573586e-05,
+      "loss": 1.682,
+      "step": 6833
+    },
+    {
+      "epoch": 0.709878466812091,
+      "grad_norm": 0.42151594161987305,
+      "learning_rate": 1.9369733916854526e-05,
+      "loss": 1.5583,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7099823413316714,
+      "grad_norm": 0.4763428866863251,
+      "learning_rate": 1.935683911001988e-05,
+      "loss": 1.6265,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7100862158512516,
+      "grad_norm": 0.4347456693649292,
+      "learning_rate": 1.934394756644285e-05,
+      "loss": 1.7359,
+      "step": 6836
+    },
+    {
+      "epoch": 0.710190090370832,
+      "grad_norm": 0.40570497512817383,
+      "learning_rate": 1.9331059287496274e-05,
+      "loss": 1.6238,
+      "step": 6837
+    },
+    {
+      "epoch": 0.7102939648904124,
+      "grad_norm": 0.41450199484825134,
+      "learning_rate": 1.9318174274552638e-05,
+      "loss": 1.7365,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7103978394099927,
+      "grad_norm": 0.44917452335357666,
+      "learning_rate": 1.930529252898415e-05,
+      "loss": 1.7917,
+      "step": 6839
+    },
+    {
+      "epoch": 0.7105017139295731,
+      "grad_norm": 0.4153517186641693,
+      "learning_rate": 1.929241405216254e-05,
+      "loss": 1.7213,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7106055884491534,
+      "grad_norm": 0.4131045639514923,
+      "learning_rate": 1.9279538845459327e-05,
+      "loss": 1.5519,
+      "step": 6841
+    },
+    {
+      "epoch": 0.7107094629687337,
+      "grad_norm": 0.38155892491340637,
+      "learning_rate": 1.92666669102456e-05,
+      "loss": 1.505,
+      "step": 6842
+    },
+    {
+      "epoch": 0.7108133374883141,
+      "grad_norm": 0.45029881596565247,
+      "learning_rate": 1.9253798247892113e-05,
+      "loss": 1.678,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7109172120078945,
+      "grad_norm": 0.4427982568740845,
+      "learning_rate": 1.924093285976929e-05,
+      "loss": 1.8031,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7110210865274748,
+      "grad_norm": 0.42872363328933716,
+      "learning_rate": 1.9228070747247185e-05,
+      "loss": 1.7211,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7111249610470551,
+      "grad_norm": 0.42956963181495667,
+      "learning_rate": 1.9215211911695512e-05,
+      "loss": 1.7405,
+      "step": 6846
+    },
+    {
+      "epoch": 0.7112288355666355,
+      "grad_norm": 0.39396312832832336,
+      "learning_rate": 1.9202356354483624e-05,
+      "loss": 1.6972,
+      "step": 6847
+    },
+    {
+      "epoch": 0.7113327100862159,
+      "grad_norm": 0.41904011368751526,
+      "learning_rate": 1.9189504076980567e-05,
+      "loss": 1.7432,
+      "step": 6848
+    },
+    {
+      "epoch": 0.7114365846057962,
+      "grad_norm": 0.37891194224357605,
+      "learning_rate": 1.9176655080555e-05,
+      "loss": 1.6582,
+      "step": 6849
+    },
+    {
+      "epoch": 0.7115404591253766,
+      "grad_norm": 0.4023973345756531,
+      "learning_rate": 1.9163809366575235e-05,
+      "loss": 1.6672,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7116443336449569,
+      "grad_norm": 0.44481852650642395,
+      "learning_rate": 1.9150966936409242e-05,
+      "loss": 1.7137,
+      "step": 6851
+    },
+    {
+      "epoch": 0.7117482081645372,
+      "grad_norm": 0.39180970191955566,
+      "learning_rate": 1.9138127791424637e-05,
+      "loss": 1.5847,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7118520826841176,
+      "grad_norm": 0.4336892068386078,
+      "learning_rate": 1.912529193298869e-05,
+      "loss": 1.7037,
+      "step": 6853
+    },
+    {
+      "epoch": 0.711955957203698,
+      "grad_norm": 0.3839171230792999,
+      "learning_rate": 1.91124593624683e-05,
+      "loss": 1.5538,
+      "step": 6854
+    },
+    {
+      "epoch": 0.7120598317232782,
+      "grad_norm": 0.40723204612731934,
+      "learning_rate": 1.9099630081230095e-05,
+      "loss": 1.4787,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7121637062428586,
+      "grad_norm": 0.449480801820755,
+      "learning_rate": 1.908680409064022e-05,
+      "loss": 1.831,
+      "step": 6856
+    },
+    {
+      "epoch": 0.712267580762439,
+      "grad_norm": 0.5175827145576477,
+      "learning_rate": 1.9073981392064587e-05,
+      "loss": 1.9043,
+      "step": 6857
+    },
+    {
+      "epoch": 0.7123714552820193,
+      "grad_norm": 0.41189754009246826,
+      "learning_rate": 1.9061161986868732e-05,
+      "loss": 1.7122,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7124753298015997,
+      "grad_norm": 0.43195226788520813,
+      "learning_rate": 1.9048345876417752e-05,
+      "loss": 1.6308,
+      "step": 6859
+    },
+    {
+      "epoch": 0.71257920432118,
+      "grad_norm": 0.41276681423187256,
+      "learning_rate": 1.9035533062076522e-05,
+      "loss": 1.658,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7126830788407603,
+      "grad_norm": 0.4124269485473633,
+      "learning_rate": 1.9022723545209487e-05,
+      "loss": 1.5886,
+      "step": 6861
+    },
+    {
+      "epoch": 0.7127869533603407,
+      "grad_norm": 0.4499133825302124,
+      "learning_rate": 1.9009917327180766e-05,
+      "loss": 1.5387,
+      "step": 6862
+    },
+    {
+      "epoch": 0.7128908278799211,
+      "grad_norm": 0.41163170337677,
+      "learning_rate": 1.8997114409354116e-05,
+      "loss": 1.7033,
+      "step": 6863
+    },
+    {
+      "epoch": 0.7129947023995014,
+      "grad_norm": 0.4106546640396118,
+      "learning_rate": 1.898431479309295e-05,
+      "loss": 1.6542,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7130985769190817,
+      "grad_norm": 0.46902644634246826,
+      "learning_rate": 1.8971518479760324e-05,
+      "loss": 1.6021,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7132024514386621,
+      "grad_norm": 0.4623051881790161,
+      "learning_rate": 1.8958725470718937e-05,
+      "loss": 1.9381,
+      "step": 6866
+    },
+    {
+      "epoch": 0.7133063259582424,
+      "grad_norm": 0.43215009570121765,
+      "learning_rate": 1.8945935767331186e-05,
+      "loss": 1.6968,
+      "step": 6867
+    },
+    {
+      "epoch": 0.7134102004778228,
+      "grad_norm": 0.4356382191181183,
+      "learning_rate": 1.893314937095901e-05,
+      "loss": 1.6144,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7135140749974032,
+      "grad_norm": 0.456093966960907,
+      "learning_rate": 1.8920366282964115e-05,
+      "loss": 1.723,
+      "step": 6869
+    },
+    {
+      "epoch": 0.7136179495169834,
+      "grad_norm": 0.5878288149833679,
+      "learning_rate": 1.8907586504707773e-05,
+      "loss": 1.5267,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7137218240365638,
+      "grad_norm": 0.3891201913356781,
+      "learning_rate": 1.8894810037550947e-05,
+      "loss": 1.53,
+      "step": 6871
+    },
+    {
+      "epoch": 0.7138256985561442,
+      "grad_norm": 0.3847010135650635,
+      "learning_rate": 1.888203688285422e-05,
+      "loss": 1.6174,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7139295730757246,
+      "grad_norm": 0.4661828577518463,
+      "learning_rate": 1.8869267041977817e-05,
+      "loss": 1.7112,
+      "step": 6873
+    },
+    {
+      "epoch": 0.7140334475953048,
+      "grad_norm": 0.4062906503677368,
+      "learning_rate": 1.8856500516281684e-05,
+      "loss": 1.6291,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7141373221148852,
+      "grad_norm": 0.5327140688896179,
+      "learning_rate": 1.8843737307125275e-05,
+      "loss": 1.4871,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7142411966344656,
+      "grad_norm": 0.4605855643749237,
+      "learning_rate": 1.883097741586784e-05,
+      "loss": 1.4438,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7143450711540459,
+      "grad_norm": 0.4348846673965454,
+      "learning_rate": 1.8818220843868178e-05,
+      "loss": 1.6393,
+      "step": 6877
+    },
+    {
+      "epoch": 0.7144489456736263,
+      "grad_norm": 0.41968026757240295,
+      "learning_rate": 1.8805467592484772e-05,
+      "loss": 1.7655,
+      "step": 6878
+    },
+    {
+      "epoch": 0.7145528201932067,
+      "grad_norm": 0.47305288910865784,
+      "learning_rate": 1.8792717663075738e-05,
+      "loss": 1.9356,
+      "step": 6879
+    },
+    {
+      "epoch": 0.7146566947127869,
+      "grad_norm": 0.4094095826148987,
+      "learning_rate": 1.8779971056998846e-05,
+      "loss": 1.5515,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7147605692323673,
+      "grad_norm": 0.41150006651878357,
+      "learning_rate": 1.876722777561151e-05,
+      "loss": 1.7381,
+      "step": 6881
+    },
+    {
+      "epoch": 0.7148644437519477,
+      "grad_norm": 0.405069500207901,
+      "learning_rate": 1.8754487820270778e-05,
+      "loss": 1.5871,
+      "step": 6882
+    },
+    {
+      "epoch": 0.714968318271528,
+      "grad_norm": 0.4334279000759125,
+      "learning_rate": 1.87417511923334e-05,
+      "loss": 1.7346,
+      "step": 6883
+    },
+    {
+      "epoch": 0.7150721927911083,
+      "grad_norm": 0.4156447947025299,
+      "learning_rate": 1.872901789315566e-05,
+      "loss": 1.7387,
+      "step": 6884
+    },
+    {
+      "epoch": 0.7151760673106887,
+      "grad_norm": 0.4099999666213989,
+      "learning_rate": 1.8716287924093613e-05,
+      "loss": 1.6588,
+      "step": 6885
+    },
+    {
+      "epoch": 0.715279941830269,
+      "grad_norm": 0.39068254828453064,
+      "learning_rate": 1.8703561286502896e-05,
+      "loss": 1.5263,
+      "step": 6886
+    },
+    {
+      "epoch": 0.7153838163498494,
+      "grad_norm": 0.4342229664325714,
+      "learning_rate": 1.8690837981738747e-05,
+      "loss": 1.6721,
+      "step": 6887
+    },
+    {
+      "epoch": 0.7154876908694298,
+      "grad_norm": 0.40513840317726135,
+      "learning_rate": 1.867811801115614e-05,
+      "loss": 1.6021,
+      "step": 6888
+    },
+    {
+      "epoch": 0.71559156538901,
+      "grad_norm": 0.3993147313594818,
+      "learning_rate": 1.8665401376109654e-05,
+      "loss": 1.7418,
+      "step": 6889
+    },
+    {
+      "epoch": 0.7156954399085904,
+      "grad_norm": 0.4815475046634674,
+      "learning_rate": 1.86526880779535e-05,
+      "loss": 1.9432,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7157993144281708,
+      "grad_norm": 0.4066917896270752,
+      "learning_rate": 1.863997811804155e-05,
+      "loss": 1.4421,
+      "step": 6891
+    },
+    {
+      "epoch": 0.7159031889477511,
+      "grad_norm": 0.4122363030910492,
+      "learning_rate": 1.862727149772731e-05,
+      "loss": 1.6838,
+      "step": 6892
+    },
+    {
+      "epoch": 0.7160070634673315,
+      "grad_norm": 0.41208595037460327,
+      "learning_rate": 1.8614568218363937e-05,
+      "loss": 1.6661,
+      "step": 6893
+    },
+    {
+      "epoch": 0.7161109379869118,
+      "grad_norm": 0.4167573153972626,
+      "learning_rate": 1.860186828130422e-05,
+      "loss": 1.7035,
+      "step": 6894
+    },
+    {
+      "epoch": 0.7162148125064921,
+      "grad_norm": 0.40936407446861267,
+      "learning_rate": 1.8589171687900646e-05,
+      "loss": 1.5622,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7163186870260725,
+      "grad_norm": 0.400158554315567,
+      "learning_rate": 1.8576478439505248e-05,
+      "loss": 1.6196,
+      "step": 6896
+    },
+    {
+      "epoch": 0.7164225615456529,
+      "grad_norm": 0.42571789026260376,
+      "learning_rate": 1.856378853746979e-05,
+      "loss": 1.7369,
+      "step": 6897
+    },
+    {
+      "epoch": 0.7165264360652333,
+      "grad_norm": 0.395346462726593,
+      "learning_rate": 1.855110198314565e-05,
+      "loss": 1.7267,
+      "step": 6898
+    },
+    {
+      "epoch": 0.7166303105848135,
+      "grad_norm": 0.43957024812698364,
+      "learning_rate": 1.8538418777883825e-05,
+      "loss": 1.7538,
+      "step": 6899
+    },
+    {
+      "epoch": 0.7167341851043939,
+      "grad_norm": 0.41253888607025146,
+      "learning_rate": 1.8525738923035003e-05,
+      "loss": 1.5747,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7168380596239743,
+      "grad_norm": 0.4174502193927765,
+      "learning_rate": 1.8513062419949452e-05,
+      "loss": 1.8062,
+      "step": 6901
+    },
+    {
+      "epoch": 0.7169419341435546,
+      "grad_norm": 0.42370733618736267,
+      "learning_rate": 1.8500389269977186e-05,
+      "loss": 1.668,
+      "step": 6902
+    },
+    {
+      "epoch": 0.7170458086631349,
+      "grad_norm": 0.41417741775512695,
+      "learning_rate": 1.848771947446772e-05,
+      "loss": 1.6166,
+      "step": 6903
+    },
+    {
+      "epoch": 0.7171496831827153,
+      "grad_norm": 0.4121987819671631,
+      "learning_rate": 1.847505303477034e-05,
+      "loss": 1.5487,
+      "step": 6904
+    },
+    {
+      "epoch": 0.7172535577022956,
+      "grad_norm": 0.39274632930755615,
+      "learning_rate": 1.8462389952233905e-05,
+      "loss": 1.5665,
+      "step": 6905
+    },
+    {
+      "epoch": 0.717357432221876,
+      "grad_norm": 0.41530582308769226,
+      "learning_rate": 1.8449730228206945e-05,
+      "loss": 1.4904,
+      "step": 6906
+    },
+    {
+      "epoch": 0.7174613067414564,
+      "grad_norm": 0.4424355626106262,
+      "learning_rate": 1.8437073864037608e-05,
+      "loss": 1.6915,
+      "step": 6907
+    },
+    {
+      "epoch": 0.7175651812610366,
+      "grad_norm": 0.41284075379371643,
+      "learning_rate": 1.842442086107371e-05,
+      "loss": 1.6279,
+      "step": 6908
+    },
+    {
+      "epoch": 0.717669055780617,
+      "grad_norm": 0.4062071442604065,
+      "learning_rate": 1.8411771220662684e-05,
+      "loss": 1.6648,
+      "step": 6909
+    },
+    {
+      "epoch": 0.7177729303001974,
+      "grad_norm": 0.4447495639324188,
+      "learning_rate": 1.8399124944151614e-05,
+      "loss": 1.755,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7178768048197777,
+      "grad_norm": 0.4060676395893097,
+      "learning_rate": 1.838648203288727e-05,
+      "loss": 1.7247,
+      "step": 6911
+    },
+    {
+      "epoch": 0.717980679339358,
+      "grad_norm": 0.4383039176464081,
+      "learning_rate": 1.8373842488215966e-05,
+      "loss": 1.5317,
+      "step": 6912
+    },
+    {
+      "epoch": 0.7180845538589384,
+      "grad_norm": 0.42534106969833374,
+      "learning_rate": 1.836120631148376e-05,
+      "loss": 1.7154,
+      "step": 6913
+    },
+    {
+      "epoch": 0.7181884283785187,
+      "grad_norm": 0.4181063175201416,
+      "learning_rate": 1.8348573504036305e-05,
+      "loss": 1.7214,
+      "step": 6914
+    },
+    {
+      "epoch": 0.7182923028980991,
+      "grad_norm": 0.40229707956314087,
+      "learning_rate": 1.8335944067218846e-05,
+      "loss": 1.652,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7183961774176795,
+      "grad_norm": 0.4175788462162018,
+      "learning_rate": 1.8323318002376372e-05,
+      "loss": 1.6764,
+      "step": 6916
+    },
+    {
+      "epoch": 0.7185000519372597,
+      "grad_norm": 0.4145587682723999,
+      "learning_rate": 1.831069531085344e-05,
+      "loss": 1.7295,
+      "step": 6917
+    },
+    {
+      "epoch": 0.7186039264568401,
+      "grad_norm": 0.4142177700996399,
+      "learning_rate": 1.829807599399428e-05,
+      "loss": 1.5781,
+      "step": 6918
+    },
+    {
+      "epoch": 0.7187078009764205,
+      "grad_norm": 0.4172203540802002,
+      "learning_rate": 1.828546005314274e-05,
+      "loss": 1.6024,
+      "step": 6919
+    },
+    {
+      "epoch": 0.7188116754960008,
+      "grad_norm": 0.42822572588920593,
+      "learning_rate": 1.8272847489642314e-05,
+      "loss": 1.7269,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7189155500155812,
+      "grad_norm": 0.4436422288417816,
+      "learning_rate": 1.826023830483616e-05,
+      "loss": 1.6872,
+      "step": 6921
+    },
+    {
+      "epoch": 0.7190194245351615,
+      "grad_norm": 0.39383989572525024,
+      "learning_rate": 1.8247632500067017e-05,
+      "loss": 1.3766,
+      "step": 6922
+    },
+    {
+      "epoch": 0.7191232990547419,
+      "grad_norm": 0.4555588364601135,
+      "learning_rate": 1.8235030076677362e-05,
+      "loss": 1.7151,
+      "step": 6923
+    },
+    {
+      "epoch": 0.7192271735743222,
+      "grad_norm": 0.42651811242103577,
+      "learning_rate": 1.8222431036009226e-05,
+      "loss": 1.6841,
+      "step": 6924
+    },
+    {
+      "epoch": 0.7193310480939026,
+      "grad_norm": 0.4434896409511566,
+      "learning_rate": 1.8209835379404304e-05,
+      "loss": 1.5951,
+      "step": 6925
+    },
+    {
+      "epoch": 0.719434922613483,
+      "grad_norm": 0.38876456022262573,
+      "learning_rate": 1.819724310820395e-05,
+      "loss": 1.6513,
+      "step": 6926
+    },
+    {
+      "epoch": 0.7195387971330632,
+      "grad_norm": 0.4082390069961548,
+      "learning_rate": 1.8184654223749126e-05,
+      "loss": 1.63,
+      "step": 6927
+    },
+    {
+      "epoch": 0.7196426716526436,
+      "grad_norm": 0.5180742740631104,
+      "learning_rate": 1.8172068727380464e-05,
+      "loss": 1.7954,
+      "step": 6928
+    },
+    {
+      "epoch": 0.719746546172224,
+      "grad_norm": 0.43790149688720703,
+      "learning_rate": 1.8159486620438193e-05,
+      "loss": 1.7982,
+      "step": 6929
+    },
+    {
+      "epoch": 0.7198504206918043,
+      "grad_norm": 0.4642382562160492,
+      "learning_rate": 1.8146907904262267e-05,
+      "loss": 1.8414,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7199542952113847,
+      "grad_norm": 0.4199405014514923,
+      "learning_rate": 1.8134332580192152e-05,
+      "loss": 1.8044,
+      "step": 6931
+    },
+    {
+      "epoch": 0.720058169730965,
+      "grad_norm": 0.4645751714706421,
+      "learning_rate": 1.8121760649567077e-05,
+      "loss": 1.6786,
+      "step": 6932
+    },
+    {
+      "epoch": 0.7201620442505453,
+      "grad_norm": 0.40321746468544006,
+      "learning_rate": 1.8109192113725826e-05,
+      "loss": 1.51,
+      "step": 6933
+    },
+    {
+      "epoch": 0.7202659187701257,
+      "grad_norm": 0.44322633743286133,
+      "learning_rate": 1.809662697400686e-05,
+      "loss": 1.5216,
+      "step": 6934
+    },
+    {
+      "epoch": 0.7203697932897061,
+      "grad_norm": 0.48086169362068176,
+      "learning_rate": 1.8084065231748265e-05,
+      "loss": 1.7657,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7204736678092863,
+      "grad_norm": 0.46815988421440125,
+      "learning_rate": 1.807150688828777e-05,
+      "loss": 1.7116,
+      "step": 6936
+    },
+    {
+      "epoch": 0.7205775423288667,
+      "grad_norm": 0.42190316319465637,
+      "learning_rate": 1.8058951944962736e-05,
+      "loss": 1.503,
+      "step": 6937
+    },
+    {
+      "epoch": 0.7206814168484471,
+      "grad_norm": 0.4263322949409485,
+      "learning_rate": 1.804640040311016e-05,
+      "loss": 1.6463,
+      "step": 6938
+    },
+    {
+      "epoch": 0.7207852913680274,
+      "grad_norm": 0.4055632948875427,
+      "learning_rate": 1.8033852264066718e-05,
+      "loss": 1.4503,
+      "step": 6939
+    },
+    {
+      "epoch": 0.7208891658876078,
+      "grad_norm": 0.40299129486083984,
+      "learning_rate": 1.802130752916864e-05,
+      "loss": 1.6289,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7209930404071881,
+      "grad_norm": 0.4563215374946594,
+      "learning_rate": 1.800876619975188e-05,
+      "loss": 1.6664,
+      "step": 6941
+    },
+    {
+      "epoch": 0.7210969149267684,
+      "grad_norm": 0.4018884599208832,
+      "learning_rate": 1.7996228277151995e-05,
+      "loss": 1.5518,
+      "step": 6942
+    },
+    {
+      "epoch": 0.7212007894463488,
+      "grad_norm": 0.4353574514389038,
+      "learning_rate": 1.7983693762704122e-05,
+      "loss": 1.5822,
+      "step": 6943
+    },
+    {
+      "epoch": 0.7213046639659292,
+      "grad_norm": 0.4123103618621826,
+      "learning_rate": 1.797116265774315e-05,
+      "loss": 1.6165,
+      "step": 6944
+    },
+    {
+      "epoch": 0.7214085384855095,
+      "grad_norm": 0.4055830240249634,
+      "learning_rate": 1.7958634963603516e-05,
+      "loss": 1.5928,
+      "step": 6945
+    },
+    {
+      "epoch": 0.7215124130050898,
+      "grad_norm": 0.439450204372406,
+      "learning_rate": 1.7946110681619327e-05,
+      "loss": 1.8438,
+      "step": 6946
+    },
+    {
+      "epoch": 0.7216162875246702,
+      "grad_norm": 0.39102089405059814,
+      "learning_rate": 1.7933589813124325e-05,
+      "loss": 1.6103,
+      "step": 6947
+    },
+    {
+      "epoch": 0.7217201620442505,
+      "grad_norm": 0.39481091499328613,
+      "learning_rate": 1.7921072359451872e-05,
+      "loss": 1.6833,
+      "step": 6948
+    },
+    {
+      "epoch": 0.7218240365638309,
+      "grad_norm": 0.4227730333805084,
+      "learning_rate": 1.7908558321934992e-05,
+      "loss": 1.5623,
+      "step": 6949
+    },
+    {
+      "epoch": 0.7219279110834113,
+      "grad_norm": 0.41327497363090515,
+      "learning_rate": 1.789604770190631e-05,
+      "loss": 1.5559,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7220317856029916,
+      "grad_norm": 0.44786590337753296,
+      "learning_rate": 1.7883540500698148e-05,
+      "loss": 1.6942,
+      "step": 6951
+    },
+    {
+      "epoch": 0.7221356601225719,
+      "grad_norm": 0.40318799018859863,
+      "learning_rate": 1.7871036719642398e-05,
+      "loss": 1.4997,
+      "step": 6952
+    },
+    {
+      "epoch": 0.7222395346421523,
+      "grad_norm": 0.43783068656921387,
+      "learning_rate": 1.7858536360070626e-05,
+      "loss": 1.8249,
+      "step": 6953
+    },
+    {
+      "epoch": 0.7223434091617327,
+      "grad_norm": 0.4172214865684509,
+      "learning_rate": 1.784603942331402e-05,
+      "loss": 1.6494,
+      "step": 6954
+    },
+    {
+      "epoch": 0.722447283681313,
+      "grad_norm": 0.3986718952655792,
+      "learning_rate": 1.7833545910703402e-05,
+      "loss": 1.5129,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7225511582008933,
+      "grad_norm": 0.4317796528339386,
+      "learning_rate": 1.7821055823569238e-05,
+      "loss": 1.4516,
+      "step": 6956
+    },
+    {
+      "epoch": 0.7226550327204737,
+      "grad_norm": 0.42882880568504333,
+      "learning_rate": 1.7808569163241605e-05,
+      "loss": 1.7454,
+      "step": 6957
+    },
+    {
+      "epoch": 0.722758907240054,
+      "grad_norm": 0.4157145917415619,
+      "learning_rate": 1.7796085931050287e-05,
+      "loss": 1.6144,
+      "step": 6958
+    },
+    {
+      "epoch": 0.7228627817596344,
+      "grad_norm": 0.4417478144168854,
+      "learning_rate": 1.778360612832458e-05,
+      "loss": 1.6667,
+      "step": 6959
+    },
+    {
+      "epoch": 0.7229666562792147,
+      "grad_norm": 0.47706910967826843,
+      "learning_rate": 1.7771129756393547e-05,
+      "loss": 1.7617,
+      "step": 6960
+    },
+    {
+      "epoch": 0.723070530798795,
+      "grad_norm": 0.38705945014953613,
+      "learning_rate": 1.77586568165858e-05,
+      "loss": 1.6515,
+      "step": 6961
+    },
+    {
+      "epoch": 0.7231744053183754,
+      "grad_norm": 0.42276665568351746,
+      "learning_rate": 1.7746187310229606e-05,
+      "loss": 1.7102,
+      "step": 6962
+    },
+    {
+      "epoch": 0.7232782798379558,
+      "grad_norm": 0.44928818941116333,
+      "learning_rate": 1.773372123865287e-05,
+      "loss": 1.8452,
+      "step": 6963
+    },
+    {
+      "epoch": 0.7233821543575361,
+      "grad_norm": 0.5049888491630554,
+      "learning_rate": 1.7721258603183144e-05,
+      "loss": 1.6948,
+      "step": 6964
+    },
+    {
+      "epoch": 0.7234860288771164,
+      "grad_norm": 0.4933055639266968,
+      "learning_rate": 1.770879940514759e-05,
+      "loss": 1.9625,
+      "step": 6965
+    },
+    {
+      "epoch": 0.7235899033966968,
+      "grad_norm": 0.4487113058567047,
+      "learning_rate": 1.7696343645873e-05,
+      "loss": 1.6516,
+      "step": 6966
+    },
+    {
+      "epoch": 0.7236937779162771,
+      "grad_norm": 0.4504063129425049,
+      "learning_rate": 1.768389132668587e-05,
+      "loss": 1.7148,
+      "step": 6967
+    },
+    {
+      "epoch": 0.7237976524358575,
+      "grad_norm": 0.4117870628833771,
+      "learning_rate": 1.7671442448912202e-05,
+      "loss": 1.457,
+      "step": 6968
+    },
+    {
+      "epoch": 0.7239015269554379,
+      "grad_norm": 0.4312496781349182,
+      "learning_rate": 1.7658997013877754e-05,
+      "loss": 1.5079,
+      "step": 6969
+    },
+    {
+      "epoch": 0.7240054014750181,
+      "grad_norm": 0.45666730403900146,
+      "learning_rate": 1.764655502290788e-05,
+      "loss": 1.7927,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7241092759945985,
+      "grad_norm": 0.41350269317626953,
+      "learning_rate": 1.763411647732749e-05,
+      "loss": 1.6194,
+      "step": 6971
+    },
+    {
+      "epoch": 0.7242131505141789,
+      "grad_norm": 0.44214317202568054,
+      "learning_rate": 1.7621681378461245e-05,
+      "loss": 1.5284,
+      "step": 6972
+    },
+    {
+      "epoch": 0.7243170250337592,
+      "grad_norm": 0.5224699974060059,
+      "learning_rate": 1.7609249727633382e-05,
+      "loss": 1.3562,
+      "step": 6973
+    },
+    {
+      "epoch": 0.7244208995533395,
+      "grad_norm": 0.4053601622581482,
+      "learning_rate": 1.759682152616776e-05,
+      "loss": 1.6075,
+      "step": 6974
+    },
+    {
+      "epoch": 0.7245247740729199,
+      "grad_norm": 0.39831477403640747,
+      "learning_rate": 1.758439677538789e-05,
+      "loss": 1.5422,
+      "step": 6975
+    },
+    {
+      "epoch": 0.7246286485925003,
+      "grad_norm": 0.4336235225200653,
+      "learning_rate": 1.7571975476616897e-05,
+      "loss": 1.6262,
+      "step": 6976
+    },
+    {
+      "epoch": 0.7247325231120806,
+      "grad_norm": 0.4514787495136261,
+      "learning_rate": 1.7559557631177605e-05,
+      "loss": 1.6682,
+      "step": 6977
+    },
+    {
+      "epoch": 0.724836397631661,
+      "grad_norm": 0.4214664101600647,
+      "learning_rate": 1.754714324039234e-05,
+      "loss": 1.836,
+      "step": 6978
+    },
+    {
+      "epoch": 0.7249402721512413,
+      "grad_norm": 0.4808892607688904,
+      "learning_rate": 1.7534732305583196e-05,
+      "loss": 1.7584,
+      "step": 6979
+    },
+    {
+      "epoch": 0.7250441466708216,
+      "grad_norm": 0.44463905692100525,
+      "learning_rate": 1.7522324828071822e-05,
+      "loss": 1.6672,
+      "step": 6980
+    },
+    {
+      "epoch": 0.725148021190402,
+      "grad_norm": 0.4203853905200958,
+      "learning_rate": 1.750992080917952e-05,
+      "loss": 1.5338,
+      "step": 6981
+    },
+    {
+      "epoch": 0.7252518957099824,
+      "grad_norm": 0.41992905735969543,
+      "learning_rate": 1.7497520250227222e-05,
+      "loss": 1.6543,
+      "step": 6982
+    },
+    {
+      "epoch": 0.7253557702295627,
+      "grad_norm": 0.42039117217063904,
+      "learning_rate": 1.7485123152535488e-05,
+      "loss": 1.6052,
+      "step": 6983
+    },
+    {
+      "epoch": 0.725459644749143,
+      "grad_norm": 0.4649543762207031,
+      "learning_rate": 1.7472729517424508e-05,
+      "loss": 1.7914,
+      "step": 6984
+    },
+    {
+      "epoch": 0.7255635192687234,
+      "grad_norm": 0.39799681305885315,
+      "learning_rate": 1.7460339346214095e-05,
+      "loss": 1.6915,
+      "step": 6985
+    },
+    {
+      "epoch": 0.7256673937883037,
+      "grad_norm": 0.4391082227230072,
+      "learning_rate": 1.744795264022376e-05,
+      "loss": 1.7437,
+      "step": 6986
+    },
+    {
+      "epoch": 0.7257712683078841,
+      "grad_norm": 0.4607831835746765,
+      "learning_rate": 1.7435569400772516e-05,
+      "loss": 1.8705,
+      "step": 6987
+    },
+    {
+      "epoch": 0.7258751428274645,
+      "grad_norm": 0.4046463966369629,
+      "learning_rate": 1.742318962917914e-05,
+      "loss": 1.7026,
+      "step": 6988
+    },
+    {
+      "epoch": 0.7259790173470447,
+      "grad_norm": 0.4511662721633911,
+      "learning_rate": 1.7410813326761955e-05,
+      "loss": 1.7673,
+      "step": 6989
+    },
+    {
+      "epoch": 0.7260828918666251,
+      "grad_norm": 0.3973452150821686,
+      "learning_rate": 1.7398440494838947e-05,
+      "loss": 1.5597,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7261867663862055,
+      "grad_norm": 0.4389094412326813,
+      "learning_rate": 1.7386071134727726e-05,
+      "loss": 1.8227,
+      "step": 6991
+    },
+    {
+      "epoch": 0.7262906409057858,
+      "grad_norm": 0.3820962905883789,
+      "learning_rate": 1.7373705247745532e-05,
+      "loss": 1.5315,
+      "step": 6992
+    },
+    {
+      "epoch": 0.7263945154253661,
+      "grad_norm": 0.4259055554866791,
+      "learning_rate": 1.7361342835209225e-05,
+      "loss": 1.598,
+      "step": 6993
+    },
+    {
+      "epoch": 0.7264983899449465,
+      "grad_norm": 0.4085586965084076,
+      "learning_rate": 1.7348983898435307e-05,
+      "loss": 1.7899,
+      "step": 6994
+    },
+    {
+      "epoch": 0.7266022644645268,
+      "grad_norm": 0.4286569654941559,
+      "learning_rate": 1.733662843873995e-05,
+      "loss": 1.7511,
+      "step": 6995
+    },
+    {
+      "epoch": 0.7267061389841072,
+      "grad_norm": 0.4160563051700592,
+      "learning_rate": 1.7324276457438843e-05,
+      "loss": 1.741,
+      "step": 6996
+    },
+    {
+      "epoch": 0.7268100135036876,
+      "grad_norm": 0.47089239954948425,
+      "learning_rate": 1.7311927955847428e-05,
+      "loss": 1.8191,
+      "step": 6997
+    },
+    {
+      "epoch": 0.7269138880232678,
+      "grad_norm": 0.41623279452323914,
+      "learning_rate": 1.729958293528073e-05,
+      "loss": 1.5197,
+      "step": 6998
+    },
+    {
+      "epoch": 0.7270177625428482,
+      "grad_norm": 0.41076546907424927,
+      "learning_rate": 1.728724139705334e-05,
+      "loss": 1.6748,
+      "step": 6999
+    },
+    {
+      "epoch": 0.7271216370624286,
+      "grad_norm": 0.4390968382358551,
+      "learning_rate": 1.7274903342479586e-05,
+      "loss": 1.6716,
+      "step": 7000
+    },
+    {
+      "epoch": 0.727225511582009,
+      "grad_norm": 0.4050026535987854,
+      "learning_rate": 1.726256877287335e-05,
+      "loss": 1.6378,
+      "step": 7001
+    },
+    {
+      "epoch": 0.7273293861015893,
+      "grad_norm": 0.41926950216293335,
+      "learning_rate": 1.725023768954821e-05,
+      "loss": 1.5495,
+      "step": 7002
+    },
+    {
+      "epoch": 0.7274332606211696,
+      "grad_norm": 0.4020102024078369,
+      "learning_rate": 1.7237910093817278e-05,
+      "loss": 1.5418,
+      "step": 7003
+    },
+    {
+      "epoch": 0.72753713514075,
+      "grad_norm": 0.42248985171318054,
+      "learning_rate": 1.7225585986993353e-05,
+      "loss": 1.5632,
+      "step": 7004
+    },
+    {
+      "epoch": 0.7276410096603303,
+      "grad_norm": 0.4645000696182251,
+      "learning_rate": 1.7213265370388898e-05,
+      "loss": 1.7221,
+      "step": 7005
+    },
+    {
+      "epoch": 0.7277448841799107,
+      "grad_norm": 0.45067256689071655,
+      "learning_rate": 1.7200948245315908e-05,
+      "loss": 1.7604,
+      "step": 7006
+    },
+    {
+      "epoch": 0.7278487586994911,
+      "grad_norm": 0.40143445134162903,
+      "learning_rate": 1.7188634613086103e-05,
+      "loss": 1.6143,
+      "step": 7007
+    },
+    {
+      "epoch": 0.7279526332190713,
+      "grad_norm": 0.37583884596824646,
+      "learning_rate": 1.717632447501077e-05,
+      "loss": 1.5589,
+      "step": 7008
+    },
+    {
+      "epoch": 0.7280565077386517,
+      "grad_norm": 0.4202621877193451,
+      "learning_rate": 1.7164017832400846e-05,
+      "loss": 1.5601,
+      "step": 7009
+    },
+    {
+      "epoch": 0.7281603822582321,
+      "grad_norm": 0.39898592233657837,
+      "learning_rate": 1.71517146865669e-05,
+      "loss": 1.6215,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7282642567778124,
+      "grad_norm": 0.4259057641029358,
+      "learning_rate": 1.7139415038819108e-05,
+      "loss": 1.8012,
+      "step": 7011
+    },
+    {
+      "epoch": 0.7283681312973928,
+      "grad_norm": 0.3842464089393616,
+      "learning_rate": 1.7127118890467292e-05,
+      "loss": 1.6024,
+      "step": 7012
+    },
+    {
+      "epoch": 0.7284720058169731,
+      "grad_norm": 0.3995085060596466,
+      "learning_rate": 1.7114826242820875e-05,
+      "loss": 1.6076,
+      "step": 7013
+    },
+    {
+      "epoch": 0.7285758803365534,
+      "grad_norm": 0.37243375182151794,
+      "learning_rate": 1.7102537097188986e-05,
+      "loss": 1.5847,
+      "step": 7014
+    },
+    {
+      "epoch": 0.7286797548561338,
+      "grad_norm": 0.46955230832099915,
+      "learning_rate": 1.7090251454880258e-05,
+      "loss": 1.8128,
+      "step": 7015
+    },
+    {
+      "epoch": 0.7287836293757142,
+      "grad_norm": 0.40310418605804443,
+      "learning_rate": 1.707796931720305e-05,
+      "loss": 1.5522,
+      "step": 7016
+    },
+    {
+      "epoch": 0.7288875038952944,
+      "grad_norm": 0.4547189176082611,
+      "learning_rate": 1.706569068546532e-05,
+      "loss": 1.8937,
+      "step": 7017
+    },
+    {
+      "epoch": 0.7289913784148748,
+      "grad_norm": 0.41194766759872437,
+      "learning_rate": 1.7053415560974623e-05,
+      "loss": 1.6175,
+      "step": 7018
+    },
+    {
+      "epoch": 0.7290952529344552,
+      "grad_norm": 0.4527081847190857,
+      "learning_rate": 1.7041143945038186e-05,
+      "loss": 1.7127,
+      "step": 7019
+    },
+    {
+      "epoch": 0.7291991274540355,
+      "grad_norm": 0.45364293456077576,
+      "learning_rate": 1.702887583896282e-05,
+      "loss": 1.822,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7293030019736159,
+      "grad_norm": 0.4415326714515686,
+      "learning_rate": 1.7016611244055e-05,
+      "loss": 1.7217,
+      "step": 7021
+    },
+    {
+      "epoch": 0.7294068764931962,
+      "grad_norm": 0.4117661714553833,
+      "learning_rate": 1.700435016162078e-05,
+      "loss": 1.7666,
+      "step": 7022
+    },
+    {
+      "epoch": 0.7295107510127765,
+      "grad_norm": 0.41793352365493774,
+      "learning_rate": 1.699209259296593e-05,
+      "loss": 1.6506,
+      "step": 7023
+    },
+    {
+      "epoch": 0.7296146255323569,
+      "grad_norm": 0.41876649856567383,
+      "learning_rate": 1.697983853939571e-05,
+      "loss": 1.6463,
+      "step": 7024
+    },
+    {
+      "epoch": 0.7297185000519373,
+      "grad_norm": 0.46522414684295654,
+      "learning_rate": 1.6967588002215134e-05,
+      "loss": 1.6875,
+      "step": 7025
+    },
+    {
+      "epoch": 0.7298223745715176,
+      "grad_norm": 0.40117374062538147,
+      "learning_rate": 1.6955340982728797e-05,
+      "loss": 1.6181,
+      "step": 7026
+    },
+    {
+      "epoch": 0.7299262490910979,
+      "grad_norm": 0.4233303368091583,
+      "learning_rate": 1.694309748224085e-05,
+      "loss": 1.7707,
+      "step": 7027
+    },
+    {
+      "epoch": 0.7300301236106783,
+      "grad_norm": 0.42827704548835754,
+      "learning_rate": 1.693085750205518e-05,
+      "loss": 1.6969,
+      "step": 7028
+    },
+    {
+      "epoch": 0.7301339981302587,
+      "grad_norm": 0.42237424850463867,
+      "learning_rate": 1.6918621043475225e-05,
+      "loss": 1.5153,
+      "step": 7029
+    },
+    {
+      "epoch": 0.730237872649839,
+      "grad_norm": 0.4134870171546936,
+      "learning_rate": 1.6906388107804118e-05,
+      "loss": 1.5392,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7303417471694194,
+      "grad_norm": 0.42920297384262085,
+      "learning_rate": 1.6894158696344515e-05,
+      "loss": 1.7474,
+      "step": 7031
+    },
+    {
+      "epoch": 0.7304456216889997,
+      "grad_norm": 0.401040643453598,
+      "learning_rate": 1.6881932810398765e-05,
+      "loss": 1.5063,
+      "step": 7032
+    },
+    {
+      "epoch": 0.73054949620858,
+      "grad_norm": 0.4182683229446411,
+      "learning_rate": 1.686971045126887e-05,
+      "loss": 1.5769,
+      "step": 7033
+    },
+    {
+      "epoch": 0.7306533707281604,
+      "grad_norm": 0.4622383713722229,
+      "learning_rate": 1.6857491620256353e-05,
+      "loss": 1.628,
+      "step": 7034
+    },
+    {
+      "epoch": 0.7307572452477408,
+      "grad_norm": 0.5033509135246277,
+      "learning_rate": 1.6845276318662472e-05,
+      "loss": 1.8296,
+      "step": 7035
+    },
+    {
+      "epoch": 0.730861119767321,
+      "grad_norm": 0.3891574442386627,
+      "learning_rate": 1.6833064547788053e-05,
+      "loss": 1.5515,
+      "step": 7036
+    },
+    {
+      "epoch": 0.7309649942869014,
+      "grad_norm": 0.44622090458869934,
+      "learning_rate": 1.6820856308933543e-05,
+      "loss": 1.7212,
+      "step": 7037
+    },
+    {
+      "epoch": 0.7310688688064818,
+      "grad_norm": 0.4310351014137268,
+      "learning_rate": 1.6808651603399024e-05,
+      "loss": 1.5695,
+      "step": 7038
+    },
+    {
+      "epoch": 0.7311727433260621,
+      "grad_norm": 0.4409294128417969,
+      "learning_rate": 1.6796450432484212e-05,
+      "loss": 1.6526,
+      "step": 7039
+    },
+    {
+      "epoch": 0.7312766178456425,
+      "grad_norm": 0.39524850249290466,
+      "learning_rate": 1.678425279748843e-05,
+      "loss": 1.5189,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7313804923652228,
+      "grad_norm": 0.42522087693214417,
+      "learning_rate": 1.6772058699710618e-05,
+      "loss": 1.6336,
+      "step": 7041
+    },
+    {
+      "epoch": 0.7314843668848031,
+      "grad_norm": 0.4241730868816376,
+      "learning_rate": 1.6759868140449397e-05,
+      "loss": 1.5303,
+      "step": 7042
+    },
+    {
+      "epoch": 0.7315882414043835,
+      "grad_norm": 0.41088756918907166,
+      "learning_rate": 1.6747681121002902e-05,
+      "loss": 1.6328,
+      "step": 7043
+    },
+    {
+      "epoch": 0.7316921159239639,
+      "grad_norm": 0.43897873163223267,
+      "learning_rate": 1.6735497642669008e-05,
+      "loss": 1.7547,
+      "step": 7044
+    },
+    {
+      "epoch": 0.7317959904435442,
+      "grad_norm": 0.42431479692459106,
+      "learning_rate": 1.6723317706745144e-05,
+      "loss": 1.6887,
+      "step": 7045
+    },
+    {
+      "epoch": 0.7318998649631245,
+      "grad_norm": 0.3968948423862457,
+      "learning_rate": 1.6711141314528374e-05,
+      "loss": 1.4962,
+      "step": 7046
+    },
+    {
+      "epoch": 0.7320037394827049,
+      "grad_norm": 0.4105996787548065,
+      "learning_rate": 1.669896846731539e-05,
+      "loss": 1.6409,
+      "step": 7047
+    },
+    {
+      "epoch": 0.7321076140022852,
+      "grad_norm": 0.40278342366218567,
+      "learning_rate": 1.6686799166402507e-05,
+      "loss": 1.5295,
+      "step": 7048
+    },
+    {
+      "epoch": 0.7322114885218656,
+      "grad_norm": 0.3999069929122925,
+      "learning_rate": 1.667463341308566e-05,
+      "loss": 1.5359,
+      "step": 7049
+    },
+    {
+      "epoch": 0.732315363041446,
+      "grad_norm": 0.3840470016002655,
+      "learning_rate": 1.666247120866039e-05,
+      "loss": 1.5948,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7324192375610262,
+      "grad_norm": 0.40649813413619995,
+      "learning_rate": 1.6650312554421928e-05,
+      "loss": 1.4639,
+      "step": 7051
+    },
+    {
+      "epoch": 0.7325231120806066,
+      "grad_norm": 0.46434590220451355,
+      "learning_rate": 1.663815745166502e-05,
+      "loss": 1.8214,
+      "step": 7052
+    },
+    {
+      "epoch": 0.732626986600187,
+      "grad_norm": 0.434417724609375,
+      "learning_rate": 1.66260059016841e-05,
+      "loss": 1.5394,
+      "step": 7053
+    },
+    {
+      "epoch": 0.7327308611197674,
+      "grad_norm": 0.39939597249031067,
+      "learning_rate": 1.6613857905773238e-05,
+      "loss": 1.5732,
+      "step": 7054
+    },
+    {
+      "epoch": 0.7328347356393476,
+      "grad_norm": 0.39410555362701416,
+      "learning_rate": 1.6601713465226086e-05,
+      "loss": 1.6296,
+      "step": 7055
+    },
+    {
+      "epoch": 0.732938610158928,
+      "grad_norm": 0.4101269245147705,
+      "learning_rate": 1.6589572581335938e-05,
+      "loss": 1.6349,
+      "step": 7056
+    },
+    {
+      "epoch": 0.7330424846785084,
+      "grad_norm": 0.4318089485168457,
+      "learning_rate": 1.6577435255395677e-05,
+      "loss": 1.6986,
+      "step": 7057
+    },
+    {
+      "epoch": 0.7331463591980887,
+      "grad_norm": 0.4095480442047119,
+      "learning_rate": 1.6565301488697893e-05,
+      "loss": 1.7157,
+      "step": 7058
+    },
+    {
+      "epoch": 0.7332502337176691,
+      "grad_norm": 0.4215502440929413,
+      "learning_rate": 1.655317128253468e-05,
+      "loss": 1.587,
+      "step": 7059
+    },
+    {
+      "epoch": 0.7333541082372494,
+      "grad_norm": 0.472887247800827,
+      "learning_rate": 1.6541044638197818e-05,
+      "loss": 1.7887,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7334579827568297,
+      "grad_norm": 0.4307557940483093,
+      "learning_rate": 1.6528921556978737e-05,
+      "loss": 1.7428,
+      "step": 7061
+    },
+    {
+      "epoch": 0.7335618572764101,
+      "grad_norm": 0.4611065685749054,
+      "learning_rate": 1.65168020401684e-05,
+      "loss": 1.79,
+      "step": 7062
+    },
+    {
+      "epoch": 0.7336657317959905,
+      "grad_norm": 0.45637285709381104,
+      "learning_rate": 1.650468608905748e-05,
+      "loss": 1.8929,
+      "step": 7063
+    },
+    {
+      "epoch": 0.7337696063155708,
+      "grad_norm": 0.41570550203323364,
+      "learning_rate": 1.649257370493622e-05,
+      "loss": 1.7476,
+      "step": 7064
+    },
+    {
+      "epoch": 0.7338734808351511,
+      "grad_norm": 0.40172329545021057,
+      "learning_rate": 1.6480464889094493e-05,
+      "loss": 1.762,
+      "step": 7065
+    },
+    {
+      "epoch": 0.7339773553547315,
+      "grad_norm": 0.4335883557796478,
+      "learning_rate": 1.646835964282179e-05,
+      "loss": 1.5714,
+      "step": 7066
+    },
+    {
+      "epoch": 0.7340812298743118,
+      "grad_norm": 0.4139271676540375,
+      "learning_rate": 1.6456257967407228e-05,
+      "loss": 1.6522,
+      "step": 7067
+    },
+    {
+      "epoch": 0.7341851043938922,
+      "grad_norm": 0.4054241478443146,
+      "learning_rate": 1.6444159864139542e-05,
+      "loss": 1.7425,
+      "step": 7068
+    },
+    {
+      "epoch": 0.7342889789134726,
+      "grad_norm": 0.4233381450176239,
+      "learning_rate": 1.6432065334307074e-05,
+      "loss": 1.6434,
+      "step": 7069
+    },
+    {
+      "epoch": 0.7343928534330528,
+      "grad_norm": 0.4486279785633087,
+      "learning_rate": 1.6419974379197834e-05,
+      "loss": 1.7726,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7344967279526332,
+      "grad_norm": 0.4719793200492859,
+      "learning_rate": 1.6407887000099354e-05,
+      "loss": 1.5734,
+      "step": 7071
+    },
+    {
+      "epoch": 0.7346006024722136,
+      "grad_norm": 0.4241810142993927,
+      "learning_rate": 1.6395803198298905e-05,
+      "loss": 1.7336,
+      "step": 7072
+    },
+    {
+      "epoch": 0.7347044769917939,
+      "grad_norm": 0.40580904483795166,
+      "learning_rate": 1.6383722975083282e-05,
+      "loss": 1.6019,
+      "step": 7073
+    },
+    {
+      "epoch": 0.7348083515113742,
+      "grad_norm": 0.46320223808288574,
+      "learning_rate": 1.6371646331738953e-05,
+      "loss": 1.7766,
+      "step": 7074
+    },
+    {
+      "epoch": 0.7349122260309546,
+      "grad_norm": 0.41690146923065186,
+      "learning_rate": 1.6359573269551977e-05,
+      "loss": 1.6961,
+      "step": 7075
+    },
+    {
+      "epoch": 0.7350161005505349,
+      "grad_norm": 0.4090801775455475,
+      "learning_rate": 1.634750378980804e-05,
+      "loss": 1.6388,
+      "step": 7076
+    },
+    {
+      "epoch": 0.7351199750701153,
+      "grad_norm": 0.3977707624435425,
+      "learning_rate": 1.633543789379245e-05,
+      "loss": 1.6002,
+      "step": 7077
+    },
+    {
+      "epoch": 0.7352238495896957,
+      "grad_norm": 0.4486248195171356,
+      "learning_rate": 1.632337558279011e-05,
+      "loss": 1.7919,
+      "step": 7078
+    },
+    {
+      "epoch": 0.735327724109276,
+      "grad_norm": 0.41269683837890625,
+      "learning_rate": 1.6311316858085607e-05,
+      "loss": 1.5934,
+      "step": 7079
+    },
+    {
+      "epoch": 0.7354315986288563,
+      "grad_norm": 0.4171505570411682,
+      "learning_rate": 1.6299261720963093e-05,
+      "loss": 1.6255,
+      "step": 7080
+    },
+    {
+      "epoch": 0.7355354731484367,
+      "grad_norm": 0.406341552734375,
+      "learning_rate": 1.6287210172706297e-05,
+      "loss": 1.7223,
+      "step": 7081
+    },
+    {
+      "epoch": 0.7356393476680171,
+      "grad_norm": 0.43116459250450134,
+      "learning_rate": 1.627516221459866e-05,
+      "loss": 1.6397,
+      "step": 7082
+    },
+    {
+      "epoch": 0.7357432221875974,
+      "grad_norm": 0.42022863030433655,
+      "learning_rate": 1.6263117847923182e-05,
+      "loss": 1.7121,
+      "step": 7083
+    },
+    {
+      "epoch": 0.7358470967071777,
+      "grad_norm": 0.4364830553531647,
+      "learning_rate": 1.62510770739625e-05,
+      "loss": 1.5572,
+      "step": 7084
+    },
+    {
+      "epoch": 0.7359509712267581,
+      "grad_norm": 0.4395456910133362,
+      "learning_rate": 1.6239039893998842e-05,
+      "loss": 1.6877,
+      "step": 7085
+    },
+    {
+      "epoch": 0.7360548457463384,
+      "grad_norm": 0.499352365732193,
+      "learning_rate": 1.622700630931412e-05,
+      "loss": 1.4045,
+      "step": 7086
+    },
+    {
+      "epoch": 0.7361587202659188,
+      "grad_norm": 0.41505447030067444,
+      "learning_rate": 1.6214976321189768e-05,
+      "loss": 1.6824,
+      "step": 7087
+    },
+    {
+      "epoch": 0.7362625947854992,
+      "grad_norm": 0.40224704146385193,
+      "learning_rate": 1.6202949930906893e-05,
+      "loss": 1.6501,
+      "step": 7088
+    },
+    {
+      "epoch": 0.7363664693050794,
+      "grad_norm": 0.4343310296535492,
+      "learning_rate": 1.6190927139746255e-05,
+      "loss": 1.7141,
+      "step": 7089
+    },
+    {
+      "epoch": 0.7364703438246598,
+      "grad_norm": 0.4633185565471649,
+      "learning_rate": 1.617890794898812e-05,
+      "loss": 1.8566,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7365742183442402,
+      "grad_norm": 0.4461994469165802,
+      "learning_rate": 1.616689235991249e-05,
+      "loss": 1.6646,
+      "step": 7091
+    },
+    {
+      "epoch": 0.7366780928638205,
+      "grad_norm": 0.4115038812160492,
+      "learning_rate": 1.6154880373798918e-05,
+      "loss": 1.6597,
+      "step": 7092
+    },
+    {
+      "epoch": 0.7367819673834008,
+      "grad_norm": 0.47499844431877136,
+      "learning_rate": 1.6142871991926584e-05,
+      "loss": 1.8035,
+      "step": 7093
+    },
+    {
+      "epoch": 0.7368858419029812,
+      "grad_norm": 0.4350105822086334,
+      "learning_rate": 1.6130867215574284e-05,
+      "loss": 1.446,
+      "step": 7094
+    },
+    {
+      "epoch": 0.7369897164225615,
+      "grad_norm": 0.39589881896972656,
+      "learning_rate": 1.6118866046020438e-05,
+      "loss": 1.5551,
+      "step": 7095
+    },
+    {
+      "epoch": 0.7370935909421419,
+      "grad_norm": 0.458922415971756,
+      "learning_rate": 1.6106868484543076e-05,
+      "loss": 1.8941,
+      "step": 7096
+    },
+    {
+      "epoch": 0.7371974654617223,
+      "grad_norm": 0.48349758982658386,
+      "learning_rate": 1.6094874532419828e-05,
+      "loss": 1.6175,
+      "step": 7097
+    },
+    {
+      "epoch": 0.7373013399813025,
+      "grad_norm": 0.43464893102645874,
+      "learning_rate": 1.6082884190928004e-05,
+      "loss": 1.7482,
+      "step": 7098
+    },
+    {
+      "epoch": 0.7374052145008829,
+      "grad_norm": 0.43286725878715515,
+      "learning_rate": 1.6070897461344413e-05,
+      "loss": 1.7566,
+      "step": 7099
+    },
+    {
+      "epoch": 0.7375090890204633,
+      "grad_norm": 0.3972510099411011,
+      "learning_rate": 1.6058914344945607e-05,
+      "loss": 1.576,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7376129635400436,
+      "grad_norm": 0.44797611236572266,
+      "learning_rate": 1.604693484300767e-05,
+      "loss": 1.6982,
+      "step": 7101
+    },
+    {
+      "epoch": 0.737716838059624,
+      "grad_norm": 0.4418253004550934,
+      "learning_rate": 1.6034958956806333e-05,
+      "loss": 1.8107,
+      "step": 7102
+    },
+    {
+      "epoch": 0.7378207125792043,
+      "grad_norm": 0.40204495191574097,
+      "learning_rate": 1.6022986687616926e-05,
+      "loss": 1.5234,
+      "step": 7103
+    },
+    {
+      "epoch": 0.7379245870987847,
+      "grad_norm": 0.407816082239151,
+      "learning_rate": 1.6011018036714388e-05,
+      "loss": 1.7615,
+      "step": 7104
+    },
+    {
+      "epoch": 0.738028461618365,
+      "grad_norm": 0.4105852544307709,
+      "learning_rate": 1.599905300537334e-05,
+      "loss": 1.7623,
+      "step": 7105
+    },
+    {
+      "epoch": 0.7381323361379454,
+      "grad_norm": 0.4015292525291443,
+      "learning_rate": 1.598709159486789e-05,
+      "loss": 1.7175,
+      "step": 7106
+    },
+    {
+      "epoch": 0.7382362106575258,
+      "grad_norm": 0.4852696359157562,
+      "learning_rate": 1.5975133806471903e-05,
+      "loss": 1.8187,
+      "step": 7107
+    },
+    {
+      "epoch": 0.738340085177106,
+      "grad_norm": 0.4263093173503876,
+      "learning_rate": 1.596317964145877e-05,
+      "loss": 1.6612,
+      "step": 7108
+    },
+    {
+      "epoch": 0.7384439596966864,
+      "grad_norm": 0.43212834000587463,
+      "learning_rate": 1.5951229101101473e-05,
+      "loss": 1.6154,
+      "step": 7109
+    },
+    {
+      "epoch": 0.7385478342162668,
+      "grad_norm": 0.4184805452823639,
+      "learning_rate": 1.5939282186672705e-05,
+      "loss": 1.706,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7386517087358471,
+      "grad_norm": 0.42199963331222534,
+      "learning_rate": 1.5927338899444706e-05,
+      "loss": 1.6324,
+      "step": 7111
+    },
+    {
+      "epoch": 0.7387555832554275,
+      "grad_norm": 0.44483646750450134,
+      "learning_rate": 1.5915399240689325e-05,
+      "loss": 1.6332,
+      "step": 7112
+    },
+    {
+      "epoch": 0.7388594577750078,
+      "grad_norm": 0.43318289518356323,
+      "learning_rate": 1.5903463211678067e-05,
+      "loss": 1.6949,
+      "step": 7113
+    },
+    {
+      "epoch": 0.7389633322945881,
+      "grad_norm": 0.5184254050254822,
+      "learning_rate": 1.5891530813682008e-05,
+      "loss": 1.6298,
+      "step": 7114
+    },
+    {
+      "epoch": 0.7390672068141685,
+      "grad_norm": 0.4334104657173157,
+      "learning_rate": 1.5879602047971865e-05,
+      "loss": 1.4948,
+      "step": 7115
+    },
+    {
+      "epoch": 0.7391710813337489,
+      "grad_norm": 0.39444056153297424,
+      "learning_rate": 1.586767691581793e-05,
+      "loss": 1.607,
+      "step": 7116
+    },
+    {
+      "epoch": 0.7392749558533291,
+      "grad_norm": 0.3886300027370453,
+      "learning_rate": 1.58557554184902e-05,
+      "loss": 1.5812,
+      "step": 7117
+    },
+    {
+      "epoch": 0.7393788303729095,
+      "grad_norm": 0.43382641673088074,
+      "learning_rate": 1.584383755725815e-05,
+      "loss": 1.6636,
+      "step": 7118
+    },
+    {
+      "epoch": 0.7394827048924899,
+      "grad_norm": 0.40782594680786133,
+      "learning_rate": 1.5831923333390985e-05,
+      "loss": 1.4104,
+      "step": 7119
+    },
+    {
+      "epoch": 0.7395865794120702,
+      "grad_norm": 0.3772667348384857,
+      "learning_rate": 1.5820012748157464e-05,
+      "loss": 1.4713,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7396904539316506,
+      "grad_norm": 0.42232704162597656,
+      "learning_rate": 1.5808105802825975e-05,
+      "loss": 1.3777,
+      "step": 7121
+    },
+    {
+      "epoch": 0.7397943284512309,
+      "grad_norm": 0.430355042219162,
+      "learning_rate": 1.579620249866451e-05,
+      "loss": 1.6707,
+      "step": 7122
+    },
+    {
+      "epoch": 0.7398982029708112,
+      "grad_norm": 0.40983816981315613,
+      "learning_rate": 1.5784302836940678e-05,
+      "loss": 1.4956,
+      "step": 7123
+    },
+    {
+      "epoch": 0.7400020774903916,
+      "grad_norm": 0.4761542081832886,
+      "learning_rate": 1.5772406818921697e-05,
+      "loss": 1.7082,
+      "step": 7124
+    },
+    {
+      "epoch": 0.740105952009972,
+      "grad_norm": 0.4703845977783203,
+      "learning_rate": 1.576051444587439e-05,
+      "loss": 1.7969,
+      "step": 7125
+    },
+    {
+      "epoch": 0.7402098265295523,
+      "grad_norm": 0.40209317207336426,
+      "learning_rate": 1.574862571906525e-05,
+      "loss": 1.7028,
+      "step": 7126
+    },
+    {
+      "epoch": 0.7403137010491326,
+      "grad_norm": 0.39198899269104004,
+      "learning_rate": 1.5736740639760272e-05,
+      "loss": 1.6769,
+      "step": 7127
+    },
+    {
+      "epoch": 0.740417575568713,
+      "grad_norm": 0.3915342390537262,
+      "learning_rate": 1.5724859209225168e-05,
+      "loss": 1.5418,
+      "step": 7128
+    },
+    {
+      "epoch": 0.7405214500882933,
+      "grad_norm": 0.4316711127758026,
+      "learning_rate": 1.5712981428725192e-05,
+      "loss": 1.5925,
+      "step": 7129
+    },
+    {
+      "epoch": 0.7406253246078737,
+      "grad_norm": 0.41193971037864685,
+      "learning_rate": 1.5701107299525253e-05,
+      "loss": 1.5975,
+      "step": 7130
+    },
+    {
+      "epoch": 0.740729199127454,
+      "grad_norm": 0.4072656035423279,
+      "learning_rate": 1.5689236822889846e-05,
+      "loss": 1.8195,
+      "step": 7131
+    },
+    {
+      "epoch": 0.7408330736470344,
+      "grad_norm": 0.40560346841812134,
+      "learning_rate": 1.5677370000083065e-05,
+      "loss": 1.4801,
+      "step": 7132
+    },
+    {
+      "epoch": 0.7409369481666147,
+      "grad_norm": 0.4122333824634552,
+      "learning_rate": 1.566550683236868e-05,
+      "loss": 1.4249,
+      "step": 7133
+    },
+    {
+      "epoch": 0.7410408226861951,
+      "grad_norm": 0.4269520938396454,
+      "learning_rate": 1.5653647321009973e-05,
+      "loss": 1.8544,
+      "step": 7134
+    },
+    {
+      "epoch": 0.7411446972057755,
+      "grad_norm": 0.4235425293445587,
+      "learning_rate": 1.5641791467269922e-05,
+      "loss": 1.5443,
+      "step": 7135
+    },
+    {
+      "epoch": 0.7412485717253557,
+      "grad_norm": 0.49636656045913696,
+      "learning_rate": 1.56299392724111e-05,
+      "loss": 1.6799,
+      "step": 7136
+    },
+    {
+      "epoch": 0.7413524462449361,
+      "grad_norm": 0.40569061040878296,
+      "learning_rate": 1.561809073769561e-05,
+      "loss": 1.5797,
+      "step": 7137
+    },
+    {
+      "epoch": 0.7414563207645165,
+      "grad_norm": 0.42479363083839417,
+      "learning_rate": 1.5606245864385287e-05,
+      "loss": 1.6916,
+      "step": 7138
+    },
+    {
+      "epoch": 0.7415601952840968,
+      "grad_norm": 0.4160860776901245,
+      "learning_rate": 1.5594404653741494e-05,
+      "loss": 1.4902,
+      "step": 7139
+    },
+    {
+      "epoch": 0.7416640698036772,
+      "grad_norm": 0.4242292642593384,
+      "learning_rate": 1.5582567107025237e-05,
+      "loss": 1.6821,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7417679443232575,
+      "grad_norm": 0.4416872262954712,
+      "learning_rate": 1.5570733225497115e-05,
+      "loss": 1.6489,
+      "step": 7141
+    },
+    {
+      "epoch": 0.7418718188428378,
+      "grad_norm": 0.4447227418422699,
+      "learning_rate": 1.5558903010417346e-05,
+      "loss": 1.766,
+      "step": 7142
+    },
+    {
+      "epoch": 0.7419756933624182,
+      "grad_norm": 0.4173499047756195,
+      "learning_rate": 1.5547076463045756e-05,
+      "loss": 1.6973,
+      "step": 7143
+    },
+    {
+      "epoch": 0.7420795678819986,
+      "grad_norm": 0.4345002770423889,
+      "learning_rate": 1.5535253584641762e-05,
+      "loss": 1.7257,
+      "step": 7144
+    },
+    {
+      "epoch": 0.7421834424015789,
+      "grad_norm": 0.45821139216423035,
+      "learning_rate": 1.552343437646446e-05,
+      "loss": 1.7526,
+      "step": 7145
+    },
+    {
+      "epoch": 0.7422873169211592,
+      "grad_norm": 0.391783744096756,
+      "learning_rate": 1.5511618839772437e-05,
+      "loss": 1.6038,
+      "step": 7146
+    },
+    {
+      "epoch": 0.7423911914407396,
+      "grad_norm": 0.43559110164642334,
+      "learning_rate": 1.5499806975824e-05,
+      "loss": 1.6295,
+      "step": 7147
+    },
+    {
+      "epoch": 0.7424950659603199,
+      "grad_norm": 0.43846607208251953,
+      "learning_rate": 1.5487998785877005e-05,
+      "loss": 1.6083,
+      "step": 7148
+    },
+    {
+      "epoch": 0.7425989404799003,
+      "grad_norm": 0.43840745091438293,
+      "learning_rate": 1.5476194271188944e-05,
+      "loss": 1.6872,
+      "step": 7149
+    },
+    {
+      "epoch": 0.7427028149994807,
+      "grad_norm": 0.44359713792800903,
+      "learning_rate": 1.5464393433016895e-05,
+      "loss": 1.6835,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7428066895190609,
+      "grad_norm": 0.46567508578300476,
+      "learning_rate": 1.5452596272617555e-05,
+      "loss": 1.6991,
+      "step": 7151
+    },
+    {
+      "epoch": 0.7429105640386413,
+      "grad_norm": 0.4170341491699219,
+      "learning_rate": 1.5440802791247234e-05,
+      "loss": 1.5401,
+      "step": 7152
+    },
+    {
+      "epoch": 0.7430144385582217,
+      "grad_norm": 0.40833204984664917,
+      "learning_rate": 1.5429012990161827e-05,
+      "loss": 1.604,
+      "step": 7153
+    },
+    {
+      "epoch": 0.743118313077802,
+      "grad_norm": 0.4111480712890625,
+      "learning_rate": 1.5417226870616902e-05,
+      "loss": 1.6432,
+      "step": 7154
+    },
+    {
+      "epoch": 0.7432221875973823,
+      "grad_norm": 0.41619327664375305,
+      "learning_rate": 1.540544443386753e-05,
+      "loss": 1.5474,
+      "step": 7155
+    },
+    {
+      "epoch": 0.7433260621169627,
+      "grad_norm": 0.41143396496772766,
+      "learning_rate": 1.5393665681168496e-05,
+      "loss": 1.8406,
+      "step": 7156
+    },
+    {
+      "epoch": 0.7434299366365431,
+      "grad_norm": 0.40822920203208923,
+      "learning_rate": 1.5381890613774126e-05,
+      "loss": 1.589,
+      "step": 7157
+    },
+    {
+      "epoch": 0.7435338111561234,
+      "grad_norm": 0.39305150508880615,
+      "learning_rate": 1.537011923293837e-05,
+      "loss": 1.4567,
+      "step": 7158
+    },
+    {
+      "epoch": 0.7436376856757038,
+      "grad_norm": 0.40565574169158936,
+      "learning_rate": 1.5358351539914794e-05,
+      "loss": 1.5574,
+      "step": 7159
+    },
+    {
+      "epoch": 0.7437415601952841,
+      "grad_norm": 0.47624900937080383,
+      "learning_rate": 1.5346587535956547e-05,
+      "loss": 1.7159,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7438454347148644,
+      "grad_norm": 0.41582754254341125,
+      "learning_rate": 1.5334827222316457e-05,
+      "loss": 1.6007,
+      "step": 7161
+    },
+    {
+      "epoch": 0.7439493092344448,
+      "grad_norm": 0.42031049728393555,
+      "learning_rate": 1.5323070600246832e-05,
+      "loss": 1.7471,
+      "step": 7162
+    },
+    {
+      "epoch": 0.7440531837540252,
+      "grad_norm": 0.43765780329704285,
+      "learning_rate": 1.5311317670999714e-05,
+      "loss": 1.6716,
+      "step": 7163
+    },
+    {
+      "epoch": 0.7441570582736055,
+      "grad_norm": 0.46914222836494446,
+      "learning_rate": 1.52995684358267e-05,
+      "loss": 1.7037,
+      "step": 7164
+    },
+    {
+      "epoch": 0.7442609327931858,
+      "grad_norm": 0.43838274478912354,
+      "learning_rate": 1.5287822895978942e-05,
+      "loss": 1.6881,
+      "step": 7165
+    },
+    {
+      "epoch": 0.7443648073127662,
+      "grad_norm": 0.43214210867881775,
+      "learning_rate": 1.52760810527073e-05,
+      "loss": 1.7643,
+      "step": 7166
+    },
+    {
+      "epoch": 0.7444686818323465,
+      "grad_norm": 0.43604573607444763,
+      "learning_rate": 1.5264342907262164e-05,
+      "loss": 1.6576,
+      "step": 7167
+    },
+    {
+      "epoch": 0.7445725563519269,
+      "grad_norm": 0.4425446391105652,
+      "learning_rate": 1.5252608460893558e-05,
+      "loss": 1.7985,
+      "step": 7168
+    },
+    {
+      "epoch": 0.7446764308715073,
+      "grad_norm": 0.42904865741729736,
+      "learning_rate": 1.5240877714851115e-05,
+      "loss": 1.6071,
+      "step": 7169
+    },
+    {
+      "epoch": 0.7447803053910875,
+      "grad_norm": 0.4095899164676666,
+      "learning_rate": 1.5229150670384056e-05,
+      "loss": 1.6072,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7448841799106679,
+      "grad_norm": 0.4281744360923767,
+      "learning_rate": 1.5217427328741235e-05,
+      "loss": 1.642,
+      "step": 7171
+    },
+    {
+      "epoch": 0.7449880544302483,
+      "grad_norm": 0.4427133798599243,
+      "learning_rate": 1.5205707691171074e-05,
+      "loss": 1.7441,
+      "step": 7172
+    },
+    {
+      "epoch": 0.7450919289498286,
+      "grad_norm": 0.4026220142841339,
+      "learning_rate": 1.5193991758921666e-05,
+      "loss": 1.7042,
+      "step": 7173
+    },
+    {
+      "epoch": 0.745195803469409,
+      "grad_norm": 0.39205819368362427,
+      "learning_rate": 1.5182279533240606e-05,
+      "loss": 1.4862,
+      "step": 7174
+    },
+    {
+      "epoch": 0.7452996779889893,
+      "grad_norm": 0.45115184783935547,
+      "learning_rate": 1.5170571015375207e-05,
+      "loss": 1.8971,
+      "step": 7175
+    },
+    {
+      "epoch": 0.7454035525085696,
+      "grad_norm": 0.4259231686592102,
+      "learning_rate": 1.5158866206572313e-05,
+      "loss": 1.5441,
+      "step": 7176
+    },
+    {
+      "epoch": 0.74550742702815,
+      "grad_norm": 0.39843812584877014,
+      "learning_rate": 1.5147165108078398e-05,
+      "loss": 1.7365,
+      "step": 7177
+    },
+    {
+      "epoch": 0.7456113015477304,
+      "grad_norm": 0.3978004455566406,
+      "learning_rate": 1.5135467721139535e-05,
+      "loss": 1.5903,
+      "step": 7178
+    },
+    {
+      "epoch": 0.7457151760673106,
+      "grad_norm": 0.43081727623939514,
+      "learning_rate": 1.5123774047001404e-05,
+      "loss": 1.6461,
+      "step": 7179
+    },
+    {
+      "epoch": 0.745819050586891,
+      "grad_norm": 0.41507482528686523,
+      "learning_rate": 1.5112084086909296e-05,
+      "loss": 1.5398,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7459229251064714,
+      "grad_norm": 0.4333358108997345,
+      "learning_rate": 1.5100397842108078e-05,
+      "loss": 1.612,
+      "step": 7181
+    },
+    {
+      "epoch": 0.7460267996260518,
+      "grad_norm": 0.4127698540687561,
+      "learning_rate": 1.5088715313842278e-05,
+      "loss": 1.5695,
+      "step": 7182
+    },
+    {
+      "epoch": 0.7461306741456321,
+      "grad_norm": 0.4184444546699524,
+      "learning_rate": 1.507703650335598e-05,
+      "loss": 1.6183,
+      "step": 7183
+    },
+    {
+      "epoch": 0.7462345486652124,
+      "grad_norm": 0.4407637417316437,
+      "learning_rate": 1.506536141189288e-05,
+      "loss": 1.6235,
+      "step": 7184
+    },
+    {
+      "epoch": 0.7463384231847928,
+      "grad_norm": 0.40334874391555786,
+      "learning_rate": 1.5053690040696295e-05,
+      "loss": 1.6717,
+      "step": 7185
+    },
+    {
+      "epoch": 0.7464422977043731,
+      "grad_norm": 0.40933892130851746,
+      "learning_rate": 1.504202239100913e-05,
+      "loss": 1.5891,
+      "step": 7186
+    },
+    {
+      "epoch": 0.7465461722239535,
+      "grad_norm": 0.41984155774116516,
+      "learning_rate": 1.5030358464073891e-05,
+      "loss": 1.6755,
+      "step": 7187
+    },
+    {
+      "epoch": 0.7466500467435339,
+      "grad_norm": 0.44677481055259705,
+      "learning_rate": 1.501869826113269e-05,
+      "loss": 1.6644,
+      "step": 7188
+    },
+    {
+      "epoch": 0.7467539212631141,
+      "grad_norm": 0.43033236265182495,
+      "learning_rate": 1.5007041783427294e-05,
+      "loss": 1.6723,
+      "step": 7189
+    },
+    {
+      "epoch": 0.7468577957826945,
+      "grad_norm": 0.4469008445739746,
+      "learning_rate": 1.4995389032198958e-05,
+      "loss": 1.7799,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7469616703022749,
+      "grad_norm": 0.44168853759765625,
+      "learning_rate": 1.4983740008688657e-05,
+      "loss": 1.6653,
+      "step": 7191
+    },
+    {
+      "epoch": 0.7470655448218552,
+      "grad_norm": 0.41871321201324463,
+      "learning_rate": 1.4972094714136926e-05,
+      "loss": 1.5746,
+      "step": 7192
+    },
+    {
+      "epoch": 0.7471694193414355,
+      "grad_norm": 0.411448210477829,
+      "learning_rate": 1.4960453149783848e-05,
+      "loss": 1.5334,
+      "step": 7193
+    },
+    {
+      "epoch": 0.7472732938610159,
+      "grad_norm": 0.4211091101169586,
+      "learning_rate": 1.4948815316869207e-05,
+      "loss": 1.7637,
+      "step": 7194
+    },
+    {
+      "epoch": 0.7473771683805962,
+      "grad_norm": 0.46923527121543884,
+      "learning_rate": 1.4937181216632323e-05,
+      "loss": 1.7789,
+      "step": 7195
+    },
+    {
+      "epoch": 0.7474810429001766,
+      "grad_norm": 0.4714355766773224,
+      "learning_rate": 1.4925550850312142e-05,
+      "loss": 1.7327,
+      "step": 7196
+    },
+    {
+      "epoch": 0.747584917419757,
+      "grad_norm": 0.39055758714675903,
+      "learning_rate": 1.4913924219147202e-05,
+      "loss": 1.6193,
+      "step": 7197
+    },
+    {
+      "epoch": 0.7476887919393372,
+      "grad_norm": 0.4228423833847046,
+      "learning_rate": 1.490230132437565e-05,
+      "loss": 1.75,
+      "step": 7198
+    },
+    {
+      "epoch": 0.7477926664589176,
+      "grad_norm": 0.42969369888305664,
+      "learning_rate": 1.4890682167235238e-05,
+      "loss": 1.5587,
+      "step": 7199
+    },
+    {
+      "epoch": 0.747896540978498,
+      "grad_norm": 0.42206692695617676,
+      "learning_rate": 1.4879066748963295e-05,
+      "loss": 1.7437,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7480004154980783,
+      "grad_norm": 0.4134342670440674,
+      "learning_rate": 1.4867455070796816e-05,
+      "loss": 1.6074,
+      "step": 7201
+    },
+    {
+      "epoch": 0.7481042900176587,
+      "grad_norm": 0.4012831747531891,
+      "learning_rate": 1.48558471339723e-05,
+      "loss": 1.5822,
+      "step": 7202
+    },
+    {
+      "epoch": 0.748208164537239,
+      "grad_norm": 0.42914196848869324,
+      "learning_rate": 1.4844242939725944e-05,
+      "loss": 1.6097,
+      "step": 7203
+    },
+    {
+      "epoch": 0.7483120390568193,
+      "grad_norm": 0.43084225058555603,
+      "learning_rate": 1.483264248929349e-05,
+      "loss": 1.7464,
+      "step": 7204
+    },
+    {
+      "epoch": 0.7484159135763997,
+      "grad_norm": 0.4075019657611847,
+      "learning_rate": 1.4821045783910298e-05,
+      "loss": 1.6601,
+      "step": 7205
+    },
+    {
+      "epoch": 0.7485197880959801,
+      "grad_norm": 0.46284353733062744,
+      "learning_rate": 1.480945282481132e-05,
+      "loss": 1.9072,
+      "step": 7206
+    },
+    {
+      "epoch": 0.7486236626155603,
+      "grad_norm": 0.4372515380382538,
+      "learning_rate": 1.4797863613231122e-05,
+      "loss": 1.7274,
+      "step": 7207
+    },
+    {
+      "epoch": 0.7487275371351407,
+      "grad_norm": 0.3990016281604767,
+      "learning_rate": 1.478627815040386e-05,
+      "loss": 1.5798,
+      "step": 7208
+    },
+    {
+      "epoch": 0.7488314116547211,
+      "grad_norm": 0.3860011696815491,
+      "learning_rate": 1.4774696437563278e-05,
+      "loss": 1.5999,
+      "step": 7209
+    },
+    {
+      "epoch": 0.7489352861743015,
+      "grad_norm": 0.4409307837486267,
+      "learning_rate": 1.4763118475942784e-05,
+      "loss": 1.7148,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7490391606938818,
+      "grad_norm": 0.41701313853263855,
+      "learning_rate": 1.4751544266775302e-05,
+      "loss": 1.7499,
+      "step": 7211
+    },
+    {
+      "epoch": 0.7491430352134621,
+      "grad_norm": 0.40917038917541504,
+      "learning_rate": 1.473997381129341e-05,
+      "loss": 1.6405,
+      "step": 7212
+    },
+    {
+      "epoch": 0.7492469097330425,
+      "grad_norm": 0.46700426936149597,
+      "learning_rate": 1.4728407110729275e-05,
+      "loss": 1.7749,
+      "step": 7213
+    },
+    {
+      "epoch": 0.7493507842526228,
+      "grad_norm": 0.4193143844604492,
+      "learning_rate": 1.4716844166314648e-05,
+      "loss": 1.6553,
+      "step": 7214
+    },
+    {
+      "epoch": 0.7494546587722032,
+      "grad_norm": 0.4625590741634369,
+      "learning_rate": 1.4705284979280898e-05,
+      "loss": 1.8334,
+      "step": 7215
+    },
+    {
+      "epoch": 0.7495585332917836,
+      "grad_norm": 0.4359906017780304,
+      "learning_rate": 1.4693729550858976e-05,
+      "loss": 1.6999,
+      "step": 7216
+    },
+    {
+      "epoch": 0.7496624078113638,
+      "grad_norm": 0.4275515079498291,
+      "learning_rate": 1.4682177882279486e-05,
+      "loss": 1.7192,
+      "step": 7217
+    },
+    {
+      "epoch": 0.7497662823309442,
+      "grad_norm": 0.457688570022583,
+      "learning_rate": 1.4670629974772532e-05,
+      "loss": 1.725,
+      "step": 7218
+    },
+    {
+      "epoch": 0.7498701568505246,
+      "grad_norm": 0.3888382613658905,
+      "learning_rate": 1.465908582956792e-05,
+      "loss": 1.5319,
+      "step": 7219
+    },
+    {
+      "epoch": 0.7499740313701049,
+      "grad_norm": 0.396145761013031,
+      "learning_rate": 1.4647545447895017e-05,
+      "loss": 1.6354,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7500779058896853,
+      "grad_norm": 0.425191730260849,
+      "learning_rate": 1.4636008830982729e-05,
+      "loss": 1.8322,
+      "step": 7221
+    },
+    {
+      "epoch": 0.7501817804092656,
+      "grad_norm": 0.4128625988960266,
+      "learning_rate": 1.4624475980059665e-05,
+      "loss": 1.5982,
+      "step": 7222
+    },
+    {
+      "epoch": 0.7502856549288459,
+      "grad_norm": 0.41751012206077576,
+      "learning_rate": 1.4612946896353974e-05,
+      "loss": 1.5379,
+      "step": 7223
+    },
+    {
+      "epoch": 0.7503895294484263,
+      "grad_norm": 0.39982759952545166,
+      "learning_rate": 1.4601421581093406e-05,
+      "loss": 1.6689,
+      "step": 7224
+    },
+    {
+      "epoch": 0.7504934039680067,
+      "grad_norm": 0.4144216775894165,
+      "learning_rate": 1.4589900035505321e-05,
+      "loss": 1.6664,
+      "step": 7225
+    },
+    {
+      "epoch": 0.750597278487587,
+      "grad_norm": 0.4442715644836426,
+      "learning_rate": 1.4578382260816669e-05,
+      "loss": 1.81,
+      "step": 7226
+    },
+    {
+      "epoch": 0.7507011530071673,
+      "grad_norm": 0.4191363751888275,
+      "learning_rate": 1.4566868258254013e-05,
+      "loss": 1.5913,
+      "step": 7227
+    },
+    {
+      "epoch": 0.7508050275267477,
+      "grad_norm": 0.41727662086486816,
+      "learning_rate": 1.455535802904348e-05,
+      "loss": 1.6758,
+      "step": 7228
+    },
+    {
+      "epoch": 0.750908902046328,
+      "grad_norm": 0.4364617168903351,
+      "learning_rate": 1.4543851574410872e-05,
+      "loss": 1.6397,
+      "step": 7229
+    },
+    {
+      "epoch": 0.7510127765659084,
+      "grad_norm": 0.45298081636428833,
+      "learning_rate": 1.4532348895581465e-05,
+      "loss": 1.6516,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7511166510854888,
+      "grad_norm": 0.4370194375514984,
+      "learning_rate": 1.4520849993780266e-05,
+      "loss": 1.6518,
+      "step": 7231
+    },
+    {
+      "epoch": 0.751220525605069,
+      "grad_norm": 0.43286606669425964,
+      "learning_rate": 1.450935487023179e-05,
+      "loss": 1.5553,
+      "step": 7232
+    },
+    {
+      "epoch": 0.7513244001246494,
+      "grad_norm": 0.40669238567352295,
+      "learning_rate": 1.4497863526160183e-05,
+      "loss": 1.6917,
+      "step": 7233
+    },
+    {
+      "epoch": 0.7514282746442298,
+      "grad_norm": 0.42103543877601624,
+      "learning_rate": 1.4486375962789184e-05,
+      "loss": 1.4978,
+      "step": 7234
+    },
+    {
+      "epoch": 0.7515321491638102,
+      "grad_norm": 0.4966675043106079,
+      "learning_rate": 1.447489218134211e-05,
+      "loss": 1.7395,
+      "step": 7235
+    },
+    {
+      "epoch": 0.7516360236833904,
+      "grad_norm": 0.43593791127204895,
+      "learning_rate": 1.4463412183041947e-05,
+      "loss": 1.771,
+      "step": 7236
+    },
+    {
+      "epoch": 0.7517398982029708,
+      "grad_norm": 0.4331296980381012,
+      "learning_rate": 1.4451935969111158e-05,
+      "loss": 1.621,
+      "step": 7237
+    },
+    {
+      "epoch": 0.7518437727225512,
+      "grad_norm": 0.46651628613471985,
+      "learning_rate": 1.4440463540771914e-05,
+      "loss": 1.8537,
+      "step": 7238
+    },
+    {
+      "epoch": 0.7519476472421315,
+      "grad_norm": 0.40865036845207214,
+      "learning_rate": 1.4428994899245934e-05,
+      "loss": 1.6614,
+      "step": 7239
+    },
+    {
+      "epoch": 0.7520515217617119,
+      "grad_norm": 0.4490216374397278,
+      "learning_rate": 1.4417530045754529e-05,
+      "loss": 1.6751,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7521553962812922,
+      "grad_norm": 0.46918779611587524,
+      "learning_rate": 1.4406068981518623e-05,
+      "loss": 1.8038,
+      "step": 7241
+    },
+    {
+      "epoch": 0.7522592708008725,
+      "grad_norm": 0.40138277411460876,
+      "learning_rate": 1.4394611707758726e-05,
+      "loss": 1.6524,
+      "step": 7242
+    },
+    {
+      "epoch": 0.7523631453204529,
+      "grad_norm": 0.43885231018066406,
+      "learning_rate": 1.4383158225694948e-05,
+      "loss": 1.7706,
+      "step": 7243
+    },
+    {
+      "epoch": 0.7524670198400333,
+      "grad_norm": 0.42752382159233093,
+      "learning_rate": 1.4371708536546985e-05,
+      "loss": 1.5644,
+      "step": 7244
+    },
+    {
+      "epoch": 0.7525708943596136,
+      "grad_norm": 0.4024030864238739,
+      "learning_rate": 1.4360262641534179e-05,
+      "loss": 1.6406,
+      "step": 7245
+    },
+    {
+      "epoch": 0.7526747688791939,
+      "grad_norm": 0.40453994274139404,
+      "learning_rate": 1.4348820541875374e-05,
+      "loss": 1.6899,
+      "step": 7246
+    },
+    {
+      "epoch": 0.7527786433987743,
+      "grad_norm": 0.4157525300979614,
+      "learning_rate": 1.4337382238789105e-05,
+      "loss": 1.4451,
+      "step": 7247
+    },
+    {
+      "epoch": 0.7528825179183546,
+      "grad_norm": 0.42336198687553406,
+      "learning_rate": 1.432594773349346e-05,
+      "loss": 1.7071,
+      "step": 7248
+    },
+    {
+      "epoch": 0.752986392437935,
+      "grad_norm": 0.4072533845901489,
+      "learning_rate": 1.4314517027206086e-05,
+      "loss": 1.6278,
+      "step": 7249
+    },
+    {
+      "epoch": 0.7530902669575154,
+      "grad_norm": 0.47410857677459717,
+      "learning_rate": 1.4303090121144302e-05,
+      "loss": 1.6032,
+      "step": 7250
+    },
+    {
+      "epoch": 0.7531941414770956,
+      "grad_norm": 0.39942529797554016,
+      "learning_rate": 1.429166701652498e-05,
+      "loss": 1.5404,
+      "step": 7251
+    },
+    {
+      "epoch": 0.753298015996676,
+      "grad_norm": 0.44464361667633057,
+      "learning_rate": 1.4280247714564582e-05,
+      "loss": 1.73,
+      "step": 7252
+    },
+    {
+      "epoch": 0.7534018905162564,
+      "grad_norm": 0.45004576444625854,
+      "learning_rate": 1.4268832216479177e-05,
+      "loss": 1.7523,
+      "step": 7253
+    },
+    {
+      "epoch": 0.7535057650358367,
+      "grad_norm": 0.4188900291919708,
+      "learning_rate": 1.4257420523484427e-05,
+      "loss": 1.605,
+      "step": 7254
+    },
+    {
+      "epoch": 0.753609639555417,
+      "grad_norm": 0.39535877108573914,
+      "learning_rate": 1.4246012636795592e-05,
+      "loss": 1.5916,
+      "step": 7255
+    },
+    {
+      "epoch": 0.7537135140749974,
+      "grad_norm": 0.44383829832077026,
+      "learning_rate": 1.42346085576275e-05,
+      "loss": 1.7938,
+      "step": 7256
+    },
+    {
+      "epoch": 0.7538173885945777,
+      "grad_norm": 0.41384685039520264,
+      "learning_rate": 1.422320828719465e-05,
+      "loss": 1.5843,
+      "step": 7257
+    },
+    {
+      "epoch": 0.7539212631141581,
+      "grad_norm": 0.41272881627082825,
+      "learning_rate": 1.4211811826711014e-05,
+      "loss": 1.5989,
+      "step": 7258
+    },
+    {
+      "epoch": 0.7540251376337385,
+      "grad_norm": 0.4295910596847534,
+      "learning_rate": 1.4200419177390279e-05,
+      "loss": 1.6587,
+      "step": 7259
+    },
+    {
+      "epoch": 0.7541290121533188,
+      "grad_norm": 0.38193991780281067,
+      "learning_rate": 1.4189030340445646e-05,
+      "loss": 1.5246,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7542328866728991,
+      "grad_norm": 0.4172359108924866,
+      "learning_rate": 1.4177645317089954e-05,
+      "loss": 1.6807,
+      "step": 7261
+    },
+    {
+      "epoch": 0.7543367611924795,
+      "grad_norm": 0.4366094172000885,
+      "learning_rate": 1.416626410853561e-05,
+      "loss": 1.6379,
+      "step": 7262
+    },
+    {
+      "epoch": 0.7544406357120599,
+      "grad_norm": 0.4305514991283417,
+      "learning_rate": 1.4154886715994603e-05,
+      "loss": 1.6603,
+      "step": 7263
+    },
+    {
+      "epoch": 0.7545445102316402,
+      "grad_norm": 0.4556597173213959,
+      "learning_rate": 1.4143513140678595e-05,
+      "loss": 1.8564,
+      "step": 7264
+    },
+    {
+      "epoch": 0.7546483847512205,
+      "grad_norm": 0.49126148223876953,
+      "learning_rate": 1.4132143383798713e-05,
+      "loss": 1.7624,
+      "step": 7265
+    },
+    {
+      "epoch": 0.7547522592708009,
+      "grad_norm": 0.46818116307258606,
+      "learning_rate": 1.4120777446565791e-05,
+      "loss": 1.6411,
+      "step": 7266
+    },
+    {
+      "epoch": 0.7548561337903812,
+      "grad_norm": 0.4020385444164276,
+      "learning_rate": 1.4109415330190213e-05,
+      "loss": 1.6907,
+      "step": 7267
+    },
+    {
+      "epoch": 0.7549600083099616,
+      "grad_norm": 0.4432200491428375,
+      "learning_rate": 1.4098057035881934e-05,
+      "loss": 1.6414,
+      "step": 7268
+    },
+    {
+      "epoch": 0.755063882829542,
+      "grad_norm": 0.43888726830482483,
+      "learning_rate": 1.408670256485054e-05,
+      "loss": 1.5862,
+      "step": 7269
+    },
+    {
+      "epoch": 0.7551677573491222,
+      "grad_norm": 0.45090728998184204,
+      "learning_rate": 1.4075351918305191e-05,
+      "loss": 1.7012,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7552716318687026,
+      "grad_norm": 0.42840370535850525,
+      "learning_rate": 1.4064005097454635e-05,
+      "loss": 1.7653,
+      "step": 7271
+    },
+    {
+      "epoch": 0.755375506388283,
+      "grad_norm": 0.4270748496055603,
+      "learning_rate": 1.4052662103507214e-05,
+      "loss": 1.6584,
+      "step": 7272
+    },
+    {
+      "epoch": 0.7554793809078633,
+      "grad_norm": 0.4157198369503021,
+      "learning_rate": 1.4041322937670903e-05,
+      "loss": 1.7567,
+      "step": 7273
+    },
+    {
+      "epoch": 0.7555832554274436,
+      "grad_norm": 0.4310022294521332,
+      "learning_rate": 1.4029987601153188e-05,
+      "loss": 1.6985,
+      "step": 7274
+    },
+    {
+      "epoch": 0.755687129947024,
+      "grad_norm": 0.41847115755081177,
+      "learning_rate": 1.4018656095161226e-05,
+      "loss": 1.6191,
+      "step": 7275
+    },
+    {
+      "epoch": 0.7557910044666043,
+      "grad_norm": 0.4358253479003906,
+      "learning_rate": 1.400732842090174e-05,
+      "loss": 1.6991,
+      "step": 7276
+    },
+    {
+      "epoch": 0.7558948789861847,
+      "grad_norm": 0.4234682321548462,
+      "learning_rate": 1.3996004579580995e-05,
+      "loss": 1.8434,
+      "step": 7277
+    },
+    {
+      "epoch": 0.7559987535057651,
+      "grad_norm": 0.5005912780761719,
+      "learning_rate": 1.3984684572404938e-05,
+      "loss": 1.7706,
+      "step": 7278
+    },
+    {
+      "epoch": 0.7561026280253453,
+      "grad_norm": 0.45514440536499023,
+      "learning_rate": 1.397336840057904e-05,
+      "loss": 1.6742,
+      "step": 7279
+    },
+    {
+      "epoch": 0.7562065025449257,
+      "grad_norm": 0.4162159562110901,
+      "learning_rate": 1.3962056065308388e-05,
+      "loss": 1.6319,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7563103770645061,
+      "grad_norm": 0.41975322365760803,
+      "learning_rate": 1.3950747567797662e-05,
+      "loss": 1.6619,
+      "step": 7281
+    },
+    {
+      "epoch": 0.7564142515840864,
+      "grad_norm": 0.3796076476573944,
+      "learning_rate": 1.3939442909251121e-05,
+      "loss": 1.4829,
+      "step": 7282
+    },
+    {
+      "epoch": 0.7565181261036668,
+      "grad_norm": 0.4343501925468445,
+      "learning_rate": 1.3928142090872632e-05,
+      "loss": 1.6641,
+      "step": 7283
+    },
+    {
+      "epoch": 0.7566220006232471,
+      "grad_norm": 0.48821020126342773,
+      "learning_rate": 1.3916845113865618e-05,
+      "loss": 2.0066,
+      "step": 7284
+    },
+    {
+      "epoch": 0.7567258751428275,
+      "grad_norm": 0.4088677763938904,
+      "learning_rate": 1.3905551979433157e-05,
+      "loss": 1.6202,
+      "step": 7285
+    },
+    {
+      "epoch": 0.7568297496624078,
+      "grad_norm": 0.4253714680671692,
+      "learning_rate": 1.3894262688777865e-05,
+      "loss": 1.595,
+      "step": 7286
+    },
+    {
+      "epoch": 0.7569336241819882,
+      "grad_norm": 0.4234124422073364,
+      "learning_rate": 1.3882977243101957e-05,
+      "loss": 1.6401,
+      "step": 7287
+    },
+    {
+      "epoch": 0.7570374987015686,
+      "grad_norm": 0.4122594892978668,
+      "learning_rate": 1.3871695643607252e-05,
+      "loss": 1.415,
+      "step": 7288
+    },
+    {
+      "epoch": 0.7571413732211488,
+      "grad_norm": 0.4564708173274994,
+      "learning_rate": 1.3860417891495154e-05,
+      "loss": 1.7223,
+      "step": 7289
+    },
+    {
+      "epoch": 0.7572452477407292,
+      "grad_norm": 0.4112136662006378,
+      "learning_rate": 1.3849143987966645e-05,
+      "loss": 1.558,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7573491222603096,
+      "grad_norm": 0.4101704955101013,
+      "learning_rate": 1.38378739342223e-05,
+      "loss": 1.5861,
+      "step": 7291
+    },
+    {
+      "epoch": 0.7574529967798899,
+      "grad_norm": 0.4257015287876129,
+      "learning_rate": 1.3826607731462332e-05,
+      "loss": 1.5489,
+      "step": 7292
+    },
+    {
+      "epoch": 0.7575568712994702,
+      "grad_norm": 0.41998380422592163,
+      "learning_rate": 1.3815345380886452e-05,
+      "loss": 1.5094,
+      "step": 7293
+    },
+    {
+      "epoch": 0.7576607458190506,
+      "grad_norm": 0.3991340398788452,
+      "learning_rate": 1.380408688369405e-05,
+      "loss": 1.5909,
+      "step": 7294
+    },
+    {
+      "epoch": 0.7577646203386309,
+      "grad_norm": 0.42524048686027527,
+      "learning_rate": 1.3792832241084058e-05,
+      "loss": 1.6435,
+      "step": 7295
+    },
+    {
+      "epoch": 0.7578684948582113,
+      "grad_norm": 0.4474044740200043,
+      "learning_rate": 1.3781581454255005e-05,
+      "loss": 1.6698,
+      "step": 7296
+    },
+    {
+      "epoch": 0.7579723693777917,
+      "grad_norm": 0.41872158646583557,
+      "learning_rate": 1.3770334524405015e-05,
+      "loss": 1.6468,
+      "step": 7297
+    },
+    {
+      "epoch": 0.7580762438973719,
+      "grad_norm": 0.44773685932159424,
+      "learning_rate": 1.3759091452731792e-05,
+      "loss": 1.7338,
+      "step": 7298
+    },
+    {
+      "epoch": 0.7581801184169523,
+      "grad_norm": 0.498161643743515,
+      "learning_rate": 1.3747852240432641e-05,
+      "loss": 1.6291,
+      "step": 7299
+    },
+    {
+      "epoch": 0.7582839929365327,
+      "grad_norm": 0.4790482223033905,
+      "learning_rate": 1.3736616888704434e-05,
+      "loss": 1.8468,
+      "step": 7300
+    },
+    {
+      "epoch": 0.758387867456113,
+      "grad_norm": 0.4306541383266449,
+      "learning_rate": 1.372538539874369e-05,
+      "loss": 1.6639,
+      "step": 7301
+    },
+    {
+      "epoch": 0.7584917419756934,
+      "grad_norm": 0.4123995304107666,
+      "learning_rate": 1.3714157771746422e-05,
+      "loss": 1.5665,
+      "step": 7302
+    },
+    {
+      "epoch": 0.7585956164952737,
+      "grad_norm": 0.45254337787628174,
+      "learning_rate": 1.3702934008908325e-05,
+      "loss": 1.8113,
+      "step": 7303
+    },
+    {
+      "epoch": 0.758699491014854,
+      "grad_norm": 0.47395646572113037,
+      "learning_rate": 1.3691714111424641e-05,
+      "loss": 1.7394,
+      "step": 7304
+    },
+    {
+      "epoch": 0.7588033655344344,
+      "grad_norm": 0.43641114234924316,
+      "learning_rate": 1.3680498080490161e-05,
+      "loss": 1.6745,
+      "step": 7305
+    },
+    {
+      "epoch": 0.7589072400540148,
+      "grad_norm": 0.4418218433856964,
+      "learning_rate": 1.3669285917299347e-05,
+      "loss": 1.6433,
+      "step": 7306
+    },
+    {
+      "epoch": 0.759011114573595,
+      "grad_norm": 0.3992539048194885,
+      "learning_rate": 1.3658077623046195e-05,
+      "loss": 1.5712,
+      "step": 7307
+    },
+    {
+      "epoch": 0.7591149890931754,
+      "grad_norm": 0.43207311630249023,
+      "learning_rate": 1.3646873198924293e-05,
+      "loss": 1.6891,
+      "step": 7308
+    },
+    {
+      "epoch": 0.7592188636127558,
+      "grad_norm": 0.43646010756492615,
+      "learning_rate": 1.3635672646126835e-05,
+      "loss": 1.7747,
+      "step": 7309
+    },
+    {
+      "epoch": 0.7593227381323361,
+      "grad_norm": 0.4463355839252472,
+      "learning_rate": 1.3624475965846584e-05,
+      "loss": 1.5922,
+      "step": 7310
+    },
+    {
+      "epoch": 0.7594266126519165,
+      "grad_norm": 0.41628172993659973,
+      "learning_rate": 1.3613283159275902e-05,
+      "loss": 1.8192,
+      "step": 7311
+    },
+    {
+      "epoch": 0.7595304871714968,
+      "grad_norm": 0.3946208953857422,
+      "learning_rate": 1.3602094227606721e-05,
+      "loss": 1.5317,
+      "step": 7312
+    },
+    {
+      "epoch": 0.7596343616910772,
+      "grad_norm": 0.41928601264953613,
+      "learning_rate": 1.3590909172030608e-05,
+      "loss": 1.7496,
+      "step": 7313
+    },
+    {
+      "epoch": 0.7597382362106575,
+      "grad_norm": 0.4171970784664154,
+      "learning_rate": 1.3579727993738666e-05,
+      "loss": 1.7095,
+      "step": 7314
+    },
+    {
+      "epoch": 0.7598421107302379,
+      "grad_norm": 0.4714401960372925,
+      "learning_rate": 1.3568550693921605e-05,
+      "loss": 1.8207,
+      "step": 7315
+    },
+    {
+      "epoch": 0.7599459852498183,
+      "grad_norm": 0.4133859872817993,
+      "learning_rate": 1.3557377273769718e-05,
+      "loss": 1.7672,
+      "step": 7316
+    },
+    {
+      "epoch": 0.7600498597693985,
+      "grad_norm": 0.40099528431892395,
+      "learning_rate": 1.3546207734472882e-05,
+      "loss": 1.565,
+      "step": 7317
+    },
+    {
+      "epoch": 0.7601537342889789,
+      "grad_norm": 0.4770098030567169,
+      "learning_rate": 1.3535042077220578e-05,
+      "loss": 1.7241,
+      "step": 7318
+    },
+    {
+      "epoch": 0.7602576088085593,
+      "grad_norm": 0.43758663535118103,
+      "learning_rate": 1.3523880303201842e-05,
+      "loss": 1.6718,
+      "step": 7319
+    },
+    {
+      "epoch": 0.7603614833281396,
+      "grad_norm": 0.4401274621486664,
+      "learning_rate": 1.3512722413605355e-05,
+      "loss": 1.7467,
+      "step": 7320
+    },
+    {
+      "epoch": 0.76046535784772,
+      "grad_norm": 0.4016100764274597,
+      "learning_rate": 1.3501568409619286e-05,
+      "loss": 1.6112,
+      "step": 7321
+    },
+    {
+      "epoch": 0.7605692323673003,
+      "grad_norm": 0.44667908549308777,
+      "learning_rate": 1.3490418292431506e-05,
+      "loss": 1.7561,
+      "step": 7322
+    },
+    {
+      "epoch": 0.7606731068868806,
+      "grad_norm": 0.3975602388381958,
+      "learning_rate": 1.3479272063229387e-05,
+      "loss": 1.7218,
+      "step": 7323
+    },
+    {
+      "epoch": 0.760776981406461,
+      "grad_norm": 0.43723657727241516,
+      "learning_rate": 1.3468129723199919e-05,
+      "loss": 1.777,
+      "step": 7324
+    },
+    {
+      "epoch": 0.7608808559260414,
+      "grad_norm": 0.46576887369155884,
+      "learning_rate": 1.3456991273529679e-05,
+      "loss": 1.7544,
+      "step": 7325
+    },
+    {
+      "epoch": 0.7609847304456216,
+      "grad_norm": 0.4114767014980316,
+      "learning_rate": 1.344585671540482e-05,
+      "loss": 1.6227,
+      "step": 7326
+    },
+    {
+      "epoch": 0.761088604965202,
+      "grad_norm": 0.4945017695426941,
+      "learning_rate": 1.3434726050011093e-05,
+      "loss": 1.6841,
+      "step": 7327
+    },
+    {
+      "epoch": 0.7611924794847824,
+      "grad_norm": 0.4236311614513397,
+      "learning_rate": 1.34235992785338e-05,
+      "loss": 1.6404,
+      "step": 7328
+    },
+    {
+      "epoch": 0.7612963540043627,
+      "grad_norm": 0.4564515948295593,
+      "learning_rate": 1.3412476402157908e-05,
+      "loss": 1.7414,
+      "step": 7329
+    },
+    {
+      "epoch": 0.7614002285239431,
+      "grad_norm": 0.41757771372795105,
+      "learning_rate": 1.3401357422067857e-05,
+      "loss": 1.7274,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7615041030435235,
+      "grad_norm": 0.4927060008049011,
+      "learning_rate": 1.3390242339447779e-05,
+      "loss": 1.7611,
+      "step": 7331
+    },
+    {
+      "epoch": 0.7616079775631037,
+      "grad_norm": 0.4198152720928192,
+      "learning_rate": 1.3379131155481334e-05,
+      "loss": 1.7494,
+      "step": 7332
+    },
+    {
+      "epoch": 0.7617118520826841,
+      "grad_norm": 0.38799822330474854,
+      "learning_rate": 1.3368023871351737e-05,
+      "loss": 1.6216,
+      "step": 7333
+    },
+    {
+      "epoch": 0.7618157266022645,
+      "grad_norm": 0.4070364236831665,
+      "learning_rate": 1.3356920488241875e-05,
+      "loss": 1.5723,
+      "step": 7334
+    },
+    {
+      "epoch": 0.7619196011218448,
+      "grad_norm": 0.41575080156326294,
+      "learning_rate": 1.3345821007334153e-05,
+      "loss": 1.6863,
+      "step": 7335
+    },
+    {
+      "epoch": 0.7620234756414251,
+      "grad_norm": 0.43388399481773376,
+      "learning_rate": 1.333472542981058e-05,
+      "loss": 1.7284,
+      "step": 7336
+    },
+    {
+      "epoch": 0.7621273501610055,
+      "grad_norm": 0.39538997411727905,
+      "learning_rate": 1.3323633756852749e-05,
+      "loss": 1.5037,
+      "step": 7337
+    },
+    {
+      "epoch": 0.7622312246805859,
+      "grad_norm": 0.39611417055130005,
+      "learning_rate": 1.3312545989641823e-05,
+      "loss": 1.5679,
+      "step": 7338
+    },
+    {
+      "epoch": 0.7623350992001662,
+      "grad_norm": 0.42863768339157104,
+      "learning_rate": 1.3301462129358605e-05,
+      "loss": 1.6756,
+      "step": 7339
+    },
+    {
+      "epoch": 0.7624389737197466,
+      "grad_norm": 0.44742724299430847,
+      "learning_rate": 1.329038217718338e-05,
+      "loss": 1.8475,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7625428482393269,
+      "grad_norm": 0.4362469017505646,
+      "learning_rate": 1.3279306134296122e-05,
+      "loss": 1.6682,
+      "step": 7341
+    },
+    {
+      "epoch": 0.7626467227589072,
+      "grad_norm": 0.41959047317504883,
+      "learning_rate": 1.3268234001876328e-05,
+      "loss": 1.6528,
+      "step": 7342
+    },
+    {
+      "epoch": 0.7627505972784876,
+      "grad_norm": 0.3941357135772705,
+      "learning_rate": 1.3257165781103093e-05,
+      "loss": 1.5253,
+      "step": 7343
+    },
+    {
+      "epoch": 0.762854471798068,
+      "grad_norm": 0.4178176522254944,
+      "learning_rate": 1.3246101473155098e-05,
+      "loss": 1.4581,
+      "step": 7344
+    },
+    {
+      "epoch": 0.7629583463176483,
+      "grad_norm": 0.42281249165534973,
+      "learning_rate": 1.3235041079210603e-05,
+      "loss": 1.5694,
+      "step": 7345
+    },
+    {
+      "epoch": 0.7630622208372286,
+      "grad_norm": 0.4281790554523468,
+      "learning_rate": 1.3223984600447459e-05,
+      "loss": 1.6775,
+      "step": 7346
+    },
+    {
+      "epoch": 0.763166095356809,
+      "grad_norm": 0.4060118496417999,
+      "learning_rate": 1.3212932038043075e-05,
+      "loss": 1.4226,
+      "step": 7347
+    },
+    {
+      "epoch": 0.7632699698763893,
+      "grad_norm": 0.41389042139053345,
+      "learning_rate": 1.3201883393174508e-05,
+      "loss": 1.6444,
+      "step": 7348
+    },
+    {
+      "epoch": 0.7633738443959697,
+      "grad_norm": 0.424570232629776,
+      "learning_rate": 1.3190838667018297e-05,
+      "loss": 1.4875,
+      "step": 7349
+    },
+    {
+      "epoch": 0.76347771891555,
+      "grad_norm": 0.39579057693481445,
+      "learning_rate": 1.3179797860750653e-05,
+      "loss": 1.7218,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7635815934351303,
+      "grad_norm": 0.4264102876186371,
+      "learning_rate": 1.3168760975547335e-05,
+      "loss": 1.7828,
+      "step": 7351
+    },
+    {
+      "epoch": 0.7636854679547107,
+      "grad_norm": 0.4306561350822449,
+      "learning_rate": 1.3157728012583675e-05,
+      "loss": 1.7086,
+      "step": 7352
+    },
+    {
+      "epoch": 0.7637893424742911,
+      "grad_norm": 0.43455207347869873,
+      "learning_rate": 1.3146698973034599e-05,
+      "loss": 1.72,
+      "step": 7353
+    },
+    {
+      "epoch": 0.7638932169938714,
+      "grad_norm": 0.4588955342769623,
+      "learning_rate": 1.3135673858074621e-05,
+      "loss": 1.7184,
+      "step": 7354
+    },
+    {
+      "epoch": 0.7639970915134517,
+      "grad_norm": 0.41876885294914246,
+      "learning_rate": 1.3124652668877819e-05,
+      "loss": 1.6402,
+      "step": 7355
+    },
+    {
+      "epoch": 0.7641009660330321,
+      "grad_norm": 0.4438285529613495,
+      "learning_rate": 1.3113635406617858e-05,
+      "loss": 1.7906,
+      "step": 7356
+    },
+    {
+      "epoch": 0.7642048405526124,
+      "grad_norm": 0.4253295063972473,
+      "learning_rate": 1.310262207246803e-05,
+      "loss": 1.6315,
+      "step": 7357
+    },
+    {
+      "epoch": 0.7643087150721928,
+      "grad_norm": 0.45891672372817993,
+      "learning_rate": 1.3091612667601111e-05,
+      "loss": 1.7554,
+      "step": 7358
+    },
+    {
+      "epoch": 0.7644125895917732,
+      "grad_norm": 0.4097495973110199,
+      "learning_rate": 1.3080607193189559e-05,
+      "loss": 1.5465,
+      "step": 7359
+    },
+    {
+      "epoch": 0.7645164641113534,
+      "grad_norm": 0.4238637685775757,
+      "learning_rate": 1.3069605650405375e-05,
+      "loss": 1.6579,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7646203386309338,
+      "grad_norm": 0.42440927028656006,
+      "learning_rate": 1.3058608040420095e-05,
+      "loss": 1.5157,
+      "step": 7361
+    },
+    {
+      "epoch": 0.7647242131505142,
+      "grad_norm": 0.42934998869895935,
+      "learning_rate": 1.3047614364404914e-05,
+      "loss": 1.6494,
+      "step": 7362
+    },
+    {
+      "epoch": 0.7648280876700946,
+      "grad_norm": 0.42825332283973694,
+      "learning_rate": 1.3036624623530552e-05,
+      "loss": 1.6405,
+      "step": 7363
+    },
+    {
+      "epoch": 0.7649319621896749,
+      "grad_norm": 0.41669198870658875,
+      "learning_rate": 1.3025638818967373e-05,
+      "loss": 1.5746,
+      "step": 7364
+    },
+    {
+      "epoch": 0.7650358367092552,
+      "grad_norm": 0.4732094705104828,
+      "learning_rate": 1.3014656951885229e-05,
+      "loss": 1.7997,
+      "step": 7365
+    },
+    {
+      "epoch": 0.7651397112288356,
+      "grad_norm": 0.38443320989608765,
+      "learning_rate": 1.3003679023453614e-05,
+      "loss": 1.52,
+      "step": 7366
+    },
+    {
+      "epoch": 0.7652435857484159,
+      "grad_norm": 0.3946983516216278,
+      "learning_rate": 1.2992705034841623e-05,
+      "loss": 1.707,
+      "step": 7367
+    },
+    {
+      "epoch": 0.7653474602679963,
+      "grad_norm": 0.4353131055831909,
+      "learning_rate": 1.2981734987217853e-05,
+      "loss": 1.788,
+      "step": 7368
+    },
+    {
+      "epoch": 0.7654513347875767,
+      "grad_norm": 0.41617780923843384,
+      "learning_rate": 1.2970768881750567e-05,
+      "loss": 1.6524,
+      "step": 7369
+    },
+    {
+      "epoch": 0.7655552093071569,
+      "grad_norm": 0.42190882563591003,
+      "learning_rate": 1.295980671960756e-05,
+      "loss": 1.6253,
+      "step": 7370
+    },
+    {
+      "epoch": 0.7656590838267373,
+      "grad_norm": 0.4275079369544983,
+      "learning_rate": 1.2948848501956207e-05,
+      "loss": 1.6816,
+      "step": 7371
+    },
+    {
+      "epoch": 0.7657629583463177,
+      "grad_norm": 0.4169149696826935,
+      "learning_rate": 1.2937894229963483e-05,
+      "loss": 1.5858,
+      "step": 7372
+    },
+    {
+      "epoch": 0.765866832865898,
+      "grad_norm": 0.4570522606372833,
+      "learning_rate": 1.2926943904795925e-05,
+      "loss": 1.7703,
+      "step": 7373
+    },
+    {
+      "epoch": 0.7659707073854783,
+      "grad_norm": 0.424277663230896,
+      "learning_rate": 1.291599752761966e-05,
+      "loss": 1.6223,
+      "step": 7374
+    },
+    {
+      "epoch": 0.7660745819050587,
+      "grad_norm": 0.41258883476257324,
+      "learning_rate": 1.2905055099600378e-05,
+      "loss": 1.6564,
+      "step": 7375
+    },
+    {
+      "epoch": 0.766178456424639,
+      "grad_norm": 0.434390127658844,
+      "learning_rate": 1.2894116621903402e-05,
+      "loss": 1.5993,
+      "step": 7376
+    },
+    {
+      "epoch": 0.7662823309442194,
+      "grad_norm": 0.44891485571861267,
+      "learning_rate": 1.2883182095693547e-05,
+      "loss": 1.5434,
+      "step": 7377
+    },
+    {
+      "epoch": 0.7663862054637998,
+      "grad_norm": 0.43780601024627686,
+      "learning_rate": 1.2872251522135286e-05,
+      "loss": 1.5887,
+      "step": 7378
+    },
+    {
+      "epoch": 0.76649007998338,
+      "grad_norm": 0.4061758518218994,
+      "learning_rate": 1.286132490239263e-05,
+      "loss": 1.7188,
+      "step": 7379
+    },
+    {
+      "epoch": 0.7665939545029604,
+      "grad_norm": 0.47357314825057983,
+      "learning_rate": 1.2850402237629183e-05,
+      "loss": 1.869,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7666978290225408,
+      "grad_norm": 0.41141971945762634,
+      "learning_rate": 1.2839483529008117e-05,
+      "loss": 1.7155,
+      "step": 7381
+    },
+    {
+      "epoch": 0.7668017035421211,
+      "grad_norm": 0.4543936252593994,
+      "learning_rate": 1.2828568777692191e-05,
+      "loss": 1.6919,
+      "step": 7382
+    },
+    {
+      "epoch": 0.7669055780617015,
+      "grad_norm": 0.4838084876537323,
+      "learning_rate": 1.2817657984843746e-05,
+      "loss": 1.7123,
+      "step": 7383
+    },
+    {
+      "epoch": 0.7670094525812818,
+      "grad_norm": 0.422134667634964,
+      "learning_rate": 1.280675115162467e-05,
+      "loss": 1.7498,
+      "step": 7384
+    },
+    {
+      "epoch": 0.7671133271008621,
+      "grad_norm": 0.44436314702033997,
+      "learning_rate": 1.2795848279196515e-05,
+      "loss": 1.6622,
+      "step": 7385
+    },
+    {
+      "epoch": 0.7672172016204425,
+      "grad_norm": 0.42208799719810486,
+      "learning_rate": 1.2784949368720295e-05,
+      "loss": 1.7493,
+      "step": 7386
+    },
+    {
+      "epoch": 0.7673210761400229,
+      "grad_norm": 0.47984281182289124,
+      "learning_rate": 1.2774054421356668e-05,
+      "loss": 1.767,
+      "step": 7387
+    },
+    {
+      "epoch": 0.7674249506596031,
+      "grad_norm": 0.4420330226421356,
+      "learning_rate": 1.2763163438265896e-05,
+      "loss": 1.7475,
+      "step": 7388
+    },
+    {
+      "epoch": 0.7675288251791835,
+      "grad_norm": 0.41784390807151794,
+      "learning_rate": 1.2752276420607735e-05,
+      "loss": 1.756,
+      "step": 7389
+    },
+    {
+      "epoch": 0.7676326996987639,
+      "grad_norm": 0.6316620707511902,
+      "learning_rate": 1.2741393369541604e-05,
+      "loss": 1.8209,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7677365742183443,
+      "grad_norm": 0.39647650718688965,
+      "learning_rate": 1.2730514286226435e-05,
+      "loss": 1.5466,
+      "step": 7391
+    },
+    {
+      "epoch": 0.7678404487379246,
+      "grad_norm": 0.41624629497528076,
+      "learning_rate": 1.2719639171820812e-05,
+      "loss": 1.6497,
+      "step": 7392
+    },
+    {
+      "epoch": 0.767944323257505,
+      "grad_norm": 0.4231652617454529,
+      "learning_rate": 1.2708768027482804e-05,
+      "loss": 1.3777,
+      "step": 7393
+    },
+    {
+      "epoch": 0.7680481977770853,
+      "grad_norm": 0.42754560708999634,
+      "learning_rate": 1.2697900854370103e-05,
+      "loss": 1.6038,
+      "step": 7394
+    },
+    {
+      "epoch": 0.7681520722966656,
+      "grad_norm": 0.4023604691028595,
+      "learning_rate": 1.2687037653640026e-05,
+      "loss": 1.6013,
+      "step": 7395
+    },
+    {
+      "epoch": 0.768255946816246,
+      "grad_norm": 0.4051744043827057,
+      "learning_rate": 1.2676178426449353e-05,
+      "loss": 1.6607,
+      "step": 7396
+    },
+    {
+      "epoch": 0.7683598213358264,
+      "grad_norm": 0.39897581934928894,
+      "learning_rate": 1.2665323173954557e-05,
+      "loss": 1.659,
+      "step": 7397
+    },
+    {
+      "epoch": 0.7684636958554066,
+      "grad_norm": 0.41026565432548523,
+      "learning_rate": 1.2654471897311627e-05,
+      "loss": 1.6306,
+      "step": 7398
+    },
+    {
+      "epoch": 0.768567570374987,
+      "grad_norm": 0.42027774453163147,
+      "learning_rate": 1.2643624597676123e-05,
+      "loss": 1.5395,
+      "step": 7399
+    },
+    {
+      "epoch": 0.7686714448945674,
+      "grad_norm": 0.3966914117336273,
+      "learning_rate": 1.2632781276203216e-05,
+      "loss": 1.6195,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7687753194141477,
+      "grad_norm": 0.40674200654029846,
+      "learning_rate": 1.2621941934047626e-05,
+      "loss": 1.5897,
+      "step": 7401
+    },
+    {
+      "epoch": 0.7688791939337281,
+      "grad_norm": 0.3980896770954132,
+      "learning_rate": 1.2611106572363663e-05,
+      "loss": 1.4635,
+      "step": 7402
+    },
+    {
+      "epoch": 0.7689830684533084,
+      "grad_norm": 0.43104100227355957,
+      "learning_rate": 1.2600275192305184e-05,
+      "loss": 1.6104,
+      "step": 7403
+    },
+    {
+      "epoch": 0.7690869429728887,
+      "grad_norm": 0.4664388597011566,
+      "learning_rate": 1.2589447795025699e-05,
+      "loss": 1.7406,
+      "step": 7404
+    },
+    {
+      "epoch": 0.7691908174924691,
+      "grad_norm": 0.4235634505748749,
+      "learning_rate": 1.2578624381678183e-05,
+      "loss": 1.7893,
+      "step": 7405
+    },
+    {
+      "epoch": 0.7692946920120495,
+      "grad_norm": 0.3953429162502289,
+      "learning_rate": 1.2567804953415286e-05,
+      "loss": 1.5755,
+      "step": 7406
+    },
+    {
+      "epoch": 0.7693985665316297,
+      "grad_norm": 0.45959627628326416,
+      "learning_rate": 1.2556989511389183e-05,
+      "loss": 1.6138,
+      "step": 7407
+    },
+    {
+      "epoch": 0.7695024410512101,
+      "grad_norm": 0.4262646436691284,
+      "learning_rate": 1.2546178056751623e-05,
+      "loss": 1.5875,
+      "step": 7408
+    },
+    {
+      "epoch": 0.7696063155707905,
+      "grad_norm": 0.4125349819660187,
+      "learning_rate": 1.2535370590653955e-05,
+      "loss": 1.5919,
+      "step": 7409
+    },
+    {
+      "epoch": 0.7697101900903708,
+      "grad_norm": 0.40705251693725586,
+      "learning_rate": 1.2524567114247082e-05,
+      "loss": 1.6622,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7698140646099512,
+      "grad_norm": 0.4188450872898102,
+      "learning_rate": 1.2513767628681489e-05,
+      "loss": 1.5001,
+      "step": 7411
+    },
+    {
+      "epoch": 0.7699179391295315,
+      "grad_norm": 0.430321604013443,
+      "learning_rate": 1.2502972135107227e-05,
+      "loss": 1.6887,
+      "step": 7412
+    },
+    {
+      "epoch": 0.7700218136491118,
+      "grad_norm": 0.4586038291454315,
+      "learning_rate": 1.2492180634673978e-05,
+      "loss": 1.6159,
+      "step": 7413
+    },
+    {
+      "epoch": 0.7701256881686922,
+      "grad_norm": 0.4472750425338745,
+      "learning_rate": 1.24813931285309e-05,
+      "loss": 1.6015,
+      "step": 7414
+    },
+    {
+      "epoch": 0.7702295626882726,
+      "grad_norm": 0.404680997133255,
+      "learning_rate": 1.2470609617826784e-05,
+      "loss": 1.3466,
+      "step": 7415
+    },
+    {
+      "epoch": 0.770333437207853,
+      "grad_norm": 0.44199302792549133,
+      "learning_rate": 1.2459830103710019e-05,
+      "loss": 1.7882,
+      "step": 7416
+    },
+    {
+      "epoch": 0.7704373117274332,
+      "grad_norm": 0.4036124050617218,
+      "learning_rate": 1.2449054587328519e-05,
+      "loss": 1.6455,
+      "step": 7417
+    },
+    {
+      "epoch": 0.7705411862470136,
+      "grad_norm": 0.43482506275177,
+      "learning_rate": 1.2438283069829799e-05,
+      "loss": 1.7585,
+      "step": 7418
+    },
+    {
+      "epoch": 0.770645060766594,
+      "grad_norm": 0.3882008194923401,
+      "learning_rate": 1.2427515552360924e-05,
+      "loss": 1.502,
+      "step": 7419
+    },
+    {
+      "epoch": 0.7707489352861743,
+      "grad_norm": 0.44999149441719055,
+      "learning_rate": 1.2416752036068596e-05,
+      "loss": 1.7382,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7708528098057547,
+      "grad_norm": 0.4068087935447693,
+      "learning_rate": 1.2405992522099002e-05,
+      "loss": 1.5567,
+      "step": 7421
+    },
+    {
+      "epoch": 0.770956684325335,
+      "grad_norm": 0.4564858675003052,
+      "learning_rate": 1.2395237011597949e-05,
+      "loss": 1.8138,
+      "step": 7422
+    },
+    {
+      "epoch": 0.7710605588449153,
+      "grad_norm": 0.4281824231147766,
+      "learning_rate": 1.2384485505710847e-05,
+      "loss": 1.6393,
+      "step": 7423
+    },
+    {
+      "epoch": 0.7711644333644957,
+      "grad_norm": 0.3930261731147766,
+      "learning_rate": 1.2373738005582603e-05,
+      "loss": 1.5276,
+      "step": 7424
+    },
+    {
+      "epoch": 0.7712683078840761,
+      "grad_norm": 0.4548739790916443,
+      "learning_rate": 1.2362994512357779e-05,
+      "loss": 1.7476,
+      "step": 7425
+    },
+    {
+      "epoch": 0.7713721824036563,
+      "grad_norm": 0.4831191897392273,
+      "learning_rate": 1.235225502718046e-05,
+      "loss": 1.8217,
+      "step": 7426
+    },
+    {
+      "epoch": 0.7714760569232367,
+      "grad_norm": 0.41000956296920776,
+      "learning_rate": 1.2341519551194314e-05,
+      "loss": 1.5592,
+      "step": 7427
+    },
+    {
+      "epoch": 0.7715799314428171,
+      "grad_norm": 0.46288132667541504,
+      "learning_rate": 1.2330788085542589e-05,
+      "loss": 1.7773,
+      "step": 7428
+    },
+    {
+      "epoch": 0.7716838059623974,
+      "grad_norm": 0.4005960524082184,
+      "learning_rate": 1.2320060631368103e-05,
+      "loss": 1.6483,
+      "step": 7429
+    },
+    {
+      "epoch": 0.7717876804819778,
+      "grad_norm": 0.4161857068538666,
+      "learning_rate": 1.2309337189813242e-05,
+      "loss": 1.5655,
+      "step": 7430
+    },
+    {
+      "epoch": 0.7718915550015581,
+      "grad_norm": 0.4241454601287842,
+      "learning_rate": 1.229861776201996e-05,
+      "loss": 1.5819,
+      "step": 7431
+    },
+    {
+      "epoch": 0.7719954295211384,
+      "grad_norm": 0.44512662291526794,
+      "learning_rate": 1.2287902349129832e-05,
+      "loss": 1.5958,
+      "step": 7432
+    },
+    {
+      "epoch": 0.7720993040407188,
+      "grad_norm": 0.42898502945899963,
+      "learning_rate": 1.2277190952283912e-05,
+      "loss": 1.6924,
+      "step": 7433
+    },
+    {
+      "epoch": 0.7722031785602992,
+      "grad_norm": 0.4218330681324005,
+      "learning_rate": 1.2266483572622916e-05,
+      "loss": 1.6906,
+      "step": 7434
+    },
+    {
+      "epoch": 0.7723070530798795,
+      "grad_norm": 0.4153681695461273,
+      "learning_rate": 1.2255780211287094e-05,
+      "loss": 1.718,
+      "step": 7435
+    },
+    {
+      "epoch": 0.7724109275994598,
+      "grad_norm": 0.443933367729187,
+      "learning_rate": 1.2245080869416258e-05,
+      "loss": 1.636,
+      "step": 7436
+    },
+    {
+      "epoch": 0.7725148021190402,
+      "grad_norm": 0.42244407534599304,
+      "learning_rate": 1.2234385548149813e-05,
+      "loss": 1.7592,
+      "step": 7437
+    },
+    {
+      "epoch": 0.7726186766386205,
+      "grad_norm": 0.4867939054965973,
+      "learning_rate": 1.2223694248626721e-05,
+      "loss": 1.7048,
+      "step": 7438
+    },
+    {
+      "epoch": 0.7727225511582009,
+      "grad_norm": 0.45962727069854736,
+      "learning_rate": 1.2213006971985524e-05,
+      "loss": 1.808,
+      "step": 7439
+    },
+    {
+      "epoch": 0.7728264256777813,
+      "grad_norm": 0.44696059823036194,
+      "learning_rate": 1.2202323719364323e-05,
+      "loss": 1.6744,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7729303001973616,
+      "grad_norm": 0.3926789462566376,
+      "learning_rate": 1.2191644491900822e-05,
+      "loss": 1.5021,
+      "step": 7441
+    },
+    {
+      "epoch": 0.7730341747169419,
+      "grad_norm": 0.4703446626663208,
+      "learning_rate": 1.2180969290732286e-05,
+      "loss": 1.7047,
+      "step": 7442
+    },
+    {
+      "epoch": 0.7731380492365223,
+      "grad_norm": 0.4337899386882782,
+      "learning_rate": 1.2170298116995493e-05,
+      "loss": 1.5239,
+      "step": 7443
+    },
+    {
+      "epoch": 0.7732419237561027,
+      "grad_norm": 0.39774975180625916,
+      "learning_rate": 1.2159630971826879e-05,
+      "loss": 1.4872,
+      "step": 7444
+    },
+    {
+      "epoch": 0.773345798275683,
+      "grad_norm": 0.47527754306793213,
+      "learning_rate": 1.2148967856362398e-05,
+      "loss": 1.813,
+      "step": 7445
+    },
+    {
+      "epoch": 0.7734496727952633,
+      "grad_norm": 0.3975711166858673,
+      "learning_rate": 1.2138308771737594e-05,
+      "loss": 1.5849,
+      "step": 7446
+    },
+    {
+      "epoch": 0.7735535473148437,
+      "grad_norm": 0.44653868675231934,
+      "learning_rate": 1.2127653719087572e-05,
+      "loss": 1.7053,
+      "step": 7447
+    },
+    {
+      "epoch": 0.773657421834424,
+      "grad_norm": 0.4855223298072815,
+      "learning_rate": 1.2117002699547015e-05,
+      "loss": 1.7733,
+      "step": 7448
+    },
+    {
+      "epoch": 0.7737612963540044,
+      "grad_norm": 0.464631050825119,
+      "learning_rate": 1.2106355714250172e-05,
+      "loss": 1.7352,
+      "step": 7449
+    },
+    {
+      "epoch": 0.7738651708735848,
+      "grad_norm": 0.44896233081817627,
+      "learning_rate": 1.2095712764330847e-05,
+      "loss": 1.9348,
+      "step": 7450
+    },
+    {
+      "epoch": 0.773969045393165,
+      "grad_norm": 0.4480723738670349,
+      "learning_rate": 1.2085073850922479e-05,
+      "loss": 1.659,
+      "step": 7451
+    },
+    {
+      "epoch": 0.7740729199127454,
+      "grad_norm": 0.469439297914505,
+      "learning_rate": 1.2074438975157964e-05,
+      "loss": 1.8332,
+      "step": 7452
+    },
+    {
+      "epoch": 0.7741767944323258,
+      "grad_norm": 0.4098644256591797,
+      "learning_rate": 1.2063808138169891e-05,
+      "loss": 1.3752,
+      "step": 7453
+    },
+    {
+      "epoch": 0.7742806689519061,
+      "grad_norm": 0.3945968449115753,
+      "learning_rate": 1.205318134109033e-05,
+      "loss": 1.632,
+      "step": 7454
+    },
+    {
+      "epoch": 0.7743845434714864,
+      "grad_norm": 0.38634219765663147,
+      "learning_rate": 1.2042558585050961e-05,
+      "loss": 1.5544,
+      "step": 7455
+    },
+    {
+      "epoch": 0.7744884179910668,
+      "grad_norm": 0.45720741152763367,
+      "learning_rate": 1.2031939871183023e-05,
+      "loss": 1.8892,
+      "step": 7456
+    },
+    {
+      "epoch": 0.7745922925106471,
+      "grad_norm": 0.44557642936706543,
+      "learning_rate": 1.2021325200617329e-05,
+      "loss": 1.6822,
+      "step": 7457
+    },
+    {
+      "epoch": 0.7746961670302275,
+      "grad_norm": 0.4418318569660187,
+      "learning_rate": 1.2010714574484255e-05,
+      "loss": 1.6621,
+      "step": 7458
+    },
+    {
+      "epoch": 0.7748000415498079,
+      "grad_norm": 0.4305969774723053,
+      "learning_rate": 1.2000107993913739e-05,
+      "loss": 1.5825,
+      "step": 7459
+    },
+    {
+      "epoch": 0.7749039160693881,
+      "grad_norm": 0.385640025138855,
+      "learning_rate": 1.1989505460035333e-05,
+      "loss": 1.573,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7750077905889685,
+      "grad_norm": 0.45117321610450745,
+      "learning_rate": 1.1978906973978083e-05,
+      "loss": 1.68,
+      "step": 7461
+    },
+    {
+      "epoch": 0.7751116651085489,
+      "grad_norm": 0.4692929983139038,
+      "learning_rate": 1.1968312536870668e-05,
+      "loss": 1.5407,
+      "step": 7462
+    },
+    {
+      "epoch": 0.7752155396281292,
+      "grad_norm": 0.38421300053596497,
+      "learning_rate": 1.1957722149841317e-05,
+      "loss": 1.4685,
+      "step": 7463
+    },
+    {
+      "epoch": 0.7753194141477096,
+      "grad_norm": 0.41444119811058044,
+      "learning_rate": 1.1947135814017812e-05,
+      "loss": 1.4937,
+      "step": 7464
+    },
+    {
+      "epoch": 0.7754232886672899,
+      "grad_norm": 0.4278319478034973,
+      "learning_rate": 1.1936553530527523e-05,
+      "loss": 1.5106,
+      "step": 7465
+    },
+    {
+      "epoch": 0.7755271631868702,
+      "grad_norm": 0.39915332198143005,
+      "learning_rate": 1.1925975300497371e-05,
+      "loss": 1.4574,
+      "step": 7466
+    },
+    {
+      "epoch": 0.7756310377064506,
+      "grad_norm": 0.44178900122642517,
+      "learning_rate": 1.1915401125053866e-05,
+      "loss": 1.749,
+      "step": 7467
+    },
+    {
+      "epoch": 0.775734912226031,
+      "grad_norm": 0.39515554904937744,
+      "learning_rate": 1.1904831005323053e-05,
+      "loss": 1.5465,
+      "step": 7468
+    },
+    {
+      "epoch": 0.7758387867456114,
+      "grad_norm": 0.3899015486240387,
+      "learning_rate": 1.18942649424306e-05,
+      "loss": 1.487,
+      "step": 7469
+    },
+    {
+      "epoch": 0.7759426612651916,
+      "grad_norm": 0.44565442204475403,
+      "learning_rate": 1.1883702937501705e-05,
+      "loss": 1.6149,
+      "step": 7470
+    },
+    {
+      "epoch": 0.776046535784772,
+      "grad_norm": 0.405678927898407,
+      "learning_rate": 1.1873144991661106e-05,
+      "loss": 1.567,
+      "step": 7471
+    },
+    {
+      "epoch": 0.7761504103043524,
+      "grad_norm": 0.46768710017204285,
+      "learning_rate": 1.1862591106033178e-05,
+      "loss": 1.7902,
+      "step": 7472
+    },
+    {
+      "epoch": 0.7762542848239327,
+      "grad_norm": 0.4093432128429413,
+      "learning_rate": 1.185204128174181e-05,
+      "loss": 1.6518,
+      "step": 7473
+    },
+    {
+      "epoch": 0.776358159343513,
+      "grad_norm": 0.4388825297355652,
+      "learning_rate": 1.1841495519910485e-05,
+      "loss": 1.6971,
+      "step": 7474
+    },
+    {
+      "epoch": 0.7764620338630934,
+      "grad_norm": 0.4254426956176758,
+      "learning_rate": 1.1830953821662232e-05,
+      "loss": 1.6166,
+      "step": 7475
+    },
+    {
+      "epoch": 0.7765659083826737,
+      "grad_norm": 0.42285576462745667,
+      "learning_rate": 1.182041618811967e-05,
+      "loss": 1.611,
+      "step": 7476
+    },
+    {
+      "epoch": 0.7766697829022541,
+      "grad_norm": 0.4236442446708679,
+      "learning_rate": 1.1809882620404972e-05,
+      "loss": 1.586,
+      "step": 7477
+    },
+    {
+      "epoch": 0.7767736574218345,
+      "grad_norm": 0.4397282600402832,
+      "learning_rate": 1.1799353119639867e-05,
+      "loss": 1.7642,
+      "step": 7478
+    },
+    {
+      "epoch": 0.7768775319414147,
+      "grad_norm": 0.41384944319725037,
+      "learning_rate": 1.1788827686945708e-05,
+      "loss": 1.6689,
+      "step": 7479
+    },
+    {
+      "epoch": 0.7769814064609951,
+      "grad_norm": 0.44808679819107056,
+      "learning_rate": 1.177830632344331e-05,
+      "loss": 1.8645,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7770852809805755,
+      "grad_norm": 0.3973611295223236,
+      "learning_rate": 1.1767789030253168e-05,
+      "loss": 1.6404,
+      "step": 7481
+    },
+    {
+      "epoch": 0.7771891555001558,
+      "grad_norm": 0.44642767310142517,
+      "learning_rate": 1.1757275808495266e-05,
+      "loss": 1.7545,
+      "step": 7482
+    },
+    {
+      "epoch": 0.7772930300197362,
+      "grad_norm": 0.4089753329753876,
+      "learning_rate": 1.1746766659289188e-05,
+      "loss": 1.6952,
+      "step": 7483
+    },
+    {
+      "epoch": 0.7773969045393165,
+      "grad_norm": 0.41882163286209106,
+      "learning_rate": 1.1736261583754072e-05,
+      "loss": 1.5784,
+      "step": 7484
+    },
+    {
+      "epoch": 0.7775007790588968,
+      "grad_norm": 0.489686518907547,
+      "learning_rate": 1.1725760583008633e-05,
+      "loss": 1.7849,
+      "step": 7485
+    },
+    {
+      "epoch": 0.7776046535784772,
+      "grad_norm": 0.41849491000175476,
+      "learning_rate": 1.171526365817114e-05,
+      "loss": 1.6088,
+      "step": 7486
+    },
+    {
+      "epoch": 0.7777085280980576,
+      "grad_norm": 0.40229809284210205,
+      "learning_rate": 1.1704770810359417e-05,
+      "loss": 1.5425,
+      "step": 7487
+    },
+    {
+      "epoch": 0.7778124026176378,
+      "grad_norm": 0.3939957916736603,
+      "learning_rate": 1.1694282040690919e-05,
+      "loss": 1.5142,
+      "step": 7488
+    },
+    {
+      "epoch": 0.7779162771372182,
+      "grad_norm": 0.46565282344818115,
+      "learning_rate": 1.168379735028256e-05,
+      "loss": 1.4844,
+      "step": 7489
+    },
+    {
+      "epoch": 0.7780201516567986,
+      "grad_norm": 0.4124033451080322,
+      "learning_rate": 1.1673316740250916e-05,
+      "loss": 1.647,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7781240261763789,
+      "grad_norm": 0.463223397731781,
+      "learning_rate": 1.1662840211712084e-05,
+      "loss": 1.6546,
+      "step": 7491
+    },
+    {
+      "epoch": 0.7782279006959593,
+      "grad_norm": 0.40807053446769714,
+      "learning_rate": 1.1652367765781724e-05,
+      "loss": 1.7317,
+      "step": 7492
+    },
+    {
+      "epoch": 0.7783317752155396,
+      "grad_norm": 0.4360339045524597,
+      "learning_rate": 1.164189940357508e-05,
+      "loss": 1.6996,
+      "step": 7493
+    },
+    {
+      "epoch": 0.77843564973512,
+      "grad_norm": 0.4180784821510315,
+      "learning_rate": 1.1631435126206924e-05,
+      "loss": 1.5694,
+      "step": 7494
+    },
+    {
+      "epoch": 0.7785395242547003,
+      "grad_norm": 0.4407843053340912,
+      "learning_rate": 1.1620974934791673e-05,
+      "loss": 1.5474,
+      "step": 7495
+    },
+    {
+      "epoch": 0.7786433987742807,
+      "grad_norm": 0.36754143238067627,
+      "learning_rate": 1.1610518830443195e-05,
+      "loss": 1.3948,
+      "step": 7496
+    },
+    {
+      "epoch": 0.7787472732938611,
+      "grad_norm": 0.4531268775463104,
+      "learning_rate": 1.1600066814275028e-05,
+      "loss": 1.7093,
+      "step": 7497
+    },
+    {
+      "epoch": 0.7788511478134413,
+      "grad_norm": 0.4329068958759308,
+      "learning_rate": 1.1589618887400223e-05,
+      "loss": 1.6267,
+      "step": 7498
+    },
+    {
+      "epoch": 0.7789550223330217,
+      "grad_norm": 0.385638564825058,
+      "learning_rate": 1.1579175050931369e-05,
+      "loss": 1.4813,
+      "step": 7499
+    },
+    {
+      "epoch": 0.7790588968526021,
+      "grad_norm": 0.4533402621746063,
+      "learning_rate": 1.1568735305980693e-05,
+      "loss": 1.6934,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7791627713721824,
+      "grad_norm": 0.41636162996292114,
+      "learning_rate": 1.1558299653659931e-05,
+      "loss": 1.7808,
+      "step": 7501
+    },
+    {
+      "epoch": 0.7792666458917628,
+      "grad_norm": 0.44357746839523315,
+      "learning_rate": 1.15478680950804e-05,
+      "loss": 1.5262,
+      "step": 7502
+    },
+    {
+      "epoch": 0.7793705204113431,
+      "grad_norm": 0.43575209379196167,
+      "learning_rate": 1.1537440631352975e-05,
+      "loss": 1.6568,
+      "step": 7503
+    },
+    {
+      "epoch": 0.7794743949309234,
+      "grad_norm": 0.4117165803909302,
+      "learning_rate": 1.1527017263588108e-05,
+      "loss": 1.5999,
+      "step": 7504
+    },
+    {
+      "epoch": 0.7795782694505038,
+      "grad_norm": 0.40848150849342346,
+      "learning_rate": 1.15165979928958e-05,
+      "loss": 1.4227,
+      "step": 7505
+    },
+    {
+      "epoch": 0.7796821439700842,
+      "grad_norm": 0.42398545145988464,
+      "learning_rate": 1.1506182820385608e-05,
+      "loss": 1.6902,
+      "step": 7506
+    },
+    {
+      "epoch": 0.7797860184896644,
+      "grad_norm": 0.4009295105934143,
+      "learning_rate": 1.149577174716671e-05,
+      "loss": 1.543,
+      "step": 7507
+    },
+    {
+      "epoch": 0.7798898930092448,
+      "grad_norm": 0.48386311531066895,
+      "learning_rate": 1.148536477434775e-05,
+      "loss": 1.6154,
+      "step": 7508
+    },
+    {
+      "epoch": 0.7799937675288252,
+      "grad_norm": 0.43705689907073975,
+      "learning_rate": 1.1474961903037029e-05,
+      "loss": 1.6337,
+      "step": 7509
+    },
+    {
+      "epoch": 0.7800976420484055,
+      "grad_norm": 0.3995164632797241,
+      "learning_rate": 1.146456313434236e-05,
+      "loss": 1.6047,
+      "step": 7510
+    },
+    {
+      "epoch": 0.7802015165679859,
+      "grad_norm": 0.40124985575675964,
+      "learning_rate": 1.145416846937113e-05,
+      "loss": 1.5502,
+      "step": 7511
+    },
+    {
+      "epoch": 0.7803053910875662,
+      "grad_norm": 0.4591343104839325,
+      "learning_rate": 1.1443777909230286e-05,
+      "loss": 1.8176,
+      "step": 7512
+    },
+    {
+      "epoch": 0.7804092656071465,
+      "grad_norm": 0.4121926724910736,
+      "learning_rate": 1.1433391455026343e-05,
+      "loss": 1.6699,
+      "step": 7513
+    },
+    {
+      "epoch": 0.7805131401267269,
+      "grad_norm": 0.4800514876842499,
+      "learning_rate": 1.1423009107865384e-05,
+      "loss": 1.9029,
+      "step": 7514
+    },
+    {
+      "epoch": 0.7806170146463073,
+      "grad_norm": 0.44299235939979553,
+      "learning_rate": 1.1412630868853024e-05,
+      "loss": 1.6599,
+      "step": 7515
+    },
+    {
+      "epoch": 0.7807208891658876,
+      "grad_norm": 0.5640523433685303,
+      "learning_rate": 1.1402256739094513e-05,
+      "loss": 1.9605,
+      "step": 7516
+    },
+    {
+      "epoch": 0.7808247636854679,
+      "grad_norm": 0.4083343744277954,
+      "learning_rate": 1.1391886719694561e-05,
+      "loss": 1.6042,
+      "step": 7517
+    },
+    {
+      "epoch": 0.7809286382050483,
+      "grad_norm": 0.4362529516220093,
+      "learning_rate": 1.1381520811757529e-05,
+      "loss": 1.6724,
+      "step": 7518
+    },
+    {
+      "epoch": 0.7810325127246287,
+      "grad_norm": 0.3871046006679535,
+      "learning_rate": 1.1371159016387296e-05,
+      "loss": 1.55,
+      "step": 7519
+    },
+    {
+      "epoch": 0.781136387244209,
+      "grad_norm": 0.41727960109710693,
+      "learning_rate": 1.136080133468731e-05,
+      "loss": 1.6557,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7812402617637894,
+      "grad_norm": 0.41626328229904175,
+      "learning_rate": 1.1350447767760586e-05,
+      "loss": 1.642,
+      "step": 7521
+    },
+    {
+      "epoch": 0.7813441362833697,
+      "grad_norm": 0.41060563921928406,
+      "learning_rate": 1.134009831670968e-05,
+      "loss": 1.6731,
+      "step": 7522
+    },
+    {
+      "epoch": 0.78144801080295,
+      "grad_norm": 0.407531201839447,
+      "learning_rate": 1.1329752982636776e-05,
+      "loss": 1.6211,
+      "step": 7523
+    },
+    {
+      "epoch": 0.7815518853225304,
+      "grad_norm": 0.38882845640182495,
+      "learning_rate": 1.1319411766643512e-05,
+      "loss": 1.5965,
+      "step": 7524
+    },
+    {
+      "epoch": 0.7816557598421108,
+      "grad_norm": 0.4430273771286011,
+      "learning_rate": 1.1309074669831189e-05,
+      "loss": 1.7404,
+      "step": 7525
+    },
+    {
+      "epoch": 0.781759634361691,
+      "grad_norm": 0.4414060115814209,
+      "learning_rate": 1.1298741693300624e-05,
+      "loss": 1.6904,
+      "step": 7526
+    },
+    {
+      "epoch": 0.7818635088812714,
+      "grad_norm": 0.4308008849620819,
+      "learning_rate": 1.1288412838152157e-05,
+      "loss": 1.8278,
+      "step": 7527
+    },
+    {
+      "epoch": 0.7819673834008518,
+      "grad_norm": 0.4290758967399597,
+      "learning_rate": 1.1278088105485778e-05,
+      "loss": 1.714,
+      "step": 7528
+    },
+    {
+      "epoch": 0.7820712579204321,
+      "grad_norm": 0.4313497841358185,
+      "learning_rate": 1.1267767496400972e-05,
+      "loss": 1.6401,
+      "step": 7529
+    },
+    {
+      "epoch": 0.7821751324400125,
+      "grad_norm": 0.4212192893028259,
+      "learning_rate": 1.1257451011996805e-05,
+      "loss": 1.6266,
+      "step": 7530
+    },
+    {
+      "epoch": 0.7822790069595928,
+      "grad_norm": 0.44428661465644836,
+      "learning_rate": 1.1247138653371903e-05,
+      "loss": 1.6743,
+      "step": 7531
+    },
+    {
+      "epoch": 0.7823828814791731,
+      "grad_norm": 0.44234758615493774,
+      "learning_rate": 1.123683042162445e-05,
+      "loss": 1.6527,
+      "step": 7532
+    },
+    {
+      "epoch": 0.7824867559987535,
+      "grad_norm": 0.4521244168281555,
+      "learning_rate": 1.1226526317852193e-05,
+      "loss": 1.8607,
+      "step": 7533
+    },
+    {
+      "epoch": 0.7825906305183339,
+      "grad_norm": 0.4284451901912689,
+      "learning_rate": 1.1216226343152424e-05,
+      "loss": 1.5447,
+      "step": 7534
+    },
+    {
+      "epoch": 0.7826945050379142,
+      "grad_norm": 0.39966773986816406,
+      "learning_rate": 1.1205930498622053e-05,
+      "loss": 1.6129,
+      "step": 7535
+    },
+    {
+      "epoch": 0.7827983795574945,
+      "grad_norm": 0.4256837069988251,
+      "learning_rate": 1.1195638785357448e-05,
+      "loss": 1.5621,
+      "step": 7536
+    },
+    {
+      "epoch": 0.7829022540770749,
+      "grad_norm": 0.4508436620235443,
+      "learning_rate": 1.1185351204454647e-05,
+      "loss": 1.7632,
+      "step": 7537
+    },
+    {
+      "epoch": 0.7830061285966552,
+      "grad_norm": 0.40548187494277954,
+      "learning_rate": 1.1175067757009171e-05,
+      "loss": 1.5052,
+      "step": 7538
+    },
+    {
+      "epoch": 0.7831100031162356,
+      "grad_norm": 0.43622684478759766,
+      "learning_rate": 1.1164788444116131e-05,
+      "loss": 1.7971,
+      "step": 7539
+    },
+    {
+      "epoch": 0.783213877635816,
+      "grad_norm": 0.4189201891422272,
+      "learning_rate": 1.1154513266870204e-05,
+      "loss": 1.5101,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7833177521553962,
+      "grad_norm": 0.4424949288368225,
+      "learning_rate": 1.1144242226365599e-05,
+      "loss": 1.8468,
+      "step": 7541
+    },
+    {
+      "epoch": 0.7834216266749766,
+      "grad_norm": 0.4143393635749817,
+      "learning_rate": 1.1133975323696111e-05,
+      "loss": 1.6425,
+      "step": 7542
+    },
+    {
+      "epoch": 0.783525501194557,
+      "grad_norm": 0.43571484088897705,
+      "learning_rate": 1.1123712559955075e-05,
+      "loss": 1.5772,
+      "step": 7543
+    },
+    {
+      "epoch": 0.7836293757141374,
+      "grad_norm": 0.42268067598342896,
+      "learning_rate": 1.1113453936235418e-05,
+      "loss": 1.632,
+      "step": 7544
+    },
+    {
+      "epoch": 0.7837332502337176,
+      "grad_norm": 0.4198543131351471,
+      "learning_rate": 1.1103199453629587e-05,
+      "loss": 1.6937,
+      "step": 7545
+    },
+    {
+      "epoch": 0.783837124753298,
+      "grad_norm": 0.4531916677951813,
+      "learning_rate": 1.1092949113229612e-05,
+      "loss": 1.7399,
+      "step": 7546
+    },
+    {
+      "epoch": 0.7839409992728784,
+      "grad_norm": 0.43174856901168823,
+      "learning_rate": 1.1082702916127063e-05,
+      "loss": 1.6816,
+      "step": 7547
+    },
+    {
+      "epoch": 0.7840448737924587,
+      "grad_norm": 0.3892672061920166,
+      "learning_rate": 1.1072460863413092e-05,
+      "loss": 1.3392,
+      "step": 7548
+    },
+    {
+      "epoch": 0.7841487483120391,
+      "grad_norm": 0.4050576388835907,
+      "learning_rate": 1.1062222956178391e-05,
+      "loss": 1.6167,
+      "step": 7549
+    },
+    {
+      "epoch": 0.7842526228316195,
+      "grad_norm": 0.43549004197120667,
+      "learning_rate": 1.1051989195513207e-05,
+      "loss": 1.7752,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7843564973511997,
+      "grad_norm": 0.4682556390762329,
+      "learning_rate": 1.1041759582507393e-05,
+      "loss": 1.6396,
+      "step": 7551
+    },
+    {
+      "epoch": 0.7844603718707801,
+      "grad_norm": 0.3986964821815491,
+      "learning_rate": 1.1031534118250269e-05,
+      "loss": 1.697,
+      "step": 7552
+    },
+    {
+      "epoch": 0.7845642463903605,
+      "grad_norm": 0.4416189193725586,
+      "learning_rate": 1.1021312803830813e-05,
+      "loss": 1.5987,
+      "step": 7553
+    },
+    {
+      "epoch": 0.7846681209099408,
+      "grad_norm": 0.39252594113349915,
+      "learning_rate": 1.101109564033751e-05,
+      "loss": 1.5662,
+      "step": 7554
+    },
+    {
+      "epoch": 0.7847719954295211,
+      "grad_norm": 0.4105452299118042,
+      "learning_rate": 1.1000882628858367e-05,
+      "loss": 1.4838,
+      "step": 7555
+    },
+    {
+      "epoch": 0.7848758699491015,
+      "grad_norm": 0.409593790769577,
+      "learning_rate": 1.0990673770481036e-05,
+      "loss": 1.6211,
+      "step": 7556
+    },
+    {
+      "epoch": 0.7849797444686818,
+      "grad_norm": 0.40311574935913086,
+      "learning_rate": 1.0980469066292659e-05,
+      "loss": 1.5268,
+      "step": 7557
+    },
+    {
+      "epoch": 0.7850836189882622,
+      "grad_norm": 0.4237530827522278,
+      "learning_rate": 1.0970268517379967e-05,
+      "loss": 1.7556,
+      "step": 7558
+    },
+    {
+      "epoch": 0.7851874935078426,
+      "grad_norm": 0.4404970407485962,
+      "learning_rate": 1.0960072124829235e-05,
+      "loss": 1.5014,
+      "step": 7559
+    },
+    {
+      "epoch": 0.7852913680274228,
+      "grad_norm": 0.42267706990242004,
+      "learning_rate": 1.0949879889726294e-05,
+      "loss": 1.6162,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7853952425470032,
+      "grad_norm": 0.44833898544311523,
+      "learning_rate": 1.0939691813156544e-05,
+      "loss": 1.7035,
+      "step": 7561
+    },
+    {
+      "epoch": 0.7854991170665836,
+      "grad_norm": 0.4051210284233093,
+      "learning_rate": 1.0929507896204921e-05,
+      "loss": 1.6337,
+      "step": 7562
+    },
+    {
+      "epoch": 0.7856029915861639,
+      "grad_norm": 0.4495202898979187,
+      "learning_rate": 1.0919328139955976e-05,
+      "loss": 1.6693,
+      "step": 7563
+    },
+    {
+      "epoch": 0.7857068661057443,
+      "grad_norm": 0.4015834331512451,
+      "learning_rate": 1.0909152545493712e-05,
+      "loss": 1.6461,
+      "step": 7564
+    },
+    {
+      "epoch": 0.7858107406253246,
+      "grad_norm": 0.4055159091949463,
+      "learning_rate": 1.089898111390179e-05,
+      "loss": 1.6278,
+      "step": 7565
+    },
+    {
+      "epoch": 0.7859146151449049,
+      "grad_norm": 0.3767836093902588,
+      "learning_rate": 1.0888813846263391e-05,
+      "loss": 1.5302,
+      "step": 7566
+    },
+    {
+      "epoch": 0.7860184896644853,
+      "grad_norm": 0.4378937780857086,
+      "learning_rate": 1.0878650743661229e-05,
+      "loss": 1.6846,
+      "step": 7567
+    },
+    {
+      "epoch": 0.7861223641840657,
+      "grad_norm": 0.43852996826171875,
+      "learning_rate": 1.086849180717761e-05,
+      "loss": 1.6535,
+      "step": 7568
+    },
+    {
+      "epoch": 0.7862262387036459,
+      "grad_norm": 0.42073825001716614,
+      "learning_rate": 1.0858337037894372e-05,
+      "loss": 1.6498,
+      "step": 7569
+    },
+    {
+      "epoch": 0.7863301132232263,
+      "grad_norm": 0.44468173384666443,
+      "learning_rate": 1.0848186436892927e-05,
+      "loss": 1.5972,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7864339877428067,
+      "grad_norm": 0.4390113055706024,
+      "learning_rate": 1.0838040005254213e-05,
+      "loss": 1.618,
+      "step": 7571
+    },
+    {
+      "epoch": 0.7865378622623871,
+      "grad_norm": 0.3942447602748871,
+      "learning_rate": 1.082789774405878e-05,
+      "loss": 1.5818,
+      "step": 7572
+    },
+    {
+      "epoch": 0.7866417367819674,
+      "grad_norm": 0.4129110276699066,
+      "learning_rate": 1.0817759654386684e-05,
+      "loss": 1.512,
+      "step": 7573
+    },
+    {
+      "epoch": 0.7867456113015477,
+      "grad_norm": 0.4303886592388153,
+      "learning_rate": 1.0807625737317545e-05,
+      "loss": 1.6786,
+      "step": 7574
+    },
+    {
+      "epoch": 0.7868494858211281,
+      "grad_norm": 0.47462910413742065,
+      "learning_rate": 1.0797495993930562e-05,
+      "loss": 1.7152,
+      "step": 7575
+    },
+    {
+      "epoch": 0.7869533603407084,
+      "grad_norm": 0.41988933086395264,
+      "learning_rate": 1.0787370425304454e-05,
+      "loss": 1.7505,
+      "step": 7576
+    },
+    {
+      "epoch": 0.7870572348602888,
+      "grad_norm": 0.41259077191352844,
+      "learning_rate": 1.0777249032517534e-05,
+      "loss": 1.6145,
+      "step": 7577
+    },
+    {
+      "epoch": 0.7871611093798692,
+      "grad_norm": 0.3950294256210327,
+      "learning_rate": 1.0767131816647624e-05,
+      "loss": 1.4186,
+      "step": 7578
+    },
+    {
+      "epoch": 0.7872649838994494,
+      "grad_norm": 0.4047811031341553,
+      "learning_rate": 1.0757018778772181e-05,
+      "loss": 1.7788,
+      "step": 7579
+    },
+    {
+      "epoch": 0.7873688584190298,
+      "grad_norm": 0.46056464314460754,
+      "learning_rate": 1.0746909919968102e-05,
+      "loss": 1.7623,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7874727329386102,
+      "grad_norm": 0.4839528203010559,
+      "learning_rate": 1.0736805241311937e-05,
+      "loss": 1.7721,
+      "step": 7581
+    },
+    {
+      "epoch": 0.7875766074581905,
+      "grad_norm": 0.44515979290008545,
+      "learning_rate": 1.0726704743879767e-05,
+      "loss": 1.6909,
+      "step": 7582
+    },
+    {
+      "epoch": 0.7876804819777709,
+      "grad_norm": 0.4091649353504181,
+      "learning_rate": 1.0716608428747177e-05,
+      "loss": 1.6288,
+      "step": 7583
+    },
+    {
+      "epoch": 0.7877843564973512,
+      "grad_norm": 0.4144968092441559,
+      "learning_rate": 1.0706516296989372e-05,
+      "loss": 1.7996,
+      "step": 7584
+    },
+    {
+      "epoch": 0.7878882310169315,
+      "grad_norm": 0.3900415301322937,
+      "learning_rate": 1.0696428349681087e-05,
+      "loss": 1.5053,
+      "step": 7585
+    },
+    {
+      "epoch": 0.7879921055365119,
+      "grad_norm": 0.4206119775772095,
+      "learning_rate": 1.0686344587896597e-05,
+      "loss": 1.6293,
+      "step": 7586
+    },
+    {
+      "epoch": 0.7880959800560923,
+      "grad_norm": 0.421841025352478,
+      "learning_rate": 1.0676265012709752e-05,
+      "loss": 1.628,
+      "step": 7587
+    },
+    {
+      "epoch": 0.7881998545756725,
+      "grad_norm": 0.45985618233680725,
+      "learning_rate": 1.066618962519395e-05,
+      "loss": 1.8276,
+      "step": 7588
+    },
+    {
+      "epoch": 0.7883037290952529,
+      "grad_norm": 0.414380818605423,
+      "learning_rate": 1.0656118426422134e-05,
+      "loss": 1.6578,
+      "step": 7589
+    },
+    {
+      "epoch": 0.7884076036148333,
+      "grad_norm": 0.4441922903060913,
+      "learning_rate": 1.06460514174668e-05,
+      "loss": 1.74,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7885114781344136,
+      "grad_norm": 0.45518070459365845,
+      "learning_rate": 1.0635988599400043e-05,
+      "loss": 1.7842,
+      "step": 7591
+    },
+    {
+      "epoch": 0.788615352653994,
+      "grad_norm": 0.3902077078819275,
+      "learning_rate": 1.0625929973293419e-05,
+      "loss": 1.5922,
+      "step": 7592
+    },
+    {
+      "epoch": 0.7887192271735743,
+      "grad_norm": 0.4476879835128784,
+      "learning_rate": 1.061587554021814e-05,
+      "loss": 1.7011,
+      "step": 7593
+    },
+    {
+      "epoch": 0.7888231016931546,
+      "grad_norm": 0.41619980335235596,
+      "learning_rate": 1.0605825301244909e-05,
+      "loss": 1.683,
+      "step": 7594
+    },
+    {
+      "epoch": 0.788926976212735,
+      "grad_norm": 0.40933772921562195,
+      "learning_rate": 1.0595779257444e-05,
+      "loss": 1.4897,
+      "step": 7595
+    },
+    {
+      "epoch": 0.7890308507323154,
+      "grad_norm": 0.4322732388973236,
+      "learning_rate": 1.0585737409885227e-05,
+      "loss": 1.6298,
+      "step": 7596
+    },
+    {
+      "epoch": 0.7891347252518958,
+      "grad_norm": 0.4134397804737091,
+      "learning_rate": 1.057569975963797e-05,
+      "loss": 1.6549,
+      "step": 7597
+    },
+    {
+      "epoch": 0.789238599771476,
+      "grad_norm": 0.44103166460990906,
+      "learning_rate": 1.0565666307771193e-05,
+      "loss": 1.7137,
+      "step": 7598
+    },
+    {
+      "epoch": 0.7893424742910564,
+      "grad_norm": 0.4072246253490448,
+      "learning_rate": 1.0555637055353324e-05,
+      "loss": 1.5784,
+      "step": 7599
+    },
+    {
+      "epoch": 0.7894463488106368,
+      "grad_norm": 0.44003406167030334,
+      "learning_rate": 1.054561200345245e-05,
+      "loss": 1.6315,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7895502233302171,
+      "grad_norm": 0.4993717074394226,
+      "learning_rate": 1.0535591153136137e-05,
+      "loss": 1.7496,
+      "step": 7601
+    },
+    {
+      "epoch": 0.7896540978497975,
+      "grad_norm": 0.41622164845466614,
+      "learning_rate": 1.0525574505471531e-05,
+      "loss": 1.6808,
+      "step": 7602
+    },
+    {
+      "epoch": 0.7897579723693778,
+      "grad_norm": 0.3730601370334625,
+      "learning_rate": 1.051556206152533e-05,
+      "loss": 1.5215,
+      "step": 7603
+    },
+    {
+      "epoch": 0.7898618468889581,
+      "grad_norm": 0.41570550203323364,
+      "learning_rate": 1.0505553822363783e-05,
+      "loss": 1.6401,
+      "step": 7604
+    },
+    {
+      "epoch": 0.7899657214085385,
+      "grad_norm": 0.41199877858161926,
+      "learning_rate": 1.0495549789052678e-05,
+      "loss": 1.613,
+      "step": 7605
+    },
+    {
+      "epoch": 0.7900695959281189,
+      "grad_norm": 0.4056338667869568,
+      "learning_rate": 1.0485549962657365e-05,
+      "loss": 1.6405,
+      "step": 7606
+    },
+    {
+      "epoch": 0.7901734704476991,
+      "grad_norm": 0.39273974299430847,
+      "learning_rate": 1.0475554344242783e-05,
+      "loss": 1.5479,
+      "step": 7607
+    },
+    {
+      "epoch": 0.7902773449672795,
+      "grad_norm": 0.454589307308197,
+      "learning_rate": 1.0465562934873335e-05,
+      "loss": 1.8367,
+      "step": 7608
+    },
+    {
+      "epoch": 0.7903812194868599,
+      "grad_norm": 0.4377838373184204,
+      "learning_rate": 1.0455575735613066e-05,
+      "loss": 1.6501,
+      "step": 7609
+    },
+    {
+      "epoch": 0.7904850940064402,
+      "grad_norm": 0.41390401124954224,
+      "learning_rate": 1.0445592747525534e-05,
+      "loss": 1.5864,
+      "step": 7610
+    },
+    {
+      "epoch": 0.7905889685260206,
+      "grad_norm": 0.40852096676826477,
+      "learning_rate": 1.0435613971673813e-05,
+      "loss": 1.5591,
+      "step": 7611
+    },
+    {
+      "epoch": 0.790692843045601,
+      "grad_norm": 0.43240541219711304,
+      "learning_rate": 1.0425639409120602e-05,
+      "loss": 1.6134,
+      "step": 7612
+    },
+    {
+      "epoch": 0.7907967175651812,
+      "grad_norm": 0.4393400549888611,
+      "learning_rate": 1.0415669060928108e-05,
+      "loss": 1.7608,
+      "step": 7613
+    },
+    {
+      "epoch": 0.7909005920847616,
+      "grad_norm": 0.4604494571685791,
+      "learning_rate": 1.0405702928158085e-05,
+      "loss": 1.7024,
+      "step": 7614
+    },
+    {
+      "epoch": 0.791004466604342,
+      "grad_norm": 0.40346240997314453,
+      "learning_rate": 1.0395741011871851e-05,
+      "loss": 1.6004,
+      "step": 7615
+    },
+    {
+      "epoch": 0.7911083411239223,
+      "grad_norm": 0.39214804768562317,
+      "learning_rate": 1.0385783313130276e-05,
+      "loss": 1.6282,
+      "step": 7616
+    },
+    {
+      "epoch": 0.7912122156435026,
+      "grad_norm": 0.44647249579429626,
+      "learning_rate": 1.0375829832993772e-05,
+      "loss": 1.7222,
+      "step": 7617
+    },
+    {
+      "epoch": 0.791316090163083,
+      "grad_norm": 0.4374501705169678,
+      "learning_rate": 1.0365880572522302e-05,
+      "loss": 1.7329,
+      "step": 7618
+    },
+    {
+      "epoch": 0.7914199646826633,
+      "grad_norm": 0.4381657838821411,
+      "learning_rate": 1.0355935532775419e-05,
+      "loss": 1.5133,
+      "step": 7619
+    },
+    {
+      "epoch": 0.7915238392022437,
+      "grad_norm": 0.4005827009677887,
+      "learning_rate": 1.0345994714812135e-05,
+      "loss": 1.4862,
+      "step": 7620
+    },
+    {
+      "epoch": 0.7916277137218241,
+      "grad_norm": 0.40875616669654846,
+      "learning_rate": 1.0336058119691122e-05,
+      "loss": 1.5222,
+      "step": 7621
+    },
+    {
+      "epoch": 0.7917315882414044,
+      "grad_norm": 0.4034888744354248,
+      "learning_rate": 1.0326125748470522e-05,
+      "loss": 1.6171,
+      "step": 7622
+    },
+    {
+      "epoch": 0.7918354627609847,
+      "grad_norm": 0.42959120869636536,
+      "learning_rate": 1.0316197602208067e-05,
+      "loss": 1.7382,
+      "step": 7623
+    },
+    {
+      "epoch": 0.7919393372805651,
+      "grad_norm": 0.5072987079620361,
+      "learning_rate": 1.0306273681961014e-05,
+      "loss": 1.8618,
+      "step": 7624
+    },
+    {
+      "epoch": 0.7920432118001455,
+      "grad_norm": 0.4025663435459137,
+      "learning_rate": 1.0296353988786184e-05,
+      "loss": 1.7562,
+      "step": 7625
+    },
+    {
+      "epoch": 0.7921470863197257,
+      "grad_norm": 0.4331921637058258,
+      "learning_rate": 1.0286438523739978e-05,
+      "loss": 1.5518,
+      "step": 7626
+    },
+    {
+      "epoch": 0.7922509608393061,
+      "grad_norm": 0.46530386805534363,
+      "learning_rate": 1.0276527287878268e-05,
+      "loss": 1.831,
+      "step": 7627
+    },
+    {
+      "epoch": 0.7923548353588865,
+      "grad_norm": 0.43432602286338806,
+      "learning_rate": 1.0266620282256556e-05,
+      "loss": 1.69,
+      "step": 7628
+    },
+    {
+      "epoch": 0.7924587098784668,
+      "grad_norm": 0.39210620522499084,
+      "learning_rate": 1.0256717507929852e-05,
+      "loss": 1.4931,
+      "step": 7629
+    },
+    {
+      "epoch": 0.7925625843980472,
+      "grad_norm": 0.4316072463989258,
+      "learning_rate": 1.0246818965952727e-05,
+      "loss": 1.5392,
+      "step": 7630
+    },
+    {
+      "epoch": 0.7926664589176275,
+      "grad_norm": 0.422419011592865,
+      "learning_rate": 1.0236924657379293e-05,
+      "loss": 1.5485,
+      "step": 7631
+    },
+    {
+      "epoch": 0.7927703334372078,
+      "grad_norm": 0.43081673979759216,
+      "learning_rate": 1.0227034583263218e-05,
+      "loss": 1.7733,
+      "step": 7632
+    },
+    {
+      "epoch": 0.7928742079567882,
+      "grad_norm": 0.4367268979549408,
+      "learning_rate": 1.0217148744657723e-05,
+      "loss": 1.7832,
+      "step": 7633
+    },
+    {
+      "epoch": 0.7929780824763686,
+      "grad_norm": 0.42134618759155273,
+      "learning_rate": 1.0207267142615552e-05,
+      "loss": 1.6996,
+      "step": 7634
+    },
+    {
+      "epoch": 0.7930819569959489,
+      "grad_norm": 0.40299269556999207,
+      "learning_rate": 1.0197389778189065e-05,
+      "loss": 1.6218,
+      "step": 7635
+    },
+    {
+      "epoch": 0.7931858315155292,
+      "grad_norm": 0.414371132850647,
+      "learning_rate": 1.0187516652430062e-05,
+      "loss": 1.6509,
+      "step": 7636
+    },
+    {
+      "epoch": 0.7932897060351096,
+      "grad_norm": 0.4730742573738098,
+      "learning_rate": 1.0177647766390002e-05,
+      "loss": 1.7145,
+      "step": 7637
+    },
+    {
+      "epoch": 0.7933935805546899,
+      "grad_norm": 0.43268048763275146,
+      "learning_rate": 1.0167783121119845e-05,
+      "loss": 1.3435,
+      "step": 7638
+    },
+    {
+      "epoch": 0.7934974550742703,
+      "grad_norm": 0.43376022577285767,
+      "learning_rate": 1.0157922717670054e-05,
+      "loss": 1.5719,
+      "step": 7639
+    },
+    {
+      "epoch": 0.7936013295938507,
+      "grad_norm": 0.40382814407348633,
+      "learning_rate": 1.0148066557090724e-05,
+      "loss": 1.6977,
+      "step": 7640
+    },
+    {
+      "epoch": 0.7937052041134309,
+      "grad_norm": 0.45663517713546753,
+      "learning_rate": 1.0138214640431454e-05,
+      "loss": 1.7793,
+      "step": 7641
+    },
+    {
+      "epoch": 0.7938090786330113,
+      "grad_norm": 0.4395454525947571,
+      "learning_rate": 1.0128366968741387e-05,
+      "loss": 1.7664,
+      "step": 7642
+    },
+    {
+      "epoch": 0.7939129531525917,
+      "grad_norm": 0.4374726116657257,
+      "learning_rate": 1.0118523543069226e-05,
+      "loss": 1.6902,
+      "step": 7643
+    },
+    {
+      "epoch": 0.794016827672172,
+      "grad_norm": 0.459427148103714,
+      "learning_rate": 1.0108684364463222e-05,
+      "loss": 1.6312,
+      "step": 7644
+    },
+    {
+      "epoch": 0.7941207021917523,
+      "grad_norm": 0.42887747287750244,
+      "learning_rate": 1.0098849433971169e-05,
+      "loss": 1.6194,
+      "step": 7645
+    },
+    {
+      "epoch": 0.7942245767113327,
+      "grad_norm": 0.4146133363246918,
+      "learning_rate": 1.0089018752640394e-05,
+      "loss": 1.6965,
+      "step": 7646
+    },
+    {
+      "epoch": 0.794328451230913,
+      "grad_norm": 0.46396344900131226,
+      "learning_rate": 1.0079192321517838e-05,
+      "loss": 1.7115,
+      "step": 7647
+    },
+    {
+      "epoch": 0.7944323257504934,
+      "grad_norm": 0.4100917875766754,
+      "learning_rate": 1.0069370141649876e-05,
+      "loss": 1.4832,
+      "step": 7648
+    },
+    {
+      "epoch": 0.7945362002700738,
+      "grad_norm": 0.4542340040206909,
+      "learning_rate": 1.0059552214082536e-05,
+      "loss": 1.6654,
+      "step": 7649
+    },
+    {
+      "epoch": 0.7946400747896541,
+      "grad_norm": 0.4383617043495178,
+      "learning_rate": 1.0049738539861332e-05,
+      "loss": 1.6049,
+      "step": 7650
+    },
+    {
+      "epoch": 0.7947439493092344,
+      "grad_norm": 0.4077033996582031,
+      "learning_rate": 1.0039929120031355e-05,
+      "loss": 1.7055,
+      "step": 7651
+    },
+    {
+      "epoch": 0.7948478238288148,
+      "grad_norm": 0.4490281641483307,
+      "learning_rate": 1.0030123955637227e-05,
+      "loss": 1.7024,
+      "step": 7652
+    },
+    {
+      "epoch": 0.7949516983483952,
+      "grad_norm": 0.4433899223804474,
+      "learning_rate": 1.0020323047723108e-05,
+      "loss": 1.7098,
+      "step": 7653
+    },
+    {
+      "epoch": 0.7950555728679755,
+      "grad_norm": 0.3918624818325043,
+      "learning_rate": 1.0010526397332754e-05,
+      "loss": 1.5875,
+      "step": 7654
+    },
+    {
+      "epoch": 0.7951594473875558,
+      "grad_norm": 0.4222068190574646,
+      "learning_rate": 1.0000734005509382e-05,
+      "loss": 1.6725,
+      "step": 7655
+    },
+    {
+      "epoch": 0.7952633219071362,
+      "grad_norm": 0.4067675471305847,
+      "learning_rate": 9.990945873295848e-06,
+      "loss": 1.6157,
+      "step": 7656
+    },
+    {
+      "epoch": 0.7953671964267165,
+      "grad_norm": 0.4570964574813843,
+      "learning_rate": 9.981162001734496e-06,
+      "loss": 1.7816,
+      "step": 7657
+    },
+    {
+      "epoch": 0.7954710709462969,
+      "grad_norm": 0.4113404452800751,
+      "learning_rate": 9.97138239186723e-06,
+      "loss": 1.6609,
+      "step": 7658
+    },
+    {
+      "epoch": 0.7955749454658773,
+      "grad_norm": 0.4180930256843567,
+      "learning_rate": 9.961607044735505e-06,
+      "loss": 1.5903,
+      "step": 7659
+    },
+    {
+      "epoch": 0.7956788199854575,
+      "grad_norm": 0.4389685094356537,
+      "learning_rate": 9.951835961380317e-06,
+      "loss": 1.655,
+      "step": 7660
+    },
+    {
+      "epoch": 0.7957826945050379,
+      "grad_norm": 0.38533127307891846,
+      "learning_rate": 9.942069142842214e-06,
+      "loss": 1.5481,
+      "step": 7661
+    },
+    {
+      "epoch": 0.7958865690246183,
+      "grad_norm": 0.4118013083934784,
+      "learning_rate": 9.932306590161266e-06,
+      "loss": 1.6456,
+      "step": 7662
+    },
+    {
+      "epoch": 0.7959904435441986,
+      "grad_norm": 0.433296263217926,
+      "learning_rate": 9.922548304377154e-06,
+      "loss": 1.6836,
+      "step": 7663
+    },
+    {
+      "epoch": 0.796094318063779,
+      "grad_norm": 0.41111084818840027,
+      "learning_rate": 9.91279428652901e-06,
+      "loss": 1.5107,
+      "step": 7664
+    },
+    {
+      "epoch": 0.7961981925833593,
+      "grad_norm": 0.4467814862728119,
+      "learning_rate": 9.903044537655587e-06,
+      "loss": 1.6828,
+      "step": 7665
+    },
+    {
+      "epoch": 0.7963020671029396,
+      "grad_norm": 0.3999803364276886,
+      "learning_rate": 9.893299058795175e-06,
+      "loss": 1.5791,
+      "step": 7666
+    },
+    {
+      "epoch": 0.79640594162252,
+      "grad_norm": 0.4356433153152466,
+      "learning_rate": 9.883557850985537e-06,
+      "loss": 1.6213,
+      "step": 7667
+    },
+    {
+      "epoch": 0.7965098161421004,
+      "grad_norm": 0.44254523515701294,
+      "learning_rate": 9.873820915264082e-06,
+      "loss": 1.7171,
+      "step": 7668
+    },
+    {
+      "epoch": 0.7966136906616806,
+      "grad_norm": 0.4416894018650055,
+      "learning_rate": 9.864088252667702e-06,
+      "loss": 1.5873,
+      "step": 7669
+    },
+    {
+      "epoch": 0.796717565181261,
+      "grad_norm": 0.46283599734306335,
+      "learning_rate": 9.854359864232854e-06,
+      "loss": 1.8697,
+      "step": 7670
+    },
+    {
+      "epoch": 0.7968214397008414,
+      "grad_norm": 0.41159945726394653,
+      "learning_rate": 9.84463575099553e-06,
+      "loss": 1.6293,
+      "step": 7671
+    },
+    {
+      "epoch": 0.7969253142204217,
+      "grad_norm": 0.4374866187572479,
+      "learning_rate": 9.834915913991272e-06,
+      "loss": 1.7457,
+      "step": 7672
+    },
+    {
+      "epoch": 0.7970291887400021,
+      "grad_norm": 0.37280136346817017,
+      "learning_rate": 9.825200354255171e-06,
+      "loss": 1.5509,
+      "step": 7673
+    },
+    {
+      "epoch": 0.7971330632595824,
+      "grad_norm": 0.41301167011260986,
+      "learning_rate": 9.815489072821848e-06,
+      "loss": 1.6039,
+      "step": 7674
+    },
+    {
+      "epoch": 0.7972369377791628,
+      "grad_norm": 0.3978497087955475,
+      "learning_rate": 9.805782070725494e-06,
+      "loss": 1.5862,
+      "step": 7675
+    },
+    {
+      "epoch": 0.7973408122987431,
+      "grad_norm": 0.42098501324653625,
+      "learning_rate": 9.796079348999831e-06,
+      "loss": 1.6685,
+      "step": 7676
+    },
+    {
+      "epoch": 0.7974446868183235,
+      "grad_norm": 0.4220602214336395,
+      "learning_rate": 9.786380908678112e-06,
+      "loss": 1.6917,
+      "step": 7677
+    },
+    {
+      "epoch": 0.7975485613379039,
+      "grad_norm": 0.44230684638023376,
+      "learning_rate": 9.77668675079314e-06,
+      "loss": 1.6786,
+      "step": 7678
+    },
+    {
+      "epoch": 0.7976524358574841,
+      "grad_norm": 0.49053019285202026,
+      "learning_rate": 9.766996876377283e-06,
+      "loss": 1.5424,
+      "step": 7679
+    },
+    {
+      "epoch": 0.7977563103770645,
+      "grad_norm": 0.41432085633277893,
+      "learning_rate": 9.757311286462429e-06,
+      "loss": 1.6682,
+      "step": 7680
+    },
+    {
+      "epoch": 0.7978601848966449,
+      "grad_norm": 0.41580092906951904,
+      "learning_rate": 9.747629982080003e-06,
+      "loss": 1.7237,
+      "step": 7681
+    },
+    {
+      "epoch": 0.7979640594162252,
+      "grad_norm": 0.4454309642314911,
+      "learning_rate": 9.737952964261027e-06,
+      "loss": 1.7689,
+      "step": 7682
+    },
+    {
+      "epoch": 0.7980679339358056,
+      "grad_norm": 0.40397220849990845,
+      "learning_rate": 9.728280234035981e-06,
+      "loss": 1.6688,
+      "step": 7683
+    },
+    {
+      "epoch": 0.7981718084553859,
+      "grad_norm": 0.43213963508605957,
+      "learning_rate": 9.718611792434973e-06,
+      "loss": 1.7833,
+      "step": 7684
+    },
+    {
+      "epoch": 0.7982756829749662,
+      "grad_norm": 0.44224637746810913,
+      "learning_rate": 9.708947640487598e-06,
+      "loss": 1.6129,
+      "step": 7685
+    },
+    {
+      "epoch": 0.7983795574945466,
+      "grad_norm": 0.42797449231147766,
+      "learning_rate": 9.699287779223022e-06,
+      "loss": 1.7152,
+      "step": 7686
+    },
+    {
+      "epoch": 0.798483432014127,
+      "grad_norm": 0.39138954877853394,
+      "learning_rate": 9.689632209669935e-06,
+      "loss": 1.5439,
+      "step": 7687
+    },
+    {
+      "epoch": 0.7985873065337072,
+      "grad_norm": 0.42455416917800903,
+      "learning_rate": 9.679980932856591e-06,
+      "loss": 1.6193,
+      "step": 7688
+    },
+    {
+      "epoch": 0.7986911810532876,
+      "grad_norm": 0.392198383808136,
+      "learning_rate": 9.670333949810761e-06,
+      "loss": 1.5908,
+      "step": 7689
+    },
+    {
+      "epoch": 0.798795055572868,
+      "grad_norm": 0.4087858200073242,
+      "learning_rate": 9.660691261559774e-06,
+      "loss": 1.6236,
+      "step": 7690
+    },
+    {
+      "epoch": 0.7988989300924483,
+      "grad_norm": 0.38556161522865295,
+      "learning_rate": 9.65105286913054e-06,
+      "loss": 1.5387,
+      "step": 7691
+    },
+    {
+      "epoch": 0.7990028046120287,
+      "grad_norm": 0.40947332978248596,
+      "learning_rate": 9.641418773549405e-06,
+      "loss": 1.583,
+      "step": 7692
+    },
+    {
+      "epoch": 0.799106679131609,
+      "grad_norm": 0.4197182059288025,
+      "learning_rate": 9.631788975842381e-06,
+      "loss": 1.593,
+      "step": 7693
+    },
+    {
+      "epoch": 0.7992105536511893,
+      "grad_norm": 0.43122637271881104,
+      "learning_rate": 9.622163477034957e-06,
+      "loss": 1.6916,
+      "step": 7694
+    },
+    {
+      "epoch": 0.7993144281707697,
+      "grad_norm": 0.4150683283805847,
+      "learning_rate": 9.612542278152142e-06,
+      "loss": 1.6192,
+      "step": 7695
+    },
+    {
+      "epoch": 0.7994183026903501,
+      "grad_norm": 0.41880491375923157,
+      "learning_rate": 9.602925380218547e-06,
+      "loss": 1.6866,
+      "step": 7696
+    },
+    {
+      "epoch": 0.7995221772099304,
+      "grad_norm": 0.40711480379104614,
+      "learning_rate": 9.593312784258297e-06,
+      "loss": 1.7376,
+      "step": 7697
+    },
+    {
+      "epoch": 0.7996260517295107,
+      "grad_norm": 0.4136184751987457,
+      "learning_rate": 9.583704491295042e-06,
+      "loss": 1.6386,
+      "step": 7698
+    },
+    {
+      "epoch": 0.7997299262490911,
+      "grad_norm": 0.41841137409210205,
+      "learning_rate": 9.574100502352007e-06,
+      "loss": 1.7478,
+      "step": 7699
+    },
+    {
+      "epoch": 0.7998338007686715,
+      "grad_norm": 0.38687604665756226,
+      "learning_rate": 9.564500818451916e-06,
+      "loss": 1.4819,
+      "step": 7700
+    },
+    {
+      "epoch": 0.7999376752882518,
+      "grad_norm": 0.40951046347618103,
+      "learning_rate": 9.554905440617108e-06,
+      "loss": 1.5952,
+      "step": 7701
+    },
+    {
+      "epoch": 0.8000415498078322,
+      "grad_norm": 0.4178246855735779,
+      "learning_rate": 9.545314369869362e-06,
+      "loss": 1.7333,
+      "step": 7702
+    },
+    {
+      "epoch": 0.8001454243274125,
+      "grad_norm": 0.4111107289791107,
+      "learning_rate": 9.535727607230088e-06,
+      "loss": 1.7021,
+      "step": 7703
+    },
+    {
+      "epoch": 0.8002492988469928,
+      "grad_norm": 0.42014679312705994,
+      "learning_rate": 9.52614515372019e-06,
+      "loss": 1.5992,
+      "step": 7704
+    },
+    {
+      "epoch": 0.8003531733665732,
+      "grad_norm": 0.45557916164398193,
+      "learning_rate": 9.516567010360127e-06,
+      "loss": 1.2576,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8004570478861536,
+      "grad_norm": 0.4034499228000641,
+      "learning_rate": 9.506993178169892e-06,
+      "loss": 1.5702,
+      "step": 7706
+    },
+    {
+      "epoch": 0.8005609224057338,
+      "grad_norm": 0.4066244661808014,
+      "learning_rate": 9.497423658169025e-06,
+      "loss": 1.5422,
+      "step": 7707
+    },
+    {
+      "epoch": 0.8006647969253142,
+      "grad_norm": 0.4864404499530792,
+      "learning_rate": 9.487858451376608e-06,
+      "loss": 1.7401,
+      "step": 7708
+    },
+    {
+      "epoch": 0.8007686714448946,
+      "grad_norm": 0.4497174024581909,
+      "learning_rate": 9.47829755881125e-06,
+      "loss": 1.7727,
+      "step": 7709
+    },
+    {
+      "epoch": 0.8008725459644749,
+      "grad_norm": 0.4629211723804474,
+      "learning_rate": 9.468740981491143e-06,
+      "loss": 1.6619,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8009764204840553,
+      "grad_norm": 0.392501562833786,
+      "learning_rate": 9.459188720433937e-06,
+      "loss": 1.5847,
+      "step": 7711
+    },
+    {
+      "epoch": 0.8010802950036356,
+      "grad_norm": 0.47376659512519836,
+      "learning_rate": 9.449640776656915e-06,
+      "loss": 1.6943,
+      "step": 7712
+    },
+    {
+      "epoch": 0.8011841695232159,
+      "grad_norm": 0.40953293442726135,
+      "learning_rate": 9.440097151176846e-06,
+      "loss": 1.5879,
+      "step": 7713
+    },
+    {
+      "epoch": 0.8012880440427963,
+      "grad_norm": 0.40460994839668274,
+      "learning_rate": 9.430557845010058e-06,
+      "loss": 1.7309,
+      "step": 7714
+    },
+    {
+      "epoch": 0.8013919185623767,
+      "grad_norm": 0.4047517776489258,
+      "learning_rate": 9.421022859172397e-06,
+      "loss": 1.7228,
+      "step": 7715
+    },
+    {
+      "epoch": 0.801495793081957,
+      "grad_norm": 0.397476464509964,
+      "learning_rate": 9.411492194679277e-06,
+      "loss": 1.6439,
+      "step": 7716
+    },
+    {
+      "epoch": 0.8015996676015373,
+      "grad_norm": 0.39920222759246826,
+      "learning_rate": 9.401965852545636e-06,
+      "loss": 1.5571,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8017035421211177,
+      "grad_norm": 0.4768927991390228,
+      "learning_rate": 9.39244383378594e-06,
+      "loss": 1.7649,
+      "step": 7718
+    },
+    {
+      "epoch": 0.801807416640698,
+      "grad_norm": 0.4142306447029114,
+      "learning_rate": 9.38292613941425e-06,
+      "loss": 1.3555,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8019112911602784,
+      "grad_norm": 0.4120739996433258,
+      "learning_rate": 9.37341277044408e-06,
+      "loss": 1.6209,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8020151656798588,
+      "grad_norm": 0.434986412525177,
+      "learning_rate": 9.363903727888556e-06,
+      "loss": 1.3753,
+      "step": 7721
+    },
+    {
+      "epoch": 0.802119040199439,
+      "grad_norm": 0.3884414732456207,
+      "learning_rate": 9.354399012760328e-06,
+      "loss": 1.4721,
+      "step": 7722
+    },
+    {
+      "epoch": 0.8022229147190194,
+      "grad_norm": 0.42061513662338257,
+      "learning_rate": 9.34489862607153e-06,
+      "loss": 1.5033,
+      "step": 7723
+    },
+    {
+      "epoch": 0.8023267892385998,
+      "grad_norm": 0.4704020023345947,
+      "learning_rate": 9.335402568833923e-06,
+      "loss": 1.8149,
+      "step": 7724
+    },
+    {
+      "epoch": 0.8024306637581802,
+      "grad_norm": 0.44218188524246216,
+      "learning_rate": 9.325910842058733e-06,
+      "loss": 1.7166,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8025345382777604,
+      "grad_norm": 0.4181729555130005,
+      "learning_rate": 9.316423446756788e-06,
+      "loss": 1.4822,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8026384127973408,
+      "grad_norm": 0.4159371256828308,
+      "learning_rate": 9.306940383938395e-06,
+      "loss": 1.7386,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8027422873169212,
+      "grad_norm": 0.45085909962654114,
+      "learning_rate": 9.297461654613415e-06,
+      "loss": 1.6013,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8028461618365015,
+      "grad_norm": 0.4149478077888489,
+      "learning_rate": 9.28798725979131e-06,
+      "loss": 1.6137,
+      "step": 7729
+    },
+    {
+      "epoch": 0.8029500363560819,
+      "grad_norm": 0.5053886771202087,
+      "learning_rate": 9.278517200480963e-06,
+      "loss": 1.8171,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8030539108756622,
+      "grad_norm": 0.4231288731098175,
+      "learning_rate": 9.26905147769091e-06,
+      "loss": 1.6836,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8031577853952425,
+      "grad_norm": 0.5030054450035095,
+      "learning_rate": 9.25959009242916e-06,
+      "loss": 1.9295,
+      "step": 7732
+    },
+    {
+      "epoch": 0.8032616599148229,
+      "grad_norm": 0.5075528025627136,
+      "learning_rate": 9.25013304570328e-06,
+      "loss": 1.8385,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8033655344344033,
+      "grad_norm": 0.43307042121887207,
+      "learning_rate": 9.24068033852037e-06,
+      "loss": 1.5867,
+      "step": 7734
+    },
+    {
+      "epoch": 0.8034694089539836,
+      "grad_norm": 0.4354676306247711,
+      "learning_rate": 9.231231971887066e-06,
+      "loss": 1.6309,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8035732834735639,
+      "grad_norm": 0.4056316912174225,
+      "learning_rate": 9.221787946809552e-06,
+      "loss": 1.5253,
+      "step": 7736
+    },
+    {
+      "epoch": 0.8036771579931443,
+      "grad_norm": 0.408130019903183,
+      "learning_rate": 9.212348264293519e-06,
+      "loss": 1.5669,
+      "step": 7737
+    },
+    {
+      "epoch": 0.8037810325127246,
+      "grad_norm": 0.4351346492767334,
+      "learning_rate": 9.202912925344276e-06,
+      "loss": 1.6559,
+      "step": 7738
+    },
+    {
+      "epoch": 0.803884907032305,
+      "grad_norm": 0.4343356788158417,
+      "learning_rate": 9.19348193096654e-06,
+      "loss": 1.7244,
+      "step": 7739
+    },
+    {
+      "epoch": 0.8039887815518854,
+      "grad_norm": 0.43080124258995056,
+      "learning_rate": 9.184055282164699e-06,
+      "loss": 1.5402,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8040926560714656,
+      "grad_norm": 0.4280051290988922,
+      "learning_rate": 9.174632979942587e-06,
+      "loss": 1.6772,
+      "step": 7741
+    },
+    {
+      "epoch": 0.804196530591046,
+      "grad_norm": 0.45156073570251465,
+      "learning_rate": 9.165215025303614e-06,
+      "loss": 1.7684,
+      "step": 7742
+    },
+    {
+      "epoch": 0.8043004051106264,
+      "grad_norm": 0.5198978185653687,
+      "learning_rate": 9.15580141925072e-06,
+      "loss": 1.8452,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8044042796302067,
+      "grad_norm": 0.4284294843673706,
+      "learning_rate": 9.146392162786377e-06,
+      "loss": 1.7204,
+      "step": 7744
+    },
+    {
+      "epoch": 0.804508154149787,
+      "grad_norm": 0.41050204634666443,
+      "learning_rate": 9.136987256912594e-06,
+      "loss": 1.4458,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8046120286693674,
+      "grad_norm": 0.5094167590141296,
+      "learning_rate": 9.127586702630913e-06,
+      "loss": 1.7218,
+      "step": 7746
+    },
+    {
+      "epoch": 0.8047159031889477,
+      "grad_norm": 0.44446441531181335,
+      "learning_rate": 9.118190500942459e-06,
+      "loss": 1.7473,
+      "step": 7747
+    },
+    {
+      "epoch": 0.8048197777085281,
+      "grad_norm": 0.4276479482650757,
+      "learning_rate": 9.108798652847811e-06,
+      "loss": 1.6519,
+      "step": 7748
+    },
+    {
+      "epoch": 0.8049236522281085,
+      "grad_norm": 0.42676594853401184,
+      "learning_rate": 9.099411159347133e-06,
+      "loss": 1.5866,
+      "step": 7749
+    },
+    {
+      "epoch": 0.8050275267476887,
+      "grad_norm": 0.43534624576568604,
+      "learning_rate": 9.090028021440155e-06,
+      "loss": 1.8477,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8051314012672691,
+      "grad_norm": 0.4531230330467224,
+      "learning_rate": 9.080649240126054e-06,
+      "loss": 1.6118,
+      "step": 7751
+    },
+    {
+      "epoch": 0.8052352757868495,
+      "grad_norm": 0.4191751182079315,
+      "learning_rate": 9.07127481640364e-06,
+      "loss": 1.5275,
+      "step": 7752
+    },
+    {
+      "epoch": 0.8053391503064299,
+      "grad_norm": 0.4555509090423584,
+      "learning_rate": 9.061904751271193e-06,
+      "loss": 1.6641,
+      "step": 7753
+    },
+    {
+      "epoch": 0.8054430248260102,
+      "grad_norm": 0.4250716269016266,
+      "learning_rate": 9.052539045726583e-06,
+      "loss": 1.7141,
+      "step": 7754
+    },
+    {
+      "epoch": 0.8055468993455905,
+      "grad_norm": 0.42984840273857117,
+      "learning_rate": 9.043177700767148e-06,
+      "loss": 1.6593,
+      "step": 7755
+    },
+    {
+      "epoch": 0.8056507738651709,
+      "grad_norm": 0.4200412929058075,
+      "learning_rate": 9.033820717389802e-06,
+      "loss": 1.6844,
+      "step": 7756
+    },
+    {
+      "epoch": 0.8057546483847512,
+      "grad_norm": 0.3976275622844696,
+      "learning_rate": 9.024468096591022e-06,
+      "loss": 1.6062,
+      "step": 7757
+    },
+    {
+      "epoch": 0.8058585229043316,
+      "grad_norm": 0.40078333020210266,
+      "learning_rate": 9.015119839366748e-06,
+      "loss": 1.6187,
+      "step": 7758
+    },
+    {
+      "epoch": 0.805962397423912,
+      "grad_norm": 0.4719092845916748,
+      "learning_rate": 9.005775946712524e-06,
+      "loss": 1.6813,
+      "step": 7759
+    },
+    {
+      "epoch": 0.8060662719434922,
+      "grad_norm": 0.48558080196380615,
+      "learning_rate": 8.9964364196234e-06,
+      "loss": 1.8797,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8061701464630726,
+      "grad_norm": 0.4854564368724823,
+      "learning_rate": 8.987101259093955e-06,
+      "loss": 1.6896,
+      "step": 7761
+    },
+    {
+      "epoch": 0.806274020982653,
+      "grad_norm": 0.4403029680252075,
+      "learning_rate": 8.977770466118313e-06,
+      "loss": 1.574,
+      "step": 7762
+    },
+    {
+      "epoch": 0.8063778955022333,
+      "grad_norm": 0.40083712339401245,
+      "learning_rate": 8.968444041690132e-06,
+      "loss": 1.5353,
+      "step": 7763
+    },
+    {
+      "epoch": 0.8064817700218136,
+      "grad_norm": 0.41334569454193115,
+      "learning_rate": 8.959121986802599e-06,
+      "loss": 1.494,
+      "step": 7764
+    },
+    {
+      "epoch": 0.806585644541394,
+      "grad_norm": 0.4323657751083374,
+      "learning_rate": 8.949804302448433e-06,
+      "loss": 1.6712,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8066895190609743,
+      "grad_norm": 0.4492452144622803,
+      "learning_rate": 8.940490989619932e-06,
+      "loss": 1.6931,
+      "step": 7766
+    },
+    {
+      "epoch": 0.8067933935805547,
+      "grad_norm": 0.39172980189323425,
+      "learning_rate": 8.931182049308833e-06,
+      "loss": 1.616,
+      "step": 7767
+    },
+    {
+      "epoch": 0.8068972681001351,
+      "grad_norm": 0.40300890803337097,
+      "learning_rate": 8.921877482506518e-06,
+      "loss": 1.5328,
+      "step": 7768
+    },
+    {
+      "epoch": 0.8070011426197153,
+      "grad_norm": 0.38939663767814636,
+      "learning_rate": 8.912577290203822e-06,
+      "loss": 1.5393,
+      "step": 7769
+    },
+    {
+      "epoch": 0.8071050171392957,
+      "grad_norm": 0.39137062430381775,
+      "learning_rate": 8.90328147339115e-06,
+      "loss": 1.5114,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8072088916588761,
+      "grad_norm": 0.39022791385650635,
+      "learning_rate": 8.893990033058442e-06,
+      "loss": 1.6231,
+      "step": 7771
+    },
+    {
+      "epoch": 0.8073127661784564,
+      "grad_norm": 0.4040229320526123,
+      "learning_rate": 8.884702970195152e-06,
+      "loss": 1.6152,
+      "step": 7772
+    },
+    {
+      "epoch": 0.8074166406980368,
+      "grad_norm": 0.45910122990608215,
+      "learning_rate": 8.875420285790281e-06,
+      "loss": 1.7177,
+      "step": 7773
+    },
+    {
+      "epoch": 0.8075205152176171,
+      "grad_norm": 0.4154045879840851,
+      "learning_rate": 8.866141980832354e-06,
+      "loss": 1.6195,
+      "step": 7774
+    },
+    {
+      "epoch": 0.8076243897371974,
+      "grad_norm": 0.397510826587677,
+      "learning_rate": 8.856868056309475e-06,
+      "loss": 1.4594,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8077282642567778,
+      "grad_norm": 0.48494023084640503,
+      "learning_rate": 8.84759851320921e-06,
+      "loss": 1.7521,
+      "step": 7776
+    },
+    {
+      "epoch": 0.8078321387763582,
+      "grad_norm": 0.4340880811214447,
+      "learning_rate": 8.838333352518685e-06,
+      "loss": 1.5742,
+      "step": 7777
+    },
+    {
+      "epoch": 0.8079360132959386,
+      "grad_norm": 0.46240857243537903,
+      "learning_rate": 8.829072575224601e-06,
+      "loss": 1.6479,
+      "step": 7778
+    },
+    {
+      "epoch": 0.8080398878155188,
+      "grad_norm": 0.4113903343677521,
+      "learning_rate": 8.819816182313139e-06,
+      "loss": 1.4088,
+      "step": 7779
+    },
+    {
+      "epoch": 0.8081437623350992,
+      "grad_norm": 0.4529806077480316,
+      "learning_rate": 8.810564174770043e-06,
+      "loss": 1.5949,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8082476368546796,
+      "grad_norm": 0.39212316274642944,
+      "learning_rate": 8.801316553580551e-06,
+      "loss": 1.6899,
+      "step": 7781
+    },
+    {
+      "epoch": 0.8083515113742599,
+      "grad_norm": 0.3910139501094818,
+      "learning_rate": 8.792073319729516e-06,
+      "loss": 1.5462,
+      "step": 7782
+    },
+    {
+      "epoch": 0.8084553858938403,
+      "grad_norm": 0.40568485856056213,
+      "learning_rate": 8.782834474201217e-06,
+      "loss": 1.8192,
+      "step": 7783
+    },
+    {
+      "epoch": 0.8085592604134206,
+      "grad_norm": 0.4968739151954651,
+      "learning_rate": 8.773600017979538e-06,
+      "loss": 1.5268,
+      "step": 7784
+    },
+    {
+      "epoch": 0.8086631349330009,
+      "grad_norm": 0.46858468651771545,
+      "learning_rate": 8.764369952047896e-06,
+      "loss": 1.7217,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8087670094525813,
+      "grad_norm": 0.4213848412036896,
+      "learning_rate": 8.755144277389183e-06,
+      "loss": 1.7892,
+      "step": 7786
+    },
+    {
+      "epoch": 0.8088708839721617,
+      "grad_norm": 0.38827750086784363,
+      "learning_rate": 8.745922994985895e-06,
+      "loss": 1.5297,
+      "step": 7787
+    },
+    {
+      "epoch": 0.8089747584917419,
+      "grad_norm": 0.45694899559020996,
+      "learning_rate": 8.736706105820008e-06,
+      "loss": 1.7274,
+      "step": 7788
+    },
+    {
+      "epoch": 0.8090786330113223,
+      "grad_norm": 0.40677914023399353,
+      "learning_rate": 8.727493610873062e-06,
+      "loss": 1.597,
+      "step": 7789
+    },
+    {
+      "epoch": 0.8091825075309027,
+      "grad_norm": 0.4284796714782715,
+      "learning_rate": 8.718285511126106e-06,
+      "loss": 1.7429,
+      "step": 7790
+    },
+    {
+      "epoch": 0.809286382050483,
+      "grad_norm": 0.4217987656593323,
+      "learning_rate": 8.709081807559732e-06,
+      "loss": 1.6746,
+      "step": 7791
+    },
+    {
+      "epoch": 0.8093902565700634,
+      "grad_norm": 0.4426036775112152,
+      "learning_rate": 8.699882501154067e-06,
+      "loss": 1.7955,
+      "step": 7792
+    },
+    {
+      "epoch": 0.8094941310896437,
+      "grad_norm": 0.43109604716300964,
+      "learning_rate": 8.690687592888747e-06,
+      "loss": 1.741,
+      "step": 7793
+    },
+    {
+      "epoch": 0.809598005609224,
+      "grad_norm": 0.4453687369823456,
+      "learning_rate": 8.681497083742995e-06,
+      "loss": 1.5883,
+      "step": 7794
+    },
+    {
+      "epoch": 0.8097018801288044,
+      "grad_norm": 0.47059452533721924,
+      "learning_rate": 8.672310974695485e-06,
+      "loss": 1.7313,
+      "step": 7795
+    },
+    {
+      "epoch": 0.8098057546483848,
+      "grad_norm": 0.39277252554893494,
+      "learning_rate": 8.663129266724501e-06,
+      "loss": 1.6054,
+      "step": 7796
+    },
+    {
+      "epoch": 0.809909629167965,
+      "grad_norm": 0.47614243626594543,
+      "learning_rate": 8.65395196080781e-06,
+      "loss": 1.9036,
+      "step": 7797
+    },
+    {
+      "epoch": 0.8100135036875454,
+      "grad_norm": 0.43281370401382446,
+      "learning_rate": 8.644779057922725e-06,
+      "loss": 1.7245,
+      "step": 7798
+    },
+    {
+      "epoch": 0.8101173782071258,
+      "grad_norm": 0.41983285546302795,
+      "learning_rate": 8.635610559046087e-06,
+      "loss": 1.6524,
+      "step": 7799
+    },
+    {
+      "epoch": 0.8102212527267061,
+      "grad_norm": 0.41546353697776794,
+      "learning_rate": 8.62644646515427e-06,
+      "loss": 1.6953,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8103251272462865,
+      "grad_norm": 0.3985000252723694,
+      "learning_rate": 8.617286777223177e-06,
+      "loss": 1.5696,
+      "step": 7801
+    },
+    {
+      "epoch": 0.8104290017658669,
+      "grad_norm": 0.4149910509586334,
+      "learning_rate": 8.608131496228228e-06,
+      "loss": 1.5954,
+      "step": 7802
+    },
+    {
+      "epoch": 0.8105328762854472,
+      "grad_norm": 0.5904526710510254,
+      "learning_rate": 8.598980623144427e-06,
+      "loss": 2.0503,
+      "step": 7803
+    },
+    {
+      "epoch": 0.8106367508050275,
+      "grad_norm": 0.44825097918510437,
+      "learning_rate": 8.589834158946253e-06,
+      "loss": 1.8498,
+      "step": 7804
+    },
+    {
+      "epoch": 0.8107406253246079,
+      "grad_norm": 0.4529954791069031,
+      "learning_rate": 8.580692104607712e-06,
+      "loss": 1.6659,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8108444998441883,
+      "grad_norm": 0.41834184527397156,
+      "learning_rate": 8.571554461102389e-06,
+      "loss": 1.625,
+      "step": 7806
+    },
+    {
+      "epoch": 0.8109483743637685,
+      "grad_norm": 0.4493601620197296,
+      "learning_rate": 8.562421229403356e-06,
+      "loss": 1.7467,
+      "step": 7807
+    },
+    {
+      "epoch": 0.8110522488833489,
+      "grad_norm": 0.4033777117729187,
+      "learning_rate": 8.553292410483243e-06,
+      "loss": 1.6585,
+      "step": 7808
+    },
+    {
+      "epoch": 0.8111561234029293,
+      "grad_norm": 0.4082568883895874,
+      "learning_rate": 8.544168005314191e-06,
+      "loss": 1.6903,
+      "step": 7809
+    },
+    {
+      "epoch": 0.8112599979225096,
+      "grad_norm": 0.44599470496177673,
+      "learning_rate": 8.535048014867875e-06,
+      "loss": 1.6525,
+      "step": 7810
+    },
+    {
+      "epoch": 0.81136387244209,
+      "grad_norm": 0.38098323345184326,
+      "learning_rate": 8.525932440115509e-06,
+      "loss": 1.4948,
+      "step": 7811
+    },
+    {
+      "epoch": 0.8114677469616703,
+      "grad_norm": 0.4076872766017914,
+      "learning_rate": 8.516821282027814e-06,
+      "loss": 1.5873,
+      "step": 7812
+    },
+    {
+      "epoch": 0.8115716214812506,
+      "grad_norm": 0.40935343503952026,
+      "learning_rate": 8.507714541575096e-06,
+      "loss": 1.5856,
+      "step": 7813
+    },
+    {
+      "epoch": 0.811675496000831,
+      "grad_norm": 0.3904459476470947,
+      "learning_rate": 8.498612219727104e-06,
+      "loss": 1.6061,
+      "step": 7814
+    },
+    {
+      "epoch": 0.8117793705204114,
+      "grad_norm": 0.41350409388542175,
+      "learning_rate": 8.4895143174532e-06,
+      "loss": 1.6078,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8118832450399917,
+      "grad_norm": 0.4305771589279175,
+      "learning_rate": 8.480420835722224e-06,
+      "loss": 1.6971,
+      "step": 7816
+    },
+    {
+      "epoch": 0.811987119559572,
+      "grad_norm": 0.43974366784095764,
+      "learning_rate": 8.471331775502566e-06,
+      "loss": 1.7409,
+      "step": 7817
+    },
+    {
+      "epoch": 0.8120909940791524,
+      "grad_norm": 0.40193986892700195,
+      "learning_rate": 8.462247137762142e-06,
+      "loss": 1.6476,
+      "step": 7818
+    },
+    {
+      "epoch": 0.8121948685987327,
+      "grad_norm": 0.42793333530426025,
+      "learning_rate": 8.453166923468387e-06,
+      "loss": 1.6937,
+      "step": 7819
+    },
+    {
+      "epoch": 0.8122987431183131,
+      "grad_norm": 0.46195828914642334,
+      "learning_rate": 8.444091133588273e-06,
+      "loss": 1.7928,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8124026176378935,
+      "grad_norm": 0.45351237058639526,
+      "learning_rate": 8.435019769088298e-06,
+      "loss": 1.6076,
+      "step": 7821
+    },
+    {
+      "epoch": 0.8125064921574737,
+      "grad_norm": 0.41227513551712036,
+      "learning_rate": 8.42595283093452e-06,
+      "loss": 1.8033,
+      "step": 7822
+    },
+    {
+      "epoch": 0.8126103666770541,
+      "grad_norm": 0.4189743101596832,
+      "learning_rate": 8.416890320092451e-06,
+      "loss": 1.613,
+      "step": 7823
+    },
+    {
+      "epoch": 0.8127142411966345,
+      "grad_norm": 0.44498223066329956,
+      "learning_rate": 8.407832237527214e-06,
+      "loss": 1.8238,
+      "step": 7824
+    },
+    {
+      "epoch": 0.8128181157162148,
+      "grad_norm": 0.4206719994544983,
+      "learning_rate": 8.398778584203414e-06,
+      "loss": 1.6164,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8129219902357951,
+      "grad_norm": 0.4327350854873657,
+      "learning_rate": 8.389729361085192e-06,
+      "loss": 1.6903,
+      "step": 7826
+    },
+    {
+      "epoch": 0.8130258647553755,
+      "grad_norm": 0.4112006425857544,
+      "learning_rate": 8.380684569136222e-06,
+      "loss": 1.5846,
+      "step": 7827
+    },
+    {
+      "epoch": 0.8131297392749558,
+      "grad_norm": 0.4160413444042206,
+      "learning_rate": 8.371644209319702e-06,
+      "loss": 1.4752,
+      "step": 7828
+    },
+    {
+      "epoch": 0.8132336137945362,
+      "grad_norm": 0.4746900796890259,
+      "learning_rate": 8.362608282598355e-06,
+      "loss": 1.6837,
+      "step": 7829
+    },
+    {
+      "epoch": 0.8133374883141166,
+      "grad_norm": 0.4345802664756775,
+      "learning_rate": 8.353576789934436e-06,
+      "loss": 1.6555,
+      "step": 7830
+    },
+    {
+      "epoch": 0.813441362833697,
+      "grad_norm": 0.43879222869873047,
+      "learning_rate": 8.344549732289741e-06,
+      "loss": 1.7406,
+      "step": 7831
+    },
+    {
+      "epoch": 0.8135452373532772,
+      "grad_norm": 0.41482990980148315,
+      "learning_rate": 8.33552711062559e-06,
+      "loss": 1.7404,
+      "step": 7832
+    },
+    {
+      "epoch": 0.8136491118728576,
+      "grad_norm": 0.4466431736946106,
+      "learning_rate": 8.326508925902781e-06,
+      "loss": 1.4813,
+      "step": 7833
+    },
+    {
+      "epoch": 0.813752986392438,
+      "grad_norm": 0.38617411255836487,
+      "learning_rate": 8.31749517908172e-06,
+      "loss": 1.3518,
+      "step": 7834
+    },
+    {
+      "epoch": 0.8138568609120183,
+      "grad_norm": 0.449827641248703,
+      "learning_rate": 8.308485871122284e-06,
+      "loss": 1.723,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8139607354315986,
+      "grad_norm": 0.4846034348011017,
+      "learning_rate": 8.2994810029839e-06,
+      "loss": 1.8044,
+      "step": 7836
+    },
+    {
+      "epoch": 0.814064609951179,
+      "grad_norm": 0.4224514663219452,
+      "learning_rate": 8.290480575625508e-06,
+      "loss": 1.6315,
+      "step": 7837
+    },
+    {
+      "epoch": 0.8141684844707593,
+      "grad_norm": 0.4300024211406708,
+      "learning_rate": 8.28148459000559e-06,
+      "loss": 1.6415,
+      "step": 7838
+    },
+    {
+      "epoch": 0.8142723589903397,
+      "grad_norm": 0.4380504786968231,
+      "learning_rate": 8.272493047082147e-06,
+      "loss": 1.7884,
+      "step": 7839
+    },
+    {
+      "epoch": 0.8143762335099201,
+      "grad_norm": 0.4354472756385803,
+      "learning_rate": 8.263505947812694e-06,
+      "loss": 1.6999,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8144801080295003,
+      "grad_norm": 0.42169252038002014,
+      "learning_rate": 8.254523293154325e-06,
+      "loss": 1.7044,
+      "step": 7841
+    },
+    {
+      "epoch": 0.8145839825490807,
+      "grad_norm": 0.4186304211616516,
+      "learning_rate": 8.245545084063572e-06,
+      "loss": 1.6382,
+      "step": 7842
+    },
+    {
+      "epoch": 0.8146878570686611,
+      "grad_norm": 0.41696295142173767,
+      "learning_rate": 8.236571321496584e-06,
+      "loss": 1.657,
+      "step": 7843
+    },
+    {
+      "epoch": 0.8147917315882414,
+      "grad_norm": 0.40641024708747864,
+      "learning_rate": 8.227602006408986e-06,
+      "loss": 1.6226,
+      "step": 7844
+    },
+    {
+      "epoch": 0.8148956061078217,
+      "grad_norm": 0.42552778124809265,
+      "learning_rate": 8.218637139755931e-06,
+      "loss": 1.6346,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8149994806274021,
+      "grad_norm": 0.4113791584968567,
+      "learning_rate": 8.209676722492116e-06,
+      "loss": 1.6784,
+      "step": 7846
+    },
+    {
+      "epoch": 0.8151033551469824,
+      "grad_norm": 0.4235158860683441,
+      "learning_rate": 8.200720755571755e-06,
+      "loss": 1.6905,
+      "step": 7847
+    },
+    {
+      "epoch": 0.8152072296665628,
+      "grad_norm": 0.47491201758384705,
+      "learning_rate": 8.191769239948588e-06,
+      "loss": 1.8831,
+      "step": 7848
+    },
+    {
+      "epoch": 0.8153111041861432,
+      "grad_norm": 0.4474998712539673,
+      "learning_rate": 8.182822176575865e-06,
+      "loss": 1.6919,
+      "step": 7849
+    },
+    {
+      "epoch": 0.8154149787057234,
+      "grad_norm": 0.43246933817863464,
+      "learning_rate": 8.173879566406418e-06,
+      "loss": 1.6247,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8155188532253038,
+      "grad_norm": 0.4082406461238861,
+      "learning_rate": 8.164941410392519e-06,
+      "loss": 1.662,
+      "step": 7851
+    },
+    {
+      "epoch": 0.8156227277448842,
+      "grad_norm": 0.3958854079246521,
+      "learning_rate": 8.15600770948604e-06,
+      "loss": 1.5751,
+      "step": 7852
+    },
+    {
+      "epoch": 0.8157266022644645,
+      "grad_norm": 0.4203862249851227,
+      "learning_rate": 8.147078464638346e-06,
+      "loss": 1.5519,
+      "step": 7853
+    },
+    {
+      "epoch": 0.8158304767840449,
+      "grad_norm": 0.5035269260406494,
+      "learning_rate": 8.138153676800336e-06,
+      "loss": 1.7116,
+      "step": 7854
+    },
+    {
+      "epoch": 0.8159343513036252,
+      "grad_norm": 0.4069529175758362,
+      "learning_rate": 8.129233346922422e-06,
+      "loss": 1.5981,
+      "step": 7855
+    },
+    {
+      "epoch": 0.8160382258232056,
+      "grad_norm": 0.44691234827041626,
+      "learning_rate": 8.120317475954536e-06,
+      "loss": 1.7425,
+      "step": 7856
+    },
+    {
+      "epoch": 0.8161421003427859,
+      "grad_norm": 0.4124663770198822,
+      "learning_rate": 8.111406064846194e-06,
+      "loss": 1.5632,
+      "step": 7857
+    },
+    {
+      "epoch": 0.8162459748623663,
+      "grad_norm": 0.4278232157230377,
+      "learning_rate": 8.102499114546335e-06,
+      "loss": 1.5863,
+      "step": 7858
+    },
+    {
+      "epoch": 0.8163498493819467,
+      "grad_norm": 0.4516265392303467,
+      "learning_rate": 8.093596626003519e-06,
+      "loss": 1.7937,
+      "step": 7859
+    },
+    {
+      "epoch": 0.8164537239015269,
+      "grad_norm": 0.42853647470474243,
+      "learning_rate": 8.084698600165796e-06,
+      "loss": 1.745,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8165575984211073,
+      "grad_norm": 0.3971802294254303,
+      "learning_rate": 8.075805037980688e-06,
+      "loss": 1.528,
+      "step": 7861
+    },
+    {
+      "epoch": 0.8166614729406877,
+      "grad_norm": 0.4206622242927551,
+      "learning_rate": 8.066915940395341e-06,
+      "loss": 1.6715,
+      "step": 7862
+    },
+    {
+      "epoch": 0.816765347460268,
+      "grad_norm": 0.4111405313014984,
+      "learning_rate": 8.058031308356357e-06,
+      "loss": 1.5871,
+      "step": 7863
+    },
+    {
+      "epoch": 0.8168692219798483,
+      "grad_norm": 0.5833050608634949,
+      "learning_rate": 8.049151142809874e-06,
+      "loss": 1.8095,
+      "step": 7864
+    },
+    {
+      "epoch": 0.8169730964994287,
+      "grad_norm": 0.42305901646614075,
+      "learning_rate": 8.040275444701567e-06,
+      "loss": 1.712,
+      "step": 7865
+    },
+    {
+      "epoch": 0.817076971019009,
+      "grad_norm": 0.46678563952445984,
+      "learning_rate": 8.031404214976628e-06,
+      "loss": 1.8676,
+      "step": 7866
+    },
+    {
+      "epoch": 0.8171808455385894,
+      "grad_norm": 0.45080798864364624,
+      "learning_rate": 8.02253745457977e-06,
+      "loss": 1.5683,
+      "step": 7867
+    },
+    {
+      "epoch": 0.8172847200581698,
+      "grad_norm": 0.41853803396224976,
+      "learning_rate": 8.013675164455225e-06,
+      "loss": 1.6236,
+      "step": 7868
+    },
+    {
+      "epoch": 0.81738859457775,
+      "grad_norm": 0.4322271943092346,
+      "learning_rate": 8.004817345546794e-06,
+      "loss": 1.677,
+      "step": 7869
+    },
+    {
+      "epoch": 0.8174924690973304,
+      "grad_norm": 0.4160740077495575,
+      "learning_rate": 7.995963998797717e-06,
+      "loss": 1.6949,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8175963436169108,
+      "grad_norm": 0.4284444749355316,
+      "learning_rate": 7.98711512515084e-06,
+      "loss": 1.6774,
+      "step": 7871
+    },
+    {
+      "epoch": 0.8177002181364911,
+      "grad_norm": 0.3733249306678772,
+      "learning_rate": 7.978270725548493e-06,
+      "loss": 1.4368,
+      "step": 7872
+    },
+    {
+      "epoch": 0.8178040926560715,
+      "grad_norm": 0.4796290695667267,
+      "learning_rate": 7.969430800932526e-06,
+      "loss": 1.6731,
+      "step": 7873
+    },
+    {
+      "epoch": 0.8179079671756518,
+      "grad_norm": 0.4326523542404175,
+      "learning_rate": 7.960595352244332e-06,
+      "loss": 1.7251,
+      "step": 7874
+    },
+    {
+      "epoch": 0.8180118416952321,
+      "grad_norm": 0.4698193669319153,
+      "learning_rate": 7.951764380424808e-06,
+      "loss": 1.8268,
+      "step": 7875
+    },
+    {
+      "epoch": 0.8181157162148125,
+      "grad_norm": 0.42041996121406555,
+      "learning_rate": 7.942937886414392e-06,
+      "loss": 1.4004,
+      "step": 7876
+    },
+    {
+      "epoch": 0.8182195907343929,
+      "grad_norm": 0.4212832450866699,
+      "learning_rate": 7.934115871153025e-06,
+      "loss": 1.6694,
+      "step": 7877
+    },
+    {
+      "epoch": 0.8183234652539731,
+      "grad_norm": 0.39500170946121216,
+      "learning_rate": 7.925298335580211e-06,
+      "loss": 1.5698,
+      "step": 7878
+    },
+    {
+      "epoch": 0.8184273397735535,
+      "grad_norm": 0.4193446934223175,
+      "learning_rate": 7.916485280634906e-06,
+      "loss": 1.6886,
+      "step": 7879
+    },
+    {
+      "epoch": 0.8185312142931339,
+      "grad_norm": 0.44972357153892517,
+      "learning_rate": 7.907676707255668e-06,
+      "loss": 1.8019,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8186350888127143,
+      "grad_norm": 0.42009323835372925,
+      "learning_rate": 7.898872616380526e-06,
+      "loss": 1.5613,
+      "step": 7881
+    },
+    {
+      "epoch": 0.8187389633322946,
+      "grad_norm": 0.4435664713382721,
+      "learning_rate": 7.890073008947047e-06,
+      "loss": 1.6753,
+      "step": 7882
+    },
+    {
+      "epoch": 0.818842837851875,
+      "grad_norm": 0.40186628699302673,
+      "learning_rate": 7.881277885892324e-06,
+      "loss": 1.4869,
+      "step": 7883
+    },
+    {
+      "epoch": 0.8189467123714553,
+      "grad_norm": 0.40945637226104736,
+      "learning_rate": 7.87248724815296e-06,
+      "loss": 1.641,
+      "step": 7884
+    },
+    {
+      "epoch": 0.8190505868910356,
+      "grad_norm": 0.42431339621543884,
+      "learning_rate": 7.863701096665116e-06,
+      "loss": 1.708,
+      "step": 7885
+    },
+    {
+      "epoch": 0.819154461410616,
+      "grad_norm": 0.3984614610671997,
+      "learning_rate": 7.854919432364406e-06,
+      "loss": 1.545,
+      "step": 7886
+    },
+    {
+      "epoch": 0.8192583359301964,
+      "grad_norm": 0.42371925711631775,
+      "learning_rate": 7.846142256186046e-06,
+      "loss": 1.6084,
+      "step": 7887
+    },
+    {
+      "epoch": 0.8193622104497766,
+      "grad_norm": 0.40940725803375244,
+      "learning_rate": 7.837369569064734e-06,
+      "loss": 1.4237,
+      "step": 7888
+    },
+    {
+      "epoch": 0.819466084969357,
+      "grad_norm": 0.4526590406894684,
+      "learning_rate": 7.828601371934657e-06,
+      "loss": 1.7202,
+      "step": 7889
+    },
+    {
+      "epoch": 0.8195699594889374,
+      "grad_norm": 0.4891595244407654,
+      "learning_rate": 7.819837665729596e-06,
+      "loss": 1.8862,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8196738340085177,
+      "grad_norm": 0.44362369179725647,
+      "learning_rate": 7.811078451382808e-06,
+      "loss": 1.776,
+      "step": 7891
+    },
+    {
+      "epoch": 0.8197777085280981,
+      "grad_norm": 0.46131572127342224,
+      "learning_rate": 7.802323729827077e-06,
+      "loss": 1.7542,
+      "step": 7892
+    },
+    {
+      "epoch": 0.8198815830476784,
+      "grad_norm": 0.45027032494544983,
+      "learning_rate": 7.79357350199471e-06,
+      "loss": 1.3346,
+      "step": 7893
+    },
+    {
+      "epoch": 0.8199854575672587,
+      "grad_norm": 0.41758599877357483,
+      "learning_rate": 7.784827768817548e-06,
+      "loss": 1.6569,
+      "step": 7894
+    },
+    {
+      "epoch": 0.8200893320868391,
+      "grad_norm": 0.3943220376968384,
+      "learning_rate": 7.77608653122694e-06,
+      "loss": 1.5361,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8201932066064195,
+      "grad_norm": 0.40477508306503296,
+      "learning_rate": 7.767349790153738e-06,
+      "loss": 1.5803,
+      "step": 7896
+    },
+    {
+      "epoch": 0.8202970811259997,
+      "grad_norm": 0.4200827479362488,
+      "learning_rate": 7.758617546528386e-06,
+      "loss": 1.6642,
+      "step": 7897
+    },
+    {
+      "epoch": 0.8204009556455801,
+      "grad_norm": 0.4615240693092346,
+      "learning_rate": 7.74988980128074e-06,
+      "loss": 1.8093,
+      "step": 7898
+    },
+    {
+      "epoch": 0.8205048301651605,
+      "grad_norm": 0.42218905687332153,
+      "learning_rate": 7.741166555340284e-06,
+      "loss": 1.7515,
+      "step": 7899
+    },
+    {
+      "epoch": 0.8206087046847408,
+      "grad_norm": 0.4505891799926758,
+      "learning_rate": 7.732447809635956e-06,
+      "loss": 1.6826,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8207125792043212,
+      "grad_norm": 0.4403258264064789,
+      "learning_rate": 7.723733565096236e-06,
+      "loss": 1.6249,
+      "step": 7901
+    },
+    {
+      "epoch": 0.8208164537239016,
+      "grad_norm": 0.4211053252220154,
+      "learning_rate": 7.715023822649126e-06,
+      "loss": 1.563,
+      "step": 7902
+    },
+    {
+      "epoch": 0.8209203282434818,
+      "grad_norm": 0.4513636529445648,
+      "learning_rate": 7.706318583222144e-06,
+      "loss": 1.787,
+      "step": 7903
+    },
+    {
+      "epoch": 0.8210242027630622,
+      "grad_norm": 0.3969414532184601,
+      "learning_rate": 7.697617847742328e-06,
+      "loss": 1.6397,
+      "step": 7904
+    },
+    {
+      "epoch": 0.8211280772826426,
+      "grad_norm": 0.43942591547966003,
+      "learning_rate": 7.688921617136224e-06,
+      "loss": 1.6617,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8212319518022229,
+      "grad_norm": 0.3998488187789917,
+      "learning_rate": 7.68022989232996e-06,
+      "loss": 1.5171,
+      "step": 7906
+    },
+    {
+      "epoch": 0.8213358263218032,
+      "grad_norm": 0.44349002838134766,
+      "learning_rate": 7.671542674249077e-06,
+      "loss": 1.7494,
+      "step": 7907
+    },
+    {
+      "epoch": 0.8214397008413836,
+      "grad_norm": 0.4117303788661957,
+      "learning_rate": 7.662859963818737e-06,
+      "loss": 1.6006,
+      "step": 7908
+    },
+    {
+      "epoch": 0.821543575360964,
+      "grad_norm": 0.4332650303840637,
+      "learning_rate": 7.65418176196357e-06,
+      "loss": 1.4621,
+      "step": 7909
+    },
+    {
+      "epoch": 0.8216474498805443,
+      "grad_norm": 0.3984694182872772,
+      "learning_rate": 7.645508069607731e-06,
+      "loss": 1.5073,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8217513244001247,
+      "grad_norm": 0.477058082818985,
+      "learning_rate": 7.636838887674908e-06,
+      "loss": 1.9018,
+      "step": 7911
+    },
+    {
+      "epoch": 0.821855198919705,
+      "grad_norm": 0.4433257281780243,
+      "learning_rate": 7.6281742170882854e-06,
+      "loss": 1.6646,
+      "step": 7912
+    },
+    {
+      "epoch": 0.8219590734392853,
+      "grad_norm": 0.4137663245201111,
+      "learning_rate": 7.619514058770622e-06,
+      "loss": 1.482,
+      "step": 7913
+    },
+    {
+      "epoch": 0.8220629479588657,
+      "grad_norm": 0.4989883005619049,
+      "learning_rate": 7.610858413644106e-06,
+      "loss": 1.7999,
+      "step": 7914
+    },
+    {
+      "epoch": 0.8221668224784461,
+      "grad_norm": 0.4211665391921997,
+      "learning_rate": 7.60220728263053e-06,
+      "loss": 1.6357,
+      "step": 7915
+    },
+    {
+      "epoch": 0.8222706969980264,
+      "grad_norm": 0.4524006247520447,
+      "learning_rate": 7.5935606666511785e-06,
+      "loss": 1.7431,
+      "step": 7916
+    },
+    {
+      "epoch": 0.8223745715176067,
+      "grad_norm": 0.42954227328300476,
+      "learning_rate": 7.584918566626808e-06,
+      "loss": 1.8379,
+      "step": 7917
+    },
+    {
+      "epoch": 0.8224784460371871,
+      "grad_norm": 0.4114231467247009,
+      "learning_rate": 7.576280983477768e-06,
+      "loss": 1.7044,
+      "step": 7918
+    },
+    {
+      "epoch": 0.8225823205567674,
+      "grad_norm": 0.43601930141448975,
+      "learning_rate": 7.5676479181238935e-06,
+      "loss": 1.6908,
+      "step": 7919
+    },
+    {
+      "epoch": 0.8226861950763478,
+      "grad_norm": 0.42451170086860657,
+      "learning_rate": 7.5590193714845215e-06,
+      "loss": 1.7343,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8227900695959282,
+      "grad_norm": 0.4314807653427124,
+      "learning_rate": 7.5503953444785395e-06,
+      "loss": 1.6858,
+      "step": 7921
+    },
+    {
+      "epoch": 0.8228939441155084,
+      "grad_norm": 0.41289034485816956,
+      "learning_rate": 7.541775838024335e-06,
+      "loss": 1.4288,
+      "step": 7922
+    },
+    {
+      "epoch": 0.8229978186350888,
+      "grad_norm": 0.4084930419921875,
+      "learning_rate": 7.533160853039811e-06,
+      "loss": 1.6902,
+      "step": 7923
+    },
+    {
+      "epoch": 0.8231016931546692,
+      "grad_norm": 0.4590340852737427,
+      "learning_rate": 7.5245503904423974e-06,
+      "loss": 1.6764,
+      "step": 7924
+    },
+    {
+      "epoch": 0.8232055676742495,
+      "grad_norm": 0.46251264214515686,
+      "learning_rate": 7.515944451149065e-06,
+      "loss": 1.5257,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8233094421938298,
+      "grad_norm": 0.4301720857620239,
+      "learning_rate": 7.507343036076236e-06,
+      "loss": 1.7414,
+      "step": 7926
+    },
+    {
+      "epoch": 0.8234133167134102,
+      "grad_norm": 0.42009609937667847,
+      "learning_rate": 7.498746146139934e-06,
+      "loss": 1.6366,
+      "step": 7927
+    },
+    {
+      "epoch": 0.8235171912329905,
+      "grad_norm": 0.40405720472335815,
+      "learning_rate": 7.490153782255643e-06,
+      "loss": 1.5959,
+      "step": 7928
+    },
+    {
+      "epoch": 0.8236210657525709,
+      "grad_norm": 0.4580763280391693,
+      "learning_rate": 7.481565945338381e-06,
+      "loss": 1.8361,
+      "step": 7929
+    },
+    {
+      "epoch": 0.8237249402721513,
+      "grad_norm": 0.43223121762275696,
+      "learning_rate": 7.472982636302694e-06,
+      "loss": 1.592,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8238288147917315,
+      "grad_norm": 0.48288586735725403,
+      "learning_rate": 7.464403856062629e-06,
+      "loss": 1.7912,
+      "step": 7931
+    },
+    {
+      "epoch": 0.8239326893113119,
+      "grad_norm": 0.4333949685096741,
+      "learning_rate": 7.455829605531767e-06,
+      "loss": 1.7699,
+      "step": 7932
+    },
+    {
+      "epoch": 0.8240365638308923,
+      "grad_norm": 0.40936028957366943,
+      "learning_rate": 7.4472598856231815e-06,
+      "loss": 1.5568,
+      "step": 7933
+    },
+    {
+      "epoch": 0.8241404383504727,
+      "grad_norm": 0.4204612672328949,
+      "learning_rate": 7.438694697249504e-06,
+      "loss": 1.5514,
+      "step": 7934
+    },
+    {
+      "epoch": 0.824244312870053,
+      "grad_norm": 0.4096861183643341,
+      "learning_rate": 7.430134041322856e-06,
+      "loss": 1.6633,
+      "step": 7935
+    },
+    {
+      "epoch": 0.8243481873896333,
+      "grad_norm": 0.4757747948169708,
+      "learning_rate": 7.421577918754868e-06,
+      "loss": 1.8486,
+      "step": 7936
+    },
+    {
+      "epoch": 0.8244520619092137,
+      "grad_norm": 0.44562995433807373,
+      "learning_rate": 7.413026330456713e-06,
+      "loss": 1.6866,
+      "step": 7937
+    },
+    {
+      "epoch": 0.824555936428794,
+      "grad_norm": 0.4388675391674042,
+      "learning_rate": 7.404479277339055e-06,
+      "loss": 1.6298,
+      "step": 7938
+    },
+    {
+      "epoch": 0.8246598109483744,
+      "grad_norm": 0.4150373041629791,
+      "learning_rate": 7.3959367603121055e-06,
+      "loss": 1.6743,
+      "step": 7939
+    },
+    {
+      "epoch": 0.8247636854679548,
+      "grad_norm": 0.40958264470100403,
+      "learning_rate": 7.387398780285548e-06,
+      "loss": 1.5683,
+      "step": 7940
+    },
+    {
+      "epoch": 0.824867559987535,
+      "grad_norm": 0.40461811423301697,
+      "learning_rate": 7.378865338168655e-06,
+      "loss": 1.5644,
+      "step": 7941
+    },
+    {
+      "epoch": 0.8249714345071154,
+      "grad_norm": 0.4974653124809265,
+      "learning_rate": 7.370336434870123e-06,
+      "loss": 1.8986,
+      "step": 7942
+    },
+    {
+      "epoch": 0.8250753090266958,
+      "grad_norm": 0.40290409326553345,
+      "learning_rate": 7.361812071298246e-06,
+      "loss": 1.6532,
+      "step": 7943
+    },
+    {
+      "epoch": 0.8251791835462761,
+      "grad_norm": 0.4083233177661896,
+      "learning_rate": 7.353292248360805e-06,
+      "loss": 1.6532,
+      "step": 7944
+    },
+    {
+      "epoch": 0.8252830580658564,
+      "grad_norm": 0.47831931710243225,
+      "learning_rate": 7.344776966965061e-06,
+      "loss": 1.7635,
+      "step": 7945
+    },
+    {
+      "epoch": 0.8253869325854368,
+      "grad_norm": 0.4501749277114868,
+      "learning_rate": 7.336266228017857e-06,
+      "loss": 1.5736,
+      "step": 7946
+    },
+    {
+      "epoch": 0.8254908071050171,
+      "grad_norm": 0.43580162525177,
+      "learning_rate": 7.327760032425507e-06,
+      "loss": 1.7677,
+      "step": 7947
+    },
+    {
+      "epoch": 0.8255946816245975,
+      "grad_norm": 0.4431309401988983,
+      "learning_rate": 7.319258381093863e-06,
+      "loss": 1.541,
+      "step": 7948
+    },
+    {
+      "epoch": 0.8256985561441779,
+      "grad_norm": 0.46647512912750244,
+      "learning_rate": 7.310761274928274e-06,
+      "loss": 1.6897,
+      "step": 7949
+    },
+    {
+      "epoch": 0.8258024306637581,
+      "grad_norm": 0.4248170554637909,
+      "learning_rate": 7.302268714833621e-06,
+      "loss": 1.6534,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8259063051833385,
+      "grad_norm": 0.461207777261734,
+      "learning_rate": 7.2937807017142975e-06,
+      "loss": 1.4829,
+      "step": 7951
+    },
+    {
+      "epoch": 0.8260101797029189,
+      "grad_norm": 0.42655470967292786,
+      "learning_rate": 7.285297236474198e-06,
+      "loss": 1.7697,
+      "step": 7952
+    },
+    {
+      "epoch": 0.8261140542224992,
+      "grad_norm": 0.40300819277763367,
+      "learning_rate": 7.27681832001677e-06,
+      "loss": 1.4717,
+      "step": 7953
+    },
+    {
+      "epoch": 0.8262179287420796,
+      "grad_norm": 0.4207182228565216,
+      "learning_rate": 7.268343953244921e-06,
+      "loss": 1.5658,
+      "step": 7954
+    },
+    {
+      "epoch": 0.8263218032616599,
+      "grad_norm": 0.516309916973114,
+      "learning_rate": 7.259874137061135e-06,
+      "loss": 1.7525,
+      "step": 7955
+    },
+    {
+      "epoch": 0.8264256777812402,
+      "grad_norm": 0.4138142466545105,
+      "learning_rate": 7.2514088723673605e-06,
+      "loss": 1.6619,
+      "step": 7956
+    },
+    {
+      "epoch": 0.8265295523008206,
+      "grad_norm": 0.4462479054927826,
+      "learning_rate": 7.2429481600650936e-06,
+      "loss": 1.7732,
+      "step": 7957
+    },
+    {
+      "epoch": 0.826633426820401,
+      "grad_norm": 0.46467795968055725,
+      "learning_rate": 7.234492001055332e-06,
+      "loss": 1.762,
+      "step": 7958
+    },
+    {
+      "epoch": 0.8267373013399814,
+      "grad_norm": 0.4307834506034851,
+      "learning_rate": 7.2260403962385706e-06,
+      "loss": 1.5749,
+      "step": 7959
+    },
+    {
+      "epoch": 0.8268411758595616,
+      "grad_norm": 0.43813708424568176,
+      "learning_rate": 7.217593346514884e-06,
+      "loss": 1.7171,
+      "step": 7960
+    },
+    {
+      "epoch": 0.826945050379142,
+      "grad_norm": 0.42140957713127136,
+      "learning_rate": 7.209150852783769e-06,
+      "loss": 1.6369,
+      "step": 7961
+    },
+    {
+      "epoch": 0.8270489248987224,
+      "grad_norm": 0.42076048254966736,
+      "learning_rate": 7.200712915944313e-06,
+      "loss": 1.6839,
+      "step": 7962
+    },
+    {
+      "epoch": 0.8271527994183027,
+      "grad_norm": 0.4501379728317261,
+      "learning_rate": 7.192279536895091e-06,
+      "loss": 1.7524,
+      "step": 7963
+    },
+    {
+      "epoch": 0.827256673937883,
+      "grad_norm": 0.39931780099868774,
+      "learning_rate": 7.183850716534179e-06,
+      "loss": 1.6049,
+      "step": 7964
+    },
+    {
+      "epoch": 0.8273605484574634,
+      "grad_norm": 0.39734897017478943,
+      "learning_rate": 7.175426455759193e-06,
+      "loss": 1.5222,
+      "step": 7965
+    },
+    {
+      "epoch": 0.8274644229770437,
+      "grad_norm": 0.434510201215744,
+      "learning_rate": 7.167006755467237e-06,
+      "loss": 1.7086,
+      "step": 7966
+    },
+    {
+      "epoch": 0.8275682974966241,
+      "grad_norm": 0.42094168066978455,
+      "learning_rate": 7.15859161655496e-06,
+      "loss": 1.7143,
+      "step": 7967
+    },
+    {
+      "epoch": 0.8276721720162045,
+      "grad_norm": 0.4152793884277344,
+      "learning_rate": 7.150181039918485e-06,
+      "loss": 1.3494,
+      "step": 7968
+    },
+    {
+      "epoch": 0.8277760465357847,
+      "grad_norm": 0.4016823470592499,
+      "learning_rate": 7.141775026453506e-06,
+      "loss": 1.5688,
+      "step": 7969
+    },
+    {
+      "epoch": 0.8278799210553651,
+      "grad_norm": 0.45909714698791504,
+      "learning_rate": 7.133373577055163e-06,
+      "loss": 1.7456,
+      "step": 7970
+    },
+    {
+      "epoch": 0.8279837955749455,
+      "grad_norm": 0.4085950255393982,
+      "learning_rate": 7.124976692618168e-06,
+      "loss": 1.5706,
+      "step": 7971
+    },
+    {
+      "epoch": 0.8280876700945258,
+      "grad_norm": 0.419292151927948,
+      "learning_rate": 7.116584374036733e-06,
+      "loss": 1.6361,
+      "step": 7972
+    },
+    {
+      "epoch": 0.8281915446141062,
+      "grad_norm": 0.4057331383228302,
+      "learning_rate": 7.108196622204538e-06,
+      "loss": 1.5935,
+      "step": 7973
+    },
+    {
+      "epoch": 0.8282954191336865,
+      "grad_norm": 0.4193360507488251,
+      "learning_rate": 7.099813438014841e-06,
+      "loss": 1.4914,
+      "step": 7974
+    },
+    {
+      "epoch": 0.8283992936532668,
+      "grad_norm": 0.42038246989250183,
+      "learning_rate": 7.091434822360377e-06,
+      "loss": 1.6914,
+      "step": 7975
+    },
+    {
+      "epoch": 0.8285031681728472,
+      "grad_norm": 0.41948550939559937,
+      "learning_rate": 7.083060776133404e-06,
+      "loss": 1.7179,
+      "step": 7976
+    },
+    {
+      "epoch": 0.8286070426924276,
+      "grad_norm": 0.46123793721199036,
+      "learning_rate": 7.0746913002257e-06,
+      "loss": 1.7011,
+      "step": 7977
+    },
+    {
+      "epoch": 0.8287109172120078,
+      "grad_norm": 0.4170924127101898,
+      "learning_rate": 7.066326395528539e-06,
+      "loss": 1.6866,
+      "step": 7978
+    },
+    {
+      "epoch": 0.8288147917315882,
+      "grad_norm": 0.4177165627479553,
+      "learning_rate": 7.05796606293272e-06,
+      "loss": 1.6695,
+      "step": 7979
+    },
+    {
+      "epoch": 0.8289186662511686,
+      "grad_norm": 0.46339476108551025,
+      "learning_rate": 7.049610303328541e-06,
+      "loss": 1.7386,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8290225407707489,
+      "grad_norm": 0.45544731616973877,
+      "learning_rate": 7.04125911760587e-06,
+      "loss": 1.6829,
+      "step": 7981
+    },
+    {
+      "epoch": 0.8291264152903293,
+      "grad_norm": 0.4197863042354584,
+      "learning_rate": 7.032912506653983e-06,
+      "loss": 1.5823,
+      "step": 7982
+    },
+    {
+      "epoch": 0.8292302898099096,
+      "grad_norm": 0.42813071608543396,
+      "learning_rate": 7.024570471361769e-06,
+      "loss": 1.5883,
+      "step": 7983
+    },
+    {
+      "epoch": 0.82933416432949,
+      "grad_norm": 0.43591973185539246,
+      "learning_rate": 7.016233012617579e-06,
+      "loss": 1.4265,
+      "step": 7984
+    },
+    {
+      "epoch": 0.8294380388490703,
+      "grad_norm": 0.41332176327705383,
+      "learning_rate": 7.0079001313092865e-06,
+      "loss": 1.7404,
+      "step": 7985
+    },
+    {
+      "epoch": 0.8295419133686507,
+      "grad_norm": 0.40137556195259094,
+      "learning_rate": 6.99957182832428e-06,
+      "loss": 1.6176,
+      "step": 7986
+    },
+    {
+      "epoch": 0.8296457878882311,
+      "grad_norm": 0.45006173849105835,
+      "learning_rate": 6.9912481045494475e-06,
+      "loss": 1.6707,
+      "step": 7987
+    },
+    {
+      "epoch": 0.8297496624078113,
+      "grad_norm": 0.4188838005065918,
+      "learning_rate": 6.9829289608712264e-06,
+      "loss": 1.6712,
+      "step": 7988
+    },
+    {
+      "epoch": 0.8298535369273917,
+      "grad_norm": 0.4287783205509186,
+      "learning_rate": 6.974614398175511e-06,
+      "loss": 1.6211,
+      "step": 7989
+    },
+    {
+      "epoch": 0.8299574114469721,
+      "grad_norm": 0.47514259815216064,
+      "learning_rate": 6.966304417347758e-06,
+      "loss": 1.6593,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8300612859665524,
+      "grad_norm": 0.40735095739364624,
+      "learning_rate": 6.95799901927291e-06,
+      "loss": 1.6225,
+      "step": 7991
+    },
+    {
+      "epoch": 0.8301651604861328,
+      "grad_norm": 0.39932069182395935,
+      "learning_rate": 6.949698204835426e-06,
+      "loss": 1.5756,
+      "step": 7992
+    },
+    {
+      "epoch": 0.8302690350057131,
+      "grad_norm": 0.43335703015327454,
+      "learning_rate": 6.941401974919276e-06,
+      "loss": 1.6939,
+      "step": 7993
+    },
+    {
+      "epoch": 0.8303729095252934,
+      "grad_norm": 0.43645966053009033,
+      "learning_rate": 6.9331103304079494e-06,
+      "loss": 1.7476,
+      "step": 7994
+    },
+    {
+      "epoch": 0.8304767840448738,
+      "grad_norm": 0.38558638095855713,
+      "learning_rate": 6.924823272184439e-06,
+      "loss": 1.449,
+      "step": 7995
+    },
+    {
+      "epoch": 0.8305806585644542,
+      "grad_norm": 0.3955431878566742,
+      "learning_rate": 6.916540801131232e-06,
+      "loss": 1.6482,
+      "step": 7996
+    },
+    {
+      "epoch": 0.8306845330840344,
+      "grad_norm": 0.3958597779273987,
+      "learning_rate": 6.908262918130392e-06,
+      "loss": 1.5417,
+      "step": 7997
+    },
+    {
+      "epoch": 0.8307884076036148,
+      "grad_norm": 0.4344055950641632,
+      "learning_rate": 6.899989624063402e-06,
+      "loss": 1.5201,
+      "step": 7998
+    },
+    {
+      "epoch": 0.8308922821231952,
+      "grad_norm": 0.4599206745624542,
+      "learning_rate": 6.8917209198113345e-06,
+      "loss": 1.7178,
+      "step": 7999
+    },
+    {
+      "epoch": 0.8309961566427755,
+      "grad_norm": 0.3974829614162445,
+      "learning_rate": 6.883456806254746e-06,
+      "loss": 1.5518,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8311000311623559,
+      "grad_norm": 0.39066168665885925,
+      "learning_rate": 6.87519728427366e-06,
+      "loss": 1.6246,
+      "step": 8001
+    },
+    {
+      "epoch": 0.8312039056819363,
+      "grad_norm": 0.4144991338253021,
+      "learning_rate": 6.866942354747685e-06,
+      "loss": 1.7103,
+      "step": 8002
+    },
+    {
+      "epoch": 0.8313077802015165,
+      "grad_norm": 0.4020610749721527,
+      "learning_rate": 6.8586920185559015e-06,
+      "loss": 1.5555,
+      "step": 8003
+    },
+    {
+      "epoch": 0.8314116547210969,
+      "grad_norm": 0.3927983045578003,
+      "learning_rate": 6.850446276576905e-06,
+      "loss": 1.5676,
+      "step": 8004
+    },
+    {
+      "epoch": 0.8315155292406773,
+      "grad_norm": 0.4108089804649353,
+      "learning_rate": 6.8422051296888e-06,
+      "loss": 1.5645,
+      "step": 8005
+    },
+    {
+      "epoch": 0.8316194037602576,
+      "grad_norm": 0.38096117973327637,
+      "learning_rate": 6.833968578769201e-06,
+      "loss": 1.4893,
+      "step": 8006
+    },
+    {
+      "epoch": 0.8317232782798379,
+      "grad_norm": 0.43577104806900024,
+      "learning_rate": 6.825736624695245e-06,
+      "loss": 1.5007,
+      "step": 8007
+    },
+    {
+      "epoch": 0.8318271527994183,
+      "grad_norm": 0.40206262469291687,
+      "learning_rate": 6.817509268343553e-06,
+      "loss": 1.3456,
+      "step": 8008
+    },
+    {
+      "epoch": 0.8319310273189986,
+      "grad_norm": 0.4439072906970978,
+      "learning_rate": 6.809286510590307e-06,
+      "loss": 1.5847,
+      "step": 8009
+    },
+    {
+      "epoch": 0.832034901838579,
+      "grad_norm": 0.42791223526000977,
+      "learning_rate": 6.80106835231113e-06,
+      "loss": 1.7015,
+      "step": 8010
+    },
+    {
+      "epoch": 0.8321387763581594,
+      "grad_norm": 0.3975352644920349,
+      "learning_rate": 6.792854794381215e-06,
+      "loss": 1.6057,
+      "step": 8011
+    },
+    {
+      "epoch": 0.8322426508777397,
+      "grad_norm": 0.42390820384025574,
+      "learning_rate": 6.784645837675241e-06,
+      "loss": 1.6555,
+      "step": 8012
+    },
+    {
+      "epoch": 0.83234652539732,
+      "grad_norm": 0.4322831928730011,
+      "learning_rate": 6.776441483067386e-06,
+      "loss": 1.6785,
+      "step": 8013
+    },
+    {
+      "epoch": 0.8324503999169004,
+      "grad_norm": 0.443946897983551,
+      "learning_rate": 6.768241731431363e-06,
+      "loss": 1.6525,
+      "step": 8014
+    },
+    {
+      "epoch": 0.8325542744364808,
+      "grad_norm": 0.4141833186149597,
+      "learning_rate": 6.760046583640362e-06,
+      "loss": 1.6921,
+      "step": 8015
+    },
+    {
+      "epoch": 0.832658148956061,
+      "grad_norm": 0.4575175344944,
+      "learning_rate": 6.751856040567134e-06,
+      "loss": 1.7602,
+      "step": 8016
+    },
+    {
+      "epoch": 0.8327620234756414,
+      "grad_norm": 0.5201651453971863,
+      "learning_rate": 6.7436701030838715e-06,
+      "loss": 1.9422,
+      "step": 8017
+    },
+    {
+      "epoch": 0.8328658979952218,
+      "grad_norm": 0.4121319353580475,
+      "learning_rate": 6.735488772062337e-06,
+      "loss": 1.7875,
+      "step": 8018
+    },
+    {
+      "epoch": 0.8329697725148021,
+      "grad_norm": 0.40832269191741943,
+      "learning_rate": 6.727312048373774e-06,
+      "loss": 1.5723,
+      "step": 8019
+    },
+    {
+      "epoch": 0.8330736470343825,
+      "grad_norm": 0.44910913705825806,
+      "learning_rate": 6.71913993288894e-06,
+      "loss": 1.536,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8331775215539629,
+      "grad_norm": 0.4083802103996277,
+      "learning_rate": 6.710972426478096e-06,
+      "loss": 1.5312,
+      "step": 8021
+    },
+    {
+      "epoch": 0.8332813960735431,
+      "grad_norm": 0.41337552666664124,
+      "learning_rate": 6.7028095300110225e-06,
+      "loss": 1.6728,
+      "step": 8022
+    },
+    {
+      "epoch": 0.8333852705931235,
+      "grad_norm": 0.39118510484695435,
+      "learning_rate": 6.6946512443570085e-06,
+      "loss": 1.4944,
+      "step": 8023
+    },
+    {
+      "epoch": 0.8334891451127039,
+      "grad_norm": 0.4005904793739319,
+      "learning_rate": 6.686497570384825e-06,
+      "loss": 1.5726,
+      "step": 8024
+    },
+    {
+      "epoch": 0.8335930196322842,
+      "grad_norm": 0.4205056428909302,
+      "learning_rate": 6.678348508962812e-06,
+      "loss": 1.6453,
+      "step": 8025
+    },
+    {
+      "epoch": 0.8336968941518645,
+      "grad_norm": 0.4279542863368988,
+      "learning_rate": 6.670204060958746e-06,
+      "loss": 1.6386,
+      "step": 8026
+    },
+    {
+      "epoch": 0.8338007686714449,
+      "grad_norm": 0.44344186782836914,
+      "learning_rate": 6.662064227239967e-06,
+      "loss": 1.6229,
+      "step": 8027
+    },
+    {
+      "epoch": 0.8339046431910252,
+      "grad_norm": 0.4007083773612976,
+      "learning_rate": 6.653929008673315e-06,
+      "loss": 1.7046,
+      "step": 8028
+    },
+    {
+      "epoch": 0.8340085177106056,
+      "grad_norm": 0.4301755130290985,
+      "learning_rate": 6.645798406125087e-06,
+      "loss": 1.6751,
+      "step": 8029
+    },
+    {
+      "epoch": 0.834112392230186,
+      "grad_norm": 0.39321601390838623,
+      "learning_rate": 6.637672420461161e-06,
+      "loss": 1.4603,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8342162667497662,
+      "grad_norm": 0.4112393856048584,
+      "learning_rate": 6.629551052546884e-06,
+      "loss": 1.4504,
+      "step": 8031
+    },
+    {
+      "epoch": 0.8343201412693466,
+      "grad_norm": 0.4138662815093994,
+      "learning_rate": 6.6214343032471204e-06,
+      "loss": 1.6946,
+      "step": 8032
+    },
+    {
+      "epoch": 0.834424015788927,
+      "grad_norm": 0.43226510286331177,
+      "learning_rate": 6.613322173426239e-06,
+      "loss": 1.7209,
+      "step": 8033
+    },
+    {
+      "epoch": 0.8345278903085073,
+      "grad_norm": 0.42375436425209045,
+      "learning_rate": 6.605214663948111e-06,
+      "loss": 1.8021,
+      "step": 8034
+    },
+    {
+      "epoch": 0.8346317648280877,
+      "grad_norm": 0.4388313889503479,
+      "learning_rate": 6.597111775676135e-06,
+      "loss": 1.6669,
+      "step": 8035
+    },
+    {
+      "epoch": 0.834735639347668,
+      "grad_norm": 0.4430168569087982,
+      "learning_rate": 6.589013509473185e-06,
+      "loss": 1.5882,
+      "step": 8036
+    },
+    {
+      "epoch": 0.8348395138672484,
+      "grad_norm": 0.38932767510414124,
+      "learning_rate": 6.58091986620169e-06,
+      "loss": 1.4626,
+      "step": 8037
+    },
+    {
+      "epoch": 0.8349433883868287,
+      "grad_norm": 0.42232656478881836,
+      "learning_rate": 6.5728308467235435e-06,
+      "loss": 1.676,
+      "step": 8038
+    },
+    {
+      "epoch": 0.8350472629064091,
+      "grad_norm": 0.39752089977264404,
+      "learning_rate": 6.5647464519001725e-06,
+      "loss": 1.3398,
+      "step": 8039
+    },
+    {
+      "epoch": 0.8351511374259895,
+      "grad_norm": 0.4039764106273651,
+      "learning_rate": 6.556666682592494e-06,
+      "loss": 1.5381,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8352550119455697,
+      "grad_norm": 0.41539785265922546,
+      "learning_rate": 6.548591539660942e-06,
+      "loss": 1.6253,
+      "step": 8041
+    },
+    {
+      "epoch": 0.8353588864651501,
+      "grad_norm": 0.45972946286201477,
+      "learning_rate": 6.540521023965457e-06,
+      "loss": 1.7125,
+      "step": 8042
+    },
+    {
+      "epoch": 0.8354627609847305,
+      "grad_norm": 0.40128108859062195,
+      "learning_rate": 6.532455136365478e-06,
+      "loss": 1.645,
+      "step": 8043
+    },
+    {
+      "epoch": 0.8355666355043108,
+      "grad_norm": 0.43023881316185,
+      "learning_rate": 6.524393877719987e-06,
+      "loss": 1.6461,
+      "step": 8044
+    },
+    {
+      "epoch": 0.8356705100238911,
+      "grad_norm": 0.4410925507545471,
+      "learning_rate": 6.516337248887399e-06,
+      "loss": 1.8415,
+      "step": 8045
+    },
+    {
+      "epoch": 0.8357743845434715,
+      "grad_norm": 0.4021972417831421,
+      "learning_rate": 6.5082852507257265e-06,
+      "loss": 1.4099,
+      "step": 8046
+    },
+    {
+      "epoch": 0.8358782590630518,
+      "grad_norm": 0.4312531352043152,
+      "learning_rate": 6.500237884092425e-06,
+      "loss": 1.6968,
+      "step": 8047
+    },
+    {
+      "epoch": 0.8359821335826322,
+      "grad_norm": 0.40367355942726135,
+      "learning_rate": 6.492195149844471e-06,
+      "loss": 1.66,
+      "step": 8048
+    },
+    {
+      "epoch": 0.8360860081022126,
+      "grad_norm": 0.4290336072444916,
+      "learning_rate": 6.484157048838368e-06,
+      "loss": 1.657,
+      "step": 8049
+    },
+    {
+      "epoch": 0.8361898826217928,
+      "grad_norm": 0.4326249063014984,
+      "learning_rate": 6.476123581930099e-06,
+      "loss": 1.7264,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8362937571413732,
+      "grad_norm": 0.42618846893310547,
+      "learning_rate": 6.468094749975167e-06,
+      "loss": 1.798,
+      "step": 8051
+    },
+    {
+      "epoch": 0.8363976316609536,
+      "grad_norm": 0.4222290515899658,
+      "learning_rate": 6.460070553828573e-06,
+      "loss": 1.5829,
+      "step": 8052
+    },
+    {
+      "epoch": 0.8365015061805339,
+      "grad_norm": 0.3985992670059204,
+      "learning_rate": 6.4520509943448656e-06,
+      "loss": 1.5542,
+      "step": 8053
+    },
+    {
+      "epoch": 0.8366053807001143,
+      "grad_norm": 0.3930535614490509,
+      "learning_rate": 6.444036072378018e-06,
+      "loss": 1.5923,
+      "step": 8054
+    },
+    {
+      "epoch": 0.8367092552196946,
+      "grad_norm": 0.4032153785228729,
+      "learning_rate": 6.436025788781586e-06,
+      "loss": 1.6131,
+      "step": 8055
+    },
+    {
+      "epoch": 0.8368131297392749,
+      "grad_norm": 0.38635045289993286,
+      "learning_rate": 6.428020144408608e-06,
+      "loss": 1.5907,
+      "step": 8056
+    },
+    {
+      "epoch": 0.8369170042588553,
+      "grad_norm": 0.43229156732559204,
+      "learning_rate": 6.420019140111588e-06,
+      "loss": 1.6621,
+      "step": 8057
+    },
+    {
+      "epoch": 0.8370208787784357,
+      "grad_norm": 0.43635618686676025,
+      "learning_rate": 6.412022776742604e-06,
+      "loss": 1.5021,
+      "step": 8058
+    },
+    {
+      "epoch": 0.8371247532980159,
+      "grad_norm": 0.395163893699646,
+      "learning_rate": 6.404031055153198e-06,
+      "loss": 1.6401,
+      "step": 8059
+    },
+    {
+      "epoch": 0.8372286278175963,
+      "grad_norm": 0.3980249762535095,
+      "learning_rate": 6.396043976194416e-06,
+      "loss": 1.4499,
+      "step": 8060
+    },
+    {
+      "epoch": 0.8373325023371767,
+      "grad_norm": 0.4174276888370514,
+      "learning_rate": 6.3880615407168335e-06,
+      "loss": 1.6367,
+      "step": 8061
+    },
+    {
+      "epoch": 0.8374363768567571,
+      "grad_norm": 0.4133899509906769,
+      "learning_rate": 6.380083749570498e-06,
+      "loss": 1.5884,
+      "step": 8062
+    },
+    {
+      "epoch": 0.8375402513763374,
+      "grad_norm": 0.41283613443374634,
+      "learning_rate": 6.372110603605014e-06,
+      "loss": 1.4089,
+      "step": 8063
+    },
+    {
+      "epoch": 0.8376441258959177,
+      "grad_norm": 0.463218629360199,
+      "learning_rate": 6.364142103669418e-06,
+      "loss": 1.6824,
+      "step": 8064
+    },
+    {
+      "epoch": 0.8377480004154981,
+      "grad_norm": 0.39653337001800537,
+      "learning_rate": 6.356178250612327e-06,
+      "loss": 1.7319,
+      "step": 8065
+    },
+    {
+      "epoch": 0.8378518749350784,
+      "grad_norm": 0.42945095896720886,
+      "learning_rate": 6.348219045281822e-06,
+      "loss": 1.6687,
+      "step": 8066
+    },
+    {
+      "epoch": 0.8379557494546588,
+      "grad_norm": 0.4201321005821228,
+      "learning_rate": 6.340264488525488e-06,
+      "loss": 1.6577,
+      "step": 8067
+    },
+    {
+      "epoch": 0.8380596239742392,
+      "grad_norm": 0.44182446599006653,
+      "learning_rate": 6.332314581190424e-06,
+      "loss": 1.7371,
+      "step": 8068
+    },
+    {
+      "epoch": 0.8381634984938194,
+      "grad_norm": 0.4279153048992157,
+      "learning_rate": 6.324369324123242e-06,
+      "loss": 1.8348,
+      "step": 8069
+    },
+    {
+      "epoch": 0.8382673730133998,
+      "grad_norm": 0.42748069763183594,
+      "learning_rate": 6.316428718170036e-06,
+      "loss": 1.7954,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8383712475329802,
+      "grad_norm": 0.4482438862323761,
+      "learning_rate": 6.308492764176421e-06,
+      "loss": 1.4829,
+      "step": 8071
+    },
+    {
+      "epoch": 0.8384751220525605,
+      "grad_norm": 0.41090673208236694,
+      "learning_rate": 6.300561462987542e-06,
+      "loss": 1.469,
+      "step": 8072
+    },
+    {
+      "epoch": 0.8385789965721409,
+      "grad_norm": 0.4080744981765747,
+      "learning_rate": 6.292634815447978e-06,
+      "loss": 1.5294,
+      "step": 8073
+    },
+    {
+      "epoch": 0.8386828710917212,
+      "grad_norm": 0.4303456246852875,
+      "learning_rate": 6.2847128224018835e-06,
+      "loss": 1.5216,
+      "step": 8074
+    },
+    {
+      "epoch": 0.8387867456113015,
+      "grad_norm": 0.44199666380882263,
+      "learning_rate": 6.276795484692882e-06,
+      "loss": 1.7961,
+      "step": 8075
+    },
+    {
+      "epoch": 0.8388906201308819,
+      "grad_norm": 0.407693088054657,
+      "learning_rate": 6.268882803164106e-06,
+      "loss": 1.6753,
+      "step": 8076
+    },
+    {
+      "epoch": 0.8389944946504623,
+      "grad_norm": 0.4050787389278412,
+      "learning_rate": 6.260974778658202e-06,
+      "loss": 1.4427,
+      "step": 8077
+    },
+    {
+      "epoch": 0.8390983691700425,
+      "grad_norm": 0.4364643692970276,
+      "learning_rate": 6.253071412017298e-06,
+      "loss": 1.7915,
+      "step": 8078
+    },
+    {
+      "epoch": 0.8392022436896229,
+      "grad_norm": 0.4825323522090912,
+      "learning_rate": 6.2451727040830525e-06,
+      "loss": 1.8103,
+      "step": 8079
+    },
+    {
+      "epoch": 0.8393061182092033,
+      "grad_norm": 0.42931079864501953,
+      "learning_rate": 6.237278655696605e-06,
+      "loss": 1.4995,
+      "step": 8080
+    },
+    {
+      "epoch": 0.8394099927287836,
+      "grad_norm": 0.43772071599960327,
+      "learning_rate": 6.229389267698638e-06,
+      "loss": 1.737,
+      "step": 8081
+    },
+    {
+      "epoch": 0.839513867248364,
+      "grad_norm": 0.4058978259563446,
+      "learning_rate": 6.2215045409292794e-06,
+      "loss": 1.5779,
+      "step": 8082
+    },
+    {
+      "epoch": 0.8396177417679443,
+      "grad_norm": 0.4402966797351837,
+      "learning_rate": 6.213624476228191e-06,
+      "loss": 1.6885,
+      "step": 8083
+    },
+    {
+      "epoch": 0.8397216162875246,
+      "grad_norm": 0.4273587167263031,
+      "learning_rate": 6.205749074434569e-06,
+      "loss": 1.5655,
+      "step": 8084
+    },
+    {
+      "epoch": 0.839825490807105,
+      "grad_norm": 0.40787985920906067,
+      "learning_rate": 6.197878336387042e-06,
+      "loss": 1.5628,
+      "step": 8085
+    },
+    {
+      "epoch": 0.8399293653266854,
+      "grad_norm": 0.42607665061950684,
+      "learning_rate": 6.190012262923811e-06,
+      "loss": 1.5581,
+      "step": 8086
+    },
+    {
+      "epoch": 0.8400332398462657,
+      "grad_norm": 0.4342617690563202,
+      "learning_rate": 6.182150854882546e-06,
+      "loss": 1.8488,
+      "step": 8087
+    },
+    {
+      "epoch": 0.840137114365846,
+      "grad_norm": 0.3898141086101532,
+      "learning_rate": 6.1742941131004205e-06,
+      "loss": 1.5469,
+      "step": 8088
+    },
+    {
+      "epoch": 0.8402409888854264,
+      "grad_norm": 0.43741726875305176,
+      "learning_rate": 6.166442038414122e-06,
+      "loss": 1.756,
+      "step": 8089
+    },
+    {
+      "epoch": 0.8403448634050068,
+      "grad_norm": 0.44850122928619385,
+      "learning_rate": 6.1585946316598195e-06,
+      "loss": 1.5767,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8404487379245871,
+      "grad_norm": 0.4059605896472931,
+      "learning_rate": 6.150751893673229e-06,
+      "loss": 1.3907,
+      "step": 8091
+    },
+    {
+      "epoch": 0.8405526124441675,
+      "grad_norm": 0.46018004417419434,
+      "learning_rate": 6.142913825289509e-06,
+      "loss": 1.8505,
+      "step": 8092
+    },
+    {
+      "epoch": 0.8406564869637478,
+      "grad_norm": 0.4265289306640625,
+      "learning_rate": 6.135080427343376e-06,
+      "loss": 1.5951,
+      "step": 8093
+    },
+    {
+      "epoch": 0.8407603614833281,
+      "grad_norm": 0.4429461658000946,
+      "learning_rate": 6.1272517006690174e-06,
+      "loss": 1.7048,
+      "step": 8094
+    },
+    {
+      "epoch": 0.8408642360029085,
+      "grad_norm": 0.4366631805896759,
+      "learning_rate": 6.1194276461001345e-06,
+      "loss": 1.6959,
+      "step": 8095
+    },
+    {
+      "epoch": 0.8409681105224889,
+      "grad_norm": 0.48970597982406616,
+      "learning_rate": 6.111608264469926e-06,
+      "loss": 1.6371,
+      "step": 8096
+    },
+    {
+      "epoch": 0.8410719850420691,
+      "grad_norm": 0.3888145983219147,
+      "learning_rate": 6.103793556611093e-06,
+      "loss": 1.6457,
+      "step": 8097
+    },
+    {
+      "epoch": 0.8411758595616495,
+      "grad_norm": 0.39202946424484253,
+      "learning_rate": 6.095983523355841e-06,
+      "loss": 1.5726,
+      "step": 8098
+    },
+    {
+      "epoch": 0.8412797340812299,
+      "grad_norm": 0.41915374994277954,
+      "learning_rate": 6.088178165535874e-06,
+      "loss": 1.6642,
+      "step": 8099
+    },
+    {
+      "epoch": 0.8413836086008102,
+      "grad_norm": 0.3962395489215851,
+      "learning_rate": 6.080377483982424e-06,
+      "loss": 1.6482,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8414874831203906,
+      "grad_norm": 0.3973849415779114,
+      "learning_rate": 6.072581479526168e-06,
+      "loss": 1.5789,
+      "step": 8101
+    },
+    {
+      "epoch": 0.841591357639971,
+      "grad_norm": 0.46176597476005554,
+      "learning_rate": 6.064790152997346e-06,
+      "loss": 1.7686,
+      "step": 8102
+    },
+    {
+      "epoch": 0.8416952321595512,
+      "grad_norm": 0.39481309056282043,
+      "learning_rate": 6.0570035052256615e-06,
+      "loss": 1.4962,
+      "step": 8103
+    },
+    {
+      "epoch": 0.8417991066791316,
+      "grad_norm": 0.42713215947151184,
+      "learning_rate": 6.049221537040339e-06,
+      "loss": 1.5899,
+      "step": 8104
+    },
+    {
+      "epoch": 0.841902981198712,
+      "grad_norm": 0.4113538861274719,
+      "learning_rate": 6.041444249270089e-06,
+      "loss": 1.7044,
+      "step": 8105
+    },
+    {
+      "epoch": 0.8420068557182923,
+      "grad_norm": 0.414631724357605,
+      "learning_rate": 6.033671642743144e-06,
+      "loss": 1.6839,
+      "step": 8106
+    },
+    {
+      "epoch": 0.8421107302378726,
+      "grad_norm": 0.48907044529914856,
+      "learning_rate": 6.025903718287212e-06,
+      "loss": 1.6526,
+      "step": 8107
+    },
+    {
+      "epoch": 0.842214604757453,
+      "grad_norm": 0.45432183146476746,
+      "learning_rate": 6.018140476729517e-06,
+      "loss": 1.6792,
+      "step": 8108
+    },
+    {
+      "epoch": 0.8423184792770333,
+      "grad_norm": 0.4079488217830658,
+      "learning_rate": 6.010381918896807e-06,
+      "loss": 1.7154,
+      "step": 8109
+    },
+    {
+      "epoch": 0.8424223537966137,
+      "grad_norm": 0.42770206928253174,
+      "learning_rate": 6.002628045615283e-06,
+      "loss": 1.6739,
+      "step": 8110
+    },
+    {
+      "epoch": 0.8425262283161941,
+      "grad_norm": 0.46013393998146057,
+      "learning_rate": 5.994878857710667e-06,
+      "loss": 1.7719,
+      "step": 8111
+    },
+    {
+      "epoch": 0.8426301028357743,
+      "grad_norm": 0.39818307757377625,
+      "learning_rate": 5.9871343560082225e-06,
+      "loss": 1.6761,
+      "step": 8112
+    },
+    {
+      "epoch": 0.8427339773553547,
+      "grad_norm": 0.3909452557563782,
+      "learning_rate": 5.9793945413326335e-06,
+      "loss": 1.4098,
+      "step": 8113
+    },
+    {
+      "epoch": 0.8428378518749351,
+      "grad_norm": 0.4382929801940918,
+      "learning_rate": 5.9716594145081625e-06,
+      "loss": 1.6621,
+      "step": 8114
+    },
+    {
+      "epoch": 0.8429417263945155,
+      "grad_norm": 0.4438725709915161,
+      "learning_rate": 5.963928976358518e-06,
+      "loss": 1.7382,
+      "step": 8115
+    },
+    {
+      "epoch": 0.8430456009140957,
+      "grad_norm": 0.4258117973804474,
+      "learning_rate": 5.956203227706963e-06,
+      "loss": 1.7194,
+      "step": 8116
+    },
+    {
+      "epoch": 0.8431494754336761,
+      "grad_norm": 0.486128032207489,
+      "learning_rate": 5.948482169376202e-06,
+      "loss": 1.7364,
+      "step": 8117
+    },
+    {
+      "epoch": 0.8432533499532565,
+      "grad_norm": 0.4413326382637024,
+      "learning_rate": 5.940765802188458e-06,
+      "loss": 1.4118,
+      "step": 8118
+    },
+    {
+      "epoch": 0.8433572244728368,
+      "grad_norm": 0.3983282148838043,
+      "learning_rate": 5.933054126965509e-06,
+      "loss": 1.3972,
+      "step": 8119
+    },
+    {
+      "epoch": 0.8434610989924172,
+      "grad_norm": 0.4958374500274658,
+      "learning_rate": 5.925347144528537e-06,
+      "loss": 1.8372,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8435649735119976,
+      "grad_norm": 0.44721078872680664,
+      "learning_rate": 5.917644855698306e-06,
+      "loss": 1.7691,
+      "step": 8121
+    },
+    {
+      "epoch": 0.8436688480315778,
+      "grad_norm": 0.40920016169548035,
+      "learning_rate": 5.9099472612950455e-06,
+      "loss": 1.6897,
+      "step": 8122
+    },
+    {
+      "epoch": 0.8437727225511582,
+      "grad_norm": 0.44964519143104553,
+      "learning_rate": 5.902254362138487e-06,
+      "loss": 1.7894,
+      "step": 8123
+    },
+    {
+      "epoch": 0.8438765970707386,
+      "grad_norm": 0.4365478456020355,
+      "learning_rate": 5.894566159047865e-06,
+      "loss": 1.5154,
+      "step": 8124
+    },
+    {
+      "epoch": 0.8439804715903189,
+      "grad_norm": 0.42581430077552795,
+      "learning_rate": 5.886882652841907e-06,
+      "loss": 1.613,
+      "step": 8125
+    },
+    {
+      "epoch": 0.8440843461098992,
+      "grad_norm": 0.4058164656162262,
+      "learning_rate": 5.879203844338848e-06,
+      "loss": 1.6085,
+      "step": 8126
+    },
+    {
+      "epoch": 0.8441882206294796,
+      "grad_norm": 0.40692541003227234,
+      "learning_rate": 5.871529734356423e-06,
+      "loss": 1.6069,
+      "step": 8127
+    },
+    {
+      "epoch": 0.8442920951490599,
+      "grad_norm": 0.4061817228794098,
+      "learning_rate": 5.863860323711879e-06,
+      "loss": 1.539,
+      "step": 8128
+    },
+    {
+      "epoch": 0.8443959696686403,
+      "grad_norm": 0.489007830619812,
+      "learning_rate": 5.856195613221921e-06,
+      "loss": 1.3045,
+      "step": 8129
+    },
+    {
+      "epoch": 0.8444998441882207,
+      "grad_norm": 0.43798279762268066,
+      "learning_rate": 5.848535603702798e-06,
+      "loss": 1.5672,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8446037187078009,
+      "grad_norm": 0.43467089533805847,
+      "learning_rate": 5.840880295970247e-06,
+      "loss": 1.7033,
+      "step": 8131
+    },
+    {
+      "epoch": 0.8447075932273813,
+      "grad_norm": 0.4537743031978607,
+      "learning_rate": 5.833229690839481e-06,
+      "loss": 1.7776,
+      "step": 8132
+    },
+    {
+      "epoch": 0.8448114677469617,
+      "grad_norm": 0.43950802087783813,
+      "learning_rate": 5.825583789125244e-06,
+      "loss": 1.7372,
+      "step": 8133
+    },
+    {
+      "epoch": 0.844915342266542,
+      "grad_norm": 0.4260973334312439,
+      "learning_rate": 5.817942591641762e-06,
+      "loss": 1.7629,
+      "step": 8134
+    },
+    {
+      "epoch": 0.8450192167861224,
+      "grad_norm": 0.40139877796173096,
+      "learning_rate": 5.810306099202755e-06,
+      "loss": 1.6475,
+      "step": 8135
+    },
+    {
+      "epoch": 0.8451230913057027,
+      "grad_norm": 0.45263901352882385,
+      "learning_rate": 5.802674312621448e-06,
+      "loss": 1.477,
+      "step": 8136
+    },
+    {
+      "epoch": 0.845226965825283,
+      "grad_norm": 0.4598293900489807,
+      "learning_rate": 5.795047232710599e-06,
+      "loss": 1.586,
+      "step": 8137
+    },
+    {
+      "epoch": 0.8453308403448634,
+      "grad_norm": 0.4252384901046753,
+      "learning_rate": 5.787424860282392e-06,
+      "loss": 1.5229,
+      "step": 8138
+    },
+    {
+      "epoch": 0.8454347148644438,
+      "grad_norm": 0.42608946561813354,
+      "learning_rate": 5.779807196148556e-06,
+      "loss": 1.7491,
+      "step": 8139
+    },
+    {
+      "epoch": 0.8455385893840242,
+      "grad_norm": 0.4330217242240906,
+      "learning_rate": 5.772194241120338e-06,
+      "loss": 1.5724,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8456424639036044,
+      "grad_norm": 0.4884481132030487,
+      "learning_rate": 5.764585996008437e-06,
+      "loss": 1.7322,
+      "step": 8141
+    },
+    {
+      "epoch": 0.8457463384231848,
+      "grad_norm": 0.45733243227005005,
+      "learning_rate": 5.756982461623084e-06,
+      "loss": 1.763,
+      "step": 8142
+    },
+    {
+      "epoch": 0.8458502129427652,
+      "grad_norm": 0.5284457802772522,
+      "learning_rate": 5.749383638773986e-06,
+      "loss": 1.9331,
+      "step": 8143
+    },
+    {
+      "epoch": 0.8459540874623455,
+      "grad_norm": 0.4064652621746063,
+      "learning_rate": 5.7417895282703635e-06,
+      "loss": 1.6423,
+      "step": 8144
+    },
+    {
+      "epoch": 0.8460579619819258,
+      "grad_norm": 0.4062190055847168,
+      "learning_rate": 5.734200130920925e-06,
+      "loss": 1.5123,
+      "step": 8145
+    },
+    {
+      "epoch": 0.8461618365015062,
+      "grad_norm": 0.4112701416015625,
+      "learning_rate": 5.726615447533878e-06,
+      "loss": 1.4982,
+      "step": 8146
+    },
+    {
+      "epoch": 0.8462657110210865,
+      "grad_norm": 0.4885147213935852,
+      "learning_rate": 5.719035478916967e-06,
+      "loss": 1.8708,
+      "step": 8147
+    },
+    {
+      "epoch": 0.8463695855406669,
+      "grad_norm": 0.44870370626449585,
+      "learning_rate": 5.711460225877346e-06,
+      "loss": 1.5868,
+      "step": 8148
+    },
+    {
+      "epoch": 0.8464734600602473,
+      "grad_norm": 0.3870431184768677,
+      "learning_rate": 5.703889689221758e-06,
+      "loss": 1.4511,
+      "step": 8149
+    },
+    {
+      "epoch": 0.8465773345798275,
+      "grad_norm": 0.41669902205467224,
+      "learning_rate": 5.696323869756398e-06,
+      "loss": 1.6854,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8466812090994079,
+      "grad_norm": 0.40349993109703064,
+      "learning_rate": 5.68876276828696e-06,
+      "loss": 1.6919,
+      "step": 8151
+    },
+    {
+      "epoch": 0.8467850836189883,
+      "grad_norm": 0.41194114089012146,
+      "learning_rate": 5.681206385618648e-06,
+      "loss": 1.5359,
+      "step": 8152
+    },
+    {
+      "epoch": 0.8468889581385686,
+      "grad_norm": 0.46073397994041443,
+      "learning_rate": 5.6736547225561545e-06,
+      "loss": 1.8919,
+      "step": 8153
+    },
+    {
+      "epoch": 0.846992832658149,
+      "grad_norm": 0.4091498851776123,
+      "learning_rate": 5.6661077799036785e-06,
+      "loss": 1.5971,
+      "step": 8154
+    },
+    {
+      "epoch": 0.8470967071777293,
+      "grad_norm": 0.4949287176132202,
+      "learning_rate": 5.65856555846489e-06,
+      "loss": 1.6858,
+      "step": 8155
+    },
+    {
+      "epoch": 0.8472005816973096,
+      "grad_norm": 0.43929752707481384,
+      "learning_rate": 5.6510280590430174e-06,
+      "loss": 1.673,
+      "step": 8156
+    },
+    {
+      "epoch": 0.84730445621689,
+      "grad_norm": 0.4475664794445038,
+      "learning_rate": 5.643495282440703e-06,
+      "loss": 1.5692,
+      "step": 8157
+    },
+    {
+      "epoch": 0.8474083307364704,
+      "grad_norm": 0.43574050068855286,
+      "learning_rate": 5.635967229460154e-06,
+      "loss": 1.5003,
+      "step": 8158
+    },
+    {
+      "epoch": 0.8475122052560506,
+      "grad_norm": 0.4088006317615509,
+      "learning_rate": 5.628443900903041e-06,
+      "loss": 1.6919,
+      "step": 8159
+    },
+    {
+      "epoch": 0.847616079775631,
+      "grad_norm": 0.41070079803466797,
+      "learning_rate": 5.62092529757054e-06,
+      "loss": 1.602,
+      "step": 8160
+    },
+    {
+      "epoch": 0.8477199542952114,
+      "grad_norm": 0.38564005494117737,
+      "learning_rate": 5.613411420263331e-06,
+      "loss": 1.6129,
+      "step": 8161
+    },
+    {
+      "epoch": 0.8478238288147917,
+      "grad_norm": 0.3986556828022003,
+      "learning_rate": 5.605902269781571e-06,
+      "loss": 1.572,
+      "step": 8162
+    },
+    {
+      "epoch": 0.8479277033343721,
+      "grad_norm": 0.41261792182922363,
+      "learning_rate": 5.598397846924935e-06,
+      "loss": 1.6821,
+      "step": 8163
+    },
+    {
+      "epoch": 0.8480315778539524,
+      "grad_norm": 0.41062214970588684,
+      "learning_rate": 5.590898152492574e-06,
+      "loss": 1.6375,
+      "step": 8164
+    },
+    {
+      "epoch": 0.8481354523735328,
+      "grad_norm": 0.4452644884586334,
+      "learning_rate": 5.583403187283159e-06,
+      "loss": 1.7255,
+      "step": 8165
+    },
+    {
+      "epoch": 0.8482393268931131,
+      "grad_norm": 0.41799211502075195,
+      "learning_rate": 5.575912952094858e-06,
+      "loss": 1.7604,
+      "step": 8166
+    },
+    {
+      "epoch": 0.8483432014126935,
+      "grad_norm": 0.41636091470718384,
+      "learning_rate": 5.568427447725283e-06,
+      "loss": 1.7037,
+      "step": 8167
+    },
+    {
+      "epoch": 0.8484470759322739,
+      "grad_norm": 0.45778560638427734,
+      "learning_rate": 5.560946674971618e-06,
+      "loss": 1.8658,
+      "step": 8168
+    },
+    {
+      "epoch": 0.8485509504518541,
+      "grad_norm": 0.4254413843154907,
+      "learning_rate": 5.553470634630492e-06,
+      "loss": 1.5678,
+      "step": 8169
+    },
+    {
+      "epoch": 0.8486548249714345,
+      "grad_norm": 0.4398033320903778,
+      "learning_rate": 5.5459993274980525e-06,
+      "loss": 1.6442,
+      "step": 8170
+    },
+    {
+      "epoch": 0.8487586994910149,
+      "grad_norm": 0.3922208547592163,
+      "learning_rate": 5.538532754369924e-06,
+      "loss": 1.6004,
+      "step": 8171
+    },
+    {
+      "epoch": 0.8488625740105952,
+      "grad_norm": 0.40766361355781555,
+      "learning_rate": 5.531070916041247e-06,
+      "loss": 1.5942,
+      "step": 8172
+    },
+    {
+      "epoch": 0.8489664485301756,
+      "grad_norm": 0.39256706833839417,
+      "learning_rate": 5.523613813306644e-06,
+      "loss": 1.6922,
+      "step": 8173
+    },
+    {
+      "epoch": 0.8490703230497559,
+      "grad_norm": 0.4119519591331482,
+      "learning_rate": 5.516161446960233e-06,
+      "loss": 1.6135,
+      "step": 8174
+    },
+    {
+      "epoch": 0.8491741975693362,
+      "grad_norm": 0.43249499797821045,
+      "learning_rate": 5.508713817795658e-06,
+      "loss": 1.5972,
+      "step": 8175
+    },
+    {
+      "epoch": 0.8492780720889166,
+      "grad_norm": 0.4471118152141571,
+      "learning_rate": 5.501270926606e-06,
+      "loss": 1.5106,
+      "step": 8176
+    },
+    {
+      "epoch": 0.849381946608497,
+      "grad_norm": 0.49427923560142517,
+      "learning_rate": 5.493832774183894e-06,
+      "loss": 1.9156,
+      "step": 8177
+    },
+    {
+      "epoch": 0.8494858211280772,
+      "grad_norm": 0.4142952263355255,
+      "learning_rate": 5.4863993613214314e-06,
+      "loss": 1.592,
+      "step": 8178
+    },
+    {
+      "epoch": 0.8495896956476576,
+      "grad_norm": 0.4335556626319885,
+      "learning_rate": 5.478970688810225e-06,
+      "loss": 1.6454,
+      "step": 8179
+    },
+    {
+      "epoch": 0.849693570167238,
+      "grad_norm": 0.42512187361717224,
+      "learning_rate": 5.471546757441359e-06,
+      "loss": 1.7903,
+      "step": 8180
+    },
+    {
+      "epoch": 0.8497974446868183,
+      "grad_norm": 0.4263908863067627,
+      "learning_rate": 5.464127568005423e-06,
+      "loss": 1.5594,
+      "step": 8181
+    },
+    {
+      "epoch": 0.8499013192063987,
+      "grad_norm": 0.4174048602581024,
+      "learning_rate": 5.456713121292517e-06,
+      "loss": 1.6509,
+      "step": 8182
+    },
+    {
+      "epoch": 0.850005193725979,
+      "grad_norm": 0.3824132978916168,
+      "learning_rate": 5.4493034180921944e-06,
+      "loss": 1.4437,
+      "step": 8183
+    },
+    {
+      "epoch": 0.8501090682455593,
+      "grad_norm": 0.42041993141174316,
+      "learning_rate": 5.441898459193573e-06,
+      "loss": 1.689,
+      "step": 8184
+    },
+    {
+      "epoch": 0.8502129427651397,
+      "grad_norm": 0.3966014087200165,
+      "learning_rate": 5.434498245385184e-06,
+      "loss": 1.6464,
+      "step": 8185
+    },
+    {
+      "epoch": 0.8503168172847201,
+      "grad_norm": 0.4592750370502472,
+      "learning_rate": 5.427102777455112e-06,
+      "loss": 1.4614,
+      "step": 8186
+    },
+    {
+      "epoch": 0.8504206918043004,
+      "grad_norm": 0.40993037819862366,
+      "learning_rate": 5.419712056190912e-06,
+      "loss": 1.5345,
+      "step": 8187
+    },
+    {
+      "epoch": 0.8505245663238807,
+      "grad_norm": 0.4124085009098053,
+      "learning_rate": 5.412326082379637e-06,
+      "loss": 1.5901,
+      "step": 8188
+    },
+    {
+      "epoch": 0.8506284408434611,
+      "grad_norm": 0.3968968391418457,
+      "learning_rate": 5.404944856807842e-06,
+      "loss": 1.5966,
+      "step": 8189
+    },
+    {
+      "epoch": 0.8507323153630414,
+      "grad_norm": 0.4331669509410858,
+      "learning_rate": 5.3975683802615586e-06,
+      "loss": 1.6621,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8508361898826218,
+      "grad_norm": 0.42043596506118774,
+      "learning_rate": 5.390196653526336e-06,
+      "loss": 1.6508,
+      "step": 8191
+    },
+    {
+      "epoch": 0.8509400644022022,
+      "grad_norm": 0.41201046109199524,
+      "learning_rate": 5.382829677387186e-06,
+      "loss": 1.6648,
+      "step": 8192
+    },
+    {
+      "epoch": 0.8510439389217825,
+      "grad_norm": 0.41224047541618347,
+      "learning_rate": 5.375467452628663e-06,
+      "loss": 1.7252,
+      "step": 8193
+    },
+    {
+      "epoch": 0.8511478134413628,
+      "grad_norm": 0.42982521653175354,
+      "learning_rate": 5.368109980034775e-06,
+      "loss": 1.6036,
+      "step": 8194
+    },
+    {
+      "epoch": 0.8512516879609432,
+      "grad_norm": 0.41480857133865356,
+      "learning_rate": 5.360757260389015e-06,
+      "loss": 1.5241,
+      "step": 8195
+    },
+    {
+      "epoch": 0.8513555624805236,
+      "grad_norm": 0.41016316413879395,
+      "learning_rate": 5.353409294474421e-06,
+      "loss": 1.568,
+      "step": 8196
+    },
+    {
+      "epoch": 0.8514594370001038,
+      "grad_norm": 0.41887757182121277,
+      "learning_rate": 5.346066083073475e-06,
+      "loss": 1.5569,
+      "step": 8197
+    },
+    {
+      "epoch": 0.8515633115196842,
+      "grad_norm": 0.43342113494873047,
+      "learning_rate": 5.338727626968182e-06,
+      "loss": 1.6933,
+      "step": 8198
+    },
+    {
+      "epoch": 0.8516671860392646,
+      "grad_norm": 0.4318787753582001,
+      "learning_rate": 5.331393926940021e-06,
+      "loss": 1.7458,
+      "step": 8199
+    },
+    {
+      "epoch": 0.8517710605588449,
+      "grad_norm": 0.4099392294883728,
+      "learning_rate": 5.324064983769978e-06,
+      "loss": 1.6325,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8518749350784253,
+      "grad_norm": 0.43265894055366516,
+      "learning_rate": 5.316740798238534e-06,
+      "loss": 1.7025,
+      "step": 8201
+    },
+    {
+      "epoch": 0.8519788095980056,
+      "grad_norm": 0.4109574854373932,
+      "learning_rate": 5.309421371125639e-06,
+      "loss": 1.7289,
+      "step": 8202
+    },
+    {
+      "epoch": 0.8520826841175859,
+      "grad_norm": 0.4493882954120636,
+      "learning_rate": 5.302106703210785e-06,
+      "loss": 1.6378,
+      "step": 8203
+    },
+    {
+      "epoch": 0.8521865586371663,
+      "grad_norm": 0.49722832441329956,
+      "learning_rate": 5.294796795272894e-06,
+      "loss": 1.8429,
+      "step": 8204
+    },
+    {
+      "epoch": 0.8522904331567467,
+      "grad_norm": 0.3835256099700928,
+      "learning_rate": 5.287491648090437e-06,
+      "loss": 1.515,
+      "step": 8205
+    },
+    {
+      "epoch": 0.852394307676327,
+      "grad_norm": 0.4009277820587158,
+      "learning_rate": 5.280191262441353e-06,
+      "loss": 1.4383,
+      "step": 8206
+    },
+    {
+      "epoch": 0.8524981821959073,
+      "grad_norm": 0.4271654486656189,
+      "learning_rate": 5.27289563910307e-06,
+      "loss": 1.6824,
+      "step": 8207
+    },
+    {
+      "epoch": 0.8526020567154877,
+      "grad_norm": 0.41742265224456787,
+      "learning_rate": 5.265604778852512e-06,
+      "loss": 1.525,
+      "step": 8208
+    },
+    {
+      "epoch": 0.852705931235068,
+      "grad_norm": 0.4177842140197754,
+      "learning_rate": 5.258318682466107e-06,
+      "loss": 1.6278,
+      "step": 8209
+    },
+    {
+      "epoch": 0.8528098057546484,
+      "grad_norm": 0.4213246703147888,
+      "learning_rate": 5.2510373507197675e-06,
+      "loss": 1.563,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8529136802742288,
+      "grad_norm": 0.40991777181625366,
+      "learning_rate": 5.243760784388879e-06,
+      "loss": 1.5489,
+      "step": 8211
+    },
+    {
+      "epoch": 0.853017554793809,
+      "grad_norm": 0.4318062961101532,
+      "learning_rate": 5.236488984248378e-06,
+      "loss": 1.7528,
+      "step": 8212
+    },
+    {
+      "epoch": 0.8531214293133894,
+      "grad_norm": 0.40052494406700134,
+      "learning_rate": 5.229221951072611e-06,
+      "loss": 1.5852,
+      "step": 8213
+    },
+    {
+      "epoch": 0.8532253038329698,
+      "grad_norm": 0.4887178838253021,
+      "learning_rate": 5.221959685635491e-06,
+      "loss": 1.7859,
+      "step": 8214
+    },
+    {
+      "epoch": 0.8533291783525501,
+      "grad_norm": 0.4212459325790405,
+      "learning_rate": 5.2147021887103796e-06,
+      "loss": 1.6043,
+      "step": 8215
+    },
+    {
+      "epoch": 0.8534330528721304,
+      "grad_norm": 0.43245670199394226,
+      "learning_rate": 5.207449461070146e-06,
+      "loss": 1.7065,
+      "step": 8216
+    },
+    {
+      "epoch": 0.8535369273917108,
+      "grad_norm": 0.48420557379722595,
+      "learning_rate": 5.2002015034871564e-06,
+      "loss": 1.8633,
+      "step": 8217
+    },
+    {
+      "epoch": 0.8536408019112912,
+      "grad_norm": 0.46730557084083557,
+      "learning_rate": 5.1929583167332375e-06,
+      "loss": 1.7325,
+      "step": 8218
+    },
+    {
+      "epoch": 0.8537446764308715,
+      "grad_norm": 0.45189306139945984,
+      "learning_rate": 5.185719901579772e-06,
+      "loss": 1.8017,
+      "step": 8219
+    },
+    {
+      "epoch": 0.8538485509504519,
+      "grad_norm": 0.46504566073417664,
+      "learning_rate": 5.178486258797555e-06,
+      "loss": 1.8846,
+      "step": 8220
+    },
+    {
+      "epoch": 0.8539524254700323,
+      "grad_norm": 0.4610961973667145,
+      "learning_rate": 5.171257389156936e-06,
+      "loss": 1.7558,
+      "step": 8221
+    },
+    {
+      "epoch": 0.8540562999896125,
+      "grad_norm": 0.4036599397659302,
+      "learning_rate": 5.164033293427739e-06,
+      "loss": 1.6539,
+      "step": 8222
+    },
+    {
+      "epoch": 0.8541601745091929,
+      "grad_norm": 0.4568711817264557,
+      "learning_rate": 5.156813972379243e-06,
+      "loss": 1.769,
+      "step": 8223
+    },
+    {
+      "epoch": 0.8542640490287733,
+      "grad_norm": 0.4027152955532074,
+      "learning_rate": 5.1495994267802775e-06,
+      "loss": 1.5509,
+      "step": 8224
+    },
+    {
+      "epoch": 0.8543679235483536,
+      "grad_norm": 0.44536474347114563,
+      "learning_rate": 5.142389657399127e-06,
+      "loss": 1.7014,
+      "step": 8225
+    },
+    {
+      "epoch": 0.8544717980679339,
+      "grad_norm": 0.43924811482429504,
+      "learning_rate": 5.135184665003573e-06,
+      "loss": 1.602,
+      "step": 8226
+    },
+    {
+      "epoch": 0.8545756725875143,
+      "grad_norm": 0.4705229699611664,
+      "learning_rate": 5.127984450360895e-06,
+      "loss": 1.7716,
+      "step": 8227
+    },
+    {
+      "epoch": 0.8546795471070946,
+      "grad_norm": 0.42262768745422363,
+      "learning_rate": 5.120789014237859e-06,
+      "loss": 1.6233,
+      "step": 8228
+    },
+    {
+      "epoch": 0.854783421626675,
+      "grad_norm": 0.41569891571998596,
+      "learning_rate": 5.113598357400723e-06,
+      "loss": 1.5956,
+      "step": 8229
+    },
+    {
+      "epoch": 0.8548872961462554,
+      "grad_norm": 0.3866024911403656,
+      "learning_rate": 5.106412480615219e-06,
+      "loss": 1.5542,
+      "step": 8230
+    },
+    {
+      "epoch": 0.8549911706658356,
+      "grad_norm": 0.4203120768070221,
+      "learning_rate": 5.09923138464663e-06,
+      "loss": 1.6801,
+      "step": 8231
+    },
+    {
+      "epoch": 0.855095045185416,
+      "grad_norm": 0.433152973651886,
+      "learning_rate": 5.092055070259632e-06,
+      "loss": 1.6589,
+      "step": 8232
+    },
+    {
+      "epoch": 0.8551989197049964,
+      "grad_norm": 0.42945393919944763,
+      "learning_rate": 5.0848835382184875e-06,
+      "loss": 1.6403,
+      "step": 8233
+    },
+    {
+      "epoch": 0.8553027942245767,
+      "grad_norm": 0.46806976199150085,
+      "learning_rate": 5.077716789286896e-06,
+      "loss": 1.7067,
+      "step": 8234
+    },
+    {
+      "epoch": 0.855406668744157,
+      "grad_norm": 0.3977743983268738,
+      "learning_rate": 5.070554824228052e-06,
+      "loss": 1.667,
+      "step": 8235
+    },
+    {
+      "epoch": 0.8555105432637374,
+      "grad_norm": 0.40513405203819275,
+      "learning_rate": 5.063397643804662e-06,
+      "loss": 1.4978,
+      "step": 8236
+    },
+    {
+      "epoch": 0.8556144177833177,
+      "grad_norm": 0.41432732343673706,
+      "learning_rate": 5.056245248778902e-06,
+      "loss": 1.5833,
+      "step": 8237
+    },
+    {
+      "epoch": 0.8557182923028981,
+      "grad_norm": 0.4109817147254944,
+      "learning_rate": 5.049097639912454e-06,
+      "loss": 1.4775,
+      "step": 8238
+    },
+    {
+      "epoch": 0.8558221668224785,
+      "grad_norm": 0.4143582582473755,
+      "learning_rate": 5.04195481796646e-06,
+      "loss": 1.612,
+      "step": 8239
+    },
+    {
+      "epoch": 0.8559260413420587,
+      "grad_norm": 0.45763957500457764,
+      "learning_rate": 5.034816783701613e-06,
+      "loss": 1.5282,
+      "step": 8240
+    },
+    {
+      "epoch": 0.8560299158616391,
+      "grad_norm": 0.4432554543018341,
+      "learning_rate": 5.027683537878014e-06,
+      "loss": 1.5708,
+      "step": 8241
+    },
+    {
+      "epoch": 0.8561337903812195,
+      "grad_norm": 0.42282775044441223,
+      "learning_rate": 5.0205550812553345e-06,
+      "loss": 1.6524,
+      "step": 8242
+    },
+    {
+      "epoch": 0.8562376649007999,
+      "grad_norm": 0.4614414870738983,
+      "learning_rate": 5.01343141459269e-06,
+      "loss": 1.8134,
+      "step": 8243
+    },
+    {
+      "epoch": 0.8563415394203802,
+      "grad_norm": 0.5265931487083435,
+      "learning_rate": 5.0063125386486834e-06,
+      "loss": 1.9862,
+      "step": 8244
+    },
+    {
+      "epoch": 0.8564454139399605,
+      "grad_norm": 0.46069347858428955,
+      "learning_rate": 4.999198454181431e-06,
+      "loss": 1.8108,
+      "step": 8245
+    },
+    {
+      "epoch": 0.8565492884595409,
+      "grad_norm": 0.397867351770401,
+      "learning_rate": 4.99208916194851e-06,
+      "loss": 1.6589,
+      "step": 8246
+    },
+    {
+      "epoch": 0.8566531629791212,
+      "grad_norm": 0.4252391457557678,
+      "learning_rate": 4.984984662707043e-06,
+      "loss": 1.7302,
+      "step": 8247
+    },
+    {
+      "epoch": 0.8567570374987016,
+      "grad_norm": 0.41175001859664917,
+      "learning_rate": 4.977884957213558e-06,
+      "loss": 1.564,
+      "step": 8248
+    },
+    {
+      "epoch": 0.856860912018282,
+      "grad_norm": 0.40351465344429016,
+      "learning_rate": 4.970790046224144e-06,
+      "loss": 1.7149,
+      "step": 8249
+    },
+    {
+      "epoch": 0.8569647865378622,
+      "grad_norm": 0.4298102557659149,
+      "learning_rate": 4.963699930494364e-06,
+      "loss": 1.5095,
+      "step": 8250
+    },
+    {
+      "epoch": 0.8570686610574426,
+      "grad_norm": 0.45123982429504395,
+      "learning_rate": 4.95661461077922e-06,
+      "loss": 1.5314,
+      "step": 8251
+    },
+    {
+      "epoch": 0.857172535577023,
+      "grad_norm": 0.41809526085853577,
+      "learning_rate": 4.94953408783328e-06,
+      "loss": 1.5747,
+      "step": 8252
+    },
+    {
+      "epoch": 0.8572764100966033,
+      "grad_norm": 0.43481943011283875,
+      "learning_rate": 4.942458362410557e-06,
+      "loss": 1.628,
+      "step": 8253
+    },
+    {
+      "epoch": 0.8573802846161837,
+      "grad_norm": 0.4104442000389099,
+      "learning_rate": 4.935387435264549e-06,
+      "loss": 1.5326,
+      "step": 8254
+    },
+    {
+      "epoch": 0.857484159135764,
+      "grad_norm": 0.47249045968055725,
+      "learning_rate": 4.928321307148265e-06,
+      "loss": 1.7157,
+      "step": 8255
+    },
+    {
+      "epoch": 0.8575880336553443,
+      "grad_norm": 0.4751419425010681,
+      "learning_rate": 4.921259978814191e-06,
+      "loss": 1.8348,
+      "step": 8256
+    },
+    {
+      "epoch": 0.8576919081749247,
+      "grad_norm": 0.44779595732688904,
+      "learning_rate": 4.914203451014299e-06,
+      "loss": 1.521,
+      "step": 8257
+    },
+    {
+      "epoch": 0.8577957826945051,
+      "grad_norm": 0.4319514036178589,
+      "learning_rate": 4.907151724500048e-06,
+      "loss": 1.6885,
+      "step": 8258
+    },
+    {
+      "epoch": 0.8578996572140853,
+      "grad_norm": 0.44196733832359314,
+      "learning_rate": 4.9001048000224205e-06,
+      "loss": 1.746,
+      "step": 8259
+    },
+    {
+      "epoch": 0.8580035317336657,
+      "grad_norm": 0.4159088134765625,
+      "learning_rate": 4.893062678331817e-06,
+      "loss": 1.794,
+      "step": 8260
+    },
+    {
+      "epoch": 0.8581074062532461,
+      "grad_norm": 0.4175340533256531,
+      "learning_rate": 4.886025360178204e-06,
+      "loss": 1.6533,
+      "step": 8261
+    },
+    {
+      "epoch": 0.8582112807728264,
+      "grad_norm": 0.41944602131843567,
+      "learning_rate": 4.878992846310987e-06,
+      "loss": 1.6403,
+      "step": 8262
+    },
+    {
+      "epoch": 0.8583151552924068,
+      "grad_norm": 0.43087661266326904,
+      "learning_rate": 4.8719651374790735e-06,
+      "loss": 1.7938,
+      "step": 8263
+    },
+    {
+      "epoch": 0.8584190298119871,
+      "grad_norm": 0.41063591837882996,
+      "learning_rate": 4.864942234430858e-06,
+      "loss": 1.5834,
+      "step": 8264
+    },
+    {
+      "epoch": 0.8585229043315674,
+      "grad_norm": 0.42360448837280273,
+      "learning_rate": 4.857924137914233e-06,
+      "loss": 1.6952,
+      "step": 8265
+    },
+    {
+      "epoch": 0.8586267788511478,
+      "grad_norm": 0.4151276648044586,
+      "learning_rate": 4.85091084867656e-06,
+      "loss": 1.658,
+      "step": 8266
+    },
+    {
+      "epoch": 0.8587306533707282,
+      "grad_norm": 0.41758909821510315,
+      "learning_rate": 4.843902367464698e-06,
+      "loss": 1.6903,
+      "step": 8267
+    },
+    {
+      "epoch": 0.8588345278903085,
+      "grad_norm": 0.4258948266506195,
+      "learning_rate": 4.8368986950250275e-06,
+      "loss": 1.7291,
+      "step": 8268
+    },
+    {
+      "epoch": 0.8589384024098888,
+      "grad_norm": 0.3859458565711975,
+      "learning_rate": 4.829899832103335e-06,
+      "loss": 1.5672,
+      "step": 8269
+    },
+    {
+      "epoch": 0.8590422769294692,
+      "grad_norm": 0.3968896269798279,
+      "learning_rate": 4.82290577944498e-06,
+      "loss": 1.5287,
+      "step": 8270
+    },
+    {
+      "epoch": 0.8591461514490496,
+      "grad_norm": 0.4189419150352478,
+      "learning_rate": 4.815916537794763e-06,
+      "loss": 1.7356,
+      "step": 8271
+    },
+    {
+      "epoch": 0.8592500259686299,
+      "grad_norm": 0.39579710364341736,
+      "learning_rate": 4.8089321078969904e-06,
+      "loss": 1.5447,
+      "step": 8272
+    },
+    {
+      "epoch": 0.8593539004882103,
+      "grad_norm": 0.42127490043640137,
+      "learning_rate": 4.801952490495437e-06,
+      "loss": 1.6652,
+      "step": 8273
+    },
+    {
+      "epoch": 0.8594577750077906,
+      "grad_norm": 0.46544432640075684,
+      "learning_rate": 4.79497768633338e-06,
+      "loss": 1.6881,
+      "step": 8274
+    },
+    {
+      "epoch": 0.8595616495273709,
+      "grad_norm": 0.39384204149246216,
+      "learning_rate": 4.788007696153607e-06,
+      "loss": 1.6703,
+      "step": 8275
+    },
+    {
+      "epoch": 0.8596655240469513,
+      "grad_norm": 0.4230908155441284,
+      "learning_rate": 4.781042520698326e-06,
+      "loss": 1.7158,
+      "step": 8276
+    },
+    {
+      "epoch": 0.8597693985665317,
+      "grad_norm": 0.4715307652950287,
+      "learning_rate": 4.774082160709309e-06,
+      "loss": 1.518,
+      "step": 8277
+    },
+    {
+      "epoch": 0.8598732730861119,
+      "grad_norm": 0.42412781715393066,
+      "learning_rate": 4.767126616927769e-06,
+      "loss": 1.5671,
+      "step": 8278
+    },
+    {
+      "epoch": 0.8599771476056923,
+      "grad_norm": 0.4136991500854492,
+      "learning_rate": 4.760175890094399e-06,
+      "loss": 1.6172,
+      "step": 8279
+    },
+    {
+      "epoch": 0.8600810221252727,
+      "grad_norm": 0.4031031131744385,
+      "learning_rate": 4.753229980949419e-06,
+      "loss": 1.5905,
+      "step": 8280
+    },
+    {
+      "epoch": 0.860184896644853,
+      "grad_norm": 0.4419058859348297,
+      "learning_rate": 4.746288890232514e-06,
+      "loss": 1.6522,
+      "step": 8281
+    },
+    {
+      "epoch": 0.8602887711644334,
+      "grad_norm": 0.4122452139854431,
+      "learning_rate": 4.739352618682846e-06,
+      "loss": 1.6796,
+      "step": 8282
+    },
+    {
+      "epoch": 0.8603926456840137,
+      "grad_norm": 0.44128134846687317,
+      "learning_rate": 4.732421167039075e-06,
+      "loss": 1.6135,
+      "step": 8283
+    },
+    {
+      "epoch": 0.860496520203594,
+      "grad_norm": 0.4065702259540558,
+      "learning_rate": 4.725494536039354e-06,
+      "loss": 1.4627,
+      "step": 8284
+    },
+    {
+      "epoch": 0.8606003947231744,
+      "grad_norm": 0.40310558676719666,
+      "learning_rate": 4.718572726421305e-06,
+      "loss": 1.782,
+      "step": 8285
+    },
+    {
+      "epoch": 0.8607042692427548,
+      "grad_norm": 0.4221704602241516,
+      "learning_rate": 4.711655738922044e-06,
+      "loss": 1.6468,
+      "step": 8286
+    },
+    {
+      "epoch": 0.860808143762335,
+      "grad_norm": 0.41662177443504333,
+      "learning_rate": 4.704743574278209e-06,
+      "loss": 1.6226,
+      "step": 8287
+    },
+    {
+      "epoch": 0.8609120182819154,
+      "grad_norm": 0.42002618312835693,
+      "learning_rate": 4.697836233225838e-06,
+      "loss": 1.6135,
+      "step": 8288
+    },
+    {
+      "epoch": 0.8610158928014958,
+      "grad_norm": 0.4459064304828644,
+      "learning_rate": 4.690933716500556e-06,
+      "loss": 1.845,
+      "step": 8289
+    },
+    {
+      "epoch": 0.8611197673210761,
+      "grad_norm": 0.43704405426979065,
+      "learning_rate": 4.684036024837402e-06,
+      "loss": 1.7574,
+      "step": 8290
+    },
+    {
+      "epoch": 0.8612236418406565,
+      "grad_norm": 0.3958377540111542,
+      "learning_rate": 4.677143158970937e-06,
+      "loss": 1.5884,
+      "step": 8291
+    },
+    {
+      "epoch": 0.8613275163602369,
+      "grad_norm": 0.4173584580421448,
+      "learning_rate": 4.670255119635192e-06,
+      "loss": 1.6082,
+      "step": 8292
+    },
+    {
+      "epoch": 0.8614313908798171,
+      "grad_norm": 0.4080117642879486,
+      "learning_rate": 4.663371907563696e-06,
+      "loss": 1.5688,
+      "step": 8293
+    },
+    {
+      "epoch": 0.8615352653993975,
+      "grad_norm": 0.4516976773738861,
+      "learning_rate": 4.656493523489447e-06,
+      "loss": 1.7237,
+      "step": 8294
+    },
+    {
+      "epoch": 0.8616391399189779,
+      "grad_norm": 0.4905436635017395,
+      "learning_rate": 4.649619968144936e-06,
+      "loss": 1.6905,
+      "step": 8295
+    },
+    {
+      "epoch": 0.8617430144385583,
+      "grad_norm": 0.4043090045452118,
+      "learning_rate": 4.642751242262161e-06,
+      "loss": 1.5359,
+      "step": 8296
+    },
+    {
+      "epoch": 0.8618468889581385,
+      "grad_norm": 0.4110516905784607,
+      "learning_rate": 4.635887346572582e-06,
+      "loss": 1.6373,
+      "step": 8297
+    },
+    {
+      "epoch": 0.8619507634777189,
+      "grad_norm": 0.4277547299861908,
+      "learning_rate": 4.629028281807146e-06,
+      "loss": 1.7373,
+      "step": 8298
+    },
+    {
+      "epoch": 0.8620546379972993,
+      "grad_norm": 0.3978832960128784,
+      "learning_rate": 4.622174048696288e-06,
+      "loss": 1.456,
+      "step": 8299
+    },
+    {
+      "epoch": 0.8621585125168796,
+      "grad_norm": 0.4189075231552124,
+      "learning_rate": 4.615324647969932e-06,
+      "loss": 1.7349,
+      "step": 8300
+    },
+    {
+      "epoch": 0.86226238703646,
+      "grad_norm": 0.4180492162704468,
+      "learning_rate": 4.608480080357491e-06,
+      "loss": 1.4529,
+      "step": 8301
+    },
+    {
+      "epoch": 0.8623662615560403,
+      "grad_norm": 0.4220123291015625,
+      "learning_rate": 4.601640346587843e-06,
+      "loss": 1.7035,
+      "step": 8302
+    },
+    {
+      "epoch": 0.8624701360756206,
+      "grad_norm": 0.4284611642360687,
+      "learning_rate": 4.5948054473894e-06,
+      "loss": 1.6631,
+      "step": 8303
+    },
+    {
+      "epoch": 0.862574010595201,
+      "grad_norm": 0.4400910437107086,
+      "learning_rate": 4.587975383489978e-06,
+      "loss": 1.6642,
+      "step": 8304
+    },
+    {
+      "epoch": 0.8626778851147814,
+      "grad_norm": 0.4507836699485779,
+      "learning_rate": 4.581150155616959e-06,
+      "loss": 1.79,
+      "step": 8305
+    },
+    {
+      "epoch": 0.8627817596343617,
+      "grad_norm": 0.4717456102371216,
+      "learning_rate": 4.574329764497182e-06,
+      "loss": 1.7097,
+      "step": 8306
+    },
+    {
+      "epoch": 0.862885634153942,
+      "grad_norm": 0.42623621225357056,
+      "learning_rate": 4.567514210856933e-06,
+      "loss": 1.6195,
+      "step": 8307
+    },
+    {
+      "epoch": 0.8629895086735224,
+      "grad_norm": 0.3888145387172699,
+      "learning_rate": 4.560703495422037e-06,
+      "loss": 1.4697,
+      "step": 8308
+    },
+    {
+      "epoch": 0.8630933831931027,
+      "grad_norm": 0.40128466486930847,
+      "learning_rate": 4.553897618917785e-06,
+      "loss": 1.6035,
+      "step": 8309
+    },
+    {
+      "epoch": 0.8631972577126831,
+      "grad_norm": 0.43197745084762573,
+      "learning_rate": 4.547096582068938e-06,
+      "loss": 1.6968,
+      "step": 8310
+    },
+    {
+      "epoch": 0.8633011322322635,
+      "grad_norm": 0.4532462954521179,
+      "learning_rate": 4.540300385599761e-06,
+      "loss": 1.8258,
+      "step": 8311
+    },
+    {
+      "epoch": 0.8634050067518437,
+      "grad_norm": 0.3836333453655243,
+      "learning_rate": 4.533509030233995e-06,
+      "loss": 1.6876,
+      "step": 8312
+    },
+    {
+      "epoch": 0.8635088812714241,
+      "grad_norm": 0.44952192902565,
+      "learning_rate": 4.5267225166948645e-06,
+      "loss": 1.7243,
+      "step": 8313
+    },
+    {
+      "epoch": 0.8636127557910045,
+      "grad_norm": 0.4556354284286499,
+      "learning_rate": 4.519940845705067e-06,
+      "loss": 1.8883,
+      "step": 8314
+    },
+    {
+      "epoch": 0.8637166303105848,
+      "grad_norm": 0.40353667736053467,
+      "learning_rate": 4.513164017986837e-06,
+      "loss": 1.4688,
+      "step": 8315
+    },
+    {
+      "epoch": 0.8638205048301651,
+      "grad_norm": 0.4035186171531677,
+      "learning_rate": 4.506392034261803e-06,
+      "loss": 1.7138,
+      "step": 8316
+    },
+    {
+      "epoch": 0.8639243793497455,
+      "grad_norm": 0.4350834786891937,
+      "learning_rate": 4.49962489525117e-06,
+      "loss": 1.6899,
+      "step": 8317
+    },
+    {
+      "epoch": 0.8640282538693258,
+      "grad_norm": 0.4328579902648926,
+      "learning_rate": 4.492862601675563e-06,
+      "loss": 1.5958,
+      "step": 8318
+    },
+    {
+      "epoch": 0.8641321283889062,
+      "grad_norm": 0.45060819387435913,
+      "learning_rate": 4.486105154255121e-06,
+      "loss": 1.8196,
+      "step": 8319
+    },
+    {
+      "epoch": 0.8642360029084866,
+      "grad_norm": 0.4311492443084717,
+      "learning_rate": 4.4793525537094595e-06,
+      "loss": 1.7509,
+      "step": 8320
+    },
+    {
+      "epoch": 0.864339877428067,
+      "grad_norm": 0.45266759395599365,
+      "learning_rate": 4.472604800757668e-06,
+      "loss": 1.813,
+      "step": 8321
+    },
+    {
+      "epoch": 0.8644437519476472,
+      "grad_norm": 0.42085716128349304,
+      "learning_rate": 4.4658618961183506e-06,
+      "loss": 1.6671,
+      "step": 8322
+    },
+    {
+      "epoch": 0.8645476264672276,
+      "grad_norm": 0.4207400381565094,
+      "learning_rate": 4.459123840509549e-06,
+      "loss": 1.5313,
+      "step": 8323
+    },
+    {
+      "epoch": 0.864651500986808,
+      "grad_norm": 0.4059183895587921,
+      "learning_rate": 4.4523906346488345e-06,
+      "loss": 1.5305,
+      "step": 8324
+    },
+    {
+      "epoch": 0.8647553755063883,
+      "grad_norm": 0.40492719411849976,
+      "learning_rate": 4.445662279253226e-06,
+      "loss": 1.5326,
+      "step": 8325
+    },
+    {
+      "epoch": 0.8648592500259686,
+      "grad_norm": 0.43756330013275146,
+      "learning_rate": 4.438938775039253e-06,
+      "loss": 1.7945,
+      "step": 8326
+    },
+    {
+      "epoch": 0.864963124545549,
+      "grad_norm": 0.44969308376312256,
+      "learning_rate": 4.4322201227229134e-06,
+      "loss": 1.7237,
+      "step": 8327
+    },
+    {
+      "epoch": 0.8650669990651293,
+      "grad_norm": 0.4445195198059082,
+      "learning_rate": 4.425506323019685e-06,
+      "loss": 1.7394,
+      "step": 8328
+    },
+    {
+      "epoch": 0.8651708735847097,
+      "grad_norm": 0.5065091252326965,
+      "learning_rate": 4.418797376644534e-06,
+      "loss": 1.6003,
+      "step": 8329
+    },
+    {
+      "epoch": 0.8652747481042901,
+      "grad_norm": 0.4191296696662903,
+      "learning_rate": 4.412093284311913e-06,
+      "loss": 1.5485,
+      "step": 8330
+    },
+    {
+      "epoch": 0.8653786226238703,
+      "grad_norm": 0.4144963026046753,
+      "learning_rate": 4.405394046735773e-06,
+      "loss": 1.5195,
+      "step": 8331
+    },
+    {
+      "epoch": 0.8654824971434507,
+      "grad_norm": 0.42345312237739563,
+      "learning_rate": 4.398699664629497e-06,
+      "loss": 1.7706,
+      "step": 8332
+    },
+    {
+      "epoch": 0.8655863716630311,
+      "grad_norm": 0.4790879189968109,
+      "learning_rate": 4.392010138706015e-06,
+      "loss": 1.6909,
+      "step": 8333
+    },
+    {
+      "epoch": 0.8656902461826114,
+      "grad_norm": 0.39938417077064514,
+      "learning_rate": 4.3853254696777036e-06,
+      "loss": 1.6691,
+      "step": 8334
+    },
+    {
+      "epoch": 0.8657941207021917,
+      "grad_norm": 0.382337361574173,
+      "learning_rate": 4.378645658256403e-06,
+      "loss": 1.5524,
+      "step": 8335
+    },
+    {
+      "epoch": 0.8658979952217721,
+      "grad_norm": 0.41304081678390503,
+      "learning_rate": 4.371970705153489e-06,
+      "loss": 1.6553,
+      "step": 8336
+    },
+    {
+      "epoch": 0.8660018697413524,
+      "grad_norm": 0.44353777170181274,
+      "learning_rate": 4.365300611079781e-06,
+      "loss": 1.7254,
+      "step": 8337
+    },
+    {
+      "epoch": 0.8661057442609328,
+      "grad_norm": 0.4305352568626404,
+      "learning_rate": 4.358635376745601e-06,
+      "loss": 1.6613,
+      "step": 8338
+    },
+    {
+      "epoch": 0.8662096187805132,
+      "grad_norm": 0.4181787371635437,
+      "learning_rate": 4.351975002860731e-06,
+      "loss": 1.6243,
+      "step": 8339
+    },
+    {
+      "epoch": 0.8663134933000934,
+      "grad_norm": 0.49580061435699463,
+      "learning_rate": 4.345319490134453e-06,
+      "loss": 1.6895,
+      "step": 8340
+    },
+    {
+      "epoch": 0.8664173678196738,
+      "grad_norm": 0.422701358795166,
+      "learning_rate": 4.338668839275534e-06,
+      "loss": 1.6746,
+      "step": 8341
+    },
+    {
+      "epoch": 0.8665212423392542,
+      "grad_norm": 0.4329741299152374,
+      "learning_rate": 4.3320230509922014e-06,
+      "loss": 1.6391,
+      "step": 8342
+    },
+    {
+      "epoch": 0.8666251168588345,
+      "grad_norm": 0.40780285000801086,
+      "learning_rate": 4.3253821259922055e-06,
+      "loss": 1.5569,
+      "step": 8343
+    },
+    {
+      "epoch": 0.8667289913784149,
+      "grad_norm": 0.42592713236808777,
+      "learning_rate": 4.318746064982721e-06,
+      "loss": 1.6763,
+      "step": 8344
+    },
+    {
+      "epoch": 0.8668328658979952,
+      "grad_norm": 0.391530305147171,
+      "learning_rate": 4.312114868670458e-06,
+      "loss": 1.4977,
+      "step": 8345
+    },
+    {
+      "epoch": 0.8669367404175756,
+      "grad_norm": 0.41570156812667847,
+      "learning_rate": 4.305488537761588e-06,
+      "loss": 1.586,
+      "step": 8346
+    },
+    {
+      "epoch": 0.8670406149371559,
+      "grad_norm": 0.42140865325927734,
+      "learning_rate": 4.298867072961754e-06,
+      "loss": 1.6925,
+      "step": 8347
+    },
+    {
+      "epoch": 0.8671444894567363,
+      "grad_norm": 0.4236295819282532,
+      "learning_rate": 4.2922504749761005e-06,
+      "loss": 1.5939,
+      "step": 8348
+    },
+    {
+      "epoch": 0.8672483639763167,
+      "grad_norm": 0.4136272966861725,
+      "learning_rate": 4.2856387445092235e-06,
+      "loss": 1.6401,
+      "step": 8349
+    },
+    {
+      "epoch": 0.8673522384958969,
+      "grad_norm": 0.4001111090183258,
+      "learning_rate": 4.2790318822652495e-06,
+      "loss": 1.4537,
+      "step": 8350
+    },
+    {
+      "epoch": 0.8674561130154773,
+      "grad_norm": 0.40965673327445984,
+      "learning_rate": 4.272429888947732e-06,
+      "loss": 1.472,
+      "step": 8351
+    },
+    {
+      "epoch": 0.8675599875350577,
+      "grad_norm": 0.4452982544898987,
+      "learning_rate": 4.265832765259748e-06,
+      "loss": 1.7643,
+      "step": 8352
+    },
+    {
+      "epoch": 0.867663862054638,
+      "grad_norm": 0.41096264123916626,
+      "learning_rate": 4.259240511903834e-06,
+      "loss": 1.6081,
+      "step": 8353
+    },
+    {
+      "epoch": 0.8677677365742184,
+      "grad_norm": 0.4857504367828369,
+      "learning_rate": 4.252653129582013e-06,
+      "loss": 1.8201,
+      "step": 8354
+    },
+    {
+      "epoch": 0.8678716110937987,
+      "grad_norm": 0.3969123959541321,
+      "learning_rate": 4.2460706189957895e-06,
+      "loss": 1.5116,
+      "step": 8355
+    },
+    {
+      "epoch": 0.867975485613379,
+      "grad_norm": 0.40044331550598145,
+      "learning_rate": 4.239492980846155e-06,
+      "loss": 1.772,
+      "step": 8356
+    },
+    {
+      "epoch": 0.8680793601329594,
+      "grad_norm": 0.47415369749069214,
+      "learning_rate": 4.232920215833569e-06,
+      "loss": 1.9206,
+      "step": 8357
+    },
+    {
+      "epoch": 0.8681832346525398,
+      "grad_norm": 0.41735053062438965,
+      "learning_rate": 4.226352324657973e-06,
+      "loss": 1.7219,
+      "step": 8358
+    },
+    {
+      "epoch": 0.86828710917212,
+      "grad_norm": 0.43952199816703796,
+      "learning_rate": 4.219789308018829e-06,
+      "loss": 1.8179,
+      "step": 8359
+    },
+    {
+      "epoch": 0.8683909836917004,
+      "grad_norm": 0.42207950353622437,
+      "learning_rate": 4.2132311666150005e-06,
+      "loss": 1.6541,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8684948582112808,
+      "grad_norm": 0.3974553942680359,
+      "learning_rate": 4.206677901144917e-06,
+      "loss": 1.5655,
+      "step": 8361
+    },
+    {
+      "epoch": 0.8685987327308611,
+      "grad_norm": 0.4383615255355835,
+      "learning_rate": 4.200129512306439e-06,
+      "loss": 1.707,
+      "step": 8362
+    },
+    {
+      "epoch": 0.8687026072504415,
+      "grad_norm": 0.43259862065315247,
+      "learning_rate": 4.1935860007969005e-06,
+      "loss": 1.6708,
+      "step": 8363
+    },
+    {
+      "epoch": 0.8688064817700218,
+      "grad_norm": 0.4056100845336914,
+      "learning_rate": 4.187047367313157e-06,
+      "loss": 1.5422,
+      "step": 8364
+    },
+    {
+      "epoch": 0.8689103562896021,
+      "grad_norm": 0.3907129466533661,
+      "learning_rate": 4.180513612551517e-06,
+      "loss": 1.5345,
+      "step": 8365
+    },
+    {
+      "epoch": 0.8690142308091825,
+      "grad_norm": 0.3913259208202362,
+      "learning_rate": 4.17398473720777e-06,
+      "loss": 1.5311,
+      "step": 8366
+    },
+    {
+      "epoch": 0.8691181053287629,
+      "grad_norm": 0.4376855790615082,
+      "learning_rate": 4.1674607419771915e-06,
+      "loss": 1.4414,
+      "step": 8367
+    },
+    {
+      "epoch": 0.8692219798483432,
+      "grad_norm": 0.44017404317855835,
+      "learning_rate": 4.160941627554538e-06,
+      "loss": 1.4987,
+      "step": 8368
+    },
+    {
+      "epoch": 0.8693258543679235,
+      "grad_norm": 0.4197766184806824,
+      "learning_rate": 4.154427394634036e-06,
+      "loss": 1.6341,
+      "step": 8369
+    },
+    {
+      "epoch": 0.8694297288875039,
+      "grad_norm": 0.41272974014282227,
+      "learning_rate": 4.147918043909404e-06,
+      "loss": 1.7277,
+      "step": 8370
+    },
+    {
+      "epoch": 0.8695336034070842,
+      "grad_norm": 0.4294355511665344,
+      "learning_rate": 4.14141357607386e-06,
+      "loss": 1.6332,
+      "step": 8371
+    },
+    {
+      "epoch": 0.8696374779266646,
+      "grad_norm": 0.43766576051712036,
+      "learning_rate": 4.1349139918200385e-06,
+      "loss": 1.6576,
+      "step": 8372
+    },
+    {
+      "epoch": 0.869741352446245,
+      "grad_norm": 0.39350345730781555,
+      "learning_rate": 4.1284192918401245e-06,
+      "loss": 1.517,
+      "step": 8373
+    },
+    {
+      "epoch": 0.8698452269658253,
+      "grad_norm": 0.41571328043937683,
+      "learning_rate": 4.121929476825742e-06,
+      "loss": 1.6047,
+      "step": 8374
+    },
+    {
+      "epoch": 0.8699491014854056,
+      "grad_norm": 0.4371244013309479,
+      "learning_rate": 4.1154445474680115e-06,
+      "loss": 1.6017,
+      "step": 8375
+    },
+    {
+      "epoch": 0.870052976004986,
+      "grad_norm": 0.4310753643512726,
+      "learning_rate": 4.108964504457524e-06,
+      "loss": 1.7602,
+      "step": 8376
+    },
+    {
+      "epoch": 0.8701568505245664,
+      "grad_norm": 0.3955596387386322,
+      "learning_rate": 4.102489348484339e-06,
+      "loss": 1.542,
+      "step": 8377
+    },
+    {
+      "epoch": 0.8702607250441466,
+      "grad_norm": 0.3818468749523163,
+      "learning_rate": 4.096019080238045e-06,
+      "loss": 1.4748,
+      "step": 8378
+    },
+    {
+      "epoch": 0.870364599563727,
+      "grad_norm": 0.4244476854801178,
+      "learning_rate": 4.089553700407634e-06,
+      "loss": 1.6876,
+      "step": 8379
+    },
+    {
+      "epoch": 0.8704684740833074,
+      "grad_norm": 0.44458895921707153,
+      "learning_rate": 4.083093209681649e-06,
+      "loss": 1.6637,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8705723486028877,
+      "grad_norm": 0.46665099263191223,
+      "learning_rate": 4.076637608748074e-06,
+      "loss": 1.4212,
+      "step": 8381
+    },
+    {
+      "epoch": 0.8706762231224681,
+      "grad_norm": 0.43539130687713623,
+      "learning_rate": 4.070186898294375e-06,
+      "loss": 1.7361,
+      "step": 8382
+    },
+    {
+      "epoch": 0.8707800976420484,
+      "grad_norm": 0.4533788561820984,
+      "learning_rate": 4.063741079007505e-06,
+      "loss": 1.9468,
+      "step": 8383
+    },
+    {
+      "epoch": 0.8708839721616287,
+      "grad_norm": 0.40480610728263855,
+      "learning_rate": 4.057300151573895e-06,
+      "loss": 1.6335,
+      "step": 8384
+    },
+    {
+      "epoch": 0.8709878466812091,
+      "grad_norm": 0.5002861618995667,
+      "learning_rate": 4.050864116679454e-06,
+      "loss": 1.8369,
+      "step": 8385
+    },
+    {
+      "epoch": 0.8710917212007895,
+      "grad_norm": 0.41242918372154236,
+      "learning_rate": 4.044432975009554e-06,
+      "loss": 1.5588,
+      "step": 8386
+    },
+    {
+      "epoch": 0.8711955957203698,
+      "grad_norm": 0.43866342306137085,
+      "learning_rate": 4.038006727249094e-06,
+      "loss": 1.6729,
+      "step": 8387
+    },
+    {
+      "epoch": 0.8712994702399501,
+      "grad_norm": 0.4091012477874756,
+      "learning_rate": 4.031585374082381e-06,
+      "loss": 1.6353,
+      "step": 8388
+    },
+    {
+      "epoch": 0.8714033447595305,
+      "grad_norm": 0.44972357153892517,
+      "learning_rate": 4.025168916193273e-06,
+      "loss": 1.7602,
+      "step": 8389
+    },
+    {
+      "epoch": 0.8715072192791108,
+      "grad_norm": 0.3911556005477905,
+      "learning_rate": 4.0187573542650625e-06,
+      "loss": 1.5152,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8716110937986912,
+      "grad_norm": 0.4245710074901581,
+      "learning_rate": 4.01235068898051e-06,
+      "loss": 1.775,
+      "step": 8391
+    },
+    {
+      "epoch": 0.8717149683182716,
+      "grad_norm": 0.41112756729125977,
+      "learning_rate": 4.005948921021896e-06,
+      "loss": 1.5729,
+      "step": 8392
+    },
+    {
+      "epoch": 0.8718188428378518,
+      "grad_norm": 0.42917004227638245,
+      "learning_rate": 3.99955205107096e-06,
+      "loss": 1.7396,
+      "step": 8393
+    },
+    {
+      "epoch": 0.8719227173574322,
+      "grad_norm": 0.44293302297592163,
+      "learning_rate": 3.99316007980891e-06,
+      "loss": 1.7512,
+      "step": 8394
+    },
+    {
+      "epoch": 0.8720265918770126,
+      "grad_norm": 0.41871437430381775,
+      "learning_rate": 3.9867730079164445e-06,
+      "loss": 1.5736,
+      "step": 8395
+    },
+    {
+      "epoch": 0.8721304663965929,
+      "grad_norm": 0.4064503312110901,
+      "learning_rate": 3.980390836073739e-06,
+      "loss": 1.4956,
+      "step": 8396
+    },
+    {
+      "epoch": 0.8722343409161732,
+      "grad_norm": 0.39449822902679443,
+      "learning_rate": 3.974013564960444e-06,
+      "loss": 1.5223,
+      "step": 8397
+    },
+    {
+      "epoch": 0.8723382154357536,
+      "grad_norm": 0.3918604850769043,
+      "learning_rate": 3.967641195255672e-06,
+      "loss": 1.5784,
+      "step": 8398
+    },
+    {
+      "epoch": 0.872442089955334,
+      "grad_norm": 0.42740651965141296,
+      "learning_rate": 3.9612737276380595e-06,
+      "loss": 1.652,
+      "step": 8399
+    },
+    {
+      "epoch": 0.8725459644749143,
+      "grad_norm": 0.45657414197921753,
+      "learning_rate": 3.9549111627856795e-06,
+      "loss": 1.863,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8726498389944947,
+      "grad_norm": 0.4107915759086609,
+      "learning_rate": 3.948553501376096e-06,
+      "loss": 1.6226,
+      "step": 8401
+    },
+    {
+      "epoch": 0.872753713514075,
+      "grad_norm": 0.43051251769065857,
+      "learning_rate": 3.942200744086349e-06,
+      "loss": 1.687,
+      "step": 8402
+    },
+    {
+      "epoch": 0.8728575880336553,
+      "grad_norm": 0.3889218270778656,
+      "learning_rate": 3.935852891592956e-06,
+      "loss": 1.6296,
+      "step": 8403
+    },
+    {
+      "epoch": 0.8729614625532357,
+      "grad_norm": 0.4343833327293396,
+      "learning_rate": 3.929509944571918e-06,
+      "loss": 1.7,
+      "step": 8404
+    },
+    {
+      "epoch": 0.8730653370728161,
+      "grad_norm": 0.43371620774269104,
+      "learning_rate": 3.9231719036987025e-06,
+      "loss": 1.627,
+      "step": 8405
+    },
+    {
+      "epoch": 0.8731692115923964,
+      "grad_norm": 0.4425232410430908,
+      "learning_rate": 3.916838769648279e-06,
+      "loss": 1.6984,
+      "step": 8406
+    },
+    {
+      "epoch": 0.8732730861119767,
+      "grad_norm": 0.3910176455974579,
+      "learning_rate": 3.910510543095047e-06,
+      "loss": 1.5953,
+      "step": 8407
+    },
+    {
+      "epoch": 0.8733769606315571,
+      "grad_norm": 0.4370116591453552,
+      "learning_rate": 3.904187224712935e-06,
+      "loss": 1.8256,
+      "step": 8408
+    },
+    {
+      "epoch": 0.8734808351511374,
+      "grad_norm": 0.4318406581878662,
+      "learning_rate": 3.897868815175326e-06,
+      "loss": 1.6616,
+      "step": 8409
+    },
+    {
+      "epoch": 0.8735847096707178,
+      "grad_norm": 0.38654428720474243,
+      "learning_rate": 3.891555315155082e-06,
+      "loss": 1.4936,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8736885841902982,
+      "grad_norm": 0.41681328415870667,
+      "learning_rate": 3.885246725324532e-06,
+      "loss": 1.5365,
+      "step": 8411
+    },
+    {
+      "epoch": 0.8737924587098784,
+      "grad_norm": 0.39945271611213684,
+      "learning_rate": 3.878943046355499e-06,
+      "loss": 1.5884,
+      "step": 8412
+    },
+    {
+      "epoch": 0.8738963332294588,
+      "grad_norm": 0.40073657035827637,
+      "learning_rate": 3.872644278919274e-06,
+      "loss": 1.6765,
+      "step": 8413
+    },
+    {
+      "epoch": 0.8740002077490392,
+      "grad_norm": 0.4032048285007477,
+      "learning_rate": 3.866350423686615e-06,
+      "loss": 1.6804,
+      "step": 8414
+    },
+    {
+      "epoch": 0.8741040822686195,
+      "grad_norm": 0.4637449085712433,
+      "learning_rate": 3.860061481327798e-06,
+      "loss": 1.7906,
+      "step": 8415
+    },
+    {
+      "epoch": 0.8742079567881998,
+      "grad_norm": 0.4216407537460327,
+      "learning_rate": 3.8537774525125195e-06,
+      "loss": 1.7432,
+      "step": 8416
+    },
+    {
+      "epoch": 0.8743118313077802,
+      "grad_norm": 0.4351256787776947,
+      "learning_rate": 3.847498337909977e-06,
+      "loss": 1.6909,
+      "step": 8417
+    },
+    {
+      "epoch": 0.8744157058273605,
+      "grad_norm": 0.4363458454608917,
+      "learning_rate": 3.8412241381888755e-06,
+      "loss": 1.5617,
+      "step": 8418
+    },
+    {
+      "epoch": 0.8745195803469409,
+      "grad_norm": 0.481900155544281,
+      "learning_rate": 3.834954854017337e-06,
+      "loss": 1.854,
+      "step": 8419
+    },
+    {
+      "epoch": 0.8746234548665213,
+      "grad_norm": 0.49199458956718445,
+      "learning_rate": 3.828690486063013e-06,
+      "loss": 1.7667,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8747273293861015,
+      "grad_norm": 0.41270673274993896,
+      "learning_rate": 3.822431034993002e-06,
+      "loss": 1.623,
+      "step": 8421
+    },
+    {
+      "epoch": 0.8748312039056819,
+      "grad_norm": 0.444240540266037,
+      "learning_rate": 3.816176501473889e-06,
+      "loss": 1.6026,
+      "step": 8422
+    },
+    {
+      "epoch": 0.8749350784252623,
+      "grad_norm": 0.4405142366886139,
+      "learning_rate": 3.8099268861717275e-06,
+      "loss": 1.7327,
+      "step": 8423
+    },
+    {
+      "epoch": 0.8750389529448427,
+      "grad_norm": 0.4171452820301056,
+      "learning_rate": 3.8036821897520502e-06,
+      "loss": 1.622,
+      "step": 8424
+    },
+    {
+      "epoch": 0.875142827464423,
+      "grad_norm": 0.4365280866622925,
+      "learning_rate": 3.7974424128798868e-06,
+      "loss": 1.5021,
+      "step": 8425
+    },
+    {
+      "epoch": 0.8752467019840033,
+      "grad_norm": 0.419349730014801,
+      "learning_rate": 3.7912075562197e-06,
+      "loss": 1.5933,
+      "step": 8426
+    },
+    {
+      "epoch": 0.8753505765035837,
+      "grad_norm": 0.41351646184921265,
+      "learning_rate": 3.784977620435476e-06,
+      "loss": 1.63,
+      "step": 8427
+    },
+    {
+      "epoch": 0.875454451023164,
+      "grad_norm": 0.4430893361568451,
+      "learning_rate": 3.7787526061906387e-06,
+      "loss": 1.655,
+      "step": 8428
+    },
+    {
+      "epoch": 0.8755583255427444,
+      "grad_norm": 0.534438967704773,
+      "learning_rate": 3.7725325141481137e-06,
+      "loss": 1.8994,
+      "step": 8429
+    },
+    {
+      "epoch": 0.8756622000623248,
+      "grad_norm": 0.44221749901771545,
+      "learning_rate": 3.766317344970288e-06,
+      "loss": 1.6722,
+      "step": 8430
+    },
+    {
+      "epoch": 0.875766074581905,
+      "grad_norm": 0.42250174283981323,
+      "learning_rate": 3.7601070993190267e-06,
+      "loss": 1.6381,
+      "step": 8431
+    },
+    {
+      "epoch": 0.8758699491014854,
+      "grad_norm": 0.45283105969429016,
+      "learning_rate": 3.753901777855673e-06,
+      "loss": 1.6305,
+      "step": 8432
+    },
+    {
+      "epoch": 0.8759738236210658,
+      "grad_norm": 0.4005245268344879,
+      "learning_rate": 3.7477013812410422e-06,
+      "loss": 1.6229,
+      "step": 8433
+    },
+    {
+      "epoch": 0.8760776981406461,
+      "grad_norm": 0.3979766368865967,
+      "learning_rate": 3.7415059101354454e-06,
+      "loss": 1.5569,
+      "step": 8434
+    },
+    {
+      "epoch": 0.8761815726602264,
+      "grad_norm": 0.4287016987800598,
+      "learning_rate": 3.735315365198622e-06,
+      "loss": 1.5411,
+      "step": 8435
+    },
+    {
+      "epoch": 0.8762854471798068,
+      "grad_norm": 0.46041205525398254,
+      "learning_rate": 3.7291297470898435e-06,
+      "loss": 1.5864,
+      "step": 8436
+    },
+    {
+      "epoch": 0.8763893216993871,
+      "grad_norm": 0.4164537787437439,
+      "learning_rate": 3.7229490564678227e-06,
+      "loss": 1.6638,
+      "step": 8437
+    },
+    {
+      "epoch": 0.8764931962189675,
+      "grad_norm": 0.3900989294052124,
+      "learning_rate": 3.7167732939907496e-06,
+      "loss": 1.497,
+      "step": 8438
+    },
+    {
+      "epoch": 0.8765970707385479,
+      "grad_norm": 0.41212278604507446,
+      "learning_rate": 3.710602460316298e-06,
+      "loss": 1.5113,
+      "step": 8439
+    },
+    {
+      "epoch": 0.8767009452581281,
+      "grad_norm": 0.39359238743782043,
+      "learning_rate": 3.704436556101615e-06,
+      "loss": 1.5233,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8768048197777085,
+      "grad_norm": 0.46984294056892395,
+      "learning_rate": 3.6982755820033186e-06,
+      "loss": 1.6151,
+      "step": 8441
+    },
+    {
+      "epoch": 0.8769086942972889,
+      "grad_norm": 0.4061124324798584,
+      "learning_rate": 3.6921195386774965e-06,
+      "loss": 1.5301,
+      "step": 8442
+    },
+    {
+      "epoch": 0.8770125688168692,
+      "grad_norm": 0.4347725808620453,
+      "learning_rate": 3.6859684267797455e-06,
+      "loss": 1.6463,
+      "step": 8443
+    },
+    {
+      "epoch": 0.8771164433364496,
+      "grad_norm": 0.46318212151527405,
+      "learning_rate": 3.679822246965087e-06,
+      "loss": 1.4855,
+      "step": 8444
+    },
+    {
+      "epoch": 0.8772203178560299,
+      "grad_norm": 0.4061053693294525,
+      "learning_rate": 3.673680999888035e-06,
+      "loss": 1.5532,
+      "step": 8445
+    },
+    {
+      "epoch": 0.8773241923756102,
+      "grad_norm": 0.3995457589626312,
+      "learning_rate": 3.6675446862026176e-06,
+      "loss": 1.5249,
+      "step": 8446
+    },
+    {
+      "epoch": 0.8774280668951906,
+      "grad_norm": 0.4302307665348053,
+      "learning_rate": 3.661413306562267e-06,
+      "loss": 1.6669,
+      "step": 8447
+    },
+    {
+      "epoch": 0.877531941414771,
+      "grad_norm": 0.45854660868644714,
+      "learning_rate": 3.65528686161995e-06,
+      "loss": 1.7169,
+      "step": 8448
+    },
+    {
+      "epoch": 0.8776358159343512,
+      "grad_norm": 0.4778292179107666,
+      "learning_rate": 3.649165352028083e-06,
+      "loss": 1.6258,
+      "step": 8449
+    },
+    {
+      "epoch": 0.8777396904539316,
+      "grad_norm": 0.3999958038330078,
+      "learning_rate": 3.6430487784385515e-06,
+      "loss": 1.6233,
+      "step": 8450
+    },
+    {
+      "epoch": 0.877843564973512,
+      "grad_norm": 0.40005001425743103,
+      "learning_rate": 3.636937141502733e-06,
+      "loss": 1.6136,
+      "step": 8451
+    },
+    {
+      "epoch": 0.8779474394930924,
+      "grad_norm": 0.4642762839794159,
+      "learning_rate": 3.6308304418714523e-06,
+      "loss": 1.8124,
+      "step": 8452
+    },
+    {
+      "epoch": 0.8780513140126727,
+      "grad_norm": 0.41780030727386475,
+      "learning_rate": 3.6247286801950497e-06,
+      "loss": 1.708,
+      "step": 8453
+    },
+    {
+      "epoch": 0.878155188532253,
+      "grad_norm": 0.3930474817752838,
+      "learning_rate": 3.6186318571232893e-06,
+      "loss": 1.5891,
+      "step": 8454
+    },
+    {
+      "epoch": 0.8782590630518334,
+      "grad_norm": 0.38920000195503235,
+      "learning_rate": 3.612539973305451e-06,
+      "loss": 1.6752,
+      "step": 8455
+    },
+    {
+      "epoch": 0.8783629375714137,
+      "grad_norm": 0.41854655742645264,
+      "learning_rate": 3.606453029390272e-06,
+      "loss": 1.6183,
+      "step": 8456
+    },
+    {
+      "epoch": 0.8784668120909941,
+      "grad_norm": 0.4399711489677429,
+      "learning_rate": 3.6003710260259616e-06,
+      "loss": 1.9086,
+      "step": 8457
+    },
+    {
+      "epoch": 0.8785706866105745,
+      "grad_norm": 0.4335370659828186,
+      "learning_rate": 3.5942939638602067e-06,
+      "loss": 1.5987,
+      "step": 8458
+    },
+    {
+      "epoch": 0.8786745611301547,
+      "grad_norm": 0.4267464280128479,
+      "learning_rate": 3.588221843540168e-06,
+      "loss": 1.7094,
+      "step": 8459
+    },
+    {
+      "epoch": 0.8787784356497351,
+      "grad_norm": 0.4027010500431061,
+      "learning_rate": 3.582154665712473e-06,
+      "loss": 1.6371,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8788823101693155,
+      "grad_norm": 0.4093351662158966,
+      "learning_rate": 3.5760924310232258e-06,
+      "loss": 1.72,
+      "step": 8461
+    },
+    {
+      "epoch": 0.8789861846888958,
+      "grad_norm": 0.4228046238422394,
+      "learning_rate": 3.5700351401180333e-06,
+      "loss": 1.769,
+      "step": 8462
+    },
+    {
+      "epoch": 0.8790900592084762,
+      "grad_norm": 0.4321579337120056,
+      "learning_rate": 3.5639827936419123e-06,
+      "loss": 1.7098,
+      "step": 8463
+    },
+    {
+      "epoch": 0.8791939337280565,
+      "grad_norm": 0.45898208022117615,
+      "learning_rate": 3.5579353922394197e-06,
+      "loss": 1.7701,
+      "step": 8464
+    },
+    {
+      "epoch": 0.8792978082476368,
+      "grad_norm": 0.44818854331970215,
+      "learning_rate": 3.5518929365545406e-06,
+      "loss": 1.7982,
+      "step": 8465
+    },
+    {
+      "epoch": 0.8794016827672172,
+      "grad_norm": 0.3915591835975647,
+      "learning_rate": 3.5458554272307597e-06,
+      "loss": 1.6386,
+      "step": 8466
+    },
+    {
+      "epoch": 0.8795055572867976,
+      "grad_norm": 0.4124680459499359,
+      "learning_rate": 3.5398228649110186e-06,
+      "loss": 1.6979,
+      "step": 8467
+    },
+    {
+      "epoch": 0.8796094318063779,
+      "grad_norm": 0.4365387558937073,
+      "learning_rate": 3.533795250237737e-06,
+      "loss": 1.4209,
+      "step": 8468
+    },
+    {
+      "epoch": 0.8797133063259582,
+      "grad_norm": 0.4055129885673523,
+      "learning_rate": 3.527772583852812e-06,
+      "loss": 1.4331,
+      "step": 8469
+    },
+    {
+      "epoch": 0.8798171808455386,
+      "grad_norm": 0.4493137001991272,
+      "learning_rate": 3.521754866397603e-06,
+      "loss": 1.8114,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8799210553651189,
+      "grad_norm": 0.40242356061935425,
+      "learning_rate": 3.5157420985129696e-06,
+      "loss": 1.5685,
+      "step": 8471
+    },
+    {
+      "epoch": 0.8800249298846993,
+      "grad_norm": 0.396705687046051,
+      "learning_rate": 3.5097342808392054e-06,
+      "loss": 1.4964,
+      "step": 8472
+    },
+    {
+      "epoch": 0.8801288044042797,
+      "grad_norm": 0.4387626647949219,
+      "learning_rate": 3.5037314140160926e-06,
+      "loss": 1.4608,
+      "step": 8473
+    },
+    {
+      "epoch": 0.8802326789238599,
+      "grad_norm": 0.47153568267822266,
+      "learning_rate": 3.4977334986829145e-06,
+      "loss": 1.6019,
+      "step": 8474
+    },
+    {
+      "epoch": 0.8803365534434403,
+      "grad_norm": 0.46217477321624756,
+      "learning_rate": 3.4917405354783772e-06,
+      "loss": 1.7727,
+      "step": 8475
+    },
+    {
+      "epoch": 0.8804404279630207,
+      "grad_norm": 0.48487722873687744,
+      "learning_rate": 3.4857525250406976e-06,
+      "loss": 1.7575,
+      "step": 8476
+    },
+    {
+      "epoch": 0.8805443024826011,
+      "grad_norm": 0.48644381761550903,
+      "learning_rate": 3.479769468007543e-06,
+      "loss": 1.6869,
+      "step": 8477
+    },
+    {
+      "epoch": 0.8806481770021813,
+      "grad_norm": 0.41009244322776794,
+      "learning_rate": 3.4737913650160826e-06,
+      "loss": 1.446,
+      "step": 8478
+    },
+    {
+      "epoch": 0.8807520515217617,
+      "grad_norm": 0.4005885720252991,
+      "learning_rate": 3.467818216702917e-06,
+      "loss": 1.5894,
+      "step": 8479
+    },
+    {
+      "epoch": 0.8808559260413421,
+      "grad_norm": 0.4276701509952545,
+      "learning_rate": 3.461850023704133e-06,
+      "loss": 1.5221,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8809598005609224,
+      "grad_norm": 0.42804211378097534,
+      "learning_rate": 3.4558867866553324e-06,
+      "loss": 1.7771,
+      "step": 8481
+    },
+    {
+      "epoch": 0.8810636750805028,
+      "grad_norm": 0.42332831025123596,
+      "learning_rate": 3.4499285061915134e-06,
+      "loss": 1.6525,
+      "step": 8482
+    },
+    {
+      "epoch": 0.8811675496000831,
+      "grad_norm": 0.4782041311264038,
+      "learning_rate": 3.4439751829472123e-06,
+      "loss": 1.808,
+      "step": 8483
+    },
+    {
+      "epoch": 0.8812714241196634,
+      "grad_norm": 0.37131935358047485,
+      "learning_rate": 3.4380268175564e-06,
+      "loss": 1.4988,
+      "step": 8484
+    },
+    {
+      "epoch": 0.8813752986392438,
+      "grad_norm": 0.46065279841423035,
+      "learning_rate": 3.4320834106525367e-06,
+      "loss": 1.7904,
+      "step": 8485
+    },
+    {
+      "epoch": 0.8814791731588242,
+      "grad_norm": 0.41007938981056213,
+      "learning_rate": 3.4261449628685492e-06,
+      "loss": 1.4656,
+      "step": 8486
+    },
+    {
+      "epoch": 0.8815830476784045,
+      "grad_norm": 0.40511950850486755,
+      "learning_rate": 3.4202114748368264e-06,
+      "loss": 1.6166,
+      "step": 8487
+    },
+    {
+      "epoch": 0.8816869221979848,
+      "grad_norm": 0.42590075731277466,
+      "learning_rate": 3.414282947189251e-06,
+      "loss": 1.681,
+      "step": 8488
+    },
+    {
+      "epoch": 0.8817907967175652,
+      "grad_norm": 0.43543121218681335,
+      "learning_rate": 3.4083593805571523e-06,
+      "loss": 1.6805,
+      "step": 8489
+    },
+    {
+      "epoch": 0.8818946712371455,
+      "grad_norm": 0.4141135513782501,
+      "learning_rate": 3.402440775571364e-06,
+      "loss": 1.7227,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8819985457567259,
+      "grad_norm": 0.4232790470123291,
+      "learning_rate": 3.3965271328621374e-06,
+      "loss": 1.792,
+      "step": 8491
+    },
+    {
+      "epoch": 0.8821024202763063,
+      "grad_norm": 0.40409740805625916,
+      "learning_rate": 3.390618453059263e-06,
+      "loss": 1.589,
+      "step": 8492
+    },
+    {
+      "epoch": 0.8822062947958865,
+      "grad_norm": 0.418250173330307,
+      "learning_rate": 3.384714736791955e-06,
+      "loss": 1.6795,
+      "step": 8493
+    },
+    {
+      "epoch": 0.8823101693154669,
+      "grad_norm": 0.42361485958099365,
+      "learning_rate": 3.3788159846889157e-06,
+      "loss": 1.6938,
+      "step": 8494
+    },
+    {
+      "epoch": 0.8824140438350473,
+      "grad_norm": 0.4593660831451416,
+      "learning_rate": 3.3729221973783144e-06,
+      "loss": 1.6643,
+      "step": 8495
+    },
+    {
+      "epoch": 0.8825179183546276,
+      "grad_norm": 0.4981370270252228,
+      "learning_rate": 3.3670333754877937e-06,
+      "loss": 1.8358,
+      "step": 8496
+    },
+    {
+      "epoch": 0.8826217928742079,
+      "grad_norm": 0.39285340905189514,
+      "learning_rate": 3.3611495196444685e-06,
+      "loss": 1.5936,
+      "step": 8497
+    },
+    {
+      "epoch": 0.8827256673937883,
+      "grad_norm": 0.41241446137428284,
+      "learning_rate": 3.355270630474916e-06,
+      "loss": 1.6163,
+      "step": 8498
+    },
+    {
+      "epoch": 0.8828295419133686,
+      "grad_norm": 0.4433367848396301,
+      "learning_rate": 3.3493967086052126e-06,
+      "loss": 1.5595,
+      "step": 8499
+    },
+    {
+      "epoch": 0.882933416432949,
+      "grad_norm": 0.4431458115577698,
+      "learning_rate": 3.343527754660858e-06,
+      "loss": 1.5006,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8830372909525294,
+      "grad_norm": 0.41516363620758057,
+      "learning_rate": 3.3376637692668632e-06,
+      "loss": 1.6199,
+      "step": 8501
+    },
+    {
+      "epoch": 0.8831411654721097,
+      "grad_norm": 0.4274922311306,
+      "learning_rate": 3.3318047530477013e-06,
+      "loss": 1.7884,
+      "step": 8502
+    },
+    {
+      "epoch": 0.88324503999169,
+      "grad_norm": 0.3764563798904419,
+      "learning_rate": 3.3259507066273064e-06,
+      "loss": 1.4395,
+      "step": 8503
+    },
+    {
+      "epoch": 0.8833489145112704,
+      "grad_norm": 0.4132494032382965,
+      "learning_rate": 3.320101630629091e-06,
+      "loss": 1.7359,
+      "step": 8504
+    },
+    {
+      "epoch": 0.8834527890308508,
+      "grad_norm": 0.42872190475463867,
+      "learning_rate": 3.3142575256759346e-06,
+      "loss": 1.6038,
+      "step": 8505
+    },
+    {
+      "epoch": 0.883556663550431,
+      "grad_norm": 0.41533058881759644,
+      "learning_rate": 3.3084183923901834e-06,
+      "loss": 1.6193,
+      "step": 8506
+    },
+    {
+      "epoch": 0.8836605380700114,
+      "grad_norm": 0.39843878149986267,
+      "learning_rate": 3.302584231393674e-06,
+      "loss": 1.636,
+      "step": 8507
+    },
+    {
+      "epoch": 0.8837644125895918,
+      "grad_norm": 0.4171253740787506,
+      "learning_rate": 3.2967550433076754e-06,
+      "loss": 1.6229,
+      "step": 8508
+    },
+    {
+      "epoch": 0.8838682871091721,
+      "grad_norm": 0.4044714570045471,
+      "learning_rate": 3.290930828752986e-06,
+      "loss": 1.639,
+      "step": 8509
+    },
+    {
+      "epoch": 0.8839721616287525,
+      "grad_norm": 0.401554673910141,
+      "learning_rate": 3.2851115883497986e-06,
+      "loss": 1.6301,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8840760361483329,
+      "grad_norm": 0.40703094005584717,
+      "learning_rate": 3.27929732271785e-06,
+      "loss": 1.5388,
+      "step": 8511
+    },
+    {
+      "epoch": 0.8841799106679131,
+      "grad_norm": 0.4279320538043976,
+      "learning_rate": 3.2734880324762962e-06,
+      "loss": 1.7176,
+      "step": 8512
+    },
+    {
+      "epoch": 0.8842837851874935,
+      "grad_norm": 0.5383460521697998,
+      "learning_rate": 3.267683718243786e-06,
+      "loss": 1.8459,
+      "step": 8513
+    },
+    {
+      "epoch": 0.8843876597070739,
+      "grad_norm": 0.4355732500553131,
+      "learning_rate": 3.261884380638436e-06,
+      "loss": 1.7073,
+      "step": 8514
+    },
+    {
+      "epoch": 0.8844915342266542,
+      "grad_norm": 0.4631870687007904,
+      "learning_rate": 3.25609002027783e-06,
+      "loss": 1.7029,
+      "step": 8515
+    },
+    {
+      "epoch": 0.8845954087462345,
+      "grad_norm": 0.3964858651161194,
+      "learning_rate": 3.250300637779019e-06,
+      "loss": 1.5025,
+      "step": 8516
+    },
+    {
+      "epoch": 0.8846992832658149,
+      "grad_norm": 0.4206657409667969,
+      "learning_rate": 3.2445162337585154e-06,
+      "loss": 1.8947,
+      "step": 8517
+    },
+    {
+      "epoch": 0.8848031577853952,
+      "grad_norm": 0.5454657077789307,
+      "learning_rate": 3.238736808832349e-06,
+      "loss": 1.9812,
+      "step": 8518
+    },
+    {
+      "epoch": 0.8849070323049756,
+      "grad_norm": 0.4348321557044983,
+      "learning_rate": 3.2329623636159435e-06,
+      "loss": 1.7276,
+      "step": 8519
+    },
+    {
+      "epoch": 0.885010906824556,
+      "grad_norm": 0.46917733550071716,
+      "learning_rate": 3.2271928987242518e-06,
+      "loss": 1.6525,
+      "step": 8520
+    },
+    {
+      "epoch": 0.8851147813441362,
+      "grad_norm": 0.45450276136398315,
+      "learning_rate": 3.2214284147716823e-06,
+      "loss": 1.7034,
+      "step": 8521
+    },
+    {
+      "epoch": 0.8852186558637166,
+      "grad_norm": 0.4001985490322113,
+      "learning_rate": 3.215668912372094e-06,
+      "loss": 1.5724,
+      "step": 8522
+    },
+    {
+      "epoch": 0.885322530383297,
+      "grad_norm": 0.41575512290000916,
+      "learning_rate": 3.209914392138835e-06,
+      "loss": 1.5867,
+      "step": 8523
+    },
+    {
+      "epoch": 0.8854264049028773,
+      "grad_norm": 0.42568308115005493,
+      "learning_rate": 3.2041648546847215e-06,
+      "loss": 1.6334,
+      "step": 8524
+    },
+    {
+      "epoch": 0.8855302794224577,
+      "grad_norm": 0.4289058744907379,
+      "learning_rate": 3.198420300622024e-06,
+      "loss": 1.6318,
+      "step": 8525
+    },
+    {
+      "epoch": 0.885634153942038,
+      "grad_norm": 0.39888161420822144,
+      "learning_rate": 3.192680730562492e-06,
+      "loss": 1.5733,
+      "step": 8526
+    },
+    {
+      "epoch": 0.8857380284616183,
+      "grad_norm": 0.42361804842948914,
+      "learning_rate": 3.18694614511737e-06,
+      "loss": 1.7382,
+      "step": 8527
+    },
+    {
+      "epoch": 0.8858419029811987,
+      "grad_norm": 0.4657709300518036,
+      "learning_rate": 3.181216544897314e-06,
+      "loss": 1.7749,
+      "step": 8528
+    },
+    {
+      "epoch": 0.8859457775007791,
+      "grad_norm": 0.4325384199619293,
+      "learning_rate": 3.175491930512492e-06,
+      "loss": 1.8558,
+      "step": 8529
+    },
+    {
+      "epoch": 0.8860496520203595,
+      "grad_norm": 0.4005754292011261,
+      "learning_rate": 3.1697723025725378e-06,
+      "loss": 1.5591,
+      "step": 8530
+    },
+    {
+      "epoch": 0.8861535265399397,
+      "grad_norm": 0.4035787880420685,
+      "learning_rate": 3.164057661686548e-06,
+      "loss": 1.6387,
+      "step": 8531
+    },
+    {
+      "epoch": 0.8862574010595201,
+      "grad_norm": 0.41289085149765015,
+      "learning_rate": 3.15834800846308e-06,
+      "loss": 1.6326,
+      "step": 8532
+    },
+    {
+      "epoch": 0.8863612755791005,
+      "grad_norm": 0.4399457573890686,
+      "learning_rate": 3.15264334351017e-06,
+      "loss": 1.6648,
+      "step": 8533
+    },
+    {
+      "epoch": 0.8864651500986808,
+      "grad_norm": 0.45032960176467896,
+      "learning_rate": 3.146943667435326e-06,
+      "loss": 1.8897,
+      "step": 8534
+    },
+    {
+      "epoch": 0.8865690246182611,
+      "grad_norm": 0.4220801293849945,
+      "learning_rate": 3.141248980845507e-06,
+      "loss": 1.6006,
+      "step": 8535
+    },
+    {
+      "epoch": 0.8866728991378415,
+      "grad_norm": 0.4171903133392334,
+      "learning_rate": 3.135559284347156e-06,
+      "loss": 1.4317,
+      "step": 8536
+    },
+    {
+      "epoch": 0.8867767736574218,
+      "grad_norm": 0.4253293573856354,
+      "learning_rate": 3.1298745785461937e-06,
+      "loss": 1.5809,
+      "step": 8537
+    },
+    {
+      "epoch": 0.8868806481770022,
+      "grad_norm": 0.4167105257511139,
+      "learning_rate": 3.1241948640479803e-06,
+      "loss": 1.6632,
+      "step": 8538
+    },
+    {
+      "epoch": 0.8869845226965826,
+      "grad_norm": 0.4174095690250397,
+      "learning_rate": 3.1185201414573707e-06,
+      "loss": 1.5574,
+      "step": 8539
+    },
+    {
+      "epoch": 0.8870883972161628,
+      "grad_norm": 0.43485039472579956,
+      "learning_rate": 3.112850411378676e-06,
+      "loss": 1.6832,
+      "step": 8540
+    },
+    {
+      "epoch": 0.8871922717357432,
+      "grad_norm": 0.4099987745285034,
+      "learning_rate": 3.10718567441568e-06,
+      "loss": 1.6025,
+      "step": 8541
+    },
+    {
+      "epoch": 0.8872961462553236,
+      "grad_norm": 0.4386424720287323,
+      "learning_rate": 3.101525931171634e-06,
+      "loss": 1.6887,
+      "step": 8542
+    },
+    {
+      "epoch": 0.8874000207749039,
+      "grad_norm": 0.3977483808994293,
+      "learning_rate": 3.095871182249255e-06,
+      "loss": 1.4197,
+      "step": 8543
+    },
+    {
+      "epoch": 0.8875038952944843,
+      "grad_norm": 0.42674514651298523,
+      "learning_rate": 3.0902214282507225e-06,
+      "loss": 1.655,
+      "step": 8544
+    },
+    {
+      "epoch": 0.8876077698140646,
+      "grad_norm": 0.4276520609855652,
+      "learning_rate": 3.0845766697776945e-06,
+      "loss": 1.6276,
+      "step": 8545
+    },
+    {
+      "epoch": 0.8877116443336449,
+      "grad_norm": 0.410883367061615,
+      "learning_rate": 3.078936907431312e-06,
+      "loss": 1.7068,
+      "step": 8546
+    },
+    {
+      "epoch": 0.8878155188532253,
+      "grad_norm": 0.4068666696548462,
+      "learning_rate": 3.0733021418121342e-06,
+      "loss": 1.5961,
+      "step": 8547
+    },
+    {
+      "epoch": 0.8879193933728057,
+      "grad_norm": 0.4384441375732422,
+      "learning_rate": 3.0676723735202416e-06,
+      "loss": 1.735,
+      "step": 8548
+    },
+    {
+      "epoch": 0.888023267892386,
+      "grad_norm": 0.41591498255729675,
+      "learning_rate": 3.0620476031551604e-06,
+      "loss": 1.7273,
+      "step": 8549
+    },
+    {
+      "epoch": 0.8881271424119663,
+      "grad_norm": 0.4390021860599518,
+      "learning_rate": 3.056427831315878e-06,
+      "loss": 1.6152,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8882310169315467,
+      "grad_norm": 0.43032291531562805,
+      "learning_rate": 3.0508130586008545e-06,
+      "loss": 1.6926,
+      "step": 8551
+    },
+    {
+      "epoch": 0.888334891451127,
+      "grad_norm": 0.43168923258781433,
+      "learning_rate": 3.0452032856080227e-06,
+      "loss": 1.6707,
+      "step": 8552
+    },
+    {
+      "epoch": 0.8884387659707074,
+      "grad_norm": 0.42715707421302795,
+      "learning_rate": 3.039598512934783e-06,
+      "loss": 1.6203,
+      "step": 8553
+    },
+    {
+      "epoch": 0.8885426404902877,
+      "grad_norm": 0.45927006006240845,
+      "learning_rate": 3.0339987411779846e-06,
+      "loss": 1.7286,
+      "step": 8554
+    },
+    {
+      "epoch": 0.8886465150098681,
+      "grad_norm": 0.3992420434951782,
+      "learning_rate": 3.0284039709339795e-06,
+      "loss": 1.5416,
+      "step": 8555
+    },
+    {
+      "epoch": 0.8887503895294484,
+      "grad_norm": 0.4219547212123871,
+      "learning_rate": 3.0228142027985684e-06,
+      "loss": 1.6695,
+      "step": 8556
+    },
+    {
+      "epoch": 0.8888542640490288,
+      "grad_norm": 0.41342371702194214,
+      "learning_rate": 3.017229437366992e-06,
+      "loss": 1.5937,
+      "step": 8557
+    },
+    {
+      "epoch": 0.8889581385686092,
+      "grad_norm": 0.3998035788536072,
+      "learning_rate": 3.0116496752340127e-06,
+      "loss": 1.6053,
+      "step": 8558
+    },
+    {
+      "epoch": 0.8890620130881894,
+      "grad_norm": 0.38309964537620544,
+      "learning_rate": 3.0060749169938174e-06,
+      "loss": 1.491,
+      "step": 8559
+    },
+    {
+      "epoch": 0.8891658876077698,
+      "grad_norm": 0.43797820806503296,
+      "learning_rate": 3.000505163240075e-06,
+      "loss": 1.775,
+      "step": 8560
+    },
+    {
+      "epoch": 0.8892697621273502,
+      "grad_norm": 0.39977407455444336,
+      "learning_rate": 2.994940414565922e-06,
+      "loss": 1.5729,
+      "step": 8561
+    },
+    {
+      "epoch": 0.8893736366469305,
+      "grad_norm": 0.410889208316803,
+      "learning_rate": 2.9893806715639626e-06,
+      "loss": 1.5631,
+      "step": 8562
+    },
+    {
+      "epoch": 0.8894775111665109,
+      "grad_norm": 0.41645950078964233,
+      "learning_rate": 2.9838259348262675e-06,
+      "loss": 1.599,
+      "step": 8563
+    },
+    {
+      "epoch": 0.8895813856860912,
+      "grad_norm": 0.43443527817726135,
+      "learning_rate": 2.978276204944358e-06,
+      "loss": 1.6522,
+      "step": 8564
+    },
+    {
+      "epoch": 0.8896852602056715,
+      "grad_norm": 0.4481103718280792,
+      "learning_rate": 2.9727314825092667e-06,
+      "loss": 1.7648,
+      "step": 8565
+    },
+    {
+      "epoch": 0.8897891347252519,
+      "grad_norm": 0.427636057138443,
+      "learning_rate": 2.967191768111427e-06,
+      "loss": 1.6324,
+      "step": 8566
+    },
+    {
+      "epoch": 0.8898930092448323,
+      "grad_norm": 0.44498735666275024,
+      "learning_rate": 2.9616570623408057e-06,
+      "loss": 1.8669,
+      "step": 8567
+    },
+    {
+      "epoch": 0.8899968837644125,
+      "grad_norm": 0.4394892156124115,
+      "learning_rate": 2.956127365786793e-06,
+      "loss": 1.6186,
+      "step": 8568
+    },
+    {
+      "epoch": 0.8901007582839929,
+      "grad_norm": 0.4512022137641907,
+      "learning_rate": 2.950602679038261e-06,
+      "loss": 1.8238,
+      "step": 8569
+    },
+    {
+      "epoch": 0.8902046328035733,
+      "grad_norm": 0.5062389969825745,
+      "learning_rate": 2.9450830026835407e-06,
+      "loss": 1.7467,
+      "step": 8570
+    },
+    {
+      "epoch": 0.8903085073231536,
+      "grad_norm": 0.46624040603637695,
+      "learning_rate": 2.939568337310439e-06,
+      "loss": 1.931,
+      "step": 8571
+    },
+    {
+      "epoch": 0.890412381842734,
+      "grad_norm": 0.41087907552719116,
+      "learning_rate": 2.9340586835062302e-06,
+      "loss": 1.6322,
+      "step": 8572
+    },
+    {
+      "epoch": 0.8905162563623144,
+      "grad_norm": 0.39506015181541443,
+      "learning_rate": 2.9285540418576283e-06,
+      "loss": 1.5128,
+      "step": 8573
+    },
+    {
+      "epoch": 0.8906201308818946,
+      "grad_norm": 0.3968523144721985,
+      "learning_rate": 2.9230544129508655e-06,
+      "loss": 1.5671,
+      "step": 8574
+    },
+    {
+      "epoch": 0.890724005401475,
+      "grad_norm": 0.4123672544956207,
+      "learning_rate": 2.9175597973715828e-06,
+      "loss": 1.4673,
+      "step": 8575
+    },
+    {
+      "epoch": 0.8908278799210554,
+      "grad_norm": 0.4277864098548889,
+      "learning_rate": 2.9120701957049246e-06,
+      "loss": 1.6937,
+      "step": 8576
+    },
+    {
+      "epoch": 0.8909317544406357,
+      "grad_norm": 0.4264744222164154,
+      "learning_rate": 2.906585608535495e-06,
+      "loss": 1.5243,
+      "step": 8577
+    },
+    {
+      "epoch": 0.891035628960216,
+      "grad_norm": 0.45255225896835327,
+      "learning_rate": 2.9011060364473595e-06,
+      "loss": 1.5281,
+      "step": 8578
+    },
+    {
+      "epoch": 0.8911395034797964,
+      "grad_norm": 0.42393550276756287,
+      "learning_rate": 2.8956314800240415e-06,
+      "loss": 1.7556,
+      "step": 8579
+    },
+    {
+      "epoch": 0.8912433779993768,
+      "grad_norm": 0.4410860240459442,
+      "learning_rate": 2.8901619398485347e-06,
+      "loss": 1.7168,
+      "step": 8580
+    },
+    {
+      "epoch": 0.8913472525189571,
+      "grad_norm": 0.43080854415893555,
+      "learning_rate": 2.8846974165033235e-06,
+      "loss": 1.5863,
+      "step": 8581
+    },
+    {
+      "epoch": 0.8914511270385375,
+      "grad_norm": 0.3852452337741852,
+      "learning_rate": 2.8792379105703094e-06,
+      "loss": 1.5794,
+      "step": 8582
+    },
+    {
+      "epoch": 0.8915550015581178,
+      "grad_norm": 0.43257638812065125,
+      "learning_rate": 2.8737834226309156e-06,
+      "loss": 1.7181,
+      "step": 8583
+    },
+    {
+      "epoch": 0.8916588760776981,
+      "grad_norm": 0.42675426602363586,
+      "learning_rate": 2.868333953265989e-06,
+      "loss": 1.6535,
+      "step": 8584
+    },
+    {
+      "epoch": 0.8917627505972785,
+      "grad_norm": 0.40293970704078674,
+      "learning_rate": 2.8628895030558435e-06,
+      "loss": 1.6133,
+      "step": 8585
+    },
+    {
+      "epoch": 0.8918666251168589,
+      "grad_norm": 0.4669196903705597,
+      "learning_rate": 2.8574500725802865e-06,
+      "loss": 1.7044,
+      "step": 8586
+    },
+    {
+      "epoch": 0.8919704996364392,
+      "grad_norm": 0.43928948044776917,
+      "learning_rate": 2.8520156624185777e-06,
+      "loss": 1.6096,
+      "step": 8587
+    },
+    {
+      "epoch": 0.8920743741560195,
+      "grad_norm": 0.4221644401550293,
+      "learning_rate": 2.8465862731494264e-06,
+      "loss": 1.7143,
+      "step": 8588
+    },
+    {
+      "epoch": 0.8921782486755999,
+      "grad_norm": 0.4293653070926666,
+      "learning_rate": 2.841161905351031e-06,
+      "loss": 1.5406,
+      "step": 8589
+    },
+    {
+      "epoch": 0.8922821231951802,
+      "grad_norm": 0.45512861013412476,
+      "learning_rate": 2.8357425596010344e-06,
+      "loss": 1.7877,
+      "step": 8590
+    },
+    {
+      "epoch": 0.8923859977147606,
+      "grad_norm": 0.4328274726867676,
+      "learning_rate": 2.830328236476559e-06,
+      "loss": 2.0108,
+      "step": 8591
+    },
+    {
+      "epoch": 0.892489872234341,
+      "grad_norm": 0.4451085925102234,
+      "learning_rate": 2.8249189365541886e-06,
+      "loss": 1.731,
+      "step": 8592
+    },
+    {
+      "epoch": 0.8925937467539212,
+      "grad_norm": 0.4539354741573334,
+      "learning_rate": 2.819514660409983e-06,
+      "loss": 1.7513,
+      "step": 8593
+    },
+    {
+      "epoch": 0.8926976212735016,
+      "grad_norm": 0.4132786691188812,
+      "learning_rate": 2.814115408619428e-06,
+      "loss": 1.5909,
+      "step": 8594
+    },
+    {
+      "epoch": 0.892801495793082,
+      "grad_norm": 0.3973241448402405,
+      "learning_rate": 2.808721181757523e-06,
+      "loss": 1.7004,
+      "step": 8595
+    },
+    {
+      "epoch": 0.8929053703126623,
+      "grad_norm": 0.4304153025150299,
+      "learning_rate": 2.80333198039871e-06,
+      "loss": 1.6818,
+      "step": 8596
+    },
+    {
+      "epoch": 0.8930092448322426,
+      "grad_norm": 0.42142942547798157,
+      "learning_rate": 2.7979478051168906e-06,
+      "loss": 1.7153,
+      "step": 8597
+    },
+    {
+      "epoch": 0.893113119351823,
+      "grad_norm": 0.4120078384876251,
+      "learning_rate": 2.7925686564854447e-06,
+      "loss": 1.6648,
+      "step": 8598
+    },
+    {
+      "epoch": 0.8932169938714033,
+      "grad_norm": 0.40456852316856384,
+      "learning_rate": 2.7871945350771977e-06,
+      "loss": 1.6177,
+      "step": 8599
+    },
+    {
+      "epoch": 0.8933208683909837,
+      "grad_norm": 0.417198121547699,
+      "learning_rate": 2.781825441464464e-06,
+      "loss": 1.6729,
+      "step": 8600
+    },
+    {
+      "epoch": 0.8934247429105641,
+      "grad_norm": 0.4824064075946808,
+      "learning_rate": 2.776461376218997e-06,
+      "loss": 1.7711,
+      "step": 8601
+    },
+    {
+      "epoch": 0.8935286174301443,
+      "grad_norm": 0.44903162121772766,
+      "learning_rate": 2.7711023399120463e-06,
+      "loss": 1.532,
+      "step": 8602
+    },
+    {
+      "epoch": 0.8936324919497247,
+      "grad_norm": 0.4706164002418518,
+      "learning_rate": 2.7657483331142875e-06,
+      "loss": 1.7887,
+      "step": 8603
+    },
+    {
+      "epoch": 0.8937363664693051,
+      "grad_norm": 0.4258130192756653,
+      "learning_rate": 2.760399356395893e-06,
+      "loss": 1.7112,
+      "step": 8604
+    },
+    {
+      "epoch": 0.8938402409888855,
+      "grad_norm": 0.4342074692249298,
+      "learning_rate": 2.7550554103264845e-06,
+      "loss": 1.6769,
+      "step": 8605
+    },
+    {
+      "epoch": 0.8939441155084658,
+      "grad_norm": 0.43605124950408936,
+      "learning_rate": 2.749716495475152e-06,
+      "loss": 1.595,
+      "step": 8606
+    },
+    {
+      "epoch": 0.8940479900280461,
+      "grad_norm": 0.4049820899963379,
+      "learning_rate": 2.7443826124104454e-06,
+      "loss": 1.5847,
+      "step": 8607
+    },
+    {
+      "epoch": 0.8941518645476265,
+      "grad_norm": 0.41089296340942383,
+      "learning_rate": 2.7390537617003777e-06,
+      "loss": 1.4398,
+      "step": 8608
+    },
+    {
+      "epoch": 0.8942557390672068,
+      "grad_norm": 0.4379929006099701,
+      "learning_rate": 2.73372994391245e-06,
+      "loss": 1.7582,
+      "step": 8609
+    },
+    {
+      "epoch": 0.8943596135867872,
+      "grad_norm": 0.44860774278640747,
+      "learning_rate": 2.72841115961357e-06,
+      "loss": 1.7001,
+      "step": 8610
+    },
+    {
+      "epoch": 0.8944634881063676,
+      "grad_norm": 0.39442136883735657,
+      "learning_rate": 2.7230974093701844e-06,
+      "loss": 1.5964,
+      "step": 8611
+    },
+    {
+      "epoch": 0.8945673626259478,
+      "grad_norm": 0.4311879873275757,
+      "learning_rate": 2.7177886937481524e-06,
+      "loss": 1.7961,
+      "step": 8612
+    },
+    {
+      "epoch": 0.8946712371455282,
+      "grad_norm": 0.40751856565475464,
+      "learning_rate": 2.7124850133127987e-06,
+      "loss": 1.4809,
+      "step": 8613
+    },
+    {
+      "epoch": 0.8947751116651086,
+      "grad_norm": 0.3862675130367279,
+      "learning_rate": 2.707186368628939e-06,
+      "loss": 1.556,
+      "step": 8614
+    },
+    {
+      "epoch": 0.8948789861846889,
+      "grad_norm": 0.43987709283828735,
+      "learning_rate": 2.7018927602608323e-06,
+      "loss": 1.4624,
+      "step": 8615
+    },
+    {
+      "epoch": 0.8949828607042692,
+      "grad_norm": 0.5344493985176086,
+      "learning_rate": 2.6966041887722117e-06,
+      "loss": 1.8081,
+      "step": 8616
+    },
+    {
+      "epoch": 0.8950867352238496,
+      "grad_norm": 0.43068018555641174,
+      "learning_rate": 2.6913206547262593e-06,
+      "loss": 1.7128,
+      "step": 8617
+    },
+    {
+      "epoch": 0.8951906097434299,
+      "grad_norm": 0.45353829860687256,
+      "learning_rate": 2.6860421586856368e-06,
+      "loss": 1.7452,
+      "step": 8618
+    },
+    {
+      "epoch": 0.8952944842630103,
+      "grad_norm": 0.4327743947505951,
+      "learning_rate": 2.6807687012124605e-06,
+      "loss": 1.773,
+      "step": 8619
+    },
+    {
+      "epoch": 0.8953983587825907,
+      "grad_norm": 0.4295056462287903,
+      "learning_rate": 2.6755002828683096e-06,
+      "loss": 1.6954,
+      "step": 8620
+    },
+    {
+      "epoch": 0.8955022333021709,
+      "grad_norm": 0.4431517720222473,
+      "learning_rate": 2.6702369042142463e-06,
+      "loss": 1.7723,
+      "step": 8621
+    },
+    {
+      "epoch": 0.8956061078217513,
+      "grad_norm": 0.4079246520996094,
+      "learning_rate": 2.6649785658107494e-06,
+      "loss": 1.6812,
+      "step": 8622
+    },
+    {
+      "epoch": 0.8957099823413317,
+      "grad_norm": 0.427082896232605,
+      "learning_rate": 2.6597252682178165e-06,
+      "loss": 1.772,
+      "step": 8623
+    },
+    {
+      "epoch": 0.895813856860912,
+      "grad_norm": 0.4852583110332489,
+      "learning_rate": 2.6544770119948714e-06,
+      "loss": 1.6485,
+      "step": 8624
+    },
+    {
+      "epoch": 0.8959177313804924,
+      "grad_norm": 0.4270947575569153,
+      "learning_rate": 2.6492337977008173e-06,
+      "loss": 1.7824,
+      "step": 8625
+    },
+    {
+      "epoch": 0.8960216059000727,
+      "grad_norm": 0.4104156792163849,
+      "learning_rate": 2.6439956258940193e-06,
+      "loss": 1.5354,
+      "step": 8626
+    },
+    {
+      "epoch": 0.896125480419653,
+      "grad_norm": 0.4451044797897339,
+      "learning_rate": 2.638762497132291e-06,
+      "loss": 1.6791,
+      "step": 8627
+    },
+    {
+      "epoch": 0.8962293549392334,
+      "grad_norm": 0.44924893975257874,
+      "learning_rate": 2.633534411972921e-06,
+      "loss": 1.7615,
+      "step": 8628
+    },
+    {
+      "epoch": 0.8963332294588138,
+      "grad_norm": 0.40664803981781006,
+      "learning_rate": 2.628311370972664e-06,
+      "loss": 1.535,
+      "step": 8629
+    },
+    {
+      "epoch": 0.896437103978394,
+      "grad_norm": 0.43392449617385864,
+      "learning_rate": 2.6230933746877406e-06,
+      "loss": 1.6903,
+      "step": 8630
+    },
+    {
+      "epoch": 0.8965409784979744,
+      "grad_norm": 0.38720911741256714,
+      "learning_rate": 2.6178804236738073e-06,
+      "loss": 1.5899,
+      "step": 8631
+    },
+    {
+      "epoch": 0.8966448530175548,
+      "grad_norm": 0.49395009875297546,
+      "learning_rate": 2.61267251848602e-06,
+      "loss": 1.8366,
+      "step": 8632
+    },
+    {
+      "epoch": 0.8967487275371352,
+      "grad_norm": 0.41654813289642334,
+      "learning_rate": 2.6074696596789793e-06,
+      "loss": 1.6235,
+      "step": 8633
+    },
+    {
+      "epoch": 0.8968526020567155,
+      "grad_norm": 0.4147179424762726,
+      "learning_rate": 2.6022718478067363e-06,
+      "loss": 1.5863,
+      "step": 8634
+    },
+    {
+      "epoch": 0.8969564765762958,
+      "grad_norm": 0.46183696389198303,
+      "learning_rate": 2.597079083422832e-06,
+      "loss": 1.7443,
+      "step": 8635
+    },
+    {
+      "epoch": 0.8970603510958762,
+      "grad_norm": 0.4648350179195404,
+      "learning_rate": 2.5918913670802346e-06,
+      "loss": 1.7862,
+      "step": 8636
+    },
+    {
+      "epoch": 0.8971642256154565,
+      "grad_norm": 0.44168201088905334,
+      "learning_rate": 2.5867086993314305e-06,
+      "loss": 1.7847,
+      "step": 8637
+    },
+    {
+      "epoch": 0.8972681001350369,
+      "grad_norm": 0.40692418813705444,
+      "learning_rate": 2.5815310807282944e-06,
+      "loss": 1.6616,
+      "step": 8638
+    },
+    {
+      "epoch": 0.8973719746546173,
+      "grad_norm": 0.4546252489089966,
+      "learning_rate": 2.5763585118222246e-06,
+      "loss": 1.8507,
+      "step": 8639
+    },
+    {
+      "epoch": 0.8974758491741975,
+      "grad_norm": 0.5330055356025696,
+      "learning_rate": 2.5711909931640634e-06,
+      "loss": 1.9273,
+      "step": 8640
+    },
+    {
+      "epoch": 0.8975797236937779,
+      "grad_norm": 0.4015289545059204,
+      "learning_rate": 2.566028525304087e-06,
+      "loss": 1.6388,
+      "step": 8641
+    },
+    {
+      "epoch": 0.8976835982133583,
+      "grad_norm": 0.4576239585876465,
+      "learning_rate": 2.5608711087920844e-06,
+      "loss": 1.7742,
+      "step": 8642
+    },
+    {
+      "epoch": 0.8977874727329386,
+      "grad_norm": 0.42561182379722595,
+      "learning_rate": 2.5557187441772655e-06,
+      "loss": 1.5785,
+      "step": 8643
+    },
+    {
+      "epoch": 0.897891347252519,
+      "grad_norm": 0.43728065490722656,
+      "learning_rate": 2.550571432008325e-06,
+      "loss": 1.6983,
+      "step": 8644
+    },
+    {
+      "epoch": 0.8979952217720993,
+      "grad_norm": 0.4214796721935272,
+      "learning_rate": 2.5454291728334022e-06,
+      "loss": 1.7185,
+      "step": 8645
+    },
+    {
+      "epoch": 0.8980990962916796,
+      "grad_norm": 0.38102787733078003,
+      "learning_rate": 2.5402919672001147e-06,
+      "loss": 1.4215,
+      "step": 8646
+    },
+    {
+      "epoch": 0.89820297081126,
+      "grad_norm": 0.4356735050678253,
+      "learning_rate": 2.5351598156555356e-06,
+      "loss": 1.5594,
+      "step": 8647
+    },
+    {
+      "epoch": 0.8983068453308404,
+      "grad_norm": 0.4779966473579407,
+      "learning_rate": 2.5300327187461837e-06,
+      "loss": 1.7134,
+      "step": 8648
+    },
+    {
+      "epoch": 0.8984107198504206,
+      "grad_norm": 0.4045184850692749,
+      "learning_rate": 2.524910677018083e-06,
+      "loss": 1.5643,
+      "step": 8649
+    },
+    {
+      "epoch": 0.898514594370001,
+      "grad_norm": 0.4227657914161682,
+      "learning_rate": 2.519793691016664e-06,
+      "loss": 1.6597,
+      "step": 8650
+    },
+    {
+      "epoch": 0.8986184688895814,
+      "grad_norm": 0.40350401401519775,
+      "learning_rate": 2.5146817612868634e-06,
+      "loss": 1.5878,
+      "step": 8651
+    },
+    {
+      "epoch": 0.8987223434091617,
+      "grad_norm": 0.42735230922698975,
+      "learning_rate": 2.5095748883730562e-06,
+      "loss": 1.6392,
+      "step": 8652
+    },
+    {
+      "epoch": 0.8988262179287421,
+      "grad_norm": 0.4346289038658142,
+      "learning_rate": 2.50447307281908e-06,
+      "loss": 1.6582,
+      "step": 8653
+    },
+    {
+      "epoch": 0.8989300924483224,
+      "grad_norm": 0.4567740559577942,
+      "learning_rate": 2.4993763151682448e-06,
+      "loss": 1.6875,
+      "step": 8654
+    },
+    {
+      "epoch": 0.8990339669679027,
+      "grad_norm": 0.4155108332633972,
+      "learning_rate": 2.4942846159633104e-06,
+      "loss": 1.6108,
+      "step": 8655
+    },
+    {
+      "epoch": 0.8991378414874831,
+      "grad_norm": 0.3986811935901642,
+      "learning_rate": 2.4891979757465102e-06,
+      "loss": 1.6022,
+      "step": 8656
+    },
+    {
+      "epoch": 0.8992417160070635,
+      "grad_norm": 0.39959198236465454,
+      "learning_rate": 2.484116395059516e-06,
+      "loss": 1.6271,
+      "step": 8657
+    },
+    {
+      "epoch": 0.8993455905266439,
+      "grad_norm": 0.3958359956741333,
+      "learning_rate": 2.479039874443495e-06,
+      "loss": 1.4597,
+      "step": 8658
+    },
+    {
+      "epoch": 0.8994494650462241,
+      "grad_norm": 0.40250423550605774,
+      "learning_rate": 2.473968414439054e-06,
+      "loss": 1.5915,
+      "step": 8659
+    },
+    {
+      "epoch": 0.8995533395658045,
+      "grad_norm": 0.4547244608402252,
+      "learning_rate": 2.468902015586255e-06,
+      "loss": 1.7938,
+      "step": 8660
+    },
+    {
+      "epoch": 0.8996572140853849,
+      "grad_norm": 0.40203481912612915,
+      "learning_rate": 2.4638406784246337e-06,
+      "loss": 1.5711,
+      "step": 8661
+    },
+    {
+      "epoch": 0.8997610886049652,
+      "grad_norm": 0.4178847074508667,
+      "learning_rate": 2.458784403493186e-06,
+      "loss": 1.5101,
+      "step": 8662
+    },
+    {
+      "epoch": 0.8998649631245456,
+      "grad_norm": 0.42161715030670166,
+      "learning_rate": 2.453733191330365e-06,
+      "loss": 1.706,
+      "step": 8663
+    },
+    {
+      "epoch": 0.8999688376441259,
+      "grad_norm": 0.4573306739330292,
+      "learning_rate": 2.448687042474074e-06,
+      "loss": 1.8456,
+      "step": 8664
+    },
+    {
+      "epoch": 0.9000727121637062,
+      "grad_norm": 0.3947295844554901,
+      "learning_rate": 2.4436459574617155e-06,
+      "loss": 1.6289,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9001765866832866,
+      "grad_norm": 0.3863442540168762,
+      "learning_rate": 2.4386099368300943e-06,
+      "loss": 1.6296,
+      "step": 8666
+    },
+    {
+      "epoch": 0.900280461202867,
+      "grad_norm": 0.4450392723083496,
+      "learning_rate": 2.4335789811155253e-06,
+      "loss": 1.7748,
+      "step": 8667
+    },
+    {
+      "epoch": 0.9003843357224472,
+      "grad_norm": 0.4072805643081665,
+      "learning_rate": 2.4285530908537747e-06,
+      "loss": 1.6025,
+      "step": 8668
+    },
+    {
+      "epoch": 0.9004882102420276,
+      "grad_norm": 0.44600561261177063,
+      "learning_rate": 2.423532266580031e-06,
+      "loss": 1.6433,
+      "step": 8669
+    },
+    {
+      "epoch": 0.900592084761608,
+      "grad_norm": 0.4108854830265045,
+      "learning_rate": 2.4185165088289996e-06,
+      "loss": 1.55,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9006959592811883,
+      "grad_norm": 0.4555318355560303,
+      "learning_rate": 2.4135058181348036e-06,
+      "loss": 1.8198,
+      "step": 8671
+    },
+    {
+      "epoch": 0.9007998338007687,
+      "grad_norm": 0.4197269082069397,
+      "learning_rate": 2.4085001950310546e-06,
+      "loss": 1.7265,
+      "step": 8672
+    },
+    {
+      "epoch": 0.900903708320349,
+      "grad_norm": 0.39955320954322815,
+      "learning_rate": 2.403499640050805e-06,
+      "loss": 1.5325,
+      "step": 8673
+    },
+    {
+      "epoch": 0.9010075828399293,
+      "grad_norm": 0.4017268717288971,
+      "learning_rate": 2.3985041537265774e-06,
+      "loss": 1.5629,
+      "step": 8674
+    },
+    {
+      "epoch": 0.9011114573595097,
+      "grad_norm": 0.4230203926563263,
+      "learning_rate": 2.3935137365903526e-06,
+      "loss": 1.6934,
+      "step": 8675
+    },
+    {
+      "epoch": 0.9012153318790901,
+      "grad_norm": 0.43759918212890625,
+      "learning_rate": 2.3885283891735664e-06,
+      "loss": 1.5736,
+      "step": 8676
+    },
+    {
+      "epoch": 0.9013192063986704,
+      "grad_norm": 0.49041640758514404,
+      "learning_rate": 2.383548112007128e-06,
+      "loss": 1.6968,
+      "step": 8677
+    },
+    {
+      "epoch": 0.9014230809182507,
+      "grad_norm": 0.4059118926525116,
+      "learning_rate": 2.3785729056213845e-06,
+      "loss": 1.6976,
+      "step": 8678
+    },
+    {
+      "epoch": 0.9015269554378311,
+      "grad_norm": 0.4104028642177582,
+      "learning_rate": 2.3736027705461737e-06,
+      "loss": 1.581,
+      "step": 8679
+    },
+    {
+      "epoch": 0.9016308299574114,
+      "grad_norm": 0.4261881709098816,
+      "learning_rate": 2.368637707310767e-06,
+      "loss": 1.6556,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9017347044769918,
+      "grad_norm": 0.43607598543167114,
+      "learning_rate": 2.3636777164439016e-06,
+      "loss": 1.802,
+      "step": 8681
+    },
+    {
+      "epoch": 0.9018385789965722,
+      "grad_norm": 0.5014055371284485,
+      "learning_rate": 2.3587227984737835e-06,
+      "loss": 1.8442,
+      "step": 8682
+    },
+    {
+      "epoch": 0.9019424535161525,
+      "grad_norm": 0.3931456208229065,
+      "learning_rate": 2.3537729539280684e-06,
+      "loss": 1.5732,
+      "step": 8683
+    },
+    {
+      "epoch": 0.9020463280357328,
+      "grad_norm": 0.3907303810119629,
+      "learning_rate": 2.3488281833338844e-06,
+      "loss": 1.4718,
+      "step": 8684
+    },
+    {
+      "epoch": 0.9021502025553132,
+      "grad_norm": 0.4133702516555786,
+      "learning_rate": 2.3438884872177992e-06,
+      "loss": 1.5745,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9022540770748936,
+      "grad_norm": 0.45526713132858276,
+      "learning_rate": 2.338953866105864e-06,
+      "loss": 1.8402,
+      "step": 8686
+    },
+    {
+      "epoch": 0.9023579515944739,
+      "grad_norm": 0.4495735764503479,
+      "learning_rate": 2.3340243205235703e-06,
+      "loss": 1.7328,
+      "step": 8687
+    },
+    {
+      "epoch": 0.9024618261140542,
+      "grad_norm": 0.43973442912101746,
+      "learning_rate": 2.3290998509958752e-06,
+      "loss": 1.7364,
+      "step": 8688
+    },
+    {
+      "epoch": 0.9025657006336346,
+      "grad_norm": 0.4277052879333496,
+      "learning_rate": 2.324180458047198e-06,
+      "loss": 1.6693,
+      "step": 8689
+    },
+    {
+      "epoch": 0.9026695751532149,
+      "grad_norm": 0.4482114315032959,
+      "learning_rate": 2.3192661422014204e-06,
+      "loss": 1.7195,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9027734496727953,
+      "grad_norm": 0.44713294506073,
+      "learning_rate": 2.3143569039818737e-06,
+      "loss": 1.5864,
+      "step": 8691
+    },
+    {
+      "epoch": 0.9028773241923757,
+      "grad_norm": 0.4385944604873657,
+      "learning_rate": 2.3094527439113446e-06,
+      "loss": 1.7092,
+      "step": 8692
+    },
+    {
+      "epoch": 0.9029811987119559,
+      "grad_norm": 0.39296817779541016,
+      "learning_rate": 2.3045536625121155e-06,
+      "loss": 1.5689,
+      "step": 8693
+    },
+    {
+      "epoch": 0.9030850732315363,
+      "grad_norm": 0.43056434392929077,
+      "learning_rate": 2.2996596603058695e-06,
+      "loss": 1.7061,
+      "step": 8694
+    },
+    {
+      "epoch": 0.9031889477511167,
+      "grad_norm": 0.49173492193222046,
+      "learning_rate": 2.294770737813795e-06,
+      "loss": 1.6226,
+      "step": 8695
+    },
+    {
+      "epoch": 0.903292822270697,
+      "grad_norm": 0.40122538805007935,
+      "learning_rate": 2.289886895556531e-06,
+      "loss": 1.5687,
+      "step": 8696
+    },
+    {
+      "epoch": 0.9033966967902773,
+      "grad_norm": 0.4927726089954376,
+      "learning_rate": 2.285008134054145e-06,
+      "loss": 1.9438,
+      "step": 8697
+    },
+    {
+      "epoch": 0.9035005713098577,
+      "grad_norm": 0.4643442928791046,
+      "learning_rate": 2.2801344538262103e-06,
+      "loss": 1.6155,
+      "step": 8698
+    },
+    {
+      "epoch": 0.903604445829438,
+      "grad_norm": 0.4130551517009735,
+      "learning_rate": 2.2752658553917226e-06,
+      "loss": 1.669,
+      "step": 8699
+    },
+    {
+      "epoch": 0.9037083203490184,
+      "grad_norm": 0.46958547830581665,
+      "learning_rate": 2.270402339269162e-06,
+      "loss": 1.7943,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9038121948685988,
+      "grad_norm": 0.46900489926338196,
+      "learning_rate": 2.2655439059764415e-06,
+      "loss": 1.5809,
+      "step": 8701
+    },
+    {
+      "epoch": 0.903916069388179,
+      "grad_norm": 0.42625823616981506,
+      "learning_rate": 2.2606905560309534e-06,
+      "loss": 1.5512,
+      "step": 8702
+    },
+    {
+      "epoch": 0.9040199439077594,
+      "grad_norm": 0.38436755537986755,
+      "learning_rate": 2.2558422899495392e-06,
+      "loss": 1.5806,
+      "step": 8703
+    },
+    {
+      "epoch": 0.9041238184273398,
+      "grad_norm": 0.4123709201812744,
+      "learning_rate": 2.2509991082484914e-06,
+      "loss": 1.71,
+      "step": 8704
+    },
+    {
+      "epoch": 0.9042276929469201,
+      "grad_norm": 0.4161551594734192,
+      "learning_rate": 2.2461610114436038e-06,
+      "loss": 1.5765,
+      "step": 8705
+    },
+    {
+      "epoch": 0.9043315674665005,
+      "grad_norm": 0.409616619348526,
+      "learning_rate": 2.241328000050058e-06,
+      "loss": 1.4221,
+      "step": 8706
+    },
+    {
+      "epoch": 0.9044354419860808,
+      "grad_norm": 0.4236219525337219,
+      "learning_rate": 2.236500074582554e-06,
+      "loss": 1.6554,
+      "step": 8707
+    },
+    {
+      "epoch": 0.9045393165056611,
+      "grad_norm": 0.4036356508731842,
+      "learning_rate": 2.2316772355552185e-06,
+      "loss": 1.4211,
+      "step": 8708
+    },
+    {
+      "epoch": 0.9046431910252415,
+      "grad_norm": 0.40732133388519287,
+      "learning_rate": 2.226859483481658e-06,
+      "loss": 1.6462,
+      "step": 8709
+    },
+    {
+      "epoch": 0.9047470655448219,
+      "grad_norm": 0.4440615773200989,
+      "learning_rate": 2.222046818874912e-06,
+      "loss": 1.6724,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9048509400644023,
+      "grad_norm": 0.39947155117988586,
+      "learning_rate": 2.217239242247493e-06,
+      "loss": 1.5371,
+      "step": 8711
+    },
+    {
+      "epoch": 0.9049548145839825,
+      "grad_norm": 0.41746270656585693,
+      "learning_rate": 2.2124367541113856e-06,
+      "loss": 1.7528,
+      "step": 8712
+    },
+    {
+      "epoch": 0.9050586891035629,
+      "grad_norm": 0.40026289224624634,
+      "learning_rate": 2.207639354977997e-06,
+      "loss": 1.5377,
+      "step": 8713
+    },
+    {
+      "epoch": 0.9051625636231433,
+      "grad_norm": 0.4574800133705139,
+      "learning_rate": 2.2028470453582253e-06,
+      "loss": 1.798,
+      "step": 8714
+    },
+    {
+      "epoch": 0.9052664381427236,
+      "grad_norm": 0.4214540123939514,
+      "learning_rate": 2.1980598257624165e-06,
+      "loss": 1.6247,
+      "step": 8715
+    },
+    {
+      "epoch": 0.9053703126623039,
+      "grad_norm": 0.4481583535671234,
+      "learning_rate": 2.193277696700363e-06,
+      "loss": 1.8091,
+      "step": 8716
+    },
+    {
+      "epoch": 0.9054741871818843,
+      "grad_norm": 0.45233625173568726,
+      "learning_rate": 2.188500658681325e-06,
+      "loss": 1.8274,
+      "step": 8717
+    },
+    {
+      "epoch": 0.9055780617014646,
+      "grad_norm": 0.41306886076927185,
+      "learning_rate": 2.183728712214028e-06,
+      "loss": 1.5299,
+      "step": 8718
+    },
+    {
+      "epoch": 0.905681936221045,
+      "grad_norm": 0.4653450846672058,
+      "learning_rate": 2.1789618578066386e-06,
+      "loss": 1.7601,
+      "step": 8719
+    },
+    {
+      "epoch": 0.9057858107406254,
+      "grad_norm": 0.4584251642227173,
+      "learning_rate": 2.1742000959667885e-06,
+      "loss": 1.624,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9058896852602056,
+      "grad_norm": 0.4341670870780945,
+      "learning_rate": 2.169443427201584e-06,
+      "loss": 1.5616,
+      "step": 8721
+    },
+    {
+      "epoch": 0.905993559779786,
+      "grad_norm": 0.408926397562027,
+      "learning_rate": 2.164691852017553e-06,
+      "loss": 1.6354,
+      "step": 8722
+    },
+    {
+      "epoch": 0.9060974342993664,
+      "grad_norm": 0.41678133606910706,
+      "learning_rate": 2.159945370920713e-06,
+      "loss": 1.7588,
+      "step": 8723
+    },
+    {
+      "epoch": 0.9062013088189467,
+      "grad_norm": 0.4230055809020996,
+      "learning_rate": 2.1552039844165316e-06,
+      "loss": 1.6595,
+      "step": 8724
+    },
+    {
+      "epoch": 0.906305183338527,
+      "grad_norm": 0.4654471278190613,
+      "learning_rate": 2.1504676930099056e-06,
+      "loss": 1.6695,
+      "step": 8725
+    },
+    {
+      "epoch": 0.9064090578581074,
+      "grad_norm": 0.4271183907985687,
+      "learning_rate": 2.145736497205242e-06,
+      "loss": 1.6587,
+      "step": 8726
+    },
+    {
+      "epoch": 0.9065129323776877,
+      "grad_norm": 0.4363931119441986,
+      "learning_rate": 2.14101039750636e-06,
+      "loss": 1.6881,
+      "step": 8727
+    },
+    {
+      "epoch": 0.9066168068972681,
+      "grad_norm": 0.4049679636955261,
+      "learning_rate": 2.1362893944165573e-06,
+      "loss": 1.6157,
+      "step": 8728
+    },
+    {
+      "epoch": 0.9067206814168485,
+      "grad_norm": 0.4094752371311188,
+      "learning_rate": 2.131573488438582e-06,
+      "loss": 1.6042,
+      "step": 8729
+    },
+    {
+      "epoch": 0.9068245559364287,
+      "grad_norm": 0.41162753105163574,
+      "learning_rate": 2.1268626800746427e-06,
+      "loss": 1.738,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9069284304560091,
+      "grad_norm": 0.4202596843242645,
+      "learning_rate": 2.1221569698264054e-06,
+      "loss": 1.662,
+      "step": 8731
+    },
+    {
+      "epoch": 0.9070323049755895,
+      "grad_norm": 0.40823620557785034,
+      "learning_rate": 2.11745635819498e-06,
+      "loss": 1.6081,
+      "step": 8732
+    },
+    {
+      "epoch": 0.9071361794951698,
+      "grad_norm": 0.4370567798614502,
+      "learning_rate": 2.1127608456809667e-06,
+      "loss": 1.6897,
+      "step": 8733
+    },
+    {
+      "epoch": 0.9072400540147502,
+      "grad_norm": 0.4072186052799225,
+      "learning_rate": 2.108070432784376e-06,
+      "loss": 1.6987,
+      "step": 8734
+    },
+    {
+      "epoch": 0.9073439285343305,
+      "grad_norm": 0.4287233352661133,
+      "learning_rate": 2.1033851200047195e-06,
+      "loss": 1.5307,
+      "step": 8735
+    },
+    {
+      "epoch": 0.9074478030539109,
+      "grad_norm": 0.40400230884552,
+      "learning_rate": 2.098704907840943e-06,
+      "loss": 1.5036,
+      "step": 8736
+    },
+    {
+      "epoch": 0.9075516775734912,
+      "grad_norm": 0.428813636302948,
+      "learning_rate": 2.094029796791441e-06,
+      "loss": 1.691,
+      "step": 8737
+    },
+    {
+      "epoch": 0.9076555520930716,
+      "grad_norm": 0.4154002368450165,
+      "learning_rate": 2.0893597873540947e-06,
+      "loss": 1.6238,
+      "step": 8738
+    },
+    {
+      "epoch": 0.907759426612652,
+      "grad_norm": 0.42926695942878723,
+      "learning_rate": 2.0846948800262e-06,
+      "loss": 1.765,
+      "step": 8739
+    },
+    {
+      "epoch": 0.9078633011322322,
+      "grad_norm": 0.41468799114227295,
+      "learning_rate": 2.080035075304565e-06,
+      "loss": 1.785,
+      "step": 8740
+    },
+    {
+      "epoch": 0.9079671756518126,
+      "grad_norm": 0.4041023254394531,
+      "learning_rate": 2.0753803736853872e-06,
+      "loss": 1.7357,
+      "step": 8741
+    },
+    {
+      "epoch": 0.908071050171393,
+      "grad_norm": 0.39929163455963135,
+      "learning_rate": 2.070730775664381e-06,
+      "loss": 1.5915,
+      "step": 8742
+    },
+    {
+      "epoch": 0.9081749246909733,
+      "grad_norm": 0.3934778571128845,
+      "learning_rate": 2.0660862817366846e-06,
+      "loss": 1.5252,
+      "step": 8743
+    },
+    {
+      "epoch": 0.9082787992105537,
+      "grad_norm": 0.4070751368999481,
+      "learning_rate": 2.061446892396901e-06,
+      "loss": 1.6859,
+      "step": 8744
+    },
+    {
+      "epoch": 0.908382673730134,
+      "grad_norm": 0.4303978979587555,
+      "learning_rate": 2.0568126081390916e-06,
+      "loss": 1.635,
+      "step": 8745
+    },
+    {
+      "epoch": 0.9084865482497143,
+      "grad_norm": 0.43727797269821167,
+      "learning_rate": 2.052183429456761e-06,
+      "loss": 1.7292,
+      "step": 8746
+    },
+    {
+      "epoch": 0.9085904227692947,
+      "grad_norm": 0.40043336153030396,
+      "learning_rate": 2.047559356842893e-06,
+      "loss": 1.6206,
+      "step": 8747
+    },
+    {
+      "epoch": 0.9086942972888751,
+      "grad_norm": 0.4222732186317444,
+      "learning_rate": 2.042940390789899e-06,
+      "loss": 1.7271,
+      "step": 8748
+    },
+    {
+      "epoch": 0.9087981718084553,
+      "grad_norm": 0.4159920811653137,
+      "learning_rate": 2.0383265317896916e-06,
+      "loss": 1.6506,
+      "step": 8749
+    },
+    {
+      "epoch": 0.9089020463280357,
+      "grad_norm": 0.4522947669029236,
+      "learning_rate": 2.033717780333577e-06,
+      "loss": 1.6477,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9090059208476161,
+      "grad_norm": 0.4436376392841339,
+      "learning_rate": 2.0291141369123746e-06,
+      "loss": 1.6716,
+      "step": 8751
+    },
+    {
+      "epoch": 0.9091097953671964,
+      "grad_norm": 0.42394956946372986,
+      "learning_rate": 2.0245156020163413e-06,
+      "loss": 1.5676,
+      "step": 8752
+    },
+    {
+      "epoch": 0.9092136698867768,
+      "grad_norm": 0.4075998365879059,
+      "learning_rate": 2.019922176135153e-06,
+      "loss": 1.7053,
+      "step": 8753
+    },
+    {
+      "epoch": 0.9093175444063571,
+      "grad_norm": 0.4617473781108856,
+      "learning_rate": 2.0153338597580006e-06,
+      "loss": 1.7169,
+      "step": 8754
+    },
+    {
+      "epoch": 0.9094214189259374,
+      "grad_norm": 0.43988341093063354,
+      "learning_rate": 2.0107506533735e-06,
+      "loss": 1.7221,
+      "step": 8755
+    },
+    {
+      "epoch": 0.9095252934455178,
+      "grad_norm": 0.3944602906703949,
+      "learning_rate": 2.0061725574697266e-06,
+      "loss": 1.4159,
+      "step": 8756
+    },
+    {
+      "epoch": 0.9096291679650982,
+      "grad_norm": 0.4070970416069031,
+      "learning_rate": 2.0015995725342073e-06,
+      "loss": 1.5355,
+      "step": 8757
+    },
+    {
+      "epoch": 0.9097330424846785,
+      "grad_norm": 0.4194161891937256,
+      "learning_rate": 1.9970316990539306e-06,
+      "loss": 1.6739,
+      "step": 8758
+    },
+    {
+      "epoch": 0.9098369170042588,
+      "grad_norm": 0.4412401020526886,
+      "learning_rate": 1.9924689375153406e-06,
+      "loss": 1.6932,
+      "step": 8759
+    },
+    {
+      "epoch": 0.9099407915238392,
+      "grad_norm": 0.483318030834198,
+      "learning_rate": 1.9879112884043315e-06,
+      "loss": 1.7716,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9100446660434196,
+      "grad_norm": 0.3971022367477417,
+      "learning_rate": 1.983358752206266e-06,
+      "loss": 1.7038,
+      "step": 8761
+    },
+    {
+      "epoch": 0.9101485405629999,
+      "grad_norm": 0.3960052728652954,
+      "learning_rate": 1.9788113294059498e-06,
+      "loss": 1.5655,
+      "step": 8762
+    },
+    {
+      "epoch": 0.9102524150825803,
+      "grad_norm": 0.4107067286968231,
+      "learning_rate": 1.974269020487646e-06,
+      "loss": 1.5848,
+      "step": 8763
+    },
+    {
+      "epoch": 0.9103562896021606,
+      "grad_norm": 0.42935070395469666,
+      "learning_rate": 1.969731825935073e-06,
+      "loss": 1.5146,
+      "step": 8764
+    },
+    {
+      "epoch": 0.9104601641217409,
+      "grad_norm": 0.4849339723587036,
+      "learning_rate": 1.9651997462314166e-06,
+      "loss": 1.7126,
+      "step": 8765
+    },
+    {
+      "epoch": 0.9105640386413213,
+      "grad_norm": 0.4564700424671173,
+      "learning_rate": 1.9606727818592963e-06,
+      "loss": 1.8559,
+      "step": 8766
+    },
+    {
+      "epoch": 0.9106679131609017,
+      "grad_norm": 0.43314129114151,
+      "learning_rate": 1.9561509333008045e-06,
+      "loss": 1.6292,
+      "step": 8767
+    },
+    {
+      "epoch": 0.910771787680482,
+      "grad_norm": 0.460445374250412,
+      "learning_rate": 1.9516342010374888e-06,
+      "loss": 1.7847,
+      "step": 8768
+    },
+    {
+      "epoch": 0.9108756622000623,
+      "grad_norm": 0.4221862554550171,
+      "learning_rate": 1.9471225855503262e-06,
+      "loss": 1.6476,
+      "step": 8769
+    },
+    {
+      "epoch": 0.9109795367196427,
+      "grad_norm": 0.3845740258693695,
+      "learning_rate": 1.9426160873197874e-06,
+      "loss": 1.5905,
+      "step": 8770
+    },
+    {
+      "epoch": 0.911083411239223,
+      "grad_norm": 0.42310813069343567,
+      "learning_rate": 1.938114706825778e-06,
+      "loss": 1.7829,
+      "step": 8771
+    },
+    {
+      "epoch": 0.9111872857588034,
+      "grad_norm": 0.42070627212524414,
+      "learning_rate": 1.9336184445476525e-06,
+      "loss": 1.5937,
+      "step": 8772
+    },
+    {
+      "epoch": 0.9112911602783837,
+      "grad_norm": 0.5164587497711182,
+      "learning_rate": 1.9291273009642287e-06,
+      "loss": 1.7094,
+      "step": 8773
+    },
+    {
+      "epoch": 0.911395034797964,
+      "grad_norm": 0.39574146270751953,
+      "learning_rate": 1.924641276553779e-06,
+      "loss": 1.5736,
+      "step": 8774
+    },
+    {
+      "epoch": 0.9114989093175444,
+      "grad_norm": 0.3940182030200958,
+      "learning_rate": 1.9201603717940385e-06,
+      "loss": 1.3872,
+      "step": 8775
+    },
+    {
+      "epoch": 0.9116027838371248,
+      "grad_norm": 0.42792201042175293,
+      "learning_rate": 1.9156845871621698e-06,
+      "loss": 1.7079,
+      "step": 8776
+    },
+    {
+      "epoch": 0.9117066583567051,
+      "grad_norm": 0.40028658509254456,
+      "learning_rate": 1.91121392313483e-06,
+      "loss": 1.564,
+      "step": 8777
+    },
+    {
+      "epoch": 0.9118105328762854,
+      "grad_norm": 0.462466835975647,
+      "learning_rate": 1.9067483801881002e-06,
+      "loss": 1.7931,
+      "step": 8778
+    },
+    {
+      "epoch": 0.9119144073958658,
+      "grad_norm": 0.42079779505729675,
+      "learning_rate": 1.9022879587975162e-06,
+      "loss": 1.6593,
+      "step": 8779
+    },
+    {
+      "epoch": 0.9120182819154461,
+      "grad_norm": 0.4041844606399536,
+      "learning_rate": 1.8978326594381035e-06,
+      "loss": 1.4404,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9121221564350265,
+      "grad_norm": 0.4102264940738678,
+      "learning_rate": 1.8933824825842828e-06,
+      "loss": 1.5394,
+      "step": 8781
+    },
+    {
+      "epoch": 0.9122260309546069,
+      "grad_norm": 0.451555073261261,
+      "learning_rate": 1.8889374287099916e-06,
+      "loss": 1.7135,
+      "step": 8782
+    },
+    {
+      "epoch": 0.9123299054741871,
+      "grad_norm": 0.399200975894928,
+      "learning_rate": 1.8844974982885843e-06,
+      "loss": 1.4235,
+      "step": 8783
+    },
+    {
+      "epoch": 0.9124337799937675,
+      "grad_norm": 0.42155495285987854,
+      "learning_rate": 1.8800626917928776e-06,
+      "loss": 1.5576,
+      "step": 8784
+    },
+    {
+      "epoch": 0.9125376545133479,
+      "grad_norm": 0.4104582965373993,
+      "learning_rate": 1.8756330096951435e-06,
+      "loss": 1.6632,
+      "step": 8785
+    },
+    {
+      "epoch": 0.9126415290329283,
+      "grad_norm": 0.4192086160182953,
+      "learning_rate": 1.8712084524671047e-06,
+      "loss": 1.6348,
+      "step": 8786
+    },
+    {
+      "epoch": 0.9127454035525085,
+      "grad_norm": 0.40471452474594116,
+      "learning_rate": 1.866789020579962e-06,
+      "loss": 1.5868,
+      "step": 8787
+    },
+    {
+      "epoch": 0.9128492780720889,
+      "grad_norm": 0.4262443482875824,
+      "learning_rate": 1.862374714504317e-06,
+      "loss": 1.6801,
+      "step": 8788
+    },
+    {
+      "epoch": 0.9129531525916693,
+      "grad_norm": 0.4280400574207306,
+      "learning_rate": 1.8579655347102875e-06,
+      "loss": 1.5403,
+      "step": 8789
+    },
+    {
+      "epoch": 0.9130570271112496,
+      "grad_norm": 0.4044937193393707,
+      "learning_rate": 1.853561481667404e-06,
+      "loss": 1.5541,
+      "step": 8790
+    },
+    {
+      "epoch": 0.91316090163083,
+      "grad_norm": 0.41494637727737427,
+      "learning_rate": 1.8491625558446634e-06,
+      "loss": 1.7474,
+      "step": 8791
+    },
+    {
+      "epoch": 0.9132647761504104,
+      "grad_norm": 0.391250878572464,
+      "learning_rate": 1.8447687577105244e-06,
+      "loss": 1.6448,
+      "step": 8792
+    },
+    {
+      "epoch": 0.9133686506699906,
+      "grad_norm": 0.4285977780818939,
+      "learning_rate": 1.8403800877328847e-06,
+      "loss": 1.6618,
+      "step": 8793
+    },
+    {
+      "epoch": 0.913472525189571,
+      "grad_norm": 0.4157012104988098,
+      "learning_rate": 1.835996546379104e-06,
+      "loss": 1.6608,
+      "step": 8794
+    },
+    {
+      "epoch": 0.9135763997091514,
+      "grad_norm": 0.41410088539123535,
+      "learning_rate": 1.8316181341159977e-06,
+      "loss": 1.7398,
+      "step": 8795
+    },
+    {
+      "epoch": 0.9136802742287317,
+      "grad_norm": 0.43308156728744507,
+      "learning_rate": 1.8272448514098373e-06,
+      "loss": 1.5286,
+      "step": 8796
+    },
+    {
+      "epoch": 0.913784148748312,
+      "grad_norm": 0.39761316776275635,
+      "learning_rate": 1.8228766987263334e-06,
+      "loss": 1.613,
+      "step": 8797
+    },
+    {
+      "epoch": 0.9138880232678924,
+      "grad_norm": 0.46875429153442383,
+      "learning_rate": 1.8185136765306643e-06,
+      "loss": 1.6326,
+      "step": 8798
+    },
+    {
+      "epoch": 0.9139918977874727,
+      "grad_norm": 0.4058031737804413,
+      "learning_rate": 1.8141557852874634e-06,
+      "loss": 1.4985,
+      "step": 8799
+    },
+    {
+      "epoch": 0.9140957723070531,
+      "grad_norm": 0.4153459668159485,
+      "learning_rate": 1.8098030254608044e-06,
+      "loss": 1.6341,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9141996468266335,
+      "grad_norm": 0.4445685148239136,
+      "learning_rate": 1.8054553975142273e-06,
+      "loss": 1.8224,
+      "step": 8801
+    },
+    {
+      "epoch": 0.9143035213462137,
+      "grad_norm": 0.4452269375324249,
+      "learning_rate": 1.8011129019107176e-06,
+      "loss": 1.7055,
+      "step": 8802
+    },
+    {
+      "epoch": 0.9144073958657941,
+      "grad_norm": 0.4562229812145233,
+      "learning_rate": 1.7967755391127161e-06,
+      "loss": 1.828,
+      "step": 8803
+    },
+    {
+      "epoch": 0.9145112703853745,
+      "grad_norm": 0.3990088701248169,
+      "learning_rate": 1.7924433095821203e-06,
+      "loss": 1.6182,
+      "step": 8804
+    },
+    {
+      "epoch": 0.9146151449049548,
+      "grad_norm": 0.4023396074771881,
+      "learning_rate": 1.788116213780283e-06,
+      "loss": 1.7746,
+      "step": 8805
+    },
+    {
+      "epoch": 0.9147190194245352,
+      "grad_norm": 0.42868486046791077,
+      "learning_rate": 1.783794252168003e-06,
+      "loss": 1.7353,
+      "step": 8806
+    },
+    {
+      "epoch": 0.9148228939441155,
+      "grad_norm": 0.43141698837280273,
+      "learning_rate": 1.7794774252055224e-06,
+      "loss": 1.6535,
+      "step": 8807
+    },
+    {
+      "epoch": 0.9149267684636958,
+      "grad_norm": 0.4295443594455719,
+      "learning_rate": 1.775165733352574e-06,
+      "loss": 1.6215,
+      "step": 8808
+    },
+    {
+      "epoch": 0.9150306429832762,
+      "grad_norm": 0.4191311299800873,
+      "learning_rate": 1.770859177068296e-06,
+      "loss": 1.6455,
+      "step": 8809
+    },
+    {
+      "epoch": 0.9151345175028566,
+      "grad_norm": 0.38304373621940613,
+      "learning_rate": 1.7665577568113212e-06,
+      "loss": 1.6085,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9152383920224368,
+      "grad_norm": 0.4099152684211731,
+      "learning_rate": 1.7622614730397114e-06,
+      "loss": 1.7182,
+      "step": 8811
+    },
+    {
+      "epoch": 0.9153422665420172,
+      "grad_norm": 0.397775262594223,
+      "learning_rate": 1.7579703262109837e-06,
+      "loss": 1.5084,
+      "step": 8812
+    },
+    {
+      "epoch": 0.9154461410615976,
+      "grad_norm": 0.4640827476978302,
+      "learning_rate": 1.7536843167821116e-06,
+      "loss": 1.8689,
+      "step": 8813
+    },
+    {
+      "epoch": 0.915550015581178,
+      "grad_norm": 0.39592549204826355,
+      "learning_rate": 1.7494034452095242e-06,
+      "loss": 1.5368,
+      "step": 8814
+    },
+    {
+      "epoch": 0.9156538901007583,
+      "grad_norm": 0.4693751633167267,
+      "learning_rate": 1.7451277119491073e-06,
+      "loss": 1.806,
+      "step": 8815
+    },
+    {
+      "epoch": 0.9157577646203386,
+      "grad_norm": 0.45381423830986023,
+      "learning_rate": 1.740857117456174e-06,
+      "loss": 1.6147,
+      "step": 8816
+    },
+    {
+      "epoch": 0.915861639139919,
+      "grad_norm": 0.40232381224632263,
+      "learning_rate": 1.7365916621855327e-06,
+      "loss": 1.5101,
+      "step": 8817
+    },
+    {
+      "epoch": 0.9159655136594993,
+      "grad_norm": 0.4090515971183777,
+      "learning_rate": 1.7323313465914093e-06,
+      "loss": 1.6499,
+      "step": 8818
+    },
+    {
+      "epoch": 0.9160693881790797,
+      "grad_norm": 0.3945220410823822,
+      "learning_rate": 1.7280761711274906e-06,
+      "loss": 1.5388,
+      "step": 8819
+    },
+    {
+      "epoch": 0.9161732626986601,
+      "grad_norm": 0.39310646057128906,
+      "learning_rate": 1.7238261362469254e-06,
+      "loss": 1.5099,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9162771372182403,
+      "grad_norm": 0.4310518205165863,
+      "learning_rate": 1.7195812424023129e-06,
+      "loss": 1.4882,
+      "step": 8821
+    },
+    {
+      "epoch": 0.9163810117378207,
+      "grad_norm": 0.4244692921638489,
+      "learning_rate": 1.715341490045691e-06,
+      "loss": 1.6063,
+      "step": 8822
+    },
+    {
+      "epoch": 0.9164848862574011,
+      "grad_norm": 0.4561496675014496,
+      "learning_rate": 1.7111068796285657e-06,
+      "loss": 1.7671,
+      "step": 8823
+    },
+    {
+      "epoch": 0.9165887607769814,
+      "grad_norm": 0.4400840103626251,
+      "learning_rate": 1.7068774116018983e-06,
+      "loss": 1.6152,
+      "step": 8824
+    },
+    {
+      "epoch": 0.9166926352965618,
+      "grad_norm": 0.46823105216026306,
+      "learning_rate": 1.7026530864160728e-06,
+      "loss": 1.5364,
+      "step": 8825
+    },
+    {
+      "epoch": 0.9167965098161421,
+      "grad_norm": 0.43917423486709595,
+      "learning_rate": 1.698433904520963e-06,
+      "loss": 1.5236,
+      "step": 8826
+    },
+    {
+      "epoch": 0.9169003843357224,
+      "grad_norm": 0.4674359858036041,
+      "learning_rate": 1.6942198663658815e-06,
+      "loss": 1.767,
+      "step": 8827
+    },
+    {
+      "epoch": 0.9170042588553028,
+      "grad_norm": 0.39382535219192505,
+      "learning_rate": 1.6900109723995805e-06,
+      "loss": 1.6415,
+      "step": 8828
+    },
+    {
+      "epoch": 0.9171081333748832,
+      "grad_norm": 0.4580112397670746,
+      "learning_rate": 1.6858072230702792e-06,
+      "loss": 1.6657,
+      "step": 8829
+    },
+    {
+      "epoch": 0.9172120078944634,
+      "grad_norm": 0.4108462929725647,
+      "learning_rate": 1.6816086188256475e-06,
+      "loss": 1.6511,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9173158824140438,
+      "grad_norm": 0.4109908938407898,
+      "learning_rate": 1.677415160112794e-06,
+      "loss": 1.5784,
+      "step": 8831
+    },
+    {
+      "epoch": 0.9174197569336242,
+      "grad_norm": 0.4298510253429413,
+      "learning_rate": 1.6732268473782953e-06,
+      "loss": 1.6897,
+      "step": 8832
+    },
+    {
+      "epoch": 0.9175236314532045,
+      "grad_norm": 0.42794668674468994,
+      "learning_rate": 1.6690436810681832e-06,
+      "loss": 1.6477,
+      "step": 8833
+    },
+    {
+      "epoch": 0.9176275059727849,
+      "grad_norm": 0.42130839824676514,
+      "learning_rate": 1.664865661627918e-06,
+      "loss": 1.6891,
+      "step": 8834
+    },
+    {
+      "epoch": 0.9177313804923652,
+      "grad_norm": 0.4458337128162384,
+      "learning_rate": 1.6606927895024217e-06,
+      "loss": 1.5716,
+      "step": 8835
+    },
+    {
+      "epoch": 0.9178352550119455,
+      "grad_norm": 0.4314666986465454,
+      "learning_rate": 1.6565250651360997e-06,
+      "loss": 1.6978,
+      "step": 8836
+    },
+    {
+      "epoch": 0.9179391295315259,
+      "grad_norm": 0.47084978222846985,
+      "learning_rate": 1.6523624889727474e-06,
+      "loss": 1.7084,
+      "step": 8837
+    },
+    {
+      "epoch": 0.9180430040511063,
+      "grad_norm": 0.42452502250671387,
+      "learning_rate": 1.6482050614556709e-06,
+      "loss": 1.4483,
+      "step": 8838
+    },
+    {
+      "epoch": 0.9181468785706867,
+      "grad_norm": 0.4466843008995056,
+      "learning_rate": 1.6440527830275997e-06,
+      "loss": 1.7525,
+      "step": 8839
+    },
+    {
+      "epoch": 0.9182507530902669,
+      "grad_norm": 0.3827952444553375,
+      "learning_rate": 1.6399056541307134e-06,
+      "loss": 1.5032,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9183546276098473,
+      "grad_norm": 0.46445831656455994,
+      "learning_rate": 1.6357636752066529e-06,
+      "loss": 1.6178,
+      "step": 8841
+    },
+    {
+      "epoch": 0.9184585021294277,
+      "grad_norm": 0.4010038375854492,
+      "learning_rate": 1.6316268466965046e-06,
+      "loss": 1.6032,
+      "step": 8842
+    },
+    {
+      "epoch": 0.918562376649008,
+      "grad_norm": 0.42369773983955383,
+      "learning_rate": 1.6274951690408158e-06,
+      "loss": 1.7141,
+      "step": 8843
+    },
+    {
+      "epoch": 0.9186662511685884,
+      "grad_norm": 0.4194977879524231,
+      "learning_rate": 1.623368642679557e-06,
+      "loss": 1.6226,
+      "step": 8844
+    },
+    {
+      "epoch": 0.9187701256881687,
+      "grad_norm": 0.5032253861427307,
+      "learning_rate": 1.6192472680521987e-06,
+      "loss": 1.9285,
+      "step": 8845
+    },
+    {
+      "epoch": 0.918874000207749,
+      "grad_norm": 0.4604244530200958,
+      "learning_rate": 1.6151310455976176e-06,
+      "loss": 1.7397,
+      "step": 8846
+    },
+    {
+      "epoch": 0.9189778747273294,
+      "grad_norm": 0.4401777386665344,
+      "learning_rate": 1.6110199757541632e-06,
+      "loss": 1.5453,
+      "step": 8847
+    },
+    {
+      "epoch": 0.9190817492469098,
+      "grad_norm": 0.48724883794784546,
+      "learning_rate": 1.6069140589596299e-06,
+      "loss": 1.6,
+      "step": 8848
+    },
+    {
+      "epoch": 0.91918562376649,
+      "grad_norm": 0.4106622338294983,
+      "learning_rate": 1.6028132956512732e-06,
+      "loss": 1.6392,
+      "step": 8849
+    },
+    {
+      "epoch": 0.9192894982860704,
+      "grad_norm": 0.4616900086402893,
+      "learning_rate": 1.5987176862657882e-06,
+      "loss": 1.6164,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9193933728056508,
+      "grad_norm": 0.4380607604980469,
+      "learning_rate": 1.5946272312393206e-06,
+      "loss": 1.5529,
+      "step": 8851
+    },
+    {
+      "epoch": 0.9194972473252311,
+      "grad_norm": 0.42449167370796204,
+      "learning_rate": 1.5905419310074831e-06,
+      "loss": 1.7714,
+      "step": 8852
+    },
+    {
+      "epoch": 0.9196011218448115,
+      "grad_norm": 0.4334729015827179,
+      "learning_rate": 1.586461786005311e-06,
+      "loss": 1.703,
+      "step": 8853
+    },
+    {
+      "epoch": 0.9197049963643918,
+      "grad_norm": 0.4081462621688843,
+      "learning_rate": 1.5823867966673288e-06,
+      "loss": 1.6907,
+      "step": 8854
+    },
+    {
+      "epoch": 0.9198088708839721,
+      "grad_norm": 0.4235094487667084,
+      "learning_rate": 1.5783169634274787e-06,
+      "loss": 1.7441,
+      "step": 8855
+    },
+    {
+      "epoch": 0.9199127454035525,
+      "grad_norm": 0.4290684759616852,
+      "learning_rate": 1.574252286719169e-06,
+      "loss": 1.5403,
+      "step": 8856
+    },
+    {
+      "epoch": 0.9200166199231329,
+      "grad_norm": 0.40172362327575684,
+      "learning_rate": 1.5701927669752537e-06,
+      "loss": 1.4616,
+      "step": 8857
+    },
+    {
+      "epoch": 0.9201204944427132,
+      "grad_norm": 0.4437967538833618,
+      "learning_rate": 1.566138404628037e-06,
+      "loss": 1.6938,
+      "step": 8858
+    },
+    {
+      "epoch": 0.9202243689622935,
+      "grad_norm": 0.42959895730018616,
+      "learning_rate": 1.5620892001092902e-06,
+      "loss": 1.5245,
+      "step": 8859
+    },
+    {
+      "epoch": 0.9203282434818739,
+      "grad_norm": 0.4243561923503876,
+      "learning_rate": 1.5580451538502016e-06,
+      "loss": 1.6801,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9204321180014542,
+      "grad_norm": 0.5258318781852722,
+      "learning_rate": 1.5540062662814603e-06,
+      "loss": 1.8436,
+      "step": 8861
+    },
+    {
+      "epoch": 0.9205359925210346,
+      "grad_norm": 0.4423414468765259,
+      "learning_rate": 1.5499725378331441e-06,
+      "loss": 1.6583,
+      "step": 8862
+    },
+    {
+      "epoch": 0.920639867040615,
+      "grad_norm": 0.4549945592880249,
+      "learning_rate": 1.5459439689348266e-06,
+      "loss": 1.6852,
+      "step": 8863
+    },
+    {
+      "epoch": 0.9207437415601953,
+      "grad_norm": 0.43814900517463684,
+      "learning_rate": 1.5419205600155306e-06,
+      "loss": 1.6821,
+      "step": 8864
+    },
+    {
+      "epoch": 0.9208476160797756,
+      "grad_norm": 0.47776687145233154,
+      "learning_rate": 1.5379023115037027e-06,
+      "loss": 1.8725,
+      "step": 8865
+    },
+    {
+      "epoch": 0.920951490599356,
+      "grad_norm": 0.422610342502594,
+      "learning_rate": 1.5338892238272562e-06,
+      "loss": 1.6768,
+      "step": 8866
+    },
+    {
+      "epoch": 0.9210553651189364,
+      "grad_norm": 0.46566253900527954,
+      "learning_rate": 1.5298812974135602e-06,
+      "loss": 1.7364,
+      "step": 8867
+    },
+    {
+      "epoch": 0.9211592396385166,
+      "grad_norm": 0.41145744919776917,
+      "learning_rate": 1.5258785326894286e-06,
+      "loss": 1.7075,
+      "step": 8868
+    },
+    {
+      "epoch": 0.921263114158097,
+      "grad_norm": 0.4216400980949402,
+      "learning_rate": 1.5218809300811155e-06,
+      "loss": 1.7723,
+      "step": 8869
+    },
+    {
+      "epoch": 0.9213669886776774,
+      "grad_norm": 0.4355829656124115,
+      "learning_rate": 1.5178884900143353e-06,
+      "loss": 1.6459,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9214708631972577,
+      "grad_norm": 0.4165591597557068,
+      "learning_rate": 1.5139012129142704e-06,
+      "loss": 1.5313,
+      "step": 8871
+    },
+    {
+      "epoch": 0.9215747377168381,
+      "grad_norm": 0.44293418526649475,
+      "learning_rate": 1.5099190992055034e-06,
+      "loss": 1.7129,
+      "step": 8872
+    },
+    {
+      "epoch": 0.9216786122364184,
+      "grad_norm": 0.45215722918510437,
+      "learning_rate": 1.505942149312123e-06,
+      "loss": 1.4032,
+      "step": 8873
+    },
+    {
+      "epoch": 0.9217824867559987,
+      "grad_norm": 0.4243696630001068,
+      "learning_rate": 1.5019703636576344e-06,
+      "loss": 1.6315,
+      "step": 8874
+    },
+    {
+      "epoch": 0.9218863612755791,
+      "grad_norm": 0.4237101078033447,
+      "learning_rate": 1.498003742665005e-06,
+      "loss": 1.6485,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9219902357951595,
+      "grad_norm": 0.4091566801071167,
+      "learning_rate": 1.4940422867566472e-06,
+      "loss": 1.671,
+      "step": 8876
+    },
+    {
+      "epoch": 0.9220941103147398,
+      "grad_norm": 0.44032323360443115,
+      "learning_rate": 1.4900859963544233e-06,
+      "loss": 1.581,
+      "step": 8877
+    },
+    {
+      "epoch": 0.9221979848343201,
+      "grad_norm": 0.41749629378318787,
+      "learning_rate": 1.486134871879652e-06,
+      "loss": 1.7884,
+      "step": 8878
+    },
+    {
+      "epoch": 0.9223018593539005,
+      "grad_norm": 0.447294682264328,
+      "learning_rate": 1.4821889137530852e-06,
+      "loss": 1.7156,
+      "step": 8879
+    },
+    {
+      "epoch": 0.9224057338734808,
+      "grad_norm": 0.45944663882255554,
+      "learning_rate": 1.4782481223949596e-06,
+      "loss": 1.8423,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9225096083930612,
+      "grad_norm": 0.39821431040763855,
+      "learning_rate": 1.4743124982249112e-06,
+      "loss": 1.5655,
+      "step": 8881
+    },
+    {
+      "epoch": 0.9226134829126416,
+      "grad_norm": 0.48018068075180054,
+      "learning_rate": 1.4703820416620718e-06,
+      "loss": 1.6241,
+      "step": 8882
+    },
+    {
+      "epoch": 0.9227173574322218,
+      "grad_norm": 0.3963991701602936,
+      "learning_rate": 1.4664567531250007e-06,
+      "loss": 1.5582,
+      "step": 8883
+    },
+    {
+      "epoch": 0.9228212319518022,
+      "grad_norm": 0.405376672744751,
+      "learning_rate": 1.462536633031708e-06,
+      "loss": 1.6288,
+      "step": 8884
+    },
+    {
+      "epoch": 0.9229251064713826,
+      "grad_norm": 0.4346674382686615,
+      "learning_rate": 1.4586216817996602e-06,
+      "loss": 1.6538,
+      "step": 8885
+    },
+    {
+      "epoch": 0.9230289809909629,
+      "grad_norm": 0.4289572536945343,
+      "learning_rate": 1.454711899845762e-06,
+      "loss": 1.8293,
+      "step": 8886
+    },
+    {
+      "epoch": 0.9231328555105432,
+      "grad_norm": 0.5129453539848328,
+      "learning_rate": 1.450807287586381e-06,
+      "loss": 1.8293,
+      "step": 8887
+    },
+    {
+      "epoch": 0.9232367300301236,
+      "grad_norm": 0.41872215270996094,
+      "learning_rate": 1.4469078454373174e-06,
+      "loss": 1.6358,
+      "step": 8888
+    },
+    {
+      "epoch": 0.9233406045497039,
+      "grad_norm": 0.4631272256374359,
+      "learning_rate": 1.4430135738138561e-06,
+      "loss": 1.824,
+      "step": 8889
+    },
+    {
+      "epoch": 0.9234444790692843,
+      "grad_norm": 0.41717085242271423,
+      "learning_rate": 1.4391244731306763e-06,
+      "loss": 1.6387,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9235483535888647,
+      "grad_norm": 0.4447776675224304,
+      "learning_rate": 1.4352405438019522e-06,
+      "loss": 1.7798,
+      "step": 8891
+    },
+    {
+      "epoch": 0.923652228108445,
+      "grad_norm": 0.48640677332878113,
+      "learning_rate": 1.431361786241292e-06,
+      "loss": 1.8333,
+      "step": 8892
+    },
+    {
+      "epoch": 0.9237561026280253,
+      "grad_norm": 0.4646976590156555,
+      "learning_rate": 1.427488200861754e-06,
+      "loss": 1.6033,
+      "step": 8893
+    },
+    {
+      "epoch": 0.9238599771476057,
+      "grad_norm": 0.39103513956069946,
+      "learning_rate": 1.4236197880758416e-06,
+      "loss": 1.5394,
+      "step": 8894
+    },
+    {
+      "epoch": 0.9239638516671861,
+      "grad_norm": 0.49602001905441284,
+      "learning_rate": 1.4197565482955144e-06,
+      "loss": 1.8753,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9240677261867664,
+      "grad_norm": 0.3964075744152069,
+      "learning_rate": 1.4158984819321707e-06,
+      "loss": 1.5401,
+      "step": 8896
+    },
+    {
+      "epoch": 0.9241716007063467,
+      "grad_norm": 0.4444807171821594,
+      "learning_rate": 1.4120455893966711e-06,
+      "loss": 1.755,
+      "step": 8897
+    },
+    {
+      "epoch": 0.9242754752259271,
+      "grad_norm": 0.48171067237854004,
+      "learning_rate": 1.4081978710993092e-06,
+      "loss": 1.7385,
+      "step": 8898
+    },
+    {
+      "epoch": 0.9243793497455074,
+      "grad_norm": 0.41400474309921265,
+      "learning_rate": 1.404355327449852e-06,
+      "loss": 1.6438,
+      "step": 8899
+    },
+    {
+      "epoch": 0.9244832242650878,
+      "grad_norm": 0.4013029932975769,
+      "learning_rate": 1.4005179588574835e-06,
+      "loss": 1.3095,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9245870987846682,
+      "grad_norm": 0.41404280066490173,
+      "learning_rate": 1.3966857657308652e-06,
+      "loss": 1.6767,
+      "step": 8901
+    },
+    {
+      "epoch": 0.9246909733042484,
+      "grad_norm": 0.4295409619808197,
+      "learning_rate": 1.392858748478093e-06,
+      "loss": 1.5915,
+      "step": 8902
+    },
+    {
+      "epoch": 0.9247948478238288,
+      "grad_norm": 0.4575590491294861,
+      "learning_rate": 1.3890369075067134e-06,
+      "loss": 1.6812,
+      "step": 8903
+    },
+    {
+      "epoch": 0.9248987223434092,
+      "grad_norm": 0.42268791794776917,
+      "learning_rate": 1.3852202432237227e-06,
+      "loss": 1.6996,
+      "step": 8904
+    },
+    {
+      "epoch": 0.9250025968629895,
+      "grad_norm": 0.4151066839694977,
+      "learning_rate": 1.3814087560355682e-06,
+      "loss": 1.5893,
+      "step": 8905
+    },
+    {
+      "epoch": 0.9251064713825699,
+      "grad_norm": 0.40030190348625183,
+      "learning_rate": 1.377602446348142e-06,
+      "loss": 1.7249,
+      "step": 8906
+    },
+    {
+      "epoch": 0.9252103459021502,
+      "grad_norm": 0.38971251249313354,
+      "learning_rate": 1.3738013145667804e-06,
+      "loss": 1.5488,
+      "step": 8907
+    },
+    {
+      "epoch": 0.9253142204217305,
+      "grad_norm": 0.39237236976623535,
+      "learning_rate": 1.3700053610962882e-06,
+      "loss": 1.5541,
+      "step": 8908
+    },
+    {
+      "epoch": 0.9254180949413109,
+      "grad_norm": 0.4266391098499298,
+      "learning_rate": 1.366214586340886e-06,
+      "loss": 1.7434,
+      "step": 8909
+    },
+    {
+      "epoch": 0.9255219694608913,
+      "grad_norm": 0.42102593183517456,
+      "learning_rate": 1.3624289907042787e-06,
+      "loss": 1.7068,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9256258439804715,
+      "grad_norm": 0.42698073387145996,
+      "learning_rate": 1.3586485745895993e-06,
+      "loss": 1.7564,
+      "step": 8911
+    },
+    {
+      "epoch": 0.9257297185000519,
+      "grad_norm": 0.43909063935279846,
+      "learning_rate": 1.354873338399426e-06,
+      "loss": 1.6685,
+      "step": 8912
+    },
+    {
+      "epoch": 0.9258335930196323,
+      "grad_norm": 0.4355563819408417,
+      "learning_rate": 1.3511032825357982e-06,
+      "loss": 1.4694,
+      "step": 8913
+    },
+    {
+      "epoch": 0.9259374675392126,
+      "grad_norm": 0.4620991349220276,
+      "learning_rate": 1.3473384074001893e-06,
+      "loss": 1.5793,
+      "step": 8914
+    },
+    {
+      "epoch": 0.926041342058793,
+      "grad_norm": 0.4601285457611084,
+      "learning_rate": 1.3435787133935397e-06,
+      "loss": 1.8184,
+      "step": 8915
+    },
+    {
+      "epoch": 0.9261452165783733,
+      "grad_norm": 0.43226033449172974,
+      "learning_rate": 1.3398242009162177e-06,
+      "loss": 1.795,
+      "step": 8916
+    },
+    {
+      "epoch": 0.9262490910979537,
+      "grad_norm": 0.4279690682888031,
+      "learning_rate": 1.3360748703680537e-06,
+      "loss": 1.6734,
+      "step": 8917
+    },
+    {
+      "epoch": 0.926352965617534,
+      "grad_norm": 0.42362403869628906,
+      "learning_rate": 1.3323307221483339e-06,
+      "loss": 1.5626,
+      "step": 8918
+    },
+    {
+      "epoch": 0.9264568401371144,
+      "grad_norm": 0.4446941614151001,
+      "learning_rate": 1.328591756655756e-06,
+      "loss": 1.6638,
+      "step": 8919
+    },
+    {
+      "epoch": 0.9265607146566948,
+      "grad_norm": 0.4361754059791565,
+      "learning_rate": 1.3248579742885126e-06,
+      "loss": 1.552,
+      "step": 8920
+    },
+    {
+      "epoch": 0.926664589176275,
+      "grad_norm": 0.4090036451816559,
+      "learning_rate": 1.3211293754442134e-06,
+      "loss": 1.6251,
+      "step": 8921
+    },
+    {
+      "epoch": 0.9267684636958554,
+      "grad_norm": 0.4072844088077545,
+      "learning_rate": 1.317405960519924e-06,
+      "loss": 1.6939,
+      "step": 8922
+    },
+    {
+      "epoch": 0.9268723382154358,
+      "grad_norm": 0.4060845673084259,
+      "learning_rate": 1.3136877299121608e-06,
+      "loss": 1.6519,
+      "step": 8923
+    },
+    {
+      "epoch": 0.9269762127350161,
+      "grad_norm": 0.4271293878555298,
+      "learning_rate": 1.3099746840168903e-06,
+      "loss": 1.5427,
+      "step": 8924
+    },
+    {
+      "epoch": 0.9270800872545965,
+      "grad_norm": 0.3999338746070862,
+      "learning_rate": 1.3062668232295128e-06,
+      "loss": 1.5207,
+      "step": 8925
+    },
+    {
+      "epoch": 0.9271839617741768,
+      "grad_norm": 0.47124266624450684,
+      "learning_rate": 1.3025641479448958e-06,
+      "loss": 1.8273,
+      "step": 8926
+    },
+    {
+      "epoch": 0.9272878362937571,
+      "grad_norm": 0.4004976451396942,
+      "learning_rate": 1.2988666585573461e-06,
+      "loss": 1.4928,
+      "step": 8927
+    },
+    {
+      "epoch": 0.9273917108133375,
+      "grad_norm": 0.4213370382785797,
+      "learning_rate": 1.2951743554606044e-06,
+      "loss": 1.6839,
+      "step": 8928
+    },
+    {
+      "epoch": 0.9274955853329179,
+      "grad_norm": 0.43283870816230774,
+      "learning_rate": 1.2914872390478894e-06,
+      "loss": 1.7069,
+      "step": 8929
+    },
+    {
+      "epoch": 0.9275994598524981,
+      "grad_norm": 0.3999961316585541,
+      "learning_rate": 1.2878053097118425e-06,
+      "loss": 1.6181,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9277033343720785,
+      "grad_norm": 0.44577065110206604,
+      "learning_rate": 1.2841285678445614e-06,
+      "loss": 1.599,
+      "step": 8931
+    },
+    {
+      "epoch": 0.9278072088916589,
+      "grad_norm": 0.40648534893989563,
+      "learning_rate": 1.2804570138375827e-06,
+      "loss": 1.6496,
+      "step": 8932
+    },
+    {
+      "epoch": 0.9279110834112392,
+      "grad_norm": 0.43543723225593567,
+      "learning_rate": 1.2767906480819104e-06,
+      "loss": 1.7878,
+      "step": 8933
+    },
+    {
+      "epoch": 0.9280149579308196,
+      "grad_norm": 0.39712443947792053,
+      "learning_rate": 1.273129470967982e-06,
+      "loss": 1.6478,
+      "step": 8934
+    },
+    {
+      "epoch": 0.9281188324503999,
+      "grad_norm": 0.43775272369384766,
+      "learning_rate": 1.2694734828856691e-06,
+      "loss": 1.6999,
+      "step": 8935
+    },
+    {
+      "epoch": 0.9282227069699802,
+      "grad_norm": 0.46979156136512756,
+      "learning_rate": 1.265822684224327e-06,
+      "loss": 1.5892,
+      "step": 8936
+    },
+    {
+      "epoch": 0.9283265814895606,
+      "grad_norm": 0.44931840896606445,
+      "learning_rate": 1.2621770753727224e-06,
+      "loss": 1.661,
+      "step": 8937
+    },
+    {
+      "epoch": 0.928430456009141,
+      "grad_norm": 0.441062867641449,
+      "learning_rate": 1.2585366567190949e-06,
+      "loss": 1.7309,
+      "step": 8938
+    },
+    {
+      "epoch": 0.9285343305287213,
+      "grad_norm": 0.45350944995880127,
+      "learning_rate": 1.2549014286511118e-06,
+      "loss": 1.5562,
+      "step": 8939
+    },
+    {
+      "epoch": 0.9286382050483016,
+      "grad_norm": 0.4538798928260803,
+      "learning_rate": 1.2512713915559027e-06,
+      "loss": 1.5712,
+      "step": 8940
+    },
+    {
+      "epoch": 0.928742079567882,
+      "grad_norm": 0.43491330742836,
+      "learning_rate": 1.247646545820036e-06,
+      "loss": 1.6381,
+      "step": 8941
+    },
+    {
+      "epoch": 0.9288459540874624,
+      "grad_norm": 0.417327880859375,
+      "learning_rate": 1.2440268918295194e-06,
+      "loss": 1.71,
+      "step": 8942
+    },
+    {
+      "epoch": 0.9289498286070427,
+      "grad_norm": 0.43641364574432373,
+      "learning_rate": 1.240412429969845e-06,
+      "loss": 1.7413,
+      "step": 8943
+    },
+    {
+      "epoch": 0.929053703126623,
+      "grad_norm": 0.4823327362537384,
+      "learning_rate": 1.2368031606258934e-06,
+      "loss": 1.8444,
+      "step": 8944
+    },
+    {
+      "epoch": 0.9291575776462034,
+      "grad_norm": 0.4450342655181885,
+      "learning_rate": 1.2331990841820407e-06,
+      "loss": 1.8187,
+      "step": 8945
+    },
+    {
+      "epoch": 0.9292614521657837,
+      "grad_norm": 0.4293372631072998,
+      "learning_rate": 1.2296002010220964e-06,
+      "loss": 1.6616,
+      "step": 8946
+    },
+    {
+      "epoch": 0.9293653266853641,
+      "grad_norm": 0.4217541217803955,
+      "learning_rate": 1.2260065115292984e-06,
+      "loss": 1.6949,
+      "step": 8947
+    },
+    {
+      "epoch": 0.9294692012049445,
+      "grad_norm": 0.4160800278186798,
+      "learning_rate": 1.2224180160863574e-06,
+      "loss": 1.535,
+      "step": 8948
+    },
+    {
+      "epoch": 0.9295730757245247,
+      "grad_norm": 0.42056944966316223,
+      "learning_rate": 1.2188347150754231e-06,
+      "loss": 1.7167,
+      "step": 8949
+    },
+    {
+      "epoch": 0.9296769502441051,
+      "grad_norm": 0.4336167871952057,
+      "learning_rate": 1.2152566088780792e-06,
+      "loss": 1.5799,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9297808247636855,
+      "grad_norm": 0.43166765570640564,
+      "learning_rate": 1.2116836978753709e-06,
+      "loss": 1.603,
+      "step": 8951
+    },
+    {
+      "epoch": 0.9298846992832658,
+      "grad_norm": 0.4209384620189667,
+      "learning_rate": 1.2081159824477883e-06,
+      "loss": 1.5984,
+      "step": 8952
+    },
+    {
+      "epoch": 0.9299885738028462,
+      "grad_norm": 0.42866531014442444,
+      "learning_rate": 1.2045534629752609e-06,
+      "loss": 1.6496,
+      "step": 8953
+    },
+    {
+      "epoch": 0.9300924483224265,
+      "grad_norm": 0.44947803020477295,
+      "learning_rate": 1.2009961398371628e-06,
+      "loss": 1.4944,
+      "step": 8954
+    },
+    {
+      "epoch": 0.9301963228420068,
+      "grad_norm": 0.4815939962863922,
+      "learning_rate": 1.197444013412341e-06,
+      "loss": 1.5212,
+      "step": 8955
+    },
+    {
+      "epoch": 0.9303001973615872,
+      "grad_norm": 0.5035466551780701,
+      "learning_rate": 1.1938970840790486e-06,
+      "loss": 1.6037,
+      "step": 8956
+    },
+    {
+      "epoch": 0.9304040718811676,
+      "grad_norm": 0.43349477648735046,
+      "learning_rate": 1.1903553522150169e-06,
+      "loss": 1.7366,
+      "step": 8957
+    },
+    {
+      "epoch": 0.9305079464007479,
+      "grad_norm": 0.4126850366592407,
+      "learning_rate": 1.186818818197416e-06,
+      "loss": 1.5468,
+      "step": 8958
+    },
+    {
+      "epoch": 0.9306118209203282,
+      "grad_norm": 0.4104941189289093,
+      "learning_rate": 1.1832874824028451e-06,
+      "loss": 1.5958,
+      "step": 8959
+    },
+    {
+      "epoch": 0.9307156954399086,
+      "grad_norm": 0.422209769487381,
+      "learning_rate": 1.1797613452073808e-06,
+      "loss": 1.5762,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9308195699594889,
+      "grad_norm": 0.4042091965675354,
+      "learning_rate": 1.1762404069865174e-06,
+      "loss": 1.6074,
+      "step": 8961
+    },
+    {
+      "epoch": 0.9309234444790693,
+      "grad_norm": 0.4372323751449585,
+      "learning_rate": 1.1727246681152049e-06,
+      "loss": 1.7324,
+      "step": 8962
+    },
+    {
+      "epoch": 0.9310273189986497,
+      "grad_norm": 0.46790817379951477,
+      "learning_rate": 1.1692141289678492e-06,
+      "loss": 1.6437,
+      "step": 8963
+    },
+    {
+      "epoch": 0.9311311935182299,
+      "grad_norm": 0.4523914158344269,
+      "learning_rate": 1.1657087899183016e-06,
+      "loss": 1.5817,
+      "step": 8964
+    },
+    {
+      "epoch": 0.9312350680378103,
+      "grad_norm": 0.4646068215370178,
+      "learning_rate": 1.162208651339841e-06,
+      "loss": 1.7894,
+      "step": 8965
+    },
+    {
+      "epoch": 0.9313389425573907,
+      "grad_norm": 0.4205915331840515,
+      "learning_rate": 1.1587137136052083e-06,
+      "loss": 1.5805,
+      "step": 8966
+    },
+    {
+      "epoch": 0.931442817076971,
+      "grad_norm": 0.40744760632514954,
+      "learning_rate": 1.1552239770865892e-06,
+      "loss": 1.6191,
+      "step": 8967
+    },
+    {
+      "epoch": 0.9315466915965513,
+      "grad_norm": 0.43079641461372375,
+      "learning_rate": 1.1517394421556138e-06,
+      "loss": 1.7234,
+      "step": 8968
+    },
+    {
+      "epoch": 0.9316505661161317,
+      "grad_norm": 0.4436144232749939,
+      "learning_rate": 1.1482601091833578e-06,
+      "loss": 1.6766,
+      "step": 8969
+    },
+    {
+      "epoch": 0.9317544406357121,
+      "grad_norm": 0.43747812509536743,
+      "learning_rate": 1.1447859785403359e-06,
+      "loss": 1.7669,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9318583151552924,
+      "grad_norm": 0.4351133108139038,
+      "learning_rate": 1.1413170505965354e-06,
+      "loss": 1.7189,
+      "step": 8971
+    },
+    {
+      "epoch": 0.9319621896748728,
+      "grad_norm": 0.4412538409233093,
+      "learning_rate": 1.1378533257213387e-06,
+      "loss": 1.8092,
+      "step": 8972
+    },
+    {
+      "epoch": 0.9320660641944531,
+      "grad_norm": 0.4323018789291382,
+      "learning_rate": 1.134394804283634e-06,
+      "loss": 1.8075,
+      "step": 8973
+    },
+    {
+      "epoch": 0.9321699387140334,
+      "grad_norm": 0.47585269808769226,
+      "learning_rate": 1.1309414866517209e-06,
+      "loss": 1.7678,
+      "step": 8974
+    },
+    {
+      "epoch": 0.9322738132336138,
+      "grad_norm": 0.5131593942642212,
+      "learning_rate": 1.127493373193339e-06,
+      "loss": 1.6321,
+      "step": 8975
+    },
+    {
+      "epoch": 0.9323776877531942,
+      "grad_norm": 0.4566580057144165,
+      "learning_rate": 1.1240504642756945e-06,
+      "loss": 1.7411,
+      "step": 8976
+    },
+    {
+      "epoch": 0.9324815622727745,
+      "grad_norm": 0.4639977514743805,
+      "learning_rate": 1.1206127602654271e-06,
+      "loss": 1.9319,
+      "step": 8977
+    },
+    {
+      "epoch": 0.9325854367923548,
+      "grad_norm": 0.4284569323062897,
+      "learning_rate": 1.1171802615286275e-06,
+      "loss": 1.6936,
+      "step": 8978
+    },
+    {
+      "epoch": 0.9326893113119352,
+      "grad_norm": 0.46785593032836914,
+      "learning_rate": 1.1137529684308311e-06,
+      "loss": 1.7217,
+      "step": 8979
+    },
+    {
+      "epoch": 0.9327931858315155,
+      "grad_norm": 0.4252128303050995,
+      "learning_rate": 1.110330881337013e-06,
+      "loss": 1.6731,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9328970603510959,
+      "grad_norm": 0.4072088599205017,
+      "learning_rate": 1.106914000611603e-06,
+      "loss": 1.5772,
+      "step": 8981
+    },
+    {
+      "epoch": 0.9330009348706763,
+      "grad_norm": 0.4445166289806366,
+      "learning_rate": 1.1035023266184663e-06,
+      "loss": 1.6541,
+      "step": 8982
+    },
+    {
+      "epoch": 0.9331048093902565,
+      "grad_norm": 0.42075401544570923,
+      "learning_rate": 1.1000958597209343e-06,
+      "loss": 1.6717,
+      "step": 8983
+    },
+    {
+      "epoch": 0.9332086839098369,
+      "grad_norm": 0.4416648745536804,
+      "learning_rate": 1.0966946002817557e-06,
+      "loss": 1.8383,
+      "step": 8984
+    },
+    {
+      "epoch": 0.9333125584294173,
+      "grad_norm": 0.4478912055492401,
+      "learning_rate": 1.0932985486631408e-06,
+      "loss": 1.6386,
+      "step": 8985
+    },
+    {
+      "epoch": 0.9334164329489976,
+      "grad_norm": 0.4134404957294464,
+      "learning_rate": 1.0899077052267393e-06,
+      "loss": 1.6147,
+      "step": 8986
+    },
+    {
+      "epoch": 0.933520307468578,
+      "grad_norm": 0.38977158069610596,
+      "learning_rate": 1.086522070333662e-06,
+      "loss": 1.6074,
+      "step": 8987
+    },
+    {
+      "epoch": 0.9336241819881583,
+      "grad_norm": 0.4343447685241699,
+      "learning_rate": 1.0831416443444376e-06,
+      "loss": 1.7838,
+      "step": 8988
+    },
+    {
+      "epoch": 0.9337280565077386,
+      "grad_norm": 0.42949017882347107,
+      "learning_rate": 1.0797664276190666e-06,
+      "loss": 1.5811,
+      "step": 8989
+    },
+    {
+      "epoch": 0.933831931027319,
+      "grad_norm": 0.4194577634334564,
+      "learning_rate": 1.0763964205169786e-06,
+      "loss": 1.7476,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9339358055468994,
+      "grad_norm": 0.42771607637405396,
+      "learning_rate": 1.0730316233970473e-06,
+      "loss": 1.7751,
+      "step": 8991
+    },
+    {
+      "epoch": 0.9340396800664796,
+      "grad_norm": 0.4049451947212219,
+      "learning_rate": 1.0696720366176138e-06,
+      "loss": 1.5934,
+      "step": 8992
+    },
+    {
+      "epoch": 0.93414355458606,
+      "grad_norm": 0.4572373628616333,
+      "learning_rate": 1.0663176605364255e-06,
+      "loss": 1.7941,
+      "step": 8993
+    },
+    {
+      "epoch": 0.9342474291056404,
+      "grad_norm": 0.3885292708873749,
+      "learning_rate": 1.0629684955107132e-06,
+      "loss": 1.4659,
+      "step": 8994
+    },
+    {
+      "epoch": 0.9343513036252208,
+      "grad_norm": 0.38825228810310364,
+      "learning_rate": 1.0596245418971307e-06,
+      "loss": 1.5177,
+      "step": 8995
+    },
+    {
+      "epoch": 0.9344551781448011,
+      "grad_norm": 0.47034990787506104,
+      "learning_rate": 1.0562858000517928e-06,
+      "loss": 1.774,
+      "step": 8996
+    },
+    {
+      "epoch": 0.9345590526643814,
+      "grad_norm": 0.5254196524620056,
+      "learning_rate": 1.0529522703302319e-06,
+      "loss": 1.578,
+      "step": 8997
+    },
+    {
+      "epoch": 0.9346629271839618,
+      "grad_norm": 0.4228893518447876,
+      "learning_rate": 1.049623953087453e-06,
+      "loss": 1.6469,
+      "step": 8998
+    },
+    {
+      "epoch": 0.9347668017035421,
+      "grad_norm": 0.4083818197250366,
+      "learning_rate": 1.0463008486779058e-06,
+      "loss": 1.5999,
+      "step": 8999
+    },
+    {
+      "epoch": 0.9348706762231225,
+      "grad_norm": 0.417376846075058,
+      "learning_rate": 1.0429829574554574e-06,
+      "loss": 1.5968,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9349745507427029,
+      "grad_norm": 0.4227767884731293,
+      "learning_rate": 1.0396702797734415e-06,
+      "loss": 1.5762,
+      "step": 9001
+    },
+    {
+      "epoch": 0.9350784252622831,
+      "grad_norm": 0.41051211953163147,
+      "learning_rate": 1.0363628159846483e-06,
+      "loss": 1.4743,
+      "step": 9002
+    },
+    {
+      "epoch": 0.9351822997818635,
+      "grad_norm": 0.41701540350914,
+      "learning_rate": 1.0330605664412684e-06,
+      "loss": 1.6244,
+      "step": 9003
+    },
+    {
+      "epoch": 0.9352861743014439,
+      "grad_norm": 0.44816792011260986,
+      "learning_rate": 1.0297635314949872e-06,
+      "loss": 1.8127,
+      "step": 9004
+    },
+    {
+      "epoch": 0.9353900488210242,
+      "grad_norm": 0.4178299307823181,
+      "learning_rate": 1.0264717114969125e-06,
+      "loss": 1.5446,
+      "step": 9005
+    },
+    {
+      "epoch": 0.9354939233406045,
+      "grad_norm": 0.4642837941646576,
+      "learning_rate": 1.023185106797586e-06,
+      "loss": 1.7789,
+      "step": 9006
+    },
+    {
+      "epoch": 0.9355977978601849,
+      "grad_norm": 0.4191594421863556,
+      "learning_rate": 1.0199037177470172e-06,
+      "loss": 1.7449,
+      "step": 9007
+    },
+    {
+      "epoch": 0.9357016723797652,
+      "grad_norm": 0.42831364274024963,
+      "learning_rate": 1.0166275446946372e-06,
+      "loss": 1.7044,
+      "step": 9008
+    },
+    {
+      "epoch": 0.9358055468993456,
+      "grad_norm": 0.44095903635025024,
+      "learning_rate": 1.0133565879893447e-06,
+      "loss": 1.7159,
+      "step": 9009
+    },
+    {
+      "epoch": 0.935909421418926,
+      "grad_norm": 0.4127524495124817,
+      "learning_rate": 1.0100908479794558e-06,
+      "loss": 1.5209,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9360132959385062,
+      "grad_norm": 0.45353084802627563,
+      "learning_rate": 1.0068303250127697e-06,
+      "loss": 1.7395,
+      "step": 9011
+    },
+    {
+      "epoch": 0.9361171704580866,
+      "grad_norm": 0.4273471236228943,
+      "learning_rate": 1.0035750194364758e-06,
+      "loss": 1.6103,
+      "step": 9012
+    },
+    {
+      "epoch": 0.936221044977667,
+      "grad_norm": 0.5556546449661255,
+      "learning_rate": 1.0003249315972684e-06,
+      "loss": 1.7848,
+      "step": 9013
+    },
+    {
+      "epoch": 0.9363249194972473,
+      "grad_norm": 0.418753057718277,
+      "learning_rate": 9.97080061841238e-07,
+      "loss": 1.7764,
+      "step": 9014
+    },
+    {
+      "epoch": 0.9364287940168277,
+      "grad_norm": 0.40560612082481384,
+      "learning_rate": 9.938404105139465e-07,
+      "loss": 1.6056,
+      "step": 9015
+    },
+    {
+      "epoch": 0.936532668536408,
+      "grad_norm": 0.402055025100708,
+      "learning_rate": 9.906059779603849e-07,
+      "loss": 1.5732,
+      "step": 9016
+    },
+    {
+      "epoch": 0.9366365430559883,
+      "grad_norm": 0.4551316499710083,
+      "learning_rate": 9.873767645250055e-07,
+      "loss": 1.8166,
+      "step": 9017
+    },
+    {
+      "epoch": 0.9367404175755687,
+      "grad_norm": 0.4371841549873352,
+      "learning_rate": 9.84152770551683e-07,
+      "loss": 1.6245,
+      "step": 9018
+    },
+    {
+      "epoch": 0.9368442920951491,
+      "grad_norm": 0.43440502882003784,
+      "learning_rate": 9.809339963837426e-07,
+      "loss": 1.6387,
+      "step": 9019
+    },
+    {
+      "epoch": 0.9369481666147295,
+      "grad_norm": 0.41472846269607544,
+      "learning_rate": 9.77720442363983e-07,
+      "loss": 1.3704,
+      "step": 9020
+    },
+    {
+      "epoch": 0.9370520411343097,
+      "grad_norm": 0.437029629945755,
+      "learning_rate": 9.745121088346021e-07,
+      "loss": 1.7435,
+      "step": 9021
+    },
+    {
+      "epoch": 0.9371559156538901,
+      "grad_norm": 0.4492699205875397,
+      "learning_rate": 9.713089961372657e-07,
+      "loss": 1.8335,
+      "step": 9022
+    },
+    {
+      "epoch": 0.9372597901734705,
+      "grad_norm": 0.41893166303634644,
+      "learning_rate": 9.681111046130843e-07,
+      "loss": 1.6002,
+      "step": 9023
+    },
+    {
+      "epoch": 0.9373636646930508,
+      "grad_norm": 0.4353455603122711,
+      "learning_rate": 9.64918434602613e-07,
+      "loss": 1.7331,
+      "step": 9024
+    },
+    {
+      "epoch": 0.9374675392126312,
+      "grad_norm": 0.44769999384880066,
+      "learning_rate": 9.617309864458358e-07,
+      "loss": 1.7173,
+      "step": 9025
+    },
+    {
+      "epoch": 0.9375714137322115,
+      "grad_norm": 0.3922676146030426,
+      "learning_rate": 9.585487604821863e-07,
+      "loss": 1.5642,
+      "step": 9026
+    },
+    {
+      "epoch": 0.9376752882517918,
+      "grad_norm": 0.4413311779499054,
+      "learning_rate": 9.553717570505716e-07,
+      "loss": 1.6186,
+      "step": 9027
+    },
+    {
+      "epoch": 0.9377791627713722,
+      "grad_norm": 0.3945287764072418,
+      "learning_rate": 9.521999764892875e-07,
+      "loss": 1.5881,
+      "step": 9028
+    },
+    {
+      "epoch": 0.9378830372909526,
+      "grad_norm": 0.41674908995628357,
+      "learning_rate": 9.49033419136125e-07,
+      "loss": 1.5853,
+      "step": 9029
+    },
+    {
+      "epoch": 0.9379869118105328,
+      "grad_norm": 0.4065210521221161,
+      "learning_rate": 9.458720853282977e-07,
+      "loss": 1.5876,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9380907863301132,
+      "grad_norm": 0.43645021319389343,
+      "learning_rate": 9.427159754024473e-07,
+      "loss": 1.694,
+      "step": 9031
+    },
+    {
+      "epoch": 0.9381946608496936,
+      "grad_norm": 0.4190250337123871,
+      "learning_rate": 9.395650896946884e-07,
+      "loss": 1.686,
+      "step": 9032
+    },
+    {
+      "epoch": 0.9382985353692739,
+      "grad_norm": 0.4211812913417816,
+      "learning_rate": 9.364194285405581e-07,
+      "loss": 1.6361,
+      "step": 9033
+    },
+    {
+      "epoch": 0.9384024098888543,
+      "grad_norm": 0.438837468624115,
+      "learning_rate": 9.332789922750495e-07,
+      "loss": 1.5855,
+      "step": 9034
+    },
+    {
+      "epoch": 0.9385062844084346,
+      "grad_norm": 0.40594640374183655,
+      "learning_rate": 9.301437812325952e-07,
+      "loss": 1.6175,
+      "step": 9035
+    },
+    {
+      "epoch": 0.9386101589280149,
+      "grad_norm": 0.4930059313774109,
+      "learning_rate": 9.27013795747067e-07,
+      "loss": 1.6779,
+      "step": 9036
+    },
+    {
+      "epoch": 0.9387140334475953,
+      "grad_norm": 0.4809900224208832,
+      "learning_rate": 9.238890361517815e-07,
+      "loss": 1.7196,
+      "step": 9037
+    },
+    {
+      "epoch": 0.9388179079671757,
+      "grad_norm": 0.4233071208000183,
+      "learning_rate": 9.20769502779506e-07,
+      "loss": 1.5481,
+      "step": 9038
+    },
+    {
+      "epoch": 0.938921782486756,
+      "grad_norm": 0.43219685554504395,
+      "learning_rate": 9.176551959624524e-07,
+      "loss": 1.5561,
+      "step": 9039
+    },
+    {
+      "epoch": 0.9390256570063363,
+      "grad_norm": 0.41195639967918396,
+      "learning_rate": 9.145461160322555e-07,
+      "loss": 1.6116,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9391295315259167,
+      "grad_norm": 0.3952676057815552,
+      "learning_rate": 9.114422633200226e-07,
+      "loss": 1.5454,
+      "step": 9041
+    },
+    {
+      "epoch": 0.939233406045497,
+      "grad_norm": 0.41172322630882263,
+      "learning_rate": 9.083436381562838e-07,
+      "loss": 1.6116,
+      "step": 9042
+    },
+    {
+      "epoch": 0.9393372805650774,
+      "grad_norm": 0.4490038752555847,
+      "learning_rate": 9.05250240871014e-07,
+      "loss": 1.7028,
+      "step": 9043
+    },
+    {
+      "epoch": 0.9394411550846578,
+      "grad_norm": 0.46638983488082886,
+      "learning_rate": 9.021620717936385e-07,
+      "loss": 1.7121,
+      "step": 9044
+    },
+    {
+      "epoch": 0.9395450296042381,
+      "grad_norm": 0.42944493889808655,
+      "learning_rate": 8.990791312530222e-07,
+      "loss": 1.6753,
+      "step": 9045
+    },
+    {
+      "epoch": 0.9396489041238184,
+      "grad_norm": 0.4450928270816803,
+      "learning_rate": 8.960014195774857e-07,
+      "loss": 1.7523,
+      "step": 9046
+    },
+    {
+      "epoch": 0.9397527786433988,
+      "grad_norm": 0.3949832022190094,
+      "learning_rate": 8.929289370947669e-07,
+      "loss": 1.5365,
+      "step": 9047
+    },
+    {
+      "epoch": 0.9398566531629792,
+      "grad_norm": 0.41552335023880005,
+      "learning_rate": 8.89861684132065e-07,
+      "loss": 1.7463,
+      "step": 9048
+    },
+    {
+      "epoch": 0.9399605276825594,
+      "grad_norm": 0.40294694900512695,
+      "learning_rate": 8.867996610160246e-07,
+      "loss": 1.6094,
+      "step": 9049
+    },
+    {
+      "epoch": 0.9400644022021398,
+      "grad_norm": 0.43670380115509033,
+      "learning_rate": 8.837428680727178e-07,
+      "loss": 1.7627,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9401682767217202,
+      "grad_norm": 0.4421059787273407,
+      "learning_rate": 8.806913056276788e-07,
+      "loss": 1.7939,
+      "step": 9051
+    },
+    {
+      "epoch": 0.9402721512413005,
+      "grad_norm": 0.48101523518562317,
+      "learning_rate": 8.776449740058645e-07,
+      "loss": 1.7245,
+      "step": 9052
+    },
+    {
+      "epoch": 0.9403760257608809,
+      "grad_norm": 0.41619133949279785,
+      "learning_rate": 8.746038735316931e-07,
+      "loss": 1.5208,
+      "step": 9053
+    },
+    {
+      "epoch": 0.9404799002804612,
+      "grad_norm": 0.46501943469047546,
+      "learning_rate": 8.715680045290165e-07,
+      "loss": 1.8578,
+      "step": 9054
+    },
+    {
+      "epoch": 0.9405837748000415,
+      "grad_norm": 0.443288654088974,
+      "learning_rate": 8.685373673211372e-07,
+      "loss": 1.6814,
+      "step": 9055
+    },
+    {
+      "epoch": 0.9406876493196219,
+      "grad_norm": 0.39399245381355286,
+      "learning_rate": 8.655119622307806e-07,
+      "loss": 1.327,
+      "step": 9056
+    },
+    {
+      "epoch": 0.9407915238392023,
+      "grad_norm": 0.40905261039733887,
+      "learning_rate": 8.624917895801388e-07,
+      "loss": 1.6839,
+      "step": 9057
+    },
+    {
+      "epoch": 0.9408953983587826,
+      "grad_norm": 0.4420108199119568,
+      "learning_rate": 8.594768496908378e-07,
+      "loss": 1.5596,
+      "step": 9058
+    },
+    {
+      "epoch": 0.9409992728783629,
+      "grad_norm": 0.3948979675769806,
+      "learning_rate": 8.564671428839321e-07,
+      "loss": 1.5437,
+      "step": 9059
+    },
+    {
+      "epoch": 0.9411031473979433,
+      "grad_norm": 0.415421724319458,
+      "learning_rate": 8.534626694799485e-07,
+      "loss": 1.5591,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9412070219175236,
+      "grad_norm": 0.41926440596580505,
+      "learning_rate": 8.504634297988312e-07,
+      "loss": 1.7288,
+      "step": 9061
+    },
+    {
+      "epoch": 0.941310896437104,
+      "grad_norm": 0.3910306990146637,
+      "learning_rate": 8.4746942415998e-07,
+      "loss": 1.5736,
+      "step": 9062
+    },
+    {
+      "epoch": 0.9414147709566844,
+      "grad_norm": 0.4041973054409027,
+      "learning_rate": 8.444806528822291e-07,
+      "loss": 1.5708,
+      "step": 9063
+    },
+    {
+      "epoch": 0.9415186454762646,
+      "grad_norm": 0.423279732465744,
+      "learning_rate": 8.41497116283857e-07,
+      "loss": 1.6112,
+      "step": 9064
+    },
+    {
+      "epoch": 0.941622519995845,
+      "grad_norm": 0.4005332589149475,
+      "learning_rate": 8.385188146825984e-07,
+      "loss": 1.6238,
+      "step": 9065
+    },
+    {
+      "epoch": 0.9417263945154254,
+      "grad_norm": 0.4571692943572998,
+      "learning_rate": 8.355457483955998e-07,
+      "loss": 1.8539,
+      "step": 9066
+    },
+    {
+      "epoch": 0.9418302690350057,
+      "grad_norm": 0.4386880695819855,
+      "learning_rate": 8.325779177394965e-07,
+      "loss": 1.709,
+      "step": 9067
+    },
+    {
+      "epoch": 0.941934143554586,
+      "grad_norm": 0.4796215891838074,
+      "learning_rate": 8.296153230303139e-07,
+      "loss": 1.8527,
+      "step": 9068
+    },
+    {
+      "epoch": 0.9420380180741664,
+      "grad_norm": 0.40631356835365295,
+      "learning_rate": 8.266579645835548e-07,
+      "loss": 1.4964,
+      "step": 9069
+    },
+    {
+      "epoch": 0.9421418925937467,
+      "grad_norm": 0.40857720375061035,
+      "learning_rate": 8.237058427141619e-07,
+      "loss": 1.5452,
+      "step": 9070
+    },
+    {
+      "epoch": 0.9422457671133271,
+      "grad_norm": 0.4160623550415039,
+      "learning_rate": 8.20758957736506e-07,
+      "loss": 1.6578,
+      "step": 9071
+    },
+    {
+      "epoch": 0.9423496416329075,
+      "grad_norm": 0.4103771448135376,
+      "learning_rate": 8.178173099644082e-07,
+      "loss": 1.4836,
+      "step": 9072
+    },
+    {
+      "epoch": 0.9424535161524878,
+      "grad_norm": 0.39245137572288513,
+      "learning_rate": 8.148808997111234e-07,
+      "loss": 1.4885,
+      "step": 9073
+    },
+    {
+      "epoch": 0.9425573906720681,
+      "grad_norm": 0.4215735197067261,
+      "learning_rate": 8.119497272893739e-07,
+      "loss": 1.4961,
+      "step": 9074
+    },
+    {
+      "epoch": 0.9426612651916485,
+      "grad_norm": 0.41241350769996643,
+      "learning_rate": 8.090237930112932e-07,
+      "loss": 1.7644,
+      "step": 9075
+    },
+    {
+      "epoch": 0.9427651397112289,
+      "grad_norm": 0.41640931367874146,
+      "learning_rate": 8.061030971884709e-07,
+      "loss": 1.7283,
+      "step": 9076
+    },
+    {
+      "epoch": 0.9428690142308092,
+      "grad_norm": 0.42004087567329407,
+      "learning_rate": 8.031876401319471e-07,
+      "loss": 1.7438,
+      "step": 9077
+    },
+    {
+      "epoch": 0.9429728887503895,
+      "grad_norm": 0.4231506884098053,
+      "learning_rate": 8.002774221521903e-07,
+      "loss": 1.5288,
+      "step": 9078
+    },
+    {
+      "epoch": 0.9430767632699699,
+      "grad_norm": 0.4708472192287445,
+      "learning_rate": 7.973724435591135e-07,
+      "loss": 1.8106,
+      "step": 9079
+    },
+    {
+      "epoch": 0.9431806377895502,
+      "grad_norm": 0.4298073649406433,
+      "learning_rate": 7.944727046620748e-07,
+      "loss": 1.7294,
+      "step": 9080
+    },
+    {
+      "epoch": 0.9432845123091306,
+      "grad_norm": 0.44677823781967163,
+      "learning_rate": 7.915782057698772e-07,
+      "loss": 1.6364,
+      "step": 9081
+    },
+    {
+      "epoch": 0.943388386828711,
+      "grad_norm": 0.4536736011505127,
+      "learning_rate": 7.886889471907521e-07,
+      "loss": 1.7965,
+      "step": 9082
+    },
+    {
+      "epoch": 0.9434922613482912,
+      "grad_norm": 0.4077182114124298,
+      "learning_rate": 7.858049292324032e-07,
+      "loss": 1.6954,
+      "step": 9083
+    },
+    {
+      "epoch": 0.9435961358678716,
+      "grad_norm": 0.4194972813129425,
+      "learning_rate": 7.829261522019349e-07,
+      "loss": 1.5585,
+      "step": 9084
+    },
+    {
+      "epoch": 0.943700010387452,
+      "grad_norm": 0.4863661825656891,
+      "learning_rate": 7.800526164059296e-07,
+      "loss": 1.6892,
+      "step": 9085
+    },
+    {
+      "epoch": 0.9438038849070323,
+      "grad_norm": 0.49641919136047363,
+      "learning_rate": 7.771843221503927e-07,
+      "loss": 1.816,
+      "step": 9086
+    },
+    {
+      "epoch": 0.9439077594266126,
+      "grad_norm": 0.3937380015850067,
+      "learning_rate": 7.743212697407631e-07,
+      "loss": 1.5559,
+      "step": 9087
+    },
+    {
+      "epoch": 0.944011633946193,
+      "grad_norm": 0.45555490255355835,
+      "learning_rate": 7.71463459481947e-07,
+      "loss": 1.7509,
+      "step": 9088
+    },
+    {
+      "epoch": 0.9441155084657733,
+      "grad_norm": 0.3873693645000458,
+      "learning_rate": 7.686108916782786e-07,
+      "loss": 1.4442,
+      "step": 9089
+    },
+    {
+      "epoch": 0.9442193829853537,
+      "grad_norm": 0.4046841859817505,
+      "learning_rate": 7.657635666335316e-07,
+      "loss": 1.5161,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9443232575049341,
+      "grad_norm": 0.41723448038101196,
+      "learning_rate": 7.629214846509191e-07,
+      "loss": 1.7279,
+      "step": 9091
+    },
+    {
+      "epoch": 0.9444271320245143,
+      "grad_norm": 0.39067700505256653,
+      "learning_rate": 7.600846460331046e-07,
+      "loss": 1.6004,
+      "step": 9092
+    },
+    {
+      "epoch": 0.9445310065440947,
+      "grad_norm": 0.434877872467041,
+      "learning_rate": 7.572530510821907e-07,
+      "loss": 1.6926,
+      "step": 9093
+    },
+    {
+      "epoch": 0.9446348810636751,
+      "grad_norm": 0.40955984592437744,
+      "learning_rate": 7.544267000997141e-07,
+      "loss": 1.6929,
+      "step": 9094
+    },
+    {
+      "epoch": 0.9447387555832554,
+      "grad_norm": 0.4047926962375641,
+      "learning_rate": 7.516055933866728e-07,
+      "loss": 1.6725,
+      "step": 9095
+    },
+    {
+      "epoch": 0.9448426301028358,
+      "grad_norm": 0.42006105184555054,
+      "learning_rate": 7.487897312434711e-07,
+      "loss": 1.76,
+      "step": 9096
+    },
+    {
+      "epoch": 0.9449465046224161,
+      "grad_norm": 0.4147946536540985,
+      "learning_rate": 7.459791139699968e-07,
+      "loss": 1.6635,
+      "step": 9097
+    },
+    {
+      "epoch": 0.9450503791419965,
+      "grad_norm": 0.440835565328598,
+      "learning_rate": 7.431737418655494e-07,
+      "loss": 1.7047,
+      "step": 9098
+    },
+    {
+      "epoch": 0.9451542536615768,
+      "grad_norm": 0.38934168219566345,
+      "learning_rate": 7.403736152288842e-07,
+      "loss": 1.4739,
+      "step": 9099
+    },
+    {
+      "epoch": 0.9452581281811572,
+      "grad_norm": 0.45924264192581177,
+      "learning_rate": 7.375787343581852e-07,
+      "loss": 1.8124,
+      "step": 9100
+    },
+    {
+      "epoch": 0.9453620027007376,
+      "grad_norm": 0.45659106969833374,
+      "learning_rate": 7.347890995510864e-07,
+      "loss": 1.7298,
+      "step": 9101
+    },
+    {
+      "epoch": 0.9454658772203178,
+      "grad_norm": 0.4247036278247833,
+      "learning_rate": 7.320047111046724e-07,
+      "loss": 1.6317,
+      "step": 9102
+    },
+    {
+      "epoch": 0.9455697517398982,
+      "grad_norm": 0.41282743215560913,
+      "learning_rate": 7.292255693154449e-07,
+      "loss": 1.6771,
+      "step": 9103
+    },
+    {
+      "epoch": 0.9456736262594786,
+      "grad_norm": 0.41268548369407654,
+      "learning_rate": 7.264516744793671e-07,
+      "loss": 1.6163,
+      "step": 9104
+    },
+    {
+      "epoch": 0.9457775007790589,
+      "grad_norm": 0.3944796621799469,
+      "learning_rate": 7.236830268918415e-07,
+      "loss": 1.6949,
+      "step": 9105
+    },
+    {
+      "epoch": 0.9458813752986392,
+      "grad_norm": 0.4076094925403595,
+      "learning_rate": 7.209196268477047e-07,
+      "loss": 1.5172,
+      "step": 9106
+    },
+    {
+      "epoch": 0.9459852498182196,
+      "grad_norm": 0.41099345684051514,
+      "learning_rate": 7.181614746412324e-07,
+      "loss": 1.5826,
+      "step": 9107
+    },
+    {
+      "epoch": 0.9460891243377999,
+      "grad_norm": 0.5668470859527588,
+      "learning_rate": 7.154085705661506e-07,
+      "loss": 1.8974,
+      "step": 9108
+    },
+    {
+      "epoch": 0.9461929988573803,
+      "grad_norm": 0.4689236879348755,
+      "learning_rate": 7.126609149156249e-07,
+      "loss": 1.7432,
+      "step": 9109
+    },
+    {
+      "epoch": 0.9462968733769607,
+      "grad_norm": 0.42818373441696167,
+      "learning_rate": 7.09918507982249e-07,
+      "loss": 1.614,
+      "step": 9110
+    },
+    {
+      "epoch": 0.9464007478965409,
+      "grad_norm": 0.40631645917892456,
+      "learning_rate": 7.071813500580837e-07,
+      "loss": 1.6488,
+      "step": 9111
+    },
+    {
+      "epoch": 0.9465046224161213,
+      "grad_norm": 0.38255685567855835,
+      "learning_rate": 7.044494414346015e-07,
+      "loss": 1.5024,
+      "step": 9112
+    },
+    {
+      "epoch": 0.9466084969357017,
+      "grad_norm": 0.42258989810943604,
+      "learning_rate": 7.017227824027305e-07,
+      "loss": 1.6628,
+      "step": 9113
+    },
+    {
+      "epoch": 0.946712371455282,
+      "grad_norm": 0.40682870149612427,
+      "learning_rate": 6.990013732528444e-07,
+      "loss": 1.6693,
+      "step": 9114
+    },
+    {
+      "epoch": 0.9468162459748624,
+      "grad_norm": 0.4137510657310486,
+      "learning_rate": 6.9628521427475e-07,
+      "loss": 1.5755,
+      "step": 9115
+    },
+    {
+      "epoch": 0.9469201204944427,
+      "grad_norm": 0.446568101644516,
+      "learning_rate": 6.935743057576938e-07,
+      "loss": 1.6624,
+      "step": 9116
+    },
+    {
+      "epoch": 0.947023995014023,
+      "grad_norm": 0.41516193747520447,
+      "learning_rate": 6.908686479903725e-07,
+      "loss": 1.6738,
+      "step": 9117
+    },
+    {
+      "epoch": 0.9471278695336034,
+      "grad_norm": 0.4611627459526062,
+      "learning_rate": 6.881682412609114e-07,
+      "loss": 1.7311,
+      "step": 9118
+    },
+    {
+      "epoch": 0.9472317440531838,
+      "grad_norm": 0.5270797610282898,
+      "learning_rate": 6.854730858568914e-07,
+      "loss": 1.7126,
+      "step": 9119
+    },
+    {
+      "epoch": 0.947335618572764,
+      "grad_norm": 0.41994228959083557,
+      "learning_rate": 6.827831820653163e-07,
+      "loss": 1.7939,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9474394930923444,
+      "grad_norm": 0.44093456864356995,
+      "learning_rate": 6.800985301726404e-07,
+      "loss": 1.6198,
+      "step": 9121
+    },
+    {
+      "epoch": 0.9475433676119248,
+      "grad_norm": 0.4290820360183716,
+      "learning_rate": 6.77419130464757e-07,
+      "loss": 1.73,
+      "step": 9122
+    },
+    {
+      "epoch": 0.9476472421315052,
+      "grad_norm": 0.4182687997817993,
+      "learning_rate": 6.747449832270158e-07,
+      "loss": 1.6677,
+      "step": 9123
+    },
+    {
+      "epoch": 0.9477511166510855,
+      "grad_norm": 0.4348759949207306,
+      "learning_rate": 6.720760887441779e-07,
+      "loss": 1.8017,
+      "step": 9124
+    },
+    {
+      "epoch": 0.9478549911706659,
+      "grad_norm": 0.40973037481307983,
+      "learning_rate": 6.694124473004604e-07,
+      "loss": 1.6173,
+      "step": 9125
+    },
+    {
+      "epoch": 0.9479588656902462,
+      "grad_norm": 0.4033758342266083,
+      "learning_rate": 6.667540591795307e-07,
+      "loss": 1.5386,
+      "step": 9126
+    },
+    {
+      "epoch": 0.9480627402098265,
+      "grad_norm": 0.4231725037097931,
+      "learning_rate": 6.641009246644736e-07,
+      "loss": 1.652,
+      "step": 9127
+    },
+    {
+      "epoch": 0.9481666147294069,
+      "grad_norm": 0.40688326954841614,
+      "learning_rate": 6.614530440378352e-07,
+      "loss": 1.6695,
+      "step": 9128
+    },
+    {
+      "epoch": 0.9482704892489873,
+      "grad_norm": 0.38211679458618164,
+      "learning_rate": 6.588104175815901e-07,
+      "loss": 1.4878,
+      "step": 9129
+    },
+    {
+      "epoch": 0.9483743637685675,
+      "grad_norm": 0.4489685595035553,
+      "learning_rate": 6.561730455771686e-07,
+      "loss": 1.6489,
+      "step": 9130
+    },
+    {
+      "epoch": 0.9484782382881479,
+      "grad_norm": 0.4730764627456665,
+      "learning_rate": 6.535409283054073e-07,
+      "loss": 1.6737,
+      "step": 9131
+    },
+    {
+      "epoch": 0.9485821128077283,
+      "grad_norm": 0.38945716619491577,
+      "learning_rate": 6.509140660466318e-07,
+      "loss": 1.6735,
+      "step": 9132
+    },
+    {
+      "epoch": 0.9486859873273086,
+      "grad_norm": 0.4208957254886627,
+      "learning_rate": 6.482924590805627e-07,
+      "loss": 1.6911,
+      "step": 9133
+    },
+    {
+      "epoch": 0.948789861846889,
+      "grad_norm": 0.5277408361434937,
+      "learning_rate": 6.456761076863937e-07,
+      "loss": 1.5338,
+      "step": 9134
+    },
+    {
+      "epoch": 0.9488937363664693,
+      "grad_norm": 0.39843320846557617,
+      "learning_rate": 6.430650121427406e-07,
+      "loss": 1.5234,
+      "step": 9135
+    },
+    {
+      "epoch": 0.9489976108860496,
+      "grad_norm": 0.42443880438804626,
+      "learning_rate": 6.404591727276588e-07,
+      "loss": 1.7075,
+      "step": 9136
+    },
+    {
+      "epoch": 0.94910148540563,
+      "grad_norm": 0.4635830223560333,
+      "learning_rate": 6.378585897186651e-07,
+      "loss": 1.6455,
+      "step": 9137
+    },
+    {
+      "epoch": 0.9492053599252104,
+      "grad_norm": 0.39471131563186646,
+      "learning_rate": 6.352632633926825e-07,
+      "loss": 1.5482,
+      "step": 9138
+    },
+    {
+      "epoch": 0.9493092344447907,
+      "grad_norm": 0.45582547783851624,
+      "learning_rate": 6.326731940261122e-07,
+      "loss": 1.7572,
+      "step": 9139
+    },
+    {
+      "epoch": 0.949413108964371,
+      "grad_norm": 0.4490654170513153,
+      "learning_rate": 6.300883818947611e-07,
+      "loss": 1.6723,
+      "step": 9140
+    },
+    {
+      "epoch": 0.9495169834839514,
+      "grad_norm": 0.38818129897117615,
+      "learning_rate": 6.275088272738927e-07,
+      "loss": 1.7014,
+      "step": 9141
+    },
+    {
+      "epoch": 0.9496208580035317,
+      "grad_norm": 0.4169076383113861,
+      "learning_rate": 6.249345304382259e-07,
+      "loss": 1.6569,
+      "step": 9142
+    },
+    {
+      "epoch": 0.9497247325231121,
+      "grad_norm": 0.39485812187194824,
+      "learning_rate": 6.223654916618804e-07,
+      "loss": 1.6363,
+      "step": 9143
+    },
+    {
+      "epoch": 0.9498286070426925,
+      "grad_norm": 0.41402456164360046,
+      "learning_rate": 6.19801711218454e-07,
+      "loss": 1.5838,
+      "step": 9144
+    },
+    {
+      "epoch": 0.9499324815622727,
+      "grad_norm": 0.4002974331378937,
+      "learning_rate": 6.172431893809616e-07,
+      "loss": 1.4252,
+      "step": 9145
+    },
+    {
+      "epoch": 0.9500363560818531,
+      "grad_norm": 0.42501163482666016,
+      "learning_rate": 6.146899264218742e-07,
+      "loss": 1.5948,
+      "step": 9146
+    },
+    {
+      "epoch": 0.9501402306014335,
+      "grad_norm": 0.4167088270187378,
+      "learning_rate": 6.121419226130909e-07,
+      "loss": 1.5962,
+      "step": 9147
+    },
+    {
+      "epoch": 0.9502441051210138,
+      "grad_norm": 0.42873653769493103,
+      "learning_rate": 6.095991782259447e-07,
+      "loss": 1.6758,
+      "step": 9148
+    },
+    {
+      "epoch": 0.9503479796405941,
+      "grad_norm": 0.37248915433883667,
+      "learning_rate": 6.070616935312357e-07,
+      "loss": 1.4357,
+      "step": 9149
+    },
+    {
+      "epoch": 0.9504518541601745,
+      "grad_norm": 0.4385361671447754,
+      "learning_rate": 6.045294687991643e-07,
+      "loss": 1.8251,
+      "step": 9150
+    },
+    {
+      "epoch": 0.9505557286797549,
+      "grad_norm": 0.44529131054878235,
+      "learning_rate": 6.020025042994148e-07,
+      "loss": 1.8014,
+      "step": 9151
+    },
+    {
+      "epoch": 0.9506596031993352,
+      "grad_norm": 0.4091431498527527,
+      "learning_rate": 5.994808003010777e-07,
+      "loss": 1.5917,
+      "step": 9152
+    },
+    {
+      "epoch": 0.9507634777189156,
+      "grad_norm": 0.4284379780292511,
+      "learning_rate": 5.969643570726935e-07,
+      "loss": 1.5155,
+      "step": 9153
+    },
+    {
+      "epoch": 0.9508673522384959,
+      "grad_norm": 0.4677438735961914,
+      "learning_rate": 5.944531748822535e-07,
+      "loss": 1.489,
+      "step": 9154
+    },
+    {
+      "epoch": 0.9509712267580762,
+      "grad_norm": 0.4163029193878174,
+      "learning_rate": 5.91947253997166e-07,
+      "loss": 1.4501,
+      "step": 9155
+    },
+    {
+      "epoch": 0.9510751012776566,
+      "grad_norm": 0.41146647930145264,
+      "learning_rate": 5.894465946843008e-07,
+      "loss": 1.5144,
+      "step": 9156
+    },
+    {
+      "epoch": 0.951178975797237,
+      "grad_norm": 0.4144413471221924,
+      "learning_rate": 5.869511972099506e-07,
+      "loss": 1.6731,
+      "step": 9157
+    },
+    {
+      "epoch": 0.9512828503168173,
+      "grad_norm": 0.4550022482872009,
+      "learning_rate": 5.844610618398693e-07,
+      "loss": 1.7321,
+      "step": 9158
+    },
+    {
+      "epoch": 0.9513867248363976,
+      "grad_norm": 0.4146331548690796,
+      "learning_rate": 5.819761888392228e-07,
+      "loss": 1.7712,
+      "step": 9159
+    },
+    {
+      "epoch": 0.951490599355978,
+      "grad_norm": 0.4827490746974945,
+      "learning_rate": 5.794965784726436e-07,
+      "loss": 1.7595,
+      "step": 9160
+    },
+    {
+      "epoch": 0.9515944738755583,
+      "grad_norm": 0.4096835255622864,
+      "learning_rate": 5.770222310041817e-07,
+      "loss": 1.7085,
+      "step": 9161
+    },
+    {
+      "epoch": 0.9516983483951387,
+      "grad_norm": 0.4233892261981964,
+      "learning_rate": 5.745531466973375e-07,
+      "loss": 1.7196,
+      "step": 9162
+    },
+    {
+      "epoch": 0.951802222914719,
+      "grad_norm": 0.457246333360672,
+      "learning_rate": 5.720893258150451e-07,
+      "loss": 1.7608,
+      "step": 9163
+    },
+    {
+      "epoch": 0.9519060974342993,
+      "grad_norm": 0.40001896023750305,
+      "learning_rate": 5.696307686196944e-07,
+      "loss": 1.6069,
+      "step": 9164
+    },
+    {
+      "epoch": 0.9520099719538797,
+      "grad_norm": 0.40769293904304504,
+      "learning_rate": 5.67177475373093e-07,
+      "loss": 1.5777,
+      "step": 9165
+    },
+    {
+      "epoch": 0.9521138464734601,
+      "grad_norm": 0.4234029948711395,
+      "learning_rate": 5.647294463364927e-07,
+      "loss": 1.8119,
+      "step": 9166
+    },
+    {
+      "epoch": 0.9522177209930404,
+      "grad_norm": 0.4327137768268585,
+      "learning_rate": 5.622866817706074e-07,
+      "loss": 1.5702,
+      "step": 9167
+    },
+    {
+      "epoch": 0.9523215955126207,
+      "grad_norm": 0.4188234806060791,
+      "learning_rate": 5.598491819355567e-07,
+      "loss": 1.679,
+      "step": 9168
+    },
+    {
+      "epoch": 0.9524254700322011,
+      "grad_norm": 0.44610387086868286,
+      "learning_rate": 5.574169470909163e-07,
+      "loss": 1.6981,
+      "step": 9169
+    },
+    {
+      "epoch": 0.9525293445517814,
+      "grad_norm": 0.4079190194606781,
+      "learning_rate": 5.549899774957123e-07,
+      "loss": 1.5524,
+      "step": 9170
+    },
+    {
+      "epoch": 0.9526332190713618,
+      "grad_norm": 0.4039902985095978,
+      "learning_rate": 5.52568273408377e-07,
+      "loss": 1.7129,
+      "step": 9171
+    },
+    {
+      "epoch": 0.9527370935909422,
+      "grad_norm": 0.4110052287578583,
+      "learning_rate": 5.501518350868262e-07,
+      "loss": 1.5757,
+      "step": 9172
+    },
+    {
+      "epoch": 0.9528409681105224,
+      "grad_norm": 0.3944413959980011,
+      "learning_rate": 5.477406627883763e-07,
+      "loss": 1.6487,
+      "step": 9173
+    },
+    {
+      "epoch": 0.9529448426301028,
+      "grad_norm": 0.4401375949382782,
+      "learning_rate": 5.453347567698108e-07,
+      "loss": 1.7217,
+      "step": 9174
+    },
+    {
+      "epoch": 0.9530487171496832,
+      "grad_norm": 0.40892985463142395,
+      "learning_rate": 5.429341172873248e-07,
+      "loss": 1.644,
+      "step": 9175
+    },
+    {
+      "epoch": 0.9531525916692636,
+      "grad_norm": 0.40419328212738037,
+      "learning_rate": 5.405387445965748e-07,
+      "loss": 1.5666,
+      "step": 9176
+    },
+    {
+      "epoch": 0.9532564661888439,
+      "grad_norm": 0.4269610643386841,
+      "learning_rate": 5.381486389526569e-07,
+      "loss": 1.5313,
+      "step": 9177
+    },
+    {
+      "epoch": 0.9533603407084242,
+      "grad_norm": 0.4391513764858246,
+      "learning_rate": 5.35763800610084e-07,
+      "loss": 1.6146,
+      "step": 9178
+    },
+    {
+      "epoch": 0.9534642152280046,
+      "grad_norm": 0.4620340168476105,
+      "learning_rate": 5.333842298228364e-07,
+      "loss": 1.7637,
+      "step": 9179
+    },
+    {
+      "epoch": 0.9535680897475849,
+      "grad_norm": 0.44196537137031555,
+      "learning_rate": 5.310099268443114e-07,
+      "loss": 1.7589,
+      "step": 9180
+    },
+    {
+      "epoch": 0.9536719642671653,
+      "grad_norm": 0.42510437965393066,
+      "learning_rate": 5.286408919273567e-07,
+      "loss": 1.7186,
+      "step": 9181
+    },
+    {
+      "epoch": 0.9537758387867457,
+      "grad_norm": 0.4449099600315094,
+      "learning_rate": 5.262771253242593e-07,
+      "loss": 1.698,
+      "step": 9182
+    },
+    {
+      "epoch": 0.9538797133063259,
+      "grad_norm": 0.4278699457645416,
+      "learning_rate": 5.239186272867347e-07,
+      "loss": 1.4785,
+      "step": 9183
+    },
+    {
+      "epoch": 0.9539835878259063,
+      "grad_norm": 0.40362176299095154,
+      "learning_rate": 5.215653980659485e-07,
+      "loss": 1.7033,
+      "step": 9184
+    },
+    {
+      "epoch": 0.9540874623454867,
+      "grad_norm": 0.4301554262638092,
+      "learning_rate": 5.192174379124948e-07,
+      "loss": 1.6292,
+      "step": 9185
+    },
+    {
+      "epoch": 0.954191336865067,
+      "grad_norm": 0.4313088655471802,
+      "learning_rate": 5.16874747076429e-07,
+      "loss": 1.5956,
+      "step": 9186
+    },
+    {
+      "epoch": 0.9542952113846473,
+      "grad_norm": 0.43892988562583923,
+      "learning_rate": 5.145373258072129e-07,
+      "loss": 1.7799,
+      "step": 9187
+    },
+    {
+      "epoch": 0.9543990859042277,
+      "grad_norm": 0.4435426890850067,
+      "learning_rate": 5.122051743537692e-07,
+      "loss": 1.7143,
+      "step": 9188
+    },
+    {
+      "epoch": 0.954502960423808,
+      "grad_norm": 0.4262649416923523,
+      "learning_rate": 5.098782929644607e-07,
+      "loss": 1.5772,
+      "step": 9189
+    },
+    {
+      "epoch": 0.9546068349433884,
+      "grad_norm": 0.4397432804107666,
+      "learning_rate": 5.075566818870725e-07,
+      "loss": 1.7397,
+      "step": 9190
+    },
+    {
+      "epoch": 0.9547107094629688,
+      "grad_norm": 0.42559805512428284,
+      "learning_rate": 5.0524034136884e-07,
+      "loss": 1.6717,
+      "step": 9191
+    },
+    {
+      "epoch": 0.954814583982549,
+      "grad_norm": 0.40015941858291626,
+      "learning_rate": 5.029292716564382e-07,
+      "loss": 1.6091,
+      "step": 9192
+    },
+    {
+      "epoch": 0.9549184585021294,
+      "grad_norm": 0.43587568402290344,
+      "learning_rate": 5.006234729959813e-07,
+      "loss": 1.6914,
+      "step": 9193
+    },
+    {
+      "epoch": 0.9550223330217098,
+      "grad_norm": 0.40942129492759705,
+      "learning_rate": 4.983229456330063e-07,
+      "loss": 1.5895,
+      "step": 9194
+    },
+    {
+      "epoch": 0.9551262075412901,
+      "grad_norm": 0.45830440521240234,
+      "learning_rate": 4.960276898125227e-07,
+      "loss": 1.7004,
+      "step": 9195
+    },
+    {
+      "epoch": 0.9552300820608705,
+      "grad_norm": 0.45352327823638916,
+      "learning_rate": 4.93737705778935e-07,
+      "loss": 1.684,
+      "step": 9196
+    },
+    {
+      "epoch": 0.9553339565804508,
+      "grad_norm": 0.4297914206981659,
+      "learning_rate": 4.914529937761147e-07,
+      "loss": 1.8311,
+      "step": 9197
+    },
+    {
+      "epoch": 0.9554378311000311,
+      "grad_norm": 0.40148767828941345,
+      "learning_rate": 4.891735540473785e-07,
+      "loss": 1.5433,
+      "step": 9198
+    },
+    {
+      "epoch": 0.9555417056196115,
+      "grad_norm": 0.433064341545105,
+      "learning_rate": 4.868993868354544e-07,
+      "loss": 1.6219,
+      "step": 9199
+    },
+    {
+      "epoch": 0.9556455801391919,
+      "grad_norm": 0.4220326542854309,
+      "learning_rate": 4.846304923825263e-07,
+      "loss": 1.6918,
+      "step": 9200
+    },
+    {
+      "epoch": 0.9557494546587723,
+      "grad_norm": 0.43805915117263794,
+      "learning_rate": 4.823668709302176e-07,
+      "loss": 1.5888,
+      "step": 9201
+    },
+    {
+      "epoch": 0.9558533291783525,
+      "grad_norm": 0.4448419511318207,
+      "learning_rate": 4.801085227195856e-07,
+      "loss": 1.6003,
+      "step": 9202
+    },
+    {
+      "epoch": 0.9559572036979329,
+      "grad_norm": 0.42299073934555054,
+      "learning_rate": 4.778554479911268e-07,
+      "loss": 1.6384,
+      "step": 9203
+    },
+    {
+      "epoch": 0.9560610782175133,
+      "grad_norm": 0.405784547328949,
+      "learning_rate": 4.7560764698477146e-07,
+      "loss": 1.6106,
+      "step": 9204
+    },
+    {
+      "epoch": 0.9561649527370936,
+      "grad_norm": 0.41542530059814453,
+      "learning_rate": 4.7336511993990586e-07,
+      "loss": 1.6064,
+      "step": 9205
+    },
+    {
+      "epoch": 0.956268827256674,
+      "grad_norm": 0.45177707076072693,
+      "learning_rate": 4.711278670953223e-07,
+      "loss": 1.8089,
+      "step": 9206
+    },
+    {
+      "epoch": 0.9563727017762543,
+      "grad_norm": 0.45689210295677185,
+      "learning_rate": 4.688958886892858e-07,
+      "loss": 1.5888,
+      "step": 9207
+    },
+    {
+      "epoch": 0.9564765762958346,
+      "grad_norm": 0.4608011245727539,
+      "learning_rate": 4.666691849594784e-07,
+      "loss": 1.8167,
+      "step": 9208
+    },
+    {
+      "epoch": 0.956580450815415,
+      "grad_norm": 0.40499821305274963,
+      "learning_rate": 4.6444775614302714e-07,
+      "loss": 1.6155,
+      "step": 9209
+    },
+    {
+      "epoch": 0.9566843253349954,
+      "grad_norm": 0.3993094265460968,
+      "learning_rate": 4.6223160247650386e-07,
+      "loss": 1.5941,
+      "step": 9210
+    },
+    {
+      "epoch": 0.9567881998545756,
+      "grad_norm": 0.4130497872829437,
+      "learning_rate": 4.600207241958976e-07,
+      "loss": 1.5724,
+      "step": 9211
+    },
+    {
+      "epoch": 0.956892074374156,
+      "grad_norm": 0.4392036497592926,
+      "learning_rate": 4.5781512153665884e-07,
+      "loss": 1.6075,
+      "step": 9212
+    },
+    {
+      "epoch": 0.9569959488937364,
+      "grad_norm": 0.4254901111125946,
+      "learning_rate": 4.5561479473366644e-07,
+      "loss": 1.7026,
+      "step": 9213
+    },
+    {
+      "epoch": 0.9570998234133167,
+      "grad_norm": 0.4199539124965668,
+      "learning_rate": 4.53419744021244e-07,
+      "loss": 1.5035,
+      "step": 9214
+    },
+    {
+      "epoch": 0.9572036979328971,
+      "grad_norm": 0.4554733633995056,
+      "learning_rate": 4.5122996963313234e-07,
+      "loss": 1.6648,
+      "step": 9215
+    },
+    {
+      "epoch": 0.9573075724524774,
+      "grad_norm": 0.4227248430252075,
+      "learning_rate": 4.4904547180253386e-07,
+      "loss": 1.5704,
+      "step": 9216
+    },
+    {
+      "epoch": 0.9574114469720577,
+      "grad_norm": 0.37971073389053345,
+      "learning_rate": 4.468662507620847e-07,
+      "loss": 1.5082,
+      "step": 9217
+    },
+    {
+      "epoch": 0.9575153214916381,
+      "grad_norm": 0.40242883563041687,
+      "learning_rate": 4.446923067438491e-07,
+      "loss": 1.5791,
+      "step": 9218
+    },
+    {
+      "epoch": 0.9576191960112185,
+      "grad_norm": 0.43031686544418335,
+      "learning_rate": 4.425236399793364e-07,
+      "loss": 1.6155,
+      "step": 9219
+    },
+    {
+      "epoch": 0.9577230705307987,
+      "grad_norm": 0.4682821035385132,
+      "learning_rate": 4.403602506994897e-07,
+      "loss": 1.7276,
+      "step": 9220
+    },
+    {
+      "epoch": 0.9578269450503791,
+      "grad_norm": 0.4437069892883301,
+      "learning_rate": 4.382021391346969e-07,
+      "loss": 1.6508,
+      "step": 9221
+    },
+    {
+      "epoch": 0.9579308195699595,
+      "grad_norm": 0.38210955262184143,
+      "learning_rate": 4.360493055147796e-07,
+      "loss": 1.5887,
+      "step": 9222
+    },
+    {
+      "epoch": 0.9580346940895398,
+      "grad_norm": 0.4004855751991272,
+      "learning_rate": 4.339017500690046e-07,
+      "loss": 1.4538,
+      "step": 9223
+    },
+    {
+      "epoch": 0.9581385686091202,
+      "grad_norm": 0.41363096237182617,
+      "learning_rate": 4.3175947302605547e-07,
+      "loss": 1.6011,
+      "step": 9224
+    },
+    {
+      "epoch": 0.9582424431287005,
+      "grad_norm": 0.42075902223587036,
+      "learning_rate": 4.296224746140776e-07,
+      "loss": 1.6147,
+      "step": 9225
+    },
+    {
+      "epoch": 0.9583463176482809,
+      "grad_norm": 0.45383062958717346,
+      "learning_rate": 4.2749075506063883e-07,
+      "loss": 1.6448,
+      "step": 9226
+    },
+    {
+      "epoch": 0.9584501921678612,
+      "grad_norm": 0.40463367104530334,
+      "learning_rate": 4.253643145927577e-07,
+      "loss": 1.663,
+      "step": 9227
+    },
+    {
+      "epoch": 0.9585540666874416,
+      "grad_norm": 0.4023562967777252,
+      "learning_rate": 4.232431534368808e-07,
+      "loss": 1.4755,
+      "step": 9228
+    },
+    {
+      "epoch": 0.958657941207022,
+      "grad_norm": 0.4149220883846283,
+      "learning_rate": 4.211272718188941e-07,
+      "loss": 1.6507,
+      "step": 9229
+    },
+    {
+      "epoch": 0.9587618157266022,
+      "grad_norm": 0.42307958006858826,
+      "learning_rate": 4.190166699641229e-07,
+      "loss": 1.7877,
+      "step": 9230
+    },
+    {
+      "epoch": 0.9588656902461826,
+      "grad_norm": 0.4115198254585266,
+      "learning_rate": 4.169113480973319e-07,
+      "loss": 1.5952,
+      "step": 9231
+    },
+    {
+      "epoch": 0.958969564765763,
+      "grad_norm": 0.4087419807910919,
+      "learning_rate": 4.148113064427139e-07,
+      "loss": 1.6713,
+      "step": 9232
+    },
+    {
+      "epoch": 0.9590734392853433,
+      "grad_norm": 0.43182262778282166,
+      "learning_rate": 4.127165452239179e-07,
+      "loss": 1.728,
+      "step": 9233
+    },
+    {
+      "epoch": 0.9591773138049237,
+      "grad_norm": 0.41632604598999023,
+      "learning_rate": 4.106270646640098e-07,
+      "loss": 1.7331,
+      "step": 9234
+    },
+    {
+      "epoch": 0.959281188324504,
+      "grad_norm": 0.4216884970664978,
+      "learning_rate": 4.085428649855061e-07,
+      "loss": 1.7273,
+      "step": 9235
+    },
+    {
+      "epoch": 0.9593850628440843,
+      "grad_norm": 0.39997923374176025,
+      "learning_rate": 4.0646394641036255e-07,
+      "loss": 1.6779,
+      "step": 9236
+    },
+    {
+      "epoch": 0.9594889373636647,
+      "grad_norm": 0.4439750015735626,
+      "learning_rate": 4.043903091599632e-07,
+      "loss": 1.8105,
+      "step": 9237
+    },
+    {
+      "epoch": 0.9595928118832451,
+      "grad_norm": 0.4040851593017578,
+      "learning_rate": 4.02321953455137e-07,
+      "loss": 1.619,
+      "step": 9238
+    },
+    {
+      "epoch": 0.9596966864028253,
+      "grad_norm": 0.40160131454467773,
+      "learning_rate": 4.002588795161466e-07,
+      "loss": 1.5693,
+      "step": 9239
+    },
+    {
+      "epoch": 0.9598005609224057,
+      "grad_norm": 0.49092456698417664,
+      "learning_rate": 3.9820108756268846e-07,
+      "loss": 1.8386,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9599044354419861,
+      "grad_norm": 0.4417165517807007,
+      "learning_rate": 3.9614857781390955e-07,
+      "loss": 1.5334,
+      "step": 9241
+    },
+    {
+      "epoch": 0.9600083099615664,
+      "grad_norm": 0.4167068302631378,
+      "learning_rate": 3.941013504883795e-07,
+      "loss": 1.7032,
+      "step": 9242
+    },
+    {
+      "epoch": 0.9601121844811468,
+      "grad_norm": 0.42593681812286377,
+      "learning_rate": 3.920594058041183e-07,
+      "loss": 1.741,
+      "step": 9243
+    },
+    {
+      "epoch": 0.9602160590007272,
+      "grad_norm": 0.4693407118320465,
+      "learning_rate": 3.9002274397856863e-07,
+      "loss": 1.8403,
+      "step": 9244
+    },
+    {
+      "epoch": 0.9603199335203074,
+      "grad_norm": 0.41874200105667114,
+      "learning_rate": 3.8799136522862933e-07,
+      "loss": 1.6718,
+      "step": 9245
+    },
+    {
+      "epoch": 0.9604238080398878,
+      "grad_norm": 0.42587071657180786,
+      "learning_rate": 3.859652697706162e-07,
+      "loss": 1.5569,
+      "step": 9246
+    },
+    {
+      "epoch": 0.9605276825594682,
+      "grad_norm": 0.4416424334049225,
+      "learning_rate": 3.8394445782030107e-07,
+      "loss": 1.5302,
+      "step": 9247
+    },
+    {
+      "epoch": 0.9606315570790485,
+      "grad_norm": 0.4183494746685028,
+      "learning_rate": 3.819289295928785e-07,
+      "loss": 1.6514,
+      "step": 9248
+    },
+    {
+      "epoch": 0.9607354315986288,
+      "grad_norm": 0.46521294116973877,
+      "learning_rate": 3.7991868530299345e-07,
+      "loss": 1.5298,
+      "step": 9249
+    },
+    {
+      "epoch": 0.9608393061182092,
+      "grad_norm": 0.424907386302948,
+      "learning_rate": 3.7791372516470803e-07,
+      "loss": 1.7081,
+      "step": 9250
+    },
+    {
+      "epoch": 0.9609431806377895,
+      "grad_norm": 0.44091981649398804,
+      "learning_rate": 3.759140493915514e-07,
+      "loss": 1.7594,
+      "step": 9251
+    },
+    {
+      "epoch": 0.9610470551573699,
+      "grad_norm": 0.4381832778453827,
+      "learning_rate": 3.7391965819646435e-07,
+      "loss": 1.5333,
+      "step": 9252
+    },
+    {
+      "epoch": 0.9611509296769503,
+      "grad_norm": 0.41242584586143494,
+      "learning_rate": 3.7193055179183256e-07,
+      "loss": 1.5122,
+      "step": 9253
+    },
+    {
+      "epoch": 0.9612548041965306,
+      "grad_norm": 0.4707975685596466,
+      "learning_rate": 3.6994673038948655e-07,
+      "loss": 1.7078,
+      "step": 9254
+    },
+    {
+      "epoch": 0.9613586787161109,
+      "grad_norm": 0.49493807554244995,
+      "learning_rate": 3.679681942006852e-07,
+      "loss": 1.5883,
+      "step": 9255
+    },
+    {
+      "epoch": 0.9614625532356913,
+      "grad_norm": 0.4006246030330658,
+      "learning_rate": 3.6599494343612095e-07,
+      "loss": 1.6098,
+      "step": 9256
+    },
+    {
+      "epoch": 0.9615664277552717,
+      "grad_norm": 0.4326263964176178,
+      "learning_rate": 3.6402697830594243e-07,
+      "loss": 1.6603,
+      "step": 9257
+    },
+    {
+      "epoch": 0.961670302274852,
+      "grad_norm": 0.39721041917800903,
+      "learning_rate": 3.6206429901970985e-07,
+      "loss": 1.493,
+      "step": 9258
+    },
+    {
+      "epoch": 0.9617741767944323,
+      "grad_norm": 0.41533419489860535,
+      "learning_rate": 3.6010690578643924e-07,
+      "loss": 1.5743,
+      "step": 9259
+    },
+    {
+      "epoch": 0.9618780513140127,
+      "grad_norm": 0.3986194431781769,
+      "learning_rate": 3.58154798814575e-07,
+      "loss": 1.4523,
+      "step": 9260
+    },
+    {
+      "epoch": 0.961981925833593,
+      "grad_norm": 0.46424034237861633,
+      "learning_rate": 3.56207978312012e-07,
+      "loss": 1.6645,
+      "step": 9261
+    },
+    {
+      "epoch": 0.9620858003531734,
+      "grad_norm": 0.44541025161743164,
+      "learning_rate": 3.542664444860566e-07,
+      "loss": 1.5695,
+      "step": 9262
+    },
+    {
+      "epoch": 0.9621896748727538,
+      "grad_norm": 0.39570924639701843,
+      "learning_rate": 3.5233019754347117e-07,
+      "loss": 1.5907,
+      "step": 9263
+    },
+    {
+      "epoch": 0.962293549392334,
+      "grad_norm": 0.4216559827327728,
+      "learning_rate": 3.503992376904575e-07,
+      "loss": 1.7391,
+      "step": 9264
+    },
+    {
+      "epoch": 0.9623974239119144,
+      "grad_norm": 0.3937249481678009,
+      "learning_rate": 3.4847356513263986e-07,
+      "loss": 1.6637,
+      "step": 9265
+    },
+    {
+      "epoch": 0.9625012984314948,
+      "grad_norm": 0.43053656816482544,
+      "learning_rate": 3.4655318007509316e-07,
+      "loss": 1.5119,
+      "step": 9266
+    },
+    {
+      "epoch": 0.9626051729510751,
+      "grad_norm": 0.423225998878479,
+      "learning_rate": 3.4463808272232055e-07,
+      "loss": 1.5577,
+      "step": 9267
+    },
+    {
+      "epoch": 0.9627090474706554,
+      "grad_norm": 0.3969104588031769,
+      "learning_rate": 3.427282732782644e-07,
+      "loss": 1.6157,
+      "step": 9268
+    },
+    {
+      "epoch": 0.9628129219902358,
+      "grad_norm": 0.42695853114128113,
+      "learning_rate": 3.4082375194630643e-07,
+      "loss": 1.5988,
+      "step": 9269
+    },
+    {
+      "epoch": 0.9629167965098161,
+      "grad_norm": 0.4294392764568329,
+      "learning_rate": 3.3892451892926227e-07,
+      "loss": 1.5232,
+      "step": 9270
+    },
+    {
+      "epoch": 0.9630206710293965,
+      "grad_norm": 0.45847055315971375,
+      "learning_rate": 3.370305744293867e-07,
+      "loss": 1.7633,
+      "step": 9271
+    },
+    {
+      "epoch": 0.9631245455489769,
+      "grad_norm": 0.4628918170928955,
+      "learning_rate": 3.351419186483684e-07,
+      "loss": 1.7437,
+      "step": 9272
+    },
+    {
+      "epoch": 0.9632284200685571,
+      "grad_norm": 0.40736517310142517,
+      "learning_rate": 3.3325855178733543e-07,
+      "loss": 1.6749,
+      "step": 9273
+    },
+    {
+      "epoch": 0.9633322945881375,
+      "grad_norm": 0.43378880620002747,
+      "learning_rate": 3.3138047404684957e-07,
+      "loss": 1.7485,
+      "step": 9274
+    },
+    {
+      "epoch": 0.9634361691077179,
+      "grad_norm": 0.40924814343452454,
+      "learning_rate": 3.295076856269119e-07,
+      "loss": 1.6725,
+      "step": 9275
+    },
+    {
+      "epoch": 0.9635400436272982,
+      "grad_norm": 0.40728503465652466,
+      "learning_rate": 3.2764018672696293e-07,
+      "loss": 1.6062,
+      "step": 9276
+    },
+    {
+      "epoch": 0.9636439181468786,
+      "grad_norm": 0.45276060700416565,
+      "learning_rate": 3.257779775458769e-07,
+      "loss": 1.7222,
+      "step": 9277
+    },
+    {
+      "epoch": 0.9637477926664589,
+      "grad_norm": 0.4360550045967102,
+      "learning_rate": 3.239210582819563e-07,
+      "loss": 1.7827,
+      "step": 9278
+    },
+    {
+      "epoch": 0.9638516671860393,
+      "grad_norm": 0.46878498792648315,
+      "learning_rate": 3.220694291329596e-07,
+      "loss": 1.7343,
+      "step": 9279
+    },
+    {
+      "epoch": 0.9639555417056196,
+      "grad_norm": 0.4349597096443176,
+      "learning_rate": 3.202230902960679e-07,
+      "loss": 1.3551,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9640594162252,
+      "grad_norm": 0.4269922971725464,
+      "learning_rate": 3.1838204196789625e-07,
+      "loss": 1.5912,
+      "step": 9281
+    },
+    {
+      "epoch": 0.9641632907447804,
+      "grad_norm": 0.42639386653900146,
+      "learning_rate": 3.1654628434450437e-07,
+      "loss": 1.6366,
+      "step": 9282
+    },
+    {
+      "epoch": 0.9642671652643606,
+      "grad_norm": 0.41460925340652466,
+      "learning_rate": 3.147158176213916e-07,
+      "loss": 1.5546,
+      "step": 9283
+    },
+    {
+      "epoch": 0.964371039783941,
+      "grad_norm": 0.4272399842739105,
+      "learning_rate": 3.1289064199347964e-07,
+      "loss": 1.6533,
+      "step": 9284
+    },
+    {
+      "epoch": 0.9644749143035214,
+      "grad_norm": 0.40159735083580017,
+      "learning_rate": 3.110707576551408e-07,
+      "loss": 1.7311,
+      "step": 9285
+    },
+    {
+      "epoch": 0.9645787888231017,
+      "grad_norm": 0.40594831109046936,
+      "learning_rate": 3.092561648001813e-07,
+      "loss": 1.534,
+      "step": 9286
+    },
+    {
+      "epoch": 0.964682663342682,
+      "grad_norm": 0.4097544550895691,
+      "learning_rate": 3.074468636218353e-07,
+      "loss": 1.6231,
+      "step": 9287
+    },
+    {
+      "epoch": 0.9647865378622624,
+      "grad_norm": 0.40390005707740784,
+      "learning_rate": 3.056428543127765e-07,
+      "loss": 1.6062,
+      "step": 9288
+    },
+    {
+      "epoch": 0.9648904123818427,
+      "grad_norm": 0.4110979735851288,
+      "learning_rate": 3.038441370651346e-07,
+      "loss": 1.6621,
+      "step": 9289
+    },
+    {
+      "epoch": 0.9649942869014231,
+      "grad_norm": 0.4974674582481384,
+      "learning_rate": 3.020507120704341e-07,
+      "loss": 1.8125,
+      "step": 9290
+    },
+    {
+      "epoch": 0.9650981614210035,
+      "grad_norm": 0.45189183950424194,
+      "learning_rate": 3.002625795196834e-07,
+      "loss": 1.6262,
+      "step": 9291
+    },
+    {
+      "epoch": 0.9652020359405837,
+      "grad_norm": 0.4480794668197632,
+      "learning_rate": 2.9847973960329124e-07,
+      "loss": 1.6016,
+      "step": 9292
+    },
+    {
+      "epoch": 0.9653059104601641,
+      "grad_norm": 0.4347269535064697,
+      "learning_rate": 2.967021925111224e-07,
+      "loss": 1.9229,
+      "step": 9293
+    },
+    {
+      "epoch": 0.9654097849797445,
+      "grad_norm": 0.4739961326122284,
+      "learning_rate": 2.949299384324644e-07,
+      "loss": 1.5661,
+      "step": 9294
+    },
+    {
+      "epoch": 0.9655136594993248,
+      "grad_norm": 0.39016592502593994,
+      "learning_rate": 2.931629775560607e-07,
+      "loss": 1.542,
+      "step": 9295
+    },
+    {
+      "epoch": 0.9656175340189052,
+      "grad_norm": 0.42430630326271057,
+      "learning_rate": 2.914013100700663e-07,
+      "loss": 1.4694,
+      "step": 9296
+    },
+    {
+      "epoch": 0.9657214085384855,
+      "grad_norm": 0.4428417682647705,
+      "learning_rate": 2.896449361620923e-07,
+      "loss": 1.7179,
+      "step": 9297
+    },
+    {
+      "epoch": 0.9658252830580658,
+      "grad_norm": 0.4339398145675659,
+      "learning_rate": 2.8789385601917797e-07,
+      "loss": 1.6341,
+      "step": 9298
+    },
+    {
+      "epoch": 0.9659291575776462,
+      "grad_norm": 0.4497127830982208,
+      "learning_rate": 2.861480698277963e-07,
+      "loss": 1.8256,
+      "step": 9299
+    },
+    {
+      "epoch": 0.9660330320972266,
+      "grad_norm": 0.47623205184936523,
+      "learning_rate": 2.844075777738597e-07,
+      "loss": 1.8018,
+      "step": 9300
+    },
+    {
+      "epoch": 0.9661369066168068,
+      "grad_norm": 0.4075833261013031,
+      "learning_rate": 2.8267238004272554e-07,
+      "loss": 1.6079,
+      "step": 9301
+    },
+    {
+      "epoch": 0.9662407811363872,
+      "grad_norm": 0.44640010595321655,
+      "learning_rate": 2.8094247681916817e-07,
+      "loss": 1.7133,
+      "step": 9302
+    },
+    {
+      "epoch": 0.9663446556559676,
+      "grad_norm": 0.4296375811100006,
+      "learning_rate": 2.79217868287418e-07,
+      "loss": 1.6657,
+      "step": 9303
+    },
+    {
+      "epoch": 0.966448530175548,
+      "grad_norm": 0.40996262431144714,
+      "learning_rate": 2.7749855463112265e-07,
+      "loss": 1.6412,
+      "step": 9304
+    },
+    {
+      "epoch": 0.9665524046951283,
+      "grad_norm": 0.45529499650001526,
+      "learning_rate": 2.757845360333855e-07,
+      "loss": 1.6247,
+      "step": 9305
+    },
+    {
+      "epoch": 0.9666562792147086,
+      "grad_norm": 0.4100009500980377,
+      "learning_rate": 2.740758126767273e-07,
+      "loss": 1.6921,
+      "step": 9306
+    },
+    {
+      "epoch": 0.966760153734289,
+      "grad_norm": 0.49910518527030945,
+      "learning_rate": 2.723723847431192e-07,
+      "loss": 1.9038,
+      "step": 9307
+    },
+    {
+      "epoch": 0.9668640282538693,
+      "grad_norm": 0.42856353521347046,
+      "learning_rate": 2.7067425241396606e-07,
+      "loss": 1.6905,
+      "step": 9308
+    },
+    {
+      "epoch": 0.9669679027734497,
+      "grad_norm": 0.4085339605808258,
+      "learning_rate": 2.689814158700954e-07,
+      "loss": 1.5855,
+      "step": 9309
+    },
+    {
+      "epoch": 0.9670717772930301,
+      "grad_norm": 0.3960736095905304,
+      "learning_rate": 2.672938752917908e-07,
+      "loss": 1.5288,
+      "step": 9310
+    },
+    {
+      "epoch": 0.9671756518126103,
+      "grad_norm": 0.4568417966365814,
+      "learning_rate": 2.6561163085875863e-07,
+      "loss": 1.6313,
+      "step": 9311
+    },
+    {
+      "epoch": 0.9672795263321907,
+      "grad_norm": 0.39653193950653076,
+      "learning_rate": 2.6393468275014434e-07,
+      "loss": 1.6358,
+      "step": 9312
+    },
+    {
+      "epoch": 0.9673834008517711,
+      "grad_norm": 0.436946302652359,
+      "learning_rate": 2.6226303114452733e-07,
+      "loss": 1.8352,
+      "step": 9313
+    },
+    {
+      "epoch": 0.9674872753713514,
+      "grad_norm": 0.4774184226989746,
+      "learning_rate": 2.605966762199319e-07,
+      "loss": 1.7763,
+      "step": 9314
+    },
+    {
+      "epoch": 0.9675911498909318,
+      "grad_norm": 0.40284937620162964,
+      "learning_rate": 2.58935618153805e-07,
+      "loss": 1.6718,
+      "step": 9315
+    },
+    {
+      "epoch": 0.9676950244105121,
+      "grad_norm": 0.4363865852355957,
+      "learning_rate": 2.5727985712303836e-07,
+      "loss": 1.7835,
+      "step": 9316
+    },
+    {
+      "epoch": 0.9677988989300924,
+      "grad_norm": 0.42425060272216797,
+      "learning_rate": 2.5562939330396327e-07,
+      "loss": 1.7533,
+      "step": 9317
+    },
+    {
+      "epoch": 0.9679027734496728,
+      "grad_norm": 0.4236680567264557,
+      "learning_rate": 2.539842268723336e-07,
+      "loss": 1.5438,
+      "step": 9318
+    },
+    {
+      "epoch": 0.9680066479692532,
+      "grad_norm": 0.41327956318855286,
+      "learning_rate": 2.5234435800335354e-07,
+      "loss": 1.6705,
+      "step": 9319
+    },
+    {
+      "epoch": 0.9681105224888334,
+      "grad_norm": 0.4201132655143738,
+      "learning_rate": 2.507097868716501e-07,
+      "loss": 1.7153,
+      "step": 9320
+    },
+    {
+      "epoch": 0.9682143970084138,
+      "grad_norm": 0.45233723521232605,
+      "learning_rate": 2.4908051365129524e-07,
+      "loss": 1.716,
+      "step": 9321
+    },
+    {
+      "epoch": 0.9683182715279942,
+      "grad_norm": 0.4188010096549988,
+      "learning_rate": 2.4745653851579453e-07,
+      "loss": 1.6264,
+      "step": 9322
+    },
+    {
+      "epoch": 0.9684221460475745,
+      "grad_norm": 0.4535348415374756,
+      "learning_rate": 2.458378616380874e-07,
+      "loss": 1.8135,
+      "step": 9323
+    },
+    {
+      "epoch": 0.9685260205671549,
+      "grad_norm": 0.5156634449958801,
+      "learning_rate": 2.442244831905471e-07,
+      "loss": 1.8665,
+      "step": 9324
+    },
+    {
+      "epoch": 0.9686298950867352,
+      "grad_norm": 0.41018423438072205,
+      "learning_rate": 2.426164033449918e-07,
+      "loss": 1.6818,
+      "step": 9325
+    },
+    {
+      "epoch": 0.9687337696063155,
+      "grad_norm": 0.4227202534675598,
+      "learning_rate": 2.4101362227266776e-07,
+      "loss": 1.5506,
+      "step": 9326
+    },
+    {
+      "epoch": 0.9688376441258959,
+      "grad_norm": 0.4453946352005005,
+      "learning_rate": 2.394161401442552e-07,
+      "loss": 1.8326,
+      "step": 9327
+    },
+    {
+      "epoch": 0.9689415186454763,
+      "grad_norm": 0.43642252683639526,
+      "learning_rate": 2.3782395712987903e-07,
+      "loss": 1.8348,
+      "step": 9328
+    },
+    {
+      "epoch": 0.9690453931650566,
+      "grad_norm": 0.4805013835430145,
+      "learning_rate": 2.3623707339909262e-07,
+      "loss": 1.7733,
+      "step": 9329
+    },
+    {
+      "epoch": 0.9691492676846369,
+      "grad_norm": 0.41811105608940125,
+      "learning_rate": 2.3465548912088296e-07,
+      "loss": 1.5234,
+      "step": 9330
+    },
+    {
+      "epoch": 0.9692531422042173,
+      "grad_norm": 0.39481353759765625,
+      "learning_rate": 2.3307920446368203e-07,
+      "loss": 1.3639,
+      "step": 9331
+    },
+    {
+      "epoch": 0.9693570167237977,
+      "grad_norm": 0.43810319900512695,
+      "learning_rate": 2.3150821959534442e-07,
+      "loss": 1.6738,
+      "step": 9332
+    },
+    {
+      "epoch": 0.969460891243378,
+      "grad_norm": 0.3864077627658844,
+      "learning_rate": 2.299425346831807e-07,
+      "loss": 1.5251,
+      "step": 9333
+    },
+    {
+      "epoch": 0.9695647657629584,
+      "grad_norm": 0.42563048005104065,
+      "learning_rate": 2.2838214989390762e-07,
+      "loss": 1.6638,
+      "step": 9334
+    },
+    {
+      "epoch": 0.9696686402825387,
+      "grad_norm": 0.4011988341808319,
+      "learning_rate": 2.2682706539370324e-07,
+      "loss": 1.4951,
+      "step": 9335
+    },
+    {
+      "epoch": 0.969772514802119,
+      "grad_norm": 0.4599732458591461,
+      "learning_rate": 2.2527728134817406e-07,
+      "loss": 1.7103,
+      "step": 9336
+    },
+    {
+      "epoch": 0.9698763893216994,
+      "grad_norm": 0.3994980752468109,
+      "learning_rate": 2.2373279792235468e-07,
+      "loss": 1.6718,
+      "step": 9337
+    },
+    {
+      "epoch": 0.9699802638412798,
+      "grad_norm": 0.42097207903862,
+      "learning_rate": 2.2219361528071912e-07,
+      "loss": 1.6202,
+      "step": 9338
+    },
+    {
+      "epoch": 0.97008413836086,
+      "grad_norm": 0.4318835139274597,
+      "learning_rate": 2.2065973358718627e-07,
+      "loss": 1.5318,
+      "step": 9339
+    },
+    {
+      "epoch": 0.9701880128804404,
+      "grad_norm": 0.4295785129070282,
+      "learning_rate": 2.191311530050977e-07,
+      "loss": 1.6927,
+      "step": 9340
+    },
+    {
+      "epoch": 0.9702918874000208,
+      "grad_norm": 0.4130423367023468,
+      "learning_rate": 2.1760787369723423e-07,
+      "loss": 1.6946,
+      "step": 9341
+    },
+    {
+      "epoch": 0.9703957619196011,
+      "grad_norm": 0.43795621395111084,
+      "learning_rate": 2.160898958258162e-07,
+      "loss": 1.4909,
+      "step": 9342
+    },
+    {
+      "epoch": 0.9704996364391815,
+      "grad_norm": 0.42696115374565125,
+      "learning_rate": 2.14577219552492e-07,
+      "loss": 1.5696,
+      "step": 9343
+    },
+    {
+      "epoch": 0.9706035109587619,
+      "grad_norm": 0.4208974242210388,
+      "learning_rate": 2.1306984503835504e-07,
+      "loss": 1.5663,
+      "step": 9344
+    },
+    {
+      "epoch": 0.9707073854783421,
+      "grad_norm": 0.4519440233707428,
+      "learning_rate": 2.115677724439269e-07,
+      "loss": 1.7372,
+      "step": 9345
+    },
+    {
+      "epoch": 0.9708112599979225,
+      "grad_norm": 0.4242643415927887,
+      "learning_rate": 2.1007100192916296e-07,
+      "loss": 1.6569,
+      "step": 9346
+    },
+    {
+      "epoch": 0.9709151345175029,
+      "grad_norm": 0.41533952951431274,
+      "learning_rate": 2.085795336534635e-07,
+      "loss": 1.5835,
+      "step": 9347
+    },
+    {
+      "epoch": 0.9710190090370832,
+      "grad_norm": 0.4785119593143463,
+      "learning_rate": 2.0709336777565702e-07,
+      "loss": 1.7689,
+      "step": 9348
+    },
+    {
+      "epoch": 0.9711228835566635,
+      "grad_norm": 0.41797420382499695,
+      "learning_rate": 2.0561250445400583e-07,
+      "loss": 1.6882,
+      "step": 9349
+    },
+    {
+      "epoch": 0.9712267580762439,
+      "grad_norm": 0.4142938256263733,
+      "learning_rate": 2.041369438462115e-07,
+      "loss": 1.7222,
+      "step": 9350
+    },
+    {
+      "epoch": 0.9713306325958242,
+      "grad_norm": 0.3736826181411743,
+      "learning_rate": 2.0266668610940952e-07,
+      "loss": 1.4399,
+      "step": 9351
+    },
+    {
+      "epoch": 0.9714345071154046,
+      "grad_norm": 0.4269476532936096,
+      "learning_rate": 2.012017314001746e-07,
+      "loss": 1.4723,
+      "step": 9352
+    },
+    {
+      "epoch": 0.971538381634985,
+      "grad_norm": 0.43131187558174133,
+      "learning_rate": 1.9974207987450422e-07,
+      "loss": 1.739,
+      "step": 9353
+    },
+    {
+      "epoch": 0.9716422561545652,
+      "grad_norm": 0.4340756833553314,
+      "learning_rate": 1.9828773168785175e-07,
+      "loss": 1.7562,
+      "step": 9354
+    },
+    {
+      "epoch": 0.9717461306741456,
+      "grad_norm": 0.46568766236305237,
+      "learning_rate": 1.9683868699508224e-07,
+      "loss": 1.6564,
+      "step": 9355
+    },
+    {
+      "epoch": 0.971850005193726,
+      "grad_norm": 0.438309907913208,
+      "learning_rate": 1.9539494595051112e-07,
+      "loss": 1.6676,
+      "step": 9356
+    },
+    {
+      "epoch": 0.9719538797133064,
+      "grad_norm": 0.40006324648857117,
+      "learning_rate": 1.9395650870789316e-07,
+      "loss": 1.6582,
+      "step": 9357
+    },
+    {
+      "epoch": 0.9720577542328867,
+      "grad_norm": 0.4404867887496948,
+      "learning_rate": 1.9252337542040034e-07,
+      "loss": 1.7257,
+      "step": 9358
+    },
+    {
+      "epoch": 0.972161628752467,
+      "grad_norm": 0.4128919541835785,
+      "learning_rate": 1.910955462406605e-07,
+      "loss": 1.6339,
+      "step": 9359
+    },
+    {
+      "epoch": 0.9722655032720474,
+      "grad_norm": 0.47197067737579346,
+      "learning_rate": 1.8967302132071317e-07,
+      "loss": 1.5516,
+      "step": 9360
+    },
+    {
+      "epoch": 0.9723693777916277,
+      "grad_norm": 0.46007785201072693,
+      "learning_rate": 1.882558008120594e-07,
+      "loss": 1.6057,
+      "step": 9361
+    },
+    {
+      "epoch": 0.9724732523112081,
+      "grad_norm": 0.4223778247833252,
+      "learning_rate": 1.8684388486561733e-07,
+      "loss": 1.6723,
+      "step": 9362
+    },
+    {
+      "epoch": 0.9725771268307885,
+      "grad_norm": 0.40712058544158936,
+      "learning_rate": 1.854372736317389e-07,
+      "loss": 1.5695,
+      "step": 9363
+    },
+    {
+      "epoch": 0.9726810013503687,
+      "grad_norm": 0.4485786557197571,
+      "learning_rate": 1.8403596726022653e-07,
+      "loss": 1.4115,
+      "step": 9364
+    },
+    {
+      "epoch": 0.9727848758699491,
+      "grad_norm": 0.44639256596565247,
+      "learning_rate": 1.8263996590029976e-07,
+      "loss": 1.6466,
+      "step": 9365
+    },
+    {
+      "epoch": 0.9728887503895295,
+      "grad_norm": 0.4222765862941742,
+      "learning_rate": 1.812492697006285e-07,
+      "loss": 1.5191,
+      "step": 9366
+    },
+    {
+      "epoch": 0.9729926249091098,
+      "grad_norm": 0.4526502192020416,
+      "learning_rate": 1.79863878809311e-07,
+      "loss": 1.8303,
+      "step": 9367
+    },
+    {
+      "epoch": 0.9730964994286901,
+      "grad_norm": 0.3991301655769348,
+      "learning_rate": 1.7848379337387922e-07,
+      "loss": 1.6194,
+      "step": 9368
+    },
+    {
+      "epoch": 0.9732003739482705,
+      "grad_norm": 0.43972858786582947,
+      "learning_rate": 1.7710901354129894e-07,
+      "loss": 1.6873,
+      "step": 9369
+    },
+    {
+      "epoch": 0.9733042484678508,
+      "grad_norm": 0.42031508684158325,
+      "learning_rate": 1.7573953945796973e-07,
+      "loss": 1.7127,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9734081229874312,
+      "grad_norm": 0.40002334117889404,
+      "learning_rate": 1.7437537126974158e-07,
+      "loss": 1.4936,
+      "step": 9371
+    },
+    {
+      "epoch": 0.9735119975070116,
+      "grad_norm": 0.4212174713611603,
+      "learning_rate": 1.7301650912188162e-07,
+      "loss": 1.7144,
+      "step": 9372
+    },
+    {
+      "epoch": 0.9736158720265918,
+      "grad_norm": 0.43451911211013794,
+      "learning_rate": 1.716629531590963e-07,
+      "loss": 1.7131,
+      "step": 9373
+    },
+    {
+      "epoch": 0.9737197465461722,
+      "grad_norm": 0.40921545028686523,
+      "learning_rate": 1.7031470352553146e-07,
+      "loss": 1.4595,
+      "step": 9374
+    },
+    {
+      "epoch": 0.9738236210657526,
+      "grad_norm": 0.41354814171791077,
+      "learning_rate": 1.6897176036476115e-07,
+      "loss": 1.6428,
+      "step": 9375
+    },
+    {
+      "epoch": 0.9739274955853329,
+      "grad_norm": 0.4084813892841339,
+      "learning_rate": 1.6763412381980425e-07,
+      "loss": 1.6586,
+      "step": 9376
+    },
+    {
+      "epoch": 0.9740313701049133,
+      "grad_norm": 0.5016772747039795,
+      "learning_rate": 1.6630179403310797e-07,
+      "loss": 1.7705,
+      "step": 9377
+    },
+    {
+      "epoch": 0.9741352446244936,
+      "grad_norm": 0.4082990288734436,
+      "learning_rate": 1.6497477114654768e-07,
+      "loss": 1.604,
+      "step": 9378
+    },
+    {
+      "epoch": 0.9742391191440739,
+      "grad_norm": 0.4223262369632721,
+      "learning_rate": 1.6365305530144926e-07,
+      "loss": 1.5545,
+      "step": 9379
+    },
+    {
+      "epoch": 0.9743429936636543,
+      "grad_norm": 0.45764049887657166,
+      "learning_rate": 1.623366466385612e-07,
+      "loss": 1.6171,
+      "step": 9380
+    },
+    {
+      "epoch": 0.9744468681832347,
+      "grad_norm": 0.4205342233181,
+      "learning_rate": 1.6102554529807134e-07,
+      "loss": 1.7076,
+      "step": 9381
+    },
+    {
+      "epoch": 0.974550742702815,
+      "grad_norm": 0.4099922478199005,
+      "learning_rate": 1.5971975141960138e-07,
+      "loss": 1.6506,
+      "step": 9382
+    },
+    {
+      "epoch": 0.9746546172223953,
+      "grad_norm": 0.43783533573150635,
+      "learning_rate": 1.5841926514220674e-07,
+      "loss": 1.526,
+      "step": 9383
+    },
+    {
+      "epoch": 0.9747584917419757,
+      "grad_norm": 0.4745118021965027,
+      "learning_rate": 1.571240866043877e-07,
+      "loss": 1.7779,
+      "step": 9384
+    },
+    {
+      "epoch": 0.9748623662615561,
+      "grad_norm": 0.4208376109600067,
+      "learning_rate": 1.5583421594405623e-07,
+      "loss": 1.7611,
+      "step": 9385
+    },
+    {
+      "epoch": 0.9749662407811364,
+      "grad_norm": 0.41122379899024963,
+      "learning_rate": 1.5454965329858573e-07,
+      "loss": 1.5896,
+      "step": 9386
+    },
+    {
+      "epoch": 0.9750701153007167,
+      "grad_norm": 0.3928976058959961,
+      "learning_rate": 1.5327039880477234e-07,
+      "loss": 1.59,
+      "step": 9387
+    },
+    {
+      "epoch": 0.9751739898202971,
+      "grad_norm": 0.4342755675315857,
+      "learning_rate": 1.5199645259883488e-07,
+      "loss": 1.6747,
+      "step": 9388
+    },
+    {
+      "epoch": 0.9752778643398774,
+      "grad_norm": 0.39169561862945557,
+      "learning_rate": 1.5072781481644816e-07,
+      "loss": 1.5841,
+      "step": 9389
+    },
+    {
+      "epoch": 0.9753817388594578,
+      "grad_norm": 0.43431976437568665,
+      "learning_rate": 1.4946448559270964e-07,
+      "loss": 1.661,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9754856133790382,
+      "grad_norm": 0.4187610149383545,
+      "learning_rate": 1.4820646506215064e-07,
+      "loss": 1.5102,
+      "step": 9391
+    },
+    {
+      "epoch": 0.9755894878986184,
+      "grad_norm": 0.415600448846817,
+      "learning_rate": 1.4695375335875282e-07,
+      "loss": 1.4286,
+      "step": 9392
+    },
+    {
+      "epoch": 0.9756933624181988,
+      "grad_norm": 0.4058886468410492,
+      "learning_rate": 1.45706350615904e-07,
+      "loss": 1.6924,
+      "step": 9393
+    },
+    {
+      "epoch": 0.9757972369377792,
+      "grad_norm": 0.41150805354118347,
+      "learning_rate": 1.444642569664534e-07,
+      "loss": 1.5943,
+      "step": 9394
+    },
+    {
+      "epoch": 0.9759011114573595,
+      "grad_norm": 0.39166486263275146,
+      "learning_rate": 1.43227472542673e-07,
+      "loss": 1.628,
+      "step": 9395
+    },
+    {
+      "epoch": 0.9760049859769399,
+      "grad_norm": 0.41981199383735657,
+      "learning_rate": 1.4199599747626303e-07,
+      "loss": 1.4475,
+      "step": 9396
+    },
+    {
+      "epoch": 0.9761088604965202,
+      "grad_norm": 0.4433886408805847,
+      "learning_rate": 1.4076983189837967e-07,
+      "loss": 1.6782,
+      "step": 9397
+    },
+    {
+      "epoch": 0.9762127350161005,
+      "grad_norm": 0.4331313967704773,
+      "learning_rate": 1.3954897593958516e-07,
+      "loss": 1.5824,
+      "step": 9398
+    },
+    {
+      "epoch": 0.9763166095356809,
+      "grad_norm": 0.4032611846923828,
+      "learning_rate": 1.3833342972990327e-07,
+      "loss": 1.5385,
+      "step": 9399
+    },
+    {
+      "epoch": 0.9764204840552613,
+      "grad_norm": 0.4217138886451721,
+      "learning_rate": 1.371231933987749e-07,
+      "loss": 1.7366,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9765243585748415,
+      "grad_norm": 0.3987552225589752,
+      "learning_rate": 1.3591826707508026e-07,
+      "loss": 1.5265,
+      "step": 9401
+    },
+    {
+      "epoch": 0.9766282330944219,
+      "grad_norm": 0.4729466140270233,
+      "learning_rate": 1.3471865088713342e-07,
+      "loss": 1.7914,
+      "step": 9402
+    },
+    {
+      "epoch": 0.9767321076140023,
+      "grad_norm": 0.4094342589378357,
+      "learning_rate": 1.3352434496268773e-07,
+      "loss": 1.6576,
+      "step": 9403
+    },
+    {
+      "epoch": 0.9768359821335826,
+      "grad_norm": 0.46844589710235596,
+      "learning_rate": 1.3233534942892477e-07,
+      "loss": 1.7209,
+      "step": 9404
+    },
+    {
+      "epoch": 0.976939856653163,
+      "grad_norm": 0.4402284026145935,
+      "learning_rate": 1.311516644124655e-07,
+      "loss": 1.6706,
+      "step": 9405
+    },
+    {
+      "epoch": 0.9770437311727433,
+      "grad_norm": 0.4201194941997528,
+      "learning_rate": 1.299732900393591e-07,
+      "loss": 1.598,
+      "step": 9406
+    },
+    {
+      "epoch": 0.9771476056923236,
+      "grad_norm": 0.41740289330482483,
+      "learning_rate": 1.2880022643509958e-07,
+      "loss": 1.6247,
+      "step": 9407
+    },
+    {
+      "epoch": 0.977251480211904,
+      "grad_norm": 0.4268152117729187,
+      "learning_rate": 1.2763247372460374e-07,
+      "loss": 1.6006,
+      "step": 9408
+    },
+    {
+      "epoch": 0.9773553547314844,
+      "grad_norm": 0.43314602971076965,
+      "learning_rate": 1.2647003203223318e-07,
+      "loss": 1.6009,
+      "step": 9409
+    },
+    {
+      "epoch": 0.9774592292510648,
+      "grad_norm": 0.42732593417167664,
+      "learning_rate": 1.2531290148177221e-07,
+      "loss": 1.6934,
+      "step": 9410
+    },
+    {
+      "epoch": 0.977563103770645,
+      "grad_norm": 0.4502175748348236,
+      "learning_rate": 1.2416108219645005e-07,
+      "loss": 1.7892,
+      "step": 9411
+    },
+    {
+      "epoch": 0.9776669782902254,
+      "grad_norm": 0.45358970761299133,
+      "learning_rate": 1.2301457429892972e-07,
+      "loss": 1.5935,
+      "step": 9412
+    },
+    {
+      "epoch": 0.9777708528098058,
+      "grad_norm": 0.43146711587905884,
+      "learning_rate": 1.2187337791129682e-07,
+      "loss": 1.457,
+      "step": 9413
+    },
+    {
+      "epoch": 0.9778747273293861,
+      "grad_norm": 0.44954973459243774,
+      "learning_rate": 1.2073749315508754e-07,
+      "loss": 1.6821,
+      "step": 9414
+    },
+    {
+      "epoch": 0.9779786018489665,
+      "grad_norm": 0.4219938814640045,
+      "learning_rate": 1.1960692015126062e-07,
+      "loss": 1.7259,
+      "step": 9415
+    },
+    {
+      "epoch": 0.9780824763685468,
+      "grad_norm": 0.41503259539604187,
+      "learning_rate": 1.1848165902021424e-07,
+      "loss": 1.4451,
+      "step": 9416
+    },
+    {
+      "epoch": 0.9781863508881271,
+      "grad_norm": 0.4230490028858185,
+      "learning_rate": 1.1736170988178585e-07,
+      "loss": 1.6452,
+      "step": 9417
+    },
+    {
+      "epoch": 0.9782902254077075,
+      "grad_norm": 0.44973745942115784,
+      "learning_rate": 1.1624707285523006e-07,
+      "loss": 1.811,
+      "step": 9418
+    },
+    {
+      "epoch": 0.9783940999272879,
+      "grad_norm": 0.4050397574901581,
+      "learning_rate": 1.1513774805925748e-07,
+      "loss": 1.4962,
+      "step": 9419
+    },
+    {
+      "epoch": 0.9784979744468681,
+      "grad_norm": 0.3766053020954132,
+      "learning_rate": 1.1403373561199582e-07,
+      "loss": 1.4053,
+      "step": 9420
+    },
+    {
+      "epoch": 0.9786018489664485,
+      "grad_norm": 0.41133883595466614,
+      "learning_rate": 1.1293503563101216e-07,
+      "loss": 1.6545,
+      "step": 9421
+    },
+    {
+      "epoch": 0.9787057234860289,
+      "grad_norm": 0.44046151638031006,
+      "learning_rate": 1.1184164823331844e-07,
+      "loss": 1.6459,
+      "step": 9422
+    },
+    {
+      "epoch": 0.9788095980056092,
+      "grad_norm": 0.44645535945892334,
+      "learning_rate": 1.1075357353534377e-07,
+      "loss": 1.5921,
+      "step": 9423
+    },
+    {
+      "epoch": 0.9789134725251896,
+      "grad_norm": 0.44139453768730164,
+      "learning_rate": 1.0967081165296211e-07,
+      "loss": 1.8106,
+      "step": 9424
+    },
+    {
+      "epoch": 0.97901734704477,
+      "grad_norm": 0.39633703231811523,
+      "learning_rate": 1.0859336270148124e-07,
+      "loss": 1.6446,
+      "step": 9425
+    },
+    {
+      "epoch": 0.9791212215643502,
+      "grad_norm": 0.4287833273410797,
+      "learning_rate": 1.0752122679563714e-07,
+      "loss": 1.6376,
+      "step": 9426
+    },
+    {
+      "epoch": 0.9792250960839306,
+      "grad_norm": 0.42582541704177856,
+      "learning_rate": 1.0645440404960517e-07,
+      "loss": 1.6317,
+      "step": 9427
+    },
+    {
+      "epoch": 0.979328970603511,
+      "grad_norm": 0.47708913683891296,
+      "learning_rate": 1.0539289457699442e-07,
+      "loss": 1.918,
+      "step": 9428
+    },
+    {
+      "epoch": 0.9794328451230913,
+      "grad_norm": 0.38468772172927856,
+      "learning_rate": 1.0433669849085336e-07,
+      "loss": 1.5551,
+      "step": 9429
+    },
+    {
+      "epoch": 0.9795367196426716,
+      "grad_norm": 0.4491479694843292,
+      "learning_rate": 1.032858159036476e-07,
+      "loss": 1.7375,
+      "step": 9430
+    },
+    {
+      "epoch": 0.979640594162252,
+      "grad_norm": 0.43696144223213196,
+      "learning_rate": 1.0224024692729317e-07,
+      "loss": 1.5334,
+      "step": 9431
+    },
+    {
+      "epoch": 0.9797444686818323,
+      "grad_norm": 0.38262155652046204,
+      "learning_rate": 1.0119999167313988e-07,
+      "loss": 1.4757,
+      "step": 9432
+    },
+    {
+      "epoch": 0.9798483432014127,
+      "grad_norm": 0.43123918771743774,
+      "learning_rate": 1.0016505025195467e-07,
+      "loss": 1.5889,
+      "step": 9433
+    },
+    {
+      "epoch": 0.9799522177209931,
+      "grad_norm": 0.44049620628356934,
+      "learning_rate": 9.913542277396604e-08,
+      "loss": 1.654,
+      "step": 9434
+    },
+    {
+      "epoch": 0.9800560922405734,
+      "grad_norm": 0.3811212182044983,
+      "learning_rate": 9.811110934880296e-08,
+      "loss": 1.4344,
+      "step": 9435
+    },
+    {
+      "epoch": 0.9801599667601537,
+      "grad_norm": 0.4550376534461975,
+      "learning_rate": 9.709211008556707e-08,
+      "loss": 1.8921,
+      "step": 9436
+    },
+    {
+      "epoch": 0.9802638412797341,
+      "grad_norm": 0.41628962755203247,
+      "learning_rate": 9.607842509275489e-08,
+      "loss": 1.6439,
+      "step": 9437
+    },
+    {
+      "epoch": 0.9803677157993145,
+      "grad_norm": 0.4357268810272217,
+      "learning_rate": 9.507005447833007e-08,
+      "loss": 1.8815,
+      "step": 9438
+    },
+    {
+      "epoch": 0.9804715903188947,
+      "grad_norm": 0.38655275106430054,
+      "learning_rate": 9.406699834967337e-08,
+      "loss": 1.4922,
+      "step": 9439
+    },
+    {
+      "epoch": 0.9805754648384751,
+      "grad_norm": 0.3913620710372925,
+      "learning_rate": 9.306925681359935e-08,
+      "loss": 1.6177,
+      "step": 9440
+    },
+    {
+      "epoch": 0.9806793393580555,
+      "grad_norm": 0.44508156180381775,
+      "learning_rate": 9.207682997635636e-08,
+      "loss": 1.6836,
+      "step": 9441
+    },
+    {
+      "epoch": 0.9807832138776358,
+      "grad_norm": 0.41767412424087524,
+      "learning_rate": 9.108971794363763e-08,
+      "loss": 1.6061,
+      "step": 9442
+    },
+    {
+      "epoch": 0.9808870883972162,
+      "grad_norm": 0.40216001868247986,
+      "learning_rate": 9.010792082055908e-08,
+      "loss": 1.5034,
+      "step": 9443
+    },
+    {
+      "epoch": 0.9809909629167965,
+      "grad_norm": 0.39269882440567017,
+      "learning_rate": 8.91314387116704e-08,
+      "loss": 1.5912,
+      "step": 9444
+    },
+    {
+      "epoch": 0.9810948374363768,
+      "grad_norm": 0.41241204738616943,
+      "learning_rate": 8.81602717209662e-08,
+      "loss": 1.6321,
+      "step": 9445
+    },
+    {
+      "epoch": 0.9811987119559572,
+      "grad_norm": 0.4276941120624542,
+      "learning_rate": 8.719441995186928e-08,
+      "loss": 1.6572,
+      "step": 9446
+    },
+    {
+      "epoch": 0.9813025864755376,
+      "grad_norm": 0.5014160871505737,
+      "learning_rate": 8.623388350722517e-08,
+      "loss": 1.6643,
+      "step": 9447
+    },
+    {
+      "epoch": 0.9814064609951179,
+      "grad_norm": 0.44852280616760254,
+      "learning_rate": 8.527866248933536e-08,
+      "loss": 1.4247,
+      "step": 9448
+    },
+    {
+      "epoch": 0.9815103355146982,
+      "grad_norm": 0.4082983136177063,
+      "learning_rate": 8.432875699991849e-08,
+      "loss": 1.4757,
+      "step": 9449
+    },
+    {
+      "epoch": 0.9816142100342786,
+      "grad_norm": 0.458987295627594,
+      "learning_rate": 8.338416714013253e-08,
+      "loss": 1.6738,
+      "step": 9450
+    },
+    {
+      "epoch": 0.9817180845538589,
+      "grad_norm": 0.42182016372680664,
+      "learning_rate": 8.244489301056368e-08,
+      "loss": 1.5149,
+      "step": 9451
+    },
+    {
+      "epoch": 0.9818219590734393,
+      "grad_norm": 0.4483654201030731,
+      "learning_rate": 8.151093471124305e-08,
+      "loss": 1.6678,
+      "step": 9452
+    },
+    {
+      "epoch": 0.9819258335930197,
+      "grad_norm": 0.4117814302444458,
+      "learning_rate": 8.058229234162995e-08,
+      "loss": 1.5014,
+      "step": 9453
+    },
+    {
+      "epoch": 0.9820297081125999,
+      "grad_norm": 0.43942439556121826,
+      "learning_rate": 7.965896600061751e-08,
+      "loss": 1.6016,
+      "step": 9454
+    },
+    {
+      "epoch": 0.9821335826321803,
+      "grad_norm": 0.4229027032852173,
+      "learning_rate": 7.874095578653263e-08,
+      "loss": 1.6684,
+      "step": 9455
+    },
+    {
+      "epoch": 0.9822374571517607,
+      "grad_norm": 0.38582631945610046,
+      "learning_rate": 7.782826179713043e-08,
+      "loss": 1.5004,
+      "step": 9456
+    },
+    {
+      "epoch": 0.982341331671341,
+      "grad_norm": 0.5087673664093018,
+      "learning_rate": 7.692088412961096e-08,
+      "loss": 1.8254,
+      "step": 9457
+    },
+    {
+      "epoch": 0.9824452061909213,
+      "grad_norm": 0.4083236753940582,
+      "learning_rate": 7.601882288060247e-08,
+      "loss": 1.577,
+      "step": 9458
+    },
+    {
+      "epoch": 0.9825490807105017,
+      "grad_norm": 0.42479801177978516,
+      "learning_rate": 7.512207814617256e-08,
+      "loss": 1.6106,
+      "step": 9459
+    },
+    {
+      "epoch": 0.9826529552300821,
+      "grad_norm": 0.4226894676685333,
+      "learning_rate": 7.423065002180596e-08,
+      "loss": 1.6185,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9827568297496624,
+      "grad_norm": 0.42085468769073486,
+      "learning_rate": 7.334453860244339e-08,
+      "loss": 1.5966,
+      "step": 9461
+    },
+    {
+      "epoch": 0.9828607042692428,
+      "grad_norm": 0.4113049805164337,
+      "learning_rate": 7.24637439824427e-08,
+      "loss": 1.5905,
+      "step": 9462
+    },
+    {
+      "epoch": 0.9829645787888232,
+      "grad_norm": 0.406465083360672,
+      "learning_rate": 7.158826625560112e-08,
+      "loss": 1.6842,
+      "step": 9463
+    },
+    {
+      "epoch": 0.9830684533084034,
+      "grad_norm": 0.4487413167953491,
+      "learning_rate": 7.071810551516067e-08,
+      "loss": 1.6072,
+      "step": 9464
+    },
+    {
+      "epoch": 0.9831723278279838,
+      "grad_norm": 0.4292948544025421,
+      "learning_rate": 6.98532618537695e-08,
+      "loss": 1.6148,
+      "step": 9465
+    },
+    {
+      "epoch": 0.9832762023475642,
+      "grad_norm": 0.4271443784236908,
+      "learning_rate": 6.899373536354282e-08,
+      "loss": 1.693,
+      "step": 9466
+    },
+    {
+      "epoch": 0.9833800768671445,
+      "grad_norm": 0.4509519636631012,
+      "learning_rate": 6.813952613600183e-08,
+      "loss": 1.896,
+      "step": 9467
+    },
+    {
+      "epoch": 0.9834839513867248,
+      "grad_norm": 0.4309237599372864,
+      "learning_rate": 6.729063426211824e-08,
+      "loss": 1.6331,
+      "step": 9468
+    },
+    {
+      "epoch": 0.9835878259063052,
+      "grad_norm": 0.42812246084213257,
+      "learning_rate": 6.644705983229194e-08,
+      "loss": 1.6337,
+      "step": 9469
+    },
+    {
+      "epoch": 0.9836917004258855,
+      "grad_norm": 0.43282395601272583,
+      "learning_rate": 6.560880293636218e-08,
+      "loss": 1.7527,
+      "step": 9470
+    },
+    {
+      "epoch": 0.9837955749454659,
+      "grad_norm": 0.40227630734443665,
+      "learning_rate": 6.477586366358535e-08,
+      "loss": 1.6945,
+      "step": 9471
+    },
+    {
+      "epoch": 0.9838994494650463,
+      "grad_norm": 0.4157799780368805,
+      "learning_rate": 6.394824210266826e-08,
+      "loss": 1.4567,
+      "step": 9472
+    },
+    {
+      "epoch": 0.9840033239846265,
+      "grad_norm": 0.39977502822875977,
+      "learning_rate": 6.312593834175152e-08,
+      "loss": 1.5229,
+      "step": 9473
+    },
+    {
+      "epoch": 0.9841071985042069,
+      "grad_norm": 0.42183637619018555,
+      "learning_rate": 6.230895246840396e-08,
+      "loss": 1.5506,
+      "step": 9474
+    },
+    {
+      "epoch": 0.9842110730237873,
+      "grad_norm": 0.4323163628578186,
+      "learning_rate": 6.149728456961712e-08,
+      "loss": 1.6544,
+      "step": 9475
+    },
+    {
+      "epoch": 0.9843149475433676,
+      "grad_norm": 0.5097078680992126,
+      "learning_rate": 6.069093473183296e-08,
+      "loss": 1.7364,
+      "step": 9476
+    },
+    {
+      "epoch": 0.984418822062948,
+      "grad_norm": 0.4150666296482086,
+      "learning_rate": 5.988990304092724e-08,
+      "loss": 1.5175,
+      "step": 9477
+    },
+    {
+      "epoch": 0.9845226965825283,
+      "grad_norm": 0.4314131736755371,
+      "learning_rate": 5.9094189582198366e-08,
+      "loss": 1.6242,
+      "step": 9478
+    },
+    {
+      "epoch": 0.9846265711021086,
+      "grad_norm": 0.4086059033870697,
+      "learning_rate": 5.830379444038414e-08,
+      "loss": 1.6943,
+      "step": 9479
+    },
+    {
+      "epoch": 0.984730445621689,
+      "grad_norm": 0.4106563329696655,
+      "learning_rate": 5.751871769965056e-08,
+      "loss": 1.5377,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9848343201412694,
+      "grad_norm": 0.4505181610584259,
+      "learning_rate": 5.673895944361407e-08,
+      "loss": 1.7265,
+      "step": 9481
+    },
+    {
+      "epoch": 0.9849381946608496,
+      "grad_norm": 0.41251927614212036,
+      "learning_rate": 5.59645197553027e-08,
+      "loss": 1.6865,
+      "step": 9482
+    },
+    {
+      "epoch": 0.98504206918043,
+      "grad_norm": 0.42645978927612305,
+      "learning_rate": 5.5195398717194926e-08,
+      "loss": 1.8202,
+      "step": 9483
+    },
+    {
+      "epoch": 0.9851459437000104,
+      "grad_norm": 0.4157099723815918,
+      "learning_rate": 5.443159641118634e-08,
+      "loss": 1.6619,
+      "step": 9484
+    },
+    {
+      "epoch": 0.9852498182195908,
+      "grad_norm": 0.42381036281585693,
+      "learning_rate": 5.367311291862853e-08,
+      "loss": 1.7641,
+      "step": 9485
+    },
+    {
+      "epoch": 0.9853536927391711,
+      "grad_norm": 0.4490167796611786,
+      "learning_rate": 5.291994832028469e-08,
+      "loss": 1.5296,
+      "step": 9486
+    },
+    {
+      "epoch": 0.9854575672587514,
+      "grad_norm": 0.4082736074924469,
+      "learning_rate": 5.217210269636286e-08,
+      "loss": 1.4804,
+      "step": 9487
+    },
+    {
+      "epoch": 0.9855614417783318,
+      "grad_norm": 0.41350072622299194,
+      "learning_rate": 5.1429576126504895e-08,
+      "loss": 1.5122,
+      "step": 9488
+    },
+    {
+      "epoch": 0.9856653162979121,
+      "grad_norm": 0.4074830412864685,
+      "learning_rate": 5.069236868978644e-08,
+      "loss": 1.5971,
+      "step": 9489
+    },
+    {
+      "epoch": 0.9857691908174925,
+      "grad_norm": 0.4387468099594116,
+      "learning_rate": 4.996048046471136e-08,
+      "loss": 1.6514,
+      "step": 9490
+    },
+    {
+      "epoch": 0.9858730653370729,
+      "grad_norm": 0.4530046582221985,
+      "learning_rate": 4.923391152921175e-08,
+      "loss": 1.6722,
+      "step": 9491
+    },
+    {
+      "epoch": 0.9859769398566531,
+      "grad_norm": 0.4661511778831482,
+      "learning_rate": 4.851266196068127e-08,
+      "loss": 1.6484,
+      "step": 9492
+    },
+    {
+      "epoch": 0.9860808143762335,
+      "grad_norm": 0.42836901545524597,
+      "learning_rate": 4.779673183590849e-08,
+      "loss": 1.8121,
+      "step": 9493
+    },
+    {
+      "epoch": 0.9861846888958139,
+      "grad_norm": 0.538507878780365,
+      "learning_rate": 4.708612123114353e-08,
+      "loss": 1.7639,
+      "step": 9494
+    },
+    {
+      "epoch": 0.9862885634153942,
+      "grad_norm": 0.4897894561290741,
+      "learning_rate": 4.638083022205364e-08,
+      "loss": 1.7797,
+      "step": 9495
+    },
+    {
+      "epoch": 0.9863924379349746,
+      "grad_norm": 0.4547121524810791,
+      "learning_rate": 4.56808588837565e-08,
+      "loss": 1.6323,
+      "step": 9496
+    },
+    {
+      "epoch": 0.9864963124545549,
+      "grad_norm": 0.4277886748313904,
+      "learning_rate": 4.4986207290792505e-08,
+      "loss": 1.7339,
+      "step": 9497
+    },
+    {
+      "epoch": 0.9866001869741352,
+      "grad_norm": 0.4299750328063965,
+      "learning_rate": 4.4296875517130243e-08,
+      "loss": 1.6974,
+      "step": 9498
+    },
+    {
+      "epoch": 0.9867040614937156,
+      "grad_norm": 0.4436101019382477,
+      "learning_rate": 4.361286363618322e-08,
+      "loss": 1.5564,
+      "step": 9499
+    },
+    {
+      "epoch": 0.986807936013296,
+      "grad_norm": 0.42672184109687805,
+      "learning_rate": 4.293417172078762e-08,
+      "loss": 1.5828,
+      "step": 9500
+    },
+    {
+      "epoch": 0.9869118105328762,
+      "grad_norm": 0.4650036096572876,
+      "learning_rate": 4.226079984322451e-08,
+      "loss": 1.5301,
+      "step": 9501
+    },
+    {
+      "epoch": 0.9870156850524566,
+      "grad_norm": 0.4062730371952057,
+      "learning_rate": 4.159274807520319e-08,
+      "loss": 1.6214,
+      "step": 9502
+    },
+    {
+      "epoch": 0.987119559572037,
+      "grad_norm": 0.4749537706375122,
+      "learning_rate": 4.0930016487861214e-08,
+      "loss": 1.7306,
+      "step": 9503
+    },
+    {
+      "epoch": 0.9872234340916173,
+      "grad_norm": 0.4257502555847168,
+      "learning_rate": 4.027260515177544e-08,
+      "loss": 1.8203,
+      "step": 9504
+    },
+    {
+      "epoch": 0.9873273086111977,
+      "grad_norm": 0.3824896514415741,
+      "learning_rate": 3.962051413695656e-08,
+      "loss": 1.5731,
+      "step": 9505
+    },
+    {
+      "epoch": 0.987431183130778,
+      "grad_norm": 0.41676878929138184,
+      "learning_rate": 3.897374351284899e-08,
+      "loss": 1.7339,
+      "step": 9506
+    },
+    {
+      "epoch": 0.9875350576503583,
+      "grad_norm": 0.4813458323478699,
+      "learning_rate": 3.8332293348325444e-08,
+      "loss": 1.993,
+      "step": 9507
+    },
+    {
+      "epoch": 0.9876389321699387,
+      "grad_norm": 0.47463130950927734,
+      "learning_rate": 3.769616371169793e-08,
+      "loss": 1.708,
+      "step": 9508
+    },
+    {
+      "epoch": 0.9877428066895191,
+      "grad_norm": 0.4519473612308502,
+      "learning_rate": 3.706535467070671e-08,
+      "loss": 1.9389,
+      "step": 9509
+    },
+    {
+      "epoch": 0.9878466812090994,
+      "grad_norm": 0.4201395809650421,
+      "learning_rate": 3.6439866292531376e-08,
+      "loss": 1.6609,
+      "step": 9510
+    },
+    {
+      "epoch": 0.9879505557286797,
+      "grad_norm": 0.4545467495918274,
+      "learning_rate": 3.5819698643779764e-08,
+      "loss": 1.7568,
+      "step": 9511
+    },
+    {
+      "epoch": 0.9880544302482601,
+      "grad_norm": 0.39346179366111755,
+      "learning_rate": 3.520485179048794e-08,
+      "loss": 1.6101,
+      "step": 9512
+    },
+    {
+      "epoch": 0.9881583047678405,
+      "grad_norm": 0.4246424436569214,
+      "learning_rate": 3.459532579814795e-08,
+      "loss": 1.7143,
+      "step": 9513
+    },
+    {
+      "epoch": 0.9882621792874208,
+      "grad_norm": 0.4520607590675354,
+      "learning_rate": 3.399112073165789e-08,
+      "loss": 1.5954,
+      "step": 9514
+    },
+    {
+      "epoch": 0.9883660538070012,
+      "grad_norm": 0.4503570795059204,
+      "learning_rate": 3.339223665536073e-08,
+      "loss": 1.8197,
+      "step": 9515
+    },
+    {
+      "epoch": 0.9884699283265815,
+      "grad_norm": 0.42323458194732666,
+      "learning_rate": 3.279867363303879e-08,
+      "loss": 1.6981,
+      "step": 9516
+    },
+    {
+      "epoch": 0.9885738028461618,
+      "grad_norm": 0.44448158144950867,
+      "learning_rate": 3.221043172789706e-08,
+      "loss": 1.5851,
+      "step": 9517
+    },
+    {
+      "epoch": 0.9886776773657422,
+      "grad_norm": 0.41508886218070984,
+      "learning_rate": 3.162751100257988e-08,
+      "loss": 1.6064,
+      "step": 9518
+    },
+    {
+      "epoch": 0.9887815518853226,
+      "grad_norm": 0.4167642891407013,
+      "learning_rate": 3.1049911519165365e-08,
+      "loss": 1.6993,
+      "step": 9519
+    },
+    {
+      "epoch": 0.9888854264049028,
+      "grad_norm": 0.4645927846431732,
+      "learning_rate": 3.047763333916542e-08,
+      "loss": 1.6729,
+      "step": 9520
+    },
+    {
+      "epoch": 0.9889893009244832,
+      "grad_norm": 0.42605888843536377,
+      "learning_rate": 2.991067652351465e-08,
+      "loss": 1.6401,
+      "step": 9521
+    },
+    {
+      "epoch": 0.9890931754440636,
+      "grad_norm": 0.3953196108341217,
+      "learning_rate": 2.9349041132598065e-08,
+      "loss": 1.6012,
+      "step": 9522
+    },
+    {
+      "epoch": 0.9891970499636439,
+      "grad_norm": 0.41111239790916443,
+      "learning_rate": 2.8792727226223393e-08,
+      "loss": 1.5285,
+      "step": 9523
+    },
+    {
+      "epoch": 0.9893009244832243,
+      "grad_norm": 0.4029116928577423,
+      "learning_rate": 2.824173486363213e-08,
+      "loss": 1.7036,
+      "step": 9524
+    },
+    {
+      "epoch": 0.9894047990028046,
+      "grad_norm": 0.44695624709129333,
+      "learning_rate": 2.7696064103505116e-08,
+      "loss": 1.7136,
+      "step": 9525
+    },
+    {
+      "epoch": 0.9895086735223849,
+      "grad_norm": 0.4348202347755432,
+      "learning_rate": 2.715571500394587e-08,
+      "loss": 1.6265,
+      "step": 9526
+    },
+    {
+      "epoch": 0.9896125480419653,
+      "grad_norm": 0.43105223774909973,
+      "learning_rate": 2.6620687622497255e-08,
+      "loss": 1.6818,
+      "step": 9527
+    },
+    {
+      "epoch": 0.9897164225615457,
+      "grad_norm": 0.4228067398071289,
+      "learning_rate": 2.609098201613591e-08,
+      "loss": 1.6076,
+      "step": 9528
+    },
+    {
+      "epoch": 0.989820297081126,
+      "grad_norm": 0.4084494113922119,
+      "learning_rate": 2.556659824127783e-08,
+      "loss": 1.6429,
+      "step": 9529
+    },
+    {
+      "epoch": 0.9899241716007063,
+      "grad_norm": 0.4497421979904175,
+      "learning_rate": 2.504753635376167e-08,
+      "loss": 1.5681,
+      "step": 9530
+    },
+    {
+      "epoch": 0.9900280461202867,
+      "grad_norm": 0.4513741731643677,
+      "learning_rate": 2.453379640886544e-08,
+      "loss": 1.6556,
+      "step": 9531
+    },
+    {
+      "epoch": 0.990131920639867,
+      "grad_norm": 0.38901105523109436,
+      "learning_rate": 2.4025378461289828e-08,
+      "loss": 1.502,
+      "step": 9532
+    },
+    {
+      "epoch": 0.9902357951594474,
+      "grad_norm": 0.4121716022491455,
+      "learning_rate": 2.352228256519151e-08,
+      "loss": 1.6831,
+      "step": 9533
+    },
+    {
+      "epoch": 0.9903396696790278,
+      "grad_norm": 0.4378368854522705,
+      "learning_rate": 2.3024508774133204e-08,
+      "loss": 1.6327,
+      "step": 9534
+    },
+    {
+      "epoch": 0.990443544198608,
+      "grad_norm": 0.39573803544044495,
+      "learning_rate": 2.2532057141128048e-08,
+      "loss": 1.5492,
+      "step": 9535
+    },
+    {
+      "epoch": 0.9905474187181884,
+      "grad_norm": 0.4019763469696045,
+      "learning_rate": 2.2044927718622987e-08,
+      "loss": 1.5991,
+      "step": 9536
+    },
+    {
+      "epoch": 0.9906512932377688,
+      "grad_norm": 0.40174970030784607,
+      "learning_rate": 2.1563120558487638e-08,
+      "loss": 1.5447,
+      "step": 9537
+    },
+    {
+      "epoch": 0.9907551677573492,
+      "grad_norm": 0.45444202423095703,
+      "learning_rate": 2.1086635712036507e-08,
+      "loss": 1.6873,
+      "step": 9538
+    },
+    {
+      "epoch": 0.9908590422769294,
+      "grad_norm": 0.42663663625717163,
+      "learning_rate": 2.061547323000679e-08,
+      "loss": 1.7317,
+      "step": 9539
+    },
+    {
+      "epoch": 0.9909629167965098,
+      "grad_norm": 0.40359655022621155,
+      "learning_rate": 2.014963316257501e-08,
+      "loss": 1.5926,
+      "step": 9540
+    },
+    {
+      "epoch": 0.9910667913160902,
+      "grad_norm": 0.40575510263442993,
+      "learning_rate": 1.9689115559345938e-08,
+      "loss": 1.5653,
+      "step": 9541
+    },
+    {
+      "epoch": 0.9911706658356705,
+      "grad_norm": 0.4161994457244873,
+      "learning_rate": 1.9233920469369225e-08,
+      "loss": 1.5923,
+      "step": 9542
+    },
+    {
+      "epoch": 0.9912745403552509,
+      "grad_norm": 0.4296528697013855,
+      "learning_rate": 1.878404794111166e-08,
+      "loss": 1.7091,
+      "step": 9543
+    },
+    {
+      "epoch": 0.9913784148748312,
+      "grad_norm": 0.4677174985408783,
+      "learning_rate": 1.8339498022490465e-08,
+      "loss": 1.7469,
+      "step": 9544
+    },
+    {
+      "epoch": 0.9914822893944115,
+      "grad_norm": 0.43237027525901794,
+      "learning_rate": 1.790027076083445e-08,
+      "loss": 1.6314,
+      "step": 9545
+    },
+    {
+      "epoch": 0.9915861639139919,
+      "grad_norm": 0.4431440830230713,
+      "learning_rate": 1.7466366202928408e-08,
+      "loss": 1.6766,
+      "step": 9546
+    },
+    {
+      "epoch": 0.9916900384335723,
+      "grad_norm": 0.43120649456977844,
+      "learning_rate": 1.7037784394968724e-08,
+      "loss": 1.6412,
+      "step": 9547
+    },
+    {
+      "epoch": 0.9917939129531526,
+      "grad_norm": 0.3956824541091919,
+      "learning_rate": 1.6614525382613323e-08,
+      "loss": 1.5854,
+      "step": 9548
+    },
+    {
+      "epoch": 0.9918977874727329,
+      "grad_norm": 0.41620564460754395,
+      "learning_rate": 1.6196589210915046e-08,
+      "loss": 1.5208,
+      "step": 9549
+    },
+    {
+      "epoch": 0.9920016619923133,
+      "grad_norm": 0.46985071897506714,
+      "learning_rate": 1.5783975924399398e-08,
+      "loss": 1.8193,
+      "step": 9550
+    },
+    {
+      "epoch": 0.9921055365118936,
+      "grad_norm": 0.44369015097618103,
+      "learning_rate": 1.5376685566992345e-08,
+      "loss": 1.5834,
+      "step": 9551
+    },
+    {
+      "epoch": 0.992209411031474,
+      "grad_norm": 0.40749645233154297,
+      "learning_rate": 1.4974718182075853e-08,
+      "loss": 1.5352,
+      "step": 9552
+    },
+    {
+      "epoch": 0.9923132855510544,
+      "grad_norm": 0.40627145767211914,
+      "learning_rate": 1.4578073812454574e-08,
+      "loss": 1.6593,
+      "step": 9553
+    },
+    {
+      "epoch": 0.9924171600706346,
+      "grad_norm": 0.4516492784023285,
+      "learning_rate": 1.418675250036694e-08,
+      "loss": 1.7582,
+      "step": 9554
+    },
+    {
+      "epoch": 0.992521034590215,
+      "grad_norm": 0.44202539324760437,
+      "learning_rate": 1.3800754287485174e-08,
+      "loss": 1.6358,
+      "step": 9555
+    },
+    {
+      "epoch": 0.9926249091097954,
+      "grad_norm": 0.4185415208339691,
+      "learning_rate": 1.3420079214915282e-08,
+      "loss": 1.5889,
+      "step": 9556
+    },
+    {
+      "epoch": 0.9927287836293757,
+      "grad_norm": 0.41641661524772644,
+      "learning_rate": 1.3044727323202611e-08,
+      "loss": 1.6185,
+      "step": 9557
+    },
+    {
+      "epoch": 0.992832658148956,
+      "grad_norm": 0.4016210436820984,
+      "learning_rate": 1.2674698652304085e-08,
+      "loss": 1.6872,
+      "step": 9558
+    },
+    {
+      "epoch": 0.9929365326685364,
+      "grad_norm": 0.4109307825565338,
+      "learning_rate": 1.2309993241638173e-08,
+      "loss": 1.7338,
+      "step": 9559
+    },
+    {
+      "epoch": 0.9930404071881167,
+      "grad_norm": 0.4159395098686218,
+      "learning_rate": 1.1950611130040479e-08,
+      "loss": 1.8372,
+      "step": 9560
+    },
+    {
+      "epoch": 0.9931442817076971,
+      "grad_norm": 0.37879422307014465,
+      "learning_rate": 1.1596552355780388e-08,
+      "loss": 1.5309,
+      "step": 9561
+    },
+    {
+      "epoch": 0.9932481562272775,
+      "grad_norm": 0.46690380573272705,
+      "learning_rate": 1.1247816956561075e-08,
+      "loss": 1.6077,
+      "step": 9562
+    },
+    {
+      "epoch": 0.9933520307468579,
+      "grad_norm": 0.444859117269516,
+      "learning_rate": 1.0904404969525051e-08,
+      "loss": 1.6831,
+      "step": 9563
+    },
+    {
+      "epoch": 0.9934559052664381,
+      "grad_norm": 0.46033644676208496,
+      "learning_rate": 1.0566316431237511e-08,
+      "loss": 1.7443,
+      "step": 9564
+    },
+    {
+      "epoch": 0.9935597797860185,
+      "grad_norm": 0.44090694189071655,
+      "learning_rate": 1.0233551377702988e-08,
+      "loss": 1.7238,
+      "step": 9565
+    },
+    {
+      "epoch": 0.9936636543055989,
+      "grad_norm": 0.4091147780418396,
+      "learning_rate": 9.906109844359802e-09,
+      "loss": 1.6286,
+      "step": 9566
+    },
+    {
+      "epoch": 0.9937675288251792,
+      "grad_norm": 0.41084280610084534,
+      "learning_rate": 9.583991866080055e-09,
+      "loss": 1.5039,
+      "step": 9567
+    },
+    {
+      "epoch": 0.9938714033447595,
+      "grad_norm": 0.4234639108181,
+      "learning_rate": 9.267197477169643e-09,
+      "loss": 1.6008,
+      "step": 9568
+    },
+    {
+      "epoch": 0.9939752778643399,
+      "grad_norm": 0.40848711133003235,
+      "learning_rate": 8.955726711351586e-09,
+      "loss": 1.6676,
+      "step": 9569
+    },
+    {
+      "epoch": 0.9940791523839202,
+      "grad_norm": 0.3974449932575226,
+      "learning_rate": 8.649579601810453e-09,
+      "loss": 1.4568,
+      "step": 9570
+    },
+    {
+      "epoch": 0.9941830269035006,
+      "grad_norm": 0.42480480670928955,
+      "learning_rate": 8.34875618113684e-09,
+      "loss": 1.7452,
+      "step": 9571
+    },
+    {
+      "epoch": 0.994286901423081,
+      "grad_norm": 0.4431557059288025,
+      "learning_rate": 8.053256481371785e-09,
+      "loss": 1.7141,
+      "step": 9572
+    },
+    {
+      "epoch": 0.9943907759426612,
+      "grad_norm": 0.46281933784484863,
+      "learning_rate": 7.763080533984557e-09,
+      "loss": 1.7414,
+      "step": 9573
+    },
+    {
+      "epoch": 0.9944946504622416,
+      "grad_norm": 0.44635605812072754,
+      "learning_rate": 7.478228369872664e-09,
+      "loss": 1.7477,
+      "step": 9574
+    },
+    {
+      "epoch": 0.994598524981822,
+      "grad_norm": 0.40345069766044617,
+      "learning_rate": 7.198700019378502e-09,
+      "loss": 1.7388,
+      "step": 9575
+    },
+    {
+      "epoch": 0.9947023995014023,
+      "grad_norm": 0.4090345799922943,
+      "learning_rate": 6.924495512256046e-09,
+      "loss": 1.6223,
+      "step": 9576
+    },
+    {
+      "epoch": 0.9948062740209827,
+      "grad_norm": 0.4181837737560272,
+      "learning_rate": 6.655614877720817e-09,
+      "loss": 1.6013,
+      "step": 9577
+    },
+    {
+      "epoch": 0.994910148540563,
+      "grad_norm": 0.4603481590747833,
+      "learning_rate": 6.392058144394364e-09,
+      "loss": 1.6791,
+      "step": 9578
+    },
+    {
+      "epoch": 0.9950140230601433,
+      "grad_norm": 0.43850967288017273,
+      "learning_rate": 6.133825340348675e-09,
+      "loss": 1.6949,
+      "step": 9579
+    },
+    {
+      "epoch": 0.9951178975797237,
+      "grad_norm": 0.42286258935928345,
+      "learning_rate": 5.8809164930839765e-09,
+      "loss": 1.6022,
+      "step": 9580
+    },
+    {
+      "epoch": 0.9952217720993041,
+      "grad_norm": 0.4476601481437683,
+      "learning_rate": 5.6333316295342775e-09,
+      "loss": 1.8264,
+      "step": 9581
+    },
+    {
+      "epoch": 0.9953256466188843,
+      "grad_norm": 0.4325719475746155,
+      "learning_rate": 5.391070776061824e-09,
+      "loss": 1.5983,
+      "step": 9582
+    },
+    {
+      "epoch": 0.9954295211384647,
+      "grad_norm": 0.41536781191825867,
+      "learning_rate": 5.154133958468199e-09,
+      "loss": 1.6543,
+      "step": 9583
+    },
+    {
+      "epoch": 0.9955333956580451,
+      "grad_norm": 0.4049381613731384,
+      "learning_rate": 4.922521201988772e-09,
+      "loss": 1.6388,
+      "step": 9584
+    },
+    {
+      "epoch": 0.9956372701776254,
+      "grad_norm": 0.3786325752735138,
+      "learning_rate": 4.6962325312760455e-09,
+      "loss": 1.5859,
+      "step": 9585
+    },
+    {
+      "epoch": 0.9957411446972058,
+      "grad_norm": 0.45597589015960693,
+      "learning_rate": 4.475267970444064e-09,
+      "loss": 1.8203,
+      "step": 9586
+    },
+    {
+      "epoch": 0.9958450192167861,
+      "grad_norm": 0.42808467149734497,
+      "learning_rate": 4.259627543012901e-09,
+      "loss": 1.6721,
+      "step": 9587
+    },
+    {
+      "epoch": 0.9959488937363664,
+      "grad_norm": 0.42117738723754883,
+      "learning_rate": 4.04931127194752e-09,
+      "loss": 1.5659,
+      "step": 9588
+    },
+    {
+      "epoch": 0.9960527682559468,
+      "grad_norm": 0.4049902856349945,
+      "learning_rate": 3.844319179646671e-09,
+      "loss": 1.6087,
+      "step": 9589
+    },
+    {
+      "epoch": 0.9961566427755272,
+      "grad_norm": 0.40100494027137756,
+      "learning_rate": 3.6446512879428907e-09,
+      "loss": 1.654,
+      "step": 9590
+    },
+    {
+      "epoch": 0.9962605172951076,
+      "grad_norm": 0.3948621153831482,
+      "learning_rate": 3.4503076180969486e-09,
+      "loss": 1.661,
+      "step": 9591
+    },
+    {
+      "epoch": 0.9963643918146878,
+      "grad_norm": 0.44757014513015747,
+      "learning_rate": 3.2612881908089533e-09,
+      "loss": 1.6552,
+      "step": 9592
+    },
+    {
+      "epoch": 0.9964682663342682,
+      "grad_norm": 0.413869172334671,
+      "learning_rate": 3.0775930261961462e-09,
+      "loss": 1.6067,
+      "step": 9593
+    },
+    {
+      "epoch": 0.9965721408538486,
+      "grad_norm": 0.46646273136138916,
+      "learning_rate": 2.89922214383731e-09,
+      "loss": 1.6194,
+      "step": 9594
+    },
+    {
+      "epoch": 0.9966760153734289,
+      "grad_norm": 0.39553186297416687,
+      "learning_rate": 2.726175562711708e-09,
+      "loss": 1.5895,
+      "step": 9595
+    },
+    {
+      "epoch": 0.9967798898930093,
+      "grad_norm": 0.42931756377220154,
+      "learning_rate": 2.558453301260144e-09,
+      "loss": 1.5636,
+      "step": 9596
+    },
+    {
+      "epoch": 0.9968837644125896,
+      "grad_norm": 0.4277789890766144,
+      "learning_rate": 2.3960553773350046e-09,
+      "loss": 1.6763,
+      "step": 9597
+    },
+    {
+      "epoch": 0.9969876389321699,
+      "grad_norm": 0.4121456444263458,
+      "learning_rate": 2.2389818082335645e-09,
+      "loss": 1.6111,
+      "step": 9598
+    },
+    {
+      "epoch": 0.9970915134517503,
+      "grad_norm": 0.4208837151527405,
+      "learning_rate": 2.0872326106868845e-09,
+      "loss": 1.7814,
+      "step": 9599
+    },
+    {
+      "epoch": 0.9971953879713307,
+      "grad_norm": 0.4421430826187134,
+      "learning_rate": 1.9408078008431586e-09,
+      "loss": 1.6691,
+      "step": 9600
+    },
+    {
+      "epoch": 0.9972992624909109,
+      "grad_norm": 0.4708458483219147,
+      "learning_rate": 1.7997073943121222e-09,
+      "loss": 1.6991,
+      "step": 9601
+    },
+    {
+      "epoch": 0.9974031370104913,
+      "grad_norm": 0.416151762008667,
+      "learning_rate": 1.6639314061095422e-09,
+      "loss": 1.6752,
+      "step": 9602
+    },
+    {
+      "epoch": 0.9975070115300717,
+      "grad_norm": 0.408317506313324,
+      "learning_rate": 1.5334798506905224e-09,
+      "loss": 1.4994,
+      "step": 9603
+    },
+    {
+      "epoch": 0.997610886049652,
+      "grad_norm": 0.4764207601547241,
+      "learning_rate": 1.4083527419606058e-09,
+      "loss": 1.8033,
+      "step": 9604
+    },
+    {
+      "epoch": 0.9977147605692324,
+      "grad_norm": 0.4480811655521393,
+      "learning_rate": 1.2885500932313666e-09,
+      "loss": 1.8291,
+      "step": 9605
+    },
+    {
+      "epoch": 0.9978186350888127,
+      "grad_norm": 0.4367046058177948,
+      "learning_rate": 1.1740719172703697e-09,
+      "loss": 1.6905,
+      "step": 9606
+    },
+    {
+      "epoch": 0.997922509608393,
+      "grad_norm": 0.43447238206863403,
+      "learning_rate": 1.0649182262623125e-09,
+      "loss": 1.6493,
+      "step": 9607
+    },
+    {
+      "epoch": 0.9980263841279734,
+      "grad_norm": 0.4337233006954193,
+      "learning_rate": 9.610890318312305e-10,
+      "loss": 1.6764,
+      "step": 9608
+    },
+    {
+      "epoch": 0.9981302586475538,
+      "grad_norm": 0.42332491278648376,
+      "learning_rate": 8.625843450404958e-10,
+      "loss": 1.6347,
+      "step": 9609
+    },
+    {
+      "epoch": 0.998234133167134,
+      "grad_norm": 0.4210823178291321,
+      "learning_rate": 7.69404176376165e-10,
+      "loss": 1.7052,
+      "step": 9610
+    },
+    {
+      "epoch": 0.9983380076867144,
+      "grad_norm": 0.3997040092945099,
+      "learning_rate": 6.81548535763632e-10,
+      "loss": 1.5105,
+      "step": 9611
+    },
+    {
+      "epoch": 0.9984418822062948,
+      "grad_norm": 0.43500977754592896,
+      "learning_rate": 5.990174325509745e-10,
+      "loss": 1.6361,
+      "step": 9612
+    },
+    {
+      "epoch": 0.9985457567258751,
+      "grad_norm": 0.459688276052475,
+      "learning_rate": 5.218108755367102e-10,
+      "loss": 1.6404,
+      "step": 9613
+    },
+    {
+      "epoch": 0.9986496312454555,
+      "grad_norm": 0.43085619807243347,
+      "learning_rate": 4.499288729364892e-10,
+      "loss": 1.642,
+      "step": 9614
+    },
+    {
+      "epoch": 0.9987535057650359,
+      "grad_norm": 0.4042317569255829,
+      "learning_rate": 3.833714324108506e-10,
+      "loss": 1.7038,
+      "step": 9615
+    },
+    {
+      "epoch": 0.9988573802846162,
+      "grad_norm": 0.43363288044929504,
+      "learning_rate": 3.221385610430172e-10,
+      "loss": 1.8699,
+      "step": 9616
+    },
+    {
+      "epoch": 0.9989612548041965,
+      "grad_norm": 0.40789350867271423,
+      "learning_rate": 2.6623026534999817e-10,
+      "loss": 1.7067,
+      "step": 9617
+    },
+    {
+      "epoch": 0.9990651293237769,
+      "grad_norm": 0.45757153630256653,
+      "learning_rate": 2.1564655129369115e-10,
+      "loss": 1.9236,
+      "step": 9618
+    },
+    {
+      "epoch": 0.9991690038433573,
+      "grad_norm": 0.4801182746887207,
+      "learning_rate": 1.7038742425867781e-10,
+      "loss": 1.8674,
+      "step": 9619
+    },
+    {
+      "epoch": 0.9992728783629375,
+      "grad_norm": 0.41760823130607605,
+      "learning_rate": 1.3045288906332608e-10,
+      "loss": 1.591,
+      "step": 9620
+    },
+    {
+      "epoch": 0.9993767528825179,
+      "grad_norm": 0.46719786524772644,
+      "learning_rate": 9.584294995979015e-11,
+      "loss": 1.3563,
+      "step": 9621
+    },
+    {
+      "epoch": 0.9994806274020983,
+      "grad_norm": 0.4825066030025482,
+      "learning_rate": 6.655761063956157e-11,
+      "loss": 1.7152,
+      "step": 9622
+    },
+    {
+      "epoch": 0.9995845019216786,
+      "grad_norm": 0.4088488817214966,
+      "learning_rate": 4.259687421126479e-11,
+      "loss": 1.6796,
+      "step": 9623
+    },
+    {
+      "epoch": 0.999688376441259,
+      "grad_norm": 0.4588576853275299,
+      "learning_rate": 2.3960743233963913e-11,
+      "loss": 1.8781,
+      "step": 9624
+    },
+    {
+      "epoch": 0.9997922509608393,
+      "grad_norm": 0.46052485704421997,
+      "learning_rate": 1.0649219683855905e-11,
+      "loss": 1.5272,
+      "step": 9625
+    },
+    {
+      "epoch": 0.9998961254804196,
+      "grad_norm": 0.4075811207294464,
+      "learning_rate": 2.6623049931284727e-12,
+      "loss": 1.7282,
+      "step": 9626
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5596733093261719,
+      "learning_rate": 0.0,
+      "loss": 1.8582,
+      "step": 9627
+    },
+    {
+      "epoch": 1.0,
+      "step": 9627,
+      "total_flos": 1.452208309419167e+19,
+      "train_loss": 1.7118117516813738,
+      "train_runtime": 275545.5724,
+      "train_samples_per_second": 0.14,
+      "train_steps_per_second": 0.035
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9627,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.452208309419167e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}