diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4151 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9987239472564866,
+  "eval_steps": 500,
+  "global_step": 587,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0017014036580178648,
+      "grad_norm": 2.168562940412461,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.1424,
+      "step": 1
+    },
+    {
+      "epoch": 0.0034028073160357296,
+      "grad_norm": 2.250032849091299,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.1697,
+      "step": 2
+    },
+    {
+      "epoch": 0.005104210974053594,
+      "grad_norm": 2.124744531823449,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.1364,
+      "step": 3
+    },
+    {
+      "epoch": 0.006805614632071459,
+      "grad_norm": 2.060484893473413,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.1471,
+      "step": 4
+    },
+    {
+      "epoch": 0.008507018290089324,
+      "grad_norm": 1.89749522427158,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 1.1181,
+      "step": 5
+    },
+    {
+      "epoch": 0.010208421948107189,
+      "grad_norm": 1.3357851941439673,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.0893,
+      "step": 6
+    },
+    {
+      "epoch": 0.011909825606125054,
+      "grad_norm": 1.2850663628144225,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.0484,
+      "step": 7
+    },
+    {
+      "epoch": 0.013611229264142918,
+      "grad_norm": 1.636445160753731,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.0308,
+      "step": 8
+    },
+    {
+      "epoch": 0.015312632922160783,
+      "grad_norm": 1.3002228280073427,
+      "learning_rate": 1e-05,
+      "loss": 1.0234,
+      "step": 9
+    },
+    {
+      "epoch": 0.017014036580178648,
+      "grad_norm": 1.2380680351381246,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.9064,
+      "step": 10
+    },
+    {
+      "epoch": 0.01871544023819651,
+      "grad_norm": 1.2765329463296096,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 0.8696,
+      "step": 11
+    },
+    {
+      "epoch": 0.020416843896214378,
+      "grad_norm": 1.060585818467504,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8444,
+      "step": 12
+    },
+    {
+      "epoch": 0.02211824755423224,
+      "grad_norm": 1.1232952035148502,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 0.841,
+      "step": 13
+    },
+    {
+      "epoch": 0.023819651212250107,
+      "grad_norm": 1.107943114775762,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 0.7655,
+      "step": 14
+    },
+    {
+      "epoch": 0.02552105487026797,
+      "grad_norm": 1.0059483053370937,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.7063,
+      "step": 15
+    },
+    {
+      "epoch": 0.027222458528285837,
+      "grad_norm": 0.9803969915500654,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.6458,
+      "step": 16
+    },
+    {
+      "epoch": 0.0289238621863037,
+      "grad_norm": 0.9105498405702808,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.6321,
+      "step": 17
+    },
+    {
+      "epoch": 0.030625265844321566,
+      "grad_norm": 0.9433546278214147,
+      "learning_rate": 2e-05,
+      "loss": 0.493,
+      "step": 18
+    },
+    {
+      "epoch": 0.03232666950233943,
+      "grad_norm": 0.8625027370409255,
+      "learning_rate": 1.9999847579243196e-05,
+      "loss": 0.5339,
+      "step": 19
+    },
+    {
+      "epoch": 0.034028073160357296,
+      "grad_norm": 0.8101152462345085,
+      "learning_rate": 1.9999390321619196e-05,
+      "loss": 0.4566,
+      "step": 20
+    },
+    {
+      "epoch": 0.03572947681837516,
+      "grad_norm": 0.7923424067716904,
+      "learning_rate": 1.9998628241067113e-05,
+      "loss": 0.4507,
+      "step": 21
+    },
+    {
+      "epoch": 0.03743088047639302,
+      "grad_norm": 0.7681394677537327,
+      "learning_rate": 1.9997561360818322e-05,
+      "loss": 0.4265,
+      "step": 22
+    },
+    {
+      "epoch": 0.03913228413441089,
+      "grad_norm": 0.7031739222010267,
+      "learning_rate": 1.999618971339577e-05,
+      "loss": 0.4174,
+      "step": 23
+    },
+    {
+      "epoch": 0.040833687792428755,
+      "grad_norm": 0.6603135403892707,
+      "learning_rate": 1.9994513340612957e-05,
+      "loss": 0.3599,
+      "step": 24
+    },
+    {
+      "epoch": 0.04253509145044662,
+      "grad_norm": 0.7260664469398356,
+      "learning_rate": 1.9992532293572688e-05,
+      "loss": 0.3398,
+      "step": 25
+    },
+    {
+      "epoch": 0.04423649510846448,
+      "grad_norm": 0.590669951494315,
+      "learning_rate": 1.9990246632665503e-05,
+      "loss": 0.2616,
+      "step": 26
+    },
+    {
+      "epoch": 0.04593789876648235,
+      "grad_norm": 0.6319158523814317,
+      "learning_rate": 1.998765642756783e-05,
+      "loss": 0.3082,
+      "step": 27
+    },
+    {
+      "epoch": 0.047639302424500214,
+      "grad_norm": 0.6331615276638973,
+      "learning_rate": 1.9984761757239878e-05,
+      "loss": 0.2933,
+      "step": 28
+    },
+    {
+      "epoch": 0.04934070608251808,
+      "grad_norm": 0.6376428367891515,
+      "learning_rate": 1.998156270992321e-05,
+      "loss": 0.2612,
+      "step": 29
+    },
+    {
+      "epoch": 0.05104210974053594,
+      "grad_norm": 0.6322085618139611,
+      "learning_rate": 1.9978059383138073e-05,
+      "loss": 0.2397,
+      "step": 30
+    },
+    {
+      "epoch": 0.05274351339855381,
+      "grad_norm": 0.6174252074442552,
+      "learning_rate": 1.997425188368041e-05,
+      "loss": 0.2566,
+      "step": 31
+    },
+    {
+      "epoch": 0.05444491705657167,
+      "grad_norm": 0.5327916753070241,
+      "learning_rate": 1.9970140327618612e-05,
+      "loss": 0.2416,
+      "step": 32
+    },
+    {
+      "epoch": 0.05614632071458953,
+      "grad_norm": 0.5479078907290673,
+      "learning_rate": 1.9965724840289972e-05,
+      "loss": 0.1781,
+      "step": 33
+    },
+    {
+      "epoch": 0.0578477243726074,
+      "grad_norm": 0.5315756790547577,
+      "learning_rate": 1.9961005556296875e-05,
+      "loss": 0.2258,
+      "step": 34
+    },
+    {
+      "epoch": 0.059549128030625266,
+      "grad_norm": 0.48386295113930106,
+      "learning_rate": 1.9955982619502693e-05,
+      "loss": 0.1679,
+      "step": 35
+    },
+    {
+      "epoch": 0.06125053168864313,
+      "grad_norm": 0.5185655350233463,
+      "learning_rate": 1.9950656183027392e-05,
+      "loss": 0.2022,
+      "step": 36
+    },
+    {
+      "epoch": 0.06295193534666099,
+      "grad_norm": 0.5425773293966369,
+      "learning_rate": 1.994502640924286e-05,
+      "loss": 0.2144,
+      "step": 37
+    },
+    {
+      "epoch": 0.06465333900467886,
+      "grad_norm": 0.39700624229353654,
+      "learning_rate": 1.993909346976798e-05,
+      "loss": 0.1361,
+      "step": 38
+    },
+    {
+      "epoch": 0.06635474266269673,
+      "grad_norm": 0.5024564814818023,
+      "learning_rate": 1.993285754546338e-05,
+      "loss": 0.1618,
+      "step": 39
+    },
+    {
+      "epoch": 0.06805614632071459,
+      "grad_norm": 0.4298514607710804,
+      "learning_rate": 1.9926318826425905e-05,
+      "loss": 0.1801,
+      "step": 40
+    },
+    {
+      "epoch": 0.06975754997873246,
+      "grad_norm": 0.4631733150959672,
+      "learning_rate": 1.9919477511982873e-05,
+      "loss": 0.1212,
+      "step": 41
+    },
+    {
+      "epoch": 0.07145895363675032,
+      "grad_norm": 0.46629563189251916,
+      "learning_rate": 1.991233381068594e-05,
+      "loss": 0.1538,
+      "step": 42
+    },
+    {
+      "epoch": 0.07316035729476818,
+      "grad_norm": 0.4077353003444054,
+      "learning_rate": 1.990488794030478e-05,
+      "loss": 0.1601,
+      "step": 43
+    },
+    {
+      "epoch": 0.07486176095278604,
+      "grad_norm": 0.3874990470468177,
+      "learning_rate": 1.9897140127820432e-05,
+      "loss": 0.125,
+      "step": 44
+    },
+    {
+      "epoch": 0.07656316461080391,
+      "grad_norm": 0.3638523211055093,
+      "learning_rate": 1.9889090609418384e-05,
+      "loss": 0.1366,
+      "step": 45
+    },
+    {
+      "epoch": 0.07826456826882178,
+      "grad_norm": 0.5006533990444182,
+      "learning_rate": 1.9880739630481376e-05,
+      "loss": 0.1466,
+      "step": 46
+    },
+    {
+      "epoch": 0.07996597192683964,
+      "grad_norm": 0.4320089119157406,
+      "learning_rate": 1.9872087445581912e-05,
+      "loss": 0.1256,
+      "step": 47
+    },
+    {
+      "epoch": 0.08166737558485751,
+      "grad_norm": 0.3330675873783333,
+      "learning_rate": 1.9863134318474504e-05,
+      "loss": 0.1066,
+      "step": 48
+    },
+    {
+      "epoch": 0.08336877924287538,
+      "grad_norm": 0.4410537592519413,
+      "learning_rate": 1.985388052208764e-05,
+      "loss": 0.1381,
+      "step": 49
+    },
+    {
+      "epoch": 0.08507018290089324,
+      "grad_norm": 0.4003560082493989,
+      "learning_rate": 1.9844326338515444e-05,
+      "loss": 0.1202,
+      "step": 50
+    },
+    {
+      "epoch": 0.0867715865589111,
+      "grad_norm": 0.41012939531437076,
+      "learning_rate": 1.9834472059009097e-05,
+      "loss": 0.1246,
+      "step": 51
+    },
+    {
+      "epoch": 0.08847299021692896,
+      "grad_norm": 0.408091381990657,
+      "learning_rate": 1.982431798396794e-05,
+      "loss": 0.1289,
+      "step": 52
+    },
+    {
+      "epoch": 0.09017439387494683,
+      "grad_norm": 0.4821914250638501,
+      "learning_rate": 1.9813864422930345e-05,
+      "loss": 0.1312,
+      "step": 53
+    },
+    {
+      "epoch": 0.0918757975329647,
+      "grad_norm": 0.40793016164627693,
+      "learning_rate": 1.9803111694564246e-05,
+      "loss": 0.1237,
+      "step": 54
+    },
+    {
+      "epoch": 0.09357720119098256,
+      "grad_norm": 0.3734013818822545,
+      "learning_rate": 1.9792060126657437e-05,
+      "loss": 0.1049,
+      "step": 55
+    },
+    {
+      "epoch": 0.09527860484900043,
+      "grad_norm": 0.3186480775193613,
+      "learning_rate": 1.9780710056107587e-05,
+      "loss": 0.0911,
+      "step": 56
+    },
+    {
+      "epoch": 0.0969800085070183,
+      "grad_norm": 0.39667217504563396,
+      "learning_rate": 1.976906182891197e-05,
+      "loss": 0.0918,
+      "step": 57
+    },
+    {
+      "epoch": 0.09868141216503616,
+      "grad_norm": 0.4482874203710604,
+      "learning_rate": 1.97571158001569e-05,
+      "loss": 0.1209,
+      "step": 58
+    },
+    {
+      "epoch": 0.10038281582305401,
+      "grad_norm": 0.2965972804159924,
+      "learning_rate": 1.9744872334006936e-05,
+      "loss": 0.0932,
+      "step": 59
+    },
+    {
+      "epoch": 0.10208421948107188,
+      "grad_norm": 0.3566484676555622,
+      "learning_rate": 1.973233180369374e-05,
+      "loss": 0.1248,
+      "step": 60
+    },
+    {
+      "epoch": 0.10378562313908975,
+      "grad_norm": 0.30896394609488814,
+      "learning_rate": 1.9719494591504747e-05,
+      "loss": 0.0851,
+      "step": 61
+    },
+    {
+      "epoch": 0.10548702679710761,
+      "grad_norm": 0.2903927724315073,
+      "learning_rate": 1.9706361088771474e-05,
+      "loss": 0.0763,
+      "step": 62
+    },
+    {
+      "epoch": 0.10718843045512548,
+      "grad_norm": 0.3405843767078732,
+      "learning_rate": 1.96929316958576e-05,
+      "loss": 0.103,
+      "step": 63
+    },
+    {
+      "epoch": 0.10888983411314335,
+      "grad_norm": 0.40152278275783043,
+      "learning_rate": 1.9679206822146776e-05,
+      "loss": 0.1084,
+      "step": 64
+    },
+    {
+      "epoch": 0.11059123777116121,
+      "grad_norm": 0.3372032580489575,
+      "learning_rate": 1.9665186886030135e-05,
+      "loss": 0.1022,
+      "step": 65
+    },
+    {
+      "epoch": 0.11229264142917907,
+      "grad_norm": 0.332966030135067,
+      "learning_rate": 1.9650872314893523e-05,
+      "loss": 0.0891,
+      "step": 66
+    },
+    {
+      "epoch": 0.11399404508719693,
+      "grad_norm": 0.3832837550171733,
+      "learning_rate": 1.9636263545104498e-05,
+      "loss": 0.1181,
+      "step": 67
+    },
+    {
+      "epoch": 0.1156954487452148,
+      "grad_norm": 0.40356329916123346,
+      "learning_rate": 1.962136102199901e-05,
+      "loss": 0.1262,
+      "step": 68
+    },
+    {
+      "epoch": 0.11739685240323267,
+      "grad_norm": 0.3689423450550846,
+      "learning_rate": 1.9606165199867822e-05,
+      "loss": 0.0826,
+      "step": 69
+    },
+    {
+      "epoch": 0.11909825606125053,
+      "grad_norm": 0.3457217662571401,
+      "learning_rate": 1.959067654194268e-05,
+      "loss": 0.1007,
+      "step": 70
+    },
+    {
+      "epoch": 0.1207996597192684,
+      "grad_norm": 0.3511006571491925,
+      "learning_rate": 1.9574895520382183e-05,
+      "loss": 0.1263,
+      "step": 71
+    },
+    {
+      "epoch": 0.12250106337728627,
+      "grad_norm": 0.36819206266348997,
+      "learning_rate": 1.955882261625737e-05,
+      "loss": 0.1137,
+      "step": 72
+    },
+    {
+      "epoch": 0.12420246703530413,
+      "grad_norm": 0.2607004742274578,
+      "learning_rate": 1.9542458319537094e-05,
+      "loss": 0.0914,
+      "step": 73
+    },
+    {
+      "epoch": 0.12590387069332198,
+      "grad_norm": 0.2773076381582836,
+      "learning_rate": 1.9525803129073046e-05,
+      "loss": 0.0592,
+      "step": 74
+    },
+    {
+      "epoch": 0.12760527435133986,
+      "grad_norm": 0.28951070788934496,
+      "learning_rate": 1.9508857552584574e-05,
+      "loss": 0.0928,
+      "step": 75
+    },
+    {
+      "epoch": 0.12930667800935772,
+      "grad_norm": 0.35718485474187367,
+      "learning_rate": 1.9491622106643195e-05,
+      "loss": 0.0952,
+      "step": 76
+    },
+    {
+      "epoch": 0.1310080816673756,
+      "grad_norm": 0.34891501382529494,
+      "learning_rate": 1.9474097316656856e-05,
+      "loss": 0.1004,
+      "step": 77
+    },
+    {
+      "epoch": 0.13270948532539345,
+      "grad_norm": 0.28733346641687174,
+      "learning_rate": 1.9456283716853906e-05,
+      "loss": 0.0823,
+      "step": 78
+    },
+    {
+      "epoch": 0.1344108889834113,
+      "grad_norm": 0.3735237899238901,
+      "learning_rate": 1.9438181850266815e-05,
+      "loss": 0.121,
+      "step": 79
+    },
+    {
+      "epoch": 0.13611229264142918,
+      "grad_norm": 0.2904974285120679,
+      "learning_rate": 1.941979226871563e-05,
+      "loss": 0.0768,
+      "step": 80
+    },
+    {
+      "epoch": 0.13781369629944704,
+      "grad_norm": 0.32117427662146836,
+      "learning_rate": 1.9401115532791134e-05,
+      "loss": 0.0837,
+      "step": 81
+    },
+    {
+      "epoch": 0.13951509995746492,
+      "grad_norm": 0.40967624537549835,
+      "learning_rate": 1.938215221183777e-05,
+      "loss": 0.1061,
+      "step": 82
+    },
+    {
+      "epoch": 0.14121650361548277,
+      "grad_norm": 0.33629708001931885,
+      "learning_rate": 1.936290288393629e-05,
+      "loss": 0.0892,
+      "step": 83
+    },
+    {
+      "epoch": 0.14291790727350065,
+      "grad_norm": 0.31834590604771723,
+      "learning_rate": 1.9343368135886112e-05,
+      "loss": 0.1088,
+      "step": 84
+    },
+    {
+      "epoch": 0.1446193109315185,
+      "grad_norm": 0.30988146926514176,
+      "learning_rate": 1.932354856318746e-05,
+      "loss": 0.0864,
+      "step": 85
+    },
+    {
+      "epoch": 0.14632071458953635,
+      "grad_norm": 0.41148714128819724,
+      "learning_rate": 1.9303444770023184e-05,
+      "loss": 0.0867,
+      "step": 86
+    },
+    {
+      "epoch": 0.14802211824755424,
+      "grad_norm": 0.3641579696865324,
+      "learning_rate": 1.9283057369240358e-05,
+      "loss": 0.1149,
+      "step": 87
+    },
+    {
+      "epoch": 0.1497235219055721,
+      "grad_norm": 0.3359968169832728,
+      "learning_rate": 1.9262386982331596e-05,
+      "loss": 0.0959,
+      "step": 88
+    },
+    {
+      "epoch": 0.15142492556358997,
+      "grad_norm": 0.2849324477084618,
+      "learning_rate": 1.9241434239416093e-05,
+      "loss": 0.0957,
+      "step": 89
+    },
+    {
+      "epoch": 0.15312632922160782,
+      "grad_norm": 0.29673544787967593,
+      "learning_rate": 1.922019977922045e-05,
+      "loss": 0.0823,
+      "step": 90
+    },
+    {
+      "epoch": 0.1548277328796257,
+      "grad_norm": 0.28046315847608455,
+      "learning_rate": 1.919868424905915e-05,
+      "loss": 0.0885,
+      "step": 91
+    },
+    {
+      "epoch": 0.15652913653764355,
+      "grad_norm": 0.3024116439443639,
+      "learning_rate": 1.9176888304814882e-05,
+      "loss": 0.0919,
+      "step": 92
+    },
+    {
+      "epoch": 0.15823054019566143,
+      "grad_norm": 0.4139703560637819,
+      "learning_rate": 1.9154812610918503e-05,
+      "loss": 0.0938,
+      "step": 93
+    },
+    {
+      "epoch": 0.1599319438536793,
+      "grad_norm": 0.2618424473915207,
+      "learning_rate": 1.913245784032881e-05,
+      "loss": 0.0684,
+      "step": 94
+    },
+    {
+      "epoch": 0.16163334751169714,
+      "grad_norm": 0.31266175001831703,
+      "learning_rate": 1.9109824674512014e-05,
+      "loss": 0.0673,
+      "step": 95
+    },
+    {
+      "epoch": 0.16333475116971502,
+      "grad_norm": 0.3741829507920355,
+      "learning_rate": 1.9086913803420966e-05,
+      "loss": 0.1097,
+      "step": 96
+    },
+    {
+      "epoch": 0.16503615482773287,
+      "grad_norm": 0.3356628202803492,
+      "learning_rate": 1.906372592547413e-05,
+      "loss": 0.1028,
+      "step": 97
+    },
+    {
+      "epoch": 0.16673755848575075,
+      "grad_norm": 0.34106497626751014,
+      "learning_rate": 1.9040261747534282e-05,
+      "loss": 0.0989,
+      "step": 98
+    },
+    {
+      "epoch": 0.1684389621437686,
+      "grad_norm": 0.29746452263621875,
+      "learning_rate": 1.9016521984886984e-05,
+      "loss": 0.0718,
+      "step": 99
+    },
+    {
+      "epoch": 0.1701403658017865,
+      "grad_norm": 0.2990092253885084,
+      "learning_rate": 1.8992507361218743e-05,
+      "loss": 0.0765,
+      "step": 100
+    },
+    {
+      "epoch": 0.17184176945980434,
+      "grad_norm": 0.2177337033832474,
+      "learning_rate": 1.8968218608594987e-05,
+      "loss": 0.056,
+      "step": 101
+    },
+    {
+      "epoch": 0.1735431731178222,
+      "grad_norm": 0.39840165284276846,
+      "learning_rate": 1.8943656467437726e-05,
+      "loss": 0.1111,
+      "step": 102
+    },
+    {
+      "epoch": 0.17524457677584007,
+      "grad_norm": 0.3378745899319231,
+      "learning_rate": 1.8918821686502992e-05,
+      "loss": 0.0931,
+      "step": 103
+    },
+    {
+      "epoch": 0.17694598043385792,
+      "grad_norm": 0.2637083690262847,
+      "learning_rate": 1.8893715022858e-05,
+      "loss": 0.0759,
+      "step": 104
+    },
+    {
+      "epoch": 0.1786473840918758,
+      "grad_norm": 0.36645660481595893,
+      "learning_rate": 1.886833724185809e-05,
+      "loss": 0.0956,
+      "step": 105
+    },
+    {
+      "epoch": 0.18034878774989366,
+      "grad_norm": 0.2626991205572612,
+      "learning_rate": 1.8842689117123377e-05,
+      "loss": 0.0622,
+      "step": 106
+    },
+    {
+      "epoch": 0.18205019140791154,
+      "grad_norm": 0.28778996504448945,
+      "learning_rate": 1.8816771430515178e-05,
+      "loss": 0.0778,
+      "step": 107
+    },
+    {
+      "epoch": 0.1837515950659294,
+      "grad_norm": 0.30016452527135956,
+      "learning_rate": 1.8790584972112174e-05,
+      "loss": 0.0658,
+      "step": 108
+    },
+    {
+      "epoch": 0.18545299872394724,
+      "grad_norm": 0.5203118503020492,
+      "learning_rate": 1.876413054018633e-05,
+      "loss": 0.146,
+      "step": 109
+    },
+    {
+      "epoch": 0.18715440238196512,
+      "grad_norm": 0.30134320864935904,
+      "learning_rate": 1.873740894117854e-05,
+      "loss": 0.0821,
+      "step": 110
+    },
+    {
+      "epoch": 0.18885580603998298,
+      "grad_norm": 0.34344915555537253,
+      "learning_rate": 1.8710420989674093e-05,
+      "loss": 0.075,
+      "step": 111
+    },
+    {
+      "epoch": 0.19055720969800086,
+      "grad_norm": 0.2458501444323483,
+      "learning_rate": 1.8683167508377775e-05,
+      "loss": 0.0596,
+      "step": 112
+    },
+    {
+      "epoch": 0.1922586133560187,
+      "grad_norm": 0.2651178569441844,
+      "learning_rate": 1.8655649328088836e-05,
+      "loss": 0.0674,
+      "step": 113
+    },
+    {
+      "epoch": 0.1939600170140366,
+      "grad_norm": 0.2662036987842302,
+      "learning_rate": 1.862786728767565e-05,
+      "loss": 0.0769,
+      "step": 114
+    },
+    {
+      "epoch": 0.19566142067205444,
+      "grad_norm": 0.28498333925442115,
+      "learning_rate": 1.8599822234050143e-05,
+      "loss": 0.0778,
+      "step": 115
+    },
+    {
+      "epoch": 0.19736282433007232,
+      "grad_norm": 0.26717816913116776,
+      "learning_rate": 1.8571515022141974e-05,
+      "loss": 0.0896,
+      "step": 116
+    },
+    {
+      "epoch": 0.19906422798809018,
+      "grad_norm": 0.3374842493414297,
+      "learning_rate": 1.8542946514872478e-05,
+      "loss": 0.0992,
+      "step": 117
+    },
+    {
+      "epoch": 0.20076563164610803,
+      "grad_norm": 0.21117730752547184,
+      "learning_rate": 1.851411758312835e-05,
+      "loss": 0.0454,
+      "step": 118
+    },
+    {
+      "epoch": 0.2024670353041259,
+      "grad_norm": 0.19203553955386238,
+      "learning_rate": 1.8485029105735112e-05,
+      "loss": 0.0611,
+      "step": 119
+    },
+    {
+      "epoch": 0.20416843896214376,
+      "grad_norm": 0.22445524786962626,
+      "learning_rate": 1.8455681969430307e-05,
+      "loss": 0.0584,
+      "step": 120
+    },
+    {
+      "epoch": 0.20586984262016164,
+      "grad_norm": 0.23148081400071185,
+      "learning_rate": 1.8426077068836487e-05,
+      "loss": 0.0629,
+      "step": 121
+    },
+    {
+      "epoch": 0.2075712462781795,
+      "grad_norm": 0.17709972528549517,
+      "learning_rate": 1.839621530643392e-05,
+      "loss": 0.0563,
+      "step": 122
+    },
+    {
+      "epoch": 0.20927264993619737,
+      "grad_norm": 0.3048264391849215,
+      "learning_rate": 1.8366097592533095e-05,
+      "loss": 0.0778,
+      "step": 123
+    },
+    {
+      "epoch": 0.21097405359421523,
+      "grad_norm": 0.32208607157512736,
+      "learning_rate": 1.8335724845246948e-05,
+      "loss": 0.1028,
+      "step": 124
+    },
+    {
+      "epoch": 0.21267545725223308,
+      "grad_norm": 0.23725153079241285,
+      "learning_rate": 1.830509799046292e-05,
+      "loss": 0.0803,
+      "step": 125
+    },
+    {
+      "epoch": 0.21437686091025096,
+      "grad_norm": 0.2996136224560952,
+      "learning_rate": 1.8274217961814682e-05,
+      "loss": 0.0718,
+      "step": 126
+    },
+    {
+      "epoch": 0.2160782645682688,
+      "grad_norm": 0.23529202737787724,
+      "learning_rate": 1.8243085700653698e-05,
+      "loss": 0.058,
+      "step": 127
+    },
+    {
+      "epoch": 0.2177796682262867,
+      "grad_norm": 0.2602674833841576,
+      "learning_rate": 1.821170215602053e-05,
+      "loss": 0.0847,
+      "step": 128
+    },
+    {
+      "epoch": 0.21948107188430455,
+      "grad_norm": 0.23134307276310323,
+      "learning_rate": 1.818006828461591e-05,
+      "loss": 0.0568,
+      "step": 129
+    },
+    {
+      "epoch": 0.22118247554232243,
+      "grad_norm": 0.26453961985179086,
+      "learning_rate": 1.8148185050771554e-05,
+      "loss": 0.0801,
+      "step": 130
+    },
+    {
+      "epoch": 0.22288387920034028,
+      "grad_norm": 0.25318378680529585,
+      "learning_rate": 1.8116053426420793e-05,
+      "loss": 0.0749,
+      "step": 131
+    },
+    {
+      "epoch": 0.22458528285835813,
+      "grad_norm": 0.20502925194857724,
+      "learning_rate": 1.8083674391068925e-05,
+      "loss": 0.0629,
+      "step": 132
+    },
+    {
+      "epoch": 0.226286686516376,
+      "grad_norm": 0.19279005026370763,
+      "learning_rate": 1.8051048931763366e-05,
+      "loss": 0.0463,
+      "step": 133
+    },
+    {
+      "epoch": 0.22798809017439386,
+      "grad_norm": 0.2986931928862026,
+      "learning_rate": 1.8018178043063554e-05,
+      "loss": 0.0869,
+      "step": 134
+    },
+    {
+      "epoch": 0.22968949383241175,
+      "grad_norm": 0.2580936970306385,
+      "learning_rate": 1.798506272701064e-05,
+      "loss": 0.0689,
+      "step": 135
+    },
+    {
+      "epoch": 0.2313908974904296,
+      "grad_norm": 0.35433155424672486,
+      "learning_rate": 1.795170399309692e-05,
+      "loss": 0.0889,
+      "step": 136
+    },
+    {
+      "epoch": 0.23309230114844748,
+      "grad_norm": 0.3342563856245416,
+      "learning_rate": 1.7918102858235103e-05,
+      "loss": 0.1088,
+      "step": 137
+    },
+    {
+      "epoch": 0.23479370480646533,
+      "grad_norm": 0.2658901230535673,
+      "learning_rate": 1.7884260346727257e-05,
+      "loss": 0.0967,
+      "step": 138
+    },
+    {
+      "epoch": 0.2364951084644832,
+      "grad_norm": 0.3116747759735481,
+      "learning_rate": 1.7850177490233635e-05,
+      "loss": 0.0763,
+      "step": 139
+    },
+    {
+      "epoch": 0.23819651212250106,
+      "grad_norm": 0.2561525037232419,
+      "learning_rate": 1.7815855327741185e-05,
+      "loss": 0.0956,
+      "step": 140
+    },
+    {
+      "epoch": 0.23989791578051892,
+      "grad_norm": 0.2246820236163285,
+      "learning_rate": 1.7781294905531908e-05,
+      "loss": 0.0792,
+      "step": 141
+    },
+    {
+      "epoch": 0.2415993194385368,
+      "grad_norm": 0.3264214270885714,
+      "learning_rate": 1.774649727715094e-05,
+      "loss": 0.0792,
+      "step": 142
+    },
+    {
+      "epoch": 0.24330072309655465,
+      "grad_norm": 0.19102790309255646,
+      "learning_rate": 1.7711463503374466e-05,
+      "loss": 0.0627,
+      "step": 143
+    },
+    {
+      "epoch": 0.24500212675457253,
+      "grad_norm": 0.29766375426258634,
+      "learning_rate": 1.7676194652177333e-05,
+      "loss": 0.0529,
+      "step": 144
+    },
+    {
+      "epoch": 0.24670353041259038,
+      "grad_norm": 0.29002494096896503,
+      "learning_rate": 1.764069179870055e-05,
+      "loss": 0.0873,
+      "step": 145
+    },
+    {
+      "epoch": 0.24840493407060826,
+      "grad_norm": 0.1961062743167957,
+      "learning_rate": 1.760495602521847e-05,
+      "loss": 0.0549,
+      "step": 146
+    },
+    {
+      "epoch": 0.2501063377286261,
+      "grad_norm": 0.3391386450150704,
+      "learning_rate": 1.756898842110582e-05,
+      "loss": 0.0855,
+      "step": 147
+    },
+    {
+      "epoch": 0.25180774138664397,
+      "grad_norm": 0.2237841919957392,
+      "learning_rate": 1.753279008280449e-05,
+      "loss": 0.0611,
+      "step": 148
+    },
+    {
+      "epoch": 0.2535091450446618,
+      "grad_norm": 0.1863997964345379,
+      "learning_rate": 1.74963621137901e-05,
+      "loss": 0.0557,
+      "step": 149
+    },
+    {
+      "epoch": 0.25521054870267973,
+      "grad_norm": 0.2322748935860178,
+      "learning_rate": 1.7459705624538383e-05,
+      "loss": 0.0744,
+      "step": 150
+    },
+    {
+      "epoch": 0.2569119523606976,
+      "grad_norm": 0.2811884338929891,
+      "learning_rate": 1.7422821732491297e-05,
+      "loss": 0.0869,
+      "step": 151
+    },
+    {
+      "epoch": 0.25861335601871543,
+      "grad_norm": 0.2336367022405289,
+      "learning_rate": 1.7385711562022988e-05,
+      "loss": 0.0662,
+      "step": 152
+    },
+    {
+      "epoch": 0.2603147596767333,
+      "grad_norm": 0.2525517790918003,
+      "learning_rate": 1.734837624440551e-05,
+      "loss": 0.0709,
+      "step": 153
+    },
+    {
+      "epoch": 0.2620161633347512,
+      "grad_norm": 0.25674452456443186,
+      "learning_rate": 1.731081691777434e-05,
+      "loss": 0.0536,
+      "step": 154
+    },
+    {
+      "epoch": 0.26371756699276905,
+      "grad_norm": 0.2380952734859674,
+      "learning_rate": 1.7273034727093677e-05,
+      "loss": 0.0797,
+      "step": 155
+    },
+    {
+      "epoch": 0.2654189706507869,
+      "grad_norm": 0.22013348660975807,
+      "learning_rate": 1.7235030824121542e-05,
+      "loss": 0.0608,
+      "step": 156
+    },
+    {
+      "epoch": 0.26712037430880475,
+      "grad_norm": 0.22281523456342697,
+      "learning_rate": 1.7196806367374656e-05,
+      "loss": 0.0635,
+      "step": 157
+    },
+    {
+      "epoch": 0.2688217779668226,
+      "grad_norm": 0.2102268959812806,
+      "learning_rate": 1.7158362522093153e-05,
+      "loss": 0.0682,
+      "step": 158
+    },
+    {
+      "epoch": 0.2705231816248405,
+      "grad_norm": 0.2981020750225053,
+      "learning_rate": 1.7119700460205026e-05,
+      "loss": 0.0748,
+      "step": 159
+    },
+    {
+      "epoch": 0.27222458528285837,
+      "grad_norm": 0.23396857925074813,
+      "learning_rate": 1.7080821360290426e-05,
+      "loss": 0.0668,
+      "step": 160
+    },
+    {
+      "epoch": 0.2739259889408762,
+      "grad_norm": 0.14039625188691493,
+      "learning_rate": 1.7041726407545716e-05,
+      "loss": 0.0334,
+      "step": 161
+    },
+    {
+      "epoch": 0.27562739259889407,
+      "grad_norm": 0.2591445596342466,
+      "learning_rate": 1.7002416793747354e-05,
+      "loss": 0.0613,
+      "step": 162
+    },
+    {
+      "epoch": 0.2773287962569119,
+      "grad_norm": 0.29367133177482335,
+      "learning_rate": 1.696289371721556e-05,
+      "loss": 0.0713,
+      "step": 163
+    },
+    {
+      "epoch": 0.27903019991492983,
+      "grad_norm": 0.3690305678977906,
+      "learning_rate": 1.692315838277778e-05,
+      "loss": 0.1274,
+      "step": 164
+    },
+    {
+      "epoch": 0.2807316035729477,
+      "grad_norm": 0.22214467316639894,
+      "learning_rate": 1.6883212001731956e-05,
+      "loss": 0.0655,
+      "step": 165
+    },
+    {
+      "epoch": 0.28243300723096554,
+      "grad_norm": 0.1574976967869379,
+      "learning_rate": 1.6843055791809623e-05,
+      "loss": 0.0322,
+      "step": 166
+    },
+    {
+      "epoch": 0.2841344108889834,
+      "grad_norm": 0.27654091402143877,
+      "learning_rate": 1.680269097713876e-05,
+      "loss": 0.0922,
+      "step": 167
+    },
+    {
+      "epoch": 0.2858358145470013,
+      "grad_norm": 0.28597652739783896,
+      "learning_rate": 1.6762118788206488e-05,
+      "loss": 0.079,
+      "step": 168
+    },
+    {
+      "epoch": 0.28753721820501915,
+      "grad_norm": 0.26938794460799176,
+      "learning_rate": 1.6721340461821555e-05,
+      "loss": 0.0871,
+      "step": 169
+    },
+    {
+      "epoch": 0.289238621863037,
+      "grad_norm": 0.32524563962835895,
+      "learning_rate": 1.6680357241076632e-05,
+      "loss": 0.1052,
+      "step": 170
+    },
+    {
+      "epoch": 0.29094002552105486,
+      "grad_norm": 0.3017728585235927,
+      "learning_rate": 1.6639170375310422e-05,
+      "loss": 0.0705,
+      "step": 171
+    },
+    {
+      "epoch": 0.2926414291790727,
+      "grad_norm": 0.23611080555948813,
+      "learning_rate": 1.6597781120069584e-05,
+      "loss": 0.0735,
+      "step": 172
+    },
+    {
+      "epoch": 0.2943428328370906,
+      "grad_norm": 0.2824684340323033,
+      "learning_rate": 1.655619073707043e-05,
+      "loss": 0.0957,
+      "step": 173
+    },
+    {
+      "epoch": 0.29604423649510847,
+      "grad_norm": 0.22439711399094156,
+      "learning_rate": 1.6514400494160498e-05,
+      "loss": 0.0572,
+      "step": 174
+    },
+    {
+      "epoch": 0.2977456401531263,
+      "grad_norm": 0.18939390048237154,
+      "learning_rate": 1.6472411665279872e-05,
+      "loss": 0.057,
+      "step": 175
+    },
+    {
+      "epoch": 0.2994470438111442,
+      "grad_norm": 0.22456676896884215,
+      "learning_rate": 1.643022553042237e-05,
+      "loss": 0.0557,
+      "step": 176
+    },
+    {
+      "epoch": 0.3011484474691621,
+      "grad_norm": 0.18664835436430186,
+      "learning_rate": 1.6387843375596513e-05,
+      "loss": 0.0494,
+      "step": 177
+    },
+    {
+      "epoch": 0.30284985112717994,
+      "grad_norm": 0.23489112753979385,
+      "learning_rate": 1.634526649278632e-05,
+      "loss": 0.0821,
+      "step": 178
+    },
+    {
+      "epoch": 0.3045512547851978,
+      "grad_norm": 0.2194109402248629,
+      "learning_rate": 1.630249617991194e-05,
+      "loss": 0.0672,
+      "step": 179
+    },
+    {
+      "epoch": 0.30625265844321564,
+      "grad_norm": 0.3321350768766143,
+      "learning_rate": 1.6259533740790055e-05,
+      "loss": 0.1135,
+      "step": 180
+    },
+    {
+      "epoch": 0.3079540621012335,
+      "grad_norm": 0.2502151994073321,
+      "learning_rate": 1.6216380485094164e-05,
+      "loss": 0.0718,
+      "step": 181
+    },
+    {
+      "epoch": 0.3096554657592514,
+      "grad_norm": 0.2670386782819874,
+      "learning_rate": 1.617303772831465e-05,
+      "loss": 0.0728,
+      "step": 182
+    },
+    {
+      "epoch": 0.31135686941726926,
+      "grad_norm": 0.26272098469205685,
+      "learning_rate": 1.6129506791718665e-05,
+      "loss": 0.0832,
+      "step": 183
+    },
+    {
+      "epoch": 0.3130582730752871,
+      "grad_norm": 0.2246929104575992,
+      "learning_rate": 1.6085789002309873e-05,
+      "loss": 0.0631,
+      "step": 184
+    },
+    {
+      "epoch": 0.31475967673330496,
+      "grad_norm": 0.21270610161779943,
+      "learning_rate": 1.6041885692787985e-05,
+      "loss": 0.067,
+      "step": 185
+    },
+    {
+      "epoch": 0.31646108039132287,
+      "grad_norm": 0.21242584741643203,
+      "learning_rate": 1.599779820150813e-05,
+      "loss": 0.068,
+      "step": 186
+    },
+    {
+      "epoch": 0.3181624840493407,
+      "grad_norm": 0.20419485410339863,
+      "learning_rate": 1.5953527872440063e-05,
+      "loss": 0.0757,
+      "step": 187
+    },
+    {
+      "epoch": 0.3198638877073586,
+      "grad_norm": 0.27210472135834674,
+      "learning_rate": 1.5909076055127202e-05,
+      "loss": 0.0853,
+      "step": 188
+    },
+    {
+      "epoch": 0.3215652913653764,
+      "grad_norm": 0.22338089059411625,
+      "learning_rate": 1.5864444104645473e-05,
+      "loss": 0.0742,
+      "step": 189
+    },
+    {
+      "epoch": 0.3232666950233943,
+      "grad_norm": 0.2732168335391855,
+      "learning_rate": 1.581963338156201e-05,
+      "loss": 0.0908,
+      "step": 190
+    },
+    {
+      "epoch": 0.3249680986814122,
+      "grad_norm": 0.2467079865485088,
+      "learning_rate": 1.5774645251893673e-05,
+      "loss": 0.0591,
+      "step": 191
+    },
+    {
+      "epoch": 0.32666950233943004,
+      "grad_norm": 0.2313614767876932,
+      "learning_rate": 1.5729481087065423e-05,
+      "loss": 0.0892,
+      "step": 192
+    },
+    {
+      "epoch": 0.3283709059974479,
+      "grad_norm": 0.2023763116089166,
+      "learning_rate": 1.5684142263868493e-05,
+      "loss": 0.0672,
+      "step": 193
+    },
+    {
+      "epoch": 0.33007230965546575,
+      "grad_norm": 0.31753011619888016,
+      "learning_rate": 1.5638630164418435e-05,
+      "loss": 0.0712,
+      "step": 194
+    },
+    {
+      "epoch": 0.3317737133134836,
+      "grad_norm": 0.2757080957715778,
+      "learning_rate": 1.5592946176112973e-05,
+      "loss": 0.1015,
+      "step": 195
+    },
+    {
+      "epoch": 0.3334751169715015,
+      "grad_norm": 0.2505973926615177,
+      "learning_rate": 1.554709169158972e-05,
+      "loss": 0.0971,
+      "step": 196
+    },
+    {
+      "epoch": 0.33517652062951936,
+      "grad_norm": 0.2226907461950715,
+      "learning_rate": 1.550106810868373e-05,
+      "loss": 0.0513,
+      "step": 197
+    },
+    {
+      "epoch": 0.3368779242875372,
+      "grad_norm": 0.15446367505194006,
+      "learning_rate": 1.5454876830384868e-05,
+      "loss": 0.057,
+      "step": 198
+    },
+    {
+      "epoch": 0.33857932794555506,
+      "grad_norm": 0.30643512505114406,
+      "learning_rate": 1.540851926479505e-05,
+      "loss": 0.0975,
+      "step": 199
+    },
+    {
+      "epoch": 0.340280731603573,
+      "grad_norm": 0.25762810344960363,
+      "learning_rate": 1.536199682508533e-05,
+      "loss": 0.0633,
+      "step": 200
+    },
+    {
+      "epoch": 0.3419821352615908,
+      "grad_norm": 0.19158691969988545,
+      "learning_rate": 1.531531092945279e-05,
+      "loss": 0.0569,
+      "step": 201
+    },
+    {
+      "epoch": 0.3436835389196087,
+      "grad_norm": 0.2834266199692826,
+      "learning_rate": 1.526846300107734e-05,
+      "loss": 0.0988,
+      "step": 202
+    },
+    {
+      "epoch": 0.34538494257762653,
+      "grad_norm": 0.2556087083863337,
+      "learning_rate": 1.5221454468078336e-05,
+      "loss": 0.0689,
+      "step": 203
+    },
+    {
+      "epoch": 0.3470863462356444,
+      "grad_norm": 0.2438742301334132,
+      "learning_rate": 1.5174286763470995e-05,
+      "loss": 0.0715,
+      "step": 204
+    },
+    {
+      "epoch": 0.3487877498936623,
+      "grad_norm": 0.21904569801568857,
+      "learning_rate": 1.5126961325122773e-05,
+      "loss": 0.0715,
+      "step": 205
+    },
+    {
+      "epoch": 0.35048915355168014,
+      "grad_norm": 0.32118887699890014,
+      "learning_rate": 1.5079479595709493e-05,
+      "loss": 0.1042,
+      "step": 206
+    },
+    {
+      "epoch": 0.352190557209698,
+      "grad_norm": 0.22503680461445366,
+      "learning_rate": 1.5031843022671377e-05,
+      "loss": 0.0516,
+      "step": 207
+    },
+    {
+      "epoch": 0.35389196086771585,
+      "grad_norm": 0.20476968089029912,
+      "learning_rate": 1.4984053058168936e-05,
+      "loss": 0.0651,
+      "step": 208
+    },
+    {
+      "epoch": 0.35559336452573376,
+      "grad_norm": 0.26260720806295024,
+      "learning_rate": 1.4936111159038677e-05,
+      "loss": 0.078,
+      "step": 209
+    },
+    {
+      "epoch": 0.3572947681837516,
+      "grad_norm": 0.21335908324799727,
+      "learning_rate": 1.4888018786748713e-05,
+      "loss": 0.0531,
+      "step": 210
+    },
+    {
+      "epoch": 0.35899617184176946,
+      "grad_norm": 0.24776977384172608,
+      "learning_rate": 1.4839777407354194e-05,
+      "loss": 0.0711,
+      "step": 211
+    },
+    {
+      "epoch": 0.3606975754997873,
+      "grad_norm": 0.26283265478228607,
+      "learning_rate": 1.4791388491452637e-05,
+      "loss": 0.1005,
+      "step": 212
+    },
+    {
+      "epoch": 0.36239897915780517,
+      "grad_norm": 0.21912083728071574,
+      "learning_rate": 1.4742853514139076e-05,
+      "loss": 0.0759,
+      "step": 213
+    },
+    {
+      "epoch": 0.3641003828158231,
+      "grad_norm": 0.30538258609344554,
+      "learning_rate": 1.4694173954961105e-05,
+      "loss": 0.1009,
+      "step": 214
+    },
+    {
+      "epoch": 0.36580178647384093,
+      "grad_norm": 0.26764402263401943,
+      "learning_rate": 1.4645351297873774e-05,
+      "loss": 0.0792,
+      "step": 215
+    },
+    {
+      "epoch": 0.3675031901318588,
+      "grad_norm": 0.23675986028315632,
+      "learning_rate": 1.4596387031194354e-05,
+      "loss": 0.0743,
+      "step": 216
+    },
+    {
+      "epoch": 0.36920459378987663,
+      "grad_norm": 0.223485373234782,
+      "learning_rate": 1.4547282647556964e-05,
+      "loss": 0.0929,
+      "step": 217
+    },
+    {
+      "epoch": 0.3709059974478945,
+      "grad_norm": 0.29133370479541304,
+      "learning_rate": 1.449803964386706e-05,
+      "loss": 0.0798,
+      "step": 218
+    },
+    {
+      "epoch": 0.3726074011059124,
+      "grad_norm": 0.18643880537818264,
+      "learning_rate": 1.4448659521255823e-05,
+      "loss": 0.0569,
+      "step": 219
+    },
+    {
+      "epoch": 0.37430880476393025,
+      "grad_norm": 0.280997156813028,
+      "learning_rate": 1.4399143785034388e-05,
+      "loss": 0.0999,
+      "step": 220
+    },
+    {
+      "epoch": 0.3760102084219481,
+      "grad_norm": 0.24715686808729184,
+      "learning_rate": 1.4349493944647953e-05,
+      "loss": 0.0627,
+      "step": 221
+    },
+    {
+      "epoch": 0.37771161207996595,
+      "grad_norm": 0.2653638113311677,
+      "learning_rate": 1.4299711513629759e-05,
+      "loss": 0.0863,
+      "step": 222
+    },
+    {
+      "epoch": 0.37941301573798386,
+      "grad_norm": 0.3152170570199493,
+      "learning_rate": 1.4249798009554979e-05,
+      "loss": 0.0962,
+      "step": 223
+    },
+    {
+      "epoch": 0.3811144193960017,
+      "grad_norm": 0.25435032990077316,
+      "learning_rate": 1.419975495399442e-05,
+      "loss": 0.0937,
+      "step": 224
+    },
+    {
+      "epoch": 0.38281582305401957,
+      "grad_norm": 0.15047418159614007,
+      "learning_rate": 1.4149583872468165e-05,
+      "loss": 0.0482,
+      "step": 225
+    },
+    {
+      "epoch": 0.3845172267120374,
+      "grad_norm": 0.15295980990137148,
+      "learning_rate": 1.4099286294399051e-05,
+      "loss": 0.0382,
+      "step": 226
+    },
+    {
+      "epoch": 0.38621863037005527,
+      "grad_norm": 0.2533178513009288,
+      "learning_rate": 1.404886375306607e-05,
+      "loss": 0.0948,
+      "step": 227
+    },
+    {
+      "epoch": 0.3879200340280732,
+      "grad_norm": 0.23107415073266793,
+      "learning_rate": 1.3998317785557597e-05,
+      "loss": 0.0556,
+      "step": 228
+    },
+    {
+      "epoch": 0.38962143768609103,
+      "grad_norm": 0.22917339521301278,
+      "learning_rate": 1.3947649932724563e-05,
+      "loss": 0.0843,
+      "step": 229
+    },
+    {
+      "epoch": 0.3913228413441089,
+      "grad_norm": 0.17860874606051796,
+      "learning_rate": 1.3896861739133456e-05,
+      "loss": 0.0488,
+      "step": 230
+    },
+    {
+      "epoch": 0.39302424500212674,
+      "grad_norm": 0.20110618464098665,
+      "learning_rate": 1.384595475301926e-05,
+      "loss": 0.0707,
+      "step": 231
+    },
+    {
+      "epoch": 0.39472564866014465,
+      "grad_norm": 0.20490708885008196,
+      "learning_rate": 1.3794930526238246e-05,
+      "loss": 0.0638,
+      "step": 232
+    },
+    {
+      "epoch": 0.3964270523181625,
+      "grad_norm": 0.23510478688343137,
+      "learning_rate": 1.3743790614220664e-05,
+      "loss": 0.0795,
+      "step": 233
+    },
+    {
+      "epoch": 0.39812845597618035,
+      "grad_norm": 0.23100407798838418,
+      "learning_rate": 1.3692536575923334e-05,
+      "loss": 0.0761,
+      "step": 234
+    },
+    {
+      "epoch": 0.3998298596341982,
+      "grad_norm": 0.256927774293719,
+      "learning_rate": 1.3641169973782117e-05,
+      "loss": 0.0669,
+      "step": 235
+    },
+    {
+      "epoch": 0.40153126329221606,
+      "grad_norm": 0.265960997604412,
+      "learning_rate": 1.3589692373664288e-05,
+      "loss": 0.0792,
+      "step": 236
+    },
+    {
+      "epoch": 0.40323266695023396,
+      "grad_norm": 0.2915525993978868,
+      "learning_rate": 1.3538105344820798e-05,
+      "loss": 0.0995,
+      "step": 237
+    },
+    {
+      "epoch": 0.4049340706082518,
+      "grad_norm": 0.2578778351844946,
+      "learning_rate": 1.3486410459838448e-05,
+      "loss": 0.0719,
+      "step": 238
+    },
+    {
+      "epoch": 0.40663547426626967,
+      "grad_norm": 0.23384387017059058,
+      "learning_rate": 1.343460929459193e-05,
+      "loss": 0.0712,
+      "step": 239
+    },
+    {
+      "epoch": 0.4083368779242875,
+      "grad_norm": 0.3019784671101925,
+      "learning_rate": 1.3382703428195812e-05,
+      "loss": 0.1115,
+      "step": 240
+    },
+    {
+      "epoch": 0.4100382815823054,
+      "grad_norm": 0.1352734841435768,
+      "learning_rate": 1.3330694442956376e-05,
+      "loss": 0.0464,
+      "step": 241
+    },
+    {
+      "epoch": 0.4117396852403233,
+      "grad_norm": 0.17202787234184866,
+      "learning_rate": 1.3278583924323405e-05,
+      "loss": 0.0454,
+      "step": 242
+    },
+    {
+      "epoch": 0.41344108889834114,
+      "grad_norm": 0.197390798554472,
+      "learning_rate": 1.3226373460841835e-05,
+      "loss": 0.0643,
+      "step": 243
+    },
+    {
+      "epoch": 0.415142492556359,
+      "grad_norm": 0.21059005106394152,
+      "learning_rate": 1.3174064644103334e-05,
+      "loss": 0.0619,
+      "step": 244
+    },
+    {
+      "epoch": 0.41684389621437684,
+      "grad_norm": 0.1722308196492277,
+      "learning_rate": 1.3121659068697797e-05,
+      "loss": 0.0454,
+      "step": 245
+    },
+    {
+      "epoch": 0.41854529987239475,
+      "grad_norm": 0.32958443926088915,
+      "learning_rate": 1.306915833216471e-05,
+      "loss": 0.1037,
+      "step": 246
+    },
+    {
+      "epoch": 0.4202467035304126,
+      "grad_norm": 0.2544357662171513,
+      "learning_rate": 1.3016564034944473e-05,
+      "loss": 0.0706,
+      "step": 247
+    },
+    {
+      "epoch": 0.42194810718843045,
+      "grad_norm": 0.2585510606952109,
+      "learning_rate": 1.29638777803296e-05,
+      "loss": 0.0825,
+      "step": 248
+    },
+    {
+      "epoch": 0.4236495108464483,
+      "grad_norm": 0.20275321109441732,
+      "learning_rate": 1.2911101174415861e-05,
+      "loss": 0.0526,
+      "step": 249
+    },
+    {
+      "epoch": 0.42535091450446616,
+      "grad_norm": 0.233981891948638,
+      "learning_rate": 1.2858235826053294e-05,
+      "loss": 0.0695,
+      "step": 250
+    },
+    {
+      "epoch": 0.42705231816248407,
+      "grad_norm": 0.21493578674491315,
+      "learning_rate": 1.2805283346797179e-05,
+      "loss": 0.0653,
+      "step": 251
+    },
+    {
+      "epoch": 0.4287537218205019,
+      "grad_norm": 0.24141724836500014,
+      "learning_rate": 1.2752245350858905e-05,
+      "loss": 0.0797,
+      "step": 252
+    },
+    {
+      "epoch": 0.4304551254785198,
+      "grad_norm": 0.1489771140649588,
+      "learning_rate": 1.2699123455056777e-05,
+      "loss": 0.03,
+      "step": 253
+    },
+    {
+      "epoch": 0.4321565291365376,
+      "grad_norm": 0.24828101075324488,
+      "learning_rate": 1.26459192787667e-05,
+      "loss": 0.0819,
+      "step": 254
+    },
+    {
+      "epoch": 0.43385793279455553,
+      "grad_norm": 0.17372013690514643,
+      "learning_rate": 1.2592634443872842e-05,
+      "loss": 0.0461,
+      "step": 255
+    },
+    {
+      "epoch": 0.4355593364525734,
+      "grad_norm": 0.2764346314356569,
+      "learning_rate": 1.2539270574718172e-05,
+      "loss": 0.0806,
+      "step": 256
+    },
+    {
+      "epoch": 0.43726074011059124,
+      "grad_norm": 0.29987546473911214,
+      "learning_rate": 1.2485829298054952e-05,
+      "loss": 0.0846,
+      "step": 257
+    },
+    {
+      "epoch": 0.4389621437686091,
+      "grad_norm": 0.31175687442320515,
+      "learning_rate": 1.2432312242995158e-05,
+      "loss": 0.0971,
+      "step": 258
+    },
+    {
+      "epoch": 0.44066354742662694,
+      "grad_norm": 0.3009904680059143,
+      "learning_rate": 1.2378721040960788e-05,
+      "loss": 0.0994,
+      "step": 259
+    },
+    {
+      "epoch": 0.44236495108464485,
+      "grad_norm": 0.2562384582969849,
+      "learning_rate": 1.232505732563416e-05,
+      "loss": 0.0759,
+      "step": 260
+    },
+    {
+      "epoch": 0.4440663547426627,
+      "grad_norm": 0.2550190452410635,
+      "learning_rate": 1.2271322732908091e-05,
+      "loss": 0.0733,
+      "step": 261
+    },
+    {
+      "epoch": 0.44576775840068056,
+      "grad_norm": 0.2515270792806656,
+      "learning_rate": 1.2217518900836045e-05,
+      "loss": 0.0708,
+      "step": 262
+    },
+    {
+      "epoch": 0.4474691620586984,
+      "grad_norm": 0.18357019713578807,
+      "learning_rate": 1.2163647469582181e-05,
+      "loss": 0.0515,
+      "step": 263
+    },
+    {
+      "epoch": 0.44917056571671626,
+      "grad_norm": 0.2671034404389676,
+      "learning_rate": 1.210971008137136e-05,
+      "loss": 0.0825,
+      "step": 264
+    },
+    {
+      "epoch": 0.45087196937473417,
+      "grad_norm": 0.2571681277129728,
+      "learning_rate": 1.2055708380439089e-05,
+      "loss": 0.1042,
+      "step": 265
+    },
+    {
+      "epoch": 0.452573373032752,
+      "grad_norm": 0.21793020282041717,
+      "learning_rate": 1.2001644012981392e-05,
+      "loss": 0.0672,
+      "step": 266
+    },
+    {
+      "epoch": 0.4542747766907699,
+      "grad_norm": 0.3515840793799933,
+      "learning_rate": 1.1947518627104637e-05,
+      "loss": 0.1232,
+      "step": 267
+    },
+    {
+      "epoch": 0.45597618034878773,
+      "grad_norm": 0.22270456303325817,
+      "learning_rate": 1.1893333872775275e-05,
+      "loss": 0.084,
+      "step": 268
+    },
+    {
+      "epoch": 0.45767758400680564,
+      "grad_norm": 0.1770352144116934,
+      "learning_rate": 1.1839091401769559e-05,
+      "loss": 0.0435,
+      "step": 269
+    },
+    {
+      "epoch": 0.4593789876648235,
+      "grad_norm": 0.19476419487806804,
+      "learning_rate": 1.1784792867623179e-05,
+      "loss": 0.0535,
+      "step": 270
+    },
+    {
+      "epoch": 0.46108039132284134,
+      "grad_norm": 0.2961488962332661,
+      "learning_rate": 1.1730439925580876e-05,
+      "loss": 0.1054,
+      "step": 271
+    },
+    {
+      "epoch": 0.4627817949808592,
+      "grad_norm": 0.23209294477341702,
+      "learning_rate": 1.1676034232545963e-05,
+      "loss": 0.0898,
+      "step": 272
+    },
+    {
+      "epoch": 0.46448319863887705,
+      "grad_norm": 0.3192995286716727,
+      "learning_rate": 1.1621577447029816e-05,
+      "loss": 0.0864,
+      "step": 273
+    },
+    {
+      "epoch": 0.46618460229689496,
+      "grad_norm": 0.22033834311649156,
+      "learning_rate": 1.1567071229101332e-05,
+      "loss": 0.061,
+      "step": 274
+    },
+    {
+      "epoch": 0.4678860059549128,
+      "grad_norm": 0.24431011933295152,
+      "learning_rate": 1.1512517240336304e-05,
+      "loss": 0.05,
+      "step": 275
+    },
+    {
+      "epoch": 0.46958740961293066,
+      "grad_norm": 0.2501002015877452,
+      "learning_rate": 1.1457917143766786e-05,
+      "loss": 0.0811,
+      "step": 276
+    },
+    {
+      "epoch": 0.4712888132709485,
+      "grad_norm": 0.15591637050035256,
+      "learning_rate": 1.1403272603830384e-05,
+      "loss": 0.0439,
+      "step": 277
+    },
+    {
+      "epoch": 0.4729902169289664,
+      "grad_norm": 0.23445891677475122,
+      "learning_rate": 1.1348585286319529e-05,
+      "loss": 0.0562,
+      "step": 278
+    },
+    {
+      "epoch": 0.4746916205869843,
+      "grad_norm": 0.22953431606642624,
+      "learning_rate": 1.1293856858330678e-05,
+      "loss": 0.0712,
+      "step": 279
+    },
+    {
+      "epoch": 0.47639302424500213,
+      "grad_norm": 0.24410355979016798,
+      "learning_rate": 1.1239088988213522e-05,
+      "loss": 0.0652,
+      "step": 280
+    },
+    {
+      "epoch": 0.47809442790302,
+      "grad_norm": 0.18183702432279936,
+      "learning_rate": 1.11842833455201e-05,
+      "loss": 0.0464,
+      "step": 281
+    },
+    {
+      "epoch": 0.47979583156103783,
+      "grad_norm": 0.3068246013758465,
+      "learning_rate": 1.1129441600953916e-05,
+      "loss": 0.101,
+      "step": 282
+    },
+    {
+      "epoch": 0.48149723521905574,
+      "grad_norm": 0.2725071360381625,
+      "learning_rate": 1.1074565426319014e-05,
+      "loss": 0.0906,
+      "step": 283
+    },
+    {
+      "epoch": 0.4831986388770736,
+      "grad_norm": 0.23078769715046143,
+      "learning_rate": 1.101965649446901e-05,
+      "loss": 0.0659,
+      "step": 284
+    },
+    {
+      "epoch": 0.48490004253509145,
+      "grad_norm": 0.24607556757801813,
+      "learning_rate": 1.0964716479256094e-05,
+      "loss": 0.0581,
+      "step": 285
+    },
+    {
+      "epoch": 0.4866014461931093,
+      "grad_norm": 0.2039783514676341,
+      "learning_rate": 1.0909747055480004e-05,
+      "loss": 0.042,
+      "step": 286
+    },
+    {
+      "epoch": 0.4883028498511272,
+      "grad_norm": 0.2132647216715679,
+      "learning_rate": 1.0854749898836974e-05,
+      "loss": 0.042,
+      "step": 287
+    },
+    {
+      "epoch": 0.49000425350914506,
+      "grad_norm": 0.17146104368359683,
+      "learning_rate": 1.0799726685868648e-05,
+      "loss": 0.0486,
+      "step": 288
+    },
+    {
+      "epoch": 0.4917056571671629,
+      "grad_norm": 0.3101626941591736,
+      "learning_rate": 1.0744679093910987e-05,
+      "loss": 0.0855,
+      "step": 289
+    },
+    {
+      "epoch": 0.49340706082518077,
+      "grad_norm": 0.26359884054130955,
+      "learning_rate": 1.0689608801043107e-05,
+      "loss": 0.0671,
+      "step": 290
+    },
+    {
+      "epoch": 0.4951084644831986,
+      "grad_norm": 0.21556126895353891,
+      "learning_rate": 1.063451748603616e-05,
+      "loss": 0.076,
+      "step": 291
+    },
+    {
+      "epoch": 0.4968098681412165,
+      "grad_norm": 0.27181671931172413,
+      "learning_rate": 1.0579406828302124e-05,
+      "loss": 0.0847,
+      "step": 292
+    },
+    {
+      "epoch": 0.4985112717992344,
+      "grad_norm": 0.31490941982013704,
+      "learning_rate": 1.0524278507842637e-05,
+      "loss": 0.1254,
+      "step": 293
+    },
+    {
+      "epoch": 0.5002126754572522,
+      "grad_norm": 0.28738576717755915,
+      "learning_rate": 1.0469134205197762e-05,
+      "loss": 0.0741,
+      "step": 294
+    },
+    {
+      "epoch": 0.5019140791152701,
+      "grad_norm": 0.24967744618311208,
+      "learning_rate": 1.0413975601394765e-05,
+      "loss": 0.0952,
+      "step": 295
+    },
+    {
+      "epoch": 0.5036154827732879,
+      "grad_norm": 0.18653535652798736,
+      "learning_rate": 1.0358804377896876e-05,
+      "loss": 0.0666,
+      "step": 296
+    },
+    {
+      "epoch": 0.5053168864313058,
+      "grad_norm": 0.31524703768298645,
+      "learning_rate": 1.0303622216552022e-05,
+      "loss": 0.0821,
+      "step": 297
+    },
+    {
+      "epoch": 0.5070182900893236,
+      "grad_norm": 0.19693363449938797,
+      "learning_rate": 1.0248430799541564e-05,
+      "loss": 0.0486,
+      "step": 298
+    },
+    {
+      "epoch": 0.5087196937473416,
+      "grad_norm": 0.3100958390915376,
+      "learning_rate": 1.019323180932901e-05,
+      "loss": 0.075,
+      "step": 299
+    },
+    {
+      "epoch": 0.5104210974053595,
+      "grad_norm": 0.2263867395185222,
+      "learning_rate": 1.013802692860873e-05,
+      "loss": 0.0729,
+      "step": 300
+    },
+    {
+      "epoch": 0.5121225010633773,
+      "grad_norm": 0.30499024558777427,
+      "learning_rate": 1.0082817840254667e-05,
+      "loss": 0.0949,
+      "step": 301
+    },
+    {
+      "epoch": 0.5138239047213952,
+      "grad_norm": 0.27412297007732506,
+      "learning_rate": 1.0027606227269026e-05,
+      "loss": 0.0711,
+      "step": 302
+    },
+    {
+      "epoch": 0.515525308379413,
+      "grad_norm": 0.2236619712267448,
+      "learning_rate": 9.972393772730975e-06,
+      "loss": 0.0711,
+      "step": 303
+    },
+    {
+      "epoch": 0.5172267120374309,
+      "grad_norm": 0.21645526531787626,
+      "learning_rate": 9.917182159745335e-06,
+      "loss": 0.0696,
+      "step": 304
+    },
+    {
+      "epoch": 0.5189281156954487,
+      "grad_norm": 0.19297627616781193,
+      "learning_rate": 9.861973071391272e-06,
+      "loss": 0.0723,
+      "step": 305
+    },
+    {
+      "epoch": 0.5206295193534666,
+      "grad_norm": 0.17935274621615926,
+      "learning_rate": 9.806768190670994e-06,
+      "loss": 0.0603,
+      "step": 306
+    },
+    {
+      "epoch": 0.5223309230114844,
+      "grad_norm": 0.36516731843883593,
+      "learning_rate": 9.751569200458438e-06,
+      "loss": 0.1183,
+      "step": 307
+    },
+    {
+      "epoch": 0.5240323266695024,
+      "grad_norm": 0.2666543854374252,
+      "learning_rate": 9.69637778344798e-06,
+      "loss": 0.0683,
+      "step": 308
+    },
+    {
+      "epoch": 0.5257337303275202,
+      "grad_norm": 0.1559589176353223,
+      "learning_rate": 9.641195622103126e-06,
+      "loss": 0.0457,
+      "step": 309
+    },
+    {
+      "epoch": 0.5274351339855381,
+      "grad_norm": 0.21822959242881637,
+      "learning_rate": 9.586024398605238e-06,
+      "loss": 0.0728,
+      "step": 310
+    },
+    {
+      "epoch": 0.529136537643556,
+      "grad_norm": 0.1910067933489864,
+      "learning_rate": 9.530865794802243e-06,
+      "loss": 0.0518,
+      "step": 311
+    },
+    {
+      "epoch": 0.5308379413015738,
+      "grad_norm": 0.2286981911174439,
+      "learning_rate": 9.475721492157365e-06,
+      "loss": 0.0538,
+      "step": 312
+    },
+    {
+      "epoch": 0.5325393449595917,
+      "grad_norm": 0.2649695673029832,
+      "learning_rate": 9.420593171697876e-06,
+      "loss": 0.086,
+      "step": 313
+    },
+    {
+      "epoch": 0.5342407486176095,
+      "grad_norm": 0.2420342613285877,
+      "learning_rate": 9.365482513963844e-06,
+      "loss": 0.0972,
+      "step": 314
+    },
+    {
+      "epoch": 0.5359421522756274,
+      "grad_norm": 0.2855237701478863,
+      "learning_rate": 9.310391198956896e-06,
+      "loss": 0.0795,
+      "step": 315
+    },
+    {
+      "epoch": 0.5376435559336452,
+      "grad_norm": 0.19062890218128994,
+      "learning_rate": 9.255320906089017e-06,
+      "loss": 0.0385,
+      "step": 316
+    },
+    {
+      "epoch": 0.5393449595916632,
+      "grad_norm": 0.12818263127205254,
+      "learning_rate": 9.200273314131356e-06,
+      "loss": 0.0358,
+      "step": 317
+    },
+    {
+      "epoch": 0.541046363249681,
+      "grad_norm": 0.21271005750639677,
+      "learning_rate": 9.145250101163032e-06,
+      "loss": 0.0511,
+      "step": 318
+    },
+    {
+      "epoch": 0.5427477669076989,
+      "grad_norm": 0.3230145385857894,
+      "learning_rate": 9.090252944520002e-06,
+      "loss": 0.1249,
+      "step": 319
+    },
+    {
+      "epoch": 0.5444491705657167,
+      "grad_norm": 0.17129666760859652,
+      "learning_rate": 9.035283520743911e-06,
+      "loss": 0.0473,
+      "step": 320
+    },
+    {
+      "epoch": 0.5461505742237346,
+      "grad_norm": 0.25326344089197617,
+      "learning_rate": 8.980343505530988e-06,
+      "loss": 0.0613,
+      "step": 321
+    },
+    {
+      "epoch": 0.5478519778817524,
+      "grad_norm": 0.2847520686151069,
+      "learning_rate": 8.925434573680986e-06,
+      "loss": 0.0883,
+      "step": 322
+    },
+    {
+      "epoch": 0.5495533815397703,
+      "grad_norm": 0.358576412450163,
+      "learning_rate": 8.870558399046086e-06,
+      "loss": 0.1097,
+      "step": 323
+    },
+    {
+      "epoch": 0.5512547851977881,
+      "grad_norm": 0.21816479879448794,
+      "learning_rate": 8.815716654479903e-06,
+      "loss": 0.0766,
+      "step": 324
+    },
+    {
+      "epoch": 0.552956188855806,
+      "grad_norm": 0.27098416317022683,
+      "learning_rate": 8.76091101178648e-06,
+      "loss": 0.0959,
+      "step": 325
+    },
+    {
+      "epoch": 0.5546575925138238,
+      "grad_norm": 0.16228239529758662,
+      "learning_rate": 8.706143141669324e-06,
+      "loss": 0.0427,
+      "step": 326
+    },
+    {
+      "epoch": 0.5563589961718418,
+      "grad_norm": 0.19457073616768888,
+      "learning_rate": 8.651414713680474e-06,
+      "loss": 0.0674,
+      "step": 327
+    },
+    {
+      "epoch": 0.5580603998298597,
+      "grad_norm": 0.21506185063350097,
+      "learning_rate": 8.59672739616962e-06,
+      "loss": 0.0725,
+      "step": 328
+    },
+    {
+      "epoch": 0.5597618034878775,
+      "grad_norm": 0.25928980929110046,
+      "learning_rate": 8.542082856233216e-06,
+      "loss": 0.0926,
+      "step": 329
+    },
+    {
+      "epoch": 0.5614632071458954,
+      "grad_norm": 0.19301809133671421,
+      "learning_rate": 8.487482759663696e-06,
+      "loss": 0.0661,
+      "step": 330
+    },
+    {
+      "epoch": 0.5631646108039132,
+      "grad_norm": 0.19336796199191325,
+      "learning_rate": 8.43292877089867e-06,
+      "loss": 0.0694,
+      "step": 331
+    },
+    {
+      "epoch": 0.5648660144619311,
+      "grad_norm": 0.2300001624245782,
+      "learning_rate": 8.378422552970185e-06,
+      "loss": 0.0746,
+      "step": 332
+    },
+    {
+      "epoch": 0.5665674181199489,
+      "grad_norm": 0.23998094239502984,
+      "learning_rate": 8.32396576745404e-06,
+      "loss": 0.0696,
+      "step": 333
+    },
+    {
+      "epoch": 0.5682688217779668,
+      "grad_norm": 0.2052361707208072,
+      "learning_rate": 8.269560074419126e-06,
+      "loss": 0.0624,
+      "step": 334
+    },
+    {
+      "epoch": 0.5699702254359846,
+      "grad_norm": 0.2962367385563096,
+      "learning_rate": 8.215207132376824e-06,
+      "loss": 0.124,
+      "step": 335
+    },
+    {
+      "epoch": 0.5716716290940026,
+      "grad_norm": 0.24752974776344203,
+      "learning_rate": 8.160908598230448e-06,
+      "loss": 0.0653,
+      "step": 336
+    },
+    {
+      "epoch": 0.5733730327520205,
+      "grad_norm": 0.18127533151541284,
+      "learning_rate": 8.10666612722473e-06,
+      "loss": 0.0591,
+      "step": 337
+    },
+    {
+      "epoch": 0.5750744364100383,
+      "grad_norm": 0.18346580605719615,
+      "learning_rate": 8.052481372895363e-06,
+      "loss": 0.0488,
+      "step": 338
+    },
+    {
+      "epoch": 0.5767758400680562,
+      "grad_norm": 0.31877710293947625,
+      "learning_rate": 7.998355987018606e-06,
+      "loss": 0.0872,
+      "step": 339
+    },
+    {
+      "epoch": 0.578477243726074,
+      "grad_norm": 0.1344651628761348,
+      "learning_rate": 7.944291619560914e-06,
+      "loss": 0.0403,
+      "step": 340
+    },
+    {
+      "epoch": 0.5801786473840919,
+      "grad_norm": 0.13295165420726127,
+      "learning_rate": 7.890289918628644e-06,
+      "loss": 0.0476,
+      "step": 341
+    },
+    {
+      "epoch": 0.5818800510421097,
+      "grad_norm": 0.27781942031149137,
+      "learning_rate": 7.836352530417824e-06,
+      "loss": 0.0925,
+      "step": 342
+    },
+    {
+      "epoch": 0.5835814547001276,
+      "grad_norm": 0.2923332990407699,
+      "learning_rate": 7.782481099163958e-06,
+      "loss": 0.1173,
+      "step": 343
+    },
+    {
+      "epoch": 0.5852828583581454,
+      "grad_norm": 0.20398201081622527,
+      "learning_rate": 7.728677267091912e-06,
+      "loss": 0.0712,
+      "step": 344
+    },
+    {
+      "epoch": 0.5869842620161634,
+      "grad_norm": 0.1827445671136079,
+      "learning_rate": 7.674942674365847e-06,
+      "loss": 0.0588,
+      "step": 345
+    },
+    {
+      "epoch": 0.5886856656741812,
+      "grad_norm": 0.16240437969547905,
+      "learning_rate": 7.621278959039217e-06,
+      "loss": 0.0637,
+      "step": 346
+    },
+    {
+      "epoch": 0.5903870693321991,
+      "grad_norm": 0.16062106806552065,
+      "learning_rate": 7.567687757004843e-06,
+      "loss": 0.0414,
+      "step": 347
+    },
+    {
+      "epoch": 0.5920884729902169,
+      "grad_norm": 0.2689581192862772,
+      "learning_rate": 7.514170701945047e-06,
+      "loss": 0.0897,
+      "step": 348
+    },
+    {
+      "epoch": 0.5937898766482348,
+      "grad_norm": 0.2681187781079101,
+      "learning_rate": 7.460729425281831e-06,
+      "loss": 0.0709,
+      "step": 349
+    },
+    {
+      "epoch": 0.5954912803062526,
+      "grad_norm": 0.17526558434710648,
+      "learning_rate": 7.407365556127162e-06,
+      "loss": 0.0539,
+      "step": 350
+    },
+    {
+      "epoch": 0.5971926839642705,
+      "grad_norm": 0.1648773603270098,
+      "learning_rate": 7.354080721233303e-06,
+      "loss": 0.0503,
+      "step": 351
+    },
+    {
+      "epoch": 0.5988940876222884,
+      "grad_norm": 0.17999080092184985,
+      "learning_rate": 7.300876544943227e-06,
+      "loss": 0.0605,
+      "step": 352
+    },
+    {
+      "epoch": 0.6005954912803062,
+      "grad_norm": 0.20308998126444186,
+      "learning_rate": 7.247754649141097e-06,
+      "loss": 0.0769,
+      "step": 353
+    },
+    {
+      "epoch": 0.6022968949383242,
+      "grad_norm": 0.16018163485236867,
+      "learning_rate": 7.194716653202826e-06,
+      "loss": 0.0545,
+      "step": 354
+    },
+    {
+      "epoch": 0.603998298596342,
+      "grad_norm": 0.24744753769790884,
+      "learning_rate": 7.1417641739467104e-06,
+      "loss": 0.0776,
+      "step": 355
+    },
+    {
+      "epoch": 0.6056997022543599,
+      "grad_norm": 0.2298041880240223,
+      "learning_rate": 7.088898825584139e-06,
+      "loss": 0.0674,
+      "step": 356
+    },
+    {
+      "epoch": 0.6074011059123777,
+      "grad_norm": 0.19841362318124559,
+      "learning_rate": 7.036122219670398e-06,
+      "loss": 0.0635,
+      "step": 357
+    },
+    {
+      "epoch": 0.6091025095703956,
+      "grad_norm": 0.18877644807321198,
+      "learning_rate": 6.9834359650555305e-06,
+      "loss": 0.0777,
+      "step": 358
+    },
+    {
+      "epoch": 0.6108039132284134,
+      "grad_norm": 0.16102641349173863,
+      "learning_rate": 6.930841667835295e-06,
+      "loss": 0.0576,
+      "step": 359
+    },
+    {
+      "epoch": 0.6125053168864313,
+      "grad_norm": 0.20224797100905906,
+      "learning_rate": 6.878340931302208e-06,
+      "loss": 0.0754,
+      "step": 360
+    },
+    {
+      "epoch": 0.6142067205444491,
+      "grad_norm": 0.2857194289415506,
+      "learning_rate": 6.825935355896669e-06,
+      "loss": 0.1052,
+      "step": 361
+    },
+    {
+      "epoch": 0.615908124202467,
+      "grad_norm": 0.20616924754434873,
+      "learning_rate": 6.773626539158171e-06,
+      "loss": 0.0716,
+      "step": 362
+    },
+    {
+      "epoch": 0.617609527860485,
+      "grad_norm": 0.23846455066099467,
+      "learning_rate": 6.721416075676601e-06,
+      "loss": 0.0847,
+      "step": 363
+    },
+    {
+      "epoch": 0.6193109315185028,
+      "grad_norm": 0.14989055759308637,
+      "learning_rate": 6.669305557043626e-06,
+      "loss": 0.0371,
+      "step": 364
+    },
+    {
+      "epoch": 0.6210123351765207,
+      "grad_norm": 0.25146318527723016,
+      "learning_rate": 6.617296571804191e-06,
+      "loss": 0.0938,
+      "step": 365
+    },
+    {
+      "epoch": 0.6227137388345385,
+      "grad_norm": 0.2795331744757292,
+      "learning_rate": 6.565390705408072e-06,
+      "loss": 0.0503,
+      "step": 366
+    },
+    {
+      "epoch": 0.6244151424925564,
+      "grad_norm": 0.17139219570184439,
+      "learning_rate": 6.513589540161556e-06,
+      "loss": 0.0578,
+      "step": 367
+    },
+    {
+      "epoch": 0.6261165461505742,
+      "grad_norm": 0.157560985399721,
+      "learning_rate": 6.461894655179204e-06,
+      "loss": 0.0582,
+      "step": 368
+    },
+    {
+      "epoch": 0.6278179498085921,
+      "grad_norm": 0.22752999438352467,
+      "learning_rate": 6.410307626335717e-06,
+      "loss": 0.0779,
+      "step": 369
+    },
+    {
+      "epoch": 0.6295193534666099,
+      "grad_norm": 0.27521908636927156,
+      "learning_rate": 6.358830026217887e-06,
+      "loss": 0.0826,
+      "step": 370
+    },
+    {
+      "epoch": 0.6312207571246278,
+      "grad_norm": 0.21611309006232896,
+      "learning_rate": 6.30746342407667e-06,
+      "loss": 0.0575,
+      "step": 371
+    },
+    {
+      "epoch": 0.6329221607826457,
+      "grad_norm": 0.2022628238984182,
+      "learning_rate": 6.256209385779341e-06,
+      "loss": 0.0597,
+      "step": 372
+    },
+    {
+      "epoch": 0.6346235644406636,
+      "grad_norm": 0.20087283890633761,
+      "learning_rate": 6.205069473761756e-06,
+      "loss": 0.0565,
+      "step": 373
+    },
+    {
+      "epoch": 0.6363249680986814,
+      "grad_norm": 0.27973559478745097,
+      "learning_rate": 6.154045246980742e-06,
+      "loss": 0.0777,
+      "step": 374
+    },
+    {
+      "epoch": 0.6380263717566993,
+      "grad_norm": 0.2153647046344647,
+      "learning_rate": 6.1031382608665456e-06,
+      "loss": 0.065,
+      "step": 375
+    },
+    {
+      "epoch": 0.6397277754147171,
+      "grad_norm": 0.15832658385378948,
+      "learning_rate": 6.052350067275441e-06,
+      "loss": 0.0463,
+      "step": 376
+    },
+    {
+      "epoch": 0.641429179072735,
+      "grad_norm": 0.24889956825525697,
+      "learning_rate": 6.001682214442406e-06,
+      "loss": 0.0868,
+      "step": 377
+    },
+    {
+      "epoch": 0.6431305827307529,
+      "grad_norm": 0.2537522589782198,
+      "learning_rate": 5.951136246933933e-06,
+      "loss": 0.0771,
+      "step": 378
+    },
+    {
+      "epoch": 0.6448319863887707,
+      "grad_norm": 0.35384965176549915,
+      "learning_rate": 5.900713705600951e-06,
+      "loss": 0.0885,
+      "step": 379
+    },
+    {
+      "epoch": 0.6465333900467886,
+      "grad_norm": 0.24583176378622248,
+      "learning_rate": 5.850416127531841e-06,
+      "loss": 0.076,
+      "step": 380
+    },
+    {
+      "epoch": 0.6482347937048064,
+      "grad_norm": 0.19401849479737754,
+      "learning_rate": 5.800245046005585e-06,
+      "loss": 0.055,
+      "step": 381
+    },
+    {
+      "epoch": 0.6499361973628244,
+      "grad_norm": 0.22292321754995165,
+      "learning_rate": 5.750201990445024e-06,
+      "loss": 0.0837,
+      "step": 382
+    },
+    {
+      "epoch": 0.6516376010208422,
+      "grad_norm": 0.15980625639550533,
+      "learning_rate": 5.70028848637024e-06,
+      "loss": 0.053,
+      "step": 383
+    },
+    {
+      "epoch": 0.6533390046788601,
+      "grad_norm": 0.17476218437373806,
+      "learning_rate": 5.650506055352052e-06,
+      "loss": 0.047,
+      "step": 384
+    },
+    {
+      "epoch": 0.6550404083368779,
+      "grad_norm": 0.26159431180356163,
+      "learning_rate": 5.600856214965613e-06,
+      "loss": 0.075,
+      "step": 385
+    },
+    {
+      "epoch": 0.6567418119948958,
+      "grad_norm": 0.24983048217170784,
+      "learning_rate": 5.551340478744176e-06,
+      "loss": 0.0819,
+      "step": 386
+    },
+    {
+      "epoch": 0.6584432156529136,
+      "grad_norm": 0.277677983790708,
+      "learning_rate": 5.501960356132945e-06,
+      "loss": 0.0743,
+      "step": 387
+    },
+    {
+      "epoch": 0.6601446193109315,
+      "grad_norm": 0.2687619515031017,
+      "learning_rate": 5.4527173524430395e-06,
+      "loss": 0.076,
+      "step": 388
+    },
+    {
+      "epoch": 0.6618460229689493,
+      "grad_norm": 0.18825889492381687,
+      "learning_rate": 5.403612968805649e-06,
+      "loss": 0.0533,
+      "step": 389
+    },
+    {
+      "epoch": 0.6635474266269672,
+      "grad_norm": 0.22722813251197366,
+      "learning_rate": 5.354648702126229e-06,
+      "loss": 0.0669,
+      "step": 390
+    },
+    {
+      "epoch": 0.6652488302849852,
+      "grad_norm": 0.15428942499610793,
+      "learning_rate": 5.305826045038899e-06,
+      "loss": 0.0496,
+      "step": 391
+    },
+    {
+      "epoch": 0.666950233943003,
+      "grad_norm": 0.29481966419649847,
+      "learning_rate": 5.257146485860927e-06,
+      "loss": 0.0871,
+      "step": 392
+    },
+    {
+      "epoch": 0.6686516376010209,
+      "grad_norm": 0.2132988743676148,
+      "learning_rate": 5.208611508547367e-06,
+      "loss": 0.072,
+      "step": 393
+    },
+    {
+      "epoch": 0.6703530412590387,
+      "grad_norm": 0.1932211736507852,
+      "learning_rate": 5.160222592645808e-06,
+      "loss": 0.0672,
+      "step": 394
+    },
+    {
+      "epoch": 0.6720544449170566,
+      "grad_norm": 0.32017863808058095,
+      "learning_rate": 5.111981213251293e-06,
+      "loss": 0.0996,
+      "step": 395
+    },
+    {
+      "epoch": 0.6737558485750744,
+      "grad_norm": 0.33076228774617505,
+      "learning_rate": 5.063888840961325e-06,
+      "loss": 0.1062,
+      "step": 396
+    },
+    {
+      "epoch": 0.6754572522330923,
+      "grad_norm": 0.2152145333210106,
+      "learning_rate": 5.015946941831064e-06,
+      "loss": 0.0682,
+      "step": 397
+    },
+    {
+      "epoch": 0.6771586558911101,
+      "grad_norm": 0.21543841192984545,
+      "learning_rate": 4.968156977328626e-06,
+      "loss": 0.0572,
+      "step": 398
+    },
+    {
+      "epoch": 0.678860059549128,
+      "grad_norm": 0.16950132912260057,
+      "learning_rate": 4.920520404290512e-06,
+      "loss": 0.0577,
+      "step": 399
+    },
+    {
+      "epoch": 0.680561463207146,
+      "grad_norm": 0.20787322030508298,
+      "learning_rate": 4.87303867487723e-06,
+      "loss": 0.0561,
+      "step": 400
+    },
+    {
+      "epoch": 0.6822628668651638,
+      "grad_norm": 0.1533621298140527,
+      "learning_rate": 4.825713236529005e-06,
+      "loss": 0.0435,
+      "step": 401
+    },
+    {
+      "epoch": 0.6839642705231816,
+      "grad_norm": 0.18296820958014204,
+      "learning_rate": 4.778545531921668e-06,
+      "loss": 0.0538,
+      "step": 402
+    },
+    {
+      "epoch": 0.6856656741811995,
+      "grad_norm": 0.24677189398080018,
+      "learning_rate": 4.731536998922657e-06,
+      "loss": 0.0715,
+      "step": 403
+    },
+    {
+      "epoch": 0.6873670778392174,
+      "grad_norm": 0.18381971512083234,
+      "learning_rate": 4.684689070547216e-06,
+      "loss": 0.0589,
+      "step": 404
+    },
+    {
+      "epoch": 0.6890684814972352,
+      "grad_norm": 0.19563535366138982,
+      "learning_rate": 4.638003174914675e-06,
+      "loss": 0.0375,
+      "step": 405
+    },
+    {
+      "epoch": 0.6907698851552531,
+      "grad_norm": 0.27920320369616836,
+      "learning_rate": 4.591480735204953e-06,
+      "loss": 0.0657,
+      "step": 406
+    },
+    {
+      "epoch": 0.6924712888132709,
+      "grad_norm": 0.22501473081164228,
+      "learning_rate": 4.545123169615134e-06,
+      "loss": 0.0754,
+      "step": 407
+    },
+    {
+      "epoch": 0.6941726924712888,
+      "grad_norm": 0.18685344173190274,
+      "learning_rate": 4.49893189131627e-06,
+      "loss": 0.0753,
+      "step": 408
+    },
+    {
+      "epoch": 0.6958740961293067,
+      "grad_norm": 0.17185385088024444,
+      "learning_rate": 4.45290830841028e-06,
+      "loss": 0.0514,
+      "step": 409
+    },
+    {
+      "epoch": 0.6975754997873246,
+      "grad_norm": 0.2702549233525611,
+      "learning_rate": 4.407053823887033e-06,
+      "loss": 0.0833,
+      "step": 410
+    },
+    {
+      "epoch": 0.6992769034453424,
+      "grad_norm": 0.23362806883478313,
+      "learning_rate": 4.361369835581569e-06,
+      "loss": 0.0769,
+      "step": 411
+    },
+    {
+      "epoch": 0.7009783071033603,
+      "grad_norm": 0.2101592538580294,
+      "learning_rate": 4.315857736131508e-06,
+      "loss": 0.0602,
+      "step": 412
+    },
+    {
+      "epoch": 0.7026797107613781,
+      "grad_norm": 0.20842960868238944,
+      "learning_rate": 4.2705189129345814e-06,
+      "loss": 0.074,
+      "step": 413
+    },
+    {
+      "epoch": 0.704381114419396,
+      "grad_norm": 0.18803427484767865,
+      "learning_rate": 4.225354748106328e-06,
+      "loss": 0.07,
+      "step": 414
+    },
+    {
+      "epoch": 0.7060825180774138,
+      "grad_norm": 0.3066569805512131,
+      "learning_rate": 4.180366618437996e-06,
+      "loss": 0.093,
+      "step": 415
+    },
+    {
+      "epoch": 0.7077839217354317,
+      "grad_norm": 0.1744953221856188,
+      "learning_rate": 4.13555589535453e-06,
+      "loss": 0.0555,
+      "step": 416
+    },
+    {
+      "epoch": 0.7094853253934496,
+      "grad_norm": 0.166243605934049,
+      "learning_rate": 4.0909239448727985e-06,
+      "loss": 0.061,
+      "step": 417
+    },
+    {
+      "epoch": 0.7111867290514675,
+      "grad_norm": 0.20351482232627222,
+      "learning_rate": 4.046472127559937e-06,
+      "loss": 0.0715,
+      "step": 418
+    },
+    {
+      "epoch": 0.7128881327094854,
+      "grad_norm": 0.21225234915881963,
+      "learning_rate": 4.002201798491875e-06,
+      "loss": 0.0502,
+      "step": 419
+    },
+    {
+      "epoch": 0.7145895363675032,
+      "grad_norm": 0.20113105383651986,
+      "learning_rate": 3.958114307212018e-06,
+      "loss": 0.0645,
+      "step": 420
+    },
+    {
+      "epoch": 0.7162909400255211,
+      "grad_norm": 0.18912603242706336,
+      "learning_rate": 3.91421099769013e-06,
+      "loss": 0.0642,
+      "step": 421
+    },
+    {
+      "epoch": 0.7179923436835389,
+      "grad_norm": 0.1879026045376002,
+      "learning_rate": 3.870493208281337e-06,
+      "loss": 0.0479,
+      "step": 422
+    },
+    {
+      "epoch": 0.7196937473415568,
+      "grad_norm": 0.26477442490530756,
+      "learning_rate": 3.826962271685351e-06,
+      "loss": 0.0831,
+      "step": 423
+    },
+    {
+      "epoch": 0.7213951509995746,
+      "grad_norm": 0.29428954187807327,
+      "learning_rate": 3.7836195149058386e-06,
+      "loss": 0.0724,
+      "step": 424
+    },
+    {
+      "epoch": 0.7230965546575925,
+      "grad_norm": 0.23675032676032767,
+      "learning_rate": 3.7404662592099483e-06,
+      "loss": 0.0854,
+      "step": 425
+    },
+    {
+      "epoch": 0.7247979583156103,
+      "grad_norm": 0.283635317471,
+      "learning_rate": 3.697503820088063e-06,
+      "loss": 0.0805,
+      "step": 426
+    },
+    {
+      "epoch": 0.7264993619736282,
+      "grad_norm": 0.20299536681350067,
+      "learning_rate": 3.654733507213678e-06,
+      "loss": 0.0629,
+      "step": 427
+    },
+    {
+      "epoch": 0.7282007656316462,
+      "grad_norm": 0.2737835974179114,
+      "learning_rate": 3.61215662440349e-06,
+      "loss": 0.0813,
+      "step": 428
+    },
+    {
+      "epoch": 0.729902169289664,
+      "grad_norm": 0.2292958769762407,
+      "learning_rate": 3.5697744695776326e-06,
+      "loss": 0.0625,
+      "step": 429
+    },
+    {
+      "epoch": 0.7316035729476819,
+      "grad_norm": 0.2690574028374729,
+      "learning_rate": 3.5275883347201336e-06,
+      "loss": 0.0895,
+      "step": 430
+    },
+    {
+      "epoch": 0.7333049766056997,
+      "grad_norm": 0.2612064488237033,
+      "learning_rate": 3.4855995058395066e-06,
+      "loss": 0.076,
+      "step": 431
+    },
+    {
+      "epoch": 0.7350063802637176,
+      "grad_norm": 0.20866850048700003,
+      "learning_rate": 3.443809262929575e-06,
+      "loss": 0.0719,
+      "step": 432
+    },
+    {
+      "epoch": 0.7367077839217354,
+      "grad_norm": 0.2274495669126989,
+      "learning_rate": 3.4022188799304214e-06,
+      "loss": 0.0754,
+      "step": 433
+    },
+    {
+      "epoch": 0.7384091875797533,
+      "grad_norm": 0.21436057034429756,
+      "learning_rate": 3.36082962468958e-06,
+      "loss": 0.0634,
+      "step": 434
+    },
+    {
+      "epoch": 0.7401105912377711,
+      "grad_norm": 0.2122758067179719,
+      "learning_rate": 3.3196427589233725e-06,
+      "loss": 0.0605,
+      "step": 435
+    },
+    {
+      "epoch": 0.741811994895789,
+      "grad_norm": 0.2387395787166068,
+      "learning_rate": 3.2786595381784512e-06,
+      "loss": 0.0679,
+      "step": 436
+    },
+    {
+      "epoch": 0.7435133985538069,
+      "grad_norm": 0.31722612816219875,
+      "learning_rate": 3.2378812117935154e-06,
+      "loss": 0.1076,
+      "step": 437
+    },
+    {
+      "epoch": 0.7452148022118248,
+      "grad_norm": 0.34631520054017056,
+      "learning_rate": 3.1973090228612404e-06,
+      "loss": 0.1121,
+      "step": 438
+    },
+    {
+      "epoch": 0.7469162058698426,
+      "grad_norm": 0.23043404441429802,
+      "learning_rate": 3.15694420819038e-06,
+      "loss": 0.0877,
+      "step": 439
+    },
+    {
+      "epoch": 0.7486176095278605,
+      "grad_norm": 0.09929972754097662,
+      "learning_rate": 3.116787998268046e-06,
+      "loss": 0.0281,
+      "step": 440
+    },
+    {
+      "epoch": 0.7503190131858783,
+      "grad_norm": 0.2540228794638467,
+      "learning_rate": 3.076841617222228e-06,
+      "loss": 0.1016,
+      "step": 441
+    },
+    {
+      "epoch": 0.7520204168438962,
+      "grad_norm": 0.20530486566659917,
+      "learning_rate": 3.0371062827844434e-06,
+      "loss": 0.0759,
+      "step": 442
+    },
+    {
+      "epoch": 0.753721820501914,
+      "grad_norm": 0.22889025759698128,
+      "learning_rate": 2.997583206252647e-06,
+      "loss": 0.0641,
+      "step": 443
+    },
+    {
+      "epoch": 0.7554232241599319,
+      "grad_norm": 0.22376344098617418,
+      "learning_rate": 2.958273592454285e-06,
+      "loss": 0.0696,
+      "step": 444
+    },
+    {
+      "epoch": 0.7571246278179498,
+      "grad_norm": 0.29335139294143503,
+      "learning_rate": 2.9191786397095778e-06,
+      "loss": 0.0722,
+      "step": 445
+    },
+    {
+      "epoch": 0.7588260314759677,
+      "grad_norm": 0.19904638249374088,
+      "learning_rate": 2.880299539794975e-06,
+      "loss": 0.0644,
+      "step": 446
+    },
+    {
+      "epoch": 0.7605274351339856,
+      "grad_norm": 0.27398415500191214,
+      "learning_rate": 2.841637477906851e-06,
+      "loss": 0.097,
+      "step": 447
+    },
+    {
+      "epoch": 0.7622288387920034,
+      "grad_norm": 0.1909061169980495,
+      "learning_rate": 2.803193632625346e-06,
+      "loss": 0.0653,
+      "step": 448
+    },
+    {
+      "epoch": 0.7639302424500213,
+      "grad_norm": 0.1839577194240098,
+      "learning_rate": 2.7649691758784603e-06,
+      "loss": 0.0612,
+      "step": 449
+    },
+    {
+      "epoch": 0.7656316461080391,
+      "grad_norm": 0.18749598930597564,
+      "learning_rate": 2.7269652729063233e-06,
+      "loss": 0.0586,
+      "step": 450
+    },
+    {
+      "epoch": 0.767333049766057,
+      "grad_norm": 0.2550319493358391,
+      "learning_rate": 2.689183082225659e-06,
+      "loss": 0.0784,
+      "step": 451
+    },
+    {
+      "epoch": 0.7690344534240748,
+      "grad_norm": 0.2702034699283639,
+      "learning_rate": 2.65162375559449e-06,
+      "loss": 0.1012,
+      "step": 452
+    },
+    {
+      "epoch": 0.7707358570820927,
+      "grad_norm": 0.22699731806268653,
+      "learning_rate": 2.614288437977014e-06,
+      "loss": 0.08,
+      "step": 453
+    },
+    {
+      "epoch": 0.7724372607401105,
+      "grad_norm": 0.28841442095335584,
+      "learning_rate": 2.5771782675087078e-06,
+      "loss": 0.105,
+      "step": 454
+    },
+    {
+      "epoch": 0.7741386643981285,
+      "grad_norm": 0.2545180037798505,
+      "learning_rate": 2.5402943754616182e-06,
+      "loss": 0.0847,
+      "step": 455
+    },
+    {
+      "epoch": 0.7758400680561464,
+      "grad_norm": 0.16486311867632228,
+      "learning_rate": 2.5036378862099e-06,
+      "loss": 0.0409,
+      "step": 456
+    },
+    {
+      "epoch": 0.7775414717141642,
+      "grad_norm": 0.263761370929647,
+      "learning_rate": 2.467209917195513e-06,
+      "loss": 0.096,
+      "step": 457
+    },
+    {
+      "epoch": 0.7792428753721821,
+      "grad_norm": 0.235530147590817,
+      "learning_rate": 2.4310115788941855e-06,
+      "loss": 0.0595,
+      "step": 458
+    },
+    {
+      "epoch": 0.7809442790301999,
+      "grad_norm": 0.2110709448726579,
+      "learning_rate": 2.3950439747815357e-06,
+      "loss": 0.07,
+      "step": 459
+    },
+    {
+      "epoch": 0.7826456826882178,
+      "grad_norm": 0.1763868737174647,
+      "learning_rate": 2.359308201299454e-06,
+      "loss": 0.0586,
+      "step": 460
+    },
+    {
+      "epoch": 0.7843470863462356,
+      "grad_norm": 0.16676117425431294,
+      "learning_rate": 2.3238053478226665e-06,
+      "loss": 0.0492,
+      "step": 461
+    },
+    {
+      "epoch": 0.7860484900042535,
+      "grad_norm": 0.15717970250735389,
+      "learning_rate": 2.2885364966255372e-06,
+      "loss": 0.0487,
+      "step": 462
+    },
+    {
+      "epoch": 0.7877498936622713,
+      "grad_norm": 0.28197077618286126,
+      "learning_rate": 2.2535027228490582e-06,
+      "loss": 0.0857,
+      "step": 463
+    },
+    {
+      "epoch": 0.7894512973202893,
+      "grad_norm": 0.264862322279995,
+      "learning_rate": 2.2187050944680942e-06,
+      "loss": 0.0937,
+      "step": 464
+    },
+    {
+      "epoch": 0.7911527009783071,
+      "grad_norm": 0.22066167775922854,
+      "learning_rate": 2.18414467225882e-06,
+      "loss": 0.0642,
+      "step": 465
+    },
+    {
+      "epoch": 0.792854104636325,
+      "grad_norm": 0.2250702122829751,
+      "learning_rate": 2.1498225097663695e-06,
+      "loss": 0.0831,
+      "step": 466
+    },
+    {
+      "epoch": 0.7945555082943428,
+      "grad_norm": 0.22295479048611572,
+      "learning_rate": 2.115739653272747e-06,
+      "loss": 0.0631,
+      "step": 467
+    },
+    {
+      "epoch": 0.7962569119523607,
+      "grad_norm": 0.24242984035739493,
+      "learning_rate": 2.0818971417649013e-06,
+      "loss": 0.0591,
+      "step": 468
+    },
+    {
+      "epoch": 0.7979583156103786,
+      "grad_norm": 0.2612637093823693,
+      "learning_rate": 2.048296006903081e-06,
+      "loss": 0.1046,
+      "step": 469
+    },
+    {
+      "epoch": 0.7996597192683964,
+      "grad_norm": 0.1792782441746806,
+      "learning_rate": 2.0149372729893646e-06,
+      "loss": 0.0445,
+      "step": 470
+    },
+    {
+      "epoch": 0.8013611229264143,
+      "grad_norm": 0.29350099593257656,
+      "learning_rate": 1.981821956936448e-06,
+      "loss": 0.0804,
+      "step": 471
+    },
+    {
+      "epoch": 0.8030625265844321,
+      "grad_norm": 0.22341777662676934,
+      "learning_rate": 1.9489510682366363e-06,
+      "loss": 0.0745,
+      "step": 472
+    },
+    {
+      "epoch": 0.8047639302424501,
+      "grad_norm": 0.1589102792801742,
+      "learning_rate": 1.916325608931079e-06,
+      "loss": 0.047,
+      "step": 473
+    },
+    {
+      "epoch": 0.8064653339004679,
+      "grad_norm": 0.2345326238035068,
+      "learning_rate": 1.8839465735792095e-06,
+      "loss": 0.0572,
+      "step": 474
+    },
+    {
+      "epoch": 0.8081667375584858,
+      "grad_norm": 0.24004839314637838,
+      "learning_rate": 1.8518149492284477e-06,
+      "loss": 0.0884,
+      "step": 475
+    },
+    {
+      "epoch": 0.8098681412165036,
+      "grad_norm": 0.29421822191095054,
+      "learning_rate": 1.8199317153840933e-06,
+      "loss": 0.0887,
+      "step": 476
+    },
+    {
+      "epoch": 0.8115695448745215,
+      "grad_norm": 0.29025654033124915,
+      "learning_rate": 1.7882978439794708e-06,
+      "loss": 0.1021,
+      "step": 477
+    },
+    {
+      "epoch": 0.8132709485325393,
+      "grad_norm": 0.25878138649406646,
+      "learning_rate": 1.756914299346304e-06,
+      "loss": 0.0616,
+      "step": 478
+    },
+    {
+      "epoch": 0.8149723521905572,
+      "grad_norm": 0.2102766744469287,
+      "learning_rate": 1.7257820381853197e-06,
+      "loss": 0.0627,
+      "step": 479
+    },
+    {
+      "epoch": 0.816673755848575,
+      "grad_norm": 0.24781694684746497,
+      "learning_rate": 1.6949020095370816e-06,
+      "loss": 0.0766,
+      "step": 480
+    },
+    {
+      "epoch": 0.8183751595065929,
+      "grad_norm": 0.1734085990747018,
+      "learning_rate": 1.6642751547530512e-06,
+      "loss": 0.0514,
+      "step": 481
+    },
+    {
+      "epoch": 0.8200765631646108,
+      "grad_norm": 0.2117204201108364,
+      "learning_rate": 1.6339024074669107e-06,
+      "loss": 0.0717,
+      "step": 482
+    },
+    {
+      "epoch": 0.8217779668226287,
+      "grad_norm": 0.23022449445835655,
+      "learning_rate": 1.6037846935660807e-06,
+      "loss": 0.0697,
+      "step": 483
+    },
+    {
+      "epoch": 0.8234793704806466,
+      "grad_norm": 0.2031147008426011,
+      "learning_rate": 1.5739229311635152e-06,
+      "loss": 0.0647,
+      "step": 484
+    },
+    {
+      "epoch": 0.8251807741386644,
+      "grad_norm": 0.16452534724080284,
+      "learning_rate": 1.5443180305696948e-06,
+      "loss": 0.0477,
+      "step": 485
+    },
+    {
+      "epoch": 0.8268821777966823,
+      "grad_norm": 0.22807646976291562,
+      "learning_rate": 1.5149708942648922e-06,
+      "loss": 0.0814,
+      "step": 486
+    },
+    {
+      "epoch": 0.8285835814547001,
+      "grad_norm": 0.24827387251547514,
+      "learning_rate": 1.4858824168716524e-06,
+      "loss": 0.0755,
+      "step": 487
+    },
+    {
+      "epoch": 0.830284985112718,
+      "grad_norm": 0.22526986180364844,
+      "learning_rate": 1.4570534851275241e-06,
+      "loss": 0.076,
+      "step": 488
+    },
+    {
+      "epoch": 0.8319863887707358,
+      "grad_norm": 0.20155351877240635,
+      "learning_rate": 1.4284849778580279e-06,
+      "loss": 0.0698,
+      "step": 489
+    },
+    {
+      "epoch": 0.8336877924287537,
+      "grad_norm": 0.1495610561199564,
+      "learning_rate": 1.4001777659498584e-06,
+      "loss": 0.04,
+      "step": 490
+    },
+    {
+      "epoch": 0.8353891960867715,
+      "grad_norm": 0.22042874488356587,
+      "learning_rate": 1.3721327123243533e-06,
+      "loss": 0.0696,
+      "step": 491
+    },
+    {
+      "epoch": 0.8370905997447895,
+      "grad_norm": 0.217650029772456,
+      "learning_rate": 1.3443506719111666e-06,
+      "loss": 0.0499,
+      "step": 492
+    },
+    {
+      "epoch": 0.8387920034028074,
+      "grad_norm": 0.28478642607874244,
+      "learning_rate": 1.3168324916222296e-06,
+      "loss": 0.1052,
+      "step": 493
+    },
+    {
+      "epoch": 0.8404934070608252,
+      "grad_norm": 0.2847136340573529,
+      "learning_rate": 1.28957901032591e-06,
+      "loss": 0.0772,
+      "step": 494
+    },
+    {
+      "epoch": 0.8421948107188431,
+      "grad_norm": 0.21335659065873505,
+      "learning_rate": 1.2625910588214608e-06,
+      "loss": 0.0651,
+      "step": 495
+    },
+    {
+      "epoch": 0.8438962143768609,
+      "grad_norm": 0.2007624647622523,
+      "learning_rate": 1.2358694598136755e-06,
+      "loss": 0.0579,
+      "step": 496
+    },
+    {
+      "epoch": 0.8455976180348788,
+      "grad_norm": 0.22052289165556443,
+      "learning_rate": 1.2094150278878303e-06,
+      "loss": 0.0564,
+      "step": 497
+    },
+    {
+      "epoch": 0.8472990216928966,
+      "grad_norm": 0.23003203856097848,
+      "learning_rate": 1.1832285694848255e-06,
+      "loss": 0.0604,
+      "step": 498
+    },
+    {
+      "epoch": 0.8490004253509145,
+      "grad_norm": 0.18674042534277024,
+      "learning_rate": 1.1573108828766255e-06,
+      "loss": 0.0442,
+      "step": 499
+    },
+    {
+      "epoch": 0.8507018290089323,
+      "grad_norm": 0.21647795156393285,
+      "learning_rate": 1.1316627581419137e-06,
+      "loss": 0.0535,
+      "step": 500
+    },
+    {
+      "epoch": 0.8524032326669503,
+      "grad_norm": 0.3119714991626263,
+      "learning_rate": 1.1062849771420025e-06,
+      "loss": 0.1191,
+      "step": 501
+    },
+    {
+      "epoch": 0.8541046363249681,
+      "grad_norm": 0.21164862051641084,
+      "learning_rate": 1.0811783134970132e-06,
+      "loss": 0.0658,
+      "step": 502
+    },
+    {
+      "epoch": 0.855806039982986,
+      "grad_norm": 0.2724719475504451,
+      "learning_rate": 1.0563435325622762e-06,
+      "loss": 0.0736,
+      "step": 503
+    },
+    {
+      "epoch": 0.8575074436410038,
+      "grad_norm": 0.22358105859093347,
+      "learning_rate": 1.0317813914050157e-06,
+      "loss": 0.0711,
+      "step": 504
+    },
+    {
+      "epoch": 0.8592088472990217,
+      "grad_norm": 0.2978912375008609,
+      "learning_rate": 1.007492638781259e-06,
+      "loss": 0.0895,
+      "step": 505
+    },
+    {
+      "epoch": 0.8609102509570395,
+      "grad_norm": 0.209790599151719,
+      "learning_rate": 9.834780151130196e-07,
+      "loss": 0.0718,
+      "step": 506
+    },
+    {
+      "epoch": 0.8626116546150574,
+      "grad_norm": 0.1681086042265431,
+      "learning_rate": 9.597382524657173e-07,
+      "loss": 0.0592,
+      "step": 507
+    },
+    {
+      "epoch": 0.8643130582730753,
+      "grad_norm": 0.20035287400774016,
+      "learning_rate": 9.362740745258736e-07,
+      "loss": 0.074,
+      "step": 508
+    },
+    {
+      "epoch": 0.8660144619310931,
+      "grad_norm": 0.17882197027245497,
+      "learning_rate": 9.13086196579035e-07,
+      "loss": 0.0481,
+      "step": 509
+    },
+    {
+      "epoch": 0.8677158655891111,
+      "grad_norm": 0.2010107765861354,
+      "learning_rate": 8.901753254879885e-07,
+      "loss": 0.0599,
+      "step": 510
+    },
+    {
+      "epoch": 0.8694172692471289,
+      "grad_norm": 0.1784357377756698,
+      "learning_rate": 8.67542159671192e-07,
+      "loss": 0.0422,
+      "step": 511
+    },
+    {
+      "epoch": 0.8711186729051468,
+      "grad_norm": 0.2595041553389473,
+      "learning_rate": 8.451873890814988e-07,
+      "loss": 0.0834,
+      "step": 512
+    },
+    {
+      "epoch": 0.8728200765631646,
+      "grad_norm": 0.18720029506475785,
+      "learning_rate": 8.231116951851204e-07,
+      "loss": 0.0441,
+      "step": 513
+    },
+    {
+      "epoch": 0.8745214802211825,
+      "grad_norm": 0.13605762437865937,
+      "learning_rate": 8.013157509408509e-07,
+      "loss": 0.0499,
+      "step": 514
+    },
+    {
+      "epoch": 0.8762228838792003,
+      "grad_norm": 0.3392074002609155,
+      "learning_rate": 7.79800220779554e-07,
+      "loss": 0.0935,
+      "step": 515
+    },
+    {
+      "epoch": 0.8779242875372182,
+      "grad_norm": 0.27563422248531844,
+      "learning_rate": 7.585657605839059e-07,
+      "loss": 0.0749,
+      "step": 516
+    },
+    {
+      "epoch": 0.879625691195236,
+      "grad_norm": 0.2928759924757975,
+      "learning_rate": 7.376130176684082e-07,
+      "loss": 0.107,
+      "step": 517
+    },
+    {
+      "epoch": 0.8813270948532539,
+      "grad_norm": 0.20820494565138964,
+      "learning_rate": 7.169426307596428e-07,
+      "loss": 0.0711,
+      "step": 518
+    },
+    {
+      "epoch": 0.8830284985112719,
+      "grad_norm": 0.1778541297140114,
+      "learning_rate": 6.965552299768186e-07,
+      "loss": 0.0548,
+      "step": 519
+    },
+    {
+      "epoch": 0.8847299021692897,
+      "grad_norm": 0.18347737015999377,
+      "learning_rate": 6.764514368125419e-07,
+      "loss": 0.0468,
+      "step": 520
+    },
+    {
+      "epoch": 0.8864313058273076,
+      "grad_norm": 0.22402928085104057,
+      "learning_rate": 6.566318641138902e-07,
+      "loss": 0.0819,
+      "step": 521
+    },
+    {
+      "epoch": 0.8881327094853254,
+      "grad_norm": 0.10087048048840686,
+      "learning_rate": 6.370971160637129e-07,
+      "loss": 0.0257,
+      "step": 522
+    },
+    {
+      "epoch": 0.8898341131433433,
+      "grad_norm": 0.24244337938824437,
+      "learning_rate": 6.178477881622325e-07,
+      "loss": 0.0929,
+      "step": 523
+    },
+    {
+      "epoch": 0.8915355168013611,
+      "grad_norm": 0.2215797215803862,
+      "learning_rate": 5.98884467208869e-07,
+      "loss": 0.0707,
+      "step": 524
+    },
+    {
+      "epoch": 0.893236920459379,
+      "grad_norm": 0.21837004790407136,
+      "learning_rate": 5.802077312843723e-07,
+      "loss": 0.0601,
+      "step": 525
+    },
+    {
+      "epoch": 0.8949383241173968,
+      "grad_norm": 0.13798341621078078,
+      "learning_rate": 5.618181497331865e-07,
+      "loss": 0.0387,
+      "step": 526
+    },
+    {
+      "epoch": 0.8966397277754147,
+      "grad_norm": 0.18913792414386757,
+      "learning_rate": 5.437162831460962e-07,
+      "loss": 0.0498,
+      "step": 527
+    },
+    {
+      "epoch": 0.8983411314334325,
+      "grad_norm": 0.22877307977873534,
+      "learning_rate": 5.259026833431468e-07,
+      "loss": 0.0704,
+      "step": 528
+    },
+    {
+      "epoch": 0.9000425350914505,
+      "grad_norm": 0.20234348876984656,
+      "learning_rate": 5.083778933568073e-07,
+      "loss": 0.0649,
+      "step": 529
+    },
+    {
+      "epoch": 0.9017439387494683,
+      "grad_norm": 0.2531050005740147,
+      "learning_rate": 4.911424474154314e-07,
+      "loss": 0.0878,
+      "step": 530
+    },
+    {
+      "epoch": 0.9034453424074862,
+      "grad_norm": 0.24193903233807051,
+      "learning_rate": 4.741968709269573e-07,
+      "loss": 0.073,
+      "step": 531
+    },
+    {
+      "epoch": 0.905146746065504,
+      "grad_norm": 0.2541146802125869,
+      "learning_rate": 4.575416804629085e-07,
+      "loss": 0.0563,
+      "step": 532
+    },
+    {
+      "epoch": 0.9068481497235219,
+      "grad_norm": 0.16762038698102058,
+      "learning_rate": 4.411773837426303e-07,
+      "loss": 0.053,
+      "step": 533
+    },
+    {
+      "epoch": 0.9085495533815398,
+      "grad_norm": 0.23076714189266884,
+      "learning_rate": 4.2510447961782055e-07,
+      "loss": 0.0687,
+      "step": 534
+    },
+    {
+      "epoch": 0.9102509570395576,
+      "grad_norm": 0.26029701749366,
+      "learning_rate": 4.093234580573202e-07,
+      "loss": 0.0765,
+      "step": 535
+    },
+    {
+      "epoch": 0.9119523606975755,
+      "grad_norm": 0.2130287074707999,
+      "learning_rate": 3.938348001321812e-07,
+      "loss": 0.062,
+      "step": 536
+    },
+    {
+      "epoch": 0.9136537643555933,
+      "grad_norm": 0.18154744728141692,
+      "learning_rate": 3.786389780009958e-07,
+      "loss": 0.0521,
+      "step": 537
+    },
+    {
+      "epoch": 0.9153551680136113,
+      "grad_norm": 0.15469238363767565,
+      "learning_rate": 3.637364548955047e-07,
+      "loss": 0.0358,
+      "step": 538
+    },
+    {
+      "epoch": 0.9170565716716291,
+      "grad_norm": 0.3678961865523596,
+      "learning_rate": 3.491276851064784e-07,
+      "loss": 0.0881,
+      "step": 539
+    },
+    {
+      "epoch": 0.918757975329647,
+      "grad_norm": 0.23428870195639928,
+      "learning_rate": 3.3481311396986626e-07,
+      "loss": 0.0844,
+      "step": 540
+    },
+    {
+      "epoch": 0.9204593789876648,
+      "grad_norm": 0.2491236742624812,
+      "learning_rate": 3.2079317785322363e-07,
+      "loss": 0.0767,
+      "step": 541
+    },
+    {
+      "epoch": 0.9221607826456827,
+      "grad_norm": 0.27866990645467393,
+      "learning_rate": 3.0706830414240164e-07,
+      "loss": 0.0862,
+      "step": 542
+    },
+    {
+      "epoch": 0.9238621863037005,
+      "grad_norm": 0.13956980085260687,
+      "learning_rate": 2.9363891122853097e-07,
+      "loss": 0.0437,
+      "step": 543
+    },
+    {
+      "epoch": 0.9255635899617184,
+      "grad_norm": 0.1975336620060032,
+      "learning_rate": 2.805054084952552e-07,
+      "loss": 0.076,
+      "step": 544
+    },
+    {
+      "epoch": 0.9272649936197362,
+      "grad_norm": 0.22649643287994412,
+      "learning_rate": 2.6766819630626216e-07,
+      "loss": 0.0647,
+      "step": 545
+    },
+    {
+      "epoch": 0.9289663972777541,
+      "grad_norm": 0.1932416005125246,
+      "learning_rate": 2.5512766599306903e-07,
+      "loss": 0.0642,
+      "step": 546
+    },
+    {
+      "epoch": 0.9306678009357721,
+      "grad_norm": 0.1779695034006232,
+      "learning_rate": 2.4288419984310086e-07,
+      "loss": 0.0439,
+      "step": 547
+    },
+    {
+      "epoch": 0.9323692045937899,
+      "grad_norm": 0.18689194535029088,
+      "learning_rate": 2.3093817108803318e-07,
+      "loss": 0.0761,
+      "step": 548
+    },
+    {
+      "epoch": 0.9340706082518078,
+      "grad_norm": 0.13597627329781425,
+      "learning_rate": 2.1928994389241454e-07,
+      "loss": 0.0369,
+      "step": 549
+    },
+    {
+      "epoch": 0.9357720119098256,
+      "grad_norm": 0.21989658377102142,
+      "learning_rate": 2.0793987334256637e-07,
+      "loss": 0.0625,
+      "step": 550
+    },
+    {
+      "epoch": 0.9374734155678435,
+      "grad_norm": 0.21266747025038635,
+      "learning_rate": 1.968883054357562e-07,
+      "loss": 0.0689,
+      "step": 551
+    },
+    {
+      "epoch": 0.9391748192258613,
+      "grad_norm": 0.25928931998255494,
+      "learning_rate": 1.861355770696549e-07,
+      "loss": 0.1025,
+      "step": 552
+    },
+    {
+      "epoch": 0.9408762228838792,
+      "grad_norm": 0.2801452061064216,
+      "learning_rate": 1.7568201603205827e-07,
+      "loss": 0.0869,
+      "step": 553
+    },
+    {
+      "epoch": 0.942577626541897,
+      "grad_norm": 0.3090393640358726,
+      "learning_rate": 1.6552794099090718e-07,
+      "loss": 0.1212,
+      "step": 554
+    },
+    {
+      "epoch": 0.9442790301999149,
+      "grad_norm": 0.17720681659522422,
+      "learning_rate": 1.5567366148455887e-07,
+      "loss": 0.0355,
+      "step": 555
+    },
+    {
+      "epoch": 0.9459804338579328,
+      "grad_norm": 0.15472295591692306,
+      "learning_rate": 1.4611947791236314e-07,
+      "loss": 0.0395,
+      "step": 556
+    },
+    {
+      "epoch": 0.9476818375159507,
+      "grad_norm": 0.26730416240977917,
+      "learning_rate": 1.3686568152549539e-07,
+      "loss": 0.0595,
+      "step": 557
+    },
+    {
+      "epoch": 0.9493832411739686,
+      "grad_norm": 0.21978673843573918,
+      "learning_rate": 1.2791255441809037e-07,
+      "loss": 0.064,
+      "step": 558
+    },
+    {
+      "epoch": 0.9510846448319864,
+      "grad_norm": 0.19683818868694625,
+      "learning_rate": 1.1926036951862563e-07,
+      "loss": 0.0672,
+      "step": 559
+    },
+    {
+      "epoch": 0.9527860484900043,
+      "grad_norm": 0.19189970865124908,
+      "learning_rate": 1.109093905816172e-07,
+      "loss": 0.0569,
+      "step": 560
+    },
+    {
+      "epoch": 0.9544874521480221,
+      "grad_norm": 0.22200476537406752,
+      "learning_rate": 1.0285987217957038e-07,
+      "loss": 0.0761,
+      "step": 561
+    },
+    {
+      "epoch": 0.95618885580604,
+      "grad_norm": 0.2253757142644252,
+      "learning_rate": 9.511205969522263e-08,
+      "loss": 0.0645,
+      "step": 562
+    },
+    {
+      "epoch": 0.9578902594640578,
+      "grad_norm": 0.218059392950448,
+      "learning_rate": 8.76661893140629e-08,
+      "loss": 0.0682,
+      "step": 563
+    },
+    {
+      "epoch": 0.9595916631220757,
+      "grad_norm": 0.2649165636045151,
+      "learning_rate": 8.052248801712958e-08,
+      "loss": 0.0912,
+      "step": 564
+    },
+    {
+      "epoch": 0.9612930667800936,
+      "grad_norm": 0.20253828553816175,
+      "learning_rate": 7.36811735740961e-08,
+      "loss": 0.0571,
+      "step": 565
+    },
+    {
+      "epoch": 0.9629944704381115,
+      "grad_norm": 0.19699087263694953,
+      "learning_rate": 6.714245453662504e-08,
+      "loss": 0.0457,
+      "step": 566
+    },
+    {
+      "epoch": 0.9646958740961293,
+      "grad_norm": 0.27159809390838907,
+      "learning_rate": 6.090653023201997e-08,
+      "loss": 0.1057,
+      "step": 567
+    },
+    {
+      "epoch": 0.9663972777541472,
+      "grad_norm": 0.2589995420361442,
+      "learning_rate": 5.497359075714026e-08,
+      "loss": 0.102,
+      "step": 568
+    },
+    {
+      "epoch": 0.968098681412165,
+      "grad_norm": 0.19790945958126713,
+      "learning_rate": 4.934381697261015e-08,
+      "loss": 0.0608,
+      "step": 569
+    },
+    {
+      "epoch": 0.9698000850701829,
+      "grad_norm": 0.24043673016274164,
+      "learning_rate": 4.401738049730653e-08,
+      "loss": 0.0609,
+      "step": 570
+    },
+    {
+      "epoch": 0.9715014887282007,
+      "grad_norm": 0.2690784354034665,
+      "learning_rate": 3.899444370312533e-08,
+      "loss": 0.0861,
+      "step": 571
+    },
+    {
+      "epoch": 0.9732028923862186,
+      "grad_norm": 0.22841785491856118,
+      "learning_rate": 3.4275159710032146e-08,
+      "loss": 0.0712,
+      "step": 572
+    },
+    {
+      "epoch": 0.9749042960442365,
+      "grad_norm": 0.1950332471845292,
+      "learning_rate": 2.9859672381392644e-08,
+      "loss": 0.0532,
+      "step": 573
+    },
+    {
+      "epoch": 0.9766056997022544,
+      "grad_norm": 0.22678060127201066,
+      "learning_rate": 2.574811631959273e-08,
+      "loss": 0.0771,
+      "step": 574
+    },
+    {
+      "epoch": 0.9783071033602723,
+      "grad_norm": 0.3698823942978684,
+      "learning_rate": 2.1940616861929608e-08,
+      "loss": 0.0693,
+      "step": 575
+    },
+    {
+      "epoch": 0.9800085070182901,
+      "grad_norm": 0.23298311393693824,
+      "learning_rate": 1.8437290076792624e-08,
+      "loss": 0.0725,
+      "step": 576
+    },
+    {
+      "epoch": 0.981709910676308,
+      "grad_norm": 0.2028362088180894,
+      "learning_rate": 1.5238242760126088e-08,
+      "loss": 0.0756,
+      "step": 577
+    },
+    {
+      "epoch": 0.9834113143343258,
+      "grad_norm": 0.29483561036206646,
+      "learning_rate": 1.234357243217188e-08,
+      "loss": 0.0988,
+      "step": 578
+    },
+    {
+      "epoch": 0.9851127179923437,
+      "grad_norm": 0.2525424387277673,
+      "learning_rate": 9.753367334499608e-09,
+      "loss": 0.0771,
+      "step": 579
+    },
+    {
+      "epoch": 0.9868141216503615,
+      "grad_norm": 0.26576715050467775,
+      "learning_rate": 7.467706427312093e-09,
+      "loss": 0.0612,
+      "step": 580
+    },
+    {
+      "epoch": 0.9885155253083794,
+      "grad_norm": 0.1729974193289849,
+      "learning_rate": 5.486659387043958e-09,
+      "loss": 0.0371,
+      "step": 581
+    },
+    {
+      "epoch": 0.9902169289663972,
+      "grad_norm": 0.19337221194799736,
+      "learning_rate": 3.810286604232216e-09,
+      "loss": 0.0548,
+      "step": 582
+    },
+    {
+      "epoch": 0.9919183326244151,
+      "grad_norm": 0.2881735818611327,
+      "learning_rate": 2.4386391816777488e-09,
+      "loss": 0.0866,
+      "step": 583
+    },
+    {
+      "epoch": 0.993619736282433,
+      "grad_norm": 0.25389258821605704,
+      "learning_rate": 1.3717589328898773e-09,
+      "loss": 0.0442,
+      "step": 584
+    },
+    {
+      "epoch": 0.9953211399404509,
+      "grad_norm": 0.2721835551378759,
+      "learning_rate": 6.096783808062778e-10,
+      "loss": 0.0977,
+      "step": 585
+    },
+    {
+      "epoch": 0.9970225435984688,
+      "grad_norm": 0.2313901379023938,
+      "learning_rate": 1.524207568059932e-10,
+      "loss": 0.0628,
+      "step": 586
+    },
+    {
+      "epoch": 0.9987239472564866,
+      "grad_norm": 0.24018936080778358,
+      "learning_rate": 0.0,
+      "loss": 0.0614,
+      "step": 587
+    },
+    {
+      "epoch": 0.9987239472564866,
+      "step": 587,
+      "total_flos": 1551551083839488.0,
+      "train_loss": 0.1089990799881683,
+      "train_runtime": 5367.2694,
+      "train_samples_per_second": 14.016,
+      "train_steps_per_second": 0.109
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 587,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1551551083839488.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}